Jarvis Hayman, Marc Oxenham - Human Body Decomposition (2016, Academic Press)
August 13, 2022 | Author: Anonymous | Category: N/A
Short Description
Download Jarvis Hayman, Marc Oxenham - Human Body Decomposition (2016, Academic Press)...
Description
Human Body Decomposition
Human Body Decomposition
Jarvis Hayman and Marc Oxenham The School of Archaeology & Anthropology, Australian National University, Canberra, Australia
AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS SAN DIEGO SAN FRANCISCO FRANCISCO SINGAPORE SYDNEY TOKYO •
•
•
•
•
•
•
•
•
Academic Press is an imprint of Elsevier
Academic Press is an imprint of Elsevier Academic Elsevier 125 London Wall, London EC2Y 5AS, UK 525 B Street, Street, Suite 1800, San Diego, CA 92101-4495 92101-4495,, USA 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, USA The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK Copyright r 2016 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval retri eval system, without without permission permission in writing writing from the publisher. publisher. Details on how to seek permission, further information about the Publisher s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: website: www.elsevier. www.elsevier.com/permissions com/permissions.. ’
This book and the individual individual contributio contributions ns contained contained in it are protected protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods or professional practices, may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating evaluatin g and using any information information or methods methods described described herein. In using such information information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library
ISBN: 978-0-12-803691-4 For Information on all Academic Press publications visit our website at at https://www.elsev https://www.elsevier.com/ ier.com/
BIOGRAPHY
Jarvis Hayman, FRCS (Ed), FRACS, MA (Hons), PhD
Jarvis Hayman graduated in Medicine at Aberdeen University. After moving to Australia in 1974, he worked for many years as a surgeon in Sydney and in a regional city in New South Wales. Upon retiring, he studied archeology at the Australian National University, obtaining a Ma Mast ster er s degr degree ee in 20 2006 06 wi with th a thes thesis is on th thee arch archeo eolo logy gy of the Scottish Highland Clearances. He then combined his archeological and ’
medica medi call kn know owle ledg dgee to unde underrta tak ke a PhD PhD in fo fore rens nsic ic arch archeo eolo logy gy.. Rese Re sear arch ch for for th thee thes thesis is in incl clud uded ed a deta detail iled ed stud study y of th thee Nati Nation onal al Coro Co roni nial al Info Inform rmat atio ion n Syst System em,, base based d at the the Vict Victor oria ian n Inst Instit itut utee of Fore Fo rens nsic ic Medi Medici cine ne (VIF (VIFM) M) in Melb Melbou ourn rne, e, stud studie iess of th thee auto autops psie iess of decomposed bodies at the VIFM and researching the decomposition of human bodies at the Grady Early Anthropology Research Facility (GEFARL) of the Texas State University in San Marcos. The thesis developed mathematical models using numerical total body scores of the decomposition process, to estimate the time of death up to 14 days, of human bodies found in many of the states of Australia. Marc Oxenham, PhD, FSA
Marc Oxenham is a Reader in Archaeology and Biological Anth An thro ropo polo logy gy an and d an Aust Austra rali lian an Futu Future re Fell Fellow ow at The The Aust Austra rali lian an Nati Na tion onal al Un Univ iver ersi sity ty.. He ha hass pl playe ayed d a lead leadin ing g role role in unde unders rsta tand ndin ing g human biological (health and disease related) and cultural responses to significant human- and climate-induced events in Southeast Asia during ing the the last last 10,0 10,000 00 year years. s. In addi additi tion on to hi hiss cl clos osee in invo volv lvem emen entt over over many ma ny year yearss as an oste osteol olog ogis istt and and arch archeo eolo logi gist st wi with th the the Aust Austra rali lian an Defence Force s Unrecovered War Casualty-Army unit, he edited the ’
first Australian book on forensic andand allied disciplines in 2008 focused Forensic Approaches to anthropology Death, Disaster Abuse. “
”
viii
Biography
Over the last decade or so he has authored and edited 7 books and some so me 75 re rese sear arch ch pa pape pers rs and and book book chap chapte ters rs on to topi pics cs rang rangin ing g from from forensic forens ic anthro anthropology pology,, paleopatholo paleopathology, gy, bioarc bioarchaeolo haeology, gy, paleoparasi paleoparasi-tology, to the archeology of children and the elderly.
PREFACE
From time immemorial when a decomposed or decomposing human body is discovered, those people with a questioning or scientific nature have ha ve aske asked d thre threee ques questi tion ons; s; how how di did d this this de deat ath h ha happ ppen en,, who who was was involved in the death, and what was the time of death. The answer to the third question will very often give clues which lead to answers to the first two questions. The firs The firstt know known n subs substa tant ntiv ivee wr writ itte ten n docu docume ment ntat atio ion n of fore forens nsic ic methods and estimation of the time of death can be traced back to a 13th-century Sung dynasty medicolegal textbook entitled Hsi Yuan Lu (The (T he Wash Washin ing g Aw Away ay of Wron Wrongs gs)) by Su Sung ng Tz u, a very very astu astute te le lega gall ’
official and death investigator in Fujian province, southern China. The mode mo dern rn era era of sc scie ient ntif ific ic fore forens nsic ic in inve vest stiga igati tion on and and the the in inve vest stig igat atio ion n into the time of death began at the beginning of the 19th century with Dr John Davy in Ceylon (Sri Lanka). Sporadic studies were published duri du ring ng th thee 19th 19th ce cent ntur ury y afte afterr whic which h ther theree was a nota notabl blee pauc paucit ity y of studies, probably caused by two World Wars, until increasing numbers of studies began to be published from the 1950s until the present time. Durin Duri ng the time time one of us (JH) (JH) wa wass carr carryi ying ng ou outt back backgr grou ound nd research for a PhD thesis on the time since death in human bodies found decomposed in Australian conditions, it became apparent that a chronology of such research was lacking. There are numerous studies which begin with a short overview of the research but none which give a detailed account of it. On the basis that researchers need to know where they have come from to know where they are going to and that ever ev ery y re rese sear arch cher er,, figu figura rati tivel vely y spea speaki king ng,, stan stands ds on the the shou should lder erss of others, this work is presented. It has five chapters. 1.
Supravital Reactions in the Estimation of the Time Since Death (TSD) deal dealss with with re rese sear arch ch in th thee stage stage imme immedi diat atel ely y af afte terr deat death h when the body undergoes changes but remains responsive to various stimuli.
“
”
x
2.
Preface
Algor Mortis and Temperature-Based Methods of Estimating the Time Since Death deals with research during the phase when the body is cooling from normal body to ambient temperature. 3. Biochemical Methods of Estimating the Time Since Death covers re rese sear arch ch into into me meth thod odss usin using g the the chan changi ging ng bi bioc oche hemi mica call re react actio ions ns whic wh ich h occu occurr and and the the ch chem emic ical al subs substa tanc nces es pr prod oduc uced ed as the the body body decomposes. 4. Re Rese sear arch ch in the the Later Later Stag Stages es of De Deco comp mpos osit itio ion n deal dealss with with th thee st stag agee of deco decomp mpos osit itio ion n afte afterr th thee on onse sett of putr putref efac acti tion on and and unti untill skeletonisation occurs. 5. Recent Research and Current Trends in research into the TSD are outlined in the final chapter. “
”
“
”
“
“
”
”
Throughout the book the terms TSD (time since death) and PMI (postmortem interval) are interchangeable.
CHAPTER
1
Supravital Reactions in the Estimation of the Time Since Death (TSD)
RIGOR MORTIS LIVOR MORTIS MEASUREMENT OF MECHANICAL AND ELECTRICAL EXCITABILITY OF MUSCLE REFERENCES
Follow Foll owin ing g th thee ce cess ssat atio ion n of the the he hear artt an and d ci circ rcul ulat ator ory y syst system em at the the moment of death, there is a period of time, generally accepted to be 34 minutes after the heart stops beating, during which some tissues remain responsive to various stimuli, and when resuscitation may be possible despite increasing anoxia of the tissues and resultant irreversible ischemia. However, it is not equivalent to the supravital period which is believed to extend from 100 to 120 minutes after cessation of thee circ th circul ulat atio ion n (Madea, Madea, 20 2002a 02a). ). Duri During ng the the supr suprav avit ital al peri period od,, the the tissues remain responsive to various electrical and mechanical stimuli and exhibit certain phenomena such as rigor mortis and livor mortis. Attemp Atte mpts ts ha have ve been been mad madee to use use thes thesee re reac acti tion onss an and d ph phen enom omen ena a to measure the time since death (TSD).
RIGOR MORTIS
The phenomenon of rigor mortis was first described in 1811 by the French physician, P.H. Nysten, but its physiological basis was not discovered until 1945 by Szent-Györgyi (2004). (2004). It consists of a sustained cont co ntra ract ctio ion n of the the musc muscle less of the the body body,, whic which h begi begins ns at 26 hours after death, persists for 2484 hours, and is then followed by gradual
Human Body Decomposition. DOI: http://dx.doi.org/10.1016/B978-0-12-803691-4.00001-7 © 2016 Elsevier Inc. All rights reserved.
2
Human Body Decomposition
re rela laxat xatio ion n unti untill th thee musc muscle less agai again n beco become me fl flac acci cid d (Gill-K Gill-King, ing, 1997 1997). ). Thee cont Th contrac racti tile le unit unitss of musc muscle le cell cells, s, sa sarc rcom omer eres es,, cons consis istt of pa para rall llel el units of two types of protein, actin and myosin. Crosslinkages on the myos my osin in unit unitss pull pull the the actin actin unit unitss towa toward rd ea each ch ot othe her, r, caus causin ing g musc muscle le contraction. The process requires calcium and energy, the latter provide vided d by aden adenos osin inee tr trip ipho hosp spha hate te (ATP (ATP)) (Bat Batee-Sm Smit ith h an and d Be Bend ndal all, l, 1947). 1947 ). The initial flaccidity of muscles after death is due to continued form fo rmat atio ion n of AT ATP P by an anae aero robi bicc glyco glycoly lysi sis, s, bu butt wi with th th thee pass passag agee of time, ATP is no longer resynthesized, energy is no longer available for the actin and myosin fibrils to remain relaxed and the fibrils contract, re resu sult ltin ing g in the the musc muscle le body body as a whol wholee cont contra ract ctin ing. g. Reso Resolu luti tion on of rigor mortis after 2484 hours is caused by proteolytic enzymes within the muscle cells disrupting the myosin/actin units, causing the crosslinkages to break down and the muscles to relax (Gill-K (Gill-King, ing, 1997 1997). ). At the beginning of the 19th century Nysten (1811) (1811),, in France, carried out experiments on criminals immediately after their decapitation on the guillotine and he observed that rigor mortis began in the muscles of the jaw jaw and and the then pro progres gresse sed d dista istall lly y to the fe feet et an and d toe oes. s. This This sequence was disputed by Shapiro (1950, 1954), 1954), who suggested that it began at the same time in all muscles but the variation in the sizes of the different joints and muscles meant that the larger muscles took longerr to de ge deve velo lop p rigo rigorr mort mortis, is, givi giving ng th thee impr impres essi sion on th that at it prog progre ress ssed ed from proximal to distal in the body. Krompecher designed an experiment to measure the intensity of rigor mortis in rat front limbs compared with rat hind limbs using different forces at different times during thee cour th course se of ri rigo gorr mor mortis tis (Kro Kromp mpeche echerr and Fryc Fryc,, 1978 1978a a). The The hin ind d limbs had a muscle mass 2.89 times the muscle mass of the front limbs. The results showed that although there was no difference between front and hind limbs with respect to the time taken to reach complete evolution tion of the the ri rigo gorr mort mortis is,, the the on onset set and and the the re rela laxa xati tion on of ri rigo gorr mo mort rtis is were more rapid in the front limbs which had the smaller muscle mass. In contr contras ast, t, Koba Kobaya yash shii an and d col collea league guess (20 (2001) 01),, exper experim imen enti ting ng with with in vitro rat erector spinae muscles, found that although the volume of muscle samples varied there was no difference in the development and resolution of rigor mortis. They concluded that it was the proportion of muscl mus clee fibe fiberr type typess in each each mu musc scle le,, di diff ffer eren ence ce in te temp mper erat atur ure, e, and and th thee dynamic characteristics of each joint that determined the speed of onset and resolution of rigor mortis.
Supravital Reactions in the Estimation of the Time Since Death (TSD)
3
Several intrinsic and extrinsic factors affect the speed of onset and duration of rigor mortis. Intrinsic factors such as violent exercise and high fever during the agonal stage will cause a rapid onset and shorter duration. The amount of skeletal muscle dictates the duration of rigor, rig or, for exampl example, e, it appear appearss and re resol solve vess early early in infan infants ts but but,, in contrast, a robust physical person will have slower onset and a prolonged duration (Gill-King, (Gill-King, 1997). 1997). This finding, however, was contradicted by Kobayashi by Kobayashi and colleagues (2001). (2001). Krompecher and Fryc (1978b),, in a stud (1978b) study y usin using g ra rats ts,, foun found d that that ph phys ysic ical al exer exerci cise se be befo fore re death dea th cause caused d an increa increased sed in inten tensit sity y of the rig rigor or wh which ich reache reached d it itss maximum intensity at the same time as normal controls but the maximum inten intensit sity y was susta sustaine ined d longe longer. r. The The rigor rigor,, howeve however, r, reach reached ed resolution at the same time as the controls. In a controlled experiment using rats, Krompecher rats, Krompecher (1981) found (1981) found that the higher the temperature, the shorter was the onset of rigor and the faster the resolution, a finding later confirmed by by Kobayashi and colleagues (2001) (2001).. At very
low temperature (6 C), development very slow at 48 60with hours and resolution very prolonged to 168 was hours. This contrasted a temp temper erat atur uree of 37 C when when deve develo lopm pmen entt oc occu curr rred ed at 3 hour hourss an and d resolved at 6 hours. In a mortuary where corpses were kept refrigerated at 4 C, rigor was found to completely persist for 10 days in all corp co rpse ses, s, beca became me part partia iall by 17 da days ys,, and and re reso solv lved ed afte afterr 28 days days (Varetto and Curto, 2005 2005). ).
Other extrinsic factors which affect the course of rigor mortis are electrocution causing death, which accelerates the onset of rigor and shortens the duration, possibly because the violent cramps experienced Krom ompe pech cher er and Be Berg rger erio ioux ux,, 198 1988 8). cause a rapid fall in ATP (Kr Strych Str ychnin ninee poison poisoning ing hasten hastenss the onset onset and dur durati ation on of rigor rigor mor mortis tis while carbon monoxide poisoning delays the resolution (Krompecher (Krompecher et al., 1983). 1983). If the rigidity of rigor mortis is broken by force it can re re-e -est stab abli lish sh itse itself lf if th thee proc proces esss is stil stilll ongo ongoin ing; g; the the re re-e -est stab abli lish shme ment nt begins immediately after being broken, the rigidity is weaker but the maximum extent of it is the same as in controls, as is the course of resolution (Krompecher (Krompecher et al., 2008 2008). ). Objective measurement of the force required to break the rigidity of ri rigo gorr mo mort rtis is wa wass atte attemp mpte ted d for for ma many ny year years, s, th thee fi firs rstt at atte temp mptt bein being g made in 1919 by Oppenheim and Wacker, but the difficulty in
4
Human Body Decomposition
measuring this force is that the strength of the force varies with the stage of development and resolution of the rigor mortis (Krompecher, ( Krompecher, 2002). 2002 ). The forces involved are initially small, rising rapidly to a maximum, and then reducing gradually over time until resolution occurs. One measurement at one period of time in the duration of the rigor will not reveal any useful information concerning the estimation of the Krompecher cher (1994) carried out experiments on groups of rats TSD. Krompe killed by a standard method and kept at the same temperature of 24 C post po st mo mort rtem em.. The The same same fo forc rce, e, in insu suff ffic icie ient nt to brea break k the the ri rigo gor, r, was was applied to a limb at varying intervals after death up to 48 hours. It was found that repeated measurements of the intensity of rigor mortis allowed a more accurate estimation of the TSD than a single measurement and Krompecher suggested certain guidelines: (1) If there was an incr increa ease se in inte intens nsit ity, y, th thee in init itia iall measu measure reme ment ntss wer weree ta take ken n no ea earl rlie ierr than 5 hours post mortem. (2) If there was a decrease in intensity the initial measurements were taken no earlier than 7 hours post mortem.
(3) (3 )ange At hour urs s sity po post mort rtem em re reso solu luti on was comp mple lete and and79nodece furt rthe her chan ch ge 24 inho in int tensi en tystmo sh shou ould ld oc occu cur. r.tion A rece re cent ntco stud study yte of defu ceas ased edr pati pa tien ents ts wa wass unde undert rtak aken en in a hosp hospit ital al mort mortua uary ry wher wheree the the ti time me of deat de ath h was was know known, n, wher wheree they they were were all all tran transp spor orte ted d to th thee mort mortua uary ry within 5 hours and kept at a temperature of 2021 C (Anders et al., 2013). 2013 ). The aims of the study were to determine if re-establishment of rigor mortis took place in loosened joints after more than 8 hours and, if so, could it be determined how many hours postmortem reestablishment of rigor mortis did occur? Deaths occurred from a variety of disease conditions but because of the small numbers, no correction was possible for disease state. Rigor mortis was loosened in 174
joints of 44 deceased persons between 7.5 and 10.5 hours post mortem to determine whether re-establishment occurred after 8 hours and 140 joints were examined after loosening at 1521 hours post mortem to determine how many hours postmortem re-establishment could occur. The study found that 121 of 314 joints (38.5%) showed reestablishment of rigor mortis between 7.5 and 19 hours and the authors conc co nclu lude ded d th that at th thee cu curr rren entl tly y acce accept pted ed view view th that at ri rigo gorr mo mort rtis is coul could d only on ly be st stud udie ied d to de dettermi ermine ne the time time of deat eath le less ss th than an 8 hour hourss post mortem, required re-evaluation by further studies. Attempts have been made to standardize the measurement of the force of rigidity in rigor mortis but they have not received widespread acceptance (Schuck ( Schuck et al al., ., 19 1979 79;; Va Vain in et al al., ., 19 1992 92). ). Be Beca caus usee of th thee subj subjec ecti tive ve natur naturee of thee as th asse sess ssme ment nt of ri rigo gorr mort mortis is and and the the numb number er of vari variab able le fact factor orss
Supravital Reactions in the Estimation of the Time Since Death (TSD)
5
determining its onset, duration, and resolution, it should only be used in conjunction with other methods when estimating TSD (Henssge ( Henssge and Madea, 2002). 2002). LIVOR MORTIS
Livorr mo Livo morrtis tis or livi livid dity ity is th thee gr grav avit itat atio iona nall pool poolin ing g of bl bloo ood d in th thee dependent parts of the body, both externally in the skin capillaries and venules but also in the internal organs. Its onset is variable but it is usually ally mo mosst evid evideent abou aboutt 2 ho hour urss afte afterr de deat ath, h, al alth thou ough gh it is stat stated ed to occur as soon as 15 minutes after death (Clark ( Clark et al., 1997). 1997). Initially the color is red but it later becomes purple as oxygen dissociates from the hemoglobin, changing it to purple-colored deoxyhemoglobin. This color change can be variable depending on the circumstances of death and the enviro env ironme nment. nt. Cold Cold tempera temperatur tures es will delay delay the dis dissoc sociat iation ion of oxygen oxygen from fro m the the hemog hemoglo lobi bin, n, dela delayi ying ng the color color ch chan ange ge from from re red d to pu purp rple le.. Carbon monoxide poisoning produces a persistent cherry red color and cyanide poisoning will also cause the red color to persist. Lividity may not be seen in bodies that are very anemic at death. Initially it is not fixed, that is, if pressure is applied to a skin area the red color changes to white as the blood is returned to the capillaries due to the pressure. Bodies lying on a hard surface will also show white blanching in the areas making contact with the surface for the same reason. Lividity is said to become fixed in 46 hours, that is, the red color no longer disappears on pressure because with cooling of the body, the fat surrounding the capilla capillaries ries solidif solidifies ies,, constri constricti cting ng the capilla capillarie riess and preven preventin ting g the return of blood into them (Clark (Clark et al., 1997). 1997). A reference to lividity occurs in the earliest extant comprehensive hand ha ndbo book ok for for midmid-13t 13thh-ce cent ntur ury y Chin Chines esee fore forens nsic ic in inves vesti tigat gator orss in into to homicides and other deaths by Sung Tz u. A clear description of livid(118 186 61249) handbook: ity ity oc occu curs rs towa towarrd the the end end of f Sung Sung Tz u s (1 Generally, dead persons have a slight red coloration on the back of the neck, on the top of the back, on the ribs, the back of the waist, the insides of the legs, the knees, the feet and the stomach. Check to determine if after death these corpses were laid out supine overnight. The coll co llap apse se of th thee bl bloo ood d vess vessel elss may may caus causee this this slig slight ht re red d co colo lora rati tion on,, which does not indicate any other cause of death. ’
’
’
“
”
Turnin Turn ing g to more more re rece cent nt time times, s, th thee ti time me se sequ quen ence ce of li livo vorr mort mortis is has been proposed as a method of estimating the TSD. However, the
6
Human Body Decomposition
physiological mechanism of blood pooling, coagulation, and hemoglobi bin n di diss ssoc ocia iati tion on is so vari variab able le as to make make it an unre unreli liab able le meth method od (Knight Knight,, 2002 2002). ). The assessment by observation with the passage of timee is subjec tim subjectiv tivee and there therefor foree prone prone to obser observer ver er error ror.. Rec Recent ently ly attempts have been made to quantify the hypostasis of lividity by colorimetry. A preliminary study carried out by Vanezis by Vanezis (1991) (1991),, using a colorimeter, showed a linear relationship between the passage of time and th thee lig light hteni ening ng color color intens intensity ity wh when en bodie bodiess showin showing g lividi lividity ty in depe de pen nde dent nt area areass were were tu turn rned ed ov over er to enab enable le bloo blood d to retu return rn to capillaries. Hypostasis reduced considerably even after 24 hours and slightly up to 3 days after death. In a follow-up study, Vanezis and Trujillo (1996) attempted (1996) attempted to quantify the rate of change in the intensity of livor mortis with time by the use of a colorimeter. Ninetythre th reee ca cada dave vers rs in whom whom the the time time of de deat ath h wa wass know known n to wi with thin in 3 ho hour urss were were subj subjec ecte ted d to colo colori rime metr tric ic stud study. y. The The bo bodi dies es,, kept kept at 4 C, were placed in the prone position and the degree of luminosity
on thei their r back backss me meas asur ured ed at 4-ho 4hour urly ly the in inte terv rval alss up to 72 hour hours. s.and A strong correlation was found between degree of luminosity thee po th post stmo mort rtem em inte interv rval al (PMI (PMI), ), livi lividi dity ty be beco comi ming ng dark darker er with with incre inc reasi asing ng PMI in an expon exponent ential ial fas fashio hion. n. After After 72 hours hours li livid vidity ity became fixed. There were only a small number of cases in this series and the authors concluded that factors such as body size, cause of death, body position, environmental temperature, and especially skin color could affect the luminosity. A simple colorimeter was developed by Masashi Inoue and colleagues (1994) to measure the change in intensity of postmortem lividity with the passage of time. The colorimeter transmitted and received light in the the 600600-nm nm wa wave vele leng ngth th reg regio ion n whic which h was was po poor orly ly absor absorbe bed d by the the water and melanin of the skin but was strongly absorbed by hemoglobin. The instrument was used to measure the intensity of lividity as pressure su re was was appl applie ied d to area areass of maxi maximu mum m an and d mini minimu mum m li livi vidi ditty in 42 corp co rpse sess whos whosee PM PMII was was know known, n, over over a peri period od of ti time me unt until li livi vidi dity ty became fixed. A strong correlation between the passage of time and the intensity of the livor mortis was found but there were variations caused by the same factors that others had noted, namely, body size, cause of death, posture, environmental temperature, and skin color. Kaatsch and colleagues collea gues (1994 (1994)) measured measured pressure pressure-ind -induced uced blanch blanching ing of livor livor mortis mortis to estimate TSD with a digital system of photometric quantification on 50 cadavers in which the time of death was known. The authors used
Supravital Reactions in the Estimation of the Time Since Death (TSD)
7
defined magnitudes of pressure, in contrast to subjective pressure by finger or forceps. They found distinct differences between the different time categories for pressure-induced color changes in lividity up to 40 hours post mortem, after which pressure no longer produced blanching. They also found that blanching could be produced for a longer period of time in bodies stored at a cooler temperature. There were wide variations in the data which they attributed to skin color, antemort ante mortem em physical physical conditio condition, n, cause cause of deat death, h, environ environment mental al factors to rs such such as ambie ambient nt temper temperatu ature, re, and storag storagee condit conditio ions ns prior prior to measu mea surem rement ent.. The author authorss concl conclude uded d that that pr provi ovided ded these these facto factors rs were taken into account, and considering this was a small study, the findings provided a basis for further research to improve the measurement of lividity. The quantitative measurements of livor mortis are only useful for 3040 hours post mortem and before the changes of putrefaction have begun (Kaatsch (Kaatsch et al., 1994). 1994). The rate of occurrence, intensity of coloration, distribution, and possible redistribution of lividity are so variable and difficult to standardize that they should not be used alone to estimate the postmortem period (Vanezis (Vanezis and Trujillo, 1996 1996). ).
MEASUREMENT OF MECHANICAL AND ELECTRICAL EXCITABILITY OF MUSCLE
The first recorded observation of the effect of mechanical muscle stimulation tio n in an is isol olat ated ed post postmo mort rtem em limb limb was was by Dowler Dowler (1846) (1846),, when he struck the biceps muscle in an extended arm with the edge of his hand, 2 hours after death, causing the arm to contract so that the hand struck thee brea th breast st.. Most Most re ressea earc rch h in into to th thee use use of th thee re resp spon onse se of mu musc scle less to mechanical stimulation in the postmortem period has been carried out by researchers in Germany and published in the German language (Madea, ( Madea, 2002b; Henssge et al., 1988). 1988). Skeletal muscle is responsive to mechanical stimulation in the immediate postmortem period and the muscular contraction can be measured. When the lower third of the thigh above the pate pa tell lla a is st strruck, uck, ther theree is a stro strong ng musc muscle le cont contra raccti tion on re resp spon onsse fo forr 1.52.5 hours, gradually becoming weaker until there is no response after about 12 hours (Madea, (Madea, 2002b 2002b). ). The upper time limit of idiomuscular response was based on a few case studies in the earlier part of the 20th century but an upper time limit of 13 hours when no further response occurs has been recently confirmed by a large study on 270 cases, the
8
Human Body Decomposition
majority of which were of sudden death and in all of which the time of death was known precisely (Warther (Warther et al., 2012 2012). ). The practical use of this method is limited and it should not be used as the sole method of estimation of the TSD (Madea, (Madea, 2002b). 2002b). In 1781, Luigi Galvani carried out a series of experiments on dissected frogs in which the thighs were entirely removed from the body, leaving the legs and feet attached to a stump of the vertebral column by the sciatic nerves alone (Ho Hoff ff,, 19 1936 36). ). In ex expe peri rime ment ntss us usin ing g a machine which produced electricity, as well as using atmospheric electric tricit ity, y, he st stim imul ulat ated ed th thee sc scia iati ticc nerv nerves es an and d th thee vert verteb ebral ral colu column mns, s, which caused the muscles of the frogs legs to contract. He attributed this to animal electricity which was distributed along the nerves to the muscles of the legs. Further studies were carried out in the 19th century and toward the end of that century and in the first half of the 20th century studies were directed toward the use of electrical excitabil ab ilit ity y as a means means of dete determ rmin inin ing g the the TSD, TSD, spec specif ific icall ally y as in indi dica cate ted d by the lack of electrical excitability of muscle (Made Madea, a, 20 2002 02cc). From the 1960s to the 1990s most studies were carried out in Germany 2002c 2c). ). Most of the and an d re repo port rted ed in the the Germ German an lang langua uage ge (Madea, Madea, 200 research in this period was confined to giving a description and subjective grading of the muscular response to electrical stimulation according to the strength of contraction and spread of movement to areas di dist stan antt from from the the stim stimul ulus us.. The The re resu sult ltss of va vari riou ouss stud studie iess were were not not comparable because the position of the electrodes, the parameters of excitation, and the grade of muscular contraction were not standardized (Madea (Madea and Henssge, 1990). Madea and Henssge (1990) (1990) reviewed reviewed 1990). Madea the use of electrical stimulation of muscle tissue and its efficacy in estimating the TSD and, for the first time, measured and quantified the force of contraction in the orbicularis oculi muscle as it declined with the passage of time. They They stat stated ed that that it was was impor mporta tant nt th that at th thee method be standardized and they formulated a protocol for its use at the scene of a death with which the time of death, especially in the period per iod 38 ho hours urs aft after er dea death, th, could could be deter determin mined. ed. By combin combining ing this method of stimulation of the orbicularis oculi muscle with the use of the nomogram method of determining the TSD by temperature measurement (Henssge, (Henssge, 1988 1988)) as well as the assessment of lividity, ity, ri rigo gorr mort mortis is,, and and the the me mech chan anic ical al stim stimul ulat atio ion n of musc muscle le,, the the ’
“
”
1988). ). accuracy of estimating the TSD was increased (Henssge ( Henssge et al., 1988 A study by Madea (1992) determined that the length of time after
Supravital Reactions in the Estimation of the Time Since Death (TSD)
9
death that the muscle remained responsive to electrical stimulation, could be up to 13 hours. There was a period of maximum contraction and a period of relaxation until the muscle response was exhausted. The decrease of the maximum force by which the muscle could be stimulated to contract and the increase in the relaxation time could both be measured. The period of responsiveness showed interindividual variation depending on the amount of glycogen in the muscles at the time of death, which in turn determined the period of elasticity of thee mu th muscl scles es bef before ore the on onset set of ri rigor gor mor mortis tis.. Furthe Furtherr stud studies ies were were required using simultaneous measurement of electrical excitability of the muscle and parameters of anaerobic glycolysis, probably lactate concentration, in order to increase the accuracy of measurement of the TSD. A second factor determining the period of muscle responsiveness was the environmental temperature; the lower the temperature the longer was the response. Other research using electrical stimulation to improve estimation of the TSD in the very early postmortem period has been carried out by McDowall and colleagues (1998) who measured the absolute refrac ref ractor tory y period period (ARP) (ARP) in th thee sci sciati aticc nerves nerves of newly newly killed killed rats. rats. The ARP is the interval immediately following an action potential when wh en a nerv nervee is un unab able le to prop propag agat atee a furt furthe herr acti action on pote potent ntia iall in response to stimulation. McDowall and colleagues found that measuring the ARP did not provide a more accurate estimation of the PMII than PM than the the use use of re rect ctal al temp temper erat atur uree bu butt wh when en both both va vari riab able less were we re us used ed in co comb mbin inat atio ion, n, th thee ov over eral alll accu accurrac acy y was was im impr prov oved ed.. Skeletal muscle only remains indirectly excitable to stimulation for about 1.5 hours after death and therefore this short time period during which the ARP could be measured meant that the practical usefulness of this technique was limited to a small number of cases in whic wh ich h it wa wass nece necess ssar ary y to dete determ rmin inee whet whethe herr or no nott de deat ath h wa wass instantaneous. Similar research was carried out by Elmas and colleagues (200 (2001) 1) and Querido Querido and Phillip Phillipss used extracel extracellula lularr abdomina abdominall impe im peda danc ncee of an elec electr tric ical al cu curr rren entt as a mean meanss of es esti tima mati ting ng the the PMI. None of these recent methods have improved the estimation of thee TSD th TSD in the earl rly y postm tmor orttem per eriiod (Que Querid rido, o, 19 1994, 94, 200 2000; 0; Querido and Phillips, 2001). 2001). In conc conclu lusi sion on,, the the use use of an any y one one supr supravi avita tall re reac acti tion on to es esti tima mate te the PMI in the early period after death, up to about 13 hours, is not to
10
Human Body Decomposition
be recommended. Instead a combination of methods as suggested by He Hens nssg sgee an and d co coll llea eagu gues es (1 (1988 988)) rem remai ains ns th thee best best ap app proac roach h at the the present time. REFERENCES Ande Anders rs,, S. S.,, Ku Kunz nz,, M. M.,, Ge Gehl hl,, A. A.,, Sehn Sehner er,, S. S.,, Ra Raup upac ach, h, T. T.,, Be Beck ck-B -Bor ornh nhol oldt dt,, H. H.-P -P., ., 20 2013 13.. Estima Est imatio tion n of the tim timee sin since ce dea deathth- rec reconsi onsider dering ing the rere-est establ ablish ishmen mentt of rig rigor or mor mortis tis.. Int Int.. J. Legal Med. 127, 127130. Bate-Smith, E.C., Bendall, J.R., 1947. Rigor mortis and adenosine triphosphate. J. Physiol. 106, 177185. Clark, M.A., Worrell, M.B., Pless, J.E., 1997. Post-mortem changes in soft tissues. In: Haglund, W.D., W.D ., Sor Sorg, g, M.H M.H.. (Ed (Eds.) s.),, For Forens ensic ic Tap Taphon honom omy: y: The Pos Post-m t-mort ortem em Fat Fatee of Hum Human an Rem Remain ains. s. CRC Press, Boca Raton, pp. 151 164. Dowler, B., 1846. Experimental researches on the post-mortem contractility of the muscles, with observations on the reflex theory. N.Y. J. Med. 8, 305 339. Elmas, I., Baslo, B., Ertas, M., Kaya, M., 2001. Analysis of gastrocnemius compound muscle action potential in rat after death: significance for the estimation of early post-mortem interval. Forensic Sci. Int. 116, 125132. Gill-King, H., 1997. Chemical and ultrastructural aspects of decomposition. In: Haglund, W.D., Sorg, Sor g, M.H M.H.. (Ed (Eds.) s.),, For Forens ensic ic Tap Taphon honomy omy:: The Pos Post-m t-mort ortem em Fat Fatee of Hum Human an Rem Remain ains. s. CRC Press, Boca Raton, pp. 93108. Henssge, Henssg e, C., 1988. Death time est estima imati tion on in cas casee wor work k par partt I: the rectal rectal tem temper peratu ature re tim timee of death nomogram. Forensic Sci. Int. 38, 209236. Henssge, C., Madea, B., 2002. Practical casework. In: Knight, B. (Ed.), Estimation of the Time since Death in the Early Post-mortem Post-mortem Period, second ed. Edwa Edward rd Arnold Arnold,, pp. 244261. Henssge, C., Madea, B., Gallenkemper, E., 1988. Death time estimation in casework II: integration of different methods. Forensic Sci. Int. 39, 7787. Hoff, H.E., 1936. Galvani and the pre-Galvanian electrophysiologists. Ann. Sci. 1 (2), 157 172. Inoue, M., et al., 1994. Development of an instrument to measure post-mortem lividity and its preliminary application to estimate the time since death. Forensic Sci. Int. 65, 185 193. Kaatsch, H.-J. Kaatsch, H.-J.,, Schmi Schmidtke, dtke, E., Niet Nietsch, sch, W., 1994. Photome Photometric tric measurement measurement of pressure induced blanching blanc hing of livor morti mortiss as an aid to estimating estimating the time of death. Int. J. Legal Med. 106, 209214. Knight, B., 2002. Hypostasis and timing of death. In: Knight, B. (Ed.), Estimation of the Time since Death in the Early Post-mortem Post-mortem Period, second ed. Edwa Edward rd Arnold Arnold,, pp. 206208. Kobaya Koba yashi shi,, M. M.,, Ik Ikeg egay aya, a, H. H.,, Ta Taka kase se,, I. I.,, Ha Hata tana naka ka,, K. K.,, Sa Saku kura rada da,, K. K.,, Iw Iwas ase, e, H. H.,, 20 2001 01.. Developme Deve lopment nt of rigor mortis is not affected by muscl musclee volum volume. e. Forensic Sci. Int. 117, 213219. Krompe Krom pech cher er,, T. T.,, 20 2002 02.. Ri Rigo gorr mo mort rtis is:: es esti tima mati tion on of th thee ti time me si sinc ncee de deat ath h by ev eval alua uati tion on of cadaveric rigidity. In: Knight, B. (Ed.), Estimation of the Time since Death in the Early Postmortem Period, second ed. Edward Arnold, pp. 144 160. Krompecher, T., 1981. Experimental evaluation of rigor mortis: V. Effect of various temperatures on the evolution of rigor mortis. Forensic Sci. Int. 17, 19 26. Krompecher, T., 1994. Experimental evaluation of rigor mortis: VIII. Estimation of time since death by repeated measurements of intensity of rigor mortis on rats. Forensic Sci. Int. 68, 149 159.
Supravital Reactions in the Estimation of the Time Since Death (TSD)
11
Krompeche echer, r, T., Bergerioux, Bergerioux, C., 1988 1988.. Exper Experime imental ntal evaluation evaluation of rigor mortis: VII. Effec Effectt of Kromp ante- and post-mortem electrocution on the evolution of rigor mortis. Forensic Sci. Int. 38, 2735. Krompeche Krompe cher, r, T., Fry Fryc, c, O., 197 1978a. 8a. Exp Experi erime menta ntall evalua evaluatio tion n of rig rigor or mor mortis tis:: III III.. Com Compar parati ative ve study of the evolution of rigor mortis in different sized muscle groups in rats. Forensic Sci. Int. 12, 97102. Krompeche Krompe cher, r, T., Fry Fryc, c, O., 197 1978b. 8b. Exp Experi erimen menta tall eva evalua luatio tion n of rig rigor or mor mortis tis:: 1V. Cha Change nge in strength and evolution of rigor mortis in the case of physical exercise preceding death. Forensic Sci. Int. 12, 103107. Krompecher, T., Bergerioux, C., Brandt-Casadevall, C., Gujer, H.-R., 1983. Experimental evaluation ati on of rig rigor or mor mortis tis:: VI. Effect Effect of var variou iouss cau causes ses of death death on the evolutio evolution n of rig rigor or mor mortis tis.. Forensic Sci. Int. 22, 19. Krompecher, T., Gilles, A., Brandt-Casadevall, C., Mangin, P., 2008. Experimental evaluation of rigor mortis: IX. The influence of the breaking (mechanical solution) on the development of rigor mortis. Forensic Sci. Int. 176, 157162. Madea, B., 1992. Estimating the time of death from measurement of the electrical excitability of skeletal muscle. Forensic Sci. Soc. J. 32 (2), 117 129. Madea, B., 200 Madea, 2002a. 2a. Sup Suprav ravita italit lity y in tis tissue sues. s. In: Kni Knight ght,, B. (Ed (Ed.), .), Est Estima imatio tion n of the Tim Timee Sin Since ce Death Deat h in the Early Post-mortem Post-mortem Period, Period, secon second d ed. Edward Arnold, pp. 135142. Madea, B., 2002b. Post-mortem mechanical excitation of skeletal muscle. In: Knight, B. (Ed.), Esti Es tima mati tion on of th thee Time Time Si Sinc ncee De Deat ath h in th thee Ea Earl rly y Po Post st-m -mor orte tem m Pe Peri riod od,, se seco cond nd ed ed.. Ed Edwa ward rd
Arnold, pp. 160 164. Madea, B., 2002c. Post mortem electrical excitability of skeletal muscle in case work. In: Knight, B. (Ed (Ed.), .), Estimati Estimation on of the Tim Timee Sin Since ce Dea Death th in the Ear Early ly Pos Post-m t-mort ortem em Pe Perio riod, d, sec second ond ed. Edward Arnold, pp. 164206. Madea, B., Henssge, C., 1990. Electrical excitability of skeletal muscle post-mortem in casework. Forensic Sci. Int. 47, 207227. McDowall, K.L., Lenihan, D.V., Busuttil, A., Glasby, M.A., 1998. The use of absolute refractory period perio d in the estim estimatio ation n of early post-mortem post-mortem interval. interval. Forensic Forensic Sci. Int. 91, 163170. Nysten, P.H., 1811. Quatrième Section, Article Premier; De la contractilité des organs musculaire au prés présum umés és tels tels chez chez l hom hommes mes et les animaux animaux sang sang rouges rouges,, après après les divers divers genres genres de mort mort violente, Recherches de physiologie et de chimie pathologiques pour faire suite a celles de Bichat sur la vie et la mort. Paris, 1811. ’
Querido, D., 1994. Time-dependent changes in electrical resistance of the intact abdomen during the 1-504 hour post-mortem period in rats. Forensic Sci. Int. 67, 17 25. Querid Que rido, o, D., 200 2000. 0. Tem Temper peratu ature re cor correc rectio tion n of abd abdomi ominal nal imp impeda edance nce:: imp improv roved ed rel relati ations onship hip between impedance and post-mortem interval. Forensic Sci. Int. 109, 39 50. Querido, D., Phillips, M.R.B., 2001. Estimation of post-mortem interval, temperature correction of extracellular abdominal impedance during the first 21 days of death. Forensic Sci. Int. 116, 133138. Schuck, M., Beier, G., Liebhardt, E., Spann, W., 1979. On the estimation of lay time by measurements of rigor mortis. Forensic Sci. Int. 14, 171 176. Shapiro, H.A., 1950. Rigor mortis. Br. Med. J. 2 (4673), 304. Shapiro, H.A., 1954. Medico-legal mythology. J. Forensic Med. 1, 144169. Sung Sung Tz u, 11 1186 8612 1249 49.. The The Washi Washing ng Aw Away ay of Wron Wrongs gs.. Tr Tran ansl slat ated ed fr from om the the Chin Chines esee by McKnight, B.E., 1981. University of Michigan, Ann Arbor, p. 152. ’
12
Human Body Decomposition
Szent-Györgyi, A.G., 2004. The early history of the biochemistry of muscle contraction. J. Gen. Physiol. 123, 631641. Vain, A., Kaupilla, R., Humal, L.-H., Vuori, E., 1992. Grading rigor mortis with myotonometry a new possibility to estimate time of death. Forensic Sci. Int. 56, 147150. Vanezis, Vanez is, P., 1991. Assessing Assessing hypost hypostasis asis by colo colorimet rimetry. ry. Forensic Sci. Int. 52, 13. Vanezis, P., Trujillo, O., 1996. Evaluation of hypostasis using a colorimeter measuring system and its application to assessment of the post-mortem interval (time of death). Forensic Sci. Int. 78, 1928. Varetto, Varett o, L., Cur Curto, to, O., 200 2005. 5. Lon Long g per persis sisten tence ce of rig rigor or mor mortis tis at con consta stant nt low tem temper peratu ature. re. Forensic Sci. Int. 147, 3134. Warther, S., et al., 2012. Estimation of the time since death: post-mortem contractions of skeletal muscles muscl es follo following wing mech mechanica anicall stim stimulati ulation on (idi (idiomusc omuscular ular contr contractio action). n). Int. J. Lega Legall Med. 126, 399405.
2
CHAPTER
Algor Mortis and Temperature-Based Methods of Estimating the Time Since Death
EARLY RESEARCH IN THE 19TH CENTURY MID-20TH CENTURY RESEARCH 1950 to 1960: Early Attempts to Determine the TSD in the Early Postmortem Period The 1960s: The Search for an Accurate Formula The 1970s to Early 1980s: Continuing the Search for an Accurate Formula The Later 1980s: Henssge s Nomogram, a Definitive Mathematical Expression of TSD ’
CONTEMPORARY RESEARCH FROM 1988 TO THE PRESENT TIME OTHER STUDIES ATTEMPTING TO IMPROVE THE ESTIMATION OF TSD CONCLUSION REFERENCES
Algorr mort Algo mortis is re refe fers rs to the the co cool olin ing g of the the body body af afte terr deat death h unti untill it reaches ambient temperature. Methods of estimating the postmortem interval (PMI) from the rate of fall in temperature have been studied moree extens mor extensive ively ly than than any other other sta stage ge in the dec decomp omposi ositio tion n pro proces cess. s. Duri Du ring ng life life,, the the norm normal al body body temp temper erat atur uree of 37 37.4 .4 C is main mainta tain ined ed.. Afte Af terr deat death h the the ra rate te of heat heat lo loss ss depe depend ndss on se seve vera rall fact factor orss wh whic ich h incl includ udee th thee ambi ambien entt temp temper eratu ature re cond condit itio ions ns,, body body mas mass, s, whet whethe herr or not there was fever immediately prior to death, the presence or absence of clot clothi hing ng,, and and the the body body posi positi tion on.. A ti time me la lag g of va vari riab able le in inte terv rval al occurs before the body begins to cool, as a result of a temperature gradient developing between the core and the surface of the body. Human Body Decomposition. DOI: http://dx.doi.org/10.1016/B978-0-12-803691-4.00002-9 © 2016 Elsevier Inc. All rights reserved.
14
Human Body Decomposition
EARLY RESEARCH IN THE 19TH CENTURY Although the development of the thermometer has been attributed to several inventors over the centuries, it was Sir Thomas Clifford Allbutt who wh o devel develop oped ed th thee clin clinic ical al th ther ermo mome mete terr in 1870, 1870, re repl plac acin ing g the the foot foot-long instrument that took 20 minutes to register a patient s tempera’
ood,, 1951 1951)). Th This is in inve vent ntio ion n provi rovide ded d an im impe petu tuss to ture (Underw Underwood re rese sear arch ch,, incl includ udin ing g the the esti estima mati tion on of ti time me si sinc ncee deat death h (TSD (TSD)) in the the early postmortem period. The first researcher in the modern era to suggest that the loss of temperature after death might be useful as a basis for the estimation of the PMI was Dr John Davy John Davy (1839) (1839),, who conducted studies on the bodies of 10 British soldiers who had died in Malta in 1828. Autopsies and observations were made at times which varied from 2 to 17.5 hours af afte terr de deat ath. h. Ob Obse serv rvin ing g that that the the temp temper eratu ature re of the the in inte tern rnal al orga organs ns seemed to be higher than the external surface of the bodies and the environme enviro nmenta ntall temper temperatu atures res,, Davy took took tem temper peratu ature re measur measureme ements nts from under the left ventricles of the heart and from under the liver and he also recorded the room temperatures. All bodies apart from two had ha d di died ed fr fro om or with with inf infecti ection on and and as a resu result lt woul would d ha have ve had had a raised temperature before death. In all cases he found the temperature of the internal body to be raised above room temperature. He repeated his observat vations on nine soldiers who had died in a hospital in Chat Ch atha ham, m, Kent Kent,, in 18 1838 38 with with si simi mila larr fi find ndin ings gs.. Auto Autops psy y ti time mess af afte terr death varied from 4.5 to 29 hours, and all cases except for one had some form of infection at the time of death. He attributed the higher inte intern rnal al temp temper erat atur ures es to an unkn unknow own n heat-g heat-generat enerating ing process, process, as “
”
well as to the well the febr febril ilee di dise seas ase. e. This This stud study y used used a li limi mite ted d numb number er of cases, employed crude methods of observation and the time intervals after death when the observations were made varied, but the conclusion that these observations may enable the enquirer. . .. . .to arrive at a tolerably positive conclusion, in doubtful cases of death, as to the time time wh whic ich h may have have elap elapse sed, d, betw betwee een n the the fa fata tall even eventt and and th thee post post mortem examination examination (Da Davy vy,, 18 1839 39,, pp pp.. 24 247 7248 248)) po poin inte ted d the the way way toward further research in this field. “
”
(1863) published published a study of a series of 100 bodies Taylor and Wilks (1863) admitted to a morgue. They recorded the age, time of death, cause of death, deat h, humi humidi dity ty of th thee air, air, an and d post postmo mort rtem em te temp mper erat atur uree of ea each ch body, two to four observations being carried out on each body at
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
15
varying intervals up to 17 hours after being admitted to the morgue. However, the time of taking the first readings varied from the time of death up to 12 hours after death. The temperatures were taken by placing the thermometer bulb on the abdominal skin and the recordings in gs coul could d ther theref efor oree have have va vari ried ed wit with the the ambi ambien entt te temp mper erat atur uree chan ch ange ge.. Th Thee temp temper erat atur uree of the the mo morg rgue ue was was al also so re reco cord rded ed but but in some of the cases there was no record of the length of time before temper tem peratu ature re measur measureme ements nts we were re comme commenc nced. ed. Altho Although ugh the the condicondition tionss of the the stu tudy dy were were no nott co cont ntro roll lled ed,, Tayl Taylor or an and d Wi Wilk lkss ma made de several pertinent observations that laid the basis for future research. Firstly, the temperature of fat bodies remained at a higher temperatur at uree for for a long longer er pe peri riod od than than thin thin or em emac acia iate ted d bodi bodies es an and d th that at moisture in the atmosphere appeared to favor decomposition much moree than mor than hea heat. t. Se Seco condl ndly, y, bodie bodiess coole cooled d sl slowl owly y and progre progressi ssivel vely, y, retaining considerable heat for upwards of 12 hours after death and thee tem th tempe perat rature ure of the in inter ternal nal bod body y organ organss re remai maine ned d at a hi highe gherr “
”
temperature the skin for longer period. Thirdly, nized that thethan environment in awhich a body was found, they and recogcoverings, or lack of, on a body could modify the rate at which a body cooled. Lastly, they observed that in certain circumstances such as a person dying of fever or found with heavy body coverings, an initial ri rise se in temp temper erat atur uree occu occurr rred ed befo before re a stea steady dy fall fall but but th they ey di did d not not obse ob serv rvee th thee temp temper erat atur uree pl plat atea eau u befo before re the the fall fall in te temp mper erat atur ure. e. Taylor and Wilks (1863, p. 202) concluded that the changes which take place in a dead body before the commencement of putrefaction, may, ma y, if accu accura rate tely ly obse observ rved ed,, enab enable le a medi medica call witn witnes esss to form form an opinion of the time at which the deceased died. “
”
Rainy (18 Rainy (1868) 68) produ produce ced d a pres presci cien entt stud study y in whic which h he re reco cogn gniz ized ed that the evidence of TSD depended not only on the excess of body temperature over that of the surrounding medium but also on the presence and duration of rigor mortis and the presence and progress of chemical decomposition. He was the first to recognize that the cooling of a cadaver does not follow Newton s law of cooling (Winterton, ( Winterton, 1999). 1999 ). Rainy detailed a study of postmortem temperature measurements on 46 bodies up to 63 hours after death. He recognized the necessity of standardizing the study conditions by attempting to take the measurements in a room with still air at a uniform temperature. ’
Therefore the observations on 54 bodies out of 100 at the beginning of the study were discarded because the temperature of the morgue
16
Human Body Decomposition
did not remain steady, which was necessary in order to eliminate as far as possible all circumstances which might complicate the law of cooling. He also would have wished the observations to have been made at the same intervals after death but admitted that this was impracticable because of the varying times over a 24-hour period at which the cases had died. Measurements were taken from the abdominal skin at the umbilicus and from the rectum. The temperature of the morgue and three rectal temperature readings were recorded in all cases and four readings in 18 cases. Rainy then calculated the rati ra tio o per per hour hour at wh whic ich h the the temp temper erat atur uree was was fo foun und d to vary vary by determining how much the temperature of the rectum varied above ambient temperature at the start of the measuring process and for each hour thereafter. Rainy observed that bodies placed in a colder medium gradually lost heat until they reached the temperature of the medi me dium um but but th thee rate rate of cool coolin ing g sl slow owed ed as the the te temp mper erat atur uree of the the body approached that of the medium. He recognized that this pattern “
”
’
of olin cooling ofh astat follow s empirical of co cool ing g wh whic ich stcadaver ates es that thatdid th theenot ra rate te of loss loNewton ss of te temp mper erat atur uree of law a hot ho t body bo dy is dire direct ctly ly pr prop opor orti tion onal al to th thee diff differ eren ence ce of te temp mper erat atur uree betw be twee een n its its surf surfac acee an and d the the su surr rrou ound ndin ing g cool cool me medi dium um (Winterton, 1999). 1999 ). Whe When n th thee differ differen ence ce be betwe tween en the the surrou surroundi nding ng envir environ onmen mentt and the body temperature was moderate, by which he presumably meant low, the rate of cooling very nearly approximated the law but when the difference was greater, body temperature was found to rise before gradually cooling at a steady rate and then at a slower rate as it approach approached ed amb ambient ient temperat temperature. ure. His desc descript ription ion was essentia essentially lly that of the temperature plateau before the fall in body temperature “
”
and then a slower fall to ambient temperature, the cooling curve first desc de scri ribe bed d in th thee 20th 20th cent centur ury y by Sh Shapir apiro o (195 (1954) 4).. He could not explain the initial temperature plateau and the later more rapid fall in temperature other than by the gradual cessation of the calorific proces pro cesses ses whi which, ch, in the earlie earlierr stage stages, s, retard retard the the coolin cooling g (Rainy, 1868 18 68,, p. 32 328 8). Rain Rainy y stre stress ssed ed the the ne need ed for for tw two o re rect ctal al te temp mper erat atur uree measurements, taken at least an hour apart, emphasizing that a precise time of death could not be given, but only a maximum and a mini mi nimu mum m pe peri riod od du duri ring ng whic which h deat death h co coul uld d have have occu occurr rred ed.. He devised a formula which calculated the number of hours since death but the time calculated by his formula would, in almost every case, be less less than than the the ac actu tual al TSD TSD beca becaus usee of the the sl slow ower er co cool olin ing g of the the body in the early stage. Rainy stated that it was more difficult to “
”
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
17
give a maximum number of hours, but that from observation of cases in which the temperature of the rectum was found to be below 85 F (29.4 C) C),, the the time time elap elapse sed d si sinc ncee deat death h was was no nott le less ss th than an the the ti time me deduced from the formula multiplied by 1.5. This figure of 1.5 was also quoted by by Glaister (1942) when he suggested the use of the following formula as a method of estimating the time of death; ðNormaltemp: 98:4 Fð36:9 CÞ Rectaltemp:Þ Approx: hourssincedeath 5 1:5 (2.1) As late as 1973 this formula was quoted as being useful, when it was stated that a body tended to lose heat at the rate of 1.5 F (0.8 C) ntou oull an and d Sm Smit ith, h, 19 1973 73). ). Even Even at the the pres presen entt ti time me this this per hour (Re (Rent figure is still used as an approximate rate of heat loss (Clark ( Clark et al., 1997). 1997 ). Burman (1880) recorded (1880) recorded the temperature of nine cadavers from the moment of death for varying periods as they cooled. Recordings were take ta ken n with with a lo long ng-h -han andl dled ed th ther ermo mome mete terr bent bent at a ri righ ghtt angl anglee and and placed in the axilla so that the temperature could be read without disturb tu rbin ing g its its pl plac acem emen entt in the the ax axil illa la.. Re Reco cord rdin ings gs were were ta take ken n at le leas astt every hour and in some cases more frequently. Burman recognized the importance of measuring internal body temperature because the internal organs remained at a higher temperature than the exterior of the body and that there was a gradient of heat loss from the interior of the body to the external body surface but he regretted that he did not do this. He also acknowledged the effect of the various different environmental ment al fact factor orss su such ch as the the en envi viro ronm nmen enta tall te temp mper erat atur ure, e, whet whethe herr th thee body was in air or water, and the amount and nature of clothing in determining the rate of cooling, and he tried to standardize environmental conditions as much as possible. He calculated that the average rate of cooling was 1.6 F (0.8 C) per hour and he concluded that the estimated number of hours since death could be calculated by subtracting the axillary temperature from the temperature at the time of death, whic wh ich h he as assu sume med d was was 98 98.4 .4 F (36. (36.9 9 C). Burm Burman an (18 (1880) 80) recognized that the temperature could, in some cases, rise above 98.4 F (36.9 C) after death but he dismissed this as being a rare occurrence. He noted an initial rapid fall in temperature in the first few hours after death and that the higher the original temperature at death, the faster was the initial fall in temperature but that the cooling of a body was slower
18
Human Body Decomposition
than ha than had d been been gene genera rall lly y th thou ough ght. t. Ev Even en af afte terr 30 hour hourss a body body ma may y retain a temperature 2 23 F above that of the room. (1887) used used a series of mathematical calculations to come Womack (1887) to the conclusion that a cadaver did not cool as quickly as a liquid would by radiation and convection under similar conditions. Remar Re markab kably ly he clai claime med d to be able able to accu accura rate tely ly dete determ rmin inee cadav cadaver er temperature to within 0.025 F by using a very thin glass thermometer with a mercury bulb strapped to the abdominal wall. He acknowledged that cooling of the cadaver was affected by a number of variable factors to rs whic which h incl includ uded ed th thee unkn unknown own temp temper erat atur uree at deat death, h, the the varyi varying ng temperature of the atmosphere, coverings or lack of on the body, and the locality of the body. He did not recognize the initial temperature plateau or subsequent rapid fall before the slower decline in temperature as it reached ambient temperature. Womack attempted to achieve a great degree of accuracy with conditions which were not controlled or standardized standardized.. From the latter part of the 19th century until the mid-20th century, no papers of practical value were published, possibly because of the intervention of the Boer War and two World Wars, and research into TSD TS D di did d not not ad adva vanc nce. e. In th this is re resp spec ectt it is in inte tere rest stin ing g to note note that that Burman (1880) (1880) recognized the inaccuracy of using the then customary method of gauging the temperature of the skin of a body by using the back of the hand. In all cases he advocated the use of a thermometer. However, even as late as 1921 research had advanced no further, as was illu illust stra rate ted d by Dr E.M. E.M. Vaug Vaugha han, n, a Medi Medica call Assi Assist stan antt Dist Distri rict ct Atto At torn rney ey in Broo Brookl klyn yn,, when when he de desc scri ribe bed d an el elab abor orat atee meth method od of approximately estimating the time of death by touching various parts of the body extremities (Vaughan, (Vaughan, 1921). 1921).
MID-20TH MID-2 0TH CENTURY CENTURY RESEARCH RESEARCH 1950 to 1960: Early Attempts to Determine the TSD in the Early Postmortem Period A summary of the state of research into the estimation of TSD in the early 1950s was given by F.J. Cairns in 1952 when he outlined the facto factors rs whi which ch modifi modified ed body body cooli cooling ng immed immediat iatel ely y after after death death (Cairns and Forbes, 1952). 1952). These were the effect of different environmental temperatures, whether the body was clothed or naked and foun fo und d in wa wate terr or on dry dry land land.. Vari Variou ouss auth author orit itie iess had had qu quot oted ed
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
19
formulae and he recognized that an important modifying factor was the temperature at the time of death. The biochemical mechanism of rigor mortis still had not been elucidated and as a method of estimating the TSD it was falling into disfavor. In fact, the estimation of TSD at this time was still an observational and qualitative exercise based on experience and only an approximation could be given. In the same article, J.A. Forbes outlined a study of some cooling curv cu rves es fr from om ca cada dave vers rs at the the Roya Royall Melb Melbou ourn rnee Ho Hosp spit ital al (Cairns and Forbes, 1952). 1952). Temperatures had been recorded from the upper par artt of the abdo bdomen men, liver iver,, th tho orax ax,, rect ctu um, thigh, gh, and brai ain n. Intr In traa aabd bdom omin inal al temp temper erat atur ures es we were re ob obta tain ined ed thro throug ugh h a punc punctu ture re woun wo und d belo below w th thee ri righ ghtt lobe lobe of the the li live verr and and re reco cord rdin ings gs from from this this posi po siti tion on sh show owed ed cont contin inue ued d heat heat pr prod oduc ucti tion on af afte terr deat death h wi with th the the curves showing a rise, then a plateau, before a fall. The environmental temperature and obesity affected the rate of fall but overall there was inconsistency in the estimation of the TSD. If a certain temperature wasofselected theredeath, was awhile greatifvariation between in the number hours since a certain TSD was cases selected the correspo corr espondin nding g intr intra-ab a-abdomi dominal nal temperat temperatures ures were widely widely scattere scattered d in co comp mpar arab able le ca case sess so that that th thee esti estima mati tion on of the the TSD TSD see eeme med d impossible. On the other hand, Forbes found that the temperature curves from the brains were logarithmic in form and ran in a parallel fashion, thus conforming to Newton s law of cooling, suggesting that further work might provide data for a more accurate estimation of the TSD. ’
A study reported by Schwarz by Schwarz and Heidenwolf (1953) set (1953) set the pattern of research that would occur during the rest of the century. Their opinion was that not enough emphasis had been given until that time, of determining the time of death from the temperature of the body. They studied 39 adult bodies, in 16 of which the time of death was known. Of the remaining 23 cases studied, only 9 had not died with a fever. They established cooling curves in 25 bodies which had not died with a fever and in which the temperature at the time of death was 37 C. The authors found that when the environmental conditions were standardized, cooling was exclusively due to the difference between the temperature of the interior of the body and that of the surrounding atmosphere. Cooling was more rapid initially because of the greater difference in temperature between the body and the surrounding air. In cases dying with a fever, they stated that no definite conclusions could
20
Human Body Decomposition
be drawn as to the time of death from postmortem cooling as the tempera pe ratu ture re ma may y cont contin inue ue to ri rise se afte afterr de deat ath h due due to the the pres presen ence ce and and action of microorganisms and this would affect the rate of cooling to an unknown extent. All cases had the temperature taken from a thermometer inserted in the anus and the cooling curves they established recognized the slow reduction of temperature at the start of cooling which they attributed to the slower cooling of the interior of the body until it reached equilibrium with the surface. In none of the cases was there an initial rise in temperature before a fall. The temperature then fell more rapidly but progressively and then more slowly as it approached equilibrium with the surrounding atmosphere, reaching ambient temperature after about 36 hours. The estimation of the TSD became less accurate as thee sp th spee eed d of cool coolin ing g decr decrea ease sed d towa toward rd am ambi bien entt te temp mper erat atur uree due due to inev inevit itab able le varia variati tion onss in room room temp temper erat atur ure. e. Apart Apart from from pr prem emor orte tem m fever, another intrinsic factor influencing the rate of cooling was the ratio between the surface area and the unit weight of the body. The study was on unclothed bodies but their judgment was that the influence of clothing seemed to cause the same variation as the difference in size of the body. Schwarz and Heidenwolf seem to have been the first to suggest the necessity of determining a mathematical formula, incorporating the various factors affecting a body cooling, in order to estimate the time of death but that an exact mathematical definition which would take into account various additional thermal influences would be too complex considering the great variability of the factors concerned in the loss of heat (Fig. (Fig. 2.1 2.1). ). The present, more scientific era of research into the estimation of TSD began when De when De Saram and colleagues (1955) recorded the temperature fall in a series of 41 bodies of executed prisoners in Ceylon (Sri Lanka). It was possible to control the conditions of the collection, storage, and recording of the bodies much more closely than in previous studies because the prisoners were all executed at 8 am in the morning. Recording of the body temperatures was standardized by taki taking ng re rect ctal al temp temper erat atur ures es fr from om 3 to 4 inch inches es (7 (7.6 .6 to 10.2 10.2 cm) cm) within wit hin the re rectu ctum. m. Readi Readings ngs were were tak taken en half-h half-hour ourly ly for for 12 hours hours.. The bodies were all examined in the same room in Colombo except for five five wh which ich wer weree execut executed ed in Kandy Kandy bu butt where where condit condition ionss we were re different but similarly controlled. Room temperatures and humidity
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
21
°C 37
20
36
18
34
16
32
14
30
12
28
10
26
8
24
6
22
4
20 18 17 16
AUSKÜHLZEIT AUS KÜHLZEIT
2 0
0
2
4
6
8 10 10 12 14 16 18 20 22 24 26 28 30 32 34 36
h
°C 37 36 34 32 30 28 26 24 22 20
AUSKÜHLZEIT AU SKÜHLZEIT
18 17 16
0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
h
Figure 2.1 Cooling curves established by Schwarz by Schwarz and Heidenwolf (1953). (1953) .
weree re wer recor corde ded. d. The obser observat vation ionss confi confirme rmed d in al alll cases cases th that at there there was a lag was lag peri period od in temp temper erat atur uree of 45 minu minute tes, s, when when it rema remain ined ed essentially the same, before a rapid fall and then a slowing as temperature reached that of the room. The authors recognized that the difference between the body temperature and room temperature did have a significant effect on the cooling rate. Conduction, radiation, and especially evaporation which depended on the ambient humidity weree als wer also o impor importan tantt fac factor tors. s. The body body size size facto factorr whi which ch in incor corpo po-rated body weight and surface area was also important. Clothing on the bodies did not significantly influence the rate of heat loss. By a comple com plex x proce process, ss, a for formul mula a was calcul calculate ated d whi which ch incor incorpor porate ated d th thee infl in flue uenc ncee of so some me of th thee modi modify fyin ing g fact factor ors, s, exce except pt th thee body body si size ze fact fa ctor or an and d hu humi midi dity ty an and d it was was sugg sugges este ted d that that th this is coul could d be used used with wi th re reas ason onab able le accu accura racy cy with within in 8 ho hour urss of de deat ath, h, if th thee room room
22
Human Body Decomposition
temperature remained constant and an initial rectal temperature of 99.6 F (37.6 C) was assumed. These conditions, however, cannot be controlled in normal practice. Shapiro Shapir o (19 (1954) 54) drew drew atte attent ntio ion n to the the in init itia iall te temp mper erat atur uree pl plat atea eau u af afte terr deat death, h, duri during ng whic which h time time th thee bo body dy te temp mper erat atur uree di did d not not fa fall ll,, Rainy ny (18 (1868) 68) had had ma made de.. Shap Shapir iro o esse essent ntia iall lly y the the same same find findin ing g that that Rai found that this plateau could be maintained for up to 4 hours and he attrib att ribute uted d this this to the tem temper peratu ature re gradie gradient nt between between the cen centra trall body body core and the external surface, the surface cooling occurring before the central organs. Lyle and Cleveland studied (1956) 69 bodies over 2 years with the purpose of determining the rate of postmortem cooling in the first 24 hours. The times of death were known, the time before measurements began was as short as possible, and thermocouples were used to measure the temperature from the exposed skin of the forehead and the covered skin of the chest. Thermocouples recording rectal temperature weree pla wer placed ced in wat waterer-tig tight ht gold-f gold-fill illed ed tub tubes es and the liv liver, er, qua quadri dricep cepss musc mu scle le of the the thi high gh,, and and brai brain n ther thermo moco coup uple less were were atta attach ched ed to 19 gauge gauge needle needles. s. Enviro Environme nmenta ntall condit condition ionss wer weree as sta standa ndardi rdized zed as possible: 24 cases were in a room kept between 75 F and 78 F (23.9 C and an d 25 25.6 .6 C) C),, 22 ca case sess be betw twee een n 65 F and 68 F (18. (18.3 3 C and 20 C), and 23 cases between 62 F and 78 F (16.7 C and 25.6 C). Time Time temperature charts were constructed for each case incorporating thee fall th fall in temp temper erat atur uree fr from om ea each ch si site te.. Lyle Lyle an and d Cl Clev evel elan and d (1 (1956 956)) showed that the rate of heat loss varied considerably depending on the site of recording. The range of temperature variability was less in the brain and progressively more variable in the rectum, liver, muscle, and skin. They concluded that the rate of heat loss depended on the environmental temperature as well as body structure, physiological activity, putrefaction, and external insulation and that the time of death by measurement of heat loss was not satisfactory after 24 hours post mortem. These factors had less effect on heat loss from the brain than from the other organs. In contrast to Shapiro s previous study 2 years o, 1954 1954), ), Lyle and Cleveland (1956) did not recognize earlier (Shapir (Shapiro, any an y temp temper erat atur uree pl plat atea eau u befo before re th thee more more rapid rapid fa fall ll in te temp mper erat atur ure. e. The authors admitted they were not primarily concerned with establishing the TSD but rather with establishing rates of heat loss under known conditions. This was because they believed that determining the ’
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
23
exact time of death could be accomplished only after the establishment of rates of heat loss from the body when conditions were controlled as well as when conditions were variable. A study of over 100 bodies during a 4-year period admitted to a morg mo rgue ue wa wass ma mad de by Fidd Fiddes es an and d Pa Patt tten en (1 (195 958) 8),, in an attempt to impr im prov ovee the the accu accura racy cy of de death ath time time es esti tima mati tion on.. Th Thee bodi bodies es were were of varying stature, weight, and clothing cover and the time of death was known precisely. The environmental temperature varied between 40 F (4.4 C) and 70 F (21.1 C). The authors recorded the rectal and atmospheric temperatures at intervals, initially with a standard thermometer but later with an electrical continuous recording apparatus. They arbitr trar aril ily y pr pres esum umed ed that that th thee re rect ctal al temp temper erat atur uree at de deat ath h was was 99.0 99.0 F (37.2 C) and and they they conf confir irme med d the the expo expone nent ntia iall fa fall ll in te temp mper erat atur uree bu butt with considerable variation in the rate of cooling of each body (a range of 0.9 F or 0.5 C per hour in a series of 20 bodies). They noted that for the first 12 hours after death the TSD conformed reasonably accurately to the formula stated by Gl Glais aister ter (19 (1942) 42) (see Eq. Eq. (2 (2.1 .1)). )). However, as a body approached within 7 F (4 C) of ambient temperature, further cooling became extremely slow and temperature recording was not accurate in determining the TSD by any formula which relied on only one temperature reading. They found that the time taken to reach ambient temperature was much longer than in previous studies, not uncommonly up to 60 6070 hours. Fiddes and Patten (1958) therefore discarded the last 15% of the temperature difference between the rectal rec tal tem temper peratu ature re at deat death h and the pre prevai vailin ling g atmosp atmospher heric ic temper temperaature and considered that the body had virtually cooled when it had fallen by 85% of the difference in time taken for the body temperature to fall from the rectal temperature at death to ambient temperature. They called this the virtual cooling time (VCT) and they justified the empi em piri rica call us usee of the the figu figurre of 85% 85% fo forr th thee VC VCT T beca becaus usee they they had had assumed a rectal temperature at time of death of 99 F rather than the normal body temperature of 98.4 F. The virtual temperature difference (VTD) at the time of death was defined as 85% of the total difference between the rectal temperature at death and atmospheric temperature. To overcome the fact that bodies cooled at different rates, the temperature fall recordings were noted as percentages of the VCT and VTD rather than in degrees and hours. “
”
The VTD was then plotted in a graph against the VCT for 50 cases selected because they did not suffer from illnesses which would have
24
Human Body Decomposition
greatly altered the temperature at death such as febrile illnesses, protrac tracte ted d de deat aths hs,, or mark marked ed var varia iati tion onss in envi enviro ronm nmen enta tall te temp mper erat atur uree after death. It was noted that there was an exponential correlation and a linear relationship was obtained when these figures were converted to logarithmic values. There was a slight flattening of the line initially which whi ch the they y sug suggest gested ed corres correspon ponded ded to the ini initia tiall platea plateau u commen commented ted (1954) 4).. Fi Fidd ddes es and and Patt Patten en sugg sugges este ted d that that th this is gr grap aph h on by Shapiro Shapiro (195 could be useful in estimating the early TSD regardless of body stature, clothi clo thing, ng, or enviro environme nmenta ntall tem temper peratu ature re provid provided ed the body body remain remained ed undisturbed and that two or more rectal temperature recordings were taken over a period of several hours. They did acknowledge that if the re rect ctal al temp temper erat atur uree at deat death h was was co cons nsid ider erab ably ly hi high gher er or lo lowe werr th than an 99 F, deductions by their method would be misleading.
The 1960s: The Search for an Accurate Formula In 1962, 1962, Thom Thomas as Mars Marsha hall ll,, a pa path thol olog ogis istt work workin ing g in coll collab abor orat atio ion n with F.E. Hoare, published theirbased workon onthe estimating time of death usinga aphysicist, mathematical formula recordingthe of rectal temperatures (Marshall (Marshall and Hoare, 1962 1962). ). Marshall and Hoare ton, 1999 1999), ), whic which h stat states es considered consid ered Newton s law of cooling (Winter (Winterton, that the rate of heat loss from a hot object cooling in air is directly proportional to the excess of temperature of the object over that of its surroundings, inadequate to explain the rate of heat loss from a human body after death. The study was carried out on more than 100 bodies over a period of 7 years, in which the temperatures were taken in the same unheated room and in which the environmental temperature varied no more than 4 F (1 (12 C) during the period of each body record’
ing, although it varied somewhat with the seasons (4.4 (4.410 C between Jan. and Mar. and 15.5 15.521.1 C between Jun. and Aug.). Because the environmental temperature varied so much they considered it unnecessary to use an elaborate temperature recording system, using instead mercury in glass thermometers sealed in copper cases. Acknowledging thee fact th fact th that at in inte tern rnal al bo body dy temp temper erat atur ures es from from di diff ffer eren entt part partss of th thee body fluctuated, and the best that could be done to measure internal body bo dy temp temper erat atur uree wa wass fr from om some some pa part rt of it it,, Mars Marsha hall ll and and Ho Hoar aree reco re cord rded ed seri serial al tem empe pera ratu ture ress fr from om the the cl clos osed ed axil axillla, 34 in inch ches es (7.6 (7.6 10.2 cm) within the rectum, and under the liver via a stab wound below the right costal margin. Weight and height were recorded and each ea ch body body wa wass nake naked d and and in insu sula late ted d from from th thee tabl tablee surf surfac acee by two two sheets she ets.. Temper Temperatu ature re record recording ingss were were taken taken halfhalf- to one one-ho -hourl urly y until until
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
25
they had fallen below 70 F (21 C). Graphs were constructed to illustrate the temperature falls. The authors showed that temperature falls were regular but varied in the rate of fall from different sites. Liver temperatures were the highest, rectal temperatures slightly lower, and both fell at about the same rate but more slowly than the axillary temperature recordings which were the lowest and most rapid to fall. The influence of environmental temperature was demonstrated by placing a body in a room 8 810 F (4 (45 C) higher than the temperature of the liver and the rectum, when it was found that the axillary temperature rose almost at once, but there was a delay of 12 hour hourss be befo fore re thee live th liverr an and d re rect ctal al temp temper erat atur ures es rose rose.. In some some case casess Mar Marsh shal alll and and Hoare noted a sharp rise in the rectal and liver temperatures occurring in the first half hour of the experiment, which they attributed to local cooling caused by the insertion of the thermometer but they cautioned against interpreting this temperature rise to continued heat production within the body after death. The cooling curves from the rectum and liver followed sigmoid form by with an initial of slow cooling lasting up to 5 a hours, followed a rapid evenplateau fall of temperature and a final slow, gradual decrease in temperature until ambient temperature was reached, after about 30 hours. In contrast, the axillary temperature often rose steeply during the first half or one hour, attributed to a rise in temperature following the placement of the thermometer after opening the axilla to the surrounding lower air temperature. The authors then proceeded mathematically to show that the cooling of a body did not follow Newton s law of cooling which assumes that the body is in temperature equilibrium and that temperature differences do not exist or are sufficiently small for their effects to be disregarded. However, temperature differences do exist within the human body and after death they may change, either rising or falling. If the body cooled according to Newton s law, the graphical representation woul wo uld d be a stra straig ight ht line line of gra gradu dual ally ly re redu duci cing ng grad gradie ient nt.. Howe Howeve verr it was known that the cooling curve followed a sigmoid pattern. Therefore Newton s law did not apply in the first 12 hours after death as the temperature maintained a plateau or even rose slightly. Marshall and Hoare attributed this to the development of temperature grad gr adie ient ntss betw betwee een n the the in inte tern rnal al body body orga organs ns an and d th thee surf surfac acee of the the body bo dy.. Afte Afterr about about 12 ho hour urss the the temp temper erat atur uree then then fe fell ll grad gradua uall lly y and and ’
’
’
’
conformed more closely to Newton s law. It was also recognized that thee ra th rate te of body body cool coolin ing g depe depend nded ed on body body size size and and surf surfac acee area area..
26
Human Body Decomposition
The authors then proceeded to devise a complex formula containing two exponential terms to express the rate of cooling of different bodies: B 5 ðθ Þ 2
C
Z 2 p
(2.2)
B , C , Z , and p were were consta constants nts for the corpse corpse und under er observ observati ation. on. B was a measure of the cooling of the body if it had cooled according to Newton s law of cooling. The value of C was the excess of body temperature over its environment at the moment of death. (θ ) was the temperature at the start of observations, Z was the cooling fact factor or whi which ch expr expres esse sed d the the co cool olin ing g prop propor orti tion onal al to the the te temp mper erat atur uree excess of the body over its environment, and p determined the rate of incr increa ease se of Z. Marsh Marshal alll and and Hoar Hoaree cl clai aime med d that that th this is fo form rmul ula, a, used in the experimental data, reproduced the cooling observed in reality with considerable accuracy. “
”
’
“
”
“
“
“
“
”
”
”
”
In a second paper by Marshall Marshall (1962a) (1962a),, the the cool coolin ing g form formul ula a was was used to construct curves showing the theoretical cooling from the time of death, of any naked body lying in still air, at a uniform temperature, once the height, weight, and external temperature were known. Four cons co nsta tant ntss were were re requ quir ired ed to cons constr truc uctt th thee fo form rmul ula a whic which h the the auth author or denoted as as B , C , Z , and and p. A size factor was calculated from the mass and surface area of the body and defined as being inversely proportion tional al to buil build, d, its its valu valuee in incr crea easi sing ng as the the body body beca became me thin thinne nerr or smaller. The cooling factor expressed the rate of cooling per degree B depended on the value of temper tem peratu ature re differ differenc ence. e. The consta constant nt B of p and its value was difficult to determine. From a study of the cooling p calculated. curves of certain anthe average of be of of 0.4 was These eventually determined which cases, enabled value value of B to to constants were then used in the cooling formula to calculate the theoretical temperature differences at various times after death. When the environmental temperature was added it gave the theoretical body temperatures at these various times. This enabled a series of cooling curves to be constructed for bodies with size factors from 170 F to 280 and in environmental temperatures from 40 F to 75 F (4.4 C to 23.8 C) at 5 F (2.8 C) intervals. When constructed, the curves graphically illustrated the influence of ambient temperature, an average body with a size factor of 210 cooled 2.5 times faster in a room at 4.4 C compared to a room at 23.8 C. The influence of body size was less pronounced in its effect on body cooling; after 18 hours in a room at 23.8 C, an
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
27
obese and a thin body would differ in temperature by 4.2 C but in a room at 4.4 C, the difference would be 9.7 C. Three sets of cooling curve graphs were constructed for size factors of 210, for an average sized body, 170 for an obese body, and 280 for a thin adult. These graphs could be used to read the time of death, if the rectal temperature and weight of the body were known. The curves reproduced the initial temperature plateau, or delay in cooling, which occurred after death dea th and which which dep depend ended ed on environ environmen mental tal temper temperatu ature re and body body size. In all the curves constructed, the delay in body cooling was most evident at higher temperatures, and a body of greater size took longer to cool than a smaller-sized body in a room at the same temperature. A constant rate of cooling was least likely to occur in thin adults cooling in a hot environment. Once the initial delay had passed the rate of cool co olin ing g prog progre ress ssiv ivel ely y decr decrea ease sed, d, clos closel ely y foll follow owin ing g the the prin princi cipl plee of Newton s law of cooling. As the body approached environmental temperature, after about 12 hours the rate of cooling became much slower. ’
The prolonged timefor toconvenience reach room Marshall temperature was not onconthe cooling curves and invoked the shown previous cept of VCT to consider the body cooled, when it had fallen through 85% of its initial temperature (Fiddes (Fiddes and Patten, 1958). 1958). Regardless of the size of the body, it took the same time to cool to 85% of the initial temp temper erat atur ure. e. Howe Howeve ver, r, th thee auth author orss foun found d th that at th thee comp comple lete te cooling time of time of a body was determined by the size of the body and not by the room temperature, a heavy body taking on average longer than a light body, an obese body taking about 41 hours, an average sized body 28 hours, and a thin body 19 hours. The rate rate of cool coolin ing g howeve howeverr was determined by the room temperature as well as by the size of the body. Any formula to predict the TSD would have no value unless both of these factors were taken into account. Marshall constructed cooling curves with temperatures taken from under the liver but found that the rate of fall in temperature differed depending on where the thermometer was placed in the upper abdomen. The author also found that although the rate of cooling increased more quickly than in the rectum, it did not reach the same rate as in the rectum. The error rate over a period of hours would therefore be greater than if temperatures were taken from the rectum. For ease of access and accuracy of recordings, the rectum was found to be the best region of the body in which to record temperatures. Finally, cooling curv cu rves es were were cons constr truc ucte ted d for for clot clothe hed d bodi bodies es and and Marsh Marshall all conc conclu lude ded d
28
Human Body Decomposition
that there was no fundamental difference in the cooling rate of clothed bodies from that of naked bodies, but that a larger study was required to confirm this. A third paper by Marshall by Marshall (1962b) critically (1962b) critically reviewed the use of the cooling formula and standard cooling curves. While initially it seemed a si simp mple le matt matter er to esti estima mate te the the time time of de deat ath h from from th thee appr approp opri riat atee curve after the rectal temperature, the height and weight of the body had been measured in ideal conditions; in practice there were too many factors which would modify the result. Firstly, the cooling curves were constructed for use with average cases, but small differences from the aver av erag agee va valu luee of th thee cool coolin ing g fact factor or and and one one of th thee cons consta tant ntss co coul uld d translate to a larger error when the final estimation of the TSD was calculated. Similarly, although the use of an average value for for p of 0.4 gave no greater average error than when it was calculated, this error could still not be ignored. Secondly, the experiments to construct the cooling curves were carried out in a room with a constant temperature, but in practice the temperature could vary due to factors such as doors and window windowss ope openin ning, g, people people s mov moveme ements nts and the their ir heat heat radiat radiation ion,, and bed clothes being removed from a body. The cooling curve of a body bo dy in such such an en envi viro ronm nmen entt coul could d vary vary so mu much ch from from a stan standa dard rd cooling curve as to preclude any comparison. A third source of error was to assume that every corpse at death had the same rectal temperature, but a survey of 2000 consecutive bodies admitted to Leeds City morgue found over 50 cases in which the rectal temperature was higher than would have been expected (Marshall, (Marshall, 1962b). 1962b). When cases with a feve fever, r, or foun found d with with heav heavy y be bedc dclo loth thes es,, were were excl exclud uded ed,, ther theree stil stilll remained 28 bodies where the cause of the raised temperature could nott be expl no explai aine ned. d. This This had had the the effe effect ct of di disp spla laci cing ng th thee curv curvee up or down or altering its slope so significantly, that the final time of death would be altered, in one case by as much as 9.2 hours. By far the greatest source of error was the variable environment to which the individual bodies had been subjected. Finally, calculation of the size factor could be difficult, because the different positions in which each body was foun found d coul could d lead lead to di diff ffer eren entt quan quanti titi ties es of he heat at lo loss ss from from ea each ch body. Therefore the size factor would be difficult to standardize and over a period of time the error in estimating the time of death would be magnified. For all these reasons the use of body temperature to esti’
mate the time of death could never give consistently accurate results and must be accepted only as being approximate.
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
29
James and Knight published a study of 110 bodies in 1965, estimating the time of death based on the method of Marshall and Hoare (Knight and James, 1965). 1965). All bodies had been subjected to changed environmental conditions, had varying types of clothing, and the times of death, which varied from 20 minutes to 75 hours, were known but nott by the no the exam examin iner ers. s. Temp Temper erat atur ures es were were take taken n from from de deep ep wi with thin in the rectum and the temperature at death was assumed to be 37 C. The outd ou tdoo oorr shad shadee temp temper erat atur uree at no noon on,, the the te temp mper erat atur uree at th thee ti time me the body arrived in the morgue, and the previous locations of the bodies ies were were re reco cord rded ed.. Usin Using g thes thesee da data ta,, a grou group p of ar arbi bitr trar ary y fact factor orss (1, 1.25, 1.5, 1.75, and 2) by which the fall in rectal temperature was multiplied were selected to obtain a first approximation of the number of hours since death depending on whether the average air temperature was 0 C, 5 C, 10 C, 15 C, or 20 C, respectively, above average environmental temperature. An exam amp ple was given of a body in an assessed average air temperature of 10 C with a rectal temperature of
3
~
27 C.estimation In that case estimated TSD was (37 (37 27) 27) 1.5 to15account hours. This wasthe then arbitrarily altered by the examiner for unusual body clothing or environmental temperature. Estimates of the time of death were all given to the quarter hour. The authors were aware that they had not strictly adhered to Marshall and Hoare s criteria. The TSD was correctly estimated in 11 bodies, underestimated in 57 and overestimated in 32. In 35 of 100 bodies the error rate was less than 10%, in 54 less than 30%, in 90 less than 40%, in 96 less than 50%, and in 2 it was 100%. James and Knight were unable to improve their results by making a series of temperature observations instead of one and they concluded that great accuracy would never be achieved ’
in estimating the TSD but systematic research could well reduce the size of errors. Shapiro reempha hasiz sized ed the imp import ortanc ancee of the the initia initiall postpostShapiro (1965) reemp mortem mor tem temper temperatu ature re platea plateau, u, which which he recogn recogniz ized ed as being being cau cause sed d by the temperature gradient between the central body core and the temperature of the immediate environment. He recognized that any estimation of the TSD must take into consideration the duration of this th is pl plate ateau au and that that New Newto ton n s law of cooling did not apply to a cooling human body. Shapiro also acknowledged the final prolonged cool co olin ing, g, wh whic ich h oc occu curr rred ed as th thee cada cadave verr reac reache hed d with within in 7 F (4 C) ’
of the en envir viron onmen menta tall tempe temperat rature ure and in re recog cogniz nizin ing g that that ru rules les of thumb sometimes gave a reasonable approximation of the TSD, he
30
Human Body Decomposition
nevertheless stated that any rule of thumb for the estimation of the TSD could only provide some degree of accuracy if used after the body had begun the rapid phase of cooling and before the prolonged final cooling phase. The difficulty was to determine when this intermediate period began and finished. (1970) stated that the cooling of a body was Joseph and Schickele (1970) stated a comp omplex lex proces cess as a tempe mperat rature gr grad adiient was was estab tabli lisshed between the interior parts of the body and the body surface and not until unt il coolin cooling g of the surfac surfacee occur occurred red di did d the the interi interior or tem temper peratu ature re begin to fall. As the distances from different parts of the body s interi rior or to the the su surf rfac acee va vari ried ed,, it foll follow owed ed that that th thee body body cool cooled ed in an uneven manner and cooling could not be expressed by just one cooling curve. For this reason they suggested abandoning the rectum as the point at which body temperature was measured and instead measuring temperature from within the center of the torso and that the term body cooling should be replaced with the term torso cooling. Rectal temperature only measured temperature at the area surroun ro undi ding ng the the plac placem emen entt of the the ther thermo mome mete terr and and di did d no nott re refl flec ectt temp temper erat atur uree drop drop fr from om any any othe otherr area area with within in th thee body body core core.. The The authors used the model of an infinite cylinder, which they defined as a cylinder, the ends of which were so far away from the point of meas me asur urem emen entt th that at an any y he heat at lo loss ss th thro roug ugh h the the ends ends wa wass negl neglig igib ible le from the point of view of the measurement being made, and which woul wo uld d thus thus si simu mula late te a huma human n body body cool coolin ing. g. They They di did d not not be beli liev evee that the use of any one formula for estimation of the TSD was useful. Instead, they computed a series of cooling curves which could be used for any given set of circumstances. ’
“
”
“
”
Marshall (1969) summarized the state of research by pointing out the inaccuracy of any formula for estimating body cooling based on a si sing ngle le expon exponen enti tial al term term.. Su Such ch a fo form rmul ula a assu assume med d that that body body cool coolin ing g was most rapid immediately after death and progressively fell to ambient temperature when, in reality, the temperature curve assumed a sigmoid mo id form form;; cool coolin ing g was was very very sl slow ow duri during ng th thee fi firs rstt 35 hour hours, s, wa wass faster and fell steadily and progressively for 9 9 12 hours and did not slow and become progressively flatter until it approached ambient temperature after 12 1215 hours. Such a cooling curve could only be represent se nted ed math mathem emat atic ical ally ly by a formu ormula la cont contai aini ning ng a sum sum ser serie iess of exponential terms; the more exponential terms in the series, the more
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
31
complex the formula and, consequently, the more unwieldy it would be to use in practice. Marshall s opinion was that sufficient accuracy could be achieved with the use of the formula given in his earlier arti1962): ): cle (Marshall (Marshall and Hoare, 1962 ’
2Zt
θ 5 B :e
1
C
:e2 pt
(2.3)
Z 2 p θ 5 the temperature excess of the rectum over the environment at time t and time and B , Z , and and p are constants for the corpse under observation. This formula was found to reproduce the initial slow rectal cooling and also took account of body size. Marshall stated that the formula had been found to be accurate for up to 18 hours after death and could also be used to construct the cooling curve of any corpse once its length, weight, and the environmental temperature were known. However, one of the constants, p constants, p,, in the formula could only be estimated and this pro3 4 hour by the time the PMI had reached 15 hours. duced an error of 6 /
The formula also assumed a constant environmental temperature, which rarely occurred, and any formula containing more than two exponential terms (constants) would be unwieldy to use. The rectal temperature at death was unknown and temperature measurements were rarely taken in the same conditions of cooling which prevailed before the body was found. These factors meant that no cooling formula would produce consistently accurate estimates of the TSD and could be no more than an investigative guide.
The 1970s to Early 1980s: Continuing the Search for an Accurate Formula Brown and Marshall (1974) (1974) continued continued to explore the use of a TSD formula mu la an and d st stat ated ed that that one one re reas ason on for for the the in init itia iall te temp mper erat atur uree pl plat ateau eau after death could be continuing metabolic processes but that in their view this was an insignificant factor. It was more likely due to the thermal conductivity and thermal capacity of the body producing a compl plex ex heat heat gr grad adie ient nt.. By a se seri ries es of math mathem emat atic ical al equa equattio ions ns th they ey explained why an equation with two exponential terms was adequate to explain the PMI, but they were still unable to improve the accuracy of estimation and no examples of its use were given. Simon onse sen n et al al.. (1 (197 977) 7) when A new new ap appr proa oach ch was was ad adop opte ted d by Sim they the y took took pos postmo tmorte rtem m temper temperatu ature re measur measureme ements nts in 20 cadave cadavers, rs, in whom whom the the time time of deat death h was was know known n wi with th ce cert rtai aint nty y to wi with thin in
32
Human Body Decomposition
15 minutes, from the brain, calf muscle, liver, axilla, and rectum, as well as the environmental temperature. A thermoelectric thermometer with six electrodes was used to record temperatures from the various organs continuously and simultaneously. No attempt was made to create create unifor uniform m enviro environme nmenta ntall condit condition ionss as the purpos purposee was to find an easy and practical method of estimating the TSD and a reliable site of temperature measurement. Measurements began between 30 minutes and 4.5 hours after death and lasted until the bodies had reached ambient temperature, which varied from 16.5 to 110 hours. Grap Gr aphs hs of th thee cool coolin ing g curv curves es fr from om each each lo loca cati tion on we were re prod produc uced ed.. The cooling curves from the rectum, axilla, and liver resembled an expo ex pone nent ntia iall curv curvee with with an init initia iall pl plat atea eau u an and d lo long ng fi fina nall cool coolin ing g period, but the brain and calf temperatures fell steadily without initial tial temp temper erat atur uree pl plat atea eaus us.. The The rang rangee of va vari riat atio ion n in te temp mper erat atur uree drop was smallest for the rectum and brain and the drop in temperature was steepest for the brain and calf muscle as these latter two sites mostcloser influenced thesurface. environmental temperature, due to their were position to the by body The other sites of measurement gave less reliable estimates of the TSD and none were reliable afte af terr 20 ho hour urss post post mort mortem em.. Beca Becaus usee br brai ain n te temp mper erat atur uree me meas asur ureement me ntss show showed ed the the leas leastt va vari riat atio ion n with within in the the fi firs rstt 20 hour hourss af afte terr death, they suggested that this might be the most useful organ to use for temperature measurement to estimate the TSD. However, if the brain brain temper temperatu ature re exceed exceeded ed 25 C, the the erro errorr in esti estima mati ting ng the the time time of de deat ath h woul would d be 2. 2.5 5 hour hourss but, but, if brai brain n te temp mper erat atur uree wa wass below 25 C, the error would increase to about 5 hours. Even though brain bra in temper temperatu ature re measu measurem rement entss showed showed the least least var variat iation ion,, they they were nevertheless so wide that this fact alone invalidated brain temperature measurement as a method of estimating TSD. The authors concluded that the variation in the temperature at the time of death in the cadavers was the crucial point in determining the TSD and tha th at as al alll the the metho ethods ds of de dete term rmin inat atiion pr pro opo possed so fa farr were ere dependent on chemical processes, which were temperaturedepend dep endent ent,, this this variat variation ion was an insolu insoluble ble proble problem. m. Finall Finally, y, they they conceded that attempts to estimate the time of death could never be more than an approximation. Olaisen Olaise n (19 (1979) 79) ca carried out a study of five cada adavers in whom the time the time of deat death h was was know known n and and in whic which h te temp mper erat atur uree meas measur ureements began between 2 and 10 hours after death. The temperature at
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
33
death was assumed to be 37 C and the rate of fall in temperature was uniform and rapid after an initial small plateau of about 2 hours which Olaisen attributed to the influence of brain size and of hair and clothing delaying cooling. The number of cases was small and the rate of fall varied from case to case, but Olaisen nevertheless concluded that measuring the rate of fall of brain temperature, after the initial temperat atur uree pl plat atea eau, u, prov provid ided ed a more more accu accura rate te meth method od of te temp mper erat atur uree measurement than that from the rectum because clothing, body stature, and weight, which caused variability in temperature measurements from the latter site, did not need to be taken into consideration in estimating the time of death ath. By contrast Al-Al Al-Alou ousi si an and d co coll lleag eague uess (2002a),, usin (2002a) using g micr microw owave ave ther thermo mogr grap aphy hy,, foun found d br brai ain n te temp mper erat atur uree approached that of the environment after about 15 hours and that the cooling curve was steep. Therefore the use of brain temperature was applicable only for the first few hours, but liver and rectal temperature fall could be used for a longer period, as temperature fall was slower in these sites. (1985a,b) devised devised a time-dependent z equation Green and Wright (1985a,b) or TDZE method to try to improve the accuracy of estimation. They ackn ac know owle ledg dged ed that that di diff ffer eren entt bodi bodies es cool cooled ed at di diff ffer eren entt rate rates, s, bu butt pointed out that the general manner was the same for all bodies, that is, an initial lag period followed by a steep linear fall and then a slow expo ex pone nent ntia iall fall fall in temp temper erat atur ure. e. The The re rela lati tive ve le leng ngth th of th thes esee peri period odss varied with body build and clothing, a naked emaciated body cooling much faster than a clothed obese body. The aim of the TDZE method was to calculate the PMI using temperature data alone, which was possible if the gradient of the cooling curve and the extent by which the rectal temperature had fallen were known. This depended on the measurement of two rectal temperatures about 1 hour apart, reference to a single standard curve and a simple mathematical equation. Green and Wright avoided the use of body parameters because previous research using these measurements had been done in standard conditions with nake na ked d bodi bodies es lyin lying g flat flat on thei theirr back backs, s, wher wherea eass in prac practi tice ce bodi bodies es cooled in different positions, different environments, and clothed and naked. Two measurements of rectal temperature were required as well as three temperature-related parameters; the temperature gradient (the fall fall in temp temper erat atur uree per per degr degree ee Cent Centig igra rade de per per ho hour ur), ), the the di diff ffer eren ence ce “
”
between rectal and ambient temperatures, and a measurement which they called the reduced theta or θ R, which was a measure of the “
”
“
”
34
Human Body Decomposition
extent by which the rectal temperature had fallen. The rectal temperature at death was arbitrarily fixed at 37.6 C. Green and Wright calculated θ R, which they defined as the fractional drop in rectal temperature, by the following equation: θ R 5
Rectal temperature at death death Measured Measured rectal temperature Rectal temperature at death death Measured Measured ambient temperature (2.4)
The authors devised a Refere Reference nce Curve with an explanation of how the TSD could be calculated from an equation. They analyzed 67 postmortem cases with an error rate of 30% at 3 hours after death and an increased error rate at 23 hours which they attributed to the slower rate of temperature fall as the bodies approached ambient temperature. Thes Th esee re resu sult ltss were were only only marg margin inal ally ly bett better er than than th thos osee of of Marshall (1962b) and of Henssge (1979, 1981), 1981), who had reported results of his meth me thod od in Germ German an.. It th ther eref efor oree of offe fere red d no im impr prov ovem emen entt on th thes esee “
”
methods. Nokes methods. (1985) emphasized that the sigmoid shape of the Nokes et al. (1985) emphasized cooling curve was the basis for all postmortem temperature investigations but all models to date had failed to take into account this plateau which they speculated could be partly caused by continuing glycogenolysis after death producing heat. However it was important to take into into accou account nt envi enviro ronm nmen enta tall temp temper eratu ature re,, surf surfac acee in insu sula lati tion on of the the body, and body size when assessing the early postmortem cooling of the body as these factors were responsible for maintaining the temperature plateau for up to 6 hours after death. ’
The Later 1980s: Henssge of s Nomogram, a Definitive Mathematical Expression TSD Claus Henssge (1988) published a seminal study in English, in which he outlined a method of determining the approximate time of death from a single rectal temperature by reading it from a computed nomogram. The study had first been reported in German (Henssg (Henssge, e, 1979, 1981). 1981 ). In the 1988 paper the results of 10 years experience of using the method were also reported. A nomogram is a diagram representing a relationship between three or more variables by means of a number of straight or curved scales, so arranged that the value of one variable can be found by a simple geometrical construction, for example, by ’
means of one or more straight lines drawn to intersect the scales at the appropriate values (Fig. (Fig. 2.2). 2.2).
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
35
Figure 2.2 An example of Henssge s nomogram (Henssge (Henssge,, 1988 1988). ). ’
Henssge s meth method od was was base based d on the the two two ex expo pone nent ntia iall equa equati tion onss develo dev eloped ped by Mar Marsh shal alll an and d Ho Hoar aree (1 (1962 962)), the the obse observ rvat atio ions ns th that at the the cooling of a body was a physical process and that the influence of biologica log icall fac factor torss such such as fever, fever, hyp hypoth otherm ermia, ia, postmo postmorte rtem m hea heatt produc produc-tion, and of physical factors such as body build and composition, were eith either er negl neglig igib ible le or easi easily ly re reco cogn gniz izab able le and and coul could d be acco accoun unte ted d for. for. Thee me Th meth thod od was was base based d on a si sing ngle le re rect ctal al te temp mper eratu ature re meas measur urem emen entt taken at least 8 cm within the rectum and the result was readily available at a scene of crime. ’
36
Human Body Decomposition
Henssg Hens sgee fixe fixed d the the re rect ctal al temp temper erat atur uree at ti time me of de deat ath h at 37.2 37.2 C, even if there was a fever. He defined the standard conditions of cooling as relating to a naked body lying extended on its back, on a thermally indifferent base in still air and in a closed room without any sources of heat radiation. He incorporated various constants into his calculations. Constant A described the postmortem temperature plateau and despite the rectal temperature at death being unknown and the duration of the plateau varying between 5 and 14 hours, Henssge found that there was a significant relationship between the duration of the plateau and the rate of temperature fall after the plateau. He calculated the value of constant A as being 1.25 in ambient temperatures up to 23 C and 1.11 in ambient temperatures above 23 C. Constant B related to the body weight to the power of 2 0.625 and Henssge calculated its value from a simple equation. The quantity of fatty tissue did not have any apparent influence on the value of B; neither did a calculation of body build based on the size factor which included surface area. Using Usin g thes thesee co cons nsta tant nts, s, the the valu valuee of of Q, name named d th thee standardized temperature was calculated. calculated. Q was considered a good measure of the progress of cooling as Henssge s opinion was that errors of computed time of death should not be plotted against the progress of death time but against the progress of cooling. Using a value of of Q as 1 at death and 0 when the rectal temperature had reached ambient temperature, Henssg Hen ssgee found found tha thatt in ambient ambient temper temperatu atures res bet betwee ween n 5 C and 22 C the error range for estimation of the TSD fell into three groups. At the 95% confidence interval, for values of of Q between 1 and 0.5 the error rate was 6 2.8 hours. For values of f Q be between 0.5 and 0.3 it was 6 3.2 hours and for values of Q between 0.3 and 0.2 the error rate was 6 4.5 hours. Large errors were found below a value of Q of 0.2 which corresponded to the prolonged cooling period before actual ambient temperature was reached and the estimation of the TSD with these values of Q was not possible. “
”
’
Studies were then carried out on bodies in different environments, clothed and naked, in still and moving air. It was concluded that the clothed body cooled like a naked body that was 1.4 times heavier and that even a slight but permanent air movement accelerated the cooling of a naked body significantly. Empiric corrective factors, based on the fact that a body with coverings would cool with a corrective factor grea gr eate terr than than 1, that that is, is, sl slow ower er,, whil whilee a body body co cool olin ing g in acce accele lera rate ted d “
”
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
37
conditions would cool with a corrective factor less than 1. Using the corrective factors, error statistics for the estimation of the TSD were calculated for the three levels of the value of Q, that is, 1 to 0.5, 0.5 to 0.3, and 0.3 to 0.2. Corrective factors were also calculated for bodies in still and flowing water. In actual casework, Henssge stated that his equations only gave an approximate value of death time. He devised a computer program (not detailed) and a series of nomograms to be used at the scene of a crime on a handheld Hewlett Packard 71B computer. The ambient temperature was calculated as the sum of the average daily temperatures from the time of death, as recorded by the nearest local weather station. The choi ch oice ce of corr correc ecti tive ve fact factor or for for bo body dy we weig ight htss re requ quir ired ed ju judgm dgmen entt and and expe ex peri rien ence ce and was was on only ly an appr approx oxim imat atio ion. n. He sugg sugges este ted d choo choosi sing ng both an upper limit and a lower limit for the value of the corrective fact fa ctor or.. Two Two line liness coul could d then then be draw drawn n on the the nomo nomogr gram am,, the the fi firs rstt betwee bet ween n rectal rectal tem temper peratu ature re and ambien ambientt tem temper peratu ature re on the vertical vertical sc scal ales es on eith either er si side de of th thee se semi mici circ rcul ular ar no nomo mogr gram am,, and and the the se seco cond nd from a central point to the value of the corrective factor on the periphery of the semicircle. The TSD was then given as the value where the lines crossed on the nomogram. If there was a range of ambient temperature and an error range of corrective factors, four lines would be draw dr awn n givi giving ng an appr approx oxim imat atee ra rang ngee of th thee ti time me of de deat ath h wh wher eree the the four lines crossed (Fig. (Fig. 2.2). 2.2). There were certain conditions where the method could not be used as the error would be too great and these were where there was strong radiation of heat or cold, where there had been hypothermia, where the place that the body had been found was not the same as the place of death, where there had been uncertain or severe changes in the cooling conditions between the time of death and the examination of the body and in unusual cooling conditions without any experience of a corr co rrec ecti tive ve fact factor or.. Spec Specia iall pr prob oble lems ms were were di disc scus usse sed; d; He Hens nssg sgee had had no experience in estimating the TSD found in temperatures above 23 C and so the nomograms did not take such temperatures into account. In bodi bo dies es dy dyin ing g with with a feve feverr th thee temp temper erat atur uree was was ofte often n abov abovee 37.2 37.2 C whic wh ich h was was th thee temp temper erat atur uree at de deat ath h assu assume med d by the the ca calc lcul ulat atio ions ns forr th fo thee nomo nomogr gram am me meth thod od.. Ther Theref efor oree a body body wi with th a te temp mper erat atur uree above this level would be considered to be still alive. For example, if the temperature was 40.2 C, the error in estimating the TSD would be
38
Human Body Decomposition
4 hours. As the temperature fell increasingly, the error rate fell into the acceptable range of 2 22.5 hours. In conditions where there had been movement of the body prior to assessment, the ambient temperature chan ch ange gess intr introd oduc uced ed an unac unacce cept ptab able le er erro rorr in into to th thee asse assess ssme ment nt of the TSD. Fina Fi nall lly, y, He Hens nssg sgee comp compar ared ed the the nomo nomogr gram am meth method od with with ot othe herr methods proposed to that date. A clear advantage was shown against using the rule of thumb method and the percentage method of Fiddes of Fiddes and Patten (1958). (1958). This latter method suffered from the disadvantage of no nott taki taking ng in into to acc accou ount nt th thee temp temper erat atur uree pl plat atea eau. u. The The nomo nomogr gram am method relied on one temperature measurement, whereas the methods aram am et al., 1955), Fidd Fiddes es an and d Pa Patt tten en (1 (195 958) 8),, of De Saram (De Sar hall ll an and d Ho Hoar are, e, 19 1962; 62; Mar Marsh shal all, l, 196 1962a 2a)) and Green Marsh Mar shal alll (Mars Marsha and Wright (1985a,b) were based on measuring the slope of the drop in temp temper erat atur uree by tw two o or more more temp temper erat atur uree meas measur urem emen ents ts at in inte terrvals. Henssge compared the nomogram method with all these methods and the results were not as accurate as the nomogram method. Green Gree n an and d Wri Wright ght s (19 (1985a 85a,b) ,b) meth method od fa fare red d wors worst. t. The There re were were two two reasons for these larger errors. Firstly, there was only a smal mall decrease in the rectal temperature in the space of 1 hour, especially if thee body th body was was clot clothe hed. d. Ther Theref efor oree on only ly a smal smalll mism mismea easu sure reme ment nt in either temperature would translate into a large error when the TSD was calc calcul ulat ated ed.. Se Seco cond ndly ly,, th thee meas measur ured ed rate rate of re rect ctal al te temp mper erat atur uree decrease was only valid for the cooling conditions during the period of temp temper erat atur uree meas measur urem emen ent. t. If the the cool coolin ing g cond condit itio ions ns had had been been changed before or during meas asu urement, larg arge errors could be expected. Commenting on the part of the body from which temperature tu re was was re reco cord rded ed,, He Hens nssg sgee stat stated ed that that the the radial radial di dist stan ance ce from from th thee body core and the position of the recording thermometer determined thee actu th actual al temp temper erat atur uree at any any on onee time time;; the the smal smalle lerr the the radi radius us,, the the st stee eepe perr the the temp temper erat atur uree fall fall.. Th This is wa wass the the re reas ason on for for a more more ex exac actt esti estima mati tion on of TSD TSD from from brai brain n re reco cord rdin ings gs bu butt it wa wass im impo port rtan antt to standardize the location. Henssge reported encouraging results in estimati ma ting ng th thee TSD TSD duri during ng th thee earl early y post postmo mort rtem em peri period od by co comb mbin inin ing g brai br ain n and and re rect ctal al temp temper erat atur ures es.. Up to 6.5 hour hourss br brai ain n te temp mper erat atur ures es gave ga ve a more more pr prec ecis isee esti estima mate te.. From From 6. 6.5 5 to 10 10.5 .5 hour hourss af afte terr de deat ath, h, comb co mbin inin ing g br brai ain n and and re rect ctal al temp temper erat atur ures es were were more more accu accura rate te and and ’
late laterr than than 10.5 10.5 ho hour urss post post mort mortem em,, re rect ctal al te temp mper erat atur ures es were were most most accurate (Henssge (Henssge et al., 1984 1984). ).
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
39
CONTEMPORARY RESEARCH FROM 1988 TO THE PRESENT TIME Since the publication by Claus Henssge of the nomogram method of estimating the TSD, most relevant research has been carried out by Claus Henssge and researchers in Germany and by Louay M. Al-A -Alo lous usii from fromwork Gl Glas asgo gow. w. Th This is contemporaneously. re rese sear arch ch wi will ll be de deta tail iled ed to toge geth ther er withAl other relevant carried out As a fo foll llow ow up to the pape paperr whic which h intr intro odu duce ced d th thee nomo nomogr gram am method (Henssge, (Henssge, 1988), 1988), Henssge Henssge et al. (1988) (1988) proposed that the error in estimating the TSD might be reduced by combining all other objective measurements of time-related measurements of the TSD with the nomogram method. These were rigor mortis, livor mortis, mechanical and electrical excitability of skeletal muscle, and chemical excitability of the iris. Instead of taking the mean of these measurements they suggested taking the upper and lower limits, which would then be combined in awere chart to estimate the TSD. However, no results using this approach presented. Henssge continued to experiment with dummies simulating human bodies and he formulated an extensive list of corrective factors to be used with bodies of different weight and with different types of clothing ing in di diff ffer eren entt ambi ambien entt temp temper erat atur uree cond condit itio ions ns (Hens Henssg sge, e, 199 1992 2). These corrective factors were to be used with the nomogram method, in order to attempt to reduce the error in estimation of the TSD. No examples of the use of these corrective factors in actual practice were presented. Alth aus an and d He Hens nssg sgee (1 (1999 999)) carri carried ed out out ex expe peri rime ment ntss wi with th vulc vulcaaAlthaus nized rubber-coated cylinders filled with gel, simulating human bodies in order to determine whether the nomogram method was applicable wher wh eree bodi bodies es ha had d cool cooled ed in a hi high gh envi enviro ronm nmen enta tall te temp mper erat atur uree and and then been moved to a lower temperature and vice versa. Two dummies were we re cons constr truc ucte ted d to si simu mula late te bodi bodies es of 58 58.7 .7 and and 24. 24.6 6 kg weig weight ht an and d heated to a body temperature of 37 C. Central core temperatures of the dummies and ambient temperatures were then measured, firstly, as they were cooled for a fixed period at 21 C ambient temperature and then moved rapidly to a cool room with an ambient temperature of 4 C for several hours. Secondly, the dummies, heated to 37 C, were st stor ored ed at 4 C for seve severa rall hou hours and and then then he heat ated ed to 21 C ambi ambien entt
40
Human Body Decomposition
temperature. In the first series of experiments where the dummies were cooled, a second plateau in the cooling phase was noted, shorter than the initial plateau, as the dummies cooled to a lower ambient temperature tu re.. Th Thee cool coolin ing g curv curvee coul could d be ex expl plai aine ned d math mathem emat atic ical ally ly with with a three-step procedure based on the two exponential term equation of Marshall and Hoare (1962) with an acceptable error range. However, in the the se seco cond nd se seri ries es of ex expe peri rime ment ntss wher wheree the the dumm dummie iess we were re heat heated ed from a lower temperature, rapidly more than 15 C, the two exponential term equation was unsuitable to describe the body cooling mathematically. This was attributed to heat flows in opposite directions; the cool co oled ed su supe perf rfic icia iall laye layers rs woul would d be re rehe heat ated ed from from ou outs tsid idee an and d at the the same sa me time time wo woul uld d be co cool oled ed by the the lo lowe werr te temp mper erat atur uree of th thee deep deeper er layer layers. s. Al Alth thau auss and and Hens Henssg sgee conc conclu lude ded d that that the the nomo nomogr gram am met metho hod d was inapplicable in bodies cooled firstly for a long period in a low ambient temperature and after that subjected to a period in a higher ambient temperature. Henssge and colleagues (2000a) published the results of the use of the nomogram in estimating the TSD in 72 cases over a 4-year period. In 61 of 67 cases where the method could be used, the estimated TSD fell within the 95% confidence limit. In 60 cases where the time could be verified exactly by other methods, the estimated TSD by the nomogram method matched exactly the TSD in 50 cases and partially in 10 cases. Problems arising in estimation of the TSD were discussed: one wass th wa thee su subj bjec ecti tive ve bi bias as in ch choi oice ce of corr correc ecti tive ve fact factor orss bu butt an even even greater error could be made in evaluating the true mean ambient temperature. These errors could only be reduced by the correct use of the method and the greater experience of the examiner. If ambient temperature varied before the body was discovered a mean value might not reflect the true value. The method could not be used where the body had ha d been been tran transp spor orte ted d from from an unkn unknow own n pl plac acee to the the pl plac acee where where it had been found. There had been no improvement in the error rates quoted by Henssge by Henssge (1988) (1988) for for these 72 cases since he had first proposed the nomogram method. Henssge and colleagues (1988) then returned to a question raised by Henssge ov oveer a decade previousl sly y in ord rdeer to ex exa ami n e t h e nomogram method and its use in combination with non-temperaturebased methods (Henssge (Henssge et al., 2000b). 2000b). The same series of cases as in thee pr th previ eviou ouss stud study y (Hen Henssg ssgee et al al., ., 200 2000a 0a)) were were us used ed to de dete term rmin inee
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
41
if any combination of methods could improve the estimation of the TSD. The non-te non-tempe mperatu raturere-bas based ed methods methods wer weree livor livor mor mortis, tis, rigor rigor mortis, mechanical and electrical excitability of skeletal muscle, and chemical excitability of the iris. The non-temperature-based methods cou co uld be used used with ith the the nomo nomogr gram am meth method od in 69 of the the 72 case casess. In 49 of the cases the use of non-temperature-based methods caused a re red duction in the limit imitss of the est estimate ated TSD compare ared with thee re th resu sult ltss of the the no nomo mogr gram am me meth thod od al alon onee and and al also so pr prov ovid ided ed a TSD TS D in 4 case casess wher wheree the the no nomo mogr gram am me meth tho od co coul uld d not not be us used ed.. In 14 cases the non-tem tempera erature-bas based met eth hods ods pro rov vided no improvement in the estimation of the TSD and in 6 cases they provided an improvement in only one of the limits of the estimated TSD. In 7 ca case sess the the no nonn-te temp mpera eratu ture re-b -bas ased ed meth method odss al alon onee prove proved d to be more accurate tha than the nom nomogram metho thod. In 49 cas asees where thee TSD th TSD coul could d be accu accura rate tely ly es esti tima mate ted d from from th thee poli police ce reco record rdss al alon one, e, the the us usee of nonnon-te temp mper erat atur uree-ba base sed d meth method odss pro rovi vid ded an improvement in inatur 44 thethod cases. Henssge and foun fo und d that that th the e estimation nonnon-te temp mper erat ureeofmeth me od wh whic ich h gave gave the thcolleagues e grea greate test st accuracy was the electrical excitability of facial muscles by impulses fr from om elect electrod rodes es in inse sert rted ed in into to the the eyeb eyebro row. w. Impr Improv oveme ement ntss ho howe wever ver were modest. Althaus and Henssge (1999) had found that while the estimation of the TSD in the case of a sudden decrease of ambient temperature during the cooling phase was possible, the estimation of TSD was not possible if a sudden increase in temperature occurred. Bisegna and colleagues (2008), (2008), experimenting on dummies, calculated a complex four-step mathematical procedure to prove that this was possible within reasonable error limits. A different approach to estimation of TSD was taken by Al-Alousi and his colleagues who published a series of papers from 1986 to 2001 researching the estimation of the TSD using microwave probes to measure su re body body orga organ n temp temper erat atur ures es.. A pr prel elim imin inar ary y stud study y on th thee use use of microwave thermography in estimating the TSD had been carried out on 100 human fatalities (Al-Alousi (Al-Alousi and Anderson, 1986 1986). ). In this initial st stud udy y it was conc conclu lude ded d that that micr microw owav avee ther thermo mogr grap aphy hy was usef useful ul in esti estima mati ting ng th thee TSD, TSD, givi giving ng a si simi mila larr de degr gree ee of ac accu cura racy cy as ot othe herr temperature-measuring devices. Cooling curves were produced for the brain, liver, and rectum which required triple exponential functions for
42
Human Body Decomposition
their expression. The error rates were acceptable but the method was insufficiently precise to be used in legal evidence. In a subsequent paper, the system and method of use were lous usii et al al., ., 19 1994 94). ). Tw Two o fl flex exib ible le micr microw owave ave prob probes es expl ex plai aine ned d (Al-A Al-Alo were placed on the skin of the temporal region to measure brain temperature and on the skin of the right hypochondrium to measure liver temper tem peratu ature. re. Two rigid rigid the thermo rmocou couple pless measur measured ed enviro environme nmenta ntall and rectal temperatures. The thermocouples were attached to a microwave radi ra dior orec ecei eive verr and and a data data lo logg gger er whic which h me meas asur ured ed and and comp comput uted ed the the temperature data. Thermal radiation consists of centimetric or microwave, millimetric and infrared waves. Microwave penetration was the less penetrative part of thermal radiation from body tissues and the micr mi crow owav avee pr prob obee was was desi design gned ed to magn magnif ify y and and re reco cord rd this this wa wave ve-leng length th.. The The grea greate terr the the wa wate terr cont conten entt of ti tiss ssue ue,, the the le less ss wa wass th thee microwave penetration. Penetration through fat and bone was greater than th an thro throug ugh h musc muscle le and and skin skin.. The The pr prob obes es pl plac aced ed on th thee skin skin were were desi de sign gned ed to over overco come me the the re refr frac acti tion on back back in into to the the body body of micr microowaves when they reached the skin/air boundary. The temperature resolution of the system was 0.1 C and the response time was 2 seco second nds. s. The The ac accu cura racy cy of temp temper erat atur uree me meas asur urem emen entt of al alll the the probes was 6 0.6 C. The penetration depth of the microwave probes was 1.6 cm in high water content tissue and 10 cm in low water content tissue, which was sufficient to measure the interior temperatures of the liver and brain. The system was assessed by measuring brain and liver temperatures in 14 living healthy subjects and 5 cadavers. In addition, the temperaturess of the ture the tiss tissu ues of the ca cada dave vers rs in the the righ rightt hy hypo poch chon ondr driu ium m were measured at different depths (skin, subcutaneous tissues, muscle, and liver) by inserted thermocouples. Calibration of the probes using glycerol and comparison with a mercury thermometer and thermocoupl plee was was succ succes essf sful ul.. Temp Temper erat atur uree re read adin ings gs gi give ven n by th thee micr microw owav avee probes depended on the temperature of the tissues and the degree of micr mi crow owav avee at atte tenu nuat atio ion, n, whic which h in turn turn was was de dete term rmin ined ed by the the type type of tiss tissue ue trave travers rsed ed.. Ther Theree was no di diff ffer eren ence ce in at atte tenu nuat atio ion n betw betwee een n living and dead tissues and the authors concluded that the microwave ther th ermog mogra raph phy y syst system em offe offere red d an ac accu cura rate te techn techniq ique ue to me meas asur uree the the temp temper erat atur uree of in inte tern rnal al body body or orga gans ns by a nonnon-in inva vasi sive ve and and more more ethically acceptable method.
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
43
Continuing this line of research, Al-Alousi and colleagues (2002a) studied 117 cadavers. These were taken as soon as possible after death to a mo morg rgue ue wh wher eree micr microw owav avee re reco cord rdin ings gs of br brai ain, n, li live ver, r, and and re rect ctal al temp temper erat atur ures es were were unde undert rtak aken en,, as the the cada cadaver verss cool cooled ed for for up to 60 hour ho urss af afte terr deat death. h. Th Thee re rect ctal al and and envi enviro ronm nmen enta tall te temp mper erat atur ures es we were re record rec orded ed wit with h electr electrica icall thermo thermocou couple ples, s, the liver liver and brain brain temper temperaature tu ress with with mi micr crow owav avee prob probes es.. Data Data were were re reco cord rded ed on a data data lo logge ggerr and then analyzed with the use of a reference graph, a reference chart ruler, or computer software (type not stated). The temperatures of the brain, liver, and rectum at the time of death were determined by measuring the temperatures for the first 3 hours, fitting them to a curve, and an d then then ex extr trap apol olat atin ing g the the temp temper erat atur ures es from from th thee di diff ffer eren entt orga organs ns backwards over the PMI. A temperature difference ratio (R (R) was calculated from the temperature at each of these specific body sites at a given time, the temperature of the environment at that same time and the temperature of these specific body sites at the moment of death. Curve-fitting procedures calculated values at atur uree di diff ffer eren ence cess vers versus us were PMI PMIthen usin using g a trip triple leforexpo exthe pone nent ntia iall ofequa eqtemperuati tion on.. Al-A Al -Alo lous usii and and coll collea eagu gues es conc conclu lude ded d that that post postmo mort rtem em cool coolin ing g was a complicated phenomenon, unable to be described by a simple model, and an d that that the the best best de desc scri ript ptio ion n was was by trip triple le expo expone nent ntia iall equa equati tion ons. s. Ther Th eree were were di diff ffer eren ence cess betw betwee een n body body site sitess with with re resp spec ectt to the the rate rate of cooling and between covered and naked bodies. The rectum as a si site te for for temp temper erat atur uree re reco cord rdin ing g was was fl flaw awed ed beca becaus usee it had had a hi high gher er temperature than other body sites and standard placement for temperat atur uree re reco cord rdin ing g was di diff ffic icul ult. t. Brai Brain n and li live verr re reco cord rdin ings gs were were more more representative of body cooling. In a second part of the study using the same 117 bodies, the authors modified the computer software for use in the field (Al-Alousi ( Al-Alousi et al., 2001). 2001 ). The computer software was not specified but the authors stated that it could be obtained by a special arr arrangement wi witth them. Temperature difference ratios (R (R) were calculated for selected values of rectal rec tal and env enviro ironme nmenta ntall temper temperatu atures res fro from m 1 C to 36 C. Thirty-six Thirty-six curves were compiled to represent the relationship between the rectal and an d th thee en envi viro ronm nmen enta tall temp temper erat atur ures es ove overr the the te temp mper erat atur uree rang rangee of 136 C. The The curv curves es were were then then pl plot otte ted d to ma matc tch h the the aver averag agee re rect ctal al cooling curves for both covered and naked bodies, thus producing a reference chart ruler. Further calculation incorporated an error range of 6 1 stand andard ard devi via ation. To use the sys ysttem in the field, the
44
Human Body Decomposition
temp temper erat atur uree of a bo body dy si site te and and the the temp temper erat atur uree of th thee envi enviro ronm nmen entt were recorded, which enabled the value of R to be read from the reference chart ruler. Using this value, the PMI, 6 1 standard deviation, could then be read from curves for either naked or clothed bodies. Al-Alousi and colleagues continued their research on the same 117 Al-A -Alo lous usii et al al., ., 200 2002a 2a,b ,b;; bodies bodi es an and d publ publis ishe hed d thre threee more more stud studie iess (Al Al-Alousi, 2002). 2002). Two studies discussed the factors influencing the precision of estimating the PMI using the triple exponential formula. In thee firs th firstt st stud udy, y, th thee bodi bodies es were were se sepa para rate ted d in into to nake naked d and and co cove vere red d groups and fat and thin groups (Al-Alousi (Al-Alousi et al., 2002a). 2002a). The separation into fat or thin depended on a formula based on surface area and weight which gave a cooling size factor. Using this cooling size factor they found that thin bodies cooled faster than fat ones, but only in 56% of cases. The correlation between the rate of body cooling and body build was strong and age had no influence on the rate of cooling. In most most case casess co cove vere red d bodi bodies es cool cooled ed more more sl slow owly ly than than na nake ked d ones ones.. Brain cooling was found to approach environmental temperature faster than the liver and rectum. The liver was the slowest to cool in the covered group and the rectal cooling was intermediate. They concluded that there was no single body parameter which could be used to quantifiably predict the cooling behavior in all cases, but covering the torso significantly slowed the rate of body cooling. In the second study, the effect of the body temperature at the moment of death on the postmortem cooling rate was studied in order to incorporate this in the triple exponential formula (Al-Alousi (Al-Alousi et al., 2002b 2002b). ). Temperature measurements for the brain, liver, and rectum were taken every 5 5 10 minutes after the body entered the morgue. The temperatures of these organs at the the mome moment nt of de deat ath h were were dete determ rmin ined ed by fi fitt ttin ing g the the te temp mper erat atur uree data for the first 3 hours to single exponential equations to form fitted curves which were then extrapolated backwards by regression analysis. The mean time between death and the start of monitoring was 32 6 15 minutes, 74 (63.3%) cases were between 15 and 45 minutes and only 2 cases started after a period of 3 hours after death. Environmental temper tem peratu atures res ranged ranged betwee between n 8.38 8.38 C and and 22.7 22.76 6 C (m (mea ean n valu valuee was was 15.2 6 3.2 C). A factor was calculated and incorporated in the triple expo ex pone nent ntia iall form formul ula a to allow allow dete determ rmin inat atio ion n of th thee te temp mper erat atur uree at death from the temperature of any organ but temperatures from the “
”
“
”
“
”
brain and liver had to be taken with the microwave probe. Al-Alousi and colleagues determined that if the temperature of the environment
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
45
was kept was kept cons consta tant nt,, the the body body temp temper erat atur uree at the the mome moment nt of de deat ath h woul wo uld d dete determ rmin inee the the temp temper erat atur uree grad gradie ient nt betw betwee een n the the body body and and it itss surrounding medium. Thus a higher temperature at death would mean a steeper rate of cooling and a greater amount of heat to be lost to brin br ing g the the bo body dy to equi equili libr briu ium m wi with th the the envi enviro ronm nmen enta tall te temp mper erat atur ure. e. The various conditions causing temperature variations at the moment of death were discussed. From their calculations they theorized that if thee same th same body body wa wass allo allowe wed d to co cool ol twic twice, e, star starti ting ng from from a di diff ffer eren entt body temperature, the time required to reach equilibrium would be the same although the rate of cooling would not be constant at all stages. They conceded, however, that this would be impossible to verify. (2002) studied studied the shapes of the brain, liver, and Finally, Al-Alousi (2002) Finally, Al-Alousi rectal cooling curves and compared them in naked and covered bodies. In the covered group the liver had the most rapid and the brain the slowest cooling rates at 6 hours post mortem, while at 12 hours the brain was slowest and the rectum the most rapid. In the naked group, thee re th rect ctum um cool cooled ed sl slow owes estt and and the the brai brain n the the most most ra rapi pid d at 6 hour hourss post mortem, but at 12 hours the brain had the slowest rate of cooling. In the naked group, in all sites, cooling was faster than in the covered group. On average, thin bodies cooled faster than obese bodies. The initial temperature plateau was found in only 22% of bodies, the rectum tu m in th thee na nake ked d grou group p havi having ng the the hi high ghes estt in inci cide denc ncee of a pl plat atea eau u (27%) compared with 7% for the brain and liver curves. The rectum was the the only only si site te wh wher eree a si sign gnif ific ican antt te temp mper erat atur uree pl plat atea eau u occu occurr rred ed,, irrespective of body weight and the presence or absence of clothing. Al-Alousi concluded that the cooling curve was of a compound type and that the rate of cooling was not uniform throughout the whole PMI. The wo The work rk of Al-A Al-Allou ousi si and and hi hiss coll collea eagu gues es demo emonstr nstrat ated ed that that because of the non-uniformity of body cooling and the fact that the body temperature at the moment of death could not be known, the estimation of the TSD could never be exact. Nokes and colleagues (1992) compared eight methods of estimating the TSD on eight cadavers in which the time of death was known. The methods were two rules of thumb using rectal temperature measurement: TSD 5
ðTemperature at death death Temperature Temperature when foundÞ foundÞ 1:5
(2.5)
46
Human Body Decomposition
TSD 5 ðTemperature at death death Temperature Temperature when foundÞ foundÞ 1 3 (2.6) 1955), ), Also compared were De Saram s method (De (De Saram et al., 1955 the method of Fiddes and Patten (1958), (1958), Marshall and Hoare s (1962) method, Gre Green and Wright s (1 (195 958a 8a,b ,b)) meth method od,, Al-A Al-Alo lous usii an and d ’
’
’
’
Anderson methsge od (1988) (Al-A Al-Alo usitwo i et cadavers al., al ., 20 2001 01) ), aTSD nd twas he nunderestiomogram method of s Henssge of Hens (1988). .lous In the mate ma ted d by all all me meth thod odss and and very very larg largee er erro rors rs we were re prod produc uced ed by AlAlAlou Al ousi si and and Ande Anders rson on s and and Fiddes and and Patt atten s methods on one cada ca dave verr an and d on aver averag agee the the se seco cond nd rule rule of thum thumb, b, th that at of of Knight (1991),, produced the best results in all cadavers. Nokes and colleagues (1991) stressed that this was a small study, but it appeared that the increasingly ing ly comple complex x mathem mathemati atical cal met method hodss offere offered d no increa increased sed accura accuracy cy over simple rules of thumb. ’
’
OTHER STUDIES ATTEMPTING TO IMPROVE THE ESTIMATION OF TSD Other approaches seeking to improve the TSD estimation are briefly explored in this final section. (1983) presented a cooling model based on four rectal Nokes et al. (1983) presented temperatures, an estimated body core temperature at time of death of 37 C, and the environmental temperature, postulating that if the environmental temperature varied by no more than 2 C it had no effect on the rectal temperature curve due to the poor conducting properties of the body. They used computer software and based the calculations on the equation of of Brown and Marshall (1974) to estimate the PMI on eight bodies. The error rate in estimating the TSD was within 51 minutes of the actual time of death for the case histories presented. (1988) presented presented a method of estimating the Morgan and colleagues (1988) PMII with PM withou outt pr prio iorr know knowle ledg dgee of the the body body te temp mper erat atur uree at de deat ath, h, thee si th size ze of the the body body,, or the the orig origin inal al envi enviro ronm nmen enta tall te temp mper erat atur ure. e. Temper Tem peratu ature re mea measur sureme ements nts were were taken taken simult simultane aneous ously ly by means means of thermocouples placed in both middle ears and on the skin of the forearm and thigh. The measurements were then incorporated in an algorithm. Observations were taken only from three corpses and although the authors stated that it wa wass intended to further develop the
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
47
technique, no follow-up studies on the use of the method are known to have been carried out. up (19 (1993) 93) pres presen ente ted d a comp comput uter er soft softwa ware re prog progra ram m Niels Lynner Lynnerup usin us ing g algo algori rith thms ms based based on the the cool coolin ing g form formul ula a devi devise sed d by Mars Marsha hall ll and Hoare. Lynnerup claimed that the program could overcome the uncertainties related to ambient temperature and rectal temperature at death and give a better estimation of the TSD. However the study did not present the practical use of the program on corpses and no followup studies on this method seem to have been presented. A di diff ffer eren entt ap appr proa oach ch was was adop adopte ted d by Bac Bacci cino no and col collea league guess (1996),, who (1996) who stud studie ied d a seri series es of 138 138 ca cada dave vers rs su subd bdiv ivid ided ed into into fo four ur groups according to the ambient temperature between 0 C and 23 C (0 (05 C, 610 C, 11 1115 C, 16 1623 C) C),, at whic which h they they were were stor stored ed while cooling. The mean of the tympanic membrane temperature in both outer ears was measured on each cadaver and the rectal temperature, 8 cm inside the rectum, was also recorded on each cadaver. The authors suggested that measuring the temperature from the tympanic membrane provided a more accurate representation of the fall in temperature of the inner body core than did the rectum, as it was recorded from a site where there was an absence of clothing or of intervening organs. The cooling curve was also more linear without the initial plateau. Their aim was to try to improve the accuracy of TSD TS D esti estima mati tion on,, but but to keep keep the the me meth thod od si simp mple le.. To do this this they they proposed combining biochemical methods and temperature measurement me nt meth method ods. s. They They th theo eori rize zed d that that en envi viro ronm nmen enta tall te temp mper erat atur uree affect aff ected ed body body tempe temperat rature ure measu measurem rement ent met metho hods, ds, resul resultin ting g in an underest undere stima imati tion on of the TSD TSD,, while while heat-g heat-gene enerat ratin ing g cytoly cytolytic tic proprocesses were speeding up the process of decomposition and overestimating the TSD. By combining the methods they hoped to improve the estimation of TSD. Using multivariate regression analysis, they compared two rule-ofthumb methods of measuring TSD with Henssge s nomogram and also with the estimation of plasma electrolytes, CSF, and vitreous humor potassium levels. Baccino and colleagues in comparing methods using outer out er ear tem temper peratu atures res wit with h the sam samee method methodss usi using ng rectal rectal temper temperaatures, found that outer ear temperature measurements correlated better ’
with TS with TSD D than than di did d re rect ctal al temp temper erat atur uree me meas asur urem emen ents ts.. One One rule rule of thumb method, Eq. (2.6) (Knight, 1991), 1991), proved to be superior to the
48
Human Body Decomposition
(2.5) (Glaister, 1942; Rentoul and Smith, 1973; Clark et al., other, Eq. (2.5) other, Eq. 1997), 1997 ), as well as Henssge s nomogram at estimating the TSD. The use of vitreous humor potassium measurement was the next best method of me meas asur urin ing g TSD, TSD, but but pl plas asma ma and and CSF CSF el elec ectr trol olyt ytee meas measur urem emen ents ts were we re inacc inaccur urat atee meth method ods. s. A comb combin inat atio ion n of me meth thod odss pr prov oved ed bett better er than any single method. ’
In di disc scus ussi sing ng the the re resu sult lts, s, Ba Bacc ccin ino o and and coll collea eagu gues es ca caut utio ione ned d against using the method of measuring outer ear temperature when there was blood in the ear canal or when the body had been lying on one side with one ear in close contact with a surface. It should also not be used where the PMI was greater than 15 hours as the error of estimation increased greatly after this time. They developed a fo form rmul ula a wh whic ich h enco encomp mpas asse sed d all all te temp mper erat atur uree su subg bgro roup ups. s. Rutty (1997) criticized the use of the outer ear canal as a source of temperature measurement, implying that the method would be difficult to standardize. He stated that foreign material, blood, the shape of the canal, and position of the temperature probe, as well as the uncertain tai n core core tempe tempera ratur turee at de death ath,, could could make make temper temperatu ature re measu measurerement me nt inac inaccu cura rate te.. On Onee of the the au auth thor ors, s, Bac Bacci cino no (19 (1997) 97),, in reply defended the method, stating that all placements of the thermometer in the outer ear canal were standardized. Yet anot Yet anothe herr algo algori rith thm m base based d on an aver averag ages es-b -bas ased ed meth method od of short-term PMI estimation using a graphing calculator and spreadsheet program, was presented by Eric Nelson Eric Nelson (2000). (2000). Nelson used data pro(1992) to to calculate an absolute relative vided by Nokes by Nokes and colleagues (1992) error (ARE) by dividing the absolute value of the difference between the actual PMI and the estimated PMI by the actual PMI. By using this th is ARE, ARE, he comp compar ared ed hi hiss meth method od with with ei eigh ghtt othe otherr meth method odss an and d found his method to be more accurate. The ARE for Nelson s method wass 0.05 wa 0.05 (SD (SD of 0.05 0.05)) ov over er 24 hour hourss whil whilee the the next next most most accu accura rate te methods were a rule-of-thumb method (Eq. (Eq. (2.4)) (2.4)) with an ARE of 0.16 (SD of 0.7) over 10 hours and Marshall and Hoare s (1962) method with an ARE of 0.20 (SD of 0.11) over 10 hours. No follow-up studies using this method appear to have been published. ’
’
Finally, Mall and colleagues in 2004 studied 35 bodies after admissi sion on to a medi medico cole lega gall in inst stit itut utee wh wher eree they they were were kept kept at a cons consta tant nt
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
49
environme enviro nmenta ntall temper temperatu ature re for 12 1236 hours (Mall (Mall et al., 2005). 2005). The times of death were known exactly but the environmental temperatures at the time of death were not known. The aim of the study was to develop a mathematical formula incorporating an additional factor to account for the unknown temperature at time of death. Cooling curves were we re cons constr truc ucte ted d whic which h gave gave an ac acce cept ptab able le er erro rorr ra rate te of 6 4 hour hourss over 12 1236 hours. This error rate compared favorably with the error rate ra te qu quot oted ed by Henss Henssge ge (198 (1988) 8) of 6 2.5 hours up to 11 hours post mortem with known environmental conditions.
CONCLUSION The investigation of estimating the TSD by temperature-based methods has produced an extensive literature in the past 200 years. Several (1839) recog recogimportant milestones are noteworthy since Dr John Davy (1839) ni nize zed d that that meas measur urem emen entt of bo body dy temp temper erat atur uree wh whil ilee it cool cooled ed af afte terr death could lead to an estimation of the PMI. Taylor PMI. Taylor and Wilks (1863) noted the initial temperature plateau after death in the cooling curve and Rainy (1868) was the first to note that cooling of a body did not follow Newton s law of cooling. ’
In the 20th century, Shapiro century, Shapiro (1954) recognized (1954) recognized the sigmoid shape of 1962b)) in the the cooling curve, while Marshall and Hoare (Marshall, ( Marshall, 1962b 1960s confirmed the complex nature of body cooling and the necessity of defining it mathematically with a formula if any accuracy in determining the time of death in the early stages was to be attained. In the 1980s Claus Henssge (Henssge, (Henssge, 1979; Nokes et al., 1985 1985)) developed his nomogram method and subsequently refined it. Although this seems to be the most accurate method at the present time there is still an error rate of 2 24 hours during the first 12 hours or so after death.
REFERENCES Al-Alousi, L.M., 2002. A study of the shape of the post-mortem cooling curve in 117 forensic cases.. Foren cases Forensic sic Sci. Int. 125, 237 237244. Al-Alousi, L.M., Anderson, R.A., 1986. Microwave thermography in forensic medicine. Police Surg. 30, 30 3042. Al-Alo AlAlousi usi,, L.M L.M., ., And Anders erson, on, R.A R.A., ., Lan Land, d, D.V D.V., ., 199 1994. 4. A non non-in -invas vasive ive met method hod of pos post-m t-mort ortem em temperature measurements using a microwave probe. Forensic Sci. Int. 64, 35 35 46.
50
Human Body Decomposition
Al-Alousi, L.M., Anderson, R.A., Worster, D.M., Land, D.V., 2001. Multiple probe thermography for estimating the post-mortem interval: part II. Practical versions of the triple exponential formulae (TEF) for estimating the time of death in the field. J. Forensic Sci. 46 (2), 323 327. Al-Alousi, L.M., Anders Al-Alousi, Anderson, on, R.A., Worster, D.M., Land, D.V., 2002a 2002a.. Fact Factors ors influ influenci encing ng the precision of estimating the post-mortem interval using the triple-exponential formulae (TEF) part I. A study of the effect of body variables and covering of the torso on the post-mortem brain, liver and rectal cooling rates in 117 forensic cases. Forensic Sci. Int. 125, 223 223230. Al-Alousi, L.M., Anderson, R.A., Worster, D.M., Land, D.V., 2002b. Factors influencing the precision of estimating the post-mortem interval using the triple-exponential formulae (TEF) part II. A study of the effect of body temperature at the moment of death on the post-mortem brain, liver and rectal cooling in 117 forensic cases. Forensic Sci. Int. 125, 231 231236. Althaus, L., Henssge, C., 1999. Rectal temperature time of death nomogram: sudden change of ambient temperature. Forensic Sci. Int. 99, 171 171 178. Baccin Bacc ino, o, E. E.,, et al al., ., 19 1996 96.. Ou Oute terr ea earr te temp mper erat atur uree an and d ti time me of de deat ath. h. Fo Fore rens nsic ic Sc Sci. i. Int Int.. 83 83,, 133 133 146. Baccino, Bacc ino, E., 1997 1997.. Lett Letter er to the edito editor. r. Forensic Sci. Int. 83, 173. Bisegna, P., Henssge, C., Althaus, L., Giusti, G., 2008. Estimation of the time since death: sudden increase of ambient temperature. Forensic Sci. Int. 176, 196 196 199. Brown, A., Marshall, T.K., 1974. Body temperature as a means of estimating the time of death. Forensic Sci. 4, 125 125133. Burman, J.W., 1880. On the rate of cooling of the human body after death. Edinburgh. Med. J. 25, 993 9931003. Cairns, F.J., Forbes, J.A., 1952. Time since death. Med. J. Aust. 2 (17), 585 585586. Clark, M.A., Worrell, M.B., Pless, J.E., 1997. Post mortem changes in soft tissues. In: Haglund, W.D., W.D ., Sor Sorg, g, M.H M.H.. (Ed (Eds.) s.),, For Forens ensic ic Tap Taphon honom omy: y: The Pos Post-m t-mort ortem em Fat Fatee of Hum Human an Rem Remain ains. s. CRC Press, Boca Raton, p. 152. Ch. 9. Davy, J., 183 Davy, 1839. 9. Obs Observ ervati ations ons on the tem temper peratu ature re of the hum human an bod body y after after dea death. th. Res Resea earch rches es Physiologi Physi ological cal and Anato Anatomica micall Vol. I. Smith Smith,, Elder and Co., London, pp. 228 228 248. De Saram, G.S.W., Webster, G., Kathirgamatamby, N., 1955. Post-mortem temperature and the time of death. J. Crim. Law Criminol. Police Sci. 46 (4), 562 562 577. Fiddes, F.S., Patten, T.D., 1958. A percentage method for representing the fall in body temperature after death. J. Forensic Med. 5 (1), 2 215. Glaister, Glaist er, J. J.,, 194 1942. 2. Med Medica icall Jur Jurisp isprude rudence nce and Tox Toxico icolog logy, y, sev sevent enth h ed. E & S Liv Living ingsto stone ne,, Edinburgh, p. 119. Ch. 4. Green, M.A., Wright, J.C., 1985a. Post-mortem interval estimation from body temperature data only. Forensic Sci. Int. 28, 35 3546. Green, M.A., Wright, J.C., 1985b. The theoretical aspects of the Time Dependent Z Equation as a means of post-mortem interval estimation using body temperature data only. Forensic Sci. Int. 28, 53 5362. Henssge, C., 1979. Precision of estimating the time of death by mathematical expression of rectal body cooling. Z. Rechtsmed. 83, 49 4967. Henssge, C., 1981. Estimation of death time by computing the rectal body cooling under various cooling conditions. Z. Rechtsmed. 87, 147 147178. Henssge, C., 1988. Death time estimation in case work I. The rectal temperature time of death nomogram. Forensic Sci. Int. 38, 209 209 236. Henssge, C., 1992. Rectal temperature time of death nomogram: dependence of corrective factors on the body weight under stronger thermic insulation conditions. Forensic Sci. Int. 54, 51 51 56.
Algor Mortis and Temperatur Temperature-Base e-Based d Methods Methods of Estimating Estimating the Time Since Death
51
Henssge, C., et al., 2000a. Experiences with a compound method for estimating the time since death: I. Rectal temperature nomogram for time since death. Int. J. Legal Med. 113, 303 319. Henssge, C., et al., 2000b. Experiences with a compound method for estimating the time since death: II. Integration of non-temperature-based methods. Int. J. Legal Med. 113, 320 331. Henssge, C., Frekers, R., Reinhardt, S., Beckman, E.-R., 1984. Determination of time of death on the basis of simultaneous measurement of brain and rectal temperatures. Z. Rechtsmed. 93, 123 123 133. Henssg Hens sge, e, C. C.,, Made Madea, a, B. B.,, Ga Gall llen enke kemp mper er,, E. E.,, 19 1988 88.. De Deat ath h ti time me es esti tima mati tion on in ca case se wo work rk II II.. Integratio Integ ration n of diffe different rent methods. Forensic Sci. Int. 39, 77 77 87. Joseph, A.E.A., Schickele, E., 1970. A general method for assessing factors controlling post mortem cooling. J. Forensic Sci. 15, 364 364 391. Knight, B., 1991 Knight, 1991.. PostPost-morte mortem m decom decomposit position, ion, Simps Simpson on s Forensic Medicine, 10th ed. Edward Arnold, London, p. 39. Ch. 3. ’
Knight, B.H., James, W.R.L., 1965. Errors in estimating the time of death. Med. Sci. Law 5, 111 116. Lyle,, H. Lyle H.P. P.,, Cl Clev evel elan and, d, F. F.P. P.,, 19 1956 56.. De Dete term rmin inat atio ion n of th thee ti time me of de deat ath h by bo body dy he heat at lo loss ss.. J. Forensic Sci. 1, 11 1124. Lynnerup, N., 1993. A computer program for the estimation of time of death. J. Forensic Sci. 38 (4), 816 816820. Mall, G., et al., 2005. Temperature based death time estimation with only partially known environmental conditions. Int. J. Legal Med. 119, 185 185 194. Marshall, T.K., 1962a. Estimating the time of death: the use of the cooling formula in the study of post-mortem body cooling. J. Forensic Sci. 7 (2), 189 189210. Marshall, T.K., 1962b. Estimating the time of death: the use of body temperature in estimating the time of death. J. Forensic Sci. 7 (2), 211 211 221. Marshall, T.K., 1969. The use of body temperature in estimating the time of death and its limitations. Med. Sci. Law 9, 178 178182. Marshall, T.K., Hoare, F.E., 1962. Estimating the time of death: the rectal cooling after death and its mathe mathemati matical cal expression. expression. J. Foren Forensic sic Sci. 7 (1), 56 5681. Morgan, C., Nokes, L.D.M., Williams, J.H., Knight, B., 1988. Estimation of the post-mortem period by multiple site temperature measurements and the use of a new algorithm. Forensic Sci. Int. 39, 89 8995. Nelson, E., 2000. Forensic Estimation short term a new algorithm. Sci.ofInt. 109, 31post 31 38. mortem interval utilising core body temperature: Nokes, L.D.M., Brown, A., Knight, B., 1983. A self-contained method for determining time since death from temperature measurements. Med. Sci. Law 23 (3), 166 166 170. Nokes, L.D.M., Hicks, B., Knight, B.H., 1985. The post- mortem temperature plateau fact or fiction? Med. Sci. Law 25 (4), 263 263 264. Nokes, L.D.M., Flint, T., Williams, J.H., Knight, B.H., 1992. The application of eight reported temperature based based alg algori orithm thmss to cal calcul culate ate the post post-mo -morte rtem m int interv erval. al. For Forens ensic ic Sci Sci.. Int Int.. 54, 109 109 125. Olaisen, B., 1979. Post-mortem decrease in brain temperature. Z. Rechtsmed. 83 (3), 253 253257, Article in German, abstract only. Rainy, H., 1868. On the cooling of dead bodies as indicating the length of time that has elapsed since death. Glasgow. Med. J. 1, 323 323330. Rentoul, Rentou l, E., Smith, Smith, H., 1973. Tim Timing ing of dea death. th. In: Ren Rentou toul, l, E., Smi Smith, th, H. (Ed (Eds.) s.),, Gla Glaist ister er s Medical Jurisprudence and Toxicology, 13th ed. Churchill Livingstone, Edinburgh, pp. 132 132 133. ’
52
Human Body Decomposition
Rutty, G.N., 1997. Letter to the editor. Forensic Sci. Int. 87, 171 171172. Schwarz, F., Heidenwolf, H., 1953. Post mortem cooling and its relation to the time of death. Int. Police Crim. Rev. 73 (339), 339 339 344. Shapiro, H.A., 1954. Medico-legal mythology: some popular forensic fallacies. J. Forensic Med. 1, 144 144169. Shapiro, Shapir o, H.A H.A., ., 196 1965. 5. The pos post-m t-mort ortem em tem temper perat ature ure pla platea teau. u. J. For Forens ensic ic Med Med.. 12 (Su (Suppl ppl.) .),, 4, 137 137 141. Simonsen, J., Voigt, J., Jeppeson, N., 1977. Determination of the time of death by continuous post mortem temperature measurements. Med. Sci. Law 17 (2), 112 112 121. Taylor,, A.S Taylor A.S., ., Wil Wilks, ks, D., 1863. 1863. On the cooling cooling of the human body aft after er dea death. th. Guy s Ho Hosp. sp. Rep.180 Rep.180 211. ’
Underwood, E.A., 1951. The history of the quantitative approach in medicine. Br. Med. Bull. 7 (4), 265 265274. Vaughan, E.M., 1921. Method of determining approximate time of death. J. Am. Med. Assoc. 76 (9), 608 608609. Winterton, R.H.S., 1999. Newton s law of cooling. Contemp. Phys. 40 (3), 205 205212. ’
Womack, F., 1887. The rate of cooling of the body after death. St. Bartholomew s Hosp. Rep. 23 (Suppl.), 193 193200. ’
CHAPTER
3
Biochemical Methods of Estimating the Time Since Death
INTRODUCTION CHEMICAL INVESTIGATION OF VITREOUS HUMOR Determining Validity and Defining Methods of Collection and Measurement ACCEPTANCE OF THE RELATIONSHIP OF VITREOUS POTASSIUM WITH THE PMI RECENT ADVANCES ANDIUM THE CURRENT STATUS OF RESEARCH INTO VITREOUS VITRE OUS POTASSIUM POTASS THE ESTIMATION OF HYPOXANTHINE IN VITREOUS HUMOR CHEMICAL INVESTIGATION OF SYNOVIAL FLUID BIOMARKERS OF THE PMI IN THE BONE MARROW AND OTHER BODY ORGANS IMMUNOHISTOCHEMISTRY HIGH-RESOLUTION PROTON MAGNETIC RESONANCE SPECTROSCOPY ELECTROLYTE ELECTR OLYTE CONCENTRAT CONCENTRATIONS IONS IN CSF ELECTROLYTE ELECTR OLYTE CONCENTRAT CONCENTRATIONS IONS IN PERICARDIAL PERICARDIAL FLUID THE USE OF DNA AND RNA IN THE ESTIMATION OF THE PMI CONCLUSIONS REFERENCES
INTRODUCTION The esti The estima mati tion on of the the po post stmo mort rtem em in inte terv rval al (PMI (PMI)) by bi bioc oche hemi mica call mean me anss is ba bassed on the anal analys ysis is of ch chem emic ical al subs substa tanc nces es whic which h are are re rele leas ased ed af afte terr deat death h and and accu accumu mula late te in the the body body.. In th theo eory ry,, if any any Human Body Decomposition. DOI: http://dx.doi.org/10.1016/B978-0-12-803691-4.00003-0 © 2016 Elsevier Inc. All rights reserved.
54
Human Body Decomposition
substa subs tanc ncee ca can n be meas measur ured ed accu accura rate tely ly and and corr correl elat ated ed with with th thee ti time me since death (TSD), it might provide a method of determining the PMI. With Wi thin in a few few mi minu nute tess of deat death, h, auto autoly lysi sis, s, or cell cell di disr srup upti tion on,, begi begins ns with the release of water, of which the body is mainly composed, and enzymes, which begin the degradation of proteins, lipids, and carbohydrates. Water containing chemical breakdown products accumulates as blisters under the skin and intestinal cell walls break down, releasing bact ba cter eria ia whic which h aid aid in the the putr putref efac acti tion on proc proces ess. s. Fu Fung ngii and and prot protoz ozoa oa colo co loni nize ze body body ca cavi viti ties es,, de dest stro royi ying ng and and cons consum umin ing g body body ti tiss ssue ues. s. Proteins, lipids, and carbohydrates are broken down by the enzymes prot pr otea ease ses, s, lipa lipase ses, s, an and d glyco glycoge gena nase se,, re resp spec ecti tive vely, ly, in into to am amin ino o acid acids, s, triglyceride trigly cerides, s, and glucose. glucose. Th Thes esee la last st thre threee subs substa tanc nces es are are then then brok broken en down further into simpler molecules, electrolytes, and gas with the help of microorganisms. Amino acids finally break down to ammonia gas and amines such as histamine, tyramine, tryptamine, and phenylethylamine as well as the volatile amines, putrescine, and cadaverine (Forbes, (Forbes, 2008 2008). ). Triglycerides are further broken down by hydrolysis and the action of lipas lipases es to fo form rm satu satura rate ted d and and unsa unsatu tura rate ted d fa fatt tty y acid acids. s. In aerob aerobic ic conditions the unsaturated fatty acids are oxidized to form odoriferous alde aldehy hyde dess and and keto ketone nes, s, bu butt in anae anaero robi bicc cond condit itio ions ns the the unsat unsatur urat ated ed fatty acids are hydrogenated to form saturated fatty acids which form a solid mass of adipocere tissue. Some of the fatty acids formed are vola vo lati tile le fatt fatty y ac acid idss such such as pr prop opio ioni nic, c, buty butyri ric, c, and and va vale leri ricc acid acids. s. Measuring the ratio of these in the soil under a buried body may be Vass ss et al al., ., 19 1992 92). ). Microorganis Microorganisms ms usef us eful ul in de dete term rmin inin ing g the the TS TSD D (Va break bre ak down down carboh carbohydr ydrate atess to form form glucos glucose. e. Fung Fungii may may form form orga organi nicc acid ac idss such such as gluc glucur uron onic ic,, ox oxal alic ic,, an and d citr citric ic acid acids. s. Oxal Oxalic ic acid acid has has been been found to be a useful determinant of the TSD (Vass ( Vass et al., 2002 2002). ). In an aerobic environment glucose will be broken down to lactic and pyruvic acids and then to carbon dioxide and water while in anaerobic conditions, butyric and acetic acids will be formed and then subsequently broken down to ethanol and butanol. Bacterial fermentation will produce methane, hydrogen sulfide, and hydrogen gases (Forbes, (Forbes, 2008). 2008). A soup soup of flui fluid d is ther theref efor oree prod produc uced ed duri during ng deco decompo mposi siti tion on which which contains electrolytes and organic compounds. Theoretically, it should be possible to sample and measure some of these from various compartments of the body, in order to equate the concentrations with the TSD. Blood is unsuitable for sampling because it pools in the dependent parts
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
55
of the the body body afte afterr de deat ath h and and cl clot ots. s. The The vari variab abil ilit ity y of anal analyt ytes es in po post st-mortem blood and autolysis of blood cells and vascular endothelium is too rapid to be of practical use for obtaining chemical markers (Coe, ( Coe, 1993; Madea and Musshoff, 2007). 2007). Similarly, cells lining the subarachnoid space in the brain also undergo too rapid autolysis, such that variability in the chem ch emic ical al co cons nsti titu tuen ents ts limi limits ts th thee use use of cere cerebr bros ospi pina nall fl flui uid d (CSF (CSF)) fo forr analysis (Naumann, (Naumann, 1959 1959). ). Fluids such as the vitreous humor from the eye and synovial fluid from joints, which are located in closed compartments, have been found to be of more use in providing samples for chemical analysis because they are protected from bacterial action, chemical interaction with other body breakdown products, and putrefaction for a longer period after death. The estimation of potassium in the vitreous humor has been studied in the postmortem period in detail (Naumann, (Naumann, 1959; Jaffe, 1962; Sturner and Gantner, 1964; Leahy and Farber, 1967; Lange et al., 1994 19 94;; Ta Tagl glia iarro et al al., ., 19 1999 99,, 20 2001 01;; Ma Mad dea ea,, 20 200 05) and and toget ogeth her with with Henssge s nomo nomogr gram am (Henss Henssge, ge, 1988 1988)) ar aree the the meth method odss em empl ploy oyed ed most most ’
often to estimate the TSD in the early postmortem period.
CHEMICAL INVESTIGATION OF VITREOUS HUMOR Determining Validity and Defining Methods of Collection and Measurement In life, an active transport mechanism maintains chemical equilibrium between plasma potassium in the retinal vessels and the vitreous humor of the eye, but at death this ceases and potassium concentration in the vitreous humor increases due to simple diffusion from the retina and to a 2002). ). Because of the lesser extent, from the lens (Madea (Madea and Henssge, 2002 vary va ryin ing g conc concent entra rati tion on thro throug ugho hout ut the globe globe of the eye, eye, coll collect ectio ion n of vitreous humor should be standardized by aspirating all the vitreous with a large-gauge needle inserted through the outer canthus of both eyes into the center of the globe (Coe, (Coe, 1993). 1993). To prevent contamination with retinal cells, high suction should not be applied and all the vitreous humor should be aspirated as there is variability between many of the solutes in the fluid next to the retina and the fluid in the center of the globe. Vitreous humor is viscid; it should be centrifuged and the supernatant fluid analyzed. Each la labo bora rato tory ry shou should ld dete determ rmin inee re refe fere renc ncee rang ranges es of no norm rmal alit ity y for for th thee chemicals being analyzed. Most chemical investigations of vitreous humor have concentrated on postmortem concentrations of potassium, but more recent rec ently, ly, hyp hypoxa oxanth nthine ine concen concentra tratio tions ns hav havee been been stu studie died d (Saugs Saugstad tad and Olaisen, 1978; Madea et al., 1994). 1994).
56
Human Body Decomposition
Na Naum uman ann n (1 (195 959) 9) meas measur ured ed conc concen entr trat atiions ons of ur urea ea,, gl gluc ucos ose, e, crea creati tini nine ne,, chlo chlori rine ne,, in inor orga gani nicc phos phosph phat ate, e, calc calciu ium, m, sodi sodium um,, and and potassium in the postmortem vitreous humor. He compared concentrations of the same substances in postmortem CSF and found a greater rise in the concentrations in CSF compared with vitreous humor. The potass pot assium ium concen concentra tratio tion n was twice twice the initia initiall concen concentra tratio tion n in vit vitreo reous us humor and seven times the initial concentration in CSF, 9 hours post mortem. This was attributed to the rapid autolysis of the ependymal cells lining the brain ventricles. Although Naumann noted the increase in potassium concentration in the vitreous humor after death he did not suggest it as a measure of the PMI. Jaffe (1962) was the first to do this when wh en he esti estima mate ted d the the pota potass ssiu ium m co conc ncen entr trat atio ion n at va vari riou ouss in inte terv rval alss after death in the vitreous humor of 31 bodies in which the PMI was know kn own. n. Pe Peop ople le dy dyin ing g with with urem uremia ia or el elec ectr trol olyt ytee abno abnorm rmal alit itie iess we were re excluded. A steady rise in potassium concentration from an initial level of 4 mEq/L up to 8 mEq/L after 9 hours was noted after which it rose more slowly, reaching concentration 10 mEq/L after 30 hours. It then continued to rise aslowly up to 125ofhours post mortem. A linear relationship between potassium concentration and the logarithm of the number of hours after death was noted. In discussing the results, Jaffe warned that excessive force of aspiration of vitreous fluid would detach re reti tina nall cell cellss and and re resu sult lt in fals falsel ely y hi high gh pota potass ssiu ium m va valu lues es.. Jaff Jaffee al also so st stat ated ed that that with withdr draw awal al of smal smalll am amou ount ntss of fl flui uid d rath rather er than than th thee whol wh olee am amou ount nt also also gave gave un unre repr pres esen enta tati tive ve re resu sult lts, s, su sugg gges esti ting ng th that at thee di th dist stri ribu buti tion on of pota potass ssiu ium m thro throug ugho hout ut th thee vitre vitreou ouss fl flui uid d was not not uniform. The levels of lactic acid, pyruvic acid, ascorbic acid, non-protein nitrogen, sodium, and chloride were also studied but found to be of less use than the measurement of potassium. This study stimulated research into using vitreous potassium concentration as a measure of the PMI. Adelson and colleagues (1963) measured potassium concentrations in the vitreous humor of 349 eyes in 209 individuals who had died from multiple causes of trauma and disease. The time of death was know kn own n to wi with thin in 15 minu minute tes. s. Th Ther eree was was no di diff ffer eren ence ce in pota potass ssiu ium m concentration from diluted or undiluted specimens, from the right or left eye, or by storing the samples for 4 45 days. They found that the potassium concentration increased in proportion to the increase in PMI, but, bu t, cont contra rary ry to Jaffe s (1 (196 962) 2) find findin ings gs,, the the re rela lati tion onsh ship ip wa wass li line near ar ’
betwee betw een n pota potass ssiu ium m conc concen entr trat atio ion n and and the the PMI, PMI, not not on a logarithmic logarithmic scale but on an arithmetic scale. The type of death did not affect the
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
57
sl slop opee of the the re regr gres essi sion on line line,, its its in inte terc rcep ept, t, or th thee av aver erag agee pota potass ssiu ium m value, but the potassium concentration was more variable where the ago ag ona nall peri period od was was grea greate terr tha han n 6 ho hou urs or in per perso sons ns dy dyin ing g of a chro ch roni nicc ling linger erin ing g di dise seas ase. e. Adel Adelso son n and and coll colleag eague uess co comm mmen ente ted d that that althou alt hough gh the techni technique que was simple simple,, inexpe inexpensi nsive, ve, and used used app appara aratus tus whic wh ich h wa wass re read adil ily y avai availa labl ble, e, the the stan standa dard rd er erro rorr in case casess of ac acut utee trauma dying within 6 hours was 6 5.75 hours for a PMI of 25 hours, but when all cases were included, the error was 6 10 hours (PMI not st stat ated ed)) whic which h they hey cons consid ider ered ed too too grea greatt to be of pr prac acti tica call use use in determining the PMI. Sturner (1963) (1963) studied studied potassium concentration in the vitreous humor of 125 125 po post stmo morrtem tem ca case sess from from both both hosp hospit ital al and and coro corone nerr s sou sources rces.. Case Ca sess were were ex excl clud uded ed if the the time time of de deat ath h was was no nott ac accu cura rate tely ly know known, n, when wh en the the flui fluid d was was colo colore red, d, cl clou oudy dy,, or conta contain ined ed part partic icul ulat atee matter matter,, when wh en les less th than an 0.5 0.5 mL was was asp aspirat irated ed,, and and when when the the spec specim imen enss we were re removed too long after autopsy. This left 69 analyses, both eyes being aspi as pira rate ted d in 15 case casess. The The PMI PMI vari varied ed wi wide dely ly from from 3 to 10 104 4 hou hours. rs. Sturner confirmed the arithmetic linear relationship between potassium concentration and the PMI up to 100 hours after death and he calculated a formula for the estimation of the PMI: ’
PMI ðin hoursÞ hoursÞ 5 ð7 14 3 K1 concentration concentrationÞÞ 2 39 1 :
:
(3.1)
The error rate was 6 4.7 hour hourss which which di did d not not inc incre reas asee with with ti time me,, making it useful for determining the PMI after 24 hours. Sturner and Gantner (1964) expanded (1964) expanded on this series of cases, examining the vitreous humor in 125 cadavers but they excluded cases where the time of death was not was not kn know own n to with within in 15 mi minu nute tes, s, wher wheree the the vi vitr treo eou us fl flu uid was was cloudy, colored, or contained particulate matter, where the amount of fluid aspirated was less than 0.5 mL, where the specimens were removed more than 24 hours after autopsy, and where technical problems arose. Sturnerr and Gantner Sturne Gantner distinguis distinguished hed coroners coroners cases from hospital cases and confirm confirmed ed Sturne Sturnerr s previous previously ly quoted quoted for formul mula a for est estima imating ting the PMI (Sturner, (Sturner, 1963 1963). ). They They confir confirmed med that that the potass potassium ium concen concentra tratio tion n in th thee co coro rone ners rs case casess in incr crea ease sed d in a li line near ar,, arit arithm hmet etic ic mann manner er with with time up to 100 hours after death and it was not affected by temperature. Thee st Th stan anda dard rd erro errorr of 6 4.7 up to 100 hours post mortem did not increase with time, and Ade Adelson lson and colle colleagues agues (1963) finding that the ’
’
’
’
vitreous vitreo us potass potassium ium con concen centra tratio tion n was much much more more var variab iable le in deaths in hospit hos pital al patien patients ts compar compared ed with with deaths deaths fro from m tra trauma uma,, was con confir firmed med
58
Human Body Decomposition
with a st with stan anda dard rd erro errorr twic twicee that that of co coro rone ners rs cases. cases. They They attrib attribute uted d this th is to an incr increa ease se in se seru rum m potas potassi sium um pr prio iorr to de deat ath, h, to en enzy zyma mati ticc infl influe uenc nce, e, and and to phys physio iolo logic gic va vari riat atio ion. n. Stur Sturne nerr an and d Gant Gantne nerr stat stated ed that the period from 24 hours after death and up to 100 hours represented the most useful working time of the procedure. ’
Sturner and Gantner also stated that it was generally accepted that vitreous potassium levels were dependent on and very nearly equal to the potassium content of serum but this could not be justified by the two studies referred to (Duke-Elder, (Duke-Elder, 1929; Reddy and Kinsey, 1960 1960). ). The first of these studies by Duke-Elder by Duke-Elder (1929) (1929) was carried out on the enucleated eyes of freshly slaughtered horses from an abattoir. The second study by Reddy and Kinsey (1960) (1960) analyzed the chemical content of the vitreous humor, aqueous humor, and plasma in living and then freshly killed rabbits. The potassium concentrations of the vitreous humor, the aqueous humor, and the plasma in Duke-Elder s study were found to be almost equal and it was postulated that the vitreous and aqueous humor were dialysates of capillary blood. Reddy and Kinsey could not explain the source, function, and distribution of the high potassium content of the vitreous fluid compared to that of the aqueous humor and plasma but postulated that there might be a positive and negative concentration gradient between posterior chamber aqueous and the vitreous humor. Henry humor. Henry and Smith (1980) stated in a review of the estimation of the PMI by chemical means that virtually no information was available about the antemortem composition of human intraocular fluids and this remains true at the present time (Madea (Madea and Musshoff, 2007 2007). ). ’
Hughes (1965) (1965) reported reported contrary findings in a study of 117 cases of sudden death, both coroners cases and hospital deaths, with PMIs up to 120 hours after death. The potassium concentration in each eye of ever ev ery y ca case se was was with within in 0.1 0.1 mE mEq/ q/L. L. Hughe Hughess co conf nfir irme med d the the in incr crea ease se in potassium concentration with increasing PMI, but found no consistent linear relationship between vitreous potassium concentration and the PMI, either with sudden deaths or with lingering hospital deaths and thee time th time of deat death h coul could d not not be put put mo more re accu accura rattel ely y than than wi with thin in 24 hours. hours. She conclu concluded ded that that meas measuri uring ng postmo postmorte rtem m vitreo vitreous us pot potasassi sium um conc concen entr trat atio ion n was an in insu suff ffic icie ient nt meth method od fo forr de dete term rmin inin ing g the the PMI. Hansson and colleagues (1966) studied 203 autopsies in which ’
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
59
the time of death was known to within 1 hour. Commenting that previouss st ou stud udie iess ha had d only only stud studie ied d pota potass ssiu ium m co conc ncen entr trat atio ions ns up to abou aboutt 40 hours, the present study was designed to determine whether estimation of the vitreous potassium was useful when the PMI was longer. There were two groups; one with an agonal period less than 6 hours and one with an agonal period longer than 6 hours and both groups included cases dying of accidental death and disease. The PMI varied from 3 to 310 hours. In 180 cases with a PMI up to 120 hours, a rising potassium concentration was observed after which time it remained constant. The potassium concent ce ntra rati tion on di did d not not di diff ffer er ac acco cord rdin ing g to the the ty type pe of deat death h or th thee agona agonall period. The standard error in the PMI determination was 6 20 hours up to 120 hours, twice as great as that reported by Adels Adelson on et al al.. (1 (196 963) 3),, after which it became even greater. Hansson and colleagues commented that the large error may have been due to technical difficulties in gathering in g and and anal analyz yzin ing g the the spec specim imen enss but, but, de desp spit itee the the re resu sult lts, s, th thee meth method od might be of some use in cases where the secondary signs of death pro’
vided nocases help dying in determining time ofwith death. Lie s (1967) 88 hospital of variousthe diseases, a known time ofstudy deathofand a PMI varying from 2 to 95 hours, found a much smaller standard error of 6 4.7 hours for the PMI with a correlation coefficient of 0.99 using (1963; see Eq. see Eq. (3.1) (3.1)). ). The arithmetic linthe formula suggested by Sturner by Sturner (1963; ear increase in potassium concentration with lengthening PMI was confirmed. Lie emphasized that all samples should be collected in a standard way wa y in or orde derr to av avoi oid d any any tech techni nica call erro errors rs.. Th Ther eree was was no di diff ffer eren ence ce in potassium concentration between both eyes of the same cadaver and potassium concentrations did not differ significantly between fresh and refrigerated specimens of vitreous humor. Leahy and Farber (1967) (1967) studied studied the postmortem chemistry of vitreous humor in a series of 52 cases. They estimated antemortem concentrations of sodi sodium um,, chlo chlori ride de,, an and d pota potass ssiu ium m as we well ll as gl gluc ucos osee and and se seve vera rall enzymes, in blood collected within 24 hours of death and compared these concentrations with postmortem vitreous concentrations of the same substances. They found that vitreous potassium concentration rose erratically after aft er de death ath and and it was not not possi possibl blee to calcu calcula late te a normal normal postm postmort ortem em range of any significance. Any mathematical relationship between vitreous potassium concentration and the PMI in 12 patients who had died sudd su dden enly ly was was not not appa appare rent nt.. Th They ey co conc nclu lude ded d th that at co conc ncen entr trat atio ions ns of
60
Human Body Decomposition
sodium, chloride, and potassium in postmortem vitreous humor did not have ha ve ev evid iden entt cl clin inic icop opat atho holo logi gica call corr correl elat atio ion n and, and, in part partic icul ular ar,, the the unpredic unpr edictabl tablee natu nature re of post postmort mortem em potassiu potassium m con concent centrati rations ons in deterdetermining the PMI did not justify its use. Coe (1969) studied 20 cases in which he analyzed postmortem vitreous electrolytes, including potassium, gluco glu cose, se, and and other other su subs bstan tances ces,, in ord order er to es esta tabli blish sh no norm rmal al range rangess fo forr thes th esee subs substa tanc nces es and and to dete determ rmin inee whic which h vari variat atio ions ns in thes thesee valu values es refl reflec ecte ted d ante antemo mort rtem em abno abnorm rmal alit itie iess and and whic which h re refl flec ecte ted d post postmo mort rtem em changes. The concentrations of these substances were compared with concentrations of the same substances in blood obtained from the heart. Coe found that the potassium concentration increased linearly for at least 100 hours after death, more rapidly within the first few hours but the rate of rise had marked individual variation. Because of this inconstant postmortem change the antemortem value of vitreous potassium would not be possible to be obtained by regression analysis. He agreed with Adelson with Adelson et al. (1963),, Hughes (1965) (1963) (1965),, and Hansson and Hansson et al. (1966) that (1966) that although the potassium rose arithmetically after death, there such anthat individualconcentration variation in the rate of rise, especially after the firstwas 24 hours, it was of very limited value in determining the PMI. Coe PMI. Coe (1969) (1969) found found a standard error rate of 6 12 hours in the first 24 hours, growing increasingly larger after that time. Adjutantis and Coutselinis (1972) proposed a method of estimating the PMI within the first 12 hours by estimating the vitreous potassium conc co ncen entr trat atio ion n and and usin using g the the line linear ar in incr creas easee with with ti time me to ex extr trap apol olat atee back ba ckwar wards ds in orde orderr to esti estima mate te th thee pota potass ssiu ium m conc concen entr trat atio ion n in li life fe.. They sampled 1 mL of vitreous humor from one eye after delivery of the body and then the same amount from the other eye, in the majority of case casess at a 3-ho 3-hour ur in inte terv rval al afte afterw rwar ards ds.. Most Most ca case sess di died ed a vi viol olen entt death but the number of cases was not specified. The vitreous fluid was centrifuged and the potassium concentration determined by flame phot ph otom omet etry ry.. In a smal smalll numb number er of case cases, s, sa samp mple less were were ta take ken n from from both eyes at the same time and the authors found that whatever variation of removal was used, there was no statistical difference in potassium concentration between the eyes. A statistical graph of potassium conc co ncen entr trat atio ion n pl plot otte ted d again against st time time afte afterr de deat ath h fo foun und d a li line near ar ri rise se in potassium concentration up to 12 hours post mortem, after which time it ceased to increase. This was attributed to cessation of diffusion of potass pota ssiu ium m from from auto autoly lyzi zing ng cell cellss in the the re reti tina na.. Extr Extrap apol olat atio ion n of the the regression slope, indicating potassium concentrations, backwards to its
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
61
intersection with the axis denoting the PMI, indicated that the potassium concentration at the moment of death was 3.4 mEq/L. Adjutantis and Coutselinis calculated that in two-thirds of cases, the PMI could be esti estima mate ted d with with an erro errorr rate rate of 6 1. 1.1 1 hour hourss wi with thin in th thee fi firs rstt 12 hours. Henry and Smith (1980), (1980), however, commented that others had found fou nd the tec techni hnique que used used by Adjuta Adjutanti ntiss and Cou Coutse tselin linis is unusab unusable le or that th at th thee erro errorr rate rate of the the PM PMII was was mu much ch grea greate ter, r, in the rang rangee of 6 10 10 26 hours at 12 hours. Balaso asoori oriya ya and col collea league guess (19 (1984a) 84a) contr contrad adic icte ted d the the fi find ndin ings gs of Bal Adelson et al. (1963) who had found no bias in potassium concentrations from right and left eyes. Cases in which the PMI could be verified fied to with within in 15 mi minu nute tess were were stud studie ied d and and it was was fo foun und d th that at the potass pot assium ium concen concentra tratio tion n in the vitreo vitreous us humor humor varied varied betwee between n each each eye, with a gradual linear increase in potassium concentration during the first 85 hours after death. When the potassium concentration was compared between the two eyes, 18.6% of the results varied by more than 10% from the mean of the two values, 50.8% were greater than 4% from the mean, and only 6 cases out of 59 had the same potassium conc co ncen entr trat atio ions ns in each each eye. eye. The The di diff ffer eren ence cess were were no nott re rela late ted d to th thee TSD or the mode of death but only a small quantity of 1 mL of fluid was withdrawn withdrawn from each eye. Ba Bala laso soor oriy iya a and and coll collea eagu gues es co conc nclu lude ded d that there were previously unreported differences in potassium concentration between the two eyes of any cadaver and this made it an unreliable estimate of the PMI. (1989) commented that the estimation of vitMadea and colleagues (1989) commented re reou ouss po pota tass ssiu ium m was was not not used used in Germ German any y to de dete term rmin inee the the PMI PMI because the error rate varied so greatly between studies that it was not possible to determine which values to use. The various studies to that time averaged error rates between 6 9.5 and up to 6 40 hours in the first 100 hours after death and between 6 6 and 6 12 hours in the first 24 hours. hours. The pot potass assium ium concen concentra tratio tion n cal calcul culated ated in various various stu studie diess also varied between 5 and 8 mEq/L and the rise in potassium concentr trat atio ion n each each hour hour afte afterr de deat ath h vari varied ed be betw twee een n 0. 0.1 1 and and 0. 0.2 2 mEq/ mEq/L. L. Madea Mad ea an and d coll collea eagu gues es po post stul ulat ated ed that that el elec ectr trol olyt ytee imba imbala lanc nces es at th thee moment of death might be responsible for such large errors but studies on this this as aspe pect ct we were re lack lackin ing. g. Th They ey th ther eref efor oree stud studie ied d 170 170 cons consec ecut utiv ivee cases with an accurately known time of death comprised of 100 cases of su sudd dden en deat death h an and d 70 case casess of de deat ath h foll follow owin ing g chro chroni nicc di dise seas ase. e.
62
Human Body Decomposition
The vitreous humor from both eyes was evacuated completely at the samee time, sam time, cen centri trifug fuged ed and analyze analyzed d measur measuring ing potass potassium ium,, sodium sodium,, chloride, urea, and calcium concentrations. They found a 10% deviation from from the the mean mean po pota tass ssiu ium m co conc ncen entr trat atio ion n in th thee vi vitr treo eous us hu humo morr of bo both th eyes ey es,, whi which was was in inde depe pend nden entt of the the TSD TSD or mo mod de of de deat ath h. In this this respect they agreed with Balasooriya with Balasooriya et al. s (1984a) findings. (1984a) findings. They thereforre took fo ook the the mean mean va valu luee of pota otass ssiu ium m co conc ncen entr trat atio ion n in both both ey eyes es in their study, in an attempt to narrow down or eliminate the wide range of er erro rorr appa appare rent nt from from di diff ffer eren entt stud studie ies. s. Th They ey agai again n conf confir irme med d the line linear ar re rela lati tion onsh ship ip be betw twee een n vitr vitreo eous us pota potass ssiu ium m co conc ncen entr trati ation on an and d TSD, TS D, pota potass ssiu ium m conc concen entr trat atio ions ns stea steadi dily ly ri risi sing ng up to 12 120 0 hour hourss post mortem. However, the error rate was 6 34 hours up to a PMI of 120 hours for all cases. To eliminate electrolyte imbalance at death as a ca caus usee for for the the erro errorr ra rate te,, ca case sess with with a urea urea conc concen entr trat atiion over over 100 mg/dL (35.7 mmol/L) were removed from the study, as all these cases suffered from chronic disease ase before death. Mad Madea and ’
coll co llea eagu gues cons consid ider ered edvitreous that that the th e urea urea co ncen entr trat atio ion nwere wa wassmost norm normal ally ly stable ines postmortem humor andconc raised levels likely a marker of antemortem electrolyte imbalance. All cases with a urea conc co ncen entr trat atio ion n over over 100 mg mg/d /dL L (35. (35.7 7 mmol mmol/L /L)) su suff ffer ered ed from from ch chro roni nicc disease. This reduced the number of cases to 138 and also reduced the erro errorr ra rate te to a stat statis isti tica callly si sign gnif ific ican antt 6 22 hours. Madea and coll co llea eagu gues es thus thus elim elimin inat ated ed mo most st of the the case casess with with ch chro roni nicc di dise seas asee in Adelson et al. (1963) had noted a greater range of scatter of the which Adelson which pota po tass ssiu ium m conc concen entr trat atio ion. n. To reduc educee the the er errror rat atee furt furthe her, r, they they eliminated cases with an agonal period greater than 6 hours; a group in which Adelson which Adelson et al. (1963) had (1963) had also noted that there was a greater range of scatter of potassium concentration. The number of cases was redu re duce ced d to 107 107 an and d the the erro errorr rate ate to 6 20 hour hours, s, whic which, h, al alth thou ough gh with wi thin in th thee 95% 95% limi limits ts of conf confid iden ence ce of the the re regr gres essi sion on li line ne,, was was not not st stat atis isti tica call lly y si sign gnif ific ican ant. t. Th They ey co conc nclu lude ded d th that at ot othe herr fa fact ctor orss su such ch as ambient temperature and blood alcohol level at the time of death could be poss possib ible le fact actor orss in prod produc ucin ing g such uch a grea greatt er erro rorr rat atee but but that that esti estima mati tion on of vitr vitreo eous us humo humorr urea urea ni nitr trog ogen en conc concen entr trat atio ion n was was a suit su itab able le st stan anda dard rd to dete determ rmin inee di dist stur urbe bed d body body meta metabo boli lism sm be befo fore re death. dea th. Howeve However, r, furthe furtherr studie studiess were were requir required ed to det determ ermine ine whethe whetherr the arbitrarily chosen urea concentration of 100 mg/dL (35.7 mmol/L) should be reduced further in order to reduce the error rate. A major prob pr oble lem m ho howe weve verr wa wass that that no norm rmal al ante antemo mort rtem em valu values es of vi vitr treo eous us potassium were unknown.
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
63
Madea and colleagues (1990) studied studied the potas potassium sium concen concentrati tration on in the the vitr vitreo eous us humo humorr of 100 100 case casess of su sudd dden en natu natura rall or trau trauma mati ticc death, with a terminal period less than 6 hours and where the ambient temperature was below 50 F (10 C). The purpose of the study was to compare their formula for estimating the PMI from the vitreous potassium concentration with that of f Sturner s (196 963 3) form formul ula a (see (see (Eq. 3.1)). 3.1)). They found that the linear rate of increase of vitreous potassium concentration was faster than that found previously by Sturner, meanin mea ning g tha thatt Sturne Sturnerr s model overestimated the PMI. In comparing their formula: formula: ’
’
PMI 5 5 26 3 K1concentration 2 30 9 :
:
;
(3.2)
(3.1) .1)), ), Ma Made dea a and and co coll llea eagu gues es foun found d a with Stu with Sturne rnerr s form formul ula a (Eq. Eq. (3 mean difference between real and extrapolated TSD of 15 hours up to 100 hours with Sturner s formula, compared with 2 0.26 hours using their formula. When they included the 138 cases in their previous study with wi th a PMI over over 30 ho hour urss and and with with ure rea a valu values es be bellow 100 100 mg/d mg/dL L 1989), ), the mean difference rose to 26 hours with an (Madea et al., 1989 erro errorr rang rangee of 6 41 hour hourss at the 95% 95% li limi mits ts of co conf nfid iden ence ce usin using g Sturner s formula. Madea and colleagues concluded that the error of estimation of PMI was much less using their formula, but the error range within the 95% limits of confidence was still 6 20 hours up to 100 hours post mortem. Stephens mortem. Stephens and Richards (1987) (1987) studied studied postmortem vitreous pota po tass ssiu ium m conc concen entr trat atio ions ns by wi with thdr drawi awing ng sa samp mple less at know known n ti time mess af afte terr de deat ath h an and d comp compar arin ing g them them wi with th th thee PMI. PMI. They They anal analy yzed zed 1427 1427 samples from an unstated number of cases up to 35 hours post mortem ’
’
’
and found a wide range of potassium concentrations for a known range of PMIs. For the first time in a study, vitreous potassium concentration was correlated with PMI as the dependent variable rather than the PMI being correlated with vitreous potassium concentration as the dependent variable. Very large or small potassium concentrations, which were attributed ut ed to ab abno norm rmal al de deat aths hs,, we were re excl exclud uded ed.. Step Stephe hens ns an and d Ri Rich char ards ds co connfirm firmed ed that that th thee pota potass ssiu ium m conc concen entr trat atio ion n ro rose se in a li line near ar fa fash shio ion n bu butt found that only 37.4% of potassium concentration variation was directly attributable to the PMI, the margin of error being 6 20 hours up to 35 hours post mortem. The other 62.6% of the variation in potassium concentration was unaccounted for, but might be due to factors such as variable ambient temperature, sampling technique, sex, age, ancestry, cause of deat death, h, or othe otherr unkn unknow own n fact factor ors. s. Step Stephe hens ns an and d Ri Rich char ards ds su sugge ggest sted ed
64
Human Body Decomposition
that studies focusing on the use of other chemical determinants in con juncti jun ction on wit with h potass potassium ium might might prove prove useful useful..
ACCEPTANCE OF THE RELATIONSHIP OF VITREOUS POTASSIUM WITH THE PMI Coe (1989) summed up progress regarding the estimation of the PMI from the vitreous potassium concentration in stating that there was no doubt there was a linear relationship between potassium concentration and an d the the PMI, PMI, bu butt whet whethe herr it was was a stra straiight ght li line ne or a bi biph phas asic ic one one remained debatable. There was great variation in the error of estimation of the PMI between investigators, most reporting a standard error ranging from 6 4.7 to as high as 6 26 hours which increased as the PMI increased. There was also a large variation in the rate of increase of pota potass ssiu ium m conc concen entr trat atio ion, n, fr from om 0.14 0.14 mE mEq/ q/L L per per hour hour fo foun und d by Sturner (Sturner (Sturner and Gantner, 1964; Sturner, 1963 1963)) to 0.55 mEq/L per ho hour ur de dete term rmin ined ed by Adj Adjut utan anti tis s and Cout Co utse seli lini niss (1 (197 972) 2), , wthese ith m ost values being about 0.17 mEq/L per hour. Coe suggested that variations were due to both external and internal factors. The external factors to rs were were the the samp sampli ling ng tech techni niqu que, e, the the anal analyt ytic ical al in inst stru rume ment nts, s, the the envi en viro ronm nmen enta tall temp temper eratu ature re duri during ng the the PM PMII and, and, most most im impo port rtan antl tly, y, the temperature of the body during the PMI. The internal factors were the age of the individual, the duration of the terminal episode, and the presence or absence of nitrogen retention. Coe stated that the test was useful as long as these factors were properly evaluated in each individual case, and that he had found it to be of greatest value when the body had lain in a cool environment for several days and factors such as livor mortis, rigor mortis, and body temperature were in conflict or not able to be defined. A different approach was adopted by Lange and colleagues (1994) when wh en they they an anal alyz yzed ed 790 case casess from from si six x di diff ffer eren entt stud studie iess carr carrie ied d out out between 1963 and 1989. All studies graphed the rise in vitreous potassium concentration against the PMI. The relationship between the rise in pota potass ssiu ium m co conc ncen entr trat atio ion n and and the the PMI PMI was not not comp comple lete tely ly li line near ar and the residual variability of the potassium concentration as a function of the PMI was not constant. All studies varied in the rate of rise in potassium concentration, that is, the slope of the regression line and in the intercept of the regression line on the y-axis, indicating that the theo th eore reti tica call po pota tass ssiu ium m conc concen entr trat atio ion n at the the ti time me of de deat ath h vari varied ed
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
65
between studies. In general, the authors suggested that determination of the PMI from measured vitreous potassium concentration was best done when the PMI was no later than 24 hours, but rough estimates could extend to 48 4872 hours, especially in cold environments. Lange and an d coll collea eagu gues es sugge suggest sted ed that that th thee larg largee er erro rorr ra rate te in es esti timat matin ing g the the PMI was due to several factors. Firstly, all the studies assumed that thee re th rela lati tion onsh ship ip betw betwee een n vitr vitreo eous us pota potass ssiu ium m conc concen entr trat atio ion n and and the the PMI was linear but Lange and colleagues found that even after statistically transforming the data it was not linear, due to the inability to adjust for factors across all studies, such as age, mode of death, and sampli sam pling ng techni technique que.. Second Secondly, ly, the confid confidenc encee limits limits had been been incorincorrectly defined as the confidence limits for the estimated slope of the linee indica lin indicatin ting g increa increasin sing g potass potassium ium concen concentra tratio tion n ins instea tead d of for the estimated value of the PMI. Thirdly, the inverse prediction method of Stephens and Richards (1987) (1987),, the only other study to have used the method, was more correct, that is, where the PMI depended on the vitreous concentration rather than potassium tr trat atio ion npotassium depe depend ndin ing g on the the PMI. PMI. Beca Becaus usee vitreous of the the nonl no nlin inea eari rity tyconcenof the the regression line in the studies examined, Lange and colleagues statistically constructed a nonlinear, smooth, loess curve and fitted the confidence limits to this. This produced a much reduced margin of error. For potassium concentrations less than 7 mEq/L the error range of the PMI up to 10 hours was 6 1 ho hour ur.. For For po potas tassi sium um conc concen entr trat atio ions ns betw be twee een n 7 and and 12 mE mEq/ q/L L it was was 6 2 hour hourss an and d for for co conc ncen entr trat atio ions ns between 12 and 18 mEq/L it was 6 3 hours for a PMI between 36 and 80 hour hourss and and for for conc concen entr trat atio ions ns great greater er than than 18 mEq/ mEq/L L it was was 6 5 hours and increasing. As the potassium concentration increased above this level, the error in the PMI greatly increased and the estimation of thee PMI th PMI beca became me unre unreli liab able le.. Co Comm mmen enti ting ng on Lange Lange and and coll collea eagu gues es'' study in his review, Madea review, Madea (2005) stated (2005) stated that if these error limits were accurate, it would be almost the perfect method but he was unable to verify the accuracy after testing it on his own forensic cases. A study was carried out by Muñoz et al. (2001) using the inverse prediction method of Stephens and Richards (1987) in order to construct a new formula using vitreous potassium concentration to estimate the PMI. They studied 201 sam amp ples of vitreous humor or,, withdrawing all the fluid from both eyes from 164 bodies in which the time of death could be verified to within 15 minutes. Opaque specime cimens ns of vitr vitreo eous us an and d thos thosee fr from om infa infant ntss le less ss th than an 6 mont months hs ol old d
66
Human Body Decomposition
were discarded. The minimum PMI was 1 hour and the maximum wass 40.4 wa 40.45 5 hou hours wi with th an av aver erag agee of 11 hour hourss an and d sam ampl ples es from from corpses where the TSD could not be established to within 6 15 minutes ut es were were excl exclud uded ed.. There There were were no signi signifi fica cant nt di diff ffere erenc nces es in pota potasssium concentration between both eyes and all samples were classified into into two group groups; s; on onee group group had no known known me meta tabo boli licc distur disturba banc nce, e, usual ually those whe where a sudde den n natural or trau aum matic death ath had occurred while the second group had a metabolic disturbance before deat de ath h such such as in a ch chro roni nicc illn illnes ess. s. Th Thee auth author orss co comp mpar ared ed the the tw two o methods of estimating the PMI, that is, when the PMI was dependent on the potassium concentration and when the potassium concentration tion was was de depe pend nden entt on the the PM PMI. I. They They foun found d a gr grea eate terr er erro rorr ra rate te when the PMI was the independent variable and the vitreous potassium concentration was the dependent variable than when the vitreous potassi potassium um concen concentrat tration ion was the indepe independe ndent nt var variab iable le and the PMI was the dependent variable. Muñoz and colleagues calculated a form fo rmula ula wh which ich expl explai ained ned 70.1 70with .1% % of va vari riat atio ion ntraditional in the the data data, , comcompared with 57.5% explained thethe other more method. Removing cases with a urea concentration . 30 mg/dL and a creatinine concentration $ 0.5 mg/dL further reduced the error rate. They suggested that the remaining error was possibly caused by race, age, sampling technique, and general lack of standardization.
RECENT ADVANCES AND THE CURRENT STATUS OF RESEARCH INTO VITREOUS POTASSIUM In a wide wide-r -ran angi ging ng re revi view ew of post postmo mort rtem em chem chemis istr try y re rese sear arch ch,, Coe (1993) stated stated that that routin routinee examin examinati ations ons of vit vitreo reous us electr electroly olytes tes,, gluglucose, and urea nitrogen alone would provide information in determining the cause of death or time of death in up to 5% of all forensic cases and proper analysis of blood, CSF, vitreous humor, pericardial fluid, and other body fluids would help in about 10% of cases. Coe re revi view ewed ed th thee post postmo mort rtem em an anal alys ysis is of car carbo bohy hydr drat ates es,, ni nitr trog ogen en comcompounds, electrolytes, lipids, proteins and immunological proteins, bile pigments, enzymes, and hormones. He concluded that the postmortem fall in glucose concentrations in vitreous humor was too erratic to be of us usee in esti estima mati tin ng PM PMI, I, and and that that meas measur urin ing g conc concen enttrat rations ions of lact lactic ic,, asc ascor orbi bic, c, and and pyru pyruvi vicc acid acids, s, an and d in inos osit itol ol was not not usef useful ul.. He commented that at various times, measurements of serum non-protein nitrogen, vitreous ammonia, CSF amino acids, vitreous creatine and
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
67
hypoxa hypo xant nthi hine ne,, vitr vitreo eous us and and CSF CSF pota potass ssiu ium, m, vi vitr treo eous us ca calc lciu ium m an and d blood pH, had all been proposed for use in the estimation of the PMI, but that most had only been investigated by a single source and there had been no follow-up studies or independent verification. The exception was the measurement of vitreous potassium concentrations which had been studied extensively. Coe concluded that this seemed to be the best biochemical test available for the estimation of PMIs longer than 24 hours, but in the first 24 hours the margin of error was too great for it to be useful. Ta Tagl glia iaro ro an and d co coll llea eagu gues es (1 (199 998) 8) publ publis ishe hed d a pape paperr ex expl plai aini ning ng capillary zone electrophoresis. Capillary electrophoresis, first reported Hjertén (1967) (1967),, was a system of dispersion of fluid through narrowby Hjertén by bore bo re tube tubess made made of chem chemic ical ally ly and and el elec ectr tric ical ally ly in iner ertt mate materi rial als. s. A aro and col collea league guess (19 (1999) 99) investig investigated ated the seco se cond nd st stud udy y by Tagli Tagliaro estima est imatio tion n of vit vitreo reous us potass potassium ium concen concentra tratio tion n by cap capill illary ary zone zone electrophoresis in order to try to eliminate the differences between thee vari th variou ouss me meth thod odss of co coll llec ecti tion on and and an anal alys ysis is of vi vitr treo eous us po pota tasssi sium um conc concen entr trat atio ion, n, whic which h had had made made the the comp compar aris ison on of di diff ffer eren entt studies difficult. Each element had different speeds of migration dispers pe rsio ion n when when an elec electr tric ical al curr curren entt was was ap appl plie ied d to the the tubi tubing ng and and the dispersions could be measured. In the early 1990s advances had made it possible to carry out more rapid analysis of ions. Only a very small quantity of vitreous humor was required for analysis. It wass in wa inje ject cted ed thro throug ugh h a buff buffer er so solu luti tion on of imid imidaz azol ole, e, 18-c 18-cro rown wn-6 -6 ether, and α-hydroxybutyric acid, buffered to a pH of 4.5 and then passed through a very fine silica capillary tube with a transparent window. By running an electric current between buffer solutions, an electric field was produced as the vitreous humor passed along the capillary tubing, causing the various ions to move at different rates according to their electrical charge. Thus the concentration of each ion could could be measur measured ed using using ul ultra travio violet let absorp absorpti tion on detect detectio ion. n. The study compared the analysis of potassium concentration in the vitreous humor of 12 real autopsy specimens by capillary electrophoresi re siss with with an anal alys ysis is by flam flamee phot photom omet etry ry.. The The resu result ltss we were re well well 2 correlate corr elated d (r 5 0.9333). It was also possible to analyze other ions with wi thin in the vitr itreou eous humor umor,, includ cludin ing g ammo ammoni nia, a, calc calciu ium, m, and and sodium. Tagliaro and colleagues then tested the potassium concentrations in 20 cadavers with a PMI varying between 5 and 96 hours and an d al also so foun found d a high high corr correl elat atio ion n (r2 5 0.9 0.904) 04) betwee between n potass potassium ium
68
Human Body Decomposition
concentra concen tratio tions ns and the PMI. PMI. The They y conclu concluded ded that that ele electr ctroph ophore oresis sis was a simple, accurate method requiring a minimal amount of fluid and it was suitable for forensic application. A study by Ferslew by Ferslew and colleagues (1998) also (1998) also found a high correlation (r 5 0.9 0.9642) 642) bet betwee ween n the potassi potassium um concen concentra tration tionss in the vitreo vitreous us humor of 25 forensic autopsy specimens when analyzed by both capillary electrophoresis and by flame photometry. They concluded that electrop tropho horresis esis offe offere red d an acc accur urat atee and and more ore auto automa matted me meth thod od of potassium analysis than flame photometry. A further study by Tagliaro and colleagues (2001) tested (2001) tested the correlation of potassium concentrations in both eyes in 57 cases of natural or violent death using capillary electrophoresis, in which the PMI was known exactly. The PMI varied from 7 to 144 144 ho hour urss an and d they they foun found d a hi high ghly ly si sign gnif ific ican antt li line near ar co corr rrel elat atio ion n between potassium concentration and the PMI (r 5 0.89, p , 0.0001). In addition addi tion,, there there was no statist statistical ically ly sig signif nifica icant nt dif differ ferenc encee in potass potassium ium concentration between the two eyes of a body at the same PMI. This finding was different from other studies which did find significant differences in potassium concentration between the two eyes from the same cadaver (Balasooriya (Balasooriya et al., 1984a; Madea et al., 1990; Pounder et al., 1998). 1998 ). Tagliaro and colleagues suggested that their anomalous finding was due to only removing a small quantity of vitreous humor (50 µL). The other studies had removed much larger quantities of vitreous humor which might have reduced pressure on the cells of the eye tissues, leading to an artificially high release of potassium. In orde orderr to try try to impr improv ovee the the stat statis isti tica call corr correl elat atio ion n betw betwee een n PMI PMI ti and and pot potass assium ium concent concentrat ration ion in vitreo vitreous us hum humor, or, Bocaz-Beneven Bocaz-Beneventi colleagues (2002) estima colleagues estimated ted the concen concentrat tration ionss of sod sodium ium,, potass potassium ium,, and ammonium in the vitreous humor of 61 cadavers by capillary zone elec el ectr trop opho hores resis is afte afterr di dilu luti ting ng the the vitr vitreo eous us in an aque aqueou ouss so solu luti tion on of barium. The PMI was known in each case, the intervals varying from 3 hours up to 87 hours with only three cases being over 100 hours. The average PMI, discounting the three cases over 100 hours, was 41 hours. Only 50 µ L of vitreous fluid was removed from each eye. They analyzed the results by using the chemical estimations as input data and the PMIs as ou outp tput ut da data ta in a comp comput uter er-g -gen eneerate rated d prog progra ram m (Neu (Neura rall Netw Networ ork k Stimulator, release 3.0D, TRAJAN software, 1998). Using this method they were able to reduce the error in estimating the PMI from
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
69
approximately 6 15.28 hours using potassium estimation by single-ion analysis and linear least squares regression to 4.69 hours using the computer software program. In a crit critic ical al re revi view ew of meth method odss of es esti tima mati ting ng the the PMI PMI by bi bioochemical methods, Madea (2005) stated that all chemical methods of determining the TSD up to the time of writing were of very limited value in practice, since they were neither precise nor reliable and did nott give no give an imm mmed edia iate te resu result lt at the scen scenee of a cr crim ime, e, but but he al also so obse ob serv rved ed that that some some prog progre ress ss ha had d been been made made in meth method odol olog ogy. y. Most Most chemic che mical al met method hodss measur measured ed concen concentra tratio tion n change changess in differ different ent fluid fluid compartments, which were caused by metabolic and autolytic changes influenced by temperature, disease, cause of death, length of the terminal phase before death, and the site and method of sample acquisition. Becau Be cause se of th thes esee fact factor ors, s, vari variat atio ion n betw betwee een n in indi divi vidu dual alss was was so gr great eat that chemical methods were of little use in practice. Although the postmortem measureme men nt of vi vittreous humor potassium concentration had been extensively studied and was widely used, Madea used, (2005) listed the deficiencies with its use, namely: Madea (2005) listed 1. A number number of linear linear regres regressio sion n equati equations ons com compar paring ing the potass potassium ium concentration with environmental temperature had been produced with varying gradients of the regression line, indicating that analysis was not standardized. 2. The The st stat atee of he heal alth th an and d chro chroni nicc ill llne ness ss of an in indi divi vidu dual al migh mightt influence the potassium concentration. By eliminating cases with a vitreo vit reous us ure urea a concen concentra tratio tion n . 10 100 0 mg mg/d /dL L an and d a te term rmin inal al epis episod odee greater than 6 hours, the PMI could be reduced from 6 34 hours, but only to 6 22 and 6 20 hours, respectively, over a PMI of up to 130 hours. 3. By usi using ng pot potass assium ium concen concentra tratio tion n as the independ independent ent variab variable le and the PMI as the dependent variable instead of the reverse, the accuracy of estimating the PMI was increased from 6 25.96 to 6 23.27 hours over a PMI up to 130 hours, but this improvement was not statistically significant. 4. The The re repo port rted ed in incr crea ease sed d accu accura racy cy of es esti timat matin ing g the the PMI PMI by using using a loess curve, as reported by Lange by Lange et al. (1994) (1994),, revealed very high false estimations of the PMI when independently tested. The reliability of this method of statistical evaluation therefore remained unclear.
70
Human Body Decomposition
Mad Madea ea (20 (2005) 05) furt furthe herr comm commen ente ted d that that by us usin ing g mult multip iple le li line near ar regression analysis of sodium, glucose, and urea as well as potassium, the error rate in estimation of the PMI was reduced from 6 16.2 to 6 14 hours and using capillary zone electrophoresis as a method of determining all ions was promising, reducing the estimation of PMI to an error rate of 6 3 hours compared to 6 15 hours over a PMI of up to 200 hours when potassium alone was used. However this method had only been used in a small study group of 61 cases (BocazBene Be neven venti ti et al al., ., 200 2002 2). The varia variati tion on in vi vitr treo eous us hu humo morr po pota tass ssiu ium m conc co ncen entr trat atio ions ns wa wass not not only only due due to in inst stru rume ment ntat atio ion, n, but but al also so the the compos com positi ition on of vit vitreo reous us humor humor itself itself and the preana preanalyt lytica icall handli handling. ng. Development of a calibrated method for analysis would be one of the tasks for the future. A further review of a sample of 492 cases consisting of 170 of their own cases analyzing vitreous potassium, 176 cases from Muñoz et al. (2001) analyzing vitreous potassium, and 198 cases in which vitreous hypo hy poxan xanth thin inee was was analy analyze zed, d, was ca carr rrie ied d out out by Ma Made dea a an and d Rö Rödi dig g (2006).. The aims of the study were firstly to examine the finding by (2006) Muñoz and colleagues that more accuracy was gained in estimating the time of death by using potassium concentration as the independent vari va riab ablle an and d the the PMI PMI as the the depe depend nden entt vari variab able le and and seco second ndly ly to examine if in fact greater accuracy in estimating the PMI was possible by using the loess curve as proposed by Lange et al. (1994) (1994).. Madea and an d Rödi Rödig g est estimat imated ed the the PMI PMI up to 13 133 3 hour hourss an and d conf confir irme med d a greater, but only marginally better accuracy, from 6 25.96 to 6 23.27 hours, in estimating the PMI when potassium concentration was used as the independent variable and the PMI as the dependent variable. Madea Mad ea an and d Rödi Rödig g s eval evalua uati tion on of esti estima mati ting ng th thee PMI PMI by us usin ing g the the (1994),, revealed a systemloess curve, as recommended by Lange by Lange et al. (1994) atic overestimation of the TSD, suggesting the use of a loess curve to estimate the PMI could not be sustained. ’
Finally, an important review by Madea and Musshoff (2007) outline lined d th thee di diff ffic icul ulti ties es an and d in inacc accur uraci acies es in th thee us usee of vi vitr treo eous us hu humo morr potass pot assium ium concen concentra tratio tion n analys analysis. is. The chemic chemical al analys analysis is of vitreo vitreous us fluid differed from that of the analysis of blood, serum, or urine in that th at st stan anda dard rd pr proc oced edur ures es of anal analysi ysiss and and cali calibr brat atio ion n in th thes esee la latt tter er flui fluids ds had had be been en re read adil ily y atta attain ined ed in life life an and d co coul uld d ther theref efor oree be co commpared in postmortem samples. However, the normal value in life of
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
71
vitreous potassium concentration could not be determined and if, and how fast, abnormal serum potassium concentrations equilibrated with vitreous potassium concentrations in life could not be ascertained. It was also not known if potassium concentration was stable in postmortem vitreous humor. In all studies to that date, correlations between antemo ant emorte rtem m serum serum and postmo postmorte rtem m vitreo vitreous us val values ues wer weree comple completel tely y absent. The concentrations of vitreous potassium determined by ionspecific electrodes tended to be higher than those determined by flame photometry. For all these reasons, the term normal vitreous value wass a misn wa misnom omer er and and in inst stea ead d the the term term ref refere erence nce value value shoul should d be used. Madea and Musshoff defined reference values as a set of values of a certain type of quantity available from a single individual or group of individuals corresponding to a stated description which must be defined and an d av avai aila labl blee if ot othe hers rs were were to us usee the the re refe fere renc ncee valu values es.. They They furt furthe herr defined reference individuals as comprising a reference population from which was selected a reference sample group on which were determined “
“
”
”
reference values and onreference which was observed a reference from which were calculated limits that might definedistribution reference intervals. The reference values should be specifically stated for the reference population, as well as the environmental and physical conditions, specimen collec collectio tion, n, transp transport ort,, prepar preparati ation, on, storag storage, e, and analyt analytica icall method method used. No study in the literature on reference values to that date fulfilled all these requirements. Another problem with the use of vitreous potassium analysis was that th at of the the de deve velo lopm pmen entt of a cali calibr brat ated ed an and d va vali lida date ted d meth method od fo forr vitr vitreo eous us hu humo morr pota potass ssiu ium m anal analys ysis is.. Mo Most st in inst stru rume ment ntss us used ed were were calibrated for serum but it was questionable whether these instruments could be used for vitreous analysis as well. Methods of preanalytical collection and handling also required to be standardized. Developing a calibrated and validated method for vitreous humor analysis would be one of the tasks for the future.
THE ESTIMATION OF HYPOXANTHINE IN VITREOUS HUMOR Anothe Anot herr meth method od of an anal alyz yzin ing g vitr vitreo eous us humo humorr as a mean meanss of de dete terrmining the PMI was investigated by Rognum and colleagues (1991), (1991), who studied the levels of hypoxanthine in vitreous humor. Hypoxanthine (6-hydroxypurine) is a naturally occurring purine derivative and a deaminated form of adenine, itself a breakdown product of
72
Human Body Decomposition
adenosine monophosphate (AMP). Hypoxanthine exists as an intermediate in the biodegradation of AMP. AMP is converted to xanthine and then uric acid by the enzyme xanthine oxidase before it is excreted as urate (Harkne Harkness, ss, 1988 1988). ). In Incr crea ease sed d le leve vels ls of hypo hypoxa xant nthi hine ne in the the plasma of the umbilical cord of newborn infants suffering from intrauter ut erin inee hypo hypoxi xia a comp compar ared ed with with the the umbi umbili lica call cord cord pl plas asma ma leve levels ls of hypoxanthine from normal newborn infants had first been reported by Saugstad (1975) (1975),, who attributed the increased levels to hypoxia which caused an increase in AMP break down and consequently an increase in hypoxan hypoxanthi thine ne con concen centra tratio tion. n. The hypoxan hypoxanthi thine ne con concen centra tratio tion n was also thought to increase due to decreased transformation of hypoxanthine to uric acid by the inhibition of the enzyme xanthine oxidase. Sau Saugst gstad ad and Ola Olaise isen n (19 (1978) 78) invest investiga igated ted hypoxan hypoxanthi thine ne levels levels in the vitr vitreo eous us humo humorr of 86 in indi divi vidu dual alss betw betwee een n 0. 0.5 5 and and 19 192 2 hour hourss af afte terr death to determine whether these levels reflected the degree of tissue hypoxia preceding death. Cases were divided into five groups depending onon theorcause of sion, death; severe death, cases ofalstrangulati gul ation suspen suspensio n, death dea th from frotrauma m myocar myocausing cardia diall infarc inf arctio tion, n, fat fatal drug drug intoxication, and the last group included cases of bronchopneumonia, gastric hemorrhage, brain catastrophes, drowning, and carbon monoxide poisoning. There was no statistical difference in hypoxanthine concent ce ntra rati tion on in any any of the the grou groups ps exce except pt in th thee grou group p dy dyin ing g of fa fata tall intoxication, in which the concentration was significantly increased. As the normal concentration of hypoxanthine in the vitreous humor in life was not known, the authors considered the values in cases of sudden death to reflect the concentration during life. They therefore took the mean mea n conce concentr ntrati ation on found found in ca cases ses dying dying from from sev severe ere traum trauma a and myocardial infarction as reflecting the normal concentration before death and considered this concentration as the reference value. It was betw be twee een n 0 and and 540 540 µ mo mol/ l/L L and and was was 10 1020 ti time mess hi high gher er th than an the the plasma concentration. Although in the cases used for the reference values valu es (trauma (trauma and myocardi myocardial al inf infarct arction) ion) there was no sign signific ificant ant difference in hypoxanthine concentration in samples obtained more than 48 hours post mortem compared with samples obtained between 0 and 48 hours; in the total samples from all groups tested, a positive correlation was found between hypoxanthine concentration and the PMI, the concentration rising in a linear manner with the passage of time when it exceeded 48 hours. Compared with the reference group
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
73
there was no increased concentration of hypoxanthine in cases dying with tissue hypoxia such as suspension or strangulation, presumably because the period of hypoxia was short but there was a significant elevat ele vation ion in ca cases ses dying dying of drug drug in into toxic xicati ation, on, which which Saugs Saugstad tad and Olais Ola isen en attrib attribute uted d to prolo prolonge nged d tissue tissue hypoxi hypoxia a before before death death.. They They concluded that the hypoxanthine concentration in the vitreous humor post mortem might give information about whether or not tissue hypoxia preceded circulatory arrest. The next important study was by Rog Rognum num and col colleag leagues ues (199 (1991) 1) who wh o st stud udie ied d the the in infl flue uenc ncee of post postmo mort rtem em ti time me and te temp mper erat atur uree on the hypoxanthine concentration in vitreous humor and compared it to the potassium concentration in order to determine if this chemical substance was less prone to error than the potassium concentration estimation. In most cases four samples of vitreous humor, twice from each eye, ey e, we were re take taken n from from 87 su subj bjec ects ts wi with thin in 120 120 hour hourss af afte terr de deat ath. h. The exact time of death was known in all cases. Causes of death varied, but were mainly from myocardial infarctions and accidents. The bodies were kept at various temperatures: 5 C (33 subjects), 10 C (16 sub jects), jects ), 15 C (16 su sub bjects cts), and and 23 C (23 (23 subj subjec ects ts). ). Hypox Hypoxant anthin hinee concentration was determined by liquid chromatography and potassium by flame photometry. The results were graphed with chemical concent con centrat ration ionss on the x-a -axi xiss and and PMI PMI in hour hourss on th thee y-a -axi xis. s. In 19 subjects in which the vitreous was extracted within 1.5 hours of death, regression analyses were performed to determine the intercept and therefore the normal levels at the time of death. These values were we re esti estima mate ted d at 5.8 5.8 mm mmol ol/L /L for for pota potass ssiu ium m an and d 7. 7.6 6 µ mo mol/ l/L L for for hypoxanthine. It was found that in all subjects the increase in hypoxanthine and potassium concentrations correlated with time, but only after 48 hours post mortem and that the longer the time interval after death, up to 192 hours, the greater the correlation. The higher the ambi am bien entt temp temper erat atur ure, e, the the grea greate terr we were re the the hypo hypoxa xant nthi hine ne and and the the potassium concentrations at a similar time, that is, the steeper were the slopes of the regression lines. For the group as a whole the potassium siu m and hypoxan hypoxanthi thine ne concent concentrat ration ionss wer weree signif significan icantly tly correla correlated ted (r 5 0.93, p , 0.001). During the first hours after death the scatter of concentration values about the regression line was less with hypoxanthine than with potassium which the authors suggested might be due
74
Human Body Decomposition
to the influence of raised urea and possibly alcohol levels affecting the vitreo vit reous us potassi potassium um concent concentrat ration ion.. Hypoxan Hypoxanthi thine ne concent concentrat ration ion in vitreous humor had been previously shown to be influenced by antemortem hypoxia (Madea (Madea et al., 1994; Saugstad, 1975) 1975) and the authors suggested that this might also be the case with potassium. Therefore the method should only be used in estimating the PMI in cases of sudden death death witho without ut antem antemort ortem em hypo hypoxia xia.. Rogn Rognum um and and co coll lleag eague uess stressed that this was only a preliminary study on a small number of cases and that further studies were desirable to determine whether the esti estima mati tion on of hypo hypoxa xant nthi hine ne in th thee vi vitr treo eous us hu humo morr wa wass us usef eful ul in estimating estimat ing the PMI. A different finding was made by Madea by Madea and colleagues (1994) (1994) when when they studied postmortem hypoxanthine and potassium concentrations in the vitreous humor of 92 bodies with a known TSD, in order to evaluate statistical parameters for determining more precisely, the PMI estimation for vitreous hypoxanthine in comparison to vitreous potassium. The vitreous humor was completely removed from each eye at the same time. In an add additional 43 bodies, vitreous humor was removed from both eyes at timed intervals between 2 and 20 hours. These cases had all died either from sudden natural causes or trauma with wi th brie brieff term termin inal al epis episod odes es.. Anal Analysi ysiss wa wass carri carried ed out out by the the same same Rognum num and col collea league guess (19 (1991) 91) had used. It was conmethod met hodss that that Rog firmed that the hypoxanthine concentration rose in a linear fashion as thee PMI th PMI incr increa ease sed, d, the the ri rise se begi beginn nnin ing g imme immedi diat atel ely y post post mort mortem em,, in contrast to the findings of of Saugstad and Olaisen (1978) who noted an interval of 48 4872 hours before the concentration rose. The potassium concentration also rose, but the potassium concentration had a much strong str onger er correl correlati ation on with with the PMI than than the hypoxa hypoxanth nthine ine concen concentra tra-tion (r 5 0.9 0.925 25 for pot potass assium ium compar compared ed wit with h r 5 0.7 0.714 14 for hypoxa hypoxannthine). The 95% confidence limits were 6 17 hours for potassium and 6 32 hou ours rs fo forr hypo ypoxant xanthi hine ne up to 12 120 0 hour hourss po post st mo mort rtem em.. The The greater scatter of potassium concentration levels compared with those of hypoxanthine was not confirmed by this study. In the cases of timed bi bila late tera rall wi with thdr draw awal al of vitr vitreo eous us humo humor, r, ther theree were were grea greate terr in inte terrindividual differences between the rises in the concentration of hypoxanth xan thin inee than than wi with th po pota tass ssiu ium. m. In comm commen enti ting ng on th thee di diff ffer eren ence cess in findings between their study and that of Rognum Rognum et al. (1991) (1991),, Madea and colleagues suggested that one reason might be the repeated taking
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
75
of small samples, which disturbed the concentration gradient between the retinal cells and the center of the globe, giving a false value when cons co nsid ider eriing the the to tota tall conc concen enttrati ration on:: a true true valu valuee coul could d onl nly y be obta ob tain ined ed by samp sampli ling ng the the whol wholee vitr vitreo eous us hu humo mor. r. A se seco cond nd re reas ason on might be the fact that hypoxanthine increased in postmortem vitreous due to irreversible circulatory arrest, anoxemia, and diffusion, but that duri du ring ng vita vitall hypox hypoxia ia ther theree was an accel acceler erat ated ed ca cata tabo boli lism sm of AMP AMP to hypo hy poxan xanth thin inee which which ac accu cumu mula late ted d in body body ti tiss ssue uess and fl flui uids ds.. A fi fina nall reason could be due to the differing metabolism of potassium and hypoxanthin xant hinee at death. death. Potassi Potassium um had differ different ent con concen centra tratio tions ns in diff differen erentt intra- and extracellular body compartments during life but with loss of selective membrane permeability at death it would diffuse along a concentration gradient, for example, in the globe of the eye from the retina into the center of the vitreous. Potassium concentration was regulated during life in narrow ranges. In contrast hypoxanthine, a degradation prod pr oduc uctt of aden adenos osin inee nucl nucleo eoti tide de degr degrad adati ation on,, was form formed ed by se seve vera rall enzymat enz ymatic reactio reactions ns the before befcenter ore diffusin diff along along the concen concentrat tration ion gra gradien dient from theicretina into ofusing thegvitreous. Theoretically, therefore, at substance (potassium) whose post mortem increase was slow due to diffusion, would have a stronger correlation with the TSD than a substance (hypoxanthine) whose post mortem increase was due to degradation as well as diffusion. Madea and colleagues concluded that the rise in vitreous potassium concentration after death was a more reliable guide to PMI estima estimatio tion n than than the vitreou vitreouss hypoxa hypoxanth nthine ine concen concentrat tration ion which which could be affected by antemortem hypoxia. Finally, M Mad adea ea (2 (200 005) 5) comm commen ente ted d th that at the the stronger linear corre co rrelat latio ion n of hypox hypoxant anthin hinee conce concent ntra ratio tions ns com compa pared red with with potass potassium ium concentrations, with the PMI, in vitreous humor, as claimed by Rognum et al. (1991) had had not not been been susta sustaine ined d by furth further er stud studies ies and and that that de death ath time estimation was more precise using vitreous potassium concentration than tha n usi using ng hypox hypoxant anthin hinee con concen centr trati ation on (Made Madea a et al al., ., 19 1994 94;; Mu Muño ñozz Barús et al., 2002). 2002). More recent studies have also confirmed this finding (Madea and Rödig, 2006; Abdel Salam et al., 2012). 2012).
CHEMICAL INVESTIGATION OF SYNOVIAL FLUID Synovial fluid in a closed joint compartment would seem to offer an alt altern ernati ative ve sit site e to vitreo vities reous usamin hum humor org for chemic mical ana is, but the there re have ha ve be been en few few stud studie s exam ex inin ing this thische fl flui uid d al as analys anlysis, al alte tern rnat ativ ive e to
76
Human Body Decomposition
estimating the time of death. In one study by Madea and colleagues (2001) the the conce concentrat ntrations ions of sodiu sodium, m, potassium potassium,, calci calcium, um, chlo chloride, ride, urea ur ea,, crea creati tini nine ne,, an and d gluc glucos osee in the the supr suprap apat atel ella larr pouc pouch h sy syno novi vial al fluid flu id wer weree compar compared ed with with the sam samee concen concentra tratio tions ns in the vit vitreo reous us humor of 74 cases of sudden death, with a PMI ranging from 6 to 126 hours, which had no joint disease (defined as rheumatism and arthritis ), no met metabo abolic lic dis disord orders ers,, and receiv received ed no re resus suscit citati ation on with intravenous fluids prior to death. Only glucose and potassium were useful with regard to estimation of the PMI but the authors concluded that synovial fluid was a useful alternative fluid to estimate ma te post postmo mort rtem em pota potass ssiu ium m co conc ncen entr trat atio ion n if the the vi vitr treo eous us hu humo morr (2007) of suprapatellar was not available. A second study by Sheikh (2007) synovial fluid from 123 cadavers using potassium analysis by flame photometry in which the PMI was known, confirmed the linear relationship of potassium concentration with the PMI up to 48 hours after death. Sheikh concluded that it was a useful alternative to vit “
“
”
”
reous in the because analysis of of the potassium although more difficult to handlehumor and analyze higher viscosity. Tumram collea league guess (201 (2011) 1) analy analyze zed d both both the syn synovi ovial al fl fluid uid Tumram and col and an d vit vitreou eous humor mor in 154 154 cas cases wher wheree the time ime of death ath wa wass known, up to 35 hours post mortem. Cases of metabolic disease and injury to the knees were excluded and analysis of sodium, potassium, chloride, calcium, creatinine, glucose, and urea was carried out by capill cap illary ary zone zone elect electrop rophor horesi esis. s. The There re was no corre correlat lation ion betwe between en these the se chemic chemical al subst substanc ances es and the the PMI except except for po potas tassi sium. um. The auth au thor orss co conf nfir irme med d that that po pota tass ssiu ium m co conc ncen entr trat atio ion n rose rose in a li line near ar fashio fas hion n aft after er death death in both both synovi synovial al fluid fluid an and d vi vitr treou eouss hum humor or bu butt thatt th tha thee st stren rength gth of the the relat relation ionshi ship p betwee between n potass potassiu ium m conce concent ntraration and PMI was greater with synovial fluid ( r 5 0.7873) than with vitreous humor (r 5 0.527). Finally a study by Siddhamsetty by Siddhamsetty and colleagues (2014), (2014), examining the synovial fluid from the suprapatellar pouch of 210 cases post mortem, confirmed the linear relationship between the rise in potassium concentration and the PMI in synovial fluid. Cases were separated into 12-hour periods of the time of death from 0 hours up to 72 hours after death. The authors found no relations tio nship hip betwe between en PMI PMI and and sodi sodium, um, chlori chloride, de, calciu calcium, m, and gl gluco ucose se concentrations. There was a positive correlation between increasing synovial fluid potassium concentration and the PMI up to 72 hours (r 5 0.840).
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
77
BIOMARKERS OF THE PMI IN THE BONE MARROW AND OTHER BODY ORGANS Porteous (1961) studied (1961) studied the viability of bone marrow cells in 50 cadavers at varying times after death in order to determine whether cadaver bone marrow could be used for transplanting. He found that motility in cells was maintained for 20 hours or more death refrigerated subjects, but the motility time was shortened if after pyrexia hadinoccurred before death. He did not, however, link this method to estimation of the PMI. In an attempt to establish a basis for correlating antemortem and postmortem bone marrow findings in patients with hematological disease, Hoffman and colleagues (1964) studied the bone marrow aspirated from 20 sub jects,, rangin jects ranging g from 30 minut minutes es to 15 hours hours after death and who had died from a variety of diseases, and from trauma, in order to establish quantitative chronological trends in the counts of the various cell types. The authors found, in common with other investigators, that mature cellular types were usually indistinguishable 10 1019 hours after death. They were unable to demonstrate a significant difference in the rate of cellular autolys lysis bas ased ed on age, age, se sex, x, or mod mode of death eath.. Des Despi pite te the hesse fi find ndin ings gs,, Hoffman and colleagues suggested that it was conceivable that the study of postmortem bone marrow could be perfected as an ancillary medicolegal method for establishing the time of death. Penttilä death. Penttilä and Laiho (1981) examined the functionality and morphology of the different cellular components of blood in 123 cadavers kept at 4 C from 1.7 to 270.4 hours after death. They concluded that for a short period after death certain cellular properties could be applied to the postmortem diagnosis of diseases but were of minor use for the estimation of the TSD. Twenty years later a study by Dokgöz and colleagues (2001) on the morphology of white blood cells after death reached similar conclusions that the speed of the changes were so variable as to make it difficult to estimate the TSD. These were also the findings of Bardale Bardale and Dixit (2007) (2007).. A second study by Penttilä by Penttilä and Laiho (1981) examined (1981) examined the viability of a number of blood and other organ cells after death in the same 123 cadavers used in their other study, after staining the cells with trypan blue in a buffer solution (Laiho and Penttilä, 1981). 1981). Cells from the spleen, lymph nodes, lung, and bone marrow as well as white blood cells and spermatozoa were examined. The loss of viability of white blood cells showed a moderate correlation with the postmortem period up to about 270 hours in cadavers kept at 1
4 C butmethod the great variabilitythe between uncertain of estimating TSD. individual results made this an
78
Human Body Decomposition
Vas Vasss and col collea league guess (20 (2002) 02) studi studied ed the the amin amino o acid acidss prod produc uced ed by autolysis of protein in the heart, kidneys, liver, brain, and muscle of 18 cadavers allowed to decompose naturally over a 4-year period in an attempt to identify biomarkers that might be useful in determining the PMI. PM I. The The auth author orss ackn acknow owle ledg dged ed that that th thee most most impo import rtan antt vari variab able less affecting putrefaction were temperature and, to a lesser extent, humidity. The time of death, the air temperature record after death, and the cause of death were known and chemical analysis of the tissue from various organs was carried out at autopsy. The PMI was measured in cumulative degree hours, defined as the average temperature in degrees Cent Ce ntig igra rade de cumu cumula lati tivel vely y adde added d for for ea each ch 12-ho 12-hour ur peri period od.. Alth Althou ough gh there was a large variability among subjects there were consistent patterns of biomarker relationships that appeared at different time intervalss an val and d whic which h pers persis iste ted d acro across ss th thee enti entire re se sett of su subj bjec ects ts.. Vass Vass an and d colleagues used these patterns as the basis for determining the PMI, by prod pr oduc ucin ing g flow flow ch chart artss wh whic ich h outl outlin ined ed th thee appe appear aran ance ce of the the vario various us amino acids was at specific timeorgan-specific, intervals after adeath. Thepattern disadvantage of this method that it was different of amino acid concentration being applicable for different organs. However, if more mo re orga organs ns were were anal analyz yzed ed,, the the accu accura racy cy of PM PMII es esti tima mati tion on was was increased. Two furth Two urther er stud studie iess whic which h have have not bee een n fol followe lowed d by furth urther er research but which showed some promise in estimating the PMI are worthy of mention. The first by Gos by Gos and Raszeja (1993) examined the post po stmo mort rtem em ac acti tivi vity ty of th thee enzy enzyme mess la lact ctat atee (LDH (LDH)) an and d ma mala late te dehydr deh ydroge ogenas nasee (MDH) (MDH) in hum human an liver. liver. Twenty Twenty-fi -five ve cadave cadavers rs whi which ch had ha d di died ed from from nont nontox oxic ic caus causes es were were stud studie ied. d. Samp Sample less of li live verr were were removed from the cadavers within 12 hours of death, each divided into five parts and stored at different temperatures and examined every day for 14 days then every 7 days for 35 days. The activity of LDH and MDH was examined in each sample and the authors found a good correlation between declining activity of both enzymes with increasing PMI when when mathem mathemati atical cally ly express expressed ed by lin linear ear regres regressio sion n equ equati ations ons.. Gos and Raszeja recommended further follow-up studies. The second study by Myo Th Thai aikk-Oo Oo an and d co coll llea eagu gues es (2 (200 002) 2) exami ex amine ned d the the conc concen entr trat atio ion n of vasc vascul ular ar endo endoth thel elia iall grow growth th fa fact ctor or (VEG (V EGF) F),, a glyc glycop opro rote tein in with with pote potent nt angi angiog ogen enic ic,, mito mitoge geni nic, c, and and vascula vasc ularr permea permeabil bility ity-en -enhan hancin cing g act activi ivitie ties, s, in the brain, brain, lungs, lungs, heart, heart,
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
79
live liver, r, an and d ki kidn dneys eys from from 19 cada cadave vers rs in or orde derr to de dete term rmin inee whet whethe herr there was a correlation with the PMI. The PMI ranged from 1 to 120 hours and the authors found that the VEGF concentration increased linearly with the PMI up to 20 hours in the lungs and kidneys, up to 15 hours in the liver, and in the brain it started to increase after 24 hours, continued up to 40 hours after which time it began to fall. In thee he th hear artt th ther eree wa wass no clea clearr co corr rrel elat atio ion n betw betwee een n the the PM PMII and and the the VEGF level. Myo Thaik-Oo and colleagues concluded that the estimation of VEGF from certain organs might be useful in the estimation of thee TSD th TSD in the the earl early y po post stmo mort rtem em pe peri riod od bu butt th that at addi additi tion onal al stud studie iess were required. required.
IMMUNOHISTOCHEMISTRY From 1999 From 1999 to 2001 2001 We Wehn hner er an and d coll colleag eague uess pu publ blis ishe hed d th thre reee pape papers rs desc de scri ribi bing ng the the im immu muno nolo logi gical cal de dete tect ctio ion n of in insu suli lin n from from panc pancre reat atic ic β-cells (Wehner 1999), ), thyroglobulin from thyroid follicular cells (Wehner et al., 1999 (Wehner et al., 2000), 2000), and calcitonin from the c-cells of the thyroid er et al al., ., 200 2001 1). Th Thee au auth thor orss us used ed anim animal al anti antibo bodi dies es to glan gland d (Wehn Wehner dete de tect ct the the anti antige gens ns in insu suli lin, n, thyr thyrog oglo lobu buli lin, n, and and ca calc lcit iton onin in in thre threee si simi mila larr st stud udie ies, s, in whic which h over over 100 100 ca cada dave vers rs were were in incl clud uded ed in each each study. The time of death was known in every cadaver. The principle behind the study was that as the PMI increases the antigens, which are proteins, become denatured and can no longer be detected by staining methods. Detection was made up to 45 days post mortem with insulin and an d up to 21 da days ys wi with th th thyr yrog oglo lobu buli lin n and and ca calc lcit iton onin in.. Insu Insuli lin n was was detected in all cases with a positive immunological reaction up to 12 days post mortem, an equivocal reaction between 13 and 29 days, and a negative reaction after 30 days. With thyroglobulin the figures were: all cases positive up to 5 days, equivocal between 6 and 12 days, and negative after 13 days. With calcitonin, all cases were positive up to 4 days, equivocal between 5 and 12 days, and negative after 13 days. The authors inferred from these studies that detection, or lack of detection, of these antigens may be of use in estimating the PMI within the stated time intervals.
(2007) investigated the PMI in beagle dogs. Erlandsson and Munro (2007) Ten Te n do dogs gs,, grou groupe ped d in pair pairss for for si simi mila larr we weigh ightt an and d se sex, x, were were euth euthaanase na sed, d, kept kept in si simi mila larr en envi viro ronm nmen enta tall cond condit itio ions ns,, and and auto autops psie ied d at intervals of 24, 48, and 72 hours, 7 days, and 23 days. Rectal
80
Human Body Decomposition
temp temper erat atur ures es were were re reco cord rded ed as well well as path pathol olog ogic ical al and and hi hist stol olog ogic ical al changes in the lungs, heart, liver, urinary bladder, thyroid, pancreas, and tonsils. Immunohistochemistry of the cervical and tracheal lymph nodes was carried out, specifically staining the B and T lymphocytes. The authors found that after 3 days and up to 23 days there was a grad gr adua uall de decr crea ease se in B lymp lympho hocy cyte te stai staini ning ng as the the PMI PMI in incr crea ease sed, d, whereas the T lymph mphocytes continued to stain strongly. It was suggested that further research might lead to immunological tools for the estimation of the time of death. Immunochemistry might therefore provide a method of estimating the TSD in the later stages of decomposition. Of interest also in this study was that over the first 10 hours post mortem, rectal temperature was useful in determining the PMI, within a range of 6 2 hours and the authors determined that it was possible to develop a table of histological changes in the heart, liver, lungs, pancreas, and tonsils that might indicate the TSD.
HIGH-RESOLUTION PROTON MAGNETIC RESONANCE SPECTROSCOPY A st stud udy y in 19 1988 88 sh show owed ed that that th thee quan quanti tita tati tive ve es esti tima mati tion on of brai brain n metabol meta bolite itess by hig high-r h-reso esolut lution ion proton proton nuc nuclea learr magneti magneticc resonanc resonancee spectroscopy in rabbit brains might be useful to identify certain metabolic bo lic subs substan tance cess which which becam becamee elev elevate ated d in the the post postmo mort rtem em period period,, while N -acetyl - acetyl aspartate and the total creatine pool, which did not beco be come me elev elevat ated ed,, migh mightt be used used as re refe fere renc ncee mata matabo boli lite tess (Petroff et al., 1988). 1988). Fineschi Fineschi and col colleag leagues ues (199 (1990) 0) confirmed the concept behind this principle in their study on rat skeletal muscle. They concluded that careful consideration of the relative peak intensities of all thee de th detec tectab table le NMR re reso sona nance ncess su such ch as lacta lactate te,, hi hist stid idin ine, e, amino amino acids, and adenine nucleotide, was very likely to provide a good estimate of the postmortem time. Ith and col colleag leagues ues (200 (2002a) 2a) extended thee study th study scann scannin ing g eight eight sheep sheep head headss an and d four four post postmor morte tem m human human heads to determine the nature and quantity of metabolites occurring up to 2 23 weeks post mortem. After 3 days new metabolites, not present in life, and which included free trimethylammonium, propionate, butyrate, and iso-butyrate, began to appear and to increase in quan qu antit tity. y. They They co concl nclud uded ed that that fu furt rther her stud study y of these these meta metabo bolit lites es might be of use as a means of quantitatively determining the PMI. In a fu furt rthe herr st stud udy y of a numb number er of meta metabo boli lite tess in she heep ep and and hu huma man n brains brai ns by It Ith h and col collea leagues gues (200 (2002b) 2b) it was found that changes in ’
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
81
concentration detected by 1 H-MR spectroscopy varied with the passagee of time sag time.. Incr Increas eased ed conc concen entra trati tion on of some some follo followe wed d a si sigm gmoid oid curve, others were unpredictable but those in which the increase in concentration followed a linear course such as propionate, alanine, acetate, and γ-amino butyric acid (GABA), could be useful in predicting the PMI up to 250 hours. The mathematical models derived from sheep were successfully applied in four human bodies but the authors concluded that further studies were required to determine the influen inf luence ce of temp tempera erature ture.. This This method method rem remains ains in the experi experiment mental al (2005) in in a review of biochemical methods field at present but Madea but Madea (2005) of predicting the PMI commented that the method showed promise. The reasons were that it was a noninvasive method of quantitative measurement in which longitudinal studies with reproducible results were possible and that influencing factors such as temperature could be easily studied.
ELECTROLYTE CONCENTRATIONS IN CSF There have been few studies on the chemical analysis of CSF and even less in conjunction with the estimation of the PMI. Mason et al. (1951) estimated the rise in potassium concentration in 46 cadavers from 1.5 to 70 hours after death in order to determine if there was a correlation with the PMI. They found that if potassium concentration was plotted against the logarithm of time, a fair approximation to a straight line was produced but the error was too great for the TSD to be estimated. Madea et al. (1994) found that both the potassium and the hypoxanthine concentrations in CSF were so confounding that neither could be recommended for practical use in the estimation of the TSD. Wyler et al. (1994) published (1994) published a study in which they were able to correlate the increasing cell count in postmortem CSF with the PMI. They studied two groups of cadavers; 35 were maintained at 20 C, placed on their backs and lumbar punctures carried out at different intervals from 3 to 39 hours post mortem. A second group of 34 bodies were maintained at 4 C and lumbar punctures carried out at intervals from 3 to 53 hours post mortem. No cases with disease of the central nervous system were included and the time of death was known in all cases. In some cases CSF was obtained by cisternal puncture. Three groups of cells could be identified, lymphocytoid, monocytoid, and reticular cells, and the authors were able to compute mathematical models to estimate the TSD. They concluded that their first results
82
Human Body Decomposition
suggested a correlation between the cell count in postmortem CSF and the PMI but further studies would be necessary to investigate the influence of temperature, body position, and to establish confidence intervals of the estimations. Yadav and col Yadav collea league guess (20 (2007) 07) claim claimed ed to ha have ve foun found d a si sign gnif ific ican antt corr co rrel elat atio ion n betw betwee een n so sodi dium um and pota potass ssiu ium m conc concen entr trat atio ions ns and and the the sodium/potassium ratio in CSF up to 25 hours after death. One hundred dr ed ca cada dave vers rs,, in whic which h the the time time of de deat ath h wa wass kno nown wn,, had had CSF CSF extrac ext racted ted from from the lat latera erall ven ventri tricle cless after after openin opening g the sku skull, ll, wherea whereass previous researchers had obtained CSF by cisternal puncture, a blind proc pr oced edur ure, e, and and one one whic which h ri risk sked ed cont contam amin inat atio ion n by bl bloo ood d pr prod oduc ucts ts.. Up to 25 hours after death, a postmortem increase in potassium concentration and decrease in sodium concentration was found. The rise in potass potassium ium concen concentra tratio tion n showed showed a statis statistic ticall ally y signif significa icant nt cor correl relaation with the PMI. The fall in sodium concentration also correlated significantly with the PMI as did the sodium and potassium ion ratio, but to a lesser extent. The cause of death or environmental temperature did not significantly alter the results. The researchers concluded that th at th thee rati ratio o of sodi sodium um and and pota potass ssiu ium m co conc ncen entr trat atio ion n was was a be bett tter er predictor of PMI than the estimation of either electrolyte concentration alone. However, the research depended on obtaining CSF from the lateral ventricles of the brain at autopsy, and Yadav and colleagues gu es conc conclu lude ded d that that fu furt rthe herr stud studie iess were were re requ quir ired ed to dete determ rmin inee the the PMI in cases where sodium and potassium concentrations were abnormal at the time of death. Ther Th eree have have be been en few few stud studie iess us usin ing g CSF CSF co comp mpon onen ents ts to es esti tima mate te the TSD because cells lining the subarachnoid space in the brain autolyze rapidly and limit the use of CSF in analysis (Naumann, (Naumann, 1959). 1959).
ELECTROLYTE CONCENTRATIONS IN PERICARDIAL FLUID Bala soo ori riya ya an and d co coll lleeag ague uess (1 (198 984b 4b)) exami examine ned d co conc ncen enttra rati tion onss of Balaso potass pot assium ium,, sodium sodium,, calciu calcium, m, phos phospha phate, te, alkalin alkalinee phosph phosphata atase, se, alanine alanine aminotransferase, gammaglutamyl transferase, hydroxybutyric dehydrogena ge nase, se, glut glutam amic ic oxal oxaloa oace ceti ticc trans transam amin inas ase, e, cr crea eati tine ne phos phosph phoki okinas nase, e, bicarbonate, urea, creatinine, direct, indirect and total bilirubin, urate, iron, and cholesterol in the pericardial fluid obtained from 74 cadavers and in whom the time of death was known to within 15 minutes, in order
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
83
to ascertain whether there was any correlation between the concentrations of these substances and the PMI. The authors instanced only one previous study of postmortem pericardial fluid in 1976 and published in German (Brinkman (Brinkman et al., 1976). 1976). Although they found an increase in the concentration of some of the electrolytes with increasing PMI the changes were not sufficiently constant or predictable to allow an accurate estimate of the PMI. Balasooriya and colleagues commented that analysis of vitreous humor would provide a better indicator of the PMI. A further study Singh and colleagues (2006) (2006) on on 311 cadavers in which the pericardial by Singh by fluid was obtained between 2.5 and 58 hours after death was carried out. They found that there was a statistically significant, but not highly significant, correlation between the PMI and the rise in potassium and phosphate concentrations and the sodium/potassium ratio in pericardial fluid. Finally, a recent study by Kawamoto by Kawamoto and colleagues (2013) analyzed a broad spectrum of electrolytes and proteins in postmortem pericardial fluid from 288 cadavers dying from a number of causes and diseases. In all all ca case sess in there the re were were moder mod atee decr decrea ease sess in were sodi sodium um an and d chlor chlorid idecorree and increases potassium anderat magnesium which not sufficient to late with the TSD but the authors found characteristic alterations in certain tain su subs bsttance ancess with with cert certai ain n caus causes es of deat death h whic which h migh mightt be us usef efu ul biomarkers to investigate the cause and process of death and to reinforce pathological and toxicological findings.
THE USE OF DNA AND RNA IN THE ESTIMATION OF THE PMI Only a small number of papers dealing with research into deoxyribonucl nu clei eicc acid acid (DNA (DNA)) and and ri ribo bonu nucl clei eicc acid acid (RNA (RNA)) degr degrad adat atio ion n as a means of estimating the TSD have been published, and none of them has suggested its use as a quantitative method of estimating the PMI. The autodegradation of DNA in human rib bone was investigated by Perry and colleagues (1988) to determine whether it was useful in esti estima mati ting ng the TSD. TSD. It was was found ound that that vari variat atio ion n in te temp mper erat atur uree affected the degradation rate, that there was some evidence that the DNA degrad degradati ation on rate rate was simila similarr among among dif differ ferent ent indivi individua duals, ls, but that different bone from the same individual showed different rates of DNA degradation. DNA from bone marrow was degraded at a different rate from DNA in the bone cortex. The findings of Perry and colleagues indicated that DNA degradation was not useful in estimating the PMI.
84
Human Body Decomposition
Di Nunno and colleagues (1998) examined splenic cells in 35 cadavers from 24 to 126 hours after death to determine whether there was any an y corr correl elat atio ion n betw betwee een n DNA DNA degr degrad adat atio ion n and and th thee ti time me of deat death. h. Flow cytometry was used to evaluate DNA degradation in the cells. They found that although there was a precise correlation in the 24 24 36hour period after death and a less precise correlation from 36 to 72 hours, the method was of no use after this time as the spleen autolyzed rapidly. (2002) discussed the analysis of postmortem DNA Johnson and Ferris (2002) discussed degradation by single-cell gel electrophoresis as a possible method of estimati ma ting ng th thee PMI. PMI. When When cell cellss di die, e, nucl nuclea ease sess are are re rele leas ased ed that that degr degrad adee chrom chr omoso osomal mal DNA into into small smaller er fragme fragments nts.. Isola Isolati ting, ng, vi visua sualiz lizin ing, g, and measur mea surin ing g the fragm fragmen ents ts might might pr prov ovee a quan quantif tifiab iable le way of es estim timati ating ng the PMI. DNA fragmentation was studied in human leukocytes up to 22 hours after death and in pig tissues up to 72 hours. It was found that there th ere was was a proc proces esss whereb whereby y nucle nuclear ar DNA was fra fragme gment nted ed fol follo lowin wing g death, and this was organ- and time-dependent. Johnson and Ferris concluded clu ded that that altho although ugh the the result resultss show showed ed a se seque quenti ntial al,, time-d time-dep epend enden entt process and potential for use as a future method of estimating the PMI, further studies were needed to determine the accuracy and precision and to dete determ rmin inee the the effe effect ct of vary varyin ing g temp temper erat atur uree on the the frag fragme ment ntat atio ion n process. Bauer and colleagues (2003) studied (2003) studied postmortem RNA degradation, in order to determine whether quantification of the fragmentation of human mRNA might be useful as a method for the estimation of the PMI. They examined 50 postmortem blood samples and 36 brain tissue samp sue sample less wher wheree the the ex exac actt time time of deat death h was was know known. n. Fa Fatt tty y ac acid id synthase synthase messenger RNA (FASN (FASNmRNA) was chosen for analysis. Their results showed that RNA analysis from 50% of the postmortem bl bloo ood d samp sample less fail failed ed to prov provid idee cons consis iste tent nt re resu sult ltss du duee to adva advanc nced ed hemo he moly lysi sis. s. In th thee re rema main inin ing g sampl amples es,, FASN FASN mRNA A mole molecu cule less mRN decreased continuously up to a PMI of 120 hours and the correlation was significant (r 5 0.808). Analysis of brain samples showed a slow and continuous decrease in FASN FASNmRNA with time which was again significant (r 5 0.791), but the error range was too large to justify the use of this method in medicolegal situations. Several animal studies have shown good correlation between DNA fragmentation and the PMI (Liu (Liu et al., 2007; Chen et al., 2007; Mona
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
85
2013) and with RNA fragmentation and the et al., 2008; Gomaa et al., 2013) 2013). ). In contrast however a recent study PMI (Sampaio-Silva (Sampaio-Silva et al., 2013 of human rib bone samples taken within 24 hours post mortem and stored for varying periods of time from 103 to 445 days was carried out (Alaeddini (Alaeddini et al., 2010). 2010). The samples were stored under four differentt co en cond ndit itio ions ns:: free freezi zing ng,, am ambi bien entt temp temper erat atur ure, e, hi high gh humi humidi dity ty,, and and underground burial. DNA was extracted and examined for degradation. No correlation was found between DNA fragmentation and the PMI. At the present time the use of DNA and RNA for estimating the PMI remains in the experimental phase and is not of use in practice. In a recent review of the use of postmortem biochemistry, Madea (2005) commented (2005) commented at length on the study of Di Nunno and colleagues. He criticized the study for the fact that it took no account of changes in body or ambient temperature, whereas previous German studies had shown that DNA degradation was faster in warmer than colder surroundings (Liebhart, (Liebhart, 1969; Liebhart and Spann, 1973). 1973 ). Di Nunno and ’
colleagues study was also criticized because, as the authors had admitted, the proximity of the bowel to the spleen could have affected the degradation process by microbial migration during the autolytic process. Madea concluded that flow cytometric evaluation of DNA degradation was not yet a reliable or precise method of assessing the PMI. In gene genera rall re rega gard rdin ing g bi bioc oche hemi mica call meth method odss of es esti tima mati ting ng th thee TSD, TSD, Made Ma dea a st stat ated ed that that fo forr prac practi tica call pu purp rpos oses es ther theree was was no re real al brea breakkthro th roug ugh h be beca caus usee of th thee comp comple lex x unde underl rlyi ying ng meta metabo boli licc proc proces esse sess involved in putrefaction and the estimation of the PMI by biochemical meth me thod odss re remai maine ned d impr imprec ecis ise. e. He also also stat stated ed that that the the impo import rtan ance ce of chemical methods could change in future if the following criteria were fulfilled: quantitative measurement, mathematical description, taking into account influencing factors quantitatively, declaration of precision and proof of precision on independent material. “
”
CONCLUSIONS It ap appe pear ared ed fro from an ear early stag stagee that that wo work rk on bi bioc oche hemi mica call mar marke kerr conce con cent ntra rati tion onss in clos closed ed co comp mpar artm tmen ents ts,, prot protec ecte ted d for th thee most most par partt from the confounding effects of autolysis in surrounding tissues, would be the ideal place to focus research into the PMI. Indeed the most intensively studied biochemical marker in PMI research has focused on the conce con cent ntra rati tion on of potas potassi sium um in the the vit vitre reou ouss hu humo mor. r. Nons Nonsta tand ndar ardi dize zed d
86
Human Body Decomposition
methods metho ds,, iss issue uess with with study study desi design gn,, th thee va vari riab able le ef effe fect ctss of the the stat statee of samples (age, sex, ancestry, disease state) and the fact that the potassium concentration in living individuals (as well as its potential variability) is unkn un know own n have have all cont contri ribu bute ted d to the somew somewhat hat li limi mite ted d valu valuee of th this is means mea ns of esti estimat matin ing g the the TSD. TSD. Othe Otherr bi bioc oche hemi mica call mark marker erss wi with thin in the the vitreous humor of the eye, such as hypoxanthine for instance, have fared no bett better er.. Work Work on anot anothe herr clos closed ed comp compar artme tment nt si site te,, syno synovia viall fl flui uid d from joint compartments, although more difficult to extract and analyze, has also shown potassium concentrations to correlate with the PMI, which would indicate some degree of usefulness in estimating the TSD if vitreous fluid was unavailable for analysis. More generally, work on postmortem cell viability, enzyme activity, growth factor concentrations, immunological reactions as well as electrolyte concentrations in CSF and pericardial fluid have had for the most part positive results, but again not to the extent that they could be used in TSD estimation. Finally, work on quantifying DNA and RNA degradation post mortem has been proven to have no value in estimating the PMI.
REFERENCES Abdel Salam, H.F., et al., 2012. Estimation of post-mortem interval using thanatochemistry and post-mortem changes. Alexandria J. Med. 48, 335 335 344. Adelson, L., Sunshine, I., Rushforth, N.B., Mankoff, M., 1963. Vitreous potassium concentration as an indicator of post-mortem interval. J. Forensic Sci. 8 (4), 503 503514. Adjutanti Adjuta ntis, s, G., Cou Coutse tselin linis, is, A., 197 1972. 2. Est Estima imatio tion n of the tim timee of dea death th by pot potass assium ium levels levels in vitreous vitre ous humour. Forensic Sci. 1, 55 55 60. Alaedd Alae ddin ini, i, R. R.,, Wa Wals lsh, h, S. S.J. J.,, Ab Abba bas, s, A. A.,, 20 2010 10.. Mo Mole lecu cula larr st stud udie iess of ti time me an and d en envi viron ronme ment nt dependent effects on bone DNA survival. Aust. J. Forensic Sci. 42 (3), 211 211220. Balasooriya, B.A.W., St Hill, C.A., Williams, A.R., 1984a. The biochemistry of vitreous humour. A comparative study of the sodium, potassium and urate concentrations in the eyes at identical time intervals after death. Forensic Sci. Int. 26, 85 85 91. Balasoo Bala soori riya ya,, B. B.A. A.W. W.,, St Hi Hill ll,, C. C.A. A.,, Wi Will llia iams ms,, A. A.R. R.,, 19 1984 84b. b. Th Thee bi bioc oche hemi mica call ch chan ange gess in pericardia peric ardiall fluid after death. An inves investiga tigation tion of the relationship relationship between the time since death and the rise or fall in electrolyte and enzyme concentrations and their possible usefulness in determining the time of death. Forensic Sci. Int. 26, 93 93 102. Bardale, Bardal e, R., Dix Dixit, it, P. P.G., G., 200 2007. 7. Eva Evalua luatio tion n of mor morpho pholog logic ical al cha change ngess in blo blood od ce cells lls of hum human an cadaver for the estimation of post-mortem interval. Med-Leg. Update 7 (2), 35 39. Bauer, M., Gramlich, I., Polzin, S., Patzelt, D., 2003. Quantification of mRNA degradation as possible indicator of post-mortem interval interval a a pilot study. Leg. Med. 5, 220 220 227. Bocaz-Beneventi, G., et al., 2002. Capillary zone electrophoresis and artificial neural networks for estimation of the post mortem interval (PMI) using electrolytes measurements in human vitreous humour. Int. J. Legal Med. 116, 5 511. Brinkman, B., May, D., Riem Brinkman, Riemann, ann, U., 1976. Postmortaler Postmortaler temperaturaus temperaturausglei gleich ch im berei bereich ch des kopfes. Z. Rechtsmed. 78, 69 6982.
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
87
Chen, X., Yi, S., Liu, L., 2007. Image analysis for degradation of DNA in retinal nuclei of rat after death. J. Huazhong Univ. Sci. Technol. 27 (1), 24 2426. Coe, J. Coe, J.I. I.,, 19 1969 69.. Po Post st-m -mor orte tem m ch chem emis istr trie iess on vi vitr treo eous us hu humo mour ur.. Am Am.. J. Cl Clin in.. Pa Path thol ol.. 51 (6 (6), ), 741 741 750. Coe, J.I., 1989. Vitreous potassium as a measure of the post-mortem interval: an historical review and critical evaluation. Forensic Sci. Int. 42, 201 201 213. Coe, J.I., 1993. Post-mortem chemistry update. Am. J. Forensic Med. Pathol. 14 (2), 91 117. Di Nun Nunno, no, N.R., et al. al.,, 199 1998. 8. Is flo flow w cyt cytome ometri tricc eva evalua luatio tion n of DNA degradat degradation ion a reliab reliable le method met hod to inv invest estiga igate te the ear early ly pos post-m t-mort ortem em per period iod?. ?. Am. J. For Forens ensic ic Med Med.. Pat Pathol hol.. 19 (1), (1), 50 50 53. Dokgöz, H., Arican, N., Elmas, I., Fincanci, S.K., 2001. Comparison of morphological changes in white blood cells after death and in vitro storage of blood for the estimation of post-mortem interval. Forensic Sci. Int. 124, 25 25 31. Duke-Elder, Duke-Elde r, W.S. W.S.,, 1929. The physi physico-ch co-chemic emical al prope properties rties of the vitreous body. J. Physiol. 68, 155 155 165. Erlandsson, M., Munro, R., 2007. Estimation of the post-mortem interval in beagle dogs. Sci. Justice 47, 150 150154. Ferslew, K.F., Hagardorn, A.N., Travis Harrison, M., McCormick, W.F., 1998. Capillary ion analysis of potassium concentrations in human vitreous humour. Electrophoresis 19, 6 610. Fineschi, V., et al., 1990. 1 H-NMR studies of post-mortem biochemical changes in rat skeletal muscle. Forensic Sci. Int. 44, 225 225 236. Forbes Forb es,, S. S.L. L.,, 20 2008 08.. Fo Fore rens nsic ic ch chem emis istr try: y: ap appl plic icat atio ions ns to de deco comp mpos osit itio ion n an and d pr pres eser erva vati tion on.. In: Oxenham, M. (Ed.), Approaches to Death, Disaster and Abuse. Australian Academic Press, Brisbane, Ch. 15. Gomaa, Goma a, M. M.S. S.,, Ab Abd d ElEl- Kh Khal alek ek,, A. A.M. M.,, Sa Same meer er,, M. M.M. M.,, 20 2013 13.. Th Thee re rela lati tion onsh ship ip be betw twee een n post-mortem interval and the DNA degradation in brain and liver of adult albino rats. J. Am. Sci. 9 (5), 535 535540. Gos, T., Raszeja, S., 1993. Post-mortem activity of lactate and malate dehydrogenase in human liver in relation to time after death. Int. J. Legal Med. 106, 25 29. Hansson, L., Uoti Hansson, Uotila, la, U., Lindf Lindfors, ors, R., Laiho Laiho,, K., 1966. Potassium Potassium conte content nt of the vitreous body as an aid in determination of time of death. J. Forensic Sci. 11 (3), 390 390394. Harkness, Harkne ss, R.A R.A., ., 198 1988. 8. Hyp Hypoxa oxanth nthine ine,, xan xanthi thine ne and uri uridin dinee in bod body y flu fluids ids,, ind indica icator torss of ATP depletion. J. Chromatogr. 429, 255 255278. Henry, J.B., Smith, F.A., 1980. Estimation of the post-mortem interval by chemical means. Am J. Forensic Med. Pathol. 1 (4), 341 341347. Henssge, C., 1988. Death time estimation in case work. I. The rectal temperature time of death nomogram. Forensic Sci. Int. 38, 209 209 236. Hjertén, S., 1967. Free zone electrophoresis. Chromatogr. Rev. 9, 122 122 219. Hoffman, S.B., Morrow, G.W., Pease, G.L., Stroebel, C.F., 1964. Rate of cellular autolysis in post-mortem bone marrow. Am. J. Clin. Pathol. 41 (3), 281 281286. Hughes, W.M.H., 1965. Levels of potassium in the vitreous humour after death. Med. Sci. Law 5, 150 150156. Ith, M., et al., 2002a. Observation and identification of metabolites emerging during decomposition of brain tissue by means of in situ 1 H Magnetic Resonance Spectroscopy. Magn. Reson. Med. 48, 915 915920.
88
Human Body Decomposition
Ith, M., et al., 2002b. Estimation of the post-mortem interval by means of 1 H-MR spectroscopy. Proc. Int. Soc. Magn. Reson. Med. 10 (Suppl.), 228. Jaffe, F.A., 1962. Chemical post-mortem changes in the intraocular fluid. J. Forensic Sci. 7 (2), 231 231 237. Johnson, L.A., Ferris, J.A.J., 2002. Analysis of post-mortem DNA degradation by single-cell gel electrophoresis. Forensic Sci. Int. 126, 43 43 47. Kawamoto, Kawamo to, O., Mic Michiu hiue, e, T., Ish Ishika ikawa, wa, T., Mae Maeda, da, H., 201 2013. 3. Com Compre prehe hensi nsive ve eva evalua luatio tion n of pericardial biochemical markers in death investigation. Forensic Sci. Int. 224, 73 73 79. Laiho, K., Penttilä, A., 1981. Autolytic changes in blood cells and other tissue cells of human cadavers. I. Viability and ion studies. Forensic Sci. Int. 17, 109 109 120. Lange, N., Swe Lange, Sweare arer, r, S., Stu Sturne rner, r, W. W.Q., Q., 199 1994. 4. Human Human pos post-m t-mort ortem em int interv erval al est estima imatio tion n fro from m vitreous potassium: an analysis of original data from six different studies. Forensic Sci. Int. 66, 159 159 174. Leahy, M.S., Farber, E.R., 1967. Post-mortem chemistry of vitreous humour. J. Forensic Sci. 12 (2), 214 214222. Lie, J.T., 1967. Changes of the potassium concentration in the vitreous humour after death. Am. J. Med. Sci. 254 (2), 136 136143. Liebhart, E., 1969. Untersuchung Liebhart, Untersuchungen en über das quantitati quantitative ve Verhalten Verhalten der DNS in Abhängigke Abhängigkeit it zu der seit dem Tode verstrichenen Zeit, Habilitation, Freiburg, Germany. Liebhart, E., Spann Liebhart, Spann,, W., 1973. Cytophotometris Cytophotometrische che Unter Untersuchu suchungen ngen zum Leic Leichenal henalter. ter. Beitr Grichtl Med. 30, 277 277280. Liu, L., et al., 2007. Determination of the early time of death by computerised image analysis of DNA degradation: which is the best quantitative indicator of DNA degradation? J. Huazhong Univ. Sci. Technol. 27 (4), 362 362 366. Madea, B., 2005. Is there recent progress in the estimation of the post-mortem interval by means of thanatochemistry? Forensic Sci. Int. 151, 139 139 149. Madea, B., Henssge, C., 2002. Eye changes after death. In: Knight, B. (Ed.), The Estimation of the Time since Death in the Early Post-mortem Period, second ed. Arnold, London, p. 103, Ch. 4. Madea, B., Musshoff, F., 2007. Post-mortem biochemistry. Forensic Sci. Int. 165, 165 165 171. Madea, B., Rödig, A., 2006. Time of death dependent criteria in vitreous humour humour accuracy of estimating the time since death. Forensic Sci. Int. 164, 87 87 92. Madea, B., Henssge, C., Hönig, W., Gerbracht, A., 1989. References for determining the time of death by potassium in vitreous humour. Forensic Sci. Int. 40, 231 231243. Madea, B., Herrmann, N., Henssge, C., 1990. Precision of estimating the time since death by vitreous potassium potassium comparison comparison of two different equations. Forensic Sci. Int. 46, 277 284. Madea, B., Käferstein, Madea, Käferstein, H., Herm Hermann, ann, N., Stich Sticht, t, G., 1994. Hypoxanthine Hypoxanthine in vitre vitreous ous humour and cerebrospinal fluid fluid a marker of post-mortem interval and prolonged (vital) hypoxia? Remarks also on hypoxanthine in SIDS. Forensic Sci. Int. 65, 19 19 31. Madea, B., Kreuser, C., Banaschak, S., 2001. Post-mortem biochemical examination of synovial fluid fluid a preli preliminar minary y study. Forensic Forensic Sci. Int. 118, 29 29 35. a Mason, J.K., Klyne, W., Lennox, B., 1951. Potassium levels in the cerebrospinal fluid after death. J. Clin. Pathol. 4, 231 231233. Mona, A. El-H., et al., 2008. The relationship between post-mortem interval and DNA degradation in different tissue of drowned rats. Mansoura J. Forensic Med. Clin. Toxicol. 16 (2), 45 45 60. Muñoz, J.I., et al., 2001. A new perspective in the estimation of the post-mortem interval (PMI) based on vitreous [K1]. J. Forensic Sci. 46 (2), 209 209214.
Biochemica Bioch emicall Methods Methods of Estimating Estimating the Time Since Death
89
Muñoz Barús, J.I., et al., 2002. Improved estimation of post-mortem interval based on differential behaviour of vitreous potassium and hypoxanthine in death by hanging. Forensic Sci. Int. 125, 67 67 74. Naumann, H.N., 1959. Post-mortem chemistry of the vitreous body in man. Am. Med. Assoc. Arch. Ophthalmol. 62, 356 356363. Pentti Pent tilä lä,, A. A.,, La Laih iho, o, K. K.,, 19 1981 81.. Au Auto toly lyti ticc ch chan ange gess in bl bloo ood d ce cell llss of hu huma man n ca cada dave vers rs II II.. Morphological studies. Forensic Sci. Int. 17, 121 121 132. Perry, W.L., Bass, W.A., Riggsby, W.S., Sirotkin, K., 1988. The autodegradation of deoxyribonucle nuc leic ic aci acid d (DN (DNA) A) in hum human an rib bone and its rel relati ations onship hip to the time int interv erval al sin since ce dea death. th. J. Forensic Sci. 33 (1), 144 144 153. Petroff, Petrof f, O.A O.A.C. .C.,, Ogi Ogino, no, T., Alg Alger, er, J. J.R., R., 198 1988. 8. Hi High gh res resolu olutio tion n pro proton ton mag magnet netic ic res resona onanc ncee spectroscop spect roscopy y of rabbit brain: regio regional nal meta metabolit bolitee leve levels ls and post-m post-mortem ortem changes. J. Neuro Neurochem chem.. 51, 163 163171. Porteous, I.B., 1961. Persistence of motility in bone marrow cells from the cadaver. Nature 192 (4802), 569 569570. Pounde Poun der, r, D. D.J. J.,, Ca Carso rson, n, D. D.O. O.,, John Johnst ston on,, K. K.,, Or Orih ihar ara, a, Y. Y.,, 19 1998 98.. El Elec ectr trol olyt ytee co conc ncen entr trat atio ion n difference diffe rencess betwe between en left and right vitre vitreous ous humour sampl samples. es. J. Foren Forensic sic Sci. 43 (3), 604 604 607. Reddy, D.V.N., Kinsey, V.E., 1960. Composition of the vitreous humour in relationship to that of plasma and aqueous humours. Arch. Ophthalmol. 83, 715. Rognum, T.O., Hauge, S., Øyasaeter, S., Saugstad, O.D., 1991. A new biochemical method for estimation of post-mortem time. Forensic Sci. Int. 51, 139 139 146. Sampaio-S Sampai o-Silv ilva, a, F., et al. al.,, 201 2013. 3. Pro Profil filing ing of RNA deg degrad radati ation on for est estima imati tion on of pos post-m t-mort ortem em interval. PLoS One 8 (2), 1 18. Saugstad, O.D., 1975. Hypoxanthine as a measure of hypoxia. Paediatr. Res. 9, 158 161. Saugstad, O.D., Olaisen, B., 1978. Post-mortem hypoxanthine levels in the vitreous humour. An introductory report. Forensic Sci. Int. 12, 33 3336. Sheikh, N., 2007. Estimation of post-mortem interval according to time course of potassium ion activity acti vity in cada cadaveric veric synovial synovial fluid. Ind. J. Forensic Med. Toxicol. 1 (1), 7 712. Siddhamse Siddha msetty tty,, A.K A.K., ., et al. al.,, 201 2014. 4. Exp Explor loring ing tim timee of dea death th from from pot potass assium ium,, sod sodium ium,, chl chlori oride, de, glucose and calcium analysis of post-mortem synovial fluid in a semi-arid climate. J. Forensic Leg. Med. 28, 11 1114. Singh, D., Prashad, R., Sharma, S.K., Pandey, A.N., 2006. Estimation of post-mortem interval from human pericardial fluid electrolyte concentrations in Chandigarh zone of India: log transformed linear regression model. Leg. Med. 8, 279 279 287. Stephens, R.J., Richards, R.G., 1987. Vitreous humour chemistry: the use of potassium concentration trati on for the prediction prediction of the post-mortem post-mortem interval. interval. J. Forensic Sci. 32 (2), 503 503509. Sturner, W.Q., 1963. The vitreous humour: post-mortem potassium changes. Lancet 281 (7285), 807 807 808. Sturner, Sturne r, W. W.Q., Q., Gan Gantne tner, r, G.E G.E., ., 196 1964. 4. The post post-mo -morte rtem m int interv erval: al: a stu study dy of pot potass assium ium in the vitreous humour. Am. J. Clin. Pathol. 42 (2), 137 137144. Tagliaro, Taglia ro, F., et al. al.,, 199 1999. 9. Cap Capill illary ary zon zonee ele electr ctroph ophore oresis sis of pot potass assium ium in vit vitreo reous us hum humour our:: validation of a new method. J. Chromatogr. B 733, 273 273 279. Tagliaro, F., et al., 2001. Potassium concentration differences in the vitreous humour from the two eye eyess rev revisi isited ted by mic microa roanal nalysi ysiss wit with h cap capill illary ary ele electr ctroph ophore oresis sis.. J. Chr Chroma omatog togr. r. A 924 924,, 493 493 498. Tagliaro, F., Manetto, G., Crivellente, F., Smith, F.P., 1998. A brief introduction to capillary electrophoresis. Forensic Sci. Int. 92, 75 75 88.
90
Human Body Decomposition
Thaik-Oo, M., et al., 2002. Estimation of post-mortem interval from hypoxic inducible levels of Vascular Endothelial Growth Factor. J. Forensic Sci. 47 (1), 186 186 189. Tumram, N.K., Bardale, R.V., Dongre, A.P., 2011. Post-mortem analysis of synovial fluid and vitreous humour for determination of death interval: a comparative study. Forensic Sci. Int. 204, 186 186 190. Vass, A.A., et al., 1992. Time since death determinations in human cadavers using soil solution. J. Forensic Sci. 37 (5), 1236 12361253. Vass, A.A Vass, A.A., ., et al. al.,, 200 2002. 2. Dec Decomp omposi ositio tion n che chemis mistry try of hum human an rem remain ains: s: a new met method hodolo ology gy for determining the post-mortem interval. J. Forensic Sci. 47 (3), 542 542 553. Wehner, F., Wehner, H.-D., Schieffer, M.C., Subke, J., 1999. Delimitation of the time of death by imm immuno unohis histoc tochem hemica icall det detec ectio tion n of ins insuli ulin n in pan pancre creati aticc β cells cells.. Fo Fore rens nsic ic Sc Sci. i. In Int. t. 10 105, 5, 161 161 169. Wehner, F., Wehner, H.-D., Schieffer, M.C., Subke, J., 2000. Delimitation of the time of death by immunohistochemical detection of thyroglobulin. Forensic Sci. Int. 110, 199 199 206. Wehner,, F., We Wehner Wehne hner, r, H.H.-D. D.,, Sub Subke, ke, J., 200 2001. 1. De Delim limita itatio tion n of the tim timee of dea death th by imm immuno uno-histochemical detection of calcitonin. Forensic Sci. Int. 122, 89 89 94. Wyler, D., Mar Wyler, Marty, ty, W. W.,, Bär Bär,, W., 199 1994. 4. Cor Correl relati ation on bet betwee ween n the post post-mo -morte rtem m ce cell ll con conten tentt of cerebrospinal fluid and time of death. Int. J. Legal Med. 106, 194 194199. Yadav, J., et al., 2007. Estimation of time since death from C.S.F. electrolyte concentration in Bhopal region of central India. Leg. Med. 9, 309 309313.
CHAPTER
4
Research in the Later Stages of Decomposition
ESTIMATION OF THE POSTMORTEM INTERVAL DURING THE STAGE OF AUTOLYSIS AND EARLY PUTREFACTION EXTRINSIC FACTORS AFFECTING DECOMPOSITION: TEMPERATURE, ENVIRONMENT, INSECTS, BURIAL, MOISTURE, SCAVENGING DELAYED DECOMPOSITION AND PRESERVATION: MUMMIFICATION AND ADIPOCERE FORMATION ESTIMATION OF THE PMI IN THE LATER STAGES OF PUTREFACTION AND SKELETONIZATION Forensic Entomology and the Estimation of the PMI ESTIMATION OF THE PMI IN SKELETAL REMAINS Indirect Methods Direct Methods METHODS MEASURING PHYSICOCHEMICAL CHANGES Nitrogen and Albumin Levels in Bones Ultraviolet Fluorescence Benzidine Staining Chemiluminescence Citrate Content of Bone Raman Spectroscopy Carbon Dating and the Use of Naturally Occurring and Artificial Radioactive Isotopes CONCLUSIONS REFERENCES
Human Body Decomposition. DOI: http://dx.doi.org/10.1016/B978-0-12-803691-4.00004-2 © 2016 Elsevier Inc. All rights reserved.
92
Human Body Decomposition
ESTIMATION OF THE POSTMORTEM INTERVAL DURING THE STAGE OF AUTOLYSIS AND EARLY PUTREFACTION From th From thee time time a ca cada daver ver cool coolss to am ambi bien entt te temp mper erat atur uree an and d unti untill it becomes a skeleton, soft tissue decomposes by aerobic and anaerobic bacterial action (Rodriguez (Rodriguez and Bass, 1985). 1985). Autolysis, the disruption and disintegration ofitio cell varies depending the surrounding envi en viro ronm nmen enta tall cond condit ions ns walls, bu butt ge gene nera rall lly y begi begins ns inonbodi bo dies es buri buried ed in graves at 4872 hours after death (Dent (Dent et al., 2004). 2004). Intestinal cell memb me mbra rane ne di disr srup upti tion on re rele leas ases es aero aerobi bicc and and anae anaero robi bicc bact bacter eria ia.. The The process of autolysis gradually blends into the stage of decomposition known as putrefaction, which is characterized by decomposition occurring in an anaerobic environment (Gill-King, (Gill-King, 1997 1997). ). Decomposition is faster in the presence of oxygen, and therefore bodies on the surface decompose faster than those buried (Dent (Dent et al., 2004 2004). ). The estimation of the postmortem interval (PMI) during this stage of decomposition is more difficult and less precise because of the multiple factors involved, which consist not only of intrinsic factors (ie, the physical and diseased state of the body) but also extrinsic factors (ie, the context in which the body decomposes). A body will decompose more rapidly if sepsis or a feve feverr was was pres presen entt befo before re deat death, h, in edem edemat atou ouss ti tiss ssue ues, s, in obes obesee individuals, while decomposition will be slower in thin individuals and 2006). ). infants (Perper, (Perper, 2006
EXTRINSIC FACTORS AFFECTING DECOMPOSITION: TEMPERATURE, ENVIRONMENT, INSECTS, BURIAL, MOISTURE, SCAVENGING The major extrinsic factor affecting decomposition is temperature, which af affe fect ctss not not only only the the chem chemic ical al proc proces esse sess co cont ntro roll llin ing g the the enzy enzyma mati ticc breakdown of proteins and carbohydrates but also the activity of insects and an d bact bacter eria ia (Man Mann et al al.., 199 990; 0; Va Vass ss et al al., ., 19 1992 92). ). Decompositio Decomposition n occurs more rapidly in air than in water or soil; it is more rapid in hot dry climates in summer than in winter (Galloway (Galloway et al., 1989) 1989) and is slow in very cold climates (Komar (Komar,, 1998 1998). ). Twenty-nine bodies discovered ered deco decomp mpos osed ed in a nu numb mber er of cont contex exts ts an and d over ver a wide widesp sprrea ead d geog ge ogrraphi aphica call area rea in Spa Spain wer were stud udiied by Pri riet eto o et al al.. (2004). 2004). Decomposition occurred more rapidly in cadavers in hot coastal regions of hi high gh humi humidi dity ty and and wher wheree prol prolif ifer erat atio ion n of mi micr croo oorg rgan anis isms ms was was grea gr eate ter, r, compa compare red d with with in inla land nd re regi gion onss wh wher eree it was was dr drye yerr an and d wh wher eree
Research Resear ch in the Later Stages of Decomposit Decomposition ion
93
there were there were more more extre extreme me se seas ason onal al chan changes ges in tempe tempera ratu ture re.. Cada Cadaver verss found at high latitudes, where snow and ice occurred, decomposed even more slowly. Cadavers located outdoors decomposed more rapidly than those indoors or buried, while buried cadavers decomposed more rapidly than those in water. The authors concluded that insects, carnivores, high temperatures, and humidity were accelerants of the decomposition proc pr oces ess, s, wh whil ilee lo low w temp tempeeratu rature res, s, bu buri rial al,, and and su subm bmer erssio ion n in wate waterr delayed decomposition. Other extrinsic factors hastening decomposition are insects and animal scavenging. Together with temperature and burial depth, insects are the next most important variable, and influence the rate of decomposition by hastening tissue disintegration and the spread of bacteria (Mann et al., 1990; Payne, 1965; Simmons et al., 2010a 2010a). ). Animal scavenging is random, opportunistic, and occurs in both aquatic and terrestr tria iall se sett ttin ings gs (Spra Spradl dley ey et al al., ., 20 2012 12;; St Stea eadm dman an an and d Wo Worn rne, e, 20 2007 07). ). Wilson and col Wilson collea league guess (200 (2007) 7) studi studied ed deco decomp mpos osin ing g buri buried ed pi pigs gs and and foun fo und d that that animal animal sc scav aven engi ging ng in incr crea ease sed d th thee rate rate of deco decomp mpos osit itio ion. n. Most Mo st st stud udie iess of an anim imal al sc scav aven engi ging ng are are case case stud studie iess whic which h do no nott adman an and advance research into the determination of the PMI (Ste (Steadm Worne, 2007; Reeves, 2009). 2009). Many factors affect the decomposition process in buried bodies. The deeper the burial, the slower is the decomposition (Rodriguez (Rodriguez and Bass, 1985). 1985 ). The longer the burial, the greater the degree of decomposition, but there is no linear relationship between the burial time and stage of deco de comp mpos osit itio ion. n. Bodi Bodies es buri buried ed in hot hot se seas ason onss de deco comp mpos osee fa fast ster er than than thos th osee buri buried ed in co cold lder er se seas ason onss (Brei Breitm tmei eirr et al al., ., 20 2005 05)). Soil type influences the rate of decomposition. In the previously mentioned study of Wilson Wilson and colleagues (2007), (2007), pigs were buried in pasture, deciduous woodland, and upland peat soil. Decomposition occurred most rapidly in pasture, less so in woodland and was slowest in upland moor soils. This was attributed to soil conditions such as moisture content, pH, and nutrient availability, which in turn controlled microbial numbers and activity at each site. Carter and colleagues (2010) (2010) identified moisture as being the dominant environmental factor determining cadaver decomposition in soil. In their heir st stud udy, y, wh whic ich h was was des design igned to tes estt the the hyp hypothe othesi siss that hat an incr increa ease se in mois moistu ture re cont conten entt of so soil il woul would d re resu sult lt in a decr decrea ease se in cadaver decomposition, rats were buried in a loam soil, a sandy soil,
94
Human Body Decomposition
and a clay soil. The availability of moisture in soil is partly determined by the matric potential; specifically, the suction capacity of the soil for water, which in turn controls microbe motility, the diffusion of nutrients and waste, and the activity of extracellular enzymes. The matric potential is measured in megaPascals (MPa), and Carter et al. (2010) defined the matric potential of dry soil as being 0.3 MPa; moist soil 0.05 0.05 MP MPa; a; an and d wet wet so soil il 0.01 0.01 MP MPa. a. Mi Micr crob obia iall acti activi vity ty in soil soil was was grea gr eate test st at a ma matr tric ic pote potent ntia iall of ap appr prox oxim imat atel ely y 0.01 MPa. MPa. Clay, Clay, loam, and sandy soils, respectively, hold decreasingly less moisture (ie, thee mat th matri ricc pote potent ntia iall grad gradua uall lly y in incr crea ease ses) s).. Cadav Cadaver er de deco comp mpos osit itio ion n increased with a decrease in matric potential (ie, it was more rapid in moister soils), but only to an optimal level (0.01 MPa). The slower rate of decomposition in dry soils was not only due to restriction of microbe motility resulting from a lack of nutrient supply but also to the lack of moisture restricting the activity of enzymes, many of which were we re hy hydr drol olyt ytic ic.. In co cont ntra rast st,, in ve very ry wet wet soil soilss li limi mite ted d gas gas di diff ffus usio ion n ca caus used ed a de decr crea ease se in ae aero robi bicc meta metabo boli lism sm an and d re resu sult lted ed in de decr crea ease sed d decomposition.
DELAYED DECOMPOSITION AND PRESERVATION: MUMMIFICATION AND ADIPOCERE FORMATION In certain environments, decomposition ceases and preservation of soft tiss tissue ue occu occurs rs.. Mumm Mummif ific icat atio ion n is one one such such proc proces ess, s, occu occurr rrin ing g in th thee whol wh olee body body or par parts of it wh when en the en envi viro ronm nmen entt is extr extrem emel ely y dry dry (Forbes Forbes,, 2008 2008). ). Rapi Rapid d desi desicc ccat atio ion n of ti tiss ssue uess in inhi hibi bits ts en ente teri ricc and and soil soil bacterial action and can occur in very cold dry environments as well as 1986). ). The process of mummivery hot dry environments (Amy (Amy et al., 1986 fication is often a balance between putrefaction and dehydration with desiccation. Mummification can also occur naturally when bodies are preserved by the action of tannic acid in water, as occurs in bodies found in peat bogs (Aufderheide, (Aufderheide, 1981). 1981). Bodies and body tissues may also be preserved by the formation of adip ad ipoc ocer eree wh whic ich h is the the post postmo mort rtem em co conv nver ersi sion on of fa fatt tty y ti tiss ssue ue in into to a soli so lid d mate materi rial al co comp mpos osed ed of sa satu tura rate ted d fatt fatty y acid acids, s, form formed ed by the the hydr hy drol olys ysis is of trig trigly lyce ceri ride dess to free free fatt fatty y acid acidss whic which h then then unde underg rgo o hydr hy drog ogen enat atio ion n to satu satura rate ted d fa fatt tty y acid acidss such such as myri myrist stic ic,, palm palmit itic ic,, and an d st stea eari ricc acid acidss (Forb Forbes es et al al., ., 20 2005 05)). The The proc proces esss is ai aide ded d by th thee 2008). ). actions of aerobic and anaerobic putrefactive bacteria (Forbes, (Forbes, 2008
Research Resear ch in the Later Stages of Decomposit Decomposition ion
95
Conjugation of the fatty acids with bivalent metallic ions forms insoluble soaps of a hard consistency (Forbes, (Forbes, 2008). 2008). Fourcroy first described adip ad ipoc ocer eree form format atio ion n in exhu exhuma mati tion onss from from a Pa Pari riss ceme cemete tery ry in 1789, 1789, gave it the name, and described it as a form of soap resulting from reactions of fat and ammonia (Ub Ubel elak aker er an and d Za Zare renk nko, o, 20 2011 11). ). Adipocere formation depends on the balance of a number of variable factors which include temperature, moisture, oxygen availability, soil type ty pe,, meth method od of buri burial al,, and and the the prese resenc ncee or abse absenc ncee of cl clot othi hing ng Ubel elak aker er an and d Za Zare renk nko, o, 20 2011; 11; Sl Sled edzi zik k an and d Mi Mico cozz zzi, i, 19 1997 97). ). Re Rece cent nt (Ub research has focused on the chemical formation of adipocere and its potential use, if any, in the determination of the PMI (Gotouda (Gotouda et al., 1988; Yan et al., 2001; Forbes et al., 2004; O Brien and Kuehner, 2007). 2007). ’
In order to study the effect of the burial factors of soil pH, temperature atu re,, mois moistu ture re,, and oxyge oxygen n conte content nt on th thee chem chemis istry try of ad adip ipoc ocer eree formation, Fo Forb rbes es et al al.. (2 (200 004) 4) buri buried ed 10 pi pigs gs in shal shallo low w grav graves es.. Fatty acid content was analyzed from cadaver tissue and from the soil beneath the cadavers. Fatty acids increased and were present in vary va ryin ing g quan quanti titi ties es at ea each ch of th thee stag stages es of de deco comp mpos osit itio ion. n. Ther Theree was a lack of correlation between the stage of adipocere formation and an d the the deco decomp mpos osit itio ion n inte interv rval al,, sugg sugges esti ting ng that that the the proc proces esss was was affected by factors in the environment other than the PMI, such as tempera tem peratur ture, e, moistu moisture, re, clothi clothing, ng, soil soil type, type, and pH. Acco Account unting ing for these factors might better determine the correlation between the PMI and adipocere formation. In extending these studies, Forbes and colleagues (2005) investigated soil factors responsible for adipocere formation. Samples of pig flesh were buried in various types of soil to test the effects of soil pH, temperature, moisture, and oxygen content on the buried pig flesh. The control soil was a damp loamy sandy soil of pH 5.2. To test soil pH, samples of soil with pH 2.4 (acid), pH 8.5 (alk (a lkal alin ine) e),, an and d pH 12.6 12.6 (p (pro rodu duce ced d by addi adding ng li lime me)) we were re te test sted ed.. To test the effect of temperature, one sample was kept at 4 C and the other at 40 C. To test the effect of moisture, one sample was kept dry and the other wet. Finally, to test oxygen content, the control sample of soil was kept in an anaerobic environment while the other samples were all tested in an aerobic environment. Adipocere formation was found in most of the burial environments although the degree of formation varied. Factors favoring adipocere formation were a mildly
alkaline pH, warm temperature, and anaerobic conditions, whereas fact fa ctor orss inhi inhibi biti ting ng it we were re cold cold temp temper erat atur ure, e, hi high gh al alka kali lini nity ty,, and and
96
Human Body Decomposition
aerobic conditions. The highly acid soil did not entirely prevent adipoce po cere re fo form rmat atio ion n but but sl slow owed ed it, it, po poss ssib ibly ly due due to a lowe lowerr bact bacter eria iall load. Most work into adipocere formation has been carried out with pigs, butt re bu rese sear arch ch usin using g pi pig g and and huma human n adipo adipose se ti tiss ssue ues, s, deco decomp mpos osin ing g in similar controlled situations, found a difference in the rate of degradation between human and pig tissues (No Nottter et al., 2009). This difference was attributed to the variation of total fatty acid composition between species; pig adipose tissue has higher levels of total fatty acids than the same weight of human adipose tissue and therefore adipocere formation occurs sooner in the decomposition process in pigs than in human tissue. Variations in the amount of sodium, potassium, calcium, and magnesium were also detected between species. Ubel elak aker er an and d Za Zare renk nko o (2 (201 011) 1) state stated d that that adip adipoc ocer eree re repr pres esen ents ts a Ub chemical form of mummification and that it can form in a variety of enviro env ironme nments nts,, both both aquati aquaticc and terres terrestri trial. al. It can form form shortl shortly y aft after er death and be present for hundreds of years. Factors favorable for the formation of adipocere are the presence of body fat, moisture, mildly alka alkali line ne pH pH,, warm warm temp temper erat atur ure, e, an anae aero robi bicc cond condiiti tion ons, s, and and th thee approp app ropria riate te bac bacter teria. ia. Aerobi Aerobicc condit condition ionss and Gram-p Gram-posi ositiv tivee bacter bacteria ia will aid in its degradation. Both mummification and adipocere formation complicate attempts to more more accu accura rate tely ly de dete term rmin inee th thee PMI PMI by pres preser ervi ving ng ti tisssue sue and and delaying decomposition. Some tentative research has been carried out into determining the PMI in early mummified cadavers by estimating the ion ratios leaching into soil beneath a cadaver (Vass ( Vass et al., 1992). 1992). At the the pres presen entt time time re rese searc arch h in into to th thee chem chemic ical al chan changes ges occu occurr rrin ing g in adipocere formation as a means of determining the PMI is in the very early stage (Ubelaker (Ubelaker and Zarenko, 2011). 2011).
ESTIMATION OF THE PMI IN THE LATER STAGES OF PUTREFACTION AND SKELETONIZATION Forensic Entomology and the Estimation of the PMI Forensic entomology is a specialty within forensic science with an extensive and growing volume of literature. The first documented case of insects being used in a forensic case is that of Sung Tz u, in mid-13th-century China (see also chapter: Supravital Reactions in the Estimation of the ’
Research Resear ch in the Later Stages of Decomposit Decomposition ion
97
Time Since Death (TSD)) who investigated the case of a man found dead by a roadside (Sung (Sung Tz u, 11861249 1249). ). The man had not been robbed but had 10 wounds inflicted by a sickle. After making inquiries of the man s wife, Sung Tz u went to the neighborhood of a man suspected of being the perp pe rpet etra rato torr of the the cr crim imee and and dema demand nded ed that that al alll the the near nearby by owne owners rs of sickles lay them on the ground. The weather was hot and the murderer was identified as being the owner of the only sickle with blood staining on it, to which flies flocked. The first modern recorded case using insects in a forensic investigation was in France when Dr Bergeret, in a report dated 1855, described attempts to determine the time of death of a baby from the fly eggs, larvae, and pupae as well as moth pupae deposited on the cke, 2001 2001). ). In the 1920s the first case studies of important body (Bene (Benecke, foren fo rensi sicc insec insectt sp speci ecies es began began to be pub publi lishe shed d an and, d, impor importan tantly tly,, it was recognized that blowflies colonized open wounds on cadavers, resulting in more rapid decomposition, but that insects failed to colonize mummified cadavers (Benecke, (Benecke, 2001 2001). ). Only since the 1980s has entomology developed ’
’
’
2002). ). as a widespread forensic tool in Australia (Wallman, (Wallman, 2002 In the class of insects, the blowfly family in the order Diptera, has a worldw wor ldwide ide dis distri tribut bution ion and is the most most for forens ensica ically lly import important ant ins insect ect family with which the PMI is determined, because the lifecycle of the blowfly is intimately tied to carrion (Wallman, (Wallman, 2002). 2002). They are the first insect to colonize a cadaver, attracted by the odors and gases exuded by decaying organic matter and the colonization, egg laying (oviposition), and subsequent growth and development depend on environmental factors, the most important of which are temperature and ease of access to the cadaver (Mann (Mann et al., 1990; Campobasso et al., 2001). 2001 ). After Aft er blowfl blowflies ies,, hou house se flies flies (famil (family y Mus Muscid cidae) ae) and flesh flesh fli flies es (famil (family y Sarcophagidae) colonize a body at the bloated stage. There follows a succession of fly and beetle families that are then attracted to the body duri du ring ng th thee vari variou ouss stag stages es of deca decay, y, by ra ranc ncid id fa fats ts and and ammon ammoniac iacal al bass sso o et al al., ., 20 2001 01;; Ro Rodr drigu iguez ez an and d Ba Bass ss,, 198 1983 3). fermentation fermen tation (Campo Campoba Body orifices are often the first area to be colonized followed by moist flexure creases, as the maggots move to areas of more favorable food opportunities (Campobasso (Campobasso et al., 2001). 2001). The eggs develop into larvae or maggots, which go through three stages of development known as instars during which time the morphology changes with growth and the passage of time. The skin of the maggots hardens and they form pupa pu pae, e, wh whic ich h late laterr form form in into to adul adultt fl flie ies. s. The The grow growth th rate rate is fast faster er when the environmental temperature is high, when there is increased
98
Human Body Decomposition
temperatu temper ature re genera generated ted from from bacter bacterial ial action action during during decay decay and when when temperature rises as a result of the growing maggot mass (Simmons ( Simmons et al., 2010a; Rivers et al., 2011). 2011). Blowflies mainly lay eggs in daylight, although they have been known to do so in artificial light and temperature is the main factor in their growth and development (Campobasso ( Campobasso et al., 2001). 2001). There is a temperature range which is optimum for blowfly larvae to develop and the range varies with each species. Below a certain temperature embryogenesis in eggs and larval development will slow down or cease but may resume or speed up as temperatures again ri rise se (Camp Campob obas asso so et al. al.,, 20 2001 01). ). Abov Abovee a cert certai ain n te temp mper erat atur uree larval larval feeding is inhibited, development of pupae slows, and if the temperature tu re is too too hi high gh cert certai ain n toxi toxicc prot protei eins ns devel develop opin ing g in the the la larv rvae ae may 2011). ). Maggot masses produce a rise prove to be lethal (Rivers (Rivers et al., 2011 in temperature dependent on the size of the maggot mass and the size of the cadaver (Campobasso (Campobasso et al., 2001 2001). ). The cause of the temperature rise is unknown but has been attributed to microbial activity, frenetic muscularprocesses movement of the et mass, and high 2011) ) andmetabolism it has been associated known to with digestive (Rivers (Rivers al., 2011 reach very high levels (Anderson (Anderson and Van Laerhoven, 1996 1996). ). After temperature, access to the body by insects is the next most Mann nn et al al., ., 19 1990 90)). The The import imp ortant ant fact factor or affect affecting ing decomp decomposi ositio tion n (Ma abse ab senc ncee of bl blow owfl fly y an and d beet beetle le activ activit ity, y, oc occu curr rrin ing g in cert certai ain n envi enviro ronnmental conditions which prevent insect access; burial of a body, excessi sive ve clot clothi hing ng co cove ver, r, an and d exce excess ssiv ivel ely y co cold ld te temp mper erat atur ures es will will dela delay y decomposition. The deeper a body is buried the more limited is insect acce ac cess ss an and d in buri burial alss deep deeper er than than 60 cm th ther eree is no si sign gn of ca carr rrio ion n insect activity (Rodriguez (Rodriguez and Bass, 1985; Campobasso et al., 2001). 2001 ). Similarly there is a different array of insect species which colonize a cada ca dave verr foun found d in indo door orss comp compar ared ed with with on onee foun found d ou outd tdoo oors rs (Goff, 1991). 1991 ). Recent studies from southern Finland and Canada found that indoor carcasses were colonized more slowly by blowflies and that th at deco decomp mpos osit itio ion n was was sl slow ower er th than an that that occu occurr rriing in ca carc rcas asse sess situated outdoors (Pohjoismäki (Pohjoismäki et al., 2010; Anderson, 2011 2011). ). In the Finn Fi nnis ish h st stud udy, y, whic which h stud studie ied d huma human n re rema main ins, s, di diff ffer eren entt sp spec ecie iess of insect ins ectss coloni colonized zed cadaver cadaverss decomp decomposi osing ng indoor indoors, s, com compar pared ed with with the species which colonized cadavers decomposing outdoors (Pohjoismäki (Pohjoismäki et al., 2010). 2010). An average delay of 3 4 days occurred before the first female blowflies found the indoor cadavers and oviposited. This was attributed to more difficult access to the indoor cadavers, time of year,
Research Resear ch in the Later Stages of Decomposit Decomposition ion
99
cooler temperatures, and the species reserves of the local insects. In the Canadi Can adian an st stud udy y usin using g pi pig g carc carcas asse ses, s, ther theree was was a 55-da day y dela delay y befo before re indoor carcasses were colonized (Anderson, (Anderson, 2011 2011). ). Many more insects colonized the outdoor carcasses and these were more rap apiidly skeletonized. Not all insects insects which coloniz colonizee a cadaver active actively ly consume decomposing tissue; some species are opportunists or parasites of other species. (1993) classified the insects attracted to decomposing cadavers into Goff (1993) classified four categories: 1. Necr Necrop opha hago gous us spec specie iess are are tho those di dire rect ctlly fe feed edin ing g on th thee cada cadave verr. Diptera (Calliphoridae and Sarcophagidae) and Coleoptera (Silphidae and Dermestidae) are the dominant species in this category. 2. Pred Predat ator orss an and d paras parasite itess of the the ne necr crop opha hagou gouss sp spec ecie iess whic which h in incl clud udee other species of Coleoptera and Calliphoridae and also Hymenoptera. Some So me Dipt Dipter era a larv larvae ae be beco come me pred predat ator orss in the the la latter stag stagee of th thei eirr development. 3. Omnivorous Omnivorous species species such as wasps, ants, and some beetles beetles which feed on the corpse as well as the necrophagous species. Large numbers of th thes esee may may dela delay y deco decomp mpos osit itio ion n by exce excesssi sive vely ly cons consum umin ing g the the necrophagous species. 4. Adve Advent ntiv ivee sp spec ecie iess us usee the the corp corpse se as an exte extens nsio ion n of th thei eirr norm normal al habitat and include spiders and centipedes. While biochemical and temperature methods of determining the PMI may give reasonable results up to about 13 days after death, entomologic log ical al evid eviden ence ce may may ext exten end d the the est estim imat atio ion n of the PM PMII up to se seve vera rall weeks or even months (Amendt (Amendt et al., 2007). 2007). From 48 hours after death forensic entomology is frequently the only way of determining the PMI with any degree of accuracy (Wallman, (Wallman, 2002). 2002). The minimum PMI (mPMI) in the first days after death is based on thee ag th agee asse assess ssme ment nt of the the ol olde dest st im imma matu ture re stag stages es of necr necrop opha hago gous us insects (eggs, larvae, or pupae) on the body, mainly blowflies, the first insects to infest a cadaver, which in turn depends on the species and thei th eirr part partic icul ular ar rate rate of grow growth th.. Thi Thiss age asses assessm smen entt wi will ll in indi dica cate te a mPMI, corresponding with the time that the insects first colonized the body but it may not correspond with the actual time of death as the period of insect activity may be less than the PMI. The reason for this is that blowflies may have either not colonized the body immediately after
100 Human Body Decomposition
death because of difficulty of access or they may have colonized the body before death if necrotic tissue had been available in a debilitated and dying individual (Amendt (Amendt et al., 2007). 2007). Later in the postmortem period the mPMI is based on the composition of the insect community; different insect species infest a body at different times during its decomposi po siti tion on,, as th thee acti activi vity ty of eac each h succ succes essi sive ve spec specie iess re rend nder erss th thee body body unsuit uns uitabl ablee for contin continuin uing g occupa occupatio tion n by itself itself (Wal Wallman lman,, 2002 2002). ). Age asse as sess ssme ment nt of the the ol olde dest st im imma matu ture re sp spec ecim imen enss wi will ll depe depend nd on co corr rrec ectt identification of the predominant local species and environmental conditions. Studies such as those of Anderson of Anderson and Van Laerhoven (1996), (1996), (2001), and Archer and Elgar (2003) Wallman (2001), (2003),, which determine the predominant species in a particular geographical area, are important in prov pr ovid idin ing g a re refe fere renc ncee da data taba base se of bl blow owfl fly y spec specie iess for for fo fore rens nsic ic case casess incl includ udin ing g meas measur urem emen ents ts of leng length th and and we weig ight ht of th thee la larv rval al in inst star arss (Amendt et al., 2007 2007). ). Identification of the different species of blowflies is made more accurate by examining the microscopic morphology of mouth parts, posteriorby body spiracles, andcells, by molecular of proteins expressed the DNA within the DNAidentification itself, or the mito mi toch chon ondr dria iall DNA DNA (Harv Harvey ey et al al., ., 20 2003 03;; Wa Walllm lman an et al al., ., 20 2005 05). ). The later stages of pupae and adults can also be aged by measuring the st stag ages es of de deve velo lopm pmen entt and and comp compar ariing them them with with table abless rela relati ting ng development to temperature (Amendt (Amendt et al., 2007). 2007). Various arthropod succession patterns have been used to determine the PMI after blowflies have migrated from a body, in one case up to 36 days (Goff ( Goff and Flynn, 1991)) but when using the greater carrion insect community in forensic 1991 investigations it is necessary to have knowledge of local successional patterns based on experimental studies on animal carcasses such as pigs (Amendt et al., 2007). 2007). As growth rate and development are dependent on temperature, the quantitative amount of heat required is measured in accumulated degree hours (ADH) or accumulated degree days (ADD). ADD was originally introduced by Arnold (1959 ( 1959)) over 50 years ago as a thermal unit measure, although although Réaumur (1735) developed the concept of heat Réaumur (1735) developed units with reference to plant development in the 18th century (Wang, (Wang, 1960)) (s 1960 (see ee an exte extend nded ed disc discus ussi sion on of the the use use of ADD ADD in chap chapte ter: r: Recent Research and Current Trends). The use of these measures was subs su bseq eque uent ntly ly prom promot oted ed by Edwa Edward rdss and and Ch Chan aney ey,, who who de defi fine ned d a degree day as the amount of biological activity caused by 1 degree above the threshold for 1 day (Edwards et al., 1987 1987). ). Amendt et al. “
”
Research Resear ch in the Later Stages of Decomposit Decomposition ion
101
(200 2007) defined ADH and ADD as the temp mpeerature in degree greess Centigrade above base temperature multiplied by time either in hours or days. The use of these measurements is based on the assumption that th at unde underr a cons consta tant nt temp temper erat atur uree th ther eree is a li line near ar rela relati tion onsh ship ip betw be twee een n temp temper erat atur uree and and grow growth th and and de deve velo lopm pmen ent. t. Th Thee base base or threshold temperature is the temperature below which development ceases and this, and the temperature range which is optimal, varies for each developmental stage and each species of blowfly in different geographical areas (Amendt (Amendt et al., 2007; Edwards et al., 1987). 1987 ). If the temperatu per ature re ra rang ngee fluc fluctu tuate atess the the devel develop opme ment nt may becom becomee le less ss lin linear ear.. The recording of hourly temperature at the site of body discovery for 5 or 10 days afterwards, with a data logger, is carried out. Temp Te mper erat atur uree re reco cord rdss for for the the ac actu tual al ti time me the the body body la lay y in si situ tu are are obtained from the nearest local weather station. A regression equation is then then ca calc lcul ulat ated ed fr from om wh whic ich h a corr correc ecti tive ve fact factor or is obta obtain ined ed an and d applied to the weather station readings in order to compensate for the geographical distance between body discovery and the weather station (Amendt (Amendt et al., 2007 2007). ). the A study by Archersite (2004) found that although the accuracy of temperature data at particular sites was usually improved by correction factors, the degree of improvement was highly variable between correlation periods, therefore generous error margins of at least 6 12 hours should be quoted and in some cases should be even much longer. Johnson et al. (2012) expanded on the 2004 study. They found that although retrospective ambient temperature correction of weather station temperature was a robust technique and was not significantly affected by season, corrected weather data were we re less less re repr pres esen enta tati tive ve of actu actual al body body detect detectio ion n si site te te temp mper erat atur ures es if correlation was attempted when there was a temperature difference of .5 C between the weather station collection data and the correlation data from the body discovery site. In such a case they recommended a collection of 10 days of correlation data in order to extend the correlation period to try to correct any difference exceeding 5 C between weather station data and the body detection site data.
Is Isom omeg egal alen en re refe fere renc ncee cu curv rves es,, whic which h gi give ve in inse sect ct grow growth th stag stages es fo forr various constant temperatures and isomorphen diagrams, which document morphological development depending on time and temperature, have ha ve be been en deve develo lope ped d to aid the the corr correl elat atio ion n of age and and morp morpho holo logy gy with temperature but should be developed for the same and different species in different geographical areas (Grassberger ( Grassberger and Reiter, 2001). 2001).
102 Human Body Decomposition
Strict Stri ct,, univ univer ersa sall lly y re reco cogn gniz ized ed,, and and cons consis iste tent nt stan standa dard rdss ar aree al also so necessary for collecting, storing, and assessing insect species in order that legal requirements are met and studies can be compared (Amendt ( Amendt et al., 2007). 2007). Insects are not only important in determining the mPMI but have been be en show shown, n, toge togeth ther er with with temp temper erat atur uree and and buri burial al dept depth, h, to be the the most impo most import rtan antt fact factor orss affe affect ctin ing g th thee rate rate of de deca cay y in bodi bodies es (Mann et al., 1990). 1990). Payne Payne (1965) (1965) had had noted this fact in a study of decomposition of baby pigs and it was further confirmed by Rodriguez and Bass (1983),, when they studied insect successions on four human cadavers (1983) decomposing in different seasons of the year. They noticed that during autumn and winter, when insect populations were at their minimum, decomposition occurred more slowly. They concluded that insects were a major factor responsible for decomposition. In a study on groups of rabb ra bbit itss dec ecom ompo posi sing ng on the the surf surfac acee of the the gr grou ound nd,, some some bein being g disturbed at intervals while others were left undisturbed, it was found that th at the the undi undist stur urbe bed d grou group p lo lost st weig weight ht,, thei theirr body body te temp mper erat atur uree was was greater, and they decomposed faster than the disturbed group (Adlam ( Adlam and Simmons, 2007). 2007). Temperature over the 3-week study period was meas me asur ured ed as AD ADDs Ds and and corr correl elat ated ed wi with th the the stag stagee of de deco comp mpos osit itio ion, n, which was measured by a quantitative scoring system. The faster rate of de deco comp mpos osit itio ion n in the the undi undist stur urbe bed d grou group p was was at attr trib ibut uted ed to hi high gher er temp temper erat atur ures es re reac ache hed d as a re resu sult lt of an in incr creas easee in bl blow owfl fly y numb number ers, s, which in turn resulted in a greater maggot mass and consequently, to a faster rate of decomposition. In a re rece cen nt stud study y usin using g a quan uantita titati tive ve body body sc sco ori ring ng sy sysste tem m to measuree decomp measur decomposi ositio tion n and ADDs ADDs to measur measuree temper temperatu ature re ove overr the time of decompositio decomposition, n, Bachmann and Simmons (2010) showed that soft tissue decomposition was significantly faster in a group of 30 rabbi bitt carc carcas asse sess expo expose sed d to in inse sect ctss for for 5 hour hourss pr prio iorr to sh shal allo low w buri burial al,, than in a group of 30 rabbit carcasses which had insects excluded by st stor orin ing g th them em in pl plas asti ticc bags bags at 4 C duri during ng tran transp spor orta tati tion on,, the the bags bags bein be ing g re remo move ved d ju just st be befo fore re imme immedi diat atee buri burial al.. Both Both gr grou oups ps of rabbi rabbits ts were buried at the same time.
Simm onss an and d co coll llea eagu gues es (2 (201 010b 0b)) un undertook an an anal aly ysis of a Simmon number of published and unpublished studies using rats, rabbits, pigs, and humans in varyi yin ng envi virronments. Only studies in which it was possible to convert the timescale to ADDs were chosen.
Research Resear ch in the Later Stages of Decomposit Decomposition ion
103
Decomp Deco mpos osit itio ion n had had prev previo ious usly ly been been show shown n to co corr rrel elat atee with with weig weight ht loss (Adlam (Adlam and Simmons, 2007 2007). ). If weight loss data were reported in a study as a percentage, they were converted into a total body score (TBS) by using a specific formula. By this means Simmons and colleagues were able to compare ADDs with TBS and found that, regardless of the environment, cadavers with insect access decomposed at a faster rate ra te than than cada cadave vers rs when when in inse sect ctss were were excl exclud uded ed.. When When in inse sect ctss were were present small carcasses decomposed faster than large carcasses, which the authors explained by there being more mass for the insects to consume but when insects were excluded carcass size had no influence on the decomposition rate. Forensic entomology is therefore a recent but increasingly important and developing specialty within forensic science which has increasing potential to be of assistance in more accurately estimating the TSD in decomposed human bodies, especially after the first 48 hours after death (Wallman, (Wallman, 2002). 2002). Its importance will increase with the number of studies carried out to identify the different species of blowflies and other oth er arthro arthropod pods, s, the their ir dev develo elopme pment, nt, and behavi behavior or in differ different ent geogeoWallman lman,, 200 2001; 1; Arc Archer her and Elg Elgar, ar, 2003 2003;; And Anders erson, on, graphi gra phical cal are areas as (Wal 2000; Matuszewski et al., 2008, 2010a,b). 2010a,b).
ESTIMATION OF THE PMI IN SKELETAL REMAINS The estimation of the PMI of a cadaver decomposed to a skeleton is much more difficult and inaccurate. In the United Kingdom it is generally accepted that human bones in which the date of death is estimated to be grea greate terr than than appr approx oxim imat atel ely y 75 ye year arss ar aree trea treate ted d as hi hist stor oric ical al cases, whereas bones with a date of death less than 75 years are investigated gat ed as fore forens nsic ic case casess alth althou ough gh th this is cu cutt-of offf da date te vari varies es in di diff ffer eren entt jurisdictions throughout the world (Swift, 2006; Yoshino et al., 1991). 1991). The methods available can be categorized as indirect and direct (Swift, (Swift, 2006; Knight and Lauder, 1969). 1969).
Indirect Methods Botanical knowledge may assist in determining the PMI with buried bodies or bodies which have lain in the open for a long time. Palynology (pollen and spores analysis) may indicate the season when thee rema th remain inss were were depo deposi site ted d (Sw Swif ift, t, 200 2006 6) and assessment of the root ro otss of pere perenn nnia iall pl plan ants ts gr grow owin ing g thro throug ugh h or ar arou ound nd th thee body body by
104 Human Body Decomposition
dendro dend roch chro rono nolo logy gy may give give a mini minimu mum m TSD TSD (Wil Willey ley and Hei Heilma lman, n, 1987). 1987 ). Assessment may be difficult because the seasonal rate of growth may vary with the environmental conditions. Roots growing vertically thro th roug ugh h the the rema remain inss make make ass assessm essmen entt more more accu accura ratte than than those hose growing horizontally or obliquely. Difficulties occur with false rings, inco incomp mple lete te ri ring ngs, s, an and d ecce eccent ntri ricc ri ring ngs. s. Ro Root otss of ann annual ual pl plan ants ts make make assessment more accurate, as they indicate deposition of the body in that season. Despite these disadvantages the method can sometimes be tin n an and d very ve ry usef useful ul as a rece recent nt cas case fr from om Cana Canada da in ind dic icat ates es (Cour Courti Fairgrieve, Fairgri eve, 2004 2004). ). A de deco comp mpos osed ed body body was foun found d dres dresse sed d in wint winter er clothes in a wooded area, lying on the ground across the branch of a North American black spruce. As this tree grows, the lower branches di dip p on to the the grou ground nd and and then then subs subseq eque uent ntly ly grow grow up to th thee li ligh ght. t. The body had fallen over a branch as it lay on the ground and the bran br anch ch had had then then grow grown n upwa upward rdss over over se seve vera rall year years. s. Th Thee weig weight ht of the body on the branch had compressed the annual growth rings on thee si th side de ofand the thecounting br bran anch ch the clos closes estt to of the thecompressed grou ground nd.. By micr micros copi pica call lly y measuring number rings, it osco was establish lished ed that that the the pers person on had died ied 10 ye year arss pr prev evio ious usly ly and and th thiis was was confirmed by further investigation.
Direct Methods Morphological Examination In th thee earl early y stag stagee of skel skelet eton oniz izat atio ion, n, li liga game ment ntss an and d te tend ndon onss ma may y remain rem ain attach attached ed and bone bone marrow marrow,, perios periosteu teum, m, and the collage collagenou nouss framework may persist in long bones. These features might be thought to give an estimate of the PMI but they are dependent on the environ-
ment in which the body was found, for example, acid soil will cause decomp dec omposi ositio tion n of these these struct structure uress rap rapidl idly, y, while while rapid rapid immers immersion ion in anoxic water and rapid burial in sediment will preserve them for longer periods (Schultz, (Schultz, 1997; White and Hannus, 1983 1983). ). Other factors such as animal scavenging will obscure these features (Spradley (Spradley et al., 2012 2012). ). Behrensmeyer (1978) studied the bones of mammals deposited in the Ambo Am bose seli li Basi Basin n of Keny Kenya, a, desc descri ribe bed d the the we weat athe heri ring ng proc proces esse ses, s, and and produced a timetable of weathering based on these descriptions up to 1015 years after death by which time most bones had decomposed beyo be yond nd re reco cogn gnit itio ion. n. Be Behr hren ensm smey eyer er co conc nclu lude ded d th that at weat weathe heri ring ng proprocess ce sses es migh mightt re reco cord rd usef useful ul in info form rmat atio ion n but but mu much ch mo more re re rese sear arch ch wa wass and d Han Hannu nuss (1 (198 983) 3),, who analyzed the requir req uired. ed. Res Resear earch ch by White White an chemical changes in bone as it weathered, illustrated the complexity of
Research Resear ch in the Later Stages of Decomposit Decomposition ion
105
the process. Estimation of the PMI from the morphology of skeletal material is one of the most important, yet unreliable aspects of forensic investigation. It is dependent on the experience of the examiner, which is op open en to obse observ rver er erro errorr and and lack lack of co corr rrob obor orati ative ve evid eviden ence ce (Swift et al., 2001). 2001). Microscopic Examination A study by Knight and Lauder (1969) was undertaken in an attempt to deter etermi mine ne th thee da dati ting ng of skel skelet etal al rema remain inss an and d in par parti ticu cula larr to dete de term rmin inee a me meth thod od of di dist stin ingu guis ishi hing ng mode modern rn from from anci ancien entt bone bone.. Several methods were researched including histology, but the authors conc co nclu lude ded d that that hi hist stol ology ogy wa wass tech techni nica call lly y impo imposs ssib ible le on ma many ny of th thee bones and unless undecalcified sectioning became available, it was not worth performing.
In a study by Yoshino and colleagues (1991), (1991), 51 bone fragments from the surgical necks of humeri were studied by light microscopy and electron microscopy. Of these, 33 were left in the open air, 14 were buried, and 4 were left in the sea for several years. Specimens buried in soil showed changes beginning after 5 10 years. Vacuoles first formed in the bone, which then coalesced to form labyrinthine struct str uctur ures. es. Thei Theirr walls walls were were for formed med by thin thin coll collage agen n fibri fibrils ls along along which wh ich bact bacter eria ia coul could d be seen. seen. Sim Simila ilarr chang changes es wer weree se seen en in bo bone ness exposed in the open and in the sea. However, the changes in bones in the open took many more years to develop and the changes in the bone bo ness in the the sea sea did did not not deve develo lop p into into la laby byri rint nthi hine ne cavi caviti ties es but but re remai maine ned d at the the stage stage of vacuo vacuole le fo forma rmati tion on.. The re resea searc rche hers rs con con-clud cluded ed that that the the laby labyri rint nthi hine ne stru struct ctur ures es were were form formed ed by bact bacter eria ia brea br eak king ing down down th thee wall wallss of the vacu vacuol oles es and and that hat the bact bacter eria ia attac att acked ked both both the the or orga gani nicc an and d in inor orga ganic nic comp compon onen ents ts of the the bone bone.. They were unable to formulate a quantifiable method of estimating the PMI from these microscopic changes. Bell ll an and d co coll llea eagu gues es (19 1996 96)) stu stud die ied d huma human n bo bone ness fro from thre threee Be envi en viro ronm nmen ents ts,, terr terres estr tria ial, l, lake lakes, s, and and the the se sea, a, in orde orderr to asse assess ss th thee speed of postmortem alteration to bone. Specimens were mainly of ribs and teeth, with one tibial shaft. The PMI was known approximately and extended from 3 months to 83 years. Sections were cut, treated, and examined by scanning electron microscope. The authors confirmed the changes described by Yoshino et al. (1991) but found that th at they they coul could d occu occurr very very soon soon afte afterr de deat ath h and and up to ma many ny year yearss
106 Human Body Decomposition
after. Too few specimens were examined to enable a TSD to be established but the authors favored endogenous bacteria associated with the body for causing the bone breakdown rather than exogenous bacteria associated with the environment. A deta detail iled ed hist histol olog ogic ical al stud study y of the the fe femo mora rall bone boness of 12 pig pig Whit itee and Bo Booth oth (2 (201 014) 4) in order to carcass carca sses es was ca carr rrie ied d out by Wh determine whether the bacteria responsible for postmortem degradation tio n of bone bone were were endo endoge geno nous us gut gut bact bacteri eria a or exog exogeno enous us from from th thee environment. There were 10 juvenile pigs and 2 stillborn. Five juveniles and a stillborn were buried in a shallow grave; the remainder were exposed in the open air for a year. Decomposition was staged on each carcass but no correlation was found with the histological chan ch ang ges in bone. Bon Bone histo stolog logy was was similar to that noted by Yoshino Yosh ino and colleag colleagues. ues. Whit Whitee an and d Bo Boot oth h (2 (201 014) 4) conclud concluded ed that enteri ent ericc bacter bacteria ia were were respon responsibl siblee for consum consuming ing the mic micros rostruc tructur turee and would probably alter most of the internal bone structure within 5 years. Cadavers on the surface showed more bone destruction than buried cadavers and stillborn pig bones remained free of destruction because of the lack of enteric bacteria. The findings of these studies are inte interes resti ting ng in eluci elucida dati ting ng the the caus causee and ti timi ming ng of post postmor morte tem m bone destruction but do not yet aid in the development of a method of estimating the TSD.
METHODS MEASURING PHYSICOCHEMICAL CHANGES Knight and Lauder (1967) investigated a number of physicochemical method meth odss of dati dating ng 11 sp spec ecim imen enss of bone bone,, the the da date tess of whic which h were were appr ap prox oxim imat atel ely y know known n and and whic which h vari varied ed from from 0 to 17 1700 00 year years. s. Th Thee methods were fluorescence under ultraviolet light, estimation of nitrogen ge n cont conten ent, t, benz benzid idin inee re reac acti tion on,, immu immune ne-e -ele lect ctro roph phor ores esis is,, pape paperr chromatography of the amino acid content, Nile blue and dichloroindophenol staining, and reaction with acid. The methods used were useful in distin distingui guishi shing ng anc ancien ientt from from modern modern bon bone. e. Recent Recently ly deposi deposited ted bone was distinguished by a nitrogen content of more than 4 g%, a positive benzidine reaction on powdered bone and the production of se seve ven n or more more amin amino o acids acids on hydr hydrol olys ysis is as de demo mons nstr trat ated ed by pape paperr chromatography. Loss of immunologic activity occurred after 5 years and an d benz benzid idin inee stai staini ning ng afte afterr 150 year years. s. Fluo Fluore resc scen ence ce was was of li limi mite ted d value, effervescence of the bone surface with acid was of no value, and
Research Resear ch in the Later Stages of Decomposit Decomposition ion
107
neither was Nile blue and dichloroindophenol staining. Further studies si sinc ncee then then ha have ve re rese sear arch ched ed va vari riou ouss meth method odss an and d thes thesee in incl clud udee the the following.
Nitrogen and Albumin Levels in Bones Nitrogen in bones from 42 buried invaried coffinsfrom in two teries withlevels similar soil profiles forbodies periods which 26cemeto 93 years were examined by Jarvis (1997). (1997). Quicklime had been added to 17 of th thee coff coffin inss and and ther theree was was fuel fuel oi oill cont contam amin inat atio ion n of 3 co coff ffin ins. s. Tota To tall ni nitr trog ogen en conc concen entr trat atio ion n of th thee lo long ng bone boness wa wass anal analyz yzed ed and and found to decrease with time. The relationship could be expressed by a regression equation when the confounding effect of other variables was removed. Below 4 feet (1.22 m), the depth of the shallowest burial, the depth made no difference to the nitrogen content, neither did the presence of lime. Three bodies contaminated with fuel oil had higher nitrogen ge n leve levels ls,, an and d a skel skelet eto on foun found d in the the open open on the su surf rfac acee afte afterr 14 years had a nitrogen level much less than the average nitrogen level in the buried bodies. Lowenstein (1981) reported that collagen, the principal protein of bone, had been detected by electron microscopy in dinosaurs 200 million years old. Using antisera to albumin as a control, he developed a solidsol id-pha phase se radioi radioimmu mmunoa noassa ssay y for collag collagen, en, albumi albumin, n, and oth other er proproteins and he discovered that serum factors as well as collagen could survive in fossil extracts from a baby mammoth, an Egyptian mummy, as well as in such hominids as Cro-Magnon, Neanderthal, Homo erectus,, and Australopithe tus Australopithecus cus robustus robustus.. Cat Catta tane neo o and co coll lleag eague uess (1 (1992 992)) demons demo nstr trat ated ed that that hu human man albu albumi min n coul could d be id iden enti tifi fied ed in bone bone,, with with good sensitivity and specificity, by inhibition enzyme-linked immunosorb so rben entt assa assay y (E (ELI LISA SA)) usin using g a mono monocl clon onal al anti antibo body dy of th thee Ig cl clas asss against human albumin. They successfully identified albumin in human bone from 2200 to 1700 BC, from skeletons dating from the English Civil War in AD 1644, from medieval (AD 1100 1400), early Saxon (AD (A D 450 45060 600) 0),, Roma Roman n (AD (AD 100 10020 200) 0),, and and from from Iron Iron Ag Agee bone one (400 BC). The results were unequivocal with no crossreactivity between humans and other species. The authors suggested that the method provided a means to reconstruct dietary, ritual, and domestic practices of anci an cien entt pe peop ople less bu butt th thee me meth thod od was was no nott su sugg gges este ted d to es esti tima mate te the the Swift (2006) (2006),, in his review of the use of nitrogen levels in estimatTSD. Swift TSD. ing the PMI, concluded that although the amino acid content of bone,
108 Human Body Decomposition
from which nitrogen is derived, may alter in a somewhat recognizable manner, the rate of decrease was unpredictable and it was prone to alteration by too many external factors such as fertilizers in soil, to be of practical use in the estimation of the PMI.
Ultraviolet Fluorescence
The use of ultraviolet fluorescence of bone as a method of determininii an and d Pe Pett tten ener er ing in g th thee ag agee of bon bones wa wass in inve vest stig igat ated ed by Facch Facchin (1977).. A bluish white fluorescence occurs when bone is subjected to (1977) ultraviolet light and this is caused by the presence of organic constituents. As bone ages it gradually loses these constituents and, theoreticall ca lly, y, th thee inte intens nsit ity y an and d dist distri ribu buti tion on of the fl fluo uore resc scen ence ce shou should ld diminish. It was suggested that if fluorescence diminished in a linear manner it might indicate the PMI. Seventy-one specimens of adult femo femora ra,, vary varyin ing g from from just just ov over er 20 ye year arss to 35 3500 00 year yearss ol old, d, were were subjected to ultraviolet light after being finely sectioned. A spectrofluorim fluo rimete eterr was used used to qua quanti ntitat tativel ively y measur measuree the ultravi ultraviole olett light light emitted. Samples were classified by age into recent (AD 1935 1945), modern (AD 16001800), medieval (AD 70010 1000 00), ), Ro Roma man n (AD 100300), Etruscan (540450 BC), archaic (600500 BC), and anci an cien entt Bron Bronze ze Ag Agee (1 (170 700 0150 1500 0 BC BC)) group groups. s. Fl Fluo uores resce cence nce,, when when ul ultr trav avio iole lett ligh lightt wa wass appl applie ied, d, was was wide widesp spre read ad ov over er the the surf surfac acee of bones up to 200350 years old, discontinuous and weaker in medieval samp sample les, s, and absent absent in ol older der bone bones. s. Qu Quan anti tita tati tive ve di diff ffer eren ences ces in fluorescence intensity were noted in the modern samples but it was not significant. The authors concluded that fluorescence was too variable to be used as a reliable dating method. (1991), found that ultraviolet fluorescence of Yoshino and colleagues (1991), bones diminished with time in a sample up to 15 years after death. They calculated regression equations to date the bone, but the standard error of all samples was 2.223 years and the confidence interval at 5% was 4.5 years. A study by Hoke and colleagues (2011) visually determined the UV-in UVindu duce ced d auto autofl fluo uore resce scenc ncee of th thee cr cros osss-se sect ctio ions ns from from 76 hors horsee bonee fr bon fragm agment entss datin dating g fr from om 15, 15,00 000 0 to 200 2000 0 BC from from pre prehi hist stor oric ic archeo arc heolo logi gical cal si site tess of varyi varying ng envir environ onme ment ntal al and and chro chrono nolo logic gical al backgro back ground und.. In each sample sample they they compar compared ed the flu fluore orescen scence ce with with thee de th degr gree ee of hi hist stol ologi ogica call pr prese eserv rvati ation on,, it itsel selff a mark marker er for for ove overa rall ll
Research Resear ch in the Later Stages of Decomposit Decomposition ion
109
biomolecular preservation in bone. They concluded that the relationship between fluorescence and unaltered microstructure appeared to be go good od an and d that that scre screen enin ing g samp sample less wi with th fl fluo uore resc scen ence ce coul could d help help identify samples with the best chances for collagen and DNA recovery. UV fluorescence was also faster and less elaborate than the histological method. However, the general cause of the fluorescence and exactly what caused the specific changes in color in degraded bone was still not known. Finally, a recent study by Hoke and colleagues (2013) examined 213 bones from two sets of bones with a known PMI in order to determine the diagnostic value of ultraviolet fluorescence cen ce in di disti stingu nguish ishing ing hi histo stori rical cal and rec recen entt ske skele letal tal remai remains. ns. The first set was 58 long bone samples from abandoned graves in a modern cemetery ranging from 8 to 60 years after death. The second set wass 155 wa 155 arch archeo eolo logi gica call spec specim imen enss from from di diff ffer eren entt si site tess thro throug ugho hout ut Germany. There were 76 horse bones ranging from 2000 to 15,000 years after death and 79 human bones ranging from 90 to 4500 years after death. They concluded that the use of ultraviolet fluorescence in distinguishing between a forensically relevant and a historical PMI was questionable and specifically the intensity of fluorescence did not show a significant correlation with the PMI. The fluorescence of bone under ultraviolet light is considered to be caus aused by residual organi nicc constituents, mai ain nly colla lag gen (Facc Facchin hinii and Pet Pettene tener, r, 1977 1977). ). Loss of collagen from bone is not only on ly cause caused d by micr microbi obial al degrad degradat ation ion bu butt it is al also so te temp mpera eratu ture re-depende depe ndent nt and both both these these causes causes depend depend on environ environmen mental tal condiconditions (Hedges, (Hedges, 2002 2002). ). Therefore, loss of collagen from bone is highly variable and hence estimation of the PMI by ultraviolet fluorescence will wi ll al also so be vari variab able le.. Thes Thesee stud studie iess indi indica cate te that that the the use use of bone bone fluo fluore resc scen ence ce is not not suff suffic icie ient ntly ly accu accura rate te to esti estima mate te the the PMI PMI in forensic forens ic cases.
Benzidine Staining Facchini and Pettener (1977) (1977) investigated investigated the use of benzidine staining in bone. The application of benzidine in acetic acid with the later addition of hydrogen peroxide will stain any residual blood pigment in the bone blue. Strong positive results were found in recently deceased sampl ples es,, wea weake kerr re reac acti tion onss in mo mode dern rn bone bones, s, but but me medi diev eval al and and Roman Roman bone bo ness only only gave gave fals falsee-po posi siti tive ve re resu sult ltss due due to ir iron on stai staini ning ng re resu sult ltin ing g from burial environment soils. Facchini and Pettener concluded that
110 Human Body Decomposition
the inte the intens nsit ity y of stai stain ning ing coul could d date ate re rema main inss from from 0 to 350 350 year years. s. Benzidine is carcinogenic; its use has been replaced by aminobenzidine and an d the the meth method od give givess too too many many fals falsee-po posi siti tive ve re resu sult ltss to be us usef eful ul 2006). ). (Swift, 2006
Chemiluminescence Introna and colleagues (1999) (1999) investigated investigated luminol, an alkaline reagent for its potential in estimating the PMI from bone. Luminol is routinely used in forensic serology to locate and identify blood traces and stains which are not visible to the naked eye and it is very sensitive. Luminol produces a bluish-white light when peroxidase is added to it in the presence of blood and it is the heme portion of hemoglobin that reacts with the peroxidase. The color and intensity of the light can be measured. Samp Sa mple less of bo bone ne from from 80 nonp nonpat atho holo logi gica call fe femu murs rs were were subj subjec ecte ted d to lumi lumino noll test testin ing. g. The The samp sample les, s, fr from om 40 ma male less and and 40 fe fema male les, s, we were re grouped into five time periods according to the known time of death, but with sufficient intervals to prevent overlaps in the positive results from the tests. The time periods were 1 month to 3 years, 10 15 years, 2535 year years, s, 5060 year years, s, an and d grea greate terr than than 80 year years. s. A came camera ra re reco cord rded ed the the ch chem emil ilum umin ines esce cenc ncee and and brig bright htne ness ss was me meas asur ured ed by a computer software program (Olivetti PC PRO 486/33 MHz8MbRAM). The test was always positive and intense in bones from the first group; 80% of the second group showed luminescence visible to the eye but luminescence gradually reduced in positivity and brightness as the PMI increased. After a PMI of 25 years it was negative in 70% and after 50 years only 10% were positive. Introna and colleagues suggested that the luminol test could be useful in determining the TSD from skeletal rem remain ains, s, especi especiall ally y when when only only isolat isolated ed or fra fragmen gmented ted bon bones es wer weree recovered, but it should be tested on a wider sample of bones with a closer PMI range in order to establish establish its statistical statistical significance. significance. In another study by Ramsthaler by Ramsthaler and colleagues (2009), (2009), 80 powdered bone samples from four known historical periods were tested with the naked nake d eye by two indepen independen dentt observ observers ers in ran random domized ized,, blind blind experiexperiments by chemiluminescence with luminol. Technical aids such as cameras eras and and ligh lightt dete detect ctor orss were were de deli libe bera rate tely ly not not us used ed.. The The hi hist stor oric ical al peri pe riod odss were were:: 1st 1st3r 3rd d cen centurie uriess AD, AD, 187 187819 190 02, 1965 196519 1972 72,, and and 19952003. Observer error was corrected statistically; assessing luminescence was simplified by visualization in a dark room and reported as being either present or absent. False-positive and false-negative results
Research Resear ch in the Later Stages of Decomposit Decomposition ion
111
were used as the main criteria for assessing the method. A false-positive result was the error of classifying a historical bone as recent, and a falsenegative was the error of rejecting a modern bone as being historical, a potentially worse error, as forensic cases could thereby be left unexamined. ine d. Th Thee resul results ts we were re ob obse serv rver er-i -ind ndep epen ende dent nt but un unexp expec ecte ted d re resu sult ltss occurred in 7.5% of all historical cases and 15% of recent samples were false fal se-n -neg egat ativ ives es.. A cor corre rect ct cla class ssif ific icat ation ion occu occurr rred ed in 88.75% 88.75%.. False False-posi po sitiv tivee re resu sult ltss occ occur urre red d in 7.5% 7.5% and and fa fals lsee-ne nega gati tive ve re resu sult ltss in 10% 10% of cases. Ramsthaler and colleagues concluded that chemiluminescence on its own was not adequate for distinguishing between recent and nonrecent skeletal remains. An estimation of the PMI using the luminol test alon alonee was was als also o not not po poss ssib ible le wi with thou outt in inte terp rpre reti ting ng envi enviro ronm nment ental al and anthropological factors, and additional methods of estimating the PMI should be used in conjunction with this test. A further study by Ramsthaler by Ramsthaler and colleagues (2011) compared the use of ultraviolet fluorescence, luminol, Hexagon-OBTI, and Combur test testss to dete determ rmin inee the the PM PMII in sk skel elet etal al re rema main inss by the the de dete tect ctio ion n of intraosseous hemin traces. The Hexagon-OBTI test is based on an antibody reaction to human hemoglobin and hemoglobin breakdown products such as hemin and is used to detect blood in feces. It can detect leve levels ls of 0.88 0.88 mg Hb Hb/g /g of fece feces. s. Comb Combur ur is a comm commer erci cial al chem chemic ical al incorporated in a test strip, and it is even more sensitive in detecting thee same th same hemo hemogl glob obin in pr prod oduc ucts ts than than the the Hexa Hexago gonn-OB OBTI TI te test st.. Th Thee hypothesis underlying the research was that the chemiluminescence of the luminol reaction was due to the presence of persisting hemin from hemoglobin molecules in bone and that this was related to the PMI. The researchers also wished to test whether or not the Hexagon-OBTI and the Combur methods of testing for chemiluminescence were more effe effect ctiv ive. e. Samp Sample le lo long ng bone boness from from 39 in indi divid vidual ualss with with kn know own n PMIs PMIs varying from 0 to 2500 years were tested with the three methods. UV fluorescence could exclude bones with a forensically relevant PMI of less less than than 50 year years, s, bu butt on only ly if re redu duct ctio ion n in re refl flec ecti tion on was was mark marked ed.. Eight percent of the samples with a PMI less than 100 years tested negative for the luminol reaction, while 35% with a PMI over 100 years tested positive. The sample size was small and burial conditions may have influenced the results. Ramsthaler and colleagues concluded that a negative luminol test, which indicated a PMI without forensic relevance, was superior to a
112 Human Body Decomposition
positi posi tive ve lumi lumino noll test test in indi dica cati ting ng a shor shortt PMI PMI (i (ie, e, le less ss th than an 50 year years) s) which might be forensically relevant. None of the bones tested positive to the Hexagon-OBTI or Combur tests despite all controls testing positive. The authors concluded that the Hexagon-OBTI and Combur tests were not suitable for estimating or reducing the error of estimation of the PMI at present, but they were unable to reach a satisfactory reason for the failure of these methods. They suggested that further research was required to determine whether the bone samples were dissolved for a sufficient length of time in the buffer solution used for these tests, or if the hemoglobin was sufficiently soluble in the buffer solution.
Citrate Content of Bone Recently a method of measuring the citrate content of bone as a means (2010).. Citrate, of estimating the PMI was reported by Schwarcz by Schwarcz et al. (2010) a tricarboxylic acid, is a key component of the Krebs cycle which is essential to the metabolism of all cells. About 90% of the citrate in the body is present in bone at a constant concentration of between 1.5% and 2% by weight and its purpose is thought to be to prevent excessive growth of hydroxyapatite crystals which are essential to the structure of all newly formed bone. While there are differences in concentration between different types of bone, the concentration within each type is constant in all adults regardless of sex and age. However, the citrate conc co ncen entr trat atio ion n may may de depe pend nd on the the stat statee of bone bone mine minera rali liza zati tion on and and may therefore be lower in immature and osteoporotic bone. Corresponding skeletal components in other vertebrates, such as cows and pigs, also contain almost identical citrate, showing that uptake of citr citrat atee fr from om bl bloo ood d is a gene genera rall char charac acte teri rist stic ic of al alll bone bone an and d no nott restricted to humans. Therefore animal bones could be used in studies as prox proxie iess for for huma human n bone bones. s. Sc Schw hwar arcz cz and and coll collea eagu gues es qu quot oted ed an unpublished study which found that citrate content in bone buried for more than 100 years was less than 1% of the initial content. If the loss of citrate was gradual it could be used as a forensic tool to determine the PMI and a study was designed to determine whether the citrate content of bone could be used as an estimate of the TSD. Sample Samp less of pi pig g and and huma human n ri rib b bone boness we were re pl plac aced ed in di diff ffer eren entt contexts: buried, on the ground, and under glass jars. Some of these samples were human bone forensic specimens from a number of locations and different contexts across North America. Some of the buried bones and some from under the glass jars were lifted and frozen to
Research Resear ch in the Later Stages of Decomposit Decomposition ion
113
determ dete rmin inee th thee exte extent nt to wh whic ich h the the cl clim imat atee or stor storag agee envi enviro ronm nmen entt might influence the loss of citrate. The citrate concentration of each sample was determined before the study began and in cortical bone it was found to be remarkably uniform, averaging 1.96 6 0.06 weight%. Thee citr Th citrat atee co conc ncen entr trat atio ion n decr decrea ease sed d with with re regu gula lari rity ty in a se seri ries es of samp sa mple less whic which h were were buri buried ed,, but but in samp sample less whic which h were were stor stored ed in a laboratory setting for an extended period of time, there was no loss of citrate. The authors speculated that soil organisms and moisture were re requ quir ired ed to allo allow w citr citrat atee to be cons consum umed ed.. A re regr gres essi sion on mode modell was was formulated which accounted for the regular loss of citrate concentration with time and with which, by extrapolation, the citrate concentration tion wa wass expe expect cted ed to re reac ach h zero zero conc concen entr trat atio ion n in abou aboutt 95 year years. s. Within broad limits the storage conditions of temperature, humidity, and an d dept depth h of bu buri rial al di did d not not in infl flue uenc ncee the the rate rate of lo loss ss al alth thou ough gh th thee study suggested that no citrate loss occurred when the samples had not been exposed to moist soil and there was no loss of citrate at tempera
tures C. A mathematical formula was developed relating the rate ofbelow citrate0 decrease to the PMI for ground-surface-deposited bones. (2010) concluded that the citrate content of bone could Schwarcz et al. (2010) concluded be used to estimate the PMI up to a limit of 100 years with a minimum error of 1% of the age of the bone specimen, but further testing of the method with samples contrasted with control samples of known PMI and uniform storage conditions was required. A re rece cent nt stud study y was was cond conduc ucte ted d by Kan Kanzz an and d col colle leag ague uess (20 (2014) 14) to examine the accuracy of the model proposed by Schwarcz and colleagues and to identify potential differences in the citrate degradation process. Twenty bodies which were buried in a cemetery, over the period from 1948 to 1973, were exhumed because the cemetery was being abandone do ned. d. The There were were equa equall numb numbeers of ea each ch se sex x and and th thee ages ages and and th thee PMIIs of ea PM each ch bod body were were kno known. wn. Half Half of th thee bod bodie iess were were buri buried ed in wooden coffins and half were also in wooden coffins but with the addition tio n of pl plas asti ticc bo body dy bags bags wi with th degr degrada adabl blee back backs, s, in orde orderr to co cont ntai ain n escaped gases while the bodies were in the mortuary. The bodies had been buried to a depth of 1.43.0 m and were exhumed at the same time but the length of burial varied between 52 and 27 years. All bodies had ha d larg largel ely y dec ecom ompo pose sed d and and had had been een in cont contac actt wit ith h soi soil when when exhumed. After removal the bodies were stored in dry conditions at a controlle contro lled d tem temper peratu ature re for 10 years years before before ske skelet letal al analys analysis is for cit citrat ratee conce con cent ntra rati tion on was was carri carried ed out. out. Kantz Kantz and coll collea eagu gues es post postul ulat ated ed that that
114 Human Body Decomposition
because the loss of citrate depended on the action of moisture and soil orga or gani nism sms, s, th thee conc concen entr trat ation ion woul would d have have remai remaine ned d stabl stablee duri during ng th thee 10-year 10-y ear per period iod of storage storage in dry con condit dition ions. s. The citrate citrate concen concentrat tration ion was then determined in two types of bone: from the temporal bone and from the mid-shaft of the femur. The results of the analysis of the temporal bones revealed that the PMI was underestimated in the group buried without body bags by about 19.2 years and overestimated in those buried in body bags by 5.2 years. This difference was statistically significant. When the femoral bones were analyzed, no significant difference was found between the two groups but there was an underestimation of about 16.3 years in those buried without body bags and a slight overestimation of 0.4 years in the body bag group. When the results were tested with the model developed by Schwarcz and colleagues, the accuracy was low. In discussing the results, the authors suggested that the underestimation of the PMI in the bodies buried without body bags contradicted the hypothesis that there needed to be direct contact of bone, after decomposition, with moisture for the process of citrate degradation, further postulating that the disintegration of the decomposa po sab ble back back pa pane nels ls of the pl plas asti ticc bo body dy bag agss al allo low wed the the soil soil and and humidity to enter the body bags and start the process of citrate degradation which might even be accelerated by the enclosing nature of the surrounding plastic keeping the soil humidity high. When femoral bone was used for citrate analysis, despite there being no significant difference in citrate decomposition rates between the two buried groups, decomposition was slightly faster in the body bag group because of the accelerant effect of the body bags. Decomposition was also slower with femoral bone compared with temporal bone. It was recommended therefore that bone with a thick outer cortex be used in citrate analysis and such bone might facilitate the estimation of the PMI for longer periods into the past. Because Bec ause citrat citratee degrad degradati ation on was slo slowe werr in buried buried bones bones than than sur surface face-deposited bones, different mathematical models might require to be calculated for bones deposited in different environments and also for the different types of bones analyzed. Kanz and colleagues concluded that citrate-based PMI determination had promising potential for use in forensic cases but that at present,, concen ent concentra tratio tions ns of cit citrat ratee above above 0.140 0.140 weight weight% % in well-p well-pres reserv erved ed bones might be taken as an indication that the bones were less than 50 years old in soil-buried remains.
Research Resear ch in the Later Stages of Decomposit Decomposition ion
115
Raman Spectroscopy A novel approach to the chemical investigation of buried bone as a method of determining the PMI was recently investigat gated by (2011).. Raman spectroscopy uses laser light McLaughlin and Lednev (2011) directed on to a sample. Small portions of photons are scattered which have a different energy than the light source. A Raman spectrum is obtained by detecting these inelastically scattered photons. The energy difference, or Raman shift, between the laser source and the scattered photon pho tonss corres correspon ponds ds to mol molecu ecular lar vib vibrat ration ional al models models,, with with dif differ ferent ent mole mo lecu cule less pr prod oduc ucin ing g di diff ffer eren entt vibr vibrat atio ions ns.. A Rama Raman n spec spectr trum um can can therefore be used to give information about the identity, structure, and properties of various materials based on their vibrational transitions. Fragments of a turkey leg bone, after removal of all flesh, were buried in soil and a different fragment removed and examined spectrosc scop opic ical ally ly at 2-we 2-week ek in inte terv rval alss up to 68 days days.. Thes Thesee fragm fragmen ents ts were were compared with a fragment which had not been buried and a fragment treated with collagenase solution. During burial there was a gradual loss of the mineral phase of the bone, exposing the collagen phase to hydrolysis. It was found that as burial time increased the Raman spectrum of certain collagenous organic molecules decreased. This was not seen with unburied bone but there were similarities between the buried bone bo ne sa samp mple less and and the the co coll llag agen enas asee-tr trea eate ted d samp sample le.. McLau McLaughl ghlin in and and Lednev in commenting on these promising results suggested research shou sh ould ld be cont contin inue ued d to deve develo lop p a Rama Raman n spec spectr tros osco copi picc mode modell to extrapolate bone burial duration and which factored in the variables of soil pH, burial temperature, humidity, and the presence of soft tissue. The study shouldwith alsolonger be extended to ascertain whether the observed trends continued burial time.
Carbon Dating and the Use of Naturally Occurring and Artificial Radioactive Isotopes Radioa Radi oact ctiv ivee isot isotop opes es oc occu curr natu natura rall lly y in ever every y envi enviro ronm nmen entt an and d can can also also be pr prod oduc uced ed ar arti tifi fici cial ally ly.. The The type type and conc concen entr trat atio ion n of na natu tura rall radi ra dioa oact ctiv ivee isot isotop opes es such such as ra rado don n (226Ra Ra), ), polo poloni nium um (210Po Po), ), lead lead (210Pb Pb)), uran uraniu ium m (238U, 235U, 232U), and thorium (232Th) have a regula reg ular, r, pre predic dictab table le backgr backgroun ound d concen concentra tratio tion n which which has remain remained ed unchanged for thousands of years and which have been shown not to produce identifiable increases when ingested into the body in industrial environments where they are more prevalent (Smith ( Smith et al., 2001). 2001). The
116 Human Body Decomposition
commonest natural radioisotope is Carbon 14 (14C), which becomes rapidly oxidized in the atmosphere to form radioactive carbon dioxide (14CO2). It enters the food chain by the processes of solution in the ocean oceanss and and photo photosyn synthe thesis sis in pl plant ants. s. It be becom comes es incorp incorpora orated ted in bone through dietary ingestion of plant and animal material and its rate of decay can be measured to give an estimate of the age of anci an cien entt bo bone ness up to 50,0 50,000 00 ye year arss BP (Kni Knight ght and Lau Lauder der,, 196 1969 9), although estimating age as old as this was later disputed ( Van der Plicht, 2004). 2004). Anderson Anderson and colleagues (1947) detected (1947) detected and identified 14 C. They later calculated the half-life of 14C to be 5720 6 47 years but made the false assumption that radiocarbon in the environment had remained constant over millions of years so that modern values could cou ld be correl correlate ated d to in initi itial al concen concentra tratio tions ns in hi histo storic rical al materi materials als 14 y et al al., ., 19 1949 49). ). Subse Subseque quentl ntly y it was found found that that C does not (Libb Libby decay in a regular manner; a radiocarbon age is not equivalent to a cale ca lend ndar ar ag agee as fluc fluctu tuat atio ions ns of atmo atmosp sphe heri ricc 14C concentr concentrati ation, on, known de 19 Vries have occurred the centuries (Tayl Taylor orasetthe al., al ., 1989 89). ). effects, A larg large e fluc fluctu tuat atio ion n throughout of 14C occu occur rre red d in the the period peri od 11,0 11,000 0014 14,5 ,500 00 BP due due to chan change gess in the the prod produc ucti tion on ra rate te,, caused by geomagnetic and solar modulation of the cosmic-ray flux and the carbon cycle, although it has also been suggested that the fluctuation was due to carbon cycle changes tied to deep ocean currents (Fairbanks (Fairbanks et al., 2005 2005). ). Hence, a calibration is required, which, to be accurate and precise, should ideally be based on an absolutely dated record that has carbon incorporated directly from the atmosphere at the time of formation. In the recent past, two major fluctuations in 14C levels have been documented for the years 19001982, both of which had an anthropoTayl ylor or et al al., ., 19 1989 89). ). Af Afte terr the the onse onsett of the the Indu Indust stri rial al geni ge nicc ca caus usee (Ta Revolution, burning of fossil fuel increased and because it was millions of years old the 14C in fossil fuel had almost completely decayed. As a result the 14C in the biosphere became diluted by the non-radioactive isotopes of carbon. This was particularly noted in the period 1910 50 when the amount of 14C in the environment decreased on average by Suess or fossil fuel effect (Taylor 3%. This fluctuation is known as the Suess or fossil et al., 1989). 1989). From about 1950 there was a large increase of 14C in the at atmo mosp sphe here re as a re resu sult lt of nu nucl clea earr weap weapon onss te test stin ing. g. This This ente entere red d the the bi bios osph pher eree and and was was even eventu tual ally ly in inco corp rpor orat ated ed in bone bone.. Atmo Atmosp sphe heri ricc nuclear testing ceased in 1963, but the atmospheric 14C did not drop to
Research Resear ch in the Later Stages of Decomposit Decomposition ion
117
pre-1950 levels until about the mid-1990s. This second recent fluctuation is known as the Libby or bomb bomb ef effe fect ct (Tayl Taylor or et al al., ., 19 1989 89). ). Atte At temp mpts ts we were re ma made de to corr correc ectt the de Vrie Vriess ef effe fect cts, s, whic which h were were prevalent before the anthropogenic effects of the 20th century and the anomalies produced by 20th century fossil fuels and nuclear testing, by correlating 14C leve levels ls in th thee en envi viro ronm nmen entt wi with th 14C in tree rings of known age from around the world. Three categories of classifying 14C in bone material were proposed by Taylor by Taylor et al. (1989): (1989): 1. Nonmod Nonmodern ern befor beforee AD 1650. 2. Premod Premodern ern from from AD 1650 to 1950. 3. Modern Modern from from AD 1950 to the present present tim time. e. Howeve Howe ver, r, the the erro errorr rate rate in the the se seco cond nd ca cate tego gory ry wa wass 6 300 years years.. Since the work of Taylor and colleagues (1989), attempts have continued ue d to bett better er esti estima mate te the the date date of mate materi rial al by prod produc ucin ing g corr correc ecti tive ve 14 calib cal ibrat ratio ion n curve curvess for the var vario ious us geogra geograph phica icall areas areas.. The C in the atmosphere is incorporated into the annual growth ring of any nearby tree. Although it is rare to find trees older than about 12,000 years BP, analysis of 14C in marine corals, foraminifera, plant macrofossils, and speleothems both in the northern and southern hemispheres, can enable estimation of the age of bones and other material by radiocarbon dating to be extended before this time period (Fairbanks (Fairbanks et al., 2005 2005). ). At the present time new calibration curves are produced every few years as new materi material al bec become omess availa available ble.. Wit With h the prese present nt cal calib ibrat ratio ion n curve curve,, devised in 2013, it is possible to date material as far back as 50,000 years BP (Reim Reimeer et al al., ., 20 2013 13). ). Th Thee adve advent nt of ac acce cele lerrated ated mas mass spectrometry has also made it easier to carbon date, as much smaller quantities of material can be used and the measurement time has been Fair irb ban anks ks et al al., ., 20 2005 05)). Whe When reduced from weeks to minutes (Fa radiocarbon dating is reported in years BP , BP (Before Present) refers to the number of years before 1950, the beginning of nuclear weapons testin tes ting g (Ubel Ubelaker aker,, 2014 2014). ). In recent times since the period of elevated atmospheric 14C re resu sult ltin ing g from from nucl nuclea earr we weap apon onss te test stin ing, g, it has has been been foun fo und d that that the the in inco corp rpor orat atio ion n of 14C in into to the the di diff ffer eren entt bo body dy ti tiss ssue uess and its elimination from them varies greatly depending on the turnover rate ra te,, mak makin ing g its us usee in forens rensic ic case casess di diff ffic icu ult (Ubela Ubelaker ker,, 201 2014 4). Ubelaker and others have shown that by using the known elevation of 14 C in vari variou ouss bod body tis tissues sues at di diff ffeerent rent ti tim me peri perio ods duri durin ng the the increased levels which occurred from 1963 to the mid-1990s, it may be “
”
118 Human Body Decomposition
of some use in recent forensic cases (Ubelaker, ( Ubelaker, 2014; Wild et al., 2000; Ubelaker and Parra, 2011; Ubelaker et al., 2015). 2015 ). Artificial radioactive isotopes, which have greatly increased in the natu na tura rall en envi viro ronm nmen entt si sinc ncee nucl nuclea earr test testin ing g began began in 1945, 1945, have have been been investigated as a means of estimating the PMI because they are less af affe fect cted ed by chan change gess with within in the the phys physic ical al envi enviro ronm nmen entt to wh whic ich h the the bones have been exposed than are methods which depend on chemical Maclaughlin-Black and colleagues (1992) inves(1992) inveschanges (Swift, (Swift, 2006). 2006). Maclaughlin-Black tigated the use of Strontium-90 (90Sr) in order to distinguish modern bone from bone deposited before nuclear testing began. Strontium and calcium, both alkaline earth metals, have a similar structure but, unlike calcium, which is an essential mammalian nutrient and is incorporated in bone, strontium has no known metabolic function. Strontium is not incorporated in the body unless calcium is deficient and then only in a lesser concentration than calcium. Both are taken up in the body by absorption through the intestine and incorporated in the hydroxyapatite matrix of bone. 90Sr, the radioactive isotope of strontium, greatly incr increa ease sed d in the the atmo atmosp sphe here re afte afterr nucl nuclea earr te test stin ing g bega began n in 1945 1945,, became distributed worldwide, and reached a peak in the early 1960s. It was eventually taken up by vegetation, entered the food chain, and gradually became incorporated in human bone. It was suggested that if a date could be defined when 90Sr entered the food chain, people who di died ed be befo fore re that that date date sh shou ould ld not not poss posses esss dete detect ctab able le le leve vels ls of 90Sr in their bones, whereas people dyi yin ng after that date would have detectable levels. 90Sr levels were measured in three medieval femora and an d comp compar ared ed wi with th leve levels ls in thr three mo mode dern rn postm ostmor orte tem m fe femo mora ra (Mac Maclau laughl ghlinin-Bla Black ck et al. al.,, 1992 1992). ). A significantly higher level of 90Sr was found in the modern bones compared with the medieval bones, butt a smal bu smalll am amou ount nt of 90Sr stil stilll pr pres esen entt in the the medi mediev eval al bone boness wa wass attributed to diagenesis which the authors defined as the postmortem exch ex chan ange ge of chem chemic ical alss be betw twee een n the the skel skelet eton on and and it itss surr surrou ound ndin ing g medium med ium.. Labora Laborator tory y contami contaminat nation ion of archeo archeolog logica icall specim specimens ens and possible contamination from surrounding soil was not accounted for in this th is st stud udy y and and Macl Maclau augh ghli linn-Bl Blac ack k and and coll collea eagu gues es admi admitt tted ed that that a larger study was required to confirm the findings. They concluded that it was possible that the increased concentration of 90Sr in the atmosphere since the 1960s and its incorporation in the diet was the reason for its for its incr increa ease sed d conc concen entr trat atio ion n in th thee mo mode dern rn bo bone ne samp sample les. s. The The method might prove to be of some value in determining whether an
Research Resear ch in the Later Stages of Decomposit Decomposition ion
119
indi indivi vidu dual al was aliv alivee befo before re or afte afterr the the ti time me peri period od when when arti artifi fici cial al radionuclides were introduced into the food chain. In a si simi mila larr stud study, y, 90Sr leve levels ls in ni nine ne samp sample less of oc occi cipi pita tall bone bone from skulls, three from the years 1931 and 1932, and the rest from the years 1989 to 1994 were analyzed by Neis and colleagues (1999) (1999).. All the samples were from bodies autopsied in a university forensic pathology department and none had been buried. There was no significant concentration of 90Sr in three samples of bone from 1931/32, befo be fore re nu nucl clea earr test testin ing g bega began. n. Ho Howe weve ver, r, no none ne of th thes esee samp sample less had had been subjected to soil diagenesis. In the group of six samples from 1989 to 1994, a single sample exhibited a small concentration of 90Sr, expl ex plai aine ned d by the the fa fact ct that that sk skel elet etal al grow growth th had had al alre read ady y been been co commpleted before the nuclear fallout. Another group of two samples with a mod modera erate te con concen centra trati tion on of 90Sr was was due due to skele keleta tall grow growth th no nott being completed before the fallout and the last group of three sampl ples es had had a high igh conc concen entr trat atio ion n of 90Sr beca becaus usee the the date datess of bi birt rth h occurred during the fallout. This was a study of a small number of sample sam ples, s, none none of whi which ch had been been buried buried and th thee author authorss su sugge ggeste sted d that the major problem was to determine the cut-off year for measurable 90Sr in skeletal remains, taking into consideration that the biologica log icall half-l half-life ife of 90Sr was 7.518 year yearss an and d the the phys physic ical al half half-l -lif ifee was 28.1 years. Another problem concerned the estimation of 90Sr in buried bodies. Most bodies were buried between 1 and 5 m below the surface, 90Sr activity could never reach more than 80 cm below the surface and its rate of penetration was only 1 cm per year. Swi Swift ft (199 (1998) 8) investi investigat gated ed the use of natura naturally lly occurr occurring ing radioa radioacti ctive ve polonium (210Po) and radioactive lead (210Pb), both part of the radon (226Ra) se seri ries es,, for for th thei eirr us usef eful ulne ness ss in es esti tima mati ting ng the PMI. PMI. 210Po is a decay product of 210Pb and neither is produced by nuclear explosions. They are mainly inhaled and ingested from food and although 1015% of 210Pb is incor ncorpo porrated ated in bone one in a few hour ours, up upta take ke is slo low. w. Eventually 90% of all lead found in the body is incorporated in bone. 210 Po has a half-life of 138.4 days and the half-life of 210Pb is 22.3 years. The ratio between the two in life is in equilibrium but after death it alters as the two isotopes decay in an exponential manner and this, theoretically, should allow a predictable PMI to be calculated from the altering ing rat atio io by re refe fere renc ncee ba back ck to the the kno known rati ratio o in li life fe.. Swif wift tes este ted d various bone samples for 210Po and 210Pb and concluded that the ratio
120 Human Body Decomposition
between the two was readily measurable, allowing quantities to be ascertained accura accurately. tely. 210Pb was also abundant naturally and given its halflife of 22.3 years, would be a useful radioisotope for forensic use. One disadvantage was that shellfish consumption and smoking were known to affect 210Pb levels. Other disadvantages were the expense of analysis, possible variation in different parts of the skeleton, individual variation in lead lead meta metabo boli lism sm,, and and the the effe effect ctss of di diag agen eneesi sis. s. Swif Swiftt sugg sugges este ted d a larger study of ancient and modern bones in which the time of death was known, together with the testing of soil samples to gauge the effects of diagenesis, in order to determine the usefulness of the estimation of 210 Pb in bone. A number of radioisotopes as well as trace elements were investigated for their potential use as indicators of PMI by Swift and colleagues (2001) in a pilot study. The authors listed the requirements of a radioisotope if it was to be of forensic use as: 1. To have have a half-l half-life ife commensu commensurat ratee wit with h the time scale of invest investiga iga-tion, that is, less than 40 years. 2. To be abun abunda dant nt enou enough gh to be dete detect cted ed by co conv nven enti tion onal al anal analyt ytic ical al techniques. 3. To have ave so some me bi biol olog ogiical cal func functi tio on so as to be inco incorp rpor orat ated ed in human bone. Fifteen samples of skeletal material exhumed from soil burials in a Port Po rtug ugue uese se ceme cemete tery, ry, wi with th PM PMIs Is ra rang ngin ing g from from 15 to 77 year years, s, were were treated and analyzed for various radionuclides. There was a correlation between the PMI and concentrations of 238U, 234U, 210Po, and 210Pb from the U r238 anium 239serie s240as well as with t h137 e non-natural radionuclides Pu, Pu, Pu (Plu (Pluto toni nium um), ), an and d Cs (Ce (Cesiu sium). m). Trace element analysis revealed intercorrelations as well as correlations with wi th th thee TSD. TSD. Swif Swiftt co conc nclu lude ded d that that thes thesee radi radion onuc ucli lide dess an and d trac tracee elem elemen ents ts co coul uld d be used used to give give some some quan quanti tita tati tive ve data data re rega gard rdin ing g the postmortem interment period but further studies were required to create cre ate a pop popula ulatio tion-s n-spec pecifi ificc cal calibr ibrati ation on scale, scale, given given the var variat iation ion in isotope exposures in different countries with different geologic strata and different diets. At the present time the status of estimation of the PMI by the use of natu natura rall lly y oc occu curr rrin ing g radi radioi oiso soto tope pess and and trac tracee el elem emen ents ts ha hass ma main inly ly been of impo been import rtan ance ce in di dist stin ingu guis ishi hing ng arch archeo eolo logi gica call from from fo fore rens nsic ic
Research Resear ch in the Later Stages of Decomposit Decomposition ion
121
material, but the recent work by Ubelaker may extend its usefulness to remains of more recent forensic interest (Ubelaker, (Ubelaker, 2014; Ubelaker and Parra, Par ra, 201 2011 1). Fu Futu ture re re rese sear arch ch wi will ll depe depend nd on th thee in infl flue uenc ncee of soil soil characteristics in different soil types, the leaching by exposure of bones to water water,, the the co comp mpar aris ison on of radi radioi oiso soto tope pe conc concen entr trati ation on in di diff ffer eren entt bone bo ne type types, s, and and the the effe effect ct of di diet et and and ot othe herr li life fest styl ylee in infl flue uenc nces es on radioisotope uptake in bone (Howard, ( Howard, 2008). 2008).
CONCLUSIONS In the latter stages of decomposition, factors such as burial environment and insect activity play a major role in the nature and speed of decomposition. One of the key variables in decomposition, temperature, will be modified by soil and water (eg, sea, lake, river) contexts, and an d de deco comp mpos osit itio ion n will will tend tend (d (dep epen endi ding ng on seas season on)) to be fast faster er above abo ve the surf surfac acee th than an below below.. Part Partic icula ularr buria buriall cont contex exts ts ca can n also also si signi gnifi fica cant ntly ly delay delay the the rate rate of decom decompo posi siti tion on throu through gh natur natural al proprocesses of mummification or saponification. Insects have a major role in determining the rate of decomposition and their activity will also be regulated by temperature. While the use of insects in forensic contexts has a long history, the study of insect life cycles and carcass colonization successions only developed into the science of forensic entomology in the 20th century. It is often the chief tool for estimating the TSD in the post-48-hour postmortem period. Considering the estimation of the PMI in skeletonized remains, a range of methods, forr the fo the most most pa part rt di dict ctat ated ed by cont contex ext, t, are are avai availa labl ble. e. Some Some of the the meth me thod odss revi review ewed ed incl includ udee exam examin inat atio ion n of as asso soci ciat ated ed bota botani nica call remains, assessment of residual soft tissues, histological examination of bone bone degr degrad adat atio ion, n, phys physic icoc oche hemi mica call chan change gess to bone bone (wit (with h th thee citrate concentration in bone having some potential value in estimating the PMI) MI), as wel well as radioactiv ivee isotope concentrat rations. However, the error ranges for the majority of these approaches are uncomfortably uncomf ortably large.
REFERENCES Adlam, R.E Adlam, R.E., ., Sim Simmon mons, s, T., 200 2007. 7. The eff effect ect of rep repeat eated ed phy physic sical al dis distur turban bance ce on sof softt tis tissue sue decomposition are taphonomic studies an accurate reflection of decomposition? J. Forensic Sci. 52 (5), 10071014. Amendt, J., et al., 2007. Best practice in forensic entomology standards and guidelines. Int. J. Legal Med. 121, 90104.
122 Human Body Decomposition Amy, R., et al., 1986. The last Franklin expedition: report of a post-mortem examination of a crew member. Can. Med. Assoc. J. 135, 115117. Anderson, Anders on, E.C E.C., ., Lib Libby, by, W.F W.F., ., et al. al.,, 194 1947. 7. Nat Natura urall rad radioc iocarb arbon on fro from m cos cosmic mic rad radiat iation ion.. Phy Phys. s. Rev. 72 (10), 931936. Anderson, G.S., 2000. Minimum and maximum development rates of some forensically important Calliphoridae (Diptera). J. Forensic Sci. 45 (4), 824 832. Anderson, G.S., 2011 Anderson, 2011.. Compa Comparison rison of deco decomposi mposition tion rates and fauna faunall colon colonizat ization ion of carri carrion on in indoor and outdoor environments. J. Forensic Sci. 56 (1), 136142. Anderson, Anders on, G.S G.S., ., Van Lae Laerhov rhoven, en, S.L S.L., ., 199 1996. 6. Ini Initia tiall stu studie diess on ins insect ect suc succes cessio sion n on car carrio rion n in south western British Columbia. J. Forensic Sci. 41 (4), 617625. Archer,, M.S Archer M.S., ., 200 2004. 4. The eff effect ect of tim timee aft after er bod body y dis discov covery ery on the acc accura uracy cy of ret retros rospec pectiv tivee weather station ambient temperature corrections in forensic entomology. J. Forensic Sci. 49 (3), 17. Archer,, M.S Archer M.S., ., Elg Elgar, ar, M.A M.A., ., 200 2003. 3. Yea Yearly rly act activi ivity ty pat patter terns ns in sou southe thern rn Vic Victor toria ia (Au (Austr strali alia) a) of seasonally active carrion insects. Forensic Sci. Int. 132, 173 176. Arnold, C.Y., 1959. The determination and significance of the base temperature in a linear heat system. Proc. Am. Soc. Hortic. Sci. 74, 430445. Aufderheide, A.C., 1981. Soft tissue paleopathology an emerging specialty. Hum. Pathol. 12 (10), 865867. Bachmann, M.D., Simmons, T., 2010. The influence of pre-burial insect access on the decomposition rate. J. Forensic Sci. 55 (4), 893 900. Behrensme Behren smeye yer, r, A.K A.K., ., 197 1978. 8. Tap Taphon honomi omicc and ec ecolo ologic gic inf inform ormati ation on fro from m bon bonee wea weathe therin ring. g. Paleobiology 4 (2), 150162. Bell, L.S Bell, L.S., ., Ski Skinne nner, r, M.F M.F., ., Jon Jones, es, S.J S.J., ., 199 1996. 6. The spe speed ed of post post-mo -morte rtem m cha change nge to the hum human an skeleton and its taphonomic significance. Forensic Sci. Int. 82, 129 140. Benecke, Bene cke, M., 2001. A brief history of foren forensic sic entomology. entomology. Forensic Forensic Sci. Int. 120, 214. Breitmeir, D., et al., 2005. Evaluation of the correlation between time corpses spent in ground gravess and findings at exhum grave exhumatio ation. n. Foren Forensic sic Sci. Int. 154, 218233. Campobasso, C.P., Di Vella, G., Introna, F., 2001. Factors affecting decomposition and Diptera colonisation. Forensic Sci. Int. 120, 18 27. Carter, D.O., Yellowlees, D., Tibbett, M., 2010. Moisture can be the dominant environmental parameter governing cadaver decomposition in soil. Forensic Sci. Int. 200, 60 66. Cattaneo, Catta neo, C., Gels Gelsthorpe thorpe,, K., Phillips, Phillips, P., Sokol, R.J. R.J.,, 1992. Reliable Reliable ident identific ificatio ation n of human albumin in ancient bone using ELISA and monoclonal antibodies. Am. J. Phys. Anthropol. 87, 365372. Courtin, Courti n, G.M G.M., ., Fai Fairgr rgriev ieve, e, S.I S.I., ., 200 2004. 4. Eva Evalua luatio tion n of pos post-m t-mort ortem em int interv erval al (PM (PMI) I) as rev reveal ealed ed through the analysis of annual growth in woody tissue. J. Forensic Sci. 49 (4), 1 3. Dent, B.B., Forbes, S.L., Stuart, B.H., 2004. Review of human decomposition processes in soil. Environ. Envir on. Geol. 45 (4), 576585. Edwards, R., Chaney, W., Berg Edwards, Bergman, man, M., 1987. Temperature Temperature developmen developmental tal units for insects. insects. Pest Crop Newslett. Purdue Univ. 2, 56. Fac Facchi chini, ni, F., Pe Pette ttener ner,, D., 197 1977. 7. Che Chemic mical al and phy physic sical al met method hodss in dat dating ing hum human an ske skelet letal al remains. Am. J. Phys. Anthropol. 47, 65 70. 2382005. Radiocarbon Fairbanks, R.G., al., curve spanning 0 to 24, 50,000 BP based on paired 230 Th/ 234etU/ U and 14C dates on calibration pristine corals. Quat. Sci. Rev. 1781years 1796.
Research Resear ch in the Later Stages of Decomposit Decomposition ion
123
Forb rbes es,, S. S.L. L.,, 20 2008 08.. Fo Fore rens nsic ic ch chem emis istr try: y: ap appl plic icat atio ions ns to de deco comp mpos osit itio ion n an and d pr pres eser erva vati tion on.. Fo In: Oxenham, M. (Ed.), Approaches to Death, Disaster and Abuse. Australian Academic Press, Brisbane, Ch. 15. Forbes, S.L., Stuart, B.H., Dadour, I., Dent, B.B., 2004. A preliminary investigation of the stages of adipocere formation. J. Forensic Sci. 49 (3), 19. Forbes, S.L., Stuart, B.H., Dent, B.B., 2005. The effect of the burial environment on adipocere formation. Forensic Sci. Int. 154, 24 34. Galloway, A., et al., 1989. Decay rates of human remains in an arid environment. J. Forensic Sci. 34 (3), 607616. Gill-King, H., 1997. Chemical and ultrastructural aspects of decomposition. In: Haglund, W.D., Sorg, Sor g, M.H M.H.. (Ed (Eds.) s.),, For Forens ensic ic Tap Taphon honomy omy:: The Pos Postt Mor Mortem tem Fat Fatee of Hum Human an Rem Remain ains. s. CRC Press, Boca Raton, pp. 93108. Goff, M.L., 1991. Comparison of insect species associated with decomposing remains recovered inside dwellings and outdoors on the Island of Oahu, Hawaii. J. Forensic Sci. 36, 748 753. Goff, M.L., 1993. Estimation of post-mortem interval using arthropod development and successional patterns. Forensic Sci. Rev. 5 (2), 8294. Goff, M.L., Flynn, M.M., 1991. Determination of post-mortem interval by arthropod succession: a case study from the Hawaiian Islands. J. Forensic Sci. 36 (2), 607 614. Gotouda, H., et al., 1988. The Mechanism of experimental adipocere formation: hydration and dehydr deh ydroge ogenat nation ion in mic microbi robial al syn synthe thesis sis of hyd hydrox roxy y and oxo fat fatty ty aci acids. ds. For Forens ensic ic Sci Sci.. Int Int.. 37, 249257. Grassb Gras sber erge ger, r, M. M.,, Re Reit iter er,, C. C.,, 20 2001 01.. Ef Effe fect ct of te temp mper erat atur uree on Lucili Lucilia a seric sericata ata (Diptera: Calliphoridae) development with special reference to the isomegalen- and isomorphen- diagram. Forensic Sci. Int. 120, 3236. Harvey, M.L., Mansell, M.W., Villet, M.H., Dadour, I.R., 2003. Molecular identification of some forensically important blowflies of southern Africa and Australia. Med. Vet. Entomol. 17, 363 369. Hedges, R.E.M., 2002. Bone diagenesis: an overview of processes. Archaeometry 44, 319 328. Hoke, N., et al., 2011. Estimating the chance of success of archaeometric analyses of bone: UV induce ind uced d bon bonee flu fluore oresce scence nce com compar pared ed to his histol tologi ogical cal scr screen eening ing.. Pal Paleog eogeog eogr. r. Pal Paleoc eoclim limat atol. ol. Paleoecol. 310, 2331. Hoke, N., Gri Hoke, Grigat gat,, A., Gru Grupe, pe, G., Har Harbec beck, k, M. M.,, 201 2013. 3. Rec Reconsi onsider derati ation on of bon bonee pos post-m t-mort ortem em interval estimation by UV-induced auto fluorescence. Forensic Sci. Int. 228, 176e1 176e6. Howard, S., 2008. A Preliminary Investigation into the Estimation of Time since Death from Human Hum an Skelet Skeletal al Remain Remainss by Rad Radioi ioisot sotope ope and Tra Trace ce Elemen Elementt Analys Analysis. is. Master Master of Forens Forensic ic Science Thesis. University of Western Australia. Introna, F., Di Vella, G., Campobasso, C.P., 1999. Determination of post-mortem interval from old skeletal remains by image analysis of luminol test results. J. Forensic Sci. 44 (3), 535538. Jarvis,, D.R Jarvis D.R., ., 199 1997. 7. Ni Nitro trogen gen lev levels els in lon long g bon bones es fro from m cof coffin fin bur burial ialss int interr erred ed for per period iodss of 2690 years. Forensic Sci. Int. 85, 199208. Johnson, Johnso n, A.P A.P., ., Wal Wallma lman, n, J. J.F., F., Arc Archer her,, M.S M.S., ., 201 2012. 2. Exp Experi erime menta ntall and ca casew sework ork val valida idatio tion n of ambient temperature corrections in forensic entomology. J. Forensic Sci. 57 (1), 215 221. Kanz, F., Reiter, C., Risser, D., 2014. Citrate content of bone for time since death estimation: resu re sult ltss fr from om bu buri rial alss wi with th di diff ffer eren entt ch char arac acte teri rist stic icss an and d kn know own n PM PMI. I. J. Fo Fore rens nsic ic Sc Sci. i. 59 (3 (3), ), 613620. Knight, B., Lauder, I., 1967. Practical methods of dating skeletal remains: a preliminary study. Med. Sci. Law 7, 205209. Knight, B., Lauder, I., 1969. Methods of dating skeletal remains. Hum. Biol. 41 (3), 322 341.
124 Human Body Decomposition Komar, D.A., 1998. Decay rates in a cold climate region: a review of cases involving advanced decomposition from the Medical Examiner s Office in Edmonton, Alberta. J. Forensic Sci. 43 (1), 5761. ’
Libby, W.F. Libby, W.F.,, Ander Anderson, son, E.C., Arnold Arnold,, J.R. J.R.,, 1949 1949.. Age dete determina rmination tion by radioc radiocarbon arbon content: worldwide assay of natural radiocarbon. Science 109 (2827), 227 228. Lowenstein,, J.M. Lowenstein J.M.,, 1981 1981.. Immun Immunolog ological ical reactions reactions from fossil mate material. rial. Philos. Trans. R. Soc. Lond. B 292, 143149. Maclaughlin-Black, S.M., et al., 1992. Strontium-90 as an indicator of time since death: a pilot investigation. Forensic Sci. Int. 57, 51 56. Mann, R.W Mann, R.W., ., Ba Bass, ss, W. W.M. M.,, Me Meado adows, ws, L. L.,, 19 1990. 90. Ti Time me sin since ce de death ath an and d de decom compos positi ition on of th thee human body: variables and observations in case and experimental field studies. J. Forensic Sci. 35 (1), 103111. Matuszewski, S., et al., 2008. An initial study of insect succession and carrion decomposition in various vario us forest habitats habitats of Centr Central al Europe. Forensic Sci. Int. 180, 6169. Matuszewski, S., et al., 2010a. Insect succession and carrion decomposition in selected forests of Central Centr al Europe. Part I: pattern and rate of decom decomposit position. ion. Forensic Forensic Sci. Int. 194, 8593. Matuszewski, S., et al., 2010b. Insect succession and carrion decomposition in selected forests of Central Europe. Part II: composition and residency patterns of carrion fauna. Forensic Sci. Int. 195, 4251. McLaughlin, G., Lednev, I.K., 2011. Potential application of Raman spectroscopy for determining burial duration of skeletal remains. Anal. Bioanal. Chem. 401, 2511 2518. Neis, P., et al., 1999. Strontium 90 for determination of time since death. Forensic Sci. Int. 99, 4751. Notter, S.J., Stuart, B.H., Rowe, R., Langlois, N., 2009. The initial changes of fat deposits during the decomposition of human and pig remains. J. Forensic Sci. 54 (1), 195 201. O Bri Brien, en, T.G T.G., ., Kue Kuehne hner, r, A.C A.C., ., 200 2007. 7. Wax Waxing ing gra grave ve abo about ut adi adipoc pocere ere:: sof softt tis tissue sue cha change nge in an aquatic environment. J. Forensic Sci. 52 (2), 294 301. ’
Payne, J.A., 1965. A summer carrion study of the baby pig Sus scrofa Linnaeus. scrofa Linnaeus. Ecology 46 (5), 592602. Perper,, J.A Perper J.A., ., 200 2006. 6. Tim Timee of dea death th and cha change ngess aft after er dea death. th. In: Spi Spitz, tz, W.U. (Ed (Ed.), .), Spitz and Fisher s Medic Medico o Lega Legall Invest Investigati igation on of Deat Death, h, fourt fourth h ed. Charles C Spring Springer, er, Springfield, Springfield, Illinois, Illinois, pp. 107108. Ch. 3. ’
Pohjoismäki, Pohjoismäk i, J.L. J.L.O., O., et al., 2010. Indoors forensic entomology: entomology: colonisation colonisation of human remains in closed environments by specific species of sarcophagous flies. Forensic Sci. Int. 199, 3842. Prieto,, J. Prieto J.L., L., Mag Magaña aña,, C., Ube Ubelak laker, er, D.H D.H., ., 200 2004. 4. Int Interp erpret retati ation on of pos post-m t-mort ortem em cha change nge in cadavers in Spain. J. Forensic Sci. 49 (5), 1 6. Ramsthaler, F., Kreutz, K., Zipp, K., Verhoff, M.A., 2009. Dating skeletal remains with luminolchemiluminescence. Validity, intra- and interobserver error. Forensic Sci. Int. 187, 47 50. Ramsthaler,, F., Ebac Ramsthaler Ebach, h, S.C. S.C.,, Birng Birngruber ruber,, C.G. C.G.,, Verhof Verhoff, f, M.A. M.A.,, 2011 2011.. PostPost-morte mortem m inter interval val of skel sk elet etal al re rema main inss th thro roug ugh h th thee de dete tect ctio ion n of in intr trao aoss ssea eall ha haem emin in tr trac aces es.. A co comp mpar aris ison on of s s UV-fluorescence, luminol, Hexagon-OBTI and Combur tests. Forensic Forensic Sci. Int. 209, 59 63. Réaumur, Réaumu r, R.A R.A.F. .F.,, 173 1735. 5. Obs Observ ervati ations ons du thé thérmo rmomèt mètre re fai faites tes à Par Paris is pen pendan dantt l année 1735 1735,, comparées avec celles qui ont été faites sous la ligne; à l Isle de France, à Alger et en quelquesunes de nos Isles de l Amérique. Mémoires de l Académie des Sciences 545576. ’
’
’
’
Reeves, N.M., 2009. Taphonomic effects of vulture scavenging. J. Forensic Sci. 54 (3), 523 528. Reimer, P.J., et al., 2013. INTCAL13 and Marine 13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon 55 (4), 18691887.
Research Resear ch in the Later Stages of Decomposit Decomposition ion
125
Rivers ers,, D.B D.B., ., Tho Thomps mpson, on, C., Bro Brogan gan,, R., 201 2011. 1. Physio Physiolog logica icall tra tradede-off offss of for formin ming g mag maggot got Riv masses by necrophagous flies on vertebrate carrion. Bull. Entomol. Res. 101, 599611. Rodriguez, W.C., Bass, W.M., 1983. Insect activity and its relationship to decay rates of human cadavers cadav ers in East Tennessee. Tennessee. J. Foren Forensic sic Sci. 28 (2), 423432. Rodriguez, W.C., Bass, W.M., 1985. Decomposition of buried bodies and methods that may aid in their location. J. Forensic Sci. 30 (3), 836 852. Schultz, M., 1997. Microscopic Schultz, Microscopic investigation investigation of exca excavate vated d skele skeletal tal rema remains: ins: a contri contribution bution to pala pa laeo eopa path thol olog ogy y an and d fo fore rens nsic ic me medi dici cine ne.. In In:: Ha Hagl glun und, d, W. W.D. D.,, So Sorg rg,, M. M.H. H. (E (Eds ds.) .),, Fo Fore rens nsic ic Taphonomy Tapho nomy:: The Postmortem Postmortem Fate of Human Remai Remains. ns. CRC Press, Boca Raton, Ch. 14. Schwarcz, H.P., Agur, K., Jantz, L.M., 2010. A new method for determination of post-mortem interval: inter val: citrate citrate conte content nt of bone. J. Foren Forensic sic Sci. 55 (6), 15161522. Simmons, T., et al., 2010a. The influence of insects on decomposition rate in buried and surface remains. J. Forensic Sci. 55 (4), 889892. Simmons, T., Adlam, R.E., Moffatt, C., 2010b. Debugging decomposition data comparative taph ta phon onom omic ic st stud udie iess an and d th thee in infl flue uenc ncee of in inse sect ctss an and d ca carc rcas asss si size ze an and d de deco comp mpos osit itio ion n ra rate te.. J. Forensic Sci. 55 (1), 813. Sledzik, P.S., Micozzi, Sledzik, Micozzi, M.S., 1997. Autopsied, embalmed and preserv preserved ed human remains: distinguishing distinguishing feat fe atur ures es in fo fore rens nsic ic an and d hi hist stor oric ic co cont ntex exts ts.. In In:: Ha Hagl glun und, d, W. W.D. D.,, So Sorg rg,, M. M.H. H. (E (Eds ds.) .),, Fo Fore rens nsic ic Taphonomy: The Post-mortem Fate of Human Remains. CRC Press, Boca Raton, FL, pp. 483 495. Smith, K.R., et al., 2001. The Radiological Impact of Coal-Fired Electricity Generation in the UK. NRPB-R327, National Radiological Protection Board Report. Spradley, M.K., Hamilton, M.D., Giordano, A., 2012. Spatial patterning of vulture scavenged human remains. Forensic Sci. Int. 219, 5763. Steadman, D.W., Worne, H., 2007. Canine scavenging of human remains in an indoor setting. Forensic Sci. Int. 173, 7882. Su Sung ng Tz u, 118 118612 1249 49.. The The Wash Washin ing g Away Away of Wron Wrongs gs.. Tr Tran ansl slat ated ed fr from om the the Chin Chines esee by McKnight, McKnight, B.E., B.E., 1981. University University of Michigan Michigan,, Ann Arbor, pp. 69 70. ’
Swift, B., 199 Swift, 1998. 8. Dat Dating ing hum human an ske skelet letal al rem remain ains: s: inv invest estiga igatin ting g the via viabil bility ity of mea measur suring ing the 210 210 equilibrium between Po and Pb as a means of estimating the post-mortem interval. Forensic Sci. Int. 98, 119126. Swift, Swif t, B. B.,, 20 2006 06.. Th Thee ti timi ming ng of de deat ath. h. In In:: Ru Rutt tty, y, G. (E (Ed. d.), ), Es Esse sent ntia ials ls of Au Auto tops psy y Pr Prac acti tice ce.. Springer-Verlag, London, Ch. 8. Swift, B., Lauder, I., Black, S., Norris, J., 2001. An estimation of the post-mortem interval in human skeletal remains: a radionuclide and trace element approach. Forensic Sci. Int. 117, 73 87. Taylor,, R.E Taylor R.E., ., Suc Suchey hey,, J.M J.M., ., Pay Payen, en, L.A L.A., ., Slo Slota, ta, P.J P.J., ., 198 1989. 9. The use of rad radioc iocarb arbon on (14C) to identify human skeletal materials of forensic scientific interest. J. Forensic Sci. 34 (5), 1196 1205. Ubelaker, D.H., 2014. Radiocarbon analysis of human remains: a review of forensic applications. J. Forensic Sci. 59 (6), 14661472. Ubelaker, D.H., Parra, R.C., 2011. Radiocarbon analysis of dental enamel and bone to evaluate date da te of bi birt rth h an and d de deat ath: h: pe pers rspe pect ctiv ivee fr from om th thee so sout uthe hern rn he hemi misph spher ere. e. Fo Fore rens nsic ic Sc Sci. i. In Int. t. 20 208, 8, 103107. Ubelaker, D.H., Zarenko, K.M., 2011. Adipocere: what is known after two centuries of research. Forensic Sci. Int. 208, 167172. Ubelaker, Ubelak er, D.H., D.H., Tho Thoma mas, s, C., Olson, Olson, J.E J.E., ., 201 2015. 5. The impact impact of age at dea death th on the lag tim timee of radiocarbon values in human bone. Forensic Sci. Int. 251, 5660. Van der Pli Plicht cht,, J., 200 2004. 4. Rad Radioc iocarb arbon on cal calibr ibrati ation on past, pre presen sentt and fut future ure.. Nuc Nucl. l. Inst Instrum rum.. Methods Phys. Res. B 223224, 353358.
126 Human Body Decomposition Vass, A.A., et al., 1992. Time since death determinations in human cadavers using soil solution. J. Forensic Sci. 37 (5), 12361253. Wallman, J.F., 2001. A key to the adults of species of blowflies in southern Australia known or suspected to breed in carrion. Med. Vet. Entomol. 15, 433437. Wallman, J.F., 2002. Winged evidence: forensic identification of blowflies. Aust. J. Forensic Sci. 34 (2), 7379. Wallman, J.F., Leys Wallman, Leys,, R., Hoge Hogendoorn ndoorn,, K., 2005. Mole Molecula cularr syste systemati matics cs of Austra Australian lian carrionbreeding breed ing blowf blowflies lies (Diptera: Calliphorida Calliphoridae) e) based on mito mitochond chondrial rial DNA. Invert Invertebr. ebr. Syst. 19, 115. Wang, J.Y., 1960. A critique of the heat unit approach to plant response studies. Ecology 41 (4), 785790. White, L., Booth, T.J., 2014. The origin of bacteria responsible for bio erosion to the internal microst mic rostruc ructur ture: e: res result ultss fro from m exp experi erimen menta tally lly dep deposi osited ted pig car carcas casses ses.. For Forens ensic ic Sci Sci.. Int Int.. 239 239,, 92102. White, E.M White, E.M., ., Han Hannus nus,, L. L.A., A., 198 1983. 3. Che Chemic mical al wea weathe therin ring g of bon bonee in arc archae haeolo ologic gical al soi soils. ls. Am. Antiq. 48 (2), 316322. Wild, E.M., et al., 2000. 14C dating with the bomb peak: and application to forensic medicine. Nucl. Instrum. Methods Phys. Res. B 172, 944950. Willey Will ey,, P. P.,, He Heil ilma man, n, A. A.,, 19 1987 87.. Es Esti tima mati ting ng ti time me si sinc ncee de deat ath h us usin ing g pl plan antt ro root otss an and d st stem ems. s. J. Forensic Sci. 32 (5), 12641270. Wilson, A.S., et al., 2007. Modelling the buried human body environment in upland climes using three contrasting field sites. Forensic Sci. Int. 169, 618. Yan, F., McNally, R., Kontanis, E.J., Sadik, O.A., 2001. Preliminary quantitative investigation of post-mortem adipocere formation. J. Forensic Sci. 46 (3), 609614. Yoshino, Yoshin o, M., et al. al.,, 199 1991. 1. Mic Micros roscop copica icall stu study dy on est estima imatio tion n of tim timee sin since ce dea death th in ske skelet letal al remains. Forensic Sci. Int. 49, 143158.
CHAPTER
5
Recent Research and Current Trends
DEVELOPMENT OF A GRADING SYSTEM OF HUMAN DECOMPOSITION CORRELATION OF A DECOMPOSITION GRADING SYSTEM WITH THE PMI RESERVATIONS CONCERNING THE USE OF ADDs AND NUMERICAL BODY SCORING SYSTEMS CONCLUSION REFERENCES
In the mid-13th century Sung Tz’u (see chapters: Supravital Reactions in the Estimation of the Time Since Death (TSD) and Research in the Later Stages of Decomposition) elaborated on the variable nature and rate of decomposition in the four seasons. In concluding this section in his handbook, which includes some detailed descriptions of the decomposition process, he notes, “In extremely cold weather, five days is equivalent to one day in a time of great heat, and half a month the equivalent of three or four extremely hot days ” (Sung (Sung Tz’u, 1186 11861249 1249). ). Not a great deal has changed over the intervening 700 plus years with, until recently, most studies into autolysis and putrefaction continuing to be of a descriptive and qualitative nature, describing the various stages, the factors involved, the biochemistry of the process and other factors af afffecti ecting ng it. Onl nly y in the the last last de deccade ade has it been been re reccogniz gnized ed tha hat, t, if advances are to be made in the estimation of the postmortem interval (PMI), quantification of the process is required. In their review paper on the estimation of the PMI, Henssge PMI, Henssge and Madea (2007) state (2007) state that the main principle of determination of the time since death (TSD) should be: the calculation a that, measurable date along a time for dependent curve back to the start point and of and most methods proposed estimation of the time ” ”
“
Human Body Decomposition. DOI: http://dx.doi.org/10.1016/B978-0-12-803691-4.00010-8 © 2016 Elsevier Inc. All rights reserved.
128 Human Body Decomposition
since death are of only academic interest since they describe just post-mortem ch chan ange ges. s. Th They ey on only ly gain gain pr prac acti tica call re rele leva vanc ncee if the the fo follllow owin ing g cr crit iter eria ia ar aree fulfil ful filled led:: qua quanti ntitat tativ ivee measu measurem rement ent,, ma mathe themat matica icall des descri cripti ption, on, tak taking ing int into o account influencing factors quantitatively, declaration of precision and proof of precision on independent material. Henssge and Madea (2007 , p. 183)
Two main Two main ap appr proa oach ches es to quan quanti tify fyin ing g deco decomp mpos osit itio ion n in the the la late terr stages have evolved in recent years: firstly, how to determine a mathemati ma tica call sc scal alee of the the de deco comp mpos osit itio ion n proc proces ess, s, an and d se seco cond ndly ly,, how how to dete de term rmin inee a mode modell wh whic ich h in inco corp rpor orat ates es th thee deco decomp mpos osit itio ion n pr proc oces esss with the main variables affecting decomposition, namely, temperature, insect access, burial depth, humidity, trauma, body size and weight, the pres pr esen ence ce or abse absenc ncee of clot clothi hing ng,, and and rand random om an and d oppo opport rtun unee anim animal al scavenging.
DEVELOPMENT OF A GRADING SYSTEM OF HUMAN DECOMPOSITION
Attemp Atte mpts ts to qu quan anti tify fy the the stag stages es of de deco comp mpos osit itio ion n we were re fi firs rstt ma made de by entomologists studying insect successions colonizing carrion during decomposition. Reed (1958) studied (1958) studied insect species successions on 43 dog carcasses in a variety of environments while they decomposed. He recognized four stages — fresh, fresh, bloated, decay, and dry — and and he quoted previous authors who had staged the decomposition process: Mégnin (1894), (1894), who recognized eight stages; Fuller stages; Fuller (1934), (1934), who recognized three stages; Howden (1950) (1950),, who recognized two stages. These were early classiand Howden and fications of the stages of decomposition. In a semi semina nall stu tudy dy,, Pay Payne ne (1 (196 965) 5) researc researched hed decomp decomposi ositio tion n in animals with and without insect access. He found that decomposition in smal smalll an anim imal alss su such ch as frogs frogs,, toad toads, s, mice mice,, shre shrews ws,, rats rats,, an and d chip chip-munks was too rapid to allow the detailed study of insect succession and the feathers in chickens caused problems in recognition and estimati ma tion on of inse insect cts. s. He fo foun und d baby baby pi pigs gs mo more re suit suitab able le for for th thee stud study. y. Recogn Rec ognizi izing ng decomp decomposi ositio tion n as a contin continuou uouss pro proces cess, s, he des descri cribed bed in deta de tail il si six x st stag ages es of deco decomp mpos osit itio ion n: fres fresh, h, bl bloa oate ted, d, acti active ve deca decay, y, advanced decay, dry, and remains. He found a defined insect successi sion on at ea each ch stag stagee of deco decomp mpos osit itio ion n and and that that ca carc rcas asse sess expo expose sed d to insects decomposed more rapidly.
Recent Research and Current Trends
129
Rod Rodrigu riguez ez and Bas Basss (198 (1983) 3) studied studied ins insect ect specie speciess success succession ion and their correlation with the stages of decomposition in human cadavers, in order to determine if this might be used as a method of determining the PMI. Four decomposing human cadavers in a wooded area were we re stage taged d usin using g the the syst system em de devi vissed by Reed Reed (1 (1958 958)) of fres fresh, h, bloated, decay, and dry, which referred to the external appearance of the body, but without a detailed description of each stage. Rodriguez and Bass found that there was a direct correlation between the rate of decay and the succession of insect families and species found in association with a decomposing cadaver and application of this knowledge ed ge mi migh ghtt lead lead to a more more ac accu cura rate te es esti tima mati tion on of the the TSD TSD of an individual. Gal Gallow loway ay and col colle leagu agues es (19 (1989) 89) publi publishe shed d a retro retrospe specti ctive ve study study of 18 189 9 bodi bodies es foun found d on the the dese desert rt floo floor, r, surr surrou ound ndin ing g moun mounta tain ins, s, or indoo ind oors rs in south southern ern Ari Arizo zona, na, in order order to pro provid videe guidel guidelin ines es for for est estiimati ma ting ng the the PM PMII base based d on avera average ge deca decay y rate ratess in cl clos osed ed stru struct ctur ures es,, burials, and in the open. The dates when last seen alive and when the remains were found were recorded in the 189 cases and gave the authors an estimation of the PMI, but they acknowledged that this method of dete de term rmin inin ing g the the PMI PMI migh mightt give give an over overes esti tima mati tion on in many many case cases. s. Gallo Gal loway way and colle colleagu agues es not only only num numer erica ically lly stage staged d the the proces processs of decomposition but, for the first time, various descriptive subcategories were listed within each stage. The five major stages were fresh, early decomposition, advanced decomposition, skeletonized, and decomposition of the skeletonized remains. One example of Galloway et al.’s subcategories, in the stage of early decomposition was: 1. Pink-white Pink-white appearance. appearance. 2. Gray to green discoloration discoloration;; some flesh relatively relatively fresh. 3. Discol Discolora oratio tion n to bro browni wnish sh shades shades,, partic particula ularly rly of the finger fingers, s, nose, nose, and ears; some flesh still relatively fresh. 4. Bloating Bloating with green discolo discoloration ration.. 5. Post bloating bloating following following rupture of abdominal gases with discoloratio discoloration n going from green to dark green. 6. Brown to black discolorat discoloration ion of arms and legs; skin having a leathery leathery appearance. Galloway and colleagues were thus able to estimate a time period when the changes would occur, for a body found in a specific locality. Thee majo Th majorr fact factor orss affe affect ctin ing g th thee PM PMII were were te temp mper erat atur ure, e, mois moistu ture re,,
130 Human Body Decomposition
and insect feeding but factors such as carnivore scavenging, location of the body, and the season of the year when the body was deposited, also also pl play ayed ed a role role.. Bodi Bodies es foun found d in indo door orss sh show owed ed a sl slow ower er onse onsett of decomposition but moisture retention then led to a rapid progression of decomposition to skeletonization. No attempt was made to correlate temperature with the PMI. CORRELATION OF A DECOMPOSITION GRADING SYSTEM WITH THE PMI
Vas Vasss an and d co colle lleag agues ues (1 (1992 992)) introdu introduced ced the concept concept of acc accumu umulate lated d degree days (ADDs) into decomposition research, in particular with the purpose of using them to estimate the PMI. They based their definition of ADDs on the definition quoted by Edwards and colleagues (1987) as being the sum of the average daily temperatures in degrees Centigrade above 0 C for however many days the corpse had been decomposing. Seven weresoil placed face down onthe thecadavers soil surface at different timescadavers of the year; samples beneath were analyzed for their content of volatile fatty acids (VFAs), as well as for a number of cations and anions, at intervals until the cadavers becam be camee skelet skeleton oniz ized ed.. Da Daily ily ma maxim ximum um and and minim minimum um te temp mpera eratu ture ress were recorded and the results of the chemical analysis of the soil were correlated with the total number of ADDs for each cadaver. When the body weight and moisture content of the soils, which determined the concentration of VFAs, were standardized, it was found that the VFA concentrations were the same for any total of ADDs, regardless of the subject or the season of year. It was possible to estimate the PMI for PMI for any any indi indivi vidu dual al,, gi give ven n the the spec specif ific ic VFA VFA rati ratios os,, a gros grosss description of the corpse, and weather station data for the environment where the corpse was found. Vass and colleagues (1992) also found that the PMI estimation was unreliable after heavy rain, with mummi mum mific ficat atio ion n or with with burnin burning g of the the corp corpse se.. The The decom decompo posi sitio tion n scal sc alee us used ed was was base based d on Reed’s (1 (195 958) 8) stagin staging g of fres fresh, h, bloate bloated, d, decay, and dry. Vass and colleagues added descriptions of the external appearance of the body at each stage, but admitted that decomposition stages were not always accurate as the bodies were exposed to widely different temperatures. In discussing the results, the authors commented that temperature was by far the most important factor determining the rate of decomposition as it not only controlled the inte in tens nsit ity y of inse insect ct acti activi vity ty but but also also the the rate rate of chem chemic ical al reac reacti tion onss
Recent Research and Current Trends
131
governing the release of VFA production. Temperature was recorded from readings obtained from the nearest national weather station and adjust adj usted ed with with a corr correc ecti tive ve fac facto torr obta obtaine ined d by re recor cordin ding g dail daily y temtemperatures, taken for a week at the site of the decomposing cadavers. If the temperature fell below 4 C it became much more difficult to estimate the TSD from a single time point as the decomposition process slowed and any temperature below 0 C was regarded as being 0 C. It took 1285 6 110 ADDs for a body to become skeletonized or for VFA production to fall below detectable limits. This enabled a graph to be constructed in which TSD was plotted against temperature tu re,, th thus us pr prov ovid idin ing g a roug rough h esti estima mate te of the the TSD TSD (Fi Fig. g. 5. 5.1 1). The The emphasis in this study was to estimate the PMI from the concentration of VFAs and to correlate this with ADDs rather than with the stage of decomposition. Later, Vass and colleagues (2002) also studied 18 bodies over a 4year period with the purpose of identifying chemical biomarkers in the hear he art, t, lung lung,, brai brain, n, ki kidn dney ey,, live liver, r, and and musc muscle le asso associ ciat ated ed with with ea earl rly y decomposition. The authors quoted the reported accuracy of the estimation of VFAs as a method of estimating the TSD as being 6 2 days in preskeletonized bodies and 6 2 weeks in skeletonized bodies. It was suggested that if the rate of breakdown in decomposition of the large biomolecules of proteins, nucleic acids, lipids, and polysaccharides into their smaller component molecules could be measured, the rate might
300
275 250 225 h t a 200 e d 175 e c 150 n i s 125 s y a 100 D 75 50 25 0
y = =
0
5
10
15
20
1285( x ) –1
25
30
35
40
45
50
Temperature (°C) Figure 5.1 Representati Representation on of maximu maximum m TSD estima estimates tes based on the presence of volatil volatilee fatty acids (Vass ( Vass et al., 1992 1992). ).
132 Human Body Decomposition
correlate with the PMI. The 18 bodies were not embalmed and not auto au tops psie ied; d; some some wer weree clot clothe hed, d, othe others rs un uncl clot othe hed, d, whil whilee othe others rs were were placed in polyethylene bags in order to simulate different environments and all were placed on the ground in a wooded area. The time and caus ca usee of de deat ath h we were re know known n and and the the envi enviro ronm nmen ental tal te temp mper erat atur uree wa wass re reco cord rded ed du duri ring ng the the deco decomp mpos osit itio ion n peri period od.. Tiss Tissue uess from from th thee brai brain, n, heart, lung, kidneys, liver, and muscle were sampled at regular intervals using a probe in order to reduce the degree of organ damage. Biom Bi omar arke kerr su subs bsta tanc nces es from from the the liver liver,, ki kidn dneys eys,, and and hear heartt pr prov oved ed the the most useful in long-term estimation of the PMI, while the brain and liver were useful in the short term. The concentrations of gamma amino butyric acid (GABA), proline, methionine, and especially oxalic acid were found to be the most important biomolecules in determination of the PMI. It was found that ADDs were no longer sufficiently accurate to describe the PMIs and that a narrower time interval was required. They Th ey th ther eref efor oree used used cumu cumula lati tive ve de degr gree ee hour hourss (C (CDH DHs) s) whic which h were were
defined the average temperature in of C for period cumulatively as added. The two main features the each study12-hour were the appearance of glycolic acid in the PMI models and that every organ studied produced a varied assortment of complex biomarker information. A number of flow charts were produced which matched various concentrations of biomolecules with the PMI and two cases using biomarkers correlating with CDHs to estimate the PMI were cited. (2011) has continued to pursue research using a formula based Vass (2011) has on the the time time that that VF VFAs As ceas ceasee to exud exudee from from a deco decomp mpos osin ing g body body,, which he found to be 12 128 85 6 11 110 0 AD ADDs Ds an and d at whic which h sof soft ti tiss ssue ue deco de comp mpos osit itio ion n esse essent ntia iall lly y ends ends.. Stat Statin ing g that that th thee four four main main fa fact ctor orss infl influe uenc ncin ing g th thee ra rate te of deco decomp mpos osit itio ion n of a bo body dy were were te temp mper erat atur ure, e, moisture, pH, and partial pressure of oxygen, Vass devised two new formulae to estimate the PMI, one for a body decomposing in aerobic conditions above ground and the other for a buried body decomposing in anaerobic conditions. The formulae incorporate an estimation of the amou am ount nt of deco decomp mpos osit itio ion, n, the the nume numera rato tor; r; as well well as the the calc calcul ulat ated ed effect of the environment on the decomposition, the denominator. The cond co ndit itio ions ns for for use use of the the firs firstt form formul ula a were were that that th thee body body mu must st be located above ground, in the preskeletonized phase and, if mummified, the tissue must be soft and pliable. The body must be at least 1 day
after death, fairly intact, the temperature should be above 0 C, and the temperature and humidity should be corrected for the discovery site.
Recent Research and Current Trends
133
There should also be little or no adipocere tissue. The first formula is (Vass, 2011 2011): ): PMI ð PMI aerobicÞÞ 5 ðaerobic
1285 3 ðDecomposition=100 100ÞÞ 0:0103 3 Temperature 3 Humidity
(5.1)
In this formula, the PMI is measured in days and “decomposition ” is a percentage estimate of the extent of soft tissue decomposition, which should be determined by an experienced forensic investigator. “1285” is a constant representing the number of ADDs at which VFA liberation from fro m soft soft tissu tissuee ceases ceases and soft soft tis tissue sue decom decompos positi ition on eff effect ectiv ively ely ends ends (Vass et al., 1992). 1992). “0.0103” is a constant, representing a line with a slope of 0.0103, an empirically determined measure based on the observati va tio on of the the eff effect ect of mois moistture ure on huma human n deco decomp mpos osiiti tion on over over a decade. “Temperature” is either the average temperature in C on the day of discovery or the average temperature over a period of time at the site of discovery of the body, corrected after reference to the nearest natio nat ional nal weath weather er statio station n data. data. “Humidity” is the the averag averagee percen percentag tagee humidity on the day of discovery or over a period of time, corrected after reference to the nearest national weather station data. The conditions for use of the second formula were that the body should be undergoing anaerobic decomposition, usually buried; it must be in the preskeletonized phase and if mummified the tissue must be soft and pliable. It should not be in an area where the soil is highly saturated with water. The second formula is (Vass, ( Vass, 2011 2011): ): 1285 3 PMI ð PMI ðanaerobic anaerobicÞÞ 5
Decomposition
100
:6 3 Adipocere
34
0:0103 3 Temperature 3 ðSoil moistureÞ moistureÞ
(5.2)
1285” and “0.0103” are the same constants as used in the previous formula and “decomposition” is determined in the same way. “4.6” is a cons consta tant nt re repr pres esen enti ting ng a sl slow owdo down wn in the the rate rate of de deco comp mpos osit itio ion n due to a lack of oxygen. Buried bodies have been estimated to take approximately eight times longer to decompose than bodies subjected to aerobic conditions. The value of 4.6 was determined from experiments by Vass over many years, in which surface-decomposing bodies were compared to buried bodies taking into account temperature and “
moisture parameters and allowing for the effects of the partial pressure of oxyg oxygen en.. “Adipocere” is an inte intege gerr repr repres esen enti ting ng a perc percen enttage age
134 Human Body Decomposition
estimate of the amount of adipocere on the body and must be determine mi ned d by a fore forens nsic ic in inve vest stig igat ator or.. The The pr pres esen ence ce of ad adip ipoc ocer eree sl slow owss decomp dec omposi ositio tion n by trappi trapping ng moistu moisture re and dec decrea reasin sing g the par partia tiall prespressure of oxygen. Each 5% range is given an integer up to the value of 14 from a table devised after comparing adipocere formation in numerous surface and buried bodies and its effect on the rate of decomposition. It is this integer which is inserted in the formula. The “temperature” is that of the soil in the grave at the level of the body. It is measured by a thermocouple and compared with the surface temperature. The value inserted in the formula is determined by comparison with the temperature from the nearest weather station for the estimated time of burial adjusted accordingly. The “soil moisture” is a percentage dete de term rmin ined ed by co coll llec ecti ting ng a soil soil samp sample le from from be bene neat ath h the the body body and and comparing the weights before and after it is dried. Vass (2011) claimed that the formulae worked well in the mid to eastern United States and he stated that he had applied them to many case ca sess wo worl rldw dwid idee with with succ succes ess. s. Howe Howeve ver, r, the the use use of the the form formul ulae ae depended on the estimation of the stage of decomposition and amount of adipocere formation by an experienced observer, using percentage figu figure ress rath rather er th than an a qu quan anti tita tati tive ve sc scal ale. e. The The comm common ones estt sour source ce of error was when the time period exceeded 1285 days and it was also important that the body was in the preskeletonized stage. The estimation of the amount of adipocere was the weakest aspect of the second formula as not all bodies produced adipocere, and its estimation was difficult. An impo import rtan antt re retr tros ospe pect ctiv ivee stud study, y, whic which h stim stimul ulat ated ed othe others rs to research in a similar vein, was published in 2005 where, for the first time, decomposition was quantifiably staged and treated as a semicontinu tinuou ouss vari variab able le that that co corr rrel elat ated ed with with te temp mper erat atur uree as meas measur ured ed by ADDs (Megyesi (Megyesi et al., 2005). 2005). Soft tissue decomposition was recognized as a sequ sequen enti tial al proc proces esss with with nume numero rous us sm smal alll chan change gess occur ccurri ring ng thro th roug ugho hout ut.. Thes Thesee chan change gess were were quan quanti tifi fied ed wi with th nume numeri rica call grad gradin ing g values to allow statistical testing of the relationship between decomposition, as the dependent variable and temperature as the independent variable. Sixty-eight cases of human remains were selected from forensic cas casee files files from from throug throughou houtt the Uni United ted States States.. FiftyFifty-seve seven n corpse corpsess were found in the open in a variety of contexts and 11 were found indoor ind oors. s. Corpse Corpsess which which wer weree bur burned ned,, buried buried,, submer submerged ged in water water or with adipocere were excluded, as were children. However all corpses
Recent Research and Current Trends
135
had a wide variety of clothing and 26 were naked or had very little clothing at all. All corpses showed evidence of insect access. The PMI was known in 20 cases from records and in 48 cases it was determined by insect evidence but it was less than a year in every case in order to select cases in which soft tissue was still present. The stage of decomposi po siti tion on was gr grad aded ed from from phot photog ogra raph phss us usin ing g a modi modifi fica cati tion on of the the st stag agin ing g used used by Ga Gall llow oway ay an and d co coll llea eagu gues es (1 (198 989) 9) (ie, (ie, fres fresh, h, earl early y decomposition, advanced decomposition, and skeletonization). Within each ea ch st stag agee ther theree were were desc descri ript ptiv ivee su subc bcat ateg egor orie iess whic which h were were gi give ven n numerical values and because not all stages of decomposition applied equally to all parts of the body, it was divided into three parts for scoring purposes: the head and neck, the trunk, and the limbs. When the nume nu meri rica call va valu lues es fr from om the the th thre reee part partss we were re su summ mmat ated ed,, the the re resu sult lt amounted to a total body score (TBS). The lowest score possible was 4 and the highest 35. If the decomposition stage varied across one anatomical region, for example, an arm and a leg, the average of the two was taken. Iffini the observed decomposition seem tos match the se sequ quen ence ce defi de niti tion ons, s, the the nume nu meri rica call valu valueedidassi asnot sign gned ed was wa one one that th at matched the earlier stage of decomposition, because it was considered that the TBS should best reflect how much decomposition had taken place overall. overall. ADDs were calculated from the national weather service station data nearest to where each body was found and were defined as the product of the average daily temperature above 0 C (average of the maximum and minimum daily temperatures) from the time of death until discovery, with no correction carried out for distance or any other variable. As the temperature at which decomposition ceases is unknown, all temperatures that fell below 0 C were treated as 0 C because temperatures below 0 severely inhibit the chemical processes of decomposition and of bacterial growth. The quantitative analysis plotted TBS as the dependent variable against PMI and ADD as the independent variables. The relationships were not linear but, by log transforming both the ADD and PMI and squaring the TBS, an effective linear relationship was found, from which an equation could be produced. To find the PMI for any given case, the TBS would be calculated and this figure then substituted in the equation to calculate the figure for ADD. Once the ADD was known, the PMI could be calculated. Megyesi calculated. Megyesi et al. (2005) stressed (2005) stressed that the study only accounted for accumulated time and temperature as the main main variabl variables es affect affecting ing decomp decomposi ositio tion n and that that other other variab variables les
136 Human Body Decomposition
such as animal scavenging, sunlight, rainfall, and clothing might not be poss po ssib ible le to qu quan anti tify. fy. Ov Over er 80% 80% of th thee obse observ rved ed vari variat atio ion n in huma human n decomposition could be accounted for by the combination of elapsed time and temperature as reflected in ADDs. The publication of this paper has stimulated researchers to use the concepts of ADDs and numerical grading systems of decomposition, in order ord er to quant quantify ify th thee rel relati ation onsh ship ip betwee between n the rat ratee of decom decompo posit sitio ion n and temperature as the main variable. (2009) proposed another quantitative method Fitzgerald and Oxenham (2009) proposed of scoring decomposition when they devised a “degree of decomposition index” (DDI). The DDI used the general categories described by Galloway by Galloway et al. (1989), (1989), but followed the method of Megyesi Megyesi et al. (2005) which (2005) which used the categories independently on separate body elements. It scored decomposi po siti tion on va valu lues es from from 0 to 5 by di divi vidi ding ng the the tota totall sc scor ores es of the the stag stages es observed on individual body elements by the number of body elements sc scor ored ed in orde orderr to ma make ke th thee DDI DDI di dire rect ctly ly comp compar arab able le to stag stages es used used in pr prev evio ious us stud tudies ies. Two Two pigs igs dec eco ompos mposiing on the the grou round nd,, one one in semishade and one in full sun, were studied. Temperature and humidity data da ta we were re reco record rded ed from from each each envi enviro ronm nmen entt and and data data from from the the near neares estt Aust Au stra rali lian an Bure Bureau au of Mete Meteor orol olog ogy y weat weathe herr stat statio ion n were were ob obta tain ined ed.. Thee DD Th DDII was was comp ompared ared with ith the the fou four sta tage gess of de deco comp mpo osi sittio ion n of Rodr drig igue uezz an and d Ba Bass ss (1 (198 983) 3) and and the the stag stages es of of Ga Gall llow oway ay et al al.. (1 (198 989) 9).. Ro Fittzger Fi zgeral ald d and Oxen xenham ham di did d not not use ADDs DDs but ins nste tead ad used used TS TSD D measured in days as the independent variable. It was found that the carcass exposed to full sun became partly preserved by desiccation and adipocere form fo rmat atio ion, n, whil whilee the the ca carc rcas asss in se semi mish shad adee di disi sint nteg egra rate ted. d. The The stag stages es of decomposition described by Galloway by Galloway et al. (1989) were (1989) were found to be unusable as they were too narrow and too environmentally specific, while the (1983) were were more usebroader categories described by Rodriguez by Rodriguez and Bass (1983) ful for assigning a DDI for the carcasses. Despite the marked variation in the environments in which the two carcasses decomposed, the DDI curves of both were similar when compared with the TSD and smoother than the other methods of scoring decomposition. The results of regression modeling suggested that TSD correlated well with the DDI and accounted for 95% of the variation in decomposition between the carcasses while variations in the environment were not significant contributing factors. However, Fitzgerald and Oxenham raised the important question as to whether or not TSD was purely a measure of transpired time or were
Recent Research and Current Trends
137
other variables concealed within the measure such as environmental variables which were site-specific and specific to that particular TSD. The authors also found that there was no consistent relationship between the temp temper erat atur ures es of the the site site and and the the temp temper erat atur ures es re reco cord rded ed by the the lo loca call weather station when calculations were being made for ADDs, a finding that did not bode well for the use of ADDs obtained retrospectively. It was also noted that although TSD and ADD both measured temperature and humidity over time, ADD had a distinct advantage over TSD in that TSD was reliant on location-specific seasonal data whereas ADD was not affected by seasonal temperature variation, only by location-specific patternss in the relatio tern relationsh nship ip bet between ween tempera temperature ture and humidity humidity.. However However,, considering that TSD not only measured transpired time but also incorporate ra ted d all all envi enviro ronm nmen enta tall vari variab able less over over th thee peri period od of de deco comp mpos osit itio ion, n, whereas there was no consistent relationship between the measurement of ADDs at the decomposition site and at the local weather station, it was concluded that TSD was more appropriate for use in decomposition models than ADD. A study by Adlam by Adlam and Simmons (2007) monitored (2007) monitored the decomposition of three groups of eight rabbit carcasses placed on soil on the ground, under cages to prevent scavenging, over a period of 3 weeks. One of each group of carcasses was disturbed at 3-day intervals while the other carcasses were only disturbed once for the purpose of taking observations. The aim of the study was to determine if repeated disturbance of carcasses affected the rate of decomposition. This study was of importance because in addition to taking environmental data measurements such as average temperature over 24-hour periods which was converted to ADDs, soil pH, weight, and temperatures of the carcasses, a visual qualitative decomposition scale was employed which was later converted to a numerical score. Adlam and Simmons showed that loss of weight was faster but the end weight was higher in the undisturbed samples although in the final analysis there was no difference in the time taken between the two groups to reach total skeletonization, that is, decomposition over time was not significantly altered by repeated physical disturbance. In both groups loss of weight correlated closely with increasing decomposition but weight loss and carcass temperature were significantly affected by repeatedly disturbing a carcass. The authors concluded that scoring decomposition directly yielded a more accurate assessment of the process and together with the usee of AD us ADDs Ds migh mightt in futu future re stan standa dard rdiz izee meas measur urem emen ents ts an and d al allo low w comparisons with other studies.
138 Human Body Decomposition
Si Simmo mmons ns and co colle lleag agues ues (2 (2010 010)) foll follow owed ed up th this is hypo hypoth thes esis is by compiling data from a number of published and unpublished studies concer con cernin ning g decomp decomposi ositio tion n and weight weight loss. loss. The author authorss com commen mented ted that in the past there had been an inability to directly compare results and observations from many studies because they varied in their methodol od olog ogy, y, geo geograp graphi hica call loca locale le,, cli clima mattic zone zone and and seas season on,, spec specie iess obse ob serv rved ed,, and and dura durati tion on of ob obse serv rvat atio ion. n. Yet Yet othe otherr stud studie iess whic which h had had been be en base based d on re retr tros ospe pect ctiv ivee fore forens nsic ic ca case se wo work rk an and d da data ta were were no nott re repo port rted ed in a stan standa dard rdiz ized ed fo form rm.. It wa wass su sugg gges este ted d that that the the use use of ADDs in documenting chronological time and temperature together in decomposition studies would allow comparison of studies across multiple and varied environments as ADDs represented the accumulation of thermal energy required for the chemical and biological reactions of decomposition to take place; when the same amount of thermal energy represented by ADDs was put into a carcass, the same amount of reaction tion,, re repr pres esen ente ted d by the the TBS, TBS, a meas measur uree of de deco comp mpos osit itio ion, n, woul would d occur. Therefore, would any animals exposed to fluctuating temperatures in any environment show the same stage of decomposition for the same accumulation of degree days. They devised a formula to convert weight loss to a TBS, standardized after converting the different systems of scoring the TBS over many studies. Timescale data were converted to ADDs; not difficult in some controlled studies but in others an approximation had to be used because temperature data were not available for the duration of the studies. Data were collected from terrestrial, buried, and aquatic environments. Comparing log ADD with standardized TBS scores, Simmons and colleagues found a significant correlation over all studies. In all studies and contexts, at comparable ADDs, the presence of insects influenced the rate of decomposition by making it significantly faster. There was no significant difference in decompo deco mposit sition ion rat ratee bet between ween indoor indoor,, buried buried,, and submerg submerged ed bodies bodies,, contexts in which insects had largely been excluded but more rapid deco de comp mpos osit itio ion n oc occu curr rred ed in bodi bodies es outd outdoo oors rs than than bodi bodies es pl plac aced ed indoors, in surface compared with buried bodies, and in surface compared with submerged bodies. Variables such as penetrating trauma and percentage of body fat had no effect on the rate of decomposition. At comparable ADDs it is primarily the presence or absence of insects which has a significant effect on the rate at which a body will deco de comp mpos ose, e, howe howeve verr Si Simm mmon onss and and coll collea eagu gues es poin pointe ted d to fu furt rthe herr research being required to refine the relationship between carcass size and the decomposition rate in the presence of insects. Simmons and
Recent Research and Current Trends
139
colleagues concluded that although insects were the prime cause of rapidi rap idity ty of decompo decomposit sition ion,, in largerlarger-siz sized ed carcass carcasses, es, heat generat generated ed by larger maggot masses could increase the rate of decomposition. By comp co mpar aris ison on,, in sm smal alle lerr ca carc rcas asse sess th thee heat heat ge gene nera rate ted d by smal smalle lerr amounts of insects would be less but this might be offset by the smaller carcass decomposing more rapidly. Carcass size, however, had no influence on the rate of decomposition when insects were excluded. In contrast, a study carried out by Sutherland and colleagues (2013) in South Africa compared the decomposition rates between large pigs (60 (60 90 kg kg)) and and smal mall pigs igs ( , 35 kg kg). ). The The resu result lt of th this is stud study, y, in which insects were not excluded, found that small pigs decomposed 2.82 times faster than the large pigs. Therefore body size did have an effect on the rate of decomposition. Increasingly, studies are being published using ADDs as a proxy for the TSD (PMI) and modeling this parameter with graded decomposition scales in attempts to quantify the estimation of the PMI. Michaud and an d Mo More reau au (2 (2011 011)) studie studied d the rel relati ations onship hip betwee between n decomp decomposi ositio tion n st stag ages es and and de degr gree ee da day y accu accumu mula lati tion on by deco decomp mpos osin ing g pi pig g carc carcas asse sess exposed over various seasons and over 3 years, in two large rural fields, 2.2 km apart. The purpose of the study was to verify the predictability of the decomposition stages used in forensic entomology and to build a degree day accumulation model applicable to various decompositionrelated processes. They used a decomposition scale of five stages, fresh, bloated, active decay, advanced decay, and dry decay but because all stages were not of equal duration, a new scale was developed which allowed stages to be represented with realistic intervals instead of rigid values. ADD values were recorded by a data logger situated in one field and these were coordinated with a weather station situated in the other fiel field. d. St Stat atis isti tical cal anal analys ysis is was carri carried ed ou outt in fi five ve stag stages es:: fi firs rstl tly, y, the the calc ca lcul ulat atio ion n of ADDs ADDs si sinc ncee deat death; h; se seco cond ndly ly,, a sc scal alee was was de defi fine ned d to re repr pres esen entt deco decomp mpos osit itio ion n with with re real alis isti ticc in inte terv rval als, s, call called ed the the degr degree ee day index; thirdly, a multiple regression analysis using ADDs and the decomposition scale was carried out to develop a decomposition index; fourthly, the probability that a carcass belonged to a given stage along the degree day index was calculated; and finally, the index was validated usin us ing g data data from from a pr previ eviou ouss stud study. y. A mu mult ltip iple le re regr gres essi sion on mo mode dell was developed in which the onset of each decomposition stage, also known as the decomposition index, could be determined from the degree day index. The study supported the prediction that the physical condition
140 Human Body Decomposition
of a carc carcas ass, s, as de desc scri ribe bed d by deco decomp mpos osit itio ion n stag stages es,, was was a re reli liab able le representation of the decomposition process and the model accounted for 97% of the variability in decomposition with respect to temperature. It also indicated that every stage of decomposition, except the active decay stage, had a high probability of occurring at a precise moment along the degree day index scale. When a degree day index had been developed for a given geographical area, the model required to be constructed from temperature records to calculate the ADDs. The ADD values could then be used to calculate decomposition indices for the same sa me inte interv rval als. s. Mich Michau aud d and and More Moreau au admi admitt tted ed that that some some le leve vell of subjectivity existed in the discrimination between decomposition stages and they stressed the need for standard stage definitions that would reduce this possibility possibility.. However because the model incorporated incorporated adjustment me ntss for for inte intery ryea ear, r, betw betwee een n se seas ason on,, and and wi with thin in se seas ason on va vari riat atio ions ns,, it allo allowe wed d for for the the deve develo lopm pmen entt of pr pred edic icti tion on mo mode dels ls thro throug ugho hout ut th thee year and in differ different ent geographical geographical areas. A longitudinal study of 10 human cadavers by Suckling by Suckling (2011) tested (2011) tested (2005) of of scoring decomposition and its relathe model of Megyesi Megyesi et al. (2005) tionship to ADD. The effect of scavenging animals on the decomposition rate and its relationship to ADD was also factored into the study. The study stud y found found that sca scaven venged ged bod bodies ies deco decompo mposed sed fas faster ter than than pro protec tected ted cadavers and the Megyesi model, which had been developed from photographs, lacked precision in certain aspects. For example, Suckling found that in the central Texas environment, the external body could mummify but the deeper layers would still be undergoing soft tissue decomposition. Under the Megyesi system, mummification was given a higher numerical sc scor oree th than an mo mois istt deco decomp mpos osit itio ion n. Howe Howeve ver, r, Suck Suckli ling ng obse observ rved ed the the resumption of moist decomposition on two bodies which had been partly mummified, a situation not accounted for by Megyesi’s method of scoring. Suckling also found a low success rate in using Megyesi’s model to score bodies in the late stage of decomposition and she concluded that the model was not to be recommended in severely decomposed or skeletonized remains. Many variables were found to affect the TBS and future research was suggested to test for interobserver error in its estimation. This study by Suckling also did not support the assertion by Simmons et al. (2010) that (2010) that data comparison from many different environments and temper tem peratu atures res regard regarding ing the stages stages of dec decomp omposi osition tion wer weree compar comparabl ablee using TBS and ADD.
Recent Research and Current Trends
141
RESERVATIONS CONCERNING THE USE OF ADDs AND NUMERICAL BODY SCORING SYSTEMS
The concept of ADDs continues to be tested; a recent study was carried out on the high veldt of South Africa by Mybu Myburgh rgh an and d colleagues (2013) in (2013) in order to test the method in that specific environment. A total of 46 pigson with bodyover weights varying between 38 pigs and 91 kg were decomposed a farm a 232-day period. Thirty were used to set up a standard to create a region-specific model and 16 pigs were used to validate the results. The state of decomposition was visually assessed three times a week until all tissues were desiccated, then once a week until skeletonized. The decomposition scoring method used was the one devised by Megyesi et al. (2005) and which resulted in a TBS. Maximum and minimum air temperatures were obtained, in order to calculate ADDs from a data logger on site and from a weather station 23 km from the site. The intraclass correlati lation on coef coeffi fici cien entt be betwe tween en the the tw two o te temp mper eratu ature re re reco cord rdin ing g sites sites wa wass 0.9, indicating excellent agreement. Interobserver bias was tested by having an independent individual assessing the stages of decomposition tion and and it was was foun found d to be negl neglig igib ible le.. Beca Becaus usee ADD ADD valu values es were were skewed on the original scale they were log transformed so as to be linearly related to TBS scores. When TB When TBS S was was pl plot ottted ag agai ains nstt ADD, ADD, th thee deco decomp mpos osiiti tion on wa wass exponential, being relatively linear in the early stages, regardless of the season (between 200 and 400 ADD) but becoming very variable with TBSs greater than 17. In these later stages of decomposition, the rate decreased to produce a plateau phase in which it remained stable and unchanged for long periods of time. Decomposition occurred faster in pi pigs gs depo deposi site ted d in summ summer er co comp mpar ared ed with with th thos osee depo deposi site ted d in wint winter er after the TBS score became greater than 17. A predictive equation was developed to calculate the ADD for an unknown case. When TBS was regressed against log ADD an an r2 value of 0.623 was obtained, meaning that 62% of the variability in decomposition as reflected by the TBS was accounted for by ADD. In order to transform an ADD value into a PMI PMI in an un unkn know own n case case,, an av aver erag agee of da dail ily y te temp mper erat atur ures es was was recorded at a site or local weather station and added together from thee day th day of di disc scov over ery y un unti till th thee actu actual al ADD ADD equa equale led d the the es esti tima mate ted d ADD. The PMI was then the number of days it took for these two valu va lues es to beco become me equa equal. l. Howe Howeve verr th thee re resu sult ltss from from th thee va vali lida dati tion on
142 Human Body Decomposition
study on 16 pigs revealed that the PMI of only one pig fell within the lower limit of the 95% prediction interval. The PMIs of 11 pigs were underestimat mated while the PMIs of 4 pigs were overestimat ateed, indicating that decomposition in that particular geographic area was too to o vari variab able le to allo allow w accu accura rate te PM PMII es esti tima mati tion ons. s. Th Thee in inte tero robs bser erve verr error in estimating the TBS, however, was excellent and was found to be repeatable 99.2% of the time. In di disc scus ussi sing ng the the re resu sult ltss, My Mybu burg rgh h an and d coll colleeague aguess stat stateed th that at alth althou ough gh 62 62% % of th thee vari varia ation tion fou found in the the TBS TBS sc scor ores es cou could be explained with ADD, the 38% of the variability not accounted for by ADD,, especi ADD especiall ally y in the later later stages stages,, wa wass pos possi sibly bly due to sc scave avengi nging, ng, season sea sonali ality, ty, dif diffe feren rences ces in humid humidity ity,, and dif differ feren entt rat rates es of ba bacte cteria riall action. These factors could not be ignored and required further research to impr improv ovee the the accu accura racy cy of PMI PMI dete deterrmin minatio ation n bu butt the the auth author orss commented that the TBS method was a good quantitative indicator of the stages of decomposition and the rate of change from one stage to another could be used to compare the rate of decomposition between specimens. Studie Stud iess in re rece cent nt ye year arss ha have ve atte attemp mpte ted d to qu quan anti tify fy the the deco decommpositi pos ition on proce process ss as it aff affect ectss the the extern external al appea appeara rance nce of a cadave cadaver. r. Deco De comp mpos osit itio ion n scor scores es have have be been en devi devise sed d fo forr pig igss (Fitz Fitzgeral gerald d and Oxenham, Oxen ham, 2009 2009), ), rabbits (Adlam (Adlam and Simmons, 2007), 2007), and humans (Megyesi et al., 2005), 2005), but comparison between species may diminish accuracy (Notter (Notter et al., 2009). 2009). As has been noted, increasing numbers of studies are designed to test Megyesi and colleagues’ (2005) model of estimating the PMI. However, the method of estimating the TBS in this model may be too inaccurate because of the way it was derived. In the first instance the TBS was derived from 68 forensic cases investigated by different workers in a variety of geographical areas in 19 states of the United States. In addition the score was derived from perusing photographs of bodies, the majority of which were clothed so that only exposed parts of the body could be judged for the degree of decomp dec ompos ositi ition on.. Only Only in a few bod bodie iess wer weree the autops autopsy y pho photog tograp raphs hs used supplemented with detailed taphonomic data. Furthermore, the bodi bo dies es we were re fo foun und d in a numb number er of di diff ffer eren entt se sett ttin ings gs:: in indo door ors, s, outoutdoors, in shade, and in the sun, and finally the stages were subjectively modifi mod ified ed into into a se seque quenti ntial al ranki ranking ng so tha thatt th thee fi final nal decomp decomposi ositio tion n scores reflected the total amount of accumulated decomposition that had occur occurred red.. In the previ previous ously ly qu quote oted d studi studies es in which which this this mod model el
Recent Research and Current Trends
143
was used, the interpretation of this method of TBS may be too broad and too readily susceptible to individual interpretation. If a quantitative estimate of decomposition is to be of use in practice it must be simple to use and much more precise in its interpretation and description. The process of decomposition is not only a continuous, dynamic onee but on but also also a holi holist stic ic on one, e, affe affect ctin ing g al alll body body ti tiss ssue uess and and in inte tern rnal al organs as well as the external appearance of the body. A further difficulty is that it is affected by many variable factors apart from temperature, atu re, suc such h as hum humidi idity, ty, env enviro ironme nmenta ntall contex context, t, scaven scavengin ging, g, insect insect activity, presence or absence of clothing, to name a few (Mann ( Mann et al., 1990). 1990 ). If a body body de deco comp mpos osit itio ion n sc scor oree can can be devi devise sed d that that can can take take these factors into account and be easily employed then it may be useful forensically (Hayman, (Hayman, 2013 2013). ). Another problem with the use of the model proposed by Megyesi et al. (2005) concerns (2005) concerns the accuracy of the compilation of ADDs, which in turn raises questions about the accuracy and precision of any model to estimate the PMI in which it is incorporated. The pr The prin inci cipl plee of us usin ing g temp temper erat atur uree and and ti time me to me meas asur uree pl plan antt development in heat units seems to have originated with Réaumur (1735).. Réaumur totaled the mean daily temperatures for 91 days in (1735) Apr., May, and Jun. of that year in his locality and found the sum to be a nearly constant value for the development of any plant from year to year and he assumed that this constant value represented the amount of heat required for a plant to reach a given stage of maturity. Since that time and up to the present time horticulturists and ento en tomo molo logi gist stss have have di disc scus usse sed d the the mos mostt acc accur urat atee way of measu measuri ring ng temperature developmental temperature developmental units. Arnold (1959) discussed the difficulties in measurement of thermal heat units when they were used to defin def inee th thee de devel velop opme ment nt of certa certain in crop crop var varie ietie ties. s. These These di diff ffic icul ulti ties es were the variation in temperatures between warm and cool parts of thee sa th same me se seas ason on,, betw between een warm warm an and d cool cool ye year ars, s, be betw tween een north northern ern and south souther ern n latitu latitude des, s, an and d betwe between en lo low w and and high high al alti titud tudes. es. Thes Thesee facto fac tors rs made made deter determi mina natio tion n of a thre thresh shol old d te temp mper eratu ature re di diff fficu icult lt to calculate. Wang (1960) detailed several difficulties in the use of heat units to determine plant growth. These included the fact that plants responded differently to the same environmental factor during various stages of their life cycle and that the threshold temperature was considered a constant applied to the entire life cycle of a plant but this was unsound since ince the thr hreeshold val alu ues change with th thee
144 Human Body Decomposition
advancing age of the plant. Wang also pointed out that heat units referred to the sum of the temperature, disguising the singularity of temperature changes and that growth was not linear with temperature change but followed a sigmoid curve. The heat unit system did not takee into tak into accou account nt many many othe otherr fact factor orss which which af affe fect cted ed pl plan antt grow growth th such as soil moisture, sunlight, solar radiation, wind and moisture, and dura duratio tion n of light light and final finally ly that that the the micr microcl oclim imati aticc prob proble lems ms regarding the representative quality of the temperature records used for heat unit evaluations were among the most complicated problems yet to be solved. Researchers continued to search for a more accurate meth me thod od to ac acco coun untt for for grow growth th as expr expres esse sed d by he heat at unit units. s. Allen (1976) published a modified sine wave method for calculating degree days but this was found to differ in different geographical areas and required a method of correction for local bias using linear regression. Pruess (1983) discussed (1983) discussed the various methods to that date of measuring d e gr e e d ay s f o r t h e e s t i mat i o n o f i n s e c t d e ve l o p me n t an d he ways which they could models be improved. Hefound commented thatsuggested one reason all in insect development had not practical application was that they provided no more accuracy than the use of calendar dates and that if degree day measurements were to be used in practice some compromise might be necessary between precision and utility. It would also be desirable if similar models could be used for both insects and plants. (1992) appear to have been the first researchers Vass and colleagues (1992) appear to introduce the concept of ADDs into the study of human decomposition and they quote the definition of ADDs given by Edwards et al. (1987).. The 1987 (1987) 1987 paper paper discu discusse ssess tempe temperat rature ure dev develo elopme pment nt un units its for for inse insect ctss whic which h coul could d also also be used used to dete deterrmine mine pl plan antt de deve velo lopm pmen ent. t. The use of these units is based on the knowledge that insects as well as plants are poikilothermic, that is, unlike mammals which incorporate their the ir own body body tem tempe perat ratur uree re regul gulato atory ry mecha mechanis nism, m, in insec sectt and pl plant ant deve de velo lopm pmen entt depe depend ndss on ambi ambien entt temp temper erat atur ure. e. Each Each spec specie iess ha hass a thresh thr eshold old tem temper peratu ature re below below which which devel developm opmen entt ceases ceases and a maximaximum temperature above which development slows or stops. Eventually the right amount of temperature is accumulated for development to be comp co mple lete te.. Ed Edwar wards ds,, Chan Chaney ey,, and and Berg Bergma man n pr pres esen entt thre threee meth method odss to compute developmental temperature units. The first method, degree days da ys,, is de defi fine ned d as the the amou amount nt of bi biol olog ogic ical al acti activi vity ty caus caused ed by on onee
Recent Research and Current Trends
145
degree above the threshold for one day. They do not, however, define the “threshold.” Vass et al. (1992) define ADDs as being determined by taki taking ng the the sum sum of the the ave average dail daily y te tem mper perat atur ures es,, in de degr gree eess Cent Ce ntig igra rade de,, fo forr howe howeve verr lo long ng a corp corpse se has has be been en de deco comp mpos osin ing g and and (2005) calculate calculate ADDs for each case by adding together Megyesi et al. (2005) all average daily temperatures from death until discovery, using 0 C as the base base tem temper perat ature ure bec becaus ausee fre freezi ezing ng temper temperatu ature ress severe severely ly inhibi inhibitt biological processes such as bacterial growth. In a similar way to the development of insects and plants, decomposing cadavers are poikilothermic with the chemical reactions driving the decomposition process depen dep enden dentt on enviro environme nment ntal al tem tempe perat rature ure;; some some re react action ionss will will sl slow ow down with falling temperatures and speed up with rising temperatures but the threshold and range of temperatures will vary depending on the reacti rea ction on.. The tempe temperat rature ure thres threshol hold d at whi which ch human human decom decompo posit sitio ion n state that deco decompos mpositio ition n still still ceases is not known. Vass et al. (1992) state occurs when the temperature falls to 0 C because of the increased salt concentrations inathe human body, Micozzi body, (1991) states states that no (2009) putreet al. faction occurs at temperature less Micozzi than 4 C(1991) and Janaway and Janaway state that decomposition is prevented below 5 C as both enzymatic and mic microb robial ial action action will will be halted halted.. Howev However, er, no none ne of these these autho authors rs present any research to substantiate these statements. Putrefactive gas was progressively found to form over a 3-day period in the heart and liver of a cadaver stored in a mortuary refrigerator at 4 C (Singh et al., 2009). 2009 ). In the the proc proces esss of hu human man deco decomp mpos osit itio ion n th ther eree are are very very ma many ny chemi che mical cal re react actio ions ns occurr occurrin ing g simult simultane aneou ously sly and not all will will depen depend d upon the same maximum and minimum temperature thresholds. The environmental temperature in which any specific body decomposes will also vary depending on many other factors (Mann ( Mann et al., 1990). 1990). At the present time, therefore, it is not possible to define single threshold temperatu per atures res bel below ow or above above which which hum human an decom decompo posit sition ion cease ceases, s, which which means that the definition of ADDs is still open to interpretation, as is its use in any study. Recent Rece nt re rese sear arch ch has has also also rais raised ed the the ques questi tion on of the the accu accura racy cy of temper tem peratu ature re dat data a col collec lectio tion n in human human decomp decomposi ositio tion n stu studie diess becaus becausee the geographical separation between the death scene and the nearest weather weat her statio station n necess necessita itates tes ambi ambient ent temper temperatu ature re correc correctio tion. n. A note note of ca caut utio ion n abou aboutt us usin ing g lo loca call we weat athe herr stat statio ion n da data ta when when co comp mpil ilin ing g ADDs to provide minimum PMI estimates when blowfly larval growth
146 Human Body Decomposition
is used to estimate the TSD was reported by Arc Arche herr (2 (200 004) 4) in Australia. Archer commented that obtaining temperature data from a body discovery site was essential in order to calculate minimum PMI estimates. One common practice was to collect temperature data from a body body di disc scov over ery y si site te for for comp compari ariso son n with with te temp mper erat atur uree da data ta from from a local weather station. A regression equation was constructed to compare pa re temp temper erat atur ures es at the the body body di disc scov over ery y site site wi with th th thos osee me meas asur ured ed si simu mult ltan aneo eous usly ly at th thee lo local cal weath weather er stat statio ion. n. The equa equati tion on was was then then used to retrospectively correct ambient temperatures measured at the weather station during the period the body was thought to have lain in situ. Other workers simply made a note of differences between the site temperature data and the weather station data, taking the differences into into ac acco coun untt when when perf perfor ormi ming ng a mini minimu mum m PMI; PMI; howe howeve ver, r, ho how w this this was done remained unclear. There were no data available on whether regression relationships between body discovery sites and weather stations were likely to vary over time, and no data examining the accura racy cy of corr co rrel elat atio ion n re rela lati tion onsh ship ipss ideally betw betwee een n th e si site tess and an d th thee weat weathe herr stations. Correlation data should be the collected immediately followi lowing ng body body re remo mova vall and and when when weath weather er pa patt tter erns ns an and d te temp mper erat atur ures es were similar to the period the body lay in situ, but this might not be possible because of logistical problems and changing weather patterns between the time of body discovery and the collection of correlation data. Archer Arch er co comp mpar ared ed temp temper erat atur uree data data from from si six x hypo hypoth thet etic ical al body body discovery test sites, recorded with data loggers, with temperature data from one local weather station for the same period. All hypothetical sites were within a 5-km radius from the weather station and the study was designed to test whether the timing of the correlation period after body bo dy re remo mova vall affe affect cted ed the the accu accura racy cy of re retr tros ospe pect ctiv ivee weat weathe herr data data correction. Five data logging periods were employed, the first period called the “body in situ period” recorded temperatures for 7 consecutive days that a hypothetical body would have actually lain at the site. Four Fo ur othe otherr corr correl elat atio ion n peri period odss of 10 da days ys co cons nsec ecut utiv ivee te temp mper erat atur uree re reco cord rdin ings gs,, wi with th in incr crea easi sing ng in inte terv rval alss be betw twee een, n, were were carr carrie ied d out. out. Statistical analysis of the data obtained was then carried out. Archer found that weather data retrospectively corrected using a correlation method, provided a more accurate representation of site temperatures than uncorrected data and a more accurate calculation of thermal units if ADDs were used. However the degree of improvement was highly
Recent Research and Current Trends
147
variable between correlation periods and did not always improve the accu ac cura racy cy of weat weathe herr data data.. Th This is find findin ing g emph emphas asiz ized ed the the need need to us usee gene ge nero rous us ma marg rgin inss when when pred predic icti ting ng an es esti tima mate te of mini minimu mum m PM PMI. I. A further finding was that weather conditions during the correlation and “body in situ” periods affected the outcome of the retrospective correction, in that the mean estimated site temperatures for the “body in si situ tu” perio period d rose rose si sign gnif ific ican antl tly y wi with th ti time me af afte terr the the event event,, thou though ghtt to be because average temperatures rose throughout the experiment as spring progressed to summer. Archer concluded that further research was required to determine the effects of season, distance from the weather station, sun and shade, and an d indo indoor or and and outd outdoo oorr se sett ttin ings gs,, all all of wh whic ich h coul could d impa impact ct on the the accu ac cura racy cy of the the tech techni niqu que. e. Incr Increa ease se in the the ti time me af afte terr body body re remo moval val befo be fore re co corr rrel elat atio ion n data data we were re co coll llec ecte ted d co coul uld d decr decrea ease se th thee bene benefi fits ts gained by retrospective correction and this decrease was due to differences between weather conditions during the correlation and “body in situ” periods. Jo John hnso son n et al al.. (2 (201 012) 2) publish published ed furthe furtherr res resear earch ch examin examining ing addiadditional factors affecting ambient temperature corrections of weather station tion data data in fore forens nsic ic ento entomo molo logy gy.. Am Ambi bien entt te temp mper erat atur uree da data ta were were coll co llec ecte ted d from from 16 hypo hypoth thet etic ical al bo body dy di disc scov over ery y si site tess in two two di diff ffer eren entt states, in two different seasons, and across a number of geographical re regi gion ons. s. Th They ey we were re then then comp compar ared ed with with te temp mper erat atur uree data data from from th thee nearest weather station which were all within 15 km of the body discovery sites. The accuracy of retrospective weather data collection was tested by four experiments: firstly, to test the length of the correlation period; secondly, to test the distance between body discovery sites and weather stations; thirdly, to test the periodicity of ambient temperature measurements; and finally, to assess correlation accuracy in casework scenarios. The auth The autho ors’ res resu ults lts sh show owed ed th that at corr correl elat atio ion n ga gave ve si sign gnif ific ican antt impr im prov ovem emen entt over over ra raw w we weat athe herr data data re rega garrdl dles esss of the the le leng ngth th of the the corr co rrel elat atio ion n peri period od,, impr improv ovin ing g the the accu accura raccy of bo body dy si site te di disc scov over ery y temperature estimation by more than 1( 6 0.5) C in 96% of correlations. Secondly, there was no effect of distance on the accuracy of correlations for either experimental season. Thirdly, results suggested that 30-minute and 3and 3-ho hour urly ly temp temper erat atur uree meas measur urem emen entt in inte terv rval alss prov provide ided d impr improve ove-ment me nt over over raw we weath ather er dat data a but but tw twic icee-da dail ily y meas measur urem emen ents ts di did d no not. t.
148 Human Body Decomposition
Lastly, when assessing correlation accuracy in casework scenarios, the greate gre aterr the differ differenc encee between between the average average weathe weatherr sta statio tion n temper temperaature tu ress for for the the in si situ tu body body peri period od and and th thee aver averag agee weat weathe herr stat statio ion n temperatures for the correlation period, the greater was the accuracy. In those cases where the accuracy was not good, it was made acce ac cept ptab able le by exte extend ndin ing g the the dura durattion ion of th thee corr correl elat atio ion n per erio iod. d. Johnson, Wallman, and Archer concluded that retrospective ambient temper tem peratu ature re correc correctio tion n of weat weather her statio station n temper temperatu ature re was a robust robust technique, the accuracy of which was not affected by season, the length of the correlation period, or distances up to 15 km between the site of body discovery and the weather station. The accuracy, however, was di dimi mini nish shed ed if ther theree wa wass . 5 C di diff ffer eren ence ce betw betwee een n weat weathe herr stat statio ion, n, body in situ, and correlation data. Extending the length of the correlation period from 2 to 5 days could overcome this error and in fact the auth au thor orss re reco comm mmen ende ded d the the coll collec ecti tion on of co corr rrel elat atio ion n data data for for 10 days days when possible, and at intervals not longer than every 30 minutes or 3 hours. A third study in Arizona by Dabbs by Dabbs (2010) addressed (2010) addressed the question of thee vali th validi dity ty of usin using g temp temper erat atur uree data data from from the the Nati Nation onal al Weat Weathe herr Service for calculating the PMI, by examining the correlation between temperature data collected at a study site with temperature data from the two closest weather stations, 5.7 and 9.9 km distant. Data loggers, which had been tested prior to commencement of the study and found to be very accurate, were placed at the study site to collect temperature data for 154 days which were then converted to a daily average temperature by averaging the highest and lowest hourly temperature over a 24-hour period. Daily average temperatures were collected from the two weather stations closest to the study site, at distances of 5.7 and 9.9 km, for the same period of 154 days. Dabbs used the ADD data obtained from the three sites, incorporating it in the model published by Megyesi by (2005) to calculate theoretical PMIs for the three sites Megyesi et al. (2005) to of data collection. PMI was calculated in days for each of the three sites for TBSs ranging from 3 to 27. The res result ultss reveal revealed ed signif significa icant nt dif differ ferenc ences es betwee between n the study study sit sitee and the weather station sites. There was an average daily temperature 0.6 C (SD 1.1 C) higher at the weather station farthest away (9.9 km) from the study site than at the study site and the largest daily difference was 5.5 C. The difference in average daily temperature between
Recent Research and Current Trends
149
the study site and the nearest weather station (5.7 km) was 2.0 C (SD 0.97 C) hi high gher er on aver averag agee at th thee weat weathe herr stat statio ion n with with a maxi maximu mum m difference of 6.5 C on 1 day. When the PMI was calculated for TBS scores between 3 and 27, the average difference between the study site and the closest weather station was 5.8 days while the average difference between the study site and the farthest away weather station was 3.55 days. Dabbs concluded that blind utilization of temperature data fr from om the the ne near ares estt we weat athe herr stat statio ion n to a body body di disc scov over ery y si site te was was no nott appr ap prop opri riat atee fo forr th thee esti estima mati tion on of the PMI PMI us usin ing g ADDs ADDs an and d that that further research was required to develop a protocol to test the accuracy of use of weather station data to estimate the PMI before such data could be used in forensic cases. In recent times researchers have been increasingly using a quantitative body decomposition score together with ADDs in modeling the TSD in humans (Ga (Gall llow oway ay et al al., ., 19 1989 89;; Va Vass ss et al al., ., 19 1992 92;; Me Megy gyes esii et al., 2005; Fitzgerald and Oxenham, 2009). 2009). However the methods of compiling and scoring vary considerably. If studies are to be comparabl blee a st stan anda dard rd me meth thod od of quan quanti tita tati tive vely ly sc scor orin ing g the the deco decomp mpos osin ing g human body is required. One such method has recently been proposed base ba sed d on th thee exami examina nati tion on of stan standa dard rdiz ized ed au auto tops psy y re repo port rtss (Hayman and Oxenham, 2016; Hayman, 2013). 2013). In Australia there is a national databa dat abase, se, the Nation National al Cor Coroni onial al Inf Inform ormati ation on Sys System tem (NCIS) (NCIS),, which which hass list ha listed ed all all ca case sess re repo port rted ed to a coro corone nerr sinc sincee the the year year 2000 2000.. This This provides a valuable source of information concerning all decomposed human bodies reported since that time. All autopsy reports are listed in a standard form giving detailed descriptions of the decomposition affecting the various organs of the body as well as the external appearance an ce of th thee body body.. Poli Police ce re repo port rtss co conc ncer erni ning ng deta detail ilss of th thee di disc scov over ery y and, in some cases, an estimation of the time of death of the deceased, aree also ar also re reco cord rded ed.. From From thes thesee re repo port rtss it was was pos possi sibl blee to co comp mpil ilee quantitative decomposition scores for certain key organs: brain, heart, live liver, r, and and sple spleen en as we well ll as ext exter erna nall appe appear aran ance ce of th thee body body,, whic which h appeared to decompose in a regular manner with the passage of time up to 14 days. An unexpected finding from the NCIS database was that the great majority of decomposed bodies found in Australian conditions are found indoors in houses or units and so the models relate to th this is envi enviro ronm nmen ent. t. Ea Each ch of th thee orga organ n sc scor ores es when when tota totale led d gave gave a TBS. A total dataset of 239 cases in which the TSD was known within 1 da day y was was comp compil iled ed fr from om th thee stat states es of Ne New w Sout South h Wa Wale les, s, Vict Victor oria ia,,
150 Human Body Decomposition
Tasmania, Tasmani a, and the Nor Northe thern rn Territ Territory ory.. Datase Datasets ts wer weree also also compil compiled ed for each of these states separately. Bivari Biva riat atee re regr gres essi sion on mode modeli ling ng was was ca carr rrie ied d out out wi with th TSD TSD as the the depe de pend nden entt vari variab able le and and the the TBS as th thee in inde depe pend nden entt vari variab able le.. Good Good correlation was found between TSD and TBS for all datasets. In the 2
total dataset (239 cases) the correlation coefficient r was 0.874 and r was 0.764, indicating that 76.4% of the variability in the determination of the the TSD TSD wa wass acco accou unted nted for for by th thee TBS. TBS. The The stan standa dard rd er erro rorr of thee esti th estima mate te (SEE (SEE)) was was 2.00 2.009. 9. Howe Howeve ver, r, 23.6 23.6% % of th thee vari variab abil ilit ity y was accounted for by some other unknown factor or factors. Results forr the fo the da data tase sets ts of each each stat statee we were re even even bett better er,, for for ex exam ampl ple, e, New New 2 South Wales, r 5 0.970, r 5 0.941 (SEE 5 0.755); Victoria, Victoria, r 5 0.976, r2 5 0.953 (SEE 5 0.733). As temperature is the other major variable affect aff ecting ing decomp decomposi ositio tion, n, variou variouss com combin binati ations ons of tem temper peratu ature re data data compiled from weather stations nearest to where each body was found, such as average daily temperature, average daily high and low temperatures, and final day temperatures as well as humidity parameters, were factored in to multiple regression models using TSD as the dependentt variab den variable le but no par parame ameter ter gave a statis statistic ticall ally y signif significa icant nt result result.. Alth Al thou ough gh AD ADDs Ds were were comp compil iled ed,, th this is meas measur uree coul could d not not be used used in a statistical model as it correlates directly with the TSD, that is, it is another measure of the passage of time. Models Mode ls we were re comp compil iled ed for for ea each ch stat statee and and te terr rrit itor ory y whic which h gave gave an estimate of the TSD but only up to 14 days; after this time the assess ass essmen mentt of the TBS became became increa increasin singly gly diffic difficult ult and inaccu inaccurat rate. e. Temp Te mper erat atur uree was was argu arguab ably ly an im impo port rtant ant va vari riab able le and and al alth thou ough gh th this is could not be shown in a regre gression mod model, its importan ancce was ill illust ustrat rated ed when when fitted fitted and observ observed ed relati relations onship hipss were were compil compiled ed for each of the states and the Northern Territory. These clearly showed that in the Northern Territory, a tropical region with a much higher average temperature year round, a body decomposed at a faster rate than any of the southern states over the same period of time. This st This stud udy y sh show owed ed the the im impo port rtan ance ce of comp compil ilin ing g accu accura rate te data data concerning not only decomposed bodies but all forensic cases generally on a nati nation onal al basi basis. s. Aust Austra rali lia a is almo almost st al alon onee in the the worl world d in th thee compilation of such data but in many countries the collection of such data may data may prove prove impo imposs ssib ible le be beca caus usee of organ organiz izati ation onal al and and po poli liti tical cal reasons.
Recent Research and Current Trends
151
CONCLUSION
The present trend is to use ADDs in conjunction with TBS in order to find find mor more accu accurrat atee mo mode dels ls to qu quan anti tify fy the the PMI; MI; howe howeve ver, r, at the present time the determination of ADDs is inaccurate because the temperature at which decomposition ceases is not known and the actual temperature during which time a body is decomposing is also difficult to measure. The decomposition process is caused by many chemical reacti rea ctions ons,, many many of which which may have have differ different ent temper temperatu ature re thresh threshold oldss cont co ntro roll llin ing g thei theirr prog progre ress ssio ion. n. In ad addi diti tion on the the ambi ambien entt te temp mper erat atur uree will also vary depending on the environment in which the body decomposes pos es and other other factor factorss such such as mag maggot got act activi ivity, ty, animal animal scaven scavengin ging, g, and wounds to a body which affect the rate of decomposition. The use of a TBS TBS suff suffer erss fr from om the the inac inaccu currac acy y of the here re bein being g no stan standa dard rd method of descriptively defining one or uniformly quantifying one so that th at di diff ffer eren entt stud studie iess can can be stan standa dard rdiz ized ed and comp compar ared ed.. It ma may y be that further improved accuracy in estimating the TSD must await the advent and development of quantum computing which may be able in futu fu ture re to fact factor or in the the mult multip iple le variab variable less in invol volved ved in the the proc proces esss of decomposition.
REFERENCES Adl Adlam, am, R.E R.E., ., Sim Simmon mons, s, T., 200 2007. 7. The eff effect ect of rep repeat eated ed phy physic sical al dis distur turban bance ce on sof softt tis tissue sue decomposition decomposition are taphonomic studies an accurate reflection of decomposition? J. Forensic Sci. 52 (5), 1007 10071014. Allen, J.C., 1976. A modified sine wave method for calculating degree days. Environ. Entomol. 5, 388 388 396. Archer,, M.S Archer M.S., ., 200 2004. 4. The eff effect ect of tim timee aft after er bod body y dis discov covery ery on the acc accura uracy cy of ret retros rospec pectiv tivee weather station ambient temperature corrections in forensic entomology. J. Forensic Sci. 49 (3), 17. Arnold, C.Y., 1959. The determination and significance of the base temperature in a linear heat system. Proc. Am. Soc. Hortic. Sci. 74, 430 430445. Dabbs, G.R Dabbs, G.R., ., 201 2010. 0. Cau Cautio tion! n! All dat data a are not cre create ated d equ equal: al: the haz hazard ardss of usi using ng Nat Nation ional al Weather Service data for calculating accumulated degree days. Forensic Sci. Int. 202, e49 e49 e52. Edwards, R., Chaney, B., Bergman, M., 1987. Temperature developmental units for insects. Pest Crop Newslett. Purdue Univ. 2, 5 56. Fitzgeral Fitzge rald, d, C.M C.M., ., Oxe Oxenha nham, m, M. M.,, 200 2009. 9. Mod Modell elling ing tim time-s e-sin incece-dea death th in Aus Austra tralia lian n tem temper perate ate conditions. Aust. J. Forensic Sci. 41 (1), 27 2741. Fuller, M.E., 1934. Insect inhabitants of carrion: a study in animal ecology. Aust. Counc. Sci. Ind. Res. Bull. 82, 1 162. Galloway, A., et al., 1989. Decay rates of human remains in an arid environment. J. Forensic Sci. 34 (3), 607 607616.
152 Human Body Decomposition Hayman, J., 2013. Towards a More Accurate Estimation of the Time Since Death in Human Bodi Bo dies es Fo Foun und d De Deco comp mpos osed ed in Au Aust stra rali lian an Co Cond ndit itio ions ns.. Ph Ph.D .D.. Th Thes esis is.. Au Aust stra rali lian an Na Nati tion onal al University, Canberra. Hayman J., Oxenham M., 2016. Estimation of the time since death in decomposed bodies found in Australian Australian conditions. conditions. Aust. J. Forensic Forensic Sci. 48 (2), 171 171185. Hens nssg sge, e, C. C.,, Ma Made dea, a, B. B.,, 20 2007 07.. Es Esti tima mati tion on of th thee ti time me si sinc ncee de deat ath. h. Fo Fore rens nsic ic Sc Sci. i. In Int. t. 16 165, 5, He 182 182 184. Howden, A.T., 1950. The Succession of Beetles on Carrion. Unpublished Thesis. North Carolina State College, Raleigh, NC. Janawa Jana way, y, R. R.C. C.,, Pe Perc rciv ival al,, S. S.L. L.,, Wi Wils lson on,, A. A.S. S.,, 20 2009 09.. De Deco comp mpos osit itio ion n of hu huma man n re rema main ins. s. In: Percival, S.L. (Ed.), Microbiology and Aging. Springer Science and Business Media, LLC. Johnson, A.P. Johnson, A.P.,, Wall Wallman, man, J.F., Archer, M.S. M.S.,, 2012 2012.. Expe Experimen rimental tal and case casework work vali validatio dation n of ambient temperature corrections in forensic entomology. J. Forensic Sci. 57 (1), 215 221. Mann, R.W Mann, R.W., ., Bas Bass, s, W.M W.M., ., Mea Meadow dows, s, L., 1990. Tim Timee sin since ce dea death th and dec decomp omposi ositio tion n of the human body: variables variables and observations observations in case and experimental experimental field studies. J. Foren Forensic sic Sci. 35 (1), 103 103111. Mégnin P., 1894. Mégnin 1894. La Faune Faune des Cadavr Cadavres. es. Applic Applicati ation on d’Entom Entomolog ologie ie à la Médecine Médecine Légale. Légale. Gauthier et fils, Paris. Megyesi, M.S., Nawrocki, S.P., Haskell, N.H., 2005. Using accumulated degree days to estimate the post-mortem post-mortem inte interval rval from decom decomposed posed human remains. J. Foren Forensic sic Sci. 50 (3), 19. Michaud, J.-P., Moreau, G., 2011. A statistical approach based on accumulated degree-days to predict decomposition-related processes in forensic studies. J. Forensic Sci. 56 (1), 229 232. Micozzi, Micozz i, M.S M.S., ., 199 1991. 1. Pos Post-m t-mort ortem em Cha Change nge in Hum Human an and Ani Animal mal Rem Remain ains: s: A Sys System temati aticc Approach. Approa ch. Charles C. Thoma Thomas, s, Springfield, Springfield, IL, p. 40. Ch. 5. Myburgh, J., et al., 2013. Estimating the post-mortem interval (PMI) using accumulated degree days (ADD) in a temperate region of South Africa. Forensic Sci. Int. 229, 165.e1 165.e1165.e6. Notter, S.J., Stuart, B.H., Rowe, R., Langlois, N., 2009. The initial changes of fat deposits during the decomposition of human and pig remains. J. Forensic Sci. 54 (1), 195 195 201. Payne, J.A., 1965. A summer carrion study of the baby pig Sus scrofa Linnaeus. scrofa Linnaeus. Ecology 46 (5), 592 592 602. Pruess, Prue ss, K. K.P. P.,, 19 1983 83.. Da Dayy-de degr gree ee me meth thod odss fo forr pe pest st ma mana nage geme ment nt.. En Envi viro ron. n. En Ento tomo mol. l. 12 (3 (3), ), 613 613 619. Réaumu Réau mur, r, R. R.A. A.F. F.,, 17 1735 35.. Obs Obser erva vati tion onss du th thér érmo momè mètre tre fa fait ites es à Pa Pari riss pe penda ndant nt l’ann année ée 173 1735, 5, comparées avec celles qui ont été faites sous la ligne; à l’Isle de France, à Alger et en quelques-unes de nos Isles de l’Amérique. Mémoires de l’Académie des Sciences 545 545576. Reed, H.B., 1958. A study of dog carcass communities in Tennessee with special reference to the insects. Am. Midl. Nat. 59 (1), 213 213245. Rodriguez, W.C., Bass, W.M., 1983. Insect activity and its relationship to decay rates of human cadavers in east Tennessee. J. Forensic Sci. 28 (2), 423 432. Simmons, T., Adlam, R.E. Simmons, R.E.,, Moffa Moffatt, tt, C., 2010 2010.. Debug Debugging ging decomposition decomposition data comparative taph ta phon onom omic ic st stud udie iess an and d th thee in infl flue uenc ncee of in inse sect ctss an and d ca carc rcas asss si size ze an and d de deco comp mpos osit itio ion n ra rate te.. J. Forensic Sci. 55 (1), 8 813. Singh, M.K.C., O’Donnell, C., Woodford, N.M., 2009. Progressive gas formation in a deceased person per son dur during ing mor mortua tuary ry sto storag ragee dem demons onstra trate ted d on com comput puted ed tom tomogr ograph aphy. y. For Forens ensic ic Sci Med Med.. Pathol. 5, 236 236242. Suckling, J.K., 2011. A Longitudinal Study on the Outdoor Human Decomposition Sequence in Central Texas. MSc Thesis. Texas State University.
Recent Research and Current Trends
153
Sung Sung Tz’u, 11 1186 86 12 1249 49.. The The Washi Washing ng Aw Away ay of Wron Wrongs gs.. Tr Tran ansl slat ated ed fr from om the the Chin Chines esee by McKnight, B.E., 1981. University of Michigan, Ann Arbor, p. 87. Sutherland, A., Myburgh, J., Steyn, M., Becker, P.J., 2013. The effect of body size on the rate of decomposit decom position ion in a tempe temperate rate region of South Africa. Africa. Foren Forensic sic Sci. Int. 231, 257 257262. Vass, A.A., et al., 1992. Time since death determinations in human cadavers using soil solution. J. Forensic Sci. 37 (5), 1236 12361253. Vass, A.A Vass, A.A., ., et al. al.,, 200 2002. 2. Dec Decomp omposi ositio tion n che chemis mistry try of hum human an rem remain ains: s: a new me metho thodol dology ogy for determining the post-mortem interval. J. Forensic Sci. 47 (3), 542 542 553. Vass, A.A., 2011. The elusive post-mortem interval. Forensic Sci. Int. 204, 34 40. Wang, J.Y., 1960. A critique of the heat unit approach to plant response studies. Ecology 41 (4), 785 785 790.
View more...
Comments
