J Forensic Sci, 2014

doi: 10.1111/1556-4029.12424
PAPER Available online at: onlinelibrary.wiley.com

TOXICOLOGY; PATHOLOGY/BIOLOGY

Shane Darke,1 Ph.D.; Michelle Torok,1 M.Soc.Sci.; and Johan Duflou,1,2,3,4 M.Med.Path. (Forens),
F.R.C.P.A.

Sudden or Unnatural Deaths Involving

Anabolic-androgenic Steroids*

ABSTRACT: Anabolic-androgenic steroids (AASs) are frequently misused. To determine causes of death, characteristics, toxicology, and
pathology of AAS positive cases, all cases (n = 24) presenting to the New South Wales Department of Forensic Medicine (1995–2012) were
retrieved. All were male, and the mean age was 31.7 years. Deaths were mainly due to accidental drug toxicity (62.5%), then suicide (16.7%)
and homicide (12.5%). Abnormal testosterone/epitestosterone ratios were reported in 62.5%, followed by metabolites of nandrolone (58.3%),
stanozolol (33.3%), and methandienone (20.8%). In 23 of 24 cases, substances other than steroids were detected, most commonly psychostimu-
lants (66.7%). In nearly half, testicular atrophy was noted, as was testicular fibrosis and arrested spermatogenesis. Left ventricular hypertrophy
was noted in 30.4%, and moderate to severe narrowing of the coronary arteries in 26.1%. To summarize, the typical case was a male polydrug
user aged in their thirties, with death due to drug toxicity. Extensive cardiovascular disease was particularly notable.

KEYWORDS: forensic science, steroids, toxicology, psychostimulants, cardiovascular disease, demographics

Anabolic-androgenic steroids (AAS) include exogenous testos- with increased risk of peliosis hepatitis, cholestasis, and hepatic
terone, synthetic testosterone, and synthetic testosterone deriva- tumors (1,7,18). AAS are also psychotropic and are associated
tives (1). While such drugs have valid medical applications, with increased aggressiveness, agitation, paranoid ideation, mood
human and animal AAS are frequently misused to enhance ath- swings, and depression (1,11,12). In the few steroid positive case
letic performance, strength, and image (e.g., body building) series that have been conducted, death by violence, from other or
(2–7). Users of these drugs are mostly male, and aged predomi- to self, formed a substantial proportion of deaths (11,12).
nantly between the late teens and mid-thirties (2–8). Apart from One final factor that may contribute to premature mortality
muscular development, the regular use of AAS is associated with among AAS users is polypharmacy. There is emerging evidence,
a number of direct physiological changes, including testicular among both current users and fatalities, that substance use
atrophy and arrested spermatogenesis, while females may experi- among this population extends far beyond AAS (3,5,20–22).
ence masculinization (1). There is also evidence of a dependence There appear to be increased rates of use of opioids and psycho-
syndrome among regular users (9,10). Indeed, in one recent study stimulants in particular but, apparently, not of cannabis or
of AAS users, 30% met DSM criteria for dependence (10). tobacco.
There is emerging evidence that the use of AAS is associated Despite their widespread use, there are few case series of ste-
with increased risk of premature mortality (7,11,12). Much of this roid positive cases of unnatural or sudden death. This study
risk is thought to arise from the physiological and psychological aimed to determine the circumstances of death, demographic
sequelae of AAS. There is mounting evidence that these drugs characteristics, toxicology, and major organ pathology of cases
cause cardiovascular disease, including cardiomegaly, left ventric- presenting to the Department of Forensic Medicine (DOFM),
ular hypertrophy, cardiomyopathy, ischemic heart disease, fibro- Sydney, over the period January 1, 1996 to December 31, 2012
sis, and contraction band necrosis (1,12–19). Indeed, it has been in which anabolic steroids were detected during quantitative
noted that the profile of cardiovascular damage from AAS abuse toxicological investigations.
is similar to the pathology seen from psychostimulant abuse
(1,7,18). Liver disease also appears higher among AAS users,
Methods
1
National Drug and Alcohol Research Centre, University of New South Case Identification
Wales, Sydney, NSW, 2052, Australia.
2
Department of Forensic Medicine, South Western Sydney Local Health All cases autopsied at the DOFM between January 1, 1996 and
District, PO Box 90, Glebe, NSW, 2037, Australia. December 31, 2012 were identified in which anabolic steroids
3
School of Medical Sciences, University of New South Wales, Sydney, were detected in urine taken at autopsy. Police death investigation
NSW, 2052, Australia. summaries and autopsy reports of all such cases were retrieved
4
Sydney Medical School, University of Sydney, NSW, 2008, Australia.
*The National Drug and Alcohol Research Centre at the University of
from the database of the DOFM. The DOFM is located in central
NSW is supported by funding from the Australian Government. Sydney and is the primary forensic pathology center in New South
Received 18 Dec. 2012; and in revised form 25 April 2013; accepted 24 Wales (NSW), conducting between 2000 and 2500 autopsies per
May 2013. year during this time period. Permission to inspect the files was

© 2014 American Academy of Forensic Sciences 1

2 JOURNAL OF FORENSIC SCIENCES

received from the Sydney Local Health District Human Research TABLE 1––Characteristics of cases of sudden or unnatural death in which
Ethics Committee. All cases were reviewed by the authors. anabolic-androgenic steroids were detected.
In NSW, a case must be reported to the coroner where a per-
(N = 24)
son dies a violent or unnatural death. The majority of such cases
undergo a standardized forensic autopsy, with examination of all Age (mean years) 31.7 years (SD 6.8, range 22–48 years)
major organs and quantitative toxicological analysis. Cause of Sex (% male) 100%
Marital status (%)
death is determined by the forensic pathologist on the basis of Single 45.8% (11)
circumstances of death, the autopsy findings, and the toxicologi- Married/cohabiting 25.0% (6)
cal analyses. Circumstances of death, and case histories, were In relationship 29.2% (7)
obtained from accompanying police reports to the coroner. Employment status
Employed 66.7% (19)
Unemployed 16.7% (4)
Toxicological Analyses Student 4.2% (1)
Bodybuilder 20.8% (5)
All autopsy blood samples were taken peripherally (femoral or BMI (mean) 29.6(SD 3.8, range 22.2–39.9)
subclavian vessels). Toxicological data were reported for Death particulars
Accidental drug toxicity 54.2% (13)
alcohol, cannabis (determined by the presence of D-9-THC), opi- Combined accidental drug 8.3% (2)
oids (e.g., morphine, methadone, buprenorphine, oxycodone, toxicity/cardiovascular
hydrocodone, tramodol), psychostimulants (methamphetamine, disease
cocaine, benzoylecgonine, 3,4methylenedioxymethamphetamine Suicide 16.7% (4)
[MDMA]), benzodiazepines (e.g., diazepam, oxazapam, temaze- Homicide 12.5% (3)
Accident 4.2% (1)
pam, flunitrazepam, alprazolam), gammahydroxybutyrate (GHB), Undetermined 4.2% (1)
antidepressants (e.g., amitriptyline, citalopram, moclobemide,
venlafaxine, sertraline, dothiepin, imipramine, fluoxetine), and
antipsychotic medications (e.g., thioridazine, chlorpromazine,
olanzapine, lithium). In cases where there was prolonged hospi- relationship. Two-thirds were employed, the most common occu-
talization prior to death, antemortem toxicology taken at admis- pations being security guard (n = 5) and personal fitness trainer
sion was reported, and drugs administered by hospital and (n = 4). Bodybuilding was noted in a fifth of cases. The mean
medical staff excluded. All samples were screened by immuno- BMI was 29.6 kg/m2, with 45% exceeding the cutoff for obesity
assay and either by gas chromatography or by high-performance (>30 kg/m2). Evidence of recent injecting drug use was noted in
liquid chromatography (HPLC) for drugs of abuse and common 54.2% of cases. Two cases were hepatitis C virus (HCV) posi-
therapeutic substances. All analyses of blood were conducted by tive, and one was HIV positive on serologic testing.
the Forensic Toxicology Laboratory of the NSW Forensic & The most common direct cause of death was drug toxicity
Analytical Science Service (formerly the Division of Analytical which, in conjunction with combined drug toxicity/cardiovascu-
Laboratories). lar disease, constituted two-thirds of cases. Psychostimulant tox-
Urine samples taken at autopsy were screened for AAS metab- icity was the direct cause of death in eight cases (33.3%) and
olites by the Australian Sports Drug Testing Laboratory, using opioid toxicity in 7 (29.2%). Arsenic poisoning was diagnosed
gas chromatography and mass spectrometry, there being no cur- in one case – the manner of death in this case was not able to
rent blood test. In addition, abnormalities in steroid profiles were be determined, but homicide was suspected. Violent death (sui-
examined for evidence of exogenous testosterone use, with a tes- cide, homicide) constituted a quarter of cases. Three of the four
tosterone/epitestosterone ratio of >4 considered evidence of suicides were by gunshot, with one hanging. All three definite
exogenous use by the Australian Sports Drug Testing Laboratory homicides were gunshot victims. The sole accidental death was
(23). As is the case in other forensic institutes (11,12), AAS are of a pilot of a light aircraft with controlled flight into terrain,
not routinely screened for by the NSW Forensic & Analytical situational unawareness being the likely cause of the crash.
Science Service. Such tests are conducted at the request of the There were no deaths solely attributable to natural causes.
forensic pathologist when there are signs (e.g., a pronounced,
overdeveloped musculature), or circumstances indicating sus-
Toxicology
pected AAS use (e.g., substances discovered at the death scene).
It should be noted that the field of AAS use, and the detection Abnormal testosterone/epitestosterone ratios >4 were reported
of such use, is in constant flux as new drugs are developed, and in 15 of 24 cases, indicative of exogenous testosterone use, the
testing menus are likely to vary over time. most common indicator of AAS, followed by with metabolites
were of nandrolone, stanozolol, and methandienone (Table 2). In
Statistical Analyses one case, a significantly increased concentration of dehydroepi-
androsterone was reported. Tamoxefin was also present, a drug
For continuous distribution means, standard deviations (SD) frequently used in AAS steroid cycles (25). The presence of
and ranges were reported. All analyses were conducted using increased dehydroepiandrosterone and of tamoxifen in this case
SPSS Statistics, 20.0 (24). is strongly suggestive of exogenous steroid use. Nandrolone may
be present in low concentrations in urine from exogenous and
endogenous sources. We did not have access to more detailed
Results data on nandrolone concentrations or specific metabolites. In 12
Case Characteristics of these 14 cases, however, metabolites of other AAS were
detected and/or increased testosterone/epitestosterone ratios, con-
A total of 24 cases were identified, all male, with a mean age sistent with AAS use, while in the remaining two cases, other
in their early thirties (Table 1). The majority were in a sexual evidence of AAS use was noted.
 DARKE ET AL. . ANABOLIC-ANDROGENIC STEROIDS 3

TABLE 2––Toxicology of cases of sudden or unnatural death in which TABLE 3––Major autopsy findings of cases of sudden or unnatural death in
anabolic-androgenic steroids were detected. which anabolic-androgenic steroids were detected.

(N = 24) Pathology (n = 23)

Abnormal testosterone/epitestosterone ratio 62.5% (15) Steroid-related pathology
Anabolic-androgenic steroid metabolites Testicular atrophy* 45.8% (11)
Nandrolone 58.3% (14) Testicular fibrosis 43.5% (10)
Stanozolol 33.3% (8) Arrested spermatogenesis 47.8% (11)
Methandienone 20.8% (5) Cardiac
Boldenone 8.4% (2) Left ventricular hypertrophy 30.4% (7)
Trenbolone 4.2% (1) Moderate–severe atherosclerotic occlusion 26.1% (6)
Dehydroepiandrosterone 4.2% (1) Fibrosis
Metenolone 4.2% (1) Myocardial fibrosis 13.0% (3)
Oxymetholone 4.2% (1) Interstitial fibrosis 8.7% (2)
Psychostimulants (%) 66.7% (16) Perivascular fibrosis 13.0% (3)
Cocaine 41.7% (10) Cardiomyopathy 4.3% (1)
Methamphetamine 37.5% (9) Myocarditis 8.7% (2)
MDMA 16.7% (4) Myocyte necrosis 13.0% (3)
Opioids (%) 37.5% (9) Hepatic
Morphine 37.5% (9) Cirrhosis 8.7% (2)
Codeine 25.0% (6) Fibrosis (excluding cirrhosis) 4.3% (1)
Oxycodone 8.4% (2) Steatosis (moderate–severe) 8.7% (2)
Tramadol 4.2% (1) Hepatomegaly 13.0% (3)
Pholcodine 4.2% (1) Pulmonary
Cannabis (%) 0.0% (0) Bronchopneumonia 21.7% (5)
Alcohol (%) 25.0% (6) Bronchitis 8.7% (2)
Benzodiazepines (%) 45.8% (11) Birefingement material 25.0% (6)
Diazepam 29.2% (7) Emphysema 4.3% (1)
Alprazolam 20.8% (5) Renal
Oxazepam 8.4% (2) Nephrosclerosis 8.7% (2)
Temazepam 8.4% (2) Fibrosis 8.7% (2)
Flunitrazepam 4.2% (1)
Antidepressants (%) 4.2% (1) *n = 24.
Venlafaxine 4.2% (1)
injection of tablet preparations or other insoluble materials) was
detected in a quarter of cases. Other pulmonary pathology was
Blood toxicology was available for all cases. In all but one noted in six cases (26.1%), most commonly bronchopneumonia.
case (95.8%), psychoactive substances other than steroids were Renal pathology was diagnosed in three cases (13.0%). No pre-
detected. The most prevalent substances were psychostimulants, existing intracranial pathology was identified, and there were no
present in two-thirds of cases. Opioids were detected in 37.5% tumors typically associated with AAS use.
of cases, with morphine present in all these cases. A quarter had
alcohol detected, with a mean concentration of 0.093 g/100 mL
Discussion
(SD 0.095, range 0.010–0.247 g/100 mL). Benzodiazepines were
detected in nearly half of cases, most commonly diazepam. Ven- The cases in this series were exclusively male and aged on
lafaxine was present in one case, as was ketamine, while no case average in their early thirties, characteristics consistent with ear-
tested positive for cannabis. lier case studies and series (7,11–13,15–17). There was extensive
physical evidence of steroid abuse, most notably testicular atro-
phy, and fibrosis, as well as arrested spermatogenesis. Muscular
Autopsy Findings
overdevelopment was present in almost all cases, with half hav-
Muscular physique was noted in 21 cases (87.5%). In nearly ing BMIs in the obese range. As reported elsewhere (4), these
half of cases, testicular atrophy, testicular fibrosis, and arrested high BMIs were not due to obesity, but to muscle volume and
spermatogenesis were noted (Table 3). No case exhibited density. A large proportion of cases were employed as security
gynecomastia, acne, or prostate enlargement. guards or as personal trainers, occupations in which AAS may
Cardiac pathology was diagnosed in 11 cases (47.8%). The offer professional advantage.
mean heart weight was 467.7 g (SD 107.7, range 304–760 g), Exogenous testosterone appeared to be the most frequent form
with left ventricular hypertrophy present in seven cases (30.4%). of AAS use, as evidenced by metabolites of nandrolone, stanozo-
Moderate to severe stenosis (>50%) of the coronary arteries was lol, and methandienone. An outstanding feature of these cases
diagnosed in a quarter of cases. Fibrosis of the myocardium was was their extensive polypharmacy. All but a single case had psy-
noted in three cases (13.0%). Psychostimulants were present in choactive substances other than AAS detected in blood, and death
all six cases who exhibited moderate–severe coronary artery ste- was due to drug toxicity nearly two-thirds. Half were injecting
nosis, 6 of 7 cases of left ventricular hypertrophy, 2 of 3 cases substances other than steroids, and a quarter showed evidence of
of fibrosis of the myocardium, 1 of 1 case of cardiomyopathy, 2 likely injection of tablet preparations. Most prominent were the
of 3 cases of myocyte necrosis, and 2 of 2 cases of myocarditis. psychostimulants, although opioids were also prevalent. In terms
Clinically significant hepatic pathology was diagnosed in six of the latter, in eight of the nine morphine positive cases, the evi-
cases (26.1%). Two cases had cirrhotic livers, one of which was dence was of the injection of opioids and of heroin in particular.
HCV positive on serological testing. Pulmonary edema was pres- In one case, it would appear that a codeine preparation was con-
ent in 16 cases (69.4%), 14 of which were drug overdoses. Pul- sumed. It is clear that there is a strong association between the
monary vascular birefringent material (indicating probable illicit use of AAS and the illicit use of other substances, and with
4 JOURNAL OF FORENSIC SCIENCES

injecting drug use more broadly. As is commonly seen among 5. Larance B, Degenhardt L, Copeland J, Dillon P. Injecting risk behaviour
psychostimulant and opioid users (26), there was widespread use and related harm among men who use performance- and image-enhanc-
ing drugs. Drug Alcohol Rev 2008;27:679–86.
of benzodiazepines. In contrast, no case was positive for canna- 6. Petersson A, Bengtsson J, Voltaire-Carlsson A, Thiblin I. Substance
bis, consistent with reports elsewhere of low rates of cannabis abusers’ motives for using anabolic androgenic steroids. Drug Alcohol
use (20–22). This is clearly not a drug that appeals to this popula- Depend 2010;111:170–2.
tion, who appear to prefer injectables. 7. Di Paolo M, Agozzino M, Chiara T, Luciani AB, Bassi A, Molendini L,
et al. Sudden anabolic steroid abuse-related death in athletes. Int J Car-
While there were no deaths solely attributable to natural
diol 2007;114:114–7.
causes, an outstanding feature of these cases was widespread car- 8. Hakansson A, Mickelsson K, Wallin C, Berglund M. Anabolic andro-
diovascular disease. These findings are consistent with earlier genic steroids in the general population: user characteristics and associa-
work associating AAS use with such pathology (1,12–19). The tions with substance use. Eur Addict Res 2012;18:83–90.
etiology of these pathologies is uncertain, however, given the 9. Kanayama G, Brower KJ, Wood RI, Hudson JI, Pope HG. Treatment of
anabolic-androgenic steroid dependence: emerging evidence and its
extensive use of psychostimulants seen in this and other studies implications. Drug Alcohol Depend 2010;109:6–13.
(20–22), and that both AAS and psychostimulants are associated 10. Kanayama G, Brower KJ, Wood RI, Hudson JI, Pope HG. Anabolic–
with similar pathology. Certainly, as noted above, there were androgenic steroid dependence: an emerging disorder. Addiction
other physical manifestations of steroid abuse. The majority of 2009;104:1966–78.
11. Petersson A, Garle M, Granath F, Thiblin I. Morbidity and mortality in
cases who exhibited cardiovascular disease were, however, posi-
patients testing positively for the presence of anabolic androgenic ste-
tive for psychostimulants. It may be that the combination of AAS roids in connection with receiving medical care. A controlled retrospec-
and psychostimulants increases the risk of cardiovascular disease. tive cohort study. Drug Alcohol Depend 2006;8:215–20.
Finally, death from violence occurred in over a quarter of 12. Thiblin I, Lindquist O, Rajs J. Cause and manner of death amongst users
these cases, similar in proportion to the few case series con- of anabolic androgenic steroids. J Forensic Sci 2000;45:16–23.
13. Petersson A, Garle M, Holmgren P, Druid H, Krantz P, Thiblin I. Toxi-
ducted to date (11,12). As noted earlier, AAS use is associated cological findings and manner of death in autopsied users of anabolic
with increased aggressiveness and depression (1,11,12). The androgenic steroids. Drug Alcohol Depend 2006;81:241–9.
association with illicit drug use may also be of relevance here, 14. Baggish AL, Weiner RB, Kanayama G, Hudson JI, Picard MH, Hutter
as death due to homicide and suicide comprises a significant AM, et al. Long-term anabolic-androgenic steroid use is associated with
proportion of fatalities among psychostimulant and opioid users, left ventricular dysfunction. Circ Heart Fail 2010;3:472–6.
15. Fineschi V, Baroldi G, Monciotti F, Reattelli LP, Turillazzi E. Anabolic
and with injecting drug use more generally (27). steroid abuse and cardiac sudden death. A pathologic study. Arch Pathol
As in all studies, caveats must be made. AAS are not routinely Lab Med 2001;125:253–5.
screened, but conducted where there is reason by the forensic 16. Fineschi V, Riezzo I, Centini F, Silingardi E, Licata M, Beduschi G,
pathologist to suspect AAS use, as is the case in other cases series et al. Sudden cardiac death during anabolic steroid abuse: morphologic
and toxicologic findings in two fatal cases of bodybuilders. Int J Legal
(11,12). It is thus possible that there were undetected cases not Med 2007;121:48–53.
contained in this series, particularly there were no overt reasons 17. Montisci M, El Mazloumn R, Cecchettom G, Terranova C, Ferrara SD,
to order such tests, for example, no pronounced muscularity or Thiene G, et al. Anabolic androgenic steroids abuse and cardiac death in
AAS found at the death scene. Finally, the field of AAS use, and athletes: morphological and toxicological findings in four fatal cases.
Forensic Sci Int 2012;217:e13–8.
the detection of such use, is in constant flux as new compound is
18. Wysoczanski M, Rachko M, Bergmann SR. Acute myocardial infarction
developed, and new test is required to detect them. Again, it is in a young man using anabolic steroids. Angiology 2008;59:376–8.
possible that cases have not been detected. In relation to this, and 19. Denham BE. Association between narcotic use and anabolic-androgenic
the previous point, it must be noted that this is not an epidemio- steroid use among American adolescents. Subst Use Misuse
logical study, but a clinical study of case presentations where such 2009;44:2043–61.
20. Dodge T, Hoagland MF. The use of anabolic androgenic steroids and
drugs have been detected. The data also do not allow us to deter- polypharmacy: a review of the literature. Drug Alcohol Depend
mine the length and severity of AAS use among cases, nor their 2011;114:100–9.
polypharmacy histories. Such data would be useful in attributing 21. Garevik N, Rane A. Dual use of anabolic-androgenic steroids and narcot-
causation to the cardiovascular disease seen in this series. ics in Sweden. Drug Alcohol Depend 2010;109:144–6.
22. Skarberg K, Nyberg F, Engstrom I. Multisubstance use as a feature of
In summary, the typical case in this series was a male poly-
addiction to anabolic-androgenic steroids. Eur Addict Res 2009;15:99–106.
drug user, aged in their early thirties, with drug toxicity the 23. Mareck U, Geyer H, Fussholler G, Schwenkem A, Haenelt N, Piper T,
being most common cause of death. Extensive cardiovascular et al. Reporting and managing elevated testosterone/epitestosterone
disease was a particularly notable feature. ratios-novel aspects after five years’ experience. Drug Test Anal
2010;2:637–42.
24. IBM. SPSS statistics, 20.0. Chicago, IL: IBM, 2011.
Acknowledgments 25. Baker JS, Graham MR, Davies B. Steroid and prescription medicine
abuse in the health and fitness community: a regional study. Eur J Intern
The authors wish to thank Vera Alderton and Briony Larance Med 2006;17:479–84.
for their assistance. 26. Darke S, Torok M, Kaye S, Ross J, McKetin R. Comparative rates of
violent crime amongst regular methamphetamine and opioid users:
offending and victimisation. Addiction 2010;105:916–9.
27. Darke S, Degenhardt L, Mattick R. Mortality amongst illicit drug users:
References epidemiology, causes and intervention. Cambridge, U.K.: Cambridge
1. Karch SB. Karch’s pathology of drug abuse, 4th edn. Boca Raton, FL: University Press, 2007.
CRC Press, 2009.
2. Brennan BP, Kanayama G, Hudson JI, Pope HG. Human growth hor- Additional information and reprint requests:
mone abuse in male weightlifters. Am J Addict 2011;20:9–13. Professor Shane Darke, Ph.D.
3. Copeland J, Peters R, Dillon P. Anabolic-androgenic steroid use disor- National Drug and Alcohol Research Centre
ders among a sample of Australian competitive and recreational users. University of New South Wales
Drug Alcohol Depend 2000;60:91–6. Sydney
4. Kao TC, Deuster PA, Burnett D, Stephens M. Health behaviors associ- NSW 2052
ated with use of body building, weight loss, and performance enhancing Australia
supplements. Ann Epidemiol 2012;22:331–9. E-mail: [email protected]