Metabolic & Cardiometabolic
anabolic-androgenic steroids causes testicular atrophy and impaired fertility
In plain terms: Do anabolic steroids cause testicular shrinkage and reduced fertility?
Strong support Metabolic & Cardiometabolic 💰 Industry COI noted
Part of: 💊 anabolic-androgenic steroids
RefutedContestedStrong support
consensus score 1.00
Yes — steroids suppress the body's own testosterone and sperm production, shrinking the testes and often causing infertility; it usually recovers over months after stopping but sometimes incompletely.
📅 Last reviewed: 2026-07-14 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: All trials, pooled (Meta-analysis)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
10 support 0 contradict 0 tested null 0 mixed · 10 sources, 10 independent groups · 1 superseded — pooled inside a review, counted once
What the evidence shows
Anabolic-androgenic steroids suppress the body's own testosterone and sperm production, shrinking the testes and often causing infertility. A meta-analysis of 32 studies in 9,371 participants and a controlled comparison of current against former users show the suppression and its partial recovery after stopping. The human evidence is observational; animal work supplies the mechanism.
The evidence (28)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Vorona E, Nieschlag E 2018 · Minerva Endocrinol | observational | supports | moderate | Review states AAS cause reversible suppression of spermatogenesis up to azoospermia via HPG-axis negative feedback. |
| Al Hashimi MA, Pinggera GM, Shah R, Agarwal A 2025 · Asian J Androl | observational | supports | moderate | Narrative clinician's guide (Asian J Androl 2025;27:330-341), not a systematic review: the full text reports no literature-search strategy, no eligibility criteria and no risk-of-bias appraisal, and its figures are all secondhand. On the claim it is unambiguous — azoospermia is 'a well-documented consequence of exogenous testosterone (ET) and anabolic-androgenic steroid (AAS) use', with HPG-axis suppression causing 'testicular atrophy by a magnitude of 16.5%-30% due to spermatogenic suppression' (citing Rasmussen: 'current ASS users exhibited significantly lower testicular volumes, averaging 12.2 (s.d.: 0.7) ml, compared to healthy controls with 22.3 (s.d.: 0.6) ml'), and 'ET-induced azoospermia is a well-recognized sequela ... and the possibility of it being a permanent complication'. The authors qualify the fertility half: 'given sufficient time for recovery, the effects of androgen abuse appear to be reversible, provided that the initial semen parameters of former users were comparable to those of the control group' (mean 14.1 months to sperm return, 37.6 months to motility recovery), while '32% of all men suffering from azoospermia due to ET or AAS remain infertile with proven azoospermia even after 1 year of treatment' (that 32% is over the azoospermic subgroup, not the 463-man series). Quality moderate rather than high: no methodological safeguards are described, the authors' own verdict is that 'there is a notable lack of well-structured, controlled, and long-term studies', and two headline numbers come from the first author's own uncontrolled series (refs 87, 91). regardless (evidence_role: review). |
| Ding JB, Ng MZ, Huang SS, Ding M, Hu K 2021 · J Sports Med (Hindawi Publ Corp) | observational | supports | moderate | Narrative review (no search strategy, no data of its own; ). States the claim's mechanism directly: 'ASIH causing testicular atrophy, infertility, and low endogenous testosterone levels' and 'Infertility following AAS use is secondary to suppression of intratesticular testosterone levels, which can lead to azoospermia or oligozoospermia.' All figures are secondhand from cited primaries: a 2021 Danish registry study of 545 AAS-positive men found a fertility rate ratio of 0.74 (95% CI 0.60-0.90, p=0.0028) in the decade before the positive test, falling to a 7% deficit after cessation with no difference in assisted-reproduction rates; a 2020 cross-sectional study of 72 users vs 31 controls found no significant difference in sperm output, concentration or motility in FORMER users (mean recovery 14, 10 and 37 months respectively), while a 2021 prospective study found 34% below 40 million total sperm count at 1 year. The review treats causation as settled and the open question as reversibility, citing a 12- to 24-month recovery window. The ~6% lifetime male misuse prevalence in the abstract is an exposure-prevalence figure, not an endpoint for this claim. |
| Mulawkar PM, Maheshwari PN, Gauhar V, Agrawal SG, Mohammed TO, Singh AG, Tak GR, Shah US, Shukla DP, Mamankar D 2023 · J Hum Reprod Sci | meta-analysis | supports | moderate | SR+meta-analysis of 32 OBSERVATIONAL studies (12 cohort, 5 case-control, 9 cross-sectional, 6 case reports; 9,371 participants incl. 2,671 AAS users) — but each pooled estimate rests on a small subset, not the headline n. Testicular atrophy: 3 studies / 170 participants, 'The testis size was significantly smaller in AAS users as compared to the naïve population (SMD −3.92, 95% CI-7.65, −0.18,' P = 0.04. Sperm motility: 3 studies / 186, SMD −2.72 (95% CI −5.46, 0.02, P = 0.05) — the authors call this 'significantly lower' although the interval includes zero. Sperm concentration: 3 studies / 154, NOT significant (SMD −1.27, 95% CI −2.84, 0.30, P = 0.11). FSH: 7 studies / 327, SMD −4.61 (95% CI −6.50, −2.73); FSH/LH still low at 6 months but no different from naive at 1 year (3 studies / 294, SMD 0.02, 95% CI −0.34, 0.38). One matched Danish cohort: fertility rate 26% lower (RR 0.74, 95% CI 0.60, 0.90), recovering to RR 0.93 after the doping sanction. Authors against themselves: 'The limitations are small sample size of most of the studies, polypharmacy, lack of information on dosing and high heterogeneity', 'considerable heterogeneity and wide confidence intervals amongst the studies', 'Finally, publication bias may exist.', and 'our results should be interpreted with caution'. Net: testicular atrophy directly supported; fertility impairment supported but partially reversible. |
| ElShebiney S, Mowaad NA, Elgohary R, Shabana ME, El-Mahdy RS, Elfiky AM 2025 · Scientific Reports 2025;15:44652 | animal | supports | low | Rat experiment: 48 male Wistar rats, n=6/group, given stanozolol 5 mg/kg s.c. weekly and/or cannabis resin 20 mg/kg i.p. daily for two months, sedentary or swim-trained. Sedentary rats on stanozolol or cannabis showed 'considerable testicular degeneration... necrosis of germinal layers'; seminiferous lumen diameter fell from 29.36 to 18.45 and epithelial height from 6.66 to 4.4 (p<0.05, n=5/group), with SIRT1 and STS mRNA downregulated. The combination arm was worst. Exercise 'could palliate the damage partially, it was not fully protective.' The paper reports no sperm count, motility or mating outcome, and gives no statistics for free testosterone. |
| Bocu K, Boeri L, Mahmutoglu AM, Vogiatzi P 2025 · Arab J Urol | observational | supports | low | Narrative review of lifestyle factors in male fertility, with no search protocol, no pooling and no primary data. On steroids it states that 'AAS users often have lower LH and FSH levels but higher testosterone and estradiol levels', that common side effects include 'reduced testis volume', and that 'AAS use decreases sperm count and quality'. Recovery of the hypothalamus-pituitary-testis axis takes 'from 24 weeks to 2 years', depending on duration and dose of use. The review reports no sample sizes, effect sizes or confidence intervals of its own. |
| Sukegawa G, Tsuji Y 2020 · Acta Urol Jpn | observational | supports | low | Case series: 6 of 4,375 men attending one infertility clinic had received therapeutic-dose testosterone for late-onset hypogonadism (3-12 months). Gonadotropins were markedly suppressed in 5 of 6; semen findings ranged from azoospermia to severe oligospermia. Testosterone was stopped in all; 3 needed hCG or clomiphene for prompt recovery, and the authors note the recovery period is highly variable. |
| Szabó A, Nyirády P, Kopa Z 2025 · Curr Opin Urol | observational | supports | low | Narrative review in Current Opinion in Urology (2025) summarising 18 months of literature on lifestyle and environmental influences on male fertility. It states that 'Anabolic-androgenic steroids (AAS) are known to suppress endogenous gonadotropin secretion and halt spermatogenesis,' and cites a review reporting that an antidoping programme prompting AAS cessation raised live birth rates among former users, with fertility approaching non-user levels over 10 years. It reports no original data, and discusses testicular atrophy only for alcohol, not for steroids. |
| Di Girolamo FG, Biasinutto C, Mangogna A, Fiotti N, Vinci P, Pisot R, Mearelli F, Simunic B, Roni C, Biolo G 2024 · Sports Med Open | observational | tested-null | low | Observational cross-sectional study of 92 Slovenian male recreational bodybuilders and 45 healthy active male controls: 'The study followed an observational cross-sectional protocol.' What was measured is stated plainly - 'We have assessed plasma concentrations of selected metabolic markers and fatty acid content in erythrocyte membranes of 92 male bodybuilders and in 45 healthy controls' - i.e. body composition, tensiomyography, fasting glucose/insulin/HOMA-IR, lipids, CETP, leptin, ALT, AST, ALT/AST ratio, CK, hs-CRP and derived elongase / Delta-9 desaturase activities (Tables 1-3). NO testicular, semen, LH/FSH or testosterone endpoint was measured anywhere. Testicular atrophy and infertility appear once, as background citing reference [42]: AAS 'suppress natural testosterone production, resulting in hypogonadotropic hypogonadism, i.e. testicular atrophy, reduced fertility, and decreased sperm production' - an assertion about the prior literature, not a finding of this study. The metabolic findings it does report are also narrower than the prior extract implied: 'HDL-cholesterol was lower in insulin and/or GH abusers' (HDL 31.3 mg/dL in insulin users n=15 and 27.1 in GH users n=12 vs 48.0 in non-using bodybuilders n=52 and 42.0 in controls n=45), while the AAS-only group (n=13) had, in the authors' own words, 'normal HDL-cholesterol'; and the raised ALT/AST ratio was selective - 'Insulin doping was selectively associated with increased plasma ALT-to-AST ratio.' Exposure was self-reported by anonymous questionnaire with no dose data, and the authors note 'the small sample sizes may limit the ability to detect small effect sizes'. |
| Christou MA, Christou PA, Markozannes G, Tsatsoulis A, Mastorakos G, Tigas S 2017 · Sports Med | observational | supports | high | Systematic review with a random-effects meta-analysis nested inside it (Ioannina/Athens): 'Thirty-three studies (three randomized clinical trials, 11 cohort, 18 cross-sectional, and one non-randomized parallel clinical trial) were included in the systematic review (3879 participants; 1766 AAS users and 2113 non-AAS users).' DENOMINATOR: the pooled arithmetic is NOT over 3,879 - 'A meta-analysis (11 studies) was conducted of studies evaluating serum gonadotropin and testosterone levels in male subjects', on subsets as small as '(six studies, n = 65 AAS users; seven studies, n = 59, evaluating gonadotropin and testosterone levels respectively)', falling to n = 17 and n = 27 for the prior-vs-post-discontinuation comparisons. What was POOLED is HPG-axis suppression, which is off-object for this claim: 'During AAS intake, significant reductions in luteinizing hormone [weighted mean difference (WMD) -3.37 IU/L, 95% confidence interval (CI) -5.05 to -1.70, p < 0.001], follicle-stimulating hormone (WMD -1.73 IU/L, 95% CI -2.67 to -0.79, p < 0.001), and endogenous testosterone levels (WMD -10.75 nmol/L, 95% CI -15.01 to -6.49, p < 0.001) were reported.' On the claim's OWN object - testicular atrophy and impaired fertility - the paper pools nothing and synthesises narratively: 'In addition, AAS abuse resulted in structural and functional sperm changes, a reduction in testicular volume, gynecomastia, as well as clitoromegaly, menstrual irregularities, and subfertility.' No pooled testicular-volume estimate and no sperm-concentration figure is reported anywhere, so the prior extract's 'reducing sperm concentration' named an endpoint the paper never quantified. evidence_role review therefore STANDS: on this object the paper re-describes primaries the vault counts elsewhere, and its 33 studies very likely overlap the 32-study meta-analysis at s38322636 - a stack flagged, not fixed. Direction is not in doubt: 'The majority of AAS users demonstrated hypogonadism with persistently low gonadotropin and testosterone levels, lasting for several weeks to months after AAS withdrawal.' Abstract-grade (NCBI efetch, no PMCID); bias testing, heterogeneity statistics and funding are the available text does not say, not absent. Re-read #772. |
| Bond P, Smit DL, de Ronde W 2022 · Frontiers in Endocrinology | observational | supports | moderate | Narrative review, section 3.10: 'spermatogenesis arrest results in testicular atrophy'; sex-steroid-induced suppression of spermatogenesis reduces testicular volume by 16.5-30.0% across the cited studies. In the HAARLEM cohort (n=100 male AAS users) mean sperm concentration fell from 46.8 to 11.7 million/mL on-cycle and 68% met oligo/azoospermia criteria (28% already did at baseline), with two-thirds azoo- or oligozoospermic at cycle end; mean time to recovery of sperm output and concentration 14.1 and 10.4 months. Reviewed evidence, not the paper's own data - no weight. |
| Kamali Hosseinzadeh F, Khayatzadeh J, Forghanifard MM, Attaranzadeh A 2024 · Translational Andrology and Urology 2024;13(12):2801-2811 | animal | supports | low | Controlled rat experiment, 24 adult male Wistar rats ('24 adult male Wistar rats, each weighing between 200 and 250 grams'), four groups (~6/group; no per-group n is ever stated), Sustanon 10, 20 or 30 mg/kg IM ONCE A WEEK for 8 weeks versus peanut-oil vehicle ('Groups 2, 3, and 4 were administered Sustanon at doses of 10, 20, and 30 mg/kg body weight, respectively, via intramuscular injection once a week for 8 weeks.'). Both halves of the claim's object were measured in the same animals. ATROPHY, dose-ordered: mean testicular weight 1.482 g control vs 0.732 / 0.672 / 0.675 g; testicular diameter 11.38 vs 8.47 / 7.74 / 7.30 mm; germinal-epithelium thickness 1.298 vs 0.655 / 0.617 / 0.608 microns. FERTILITY: sperm count 39.333 vs 24.033 / 16.65 / 15.3; normal sperm morphology 87% vs 37% / 29% / 22%; sperm DNA fragmentation 9.17 vs 28.33 / 25.58 / 22.33; histology showed 'spermatogenesis efficiency decreased in the treatment group, with differentiation not occurring at a dose of 30 mg/kg'. HORMONES: FSH 0.29 vs 0.22 / 0.21 / 0.22 IU/L and LH 0.26 vs 0.20 / 0.23 / 0.21 IU/L, with the authors noting no significant difference BETWEEN doses. BCL2 mRNA fell in all treated groups (P<0.001), read by the authors as increased apoptosis; Bax and the BAX/BCL2 ratio were not assessed 'due to resource constraints'. QUALITY LOWERED TO LOW on what the paper establishes about itself, not on its size: Tables 2 and 3 are the only quantitative results and report point means with no SD/SEM, no per-group n and no per-comparison P values, significance being asserted in prose only; and the reporting is internally contradictory in three places - the abstract says 'There was no significant increase in the expression of DAZL' while the Results say DAZL at 10 mg/kg 'was significantly higher than in the control group' (P<0.05); the Methods put the control in group 1 while the Discussion says 'doses of 10, 20, and 30 mg/kg Sustanon were injected intramuscularly into rats in groups 1, 2, and 3'; and the animals are 'housed in individual cages (six rats per cage)'. The qPCR method describes microRNA reagents ('0.5 microlitre microRNA (miRNA)-specific forward primer') for mRNA targets. Separately, the Introduction credits ref 17, 'research by Bameri et al. ( 17 ) demonstrated that Sustanon treatment resulted in significantly lower expression of the B-cell lymphoma 2 ( BCL2 ) gene', to a paper that is in fact 'Insights into the role of melatonin on the reproductive system toxicity induced by bisphenol A: a systematic review of preclinical studies.' - a BPA/melatonin review with no Sustanon arm, so the stated precedent does not exist as cited. The DIRECTION of the finding is nonetheless consistent and dose-ordered on the atrophy and sperm-count endpoints, so the row still supports the claim; it should carry little weight. Funded by Islamic Azad University Mashhad Branch (No. 183084); 'The authors have no conflicts of interest to declare.' Re-read #795, full text PMC11732302. |
| Hohl A, Lopes L, Ronsoni MF, Miranda EP, Fighera TM, Facio FN, Marchesan LB, Torres LO 2026 · Int Braz J Urol | observational | supports | moderate | Joint position statement from three Brazilian medical societies (SBEM, SBU, ABEMSS) on male hypogonadism, issuing graded recommendations rather than reporting original data. It states that anabolic-androgenic steroid use has 'increased the frequency of functional hypogonadism in Brazil', that such steroids 'block hypothalamic GnRH pulsatility and the release of pituitary gonadotropins', and that 'prolonged use may result in incomplete recovery, even years after discontinuation'. Testicular atrophy and infertility are listed among the risks of exogenous androgens. The document reports no measurements of its own. |
| Kumar N, Kakoti S, Chung E 2025 · Arab J Urol | observational | supports | moderate | Narrative review of testosterone and anabolic-androgenic steroid (AAS) abuse and male fertility; it collects other studies' results rather than reporting new data. Exogenous androgens suppress the hypothalamic-pituitary-gonadal axis so intratesticular testosterone falls, with a quoted “decrease in sperm production by 90–95%, with azoospermia in more than 40% of men with testosterone abuse” and oligozoospermia or azoospermia in “approximately 66–77%” of AAS users. In the HAARLEM cohort of 100 male athletes, testicular volume and sperm parameters “remain below baseline” one year after stopping. |
| Kanayama G, Hudson JI, DeLuca J, Isaacs S, Baggish A, Weiner R, Bhasin S, Pope HG Jr 2015 · Addiction | observational | supports | moderate | Cross-sectional comparison of 24 male former long-term AAS users (>=2 yr lifetime use, >=3 mo abstinent; mean 58.9 mo) against 36 non-AAS-using weightlifters age 35-55, Boston 2011-2014; the 5 users on testosterone replacement are excluded from every number, so the numerical denominator is 19 untreated users vs 36 controls, not 24. On the claim's ATROPHY half, measured by orchidometry: 'the 19 untreated former AAS users displayed significantly smaller testicular volumes (estimated difference [95% confidence interval (CI)]: 2.3 [0.1, 4.5] ml; p = 0.042) and lower serum testosterone levels (estimated difference: 131 [25, 227] dL; p = 0.009)', with a dose-response echo of -0.5 [-0.1, -0.9] ml per year of AAS use (p = 0.029) and 5 of 19 (26%) below 200 ng/dL after 3-26 months abstinent. Carry the fragility: the volume estimate barely excludes zero and does not survive the authors' own sensitivity analysis restricted to >=7 months abstinence, where 'for testicular volume, the difference was 2.6 [-0.1, 5.0]' (testosterone held at 127 [20, 235]). On the claim's FERTILITY half this paper is silent - measurements were SCID, hormone panel, orchidometry and IIEF; no semen analysis or sperm endpoint was performed, so it votes on atrophy only. Quality moderate: exposure verified by urine AAS, hair/urine drug screens and an FFMI plausibility screen, testosterone by CDC-HoST-certified LC-MS/MS, recruiters blinded to AAS status, adjusted for age/race/education/income - but 'the sample size in the present study was not determined in advance ... hypogonadal symptoms represented a secondary outcome of interest' and 'participants were self-selected'. Full text read (PMC4398624), re-read #769. ↩ SUPERSEDED — pooled in the review above, counted once |
| Bonnecaze AK, O'Connor T, Burns CA 2021 · J Gen Intern Med | observational | supports | moderate | Narrative review, not primary evidence: the paper states 'Prior studies of AAS use reveal an association with polycythemia, dyslipidemia, infertility, hypertension, left ventricular hypertrophy, and multiple behavioral disorders' — it re-describes other studies' observations and reports no cohort, trial or measurement of its own. It also notes 'AAS withdrawal syndrome, a state of depression, anhedonia, and sexual dysfunction after discontinuing AAS use, is a common barrier to successful cessation.' Scope caveat: only the infertility half of this claim's compound object is mentioned; testicular atrophy is the paper does not say from the available abstract text (Europe PMC returns no full text for PMC8298654). under rubric axis 2. |
| Rasmussen JJ, Selmer C, Østergren PB, Pedersen KB, Schou M, Gustafsson F, Faber J, Juul A, Kistorp C 2016 · PLoS One | observational | supports | moderate | Cross-sectional case-control study, greater Copenhagen 2014-2015: 37 CURRENT AAS abusers, 33 FORMER abusers (geometric-mean 2.5 (1.7-3.7) years since cessation) and 30 age-matched healthy controls, all recreational strength trainers aged 18-50 — every number below belongs to one of those three denominators. ATROPHY half of the object, by Prader orchidometer: 'Current AAS abusers had the smallest testicular volume (12.2 (0.7) ml) and former AAS abusers had a volume of 17.4 (0.8) ml which was 4.8 (2.9; 6.8) ml smaller than that of the control participants who had largest testicular volume' (controls 22.3 (0.6) ml), with a dose-response in former abusers of -1.3 (-2.4; -0.2) per log2 week of accumulated exposure (P = 0.02). FERTILITY half, by Sertoli-cell biomarker proxy only — no semen was analysed ('We did not obtain sperm samples') — in CURRENT abusers: inhibin B 81 (8) vs 175 (9) pg/ml and AMH 21.6 (16.3-28.7) vs 49.5 (41.6-59.0) pmol/l in controls, and 'Current AAS abusers exhibited severely decreased AMH and inhibin B indicative of impaired spermatogenesis', with 56.8% (39.5; 72.7) of current abusers below the impaired-spermatogenesis cut-off of 92 pg/ml vs 3.3% of controls. CORRECTION to the prior extract, which had this backwards: current abusers were NOT testosterone-suppressed — 'Plasma total and free testosterone levels were significantly lower among former AAS abusers than among control participants and current AAS abusers, the latter of whom exhibited significantly increased plasma testosterone levels, as expected' (98.3 nmol/l current vs 18.8 control vs 14.4 former). And the fertility impairment did NOT persist: 'Plasma gonadotropins, SHBG, 17-hydroxyprogesterone, serum AMH and inhibin B did not differ significantly between former AAS abusers and control participants'; what persisted at 2.5 years was the testicular-size deficit and lower testosterone, the latter reference-limit dependent (27.2% of former abusers below 12.1 nmol/l but only 3.3% below the population-representative 6.6 nmol/l). Quality moderate: LC-MS/MS steroids (CV 10.0%), standardised fasted morning supine sampling, objective orchidometry, a priori trend test, dose-response modelling — against self-reported exposure with 'We did not screen the urine of the participants for AAS metabolites', gym-volunteer selection, an unadjusted education gap (30.0% vs 0.0% university degree), and the authors' own 'the cross-sectional study design limited our ability to determine causality'. Full text read (PMC4988681), re-read #770. |
| Tatem AJ, Beilan J, Kovac JR, Lipshultz LI 2020 · World J Mens Health | observational | supports | moderate | Narrative review (no original data). Abstract: exogenous testosterone 'almost universally impairs spermatogenesis and can even completely eradicate it in some men'; body adds 'atrophy of the germinal epithelium and subsequent azoospermia in upwards of 40% of men'. All figures are CITED, not measured here: 40% azoospermia from de Souza & Hallak 2011; 75% azoospermia after 6 months of 200 mg weekly testosterone enanthate from the WHO 1996 contraceptive task force; 'only 67% of previously eugonadal men will recover spermatogenesis at 6 months' from Liu 2006's pooled analysis of 30 contraception studies. as a review; the primaries should be appraised directly. |
| Dos Santos JC, Souza E, Meneses-Santos D, Carvalho CRO, Dos Santos JL, Aidar FJ, Marçal AC 2024 · Eur J Investig Health Psychol Educ. 2024;14(5):1451-1469 | observational | supports | low | Cross-sectional self-report questionnaire survey of 71 competitive bodybuilders in Sergipe, Brazil (210 invited, 104 responded, 71 analysable; 94.4% n=67 ever-users). CORRECTS THE PRIOR EXTRACT, which stated no fertility endpoint was reported: a fertility endpoint IS tabulated - the acute side effects included 'decreased spermatogenesis (difficulty in having children), at 11.9% ( n = 7)' (Table 6), alongside acne 33.8% (n=20), irritability 32.1% (n=19), alopecia 23.7% (n=14) and gynecomastia 15.2% (n=9). DENOMINATOR IS A SUBSET, NOT THE 71: acute-effect percentages imply about 59 responders (the footnote states '11 out of the 69 participants did not respond to whether they experienced acute side effects') and chronic-effect percentages imply about 25 (46 did not respond), and the authors' arithmetic does not close (58 vs 59 vs 60). Testicular atrophy was NEVER MEASURED and never offered as an item - the authors write that reduced testicular volume and erectile dysfunction reported in prior syntheses 'were not mentioned by athletes, which does not mean that they will not develop these disorders if continued use occurs' - so this row supports only the impaired-fertility half of the object, and the atrophy half is absence of data, not a null. Entirely self-reported: no semen analysis, no hormone assay, no examination or imaging, no non-user comparator, convenience sample with 50% invitation response. The sentence 'A systematic review and meta-analysis reported that the excessive and long-term use of AASs results in significant sperm changes, reduced testicular volume, and erectile dysfunction' is cited background (refs 40-42), not a finding of this study. |
| Al Hashimi M, Farahat Y, Kandil H, Al Khalidi I 2023 · Health Science Reports | observational | supports | low | Cross-sectional online survey of CLINICIANS, not patients: 151 of 370 invited urologists/andrologists/endocrinologists (40.8% analysable) recruited via LinkedIn and the Arab Society of Urology list. Every percentage is a proportion of the 151 RESPONDENTS reporting an impression of their AAS-abusing patients, NOT patient-level prevalence: infertility named the commonest presentation by 64.24%, 77.48% said semen analysis is commonly abnormal, 94.70% reported any hormonal abnormality, and 50.33% perceived smaller testes -- explicit atrophy of testes was reported by only 3.97%, and 45.70% noted no change in testicular size. The 6-12 month recovery figure (42.38%) is what respondents expected, not measured recovery. Uncontrolled physician recall, convenience sample, no patient denominator. |
| Hammood MN, AL-Harbi HJ, Alhamed TA, Hammod HJ 2026 · Int J Reprod Biomed | animal | supports | moderate | Experimental study: 20 male Wistar rats (n=5/group) given oral Stanozolol 5, 10 or 15 mg/kg/day or distilled water for 60 days. Every dose significantly cut testicular weight, testicular and epididymal sperm concentration, viability and motility and raised abnormal sperm morphology (mostly p<0.001). Johnsen spermatogenesis scores fell at all doses (5 mg/kg p<0.01; 10 and 15 mg/kg p<0.0001); seminiferous tubule diameter fell at 5 mg/kg (p=0.002) and 15 mg/kg (p=0.008) but not at 10 mg/kg (p=0.084). Hormones, oxidative stress and mating fertility were not measured. |
| Easton J, Grant B, Spurling G, Magnolini R, Piatkowski T 2025 · Drug and Alcohol Review | observational | mixed | low | Qualitative study of Australian online steroid forums: 150 threads and 5059 posts from 580 users were coded against the Health Belief Model to map beliefs about post-cycle therapy. Posters described fear of hormonal imbalance, testicular shrinkage and infertility as reasons to seek post-cycle therapy, while others dismissed it. No testicular volume, semen parameter or hormone level was measured on anyone; the authors note forum data 'likely introduces selection bias, as individuals who post may be those experiencing concerns or adverse effects.' The paper reports no clinical outcome data. |
| Solanki, Eu, Smith, Allan, Lee 2023 · Endocr Connect | observational | supports | moderate | Scoping review (Monash/Alfred Health), PubMed only, inception to 26 April 2022, 811 records screened; human studies of AAS users plus testosterone contraceptive trials as models, with 'Excluded studies were those that were case studies or series, studies that were not in the English language, or studies of animal models.' No pooling, no risk-of-bias tool, no PRISMA flow — every number below is re-passed from a cited primary, not measured here, which is why this row is zero-weighted as a review. DIRECTION is not in doubt on either half of the object. ATROPHY half: 'For many individuals, testicular volume is not fully reversible after AAS cessation.' — Smit found volume 'a mean 0.7 mL lower than baseline 3 months after AAS cessation, although most individuals recovered to their baseline at 9 months'; Rasmussen's orchidometry gave controls 22.3 mL vs 17.4 mL in past abusers abstinent >2 years (current abusers 12.2 mL); Shankara-Narayana found ultrasound volume 'reduction of 32% (9 mL) compared with healthy controls' at a median 10.7 months abstinent; Kanayama 'a reduced testicular volume of 10% (2.3 mL) up to a mean of 4.9 years of previous AAS abuse'. The authors' own summary: 'Taken together, the evidence suggests that most testicular volume lost does recover after ceasing AAS use, but it may take years and often is not complete.' FERTILITY half, and this is where the prior extract flattened two different verdicts: the review concludes spermatogenesis recovery is likely COMPLETE, not incomplete — contraceptive-trial pooled data show '67% recovering by 6 months, 90% by 12 months, and 100% by 24 months', with recovery time scaling with exposure ('a mean 10 months for sperm concentration and a mean 38 months for sperm motility'), and the one prospective controlled fertility-outcome study cited reports that 'though fertility rate is significantly lower in the group of androgen users at baseline, catch up is achieved in the 2-3 years following'. So it is testicular VOLUME, not fertility, that the review leaves persistently short. Quality moderate: pre-specified search strategy, explicit eligibility criteria and reference-list scanning, no COI and no funding ('This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector'), against a single database, English-only, no formal appraisal, and the authors' own 'there is a paucity of prospective controlled data characterising the trajectory of natural recovery following cessation'. INDEPENDENCE FLAG (not fixed): re-passes Rasmussen 2016 (= s27532478, same 22.3/17.4/12.2 mL figures) and Christou 2017 (= s28258581), both already separate sources on this claim — inert while all are review-role. Full text read (PMC10620455), re-read #786. |
| Jafari H, Mirzaiinajmabadi K, Roudsari RL, Rakhshkhorshid M 2021 · Int J Reprod Biomed | observational | supports | low | Narrative systematic review (no pooling) of 14 Iranian primary studies, 26,324 infertile men. Anabolic steroids appear on the factor list from only 2 of the 14 studies ('anabolic steroids (17, 22)'), each as an uncontrolled prevalence among infertile men ('Anabolic steroids (3.5)' percent in Safarinejad 2008, n=12,285, and again in Keshavarzi 2013, n=514) with no fertile comparator group, so no association is estimated. The 26,324 figure is the review's total, not this factor's denominator. Testicular atrophy is not examined in relation to steroid exposure. The authors themselves decline to call the listed factors established: 'it is not possible to say with certainty whether these are true' contributing factors, and 'More high-quality research with well-designed studies is needed to reach a strong evidence base on the factors affecting male fertility'. Zero-weight (review). |
| Magnolini R, Bottoni SL, Hammer H, Capraro J, Bruggmann P, Senn O 2025 · Harm Reduction Journal | n-of-1 | supports | low | Case report of a 38-year-old bodybuilder after 8 years of non-medical androgen use: testicular atrophy (12-13 mL on ultrasound), suppressed LH and FSH (both below 0.3 IU/L) and androgen-induced hypogonadism. A 16-week protocol of tapered transdermal testosterone, tamoxifen and hCG failed to restore endogenous testosterone over 6 months, and high-dose hCG stimulation raised total testosterone only 1.2-fold (normal 1.8-2.5x), indicating gonadal failure. The paper does not report a semen analysis; the authors call the hypogonadism idiopathic, with HIV and other causes possible contributors. |
| Skrzypiec-Spring M, Rozmus J, Abu Faraj G, Brawańska-Maśluch K, Kujawa K, Szeląg A 2024 · J Clin Med | observational | supports | low | Aggregation of 397 PUBLISHED CLINICAL CASE REPORTS of AAS complications (PubMed 1973-2022; 88.2% male, median age 30, Q1 23.5 / Q3 39.0); no control group, no denominator of AAS users. ON-OBJECT COUNTS, Table 10 ('The share of individual diseases among all cases'), whose percentages resolve to a denominator of about 227 disease entities, NOT the 397 cases analysed: 'Male infertility 15 6.6', 'Testicular athrophy 3 1.3' and 'Gynecomastia, testicular athrophy 3 1.3' - i.e. 15 infertility and 6 testicular-atrophy cases. Andrological complications overall 32/397 (8.1%, Table 7). Of 22 substance-complication pairs tested, only two survive Benjamini-Hochberg, and one is on-object: endocrine complications in 25.0% of testosterone cases vs 5.9% of non-testosterone cases (p_adj = 0.0000), which the authors read as 'a relationship between the use of T and endocrine complications in the form of gynecomastia and testicular atrophy in men or virilization in women'; for the rest 'there is a probability that the indicated dependencies are false'. Authors against themselves on reliability: 'therefore, the number of cases described does not necessarily reflect the actual number', and the andrological series ends abruptly - 'The number of andrological complications showed a consistent increase until 2012; during the period 2013-2022, no clinical cases of this type were described' - which they attribute to publication practice, not to a fall in harm. Their own conclusion is hedged: 'although we cannot predict the outcomes with certainty, these problems may significantly affect the reproductive health of these people'. CORRECTS the previous extract, which reported only the paper's framing (truncated mid-word at 'infe') without any on-object number, and stated that endocrinological, hepatic, oncological and cardiological events were all increasing over decades - false for three of the four: 'Clinical cases involving cardiac complications were the only ones that showed continuous growth', hepatic 'remained at a similar level', oncological peaked 1993-2002 'after which it decreased', endocrine rose to 2012 'and then decreased in the last decade examined'. Directionally supportive of the claim, at the weakest observational tier: uncontrolled case reports establish that these outcomes occur in AAS users, never their rate or their causal share. |
| Oduwole OO, Vydra N, Wood NEM, Samanta L, Owen L, Keevil B, Donaldson M, Naresh K, Huhtaniemi IT 2014 · FASEB J | animal | supports | moderate | In the hypogonadal LHR-/- mouse the paper reports the OPPOSITE: 'the doses of exogenous T that effectively suppressed gonadotropins and maintained sexual function also activated spermatogenesis' - elongated spermatids rose from none to WT-indistinguishable and 3/8 and 5/8 treated mice sired litters. The authors state their design tests 'T induced stimulation of spermatogenesis in hypogonadism rather than T induced suppression of spermatogenesis in eugonadism', so it cannot vote on steroid-induced atrophy in eugonadal men. Zero-weighted via evidence_role rather than flipped to contradicts, which would overstate it. Re-read #459, full text PMC4376501. |
| Kowalik K, Harasny P, Kaliczyńska L, Reweda-Kwiatkowski K, Starzyński D, Pawlak M, Waloryszak A, Ptak M, Modrzejewski A, Lisman D 2026 · Front Endocrinol (Lausanne) | observational | supports | low | Narrative review (Frontiers in Endocrinology, 2026) of non-medical testosterone and anabolic-androgenic steroid use in young men; it synthesises other studies and reports no new data. It states that suppressed GnRH, LH and FSH give 'reduced intratesticular testosterone concentrations, testicular atrophy, and impaired spermatogenesis', and that after cessation 'complete recovery of reproductive function is not invariably achieved'. The authors call the underlying evidence 'highly heterogeneous and predominantly observational' and largely self-reported. |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.