How Substance Use Patterns Shape Perceived Cardiac Concern Among An International Sample of Men Who Use Anabolic-Androgenic Steroids

Abstract

Background:

Cardiovascular side effects are commonly reported by men who use anabolic-androgenic steroids (AAS), yet little is known about the factors that increase risk. This study examines data from the 2024 Global Drug Survey (GDS2024) to identify whether route of administration (ROA), image-and performance enhancing drug (IPED) polypharmacy and licit substance use are associated with self-reported cardiovascular-related concerns.

Methods:

The dataset comprised 1167 males (≥16 years) who had used AAS within the previous 12 months. Descriptive statistics summarized demographic and substance use characteristics. Chi-square tests explored associations between self-reported cardiovascular side effects and ROA (oral vs injectable), IPED polypharmacy, and licit substance use respectively. Independent predictors were identified through multivariable binary logistic regression. Associations between substances were investigated using pairwise models. Visualizations included heatmaps, bar charts, forest plots, trend stability, and performance comparisons across models.

Results:

Showed that, among AAS consumers, 35.7% reported a self-reported cardiovascular-related concern. Alcohol (79.4%), tobacco (42.0%), and E-cigarette use (40.5%) were prevalent among AAS consumers. Clenbuterol use differed significantly by ROA, reported by 28.4% of oral and 24.3% of injectable AAS consumers. Injectable AAS use was strongly associated with concurrent human hormone growth (HGH) and insulin use. In adjusted models, injectable AAS use showed the strongest association with self-reported cardiovascular-related concerns (OR = 10.57, P < .001), followed by oral AAS use (OR = 1.92, P < .001) and clenbuterol use (OR = 1.44, P = .044). Age, HGH, insulin, alcohol, tobacco, and e-cigarette use were not significantly associated with cardiovascular-related concerns. Pairwise analyses indicated that clenbuterol-containing substance combinations were associated with higher odds of reporting cardiovascular-related concerns, highlighting the relevance of ROA and patterns of concurrent substance use.

Conclusions:

Negative cardiovascular health concerns among AAS consumers are most strongly associated with oral routes and use of substances like clenbuterol.

Keywords

anabolic-androgenic steroid cardiovascular concern polypharmacy co-substance use clenbuterol alcohol tobacco

Highlights

Cardiovascular-related concerns were common among AAS consumers.

Route of administration influenced cardiovascular risk profiles.

Clenbuterol-containing drug combinations showed elevated risk.

Introduction

Use of anabolic-androgenic steroids (AAS) has expanded over recent years from competitive sports to recreational fitness environments across the world.¹ Although AAS were originally developed for clinical applications (eg, treatment of hypogonadism),^2,3 they are now used to improve their physique and performance.⁴ This “off-label” use typically involves supraphysiological dosages of AAS and conjoint use of other image-and performance-enhancing drugs (IPEDs; eg, human growth hormone).^5,6 Prevalence rates worldwide are around 3%, with markedly higher use in males,^7,8 and initiation of use typically occurs before the age of 30.⁹ AAS use has been associated with a wide range of adverse psychosocial concerns, including aggression, mood destabilization, and schizophrenia-related disorders,^10,11 alongside significant physical harms such as endocrine disruption, gynecomastia, and cardiovascular dysfunction.^12,13 These adverse outcomes warrant further investigation.

Cardiovascular complications are among the most frequently reported and clinically important concerns associated with AAS exposure.^14,15 Cardiology-focused studies have demonstrated associations between AAS exposure and left ventricular dysfunction, reduced ejection fraction, impaired systolic function, and myocardial structural changes.^16,17 Other alterations that have been reported include things such as elevated blood pressure, increased heart rate, maladaptive lipid profiles, left ventricular wall thickening, inflammatory activation response, and continuing atrial remodeling.^18,19 These abnormalities contribute to heightened risks of arrhythmias, cardiomyopathy, and sudden cardiac death.^15,20 Long-term AAS exposure is associated with elevated risks of cardiovascular-related risks,²¹ with most documented cases presenting with left ventricular dilation and reduced ejection fraction.¹⁵ Despite growing awareness of physiological risks, little is known about which patterns of AAS use and co-substance involvement place the greatest burden on the cardiovascular-related concerns.

A notable yet currently underexplored aspect of AAS use is route of administration (ROA), most commonly oral or injectable.²² Oral AAS are commonly used during initial exposure because they are convenient and have a rapid onset of action, but they also offer more hepatotoxic effects, profound hormonal changes and an impact on lipid metabolism.^23,24 Severe hepatic complications, such as cholestasis, peliosis hepatis, and hepatocellular carcinoma, along with neuropsychiatric effects like mood instability and libido dysregulation have been reported with oral use.^25-28 In contrast, long entered injectable AAS have greater half-lives and bioavailability and more uniform pharmacokinetics with less hepatic compromise,^23,29 but also result in gonadotropin suppression, hypogonadism, erectile dysfunction, atherogenic lipid profiles, abnormal diastolic blood pressure and an increased risk of cardiovascular disease.^29-33 Though injectable formulations may have a more favorable hepatic profile,^30,31 their long-term use is linked to adverse cardiometabolic and neuroendocrine outcomes.^33-35

People who consume AAS frequently engage in polypharmacy, combining AAS with substances such as HGH, insulin, amphetamines, and clenbuterol to potentiate anabolic effects, accelerate fat loss, or counteract perceived adverse consequences of AAS use.⁶ These combinations substantially elevate cardiovascular risk markers.^36,37 Insulin doping, for example, has been associated with reduced HDL-cholesterol, elevated liver transaminases, and disruptions in fatty acid metabolism,³⁸ while clenbuterol, despite limited evidence suggesting improved glucose disposal,³⁹ is strongly linked to severe cardiovascular complications,⁴⁰ including supraventricular tachycardia, atrial fibrillation, and potential myocardial injury among athletes and IPEDs consumers.^39,40 HGH use further contributes to cardiovascular vulnerability; although short-term recombinant growth hormone (GH) may transiently improve endothelial markers, its broader cardiovascular effects remain uncertain and potentially harmful.^41,42 Such polypharmacy complicates causal attribution, as co-used substances may mask, mimic, or exacerbate AAS-related cardiovascular risk markers.^14,18,37,43

In addition to IPED polypharmacy, many AAS consumers also use licit substances including alcohol, cigarettes, and tobacco, which independently increase cardiovascular risk.⁴⁴ Even minimal tobacco and cigarette exposure are associated with markedly elevated cardiovascular risk.^45-47 E-cigarette use impairs endothelial function and elevates long-term coronary risk,⁴⁷ while tobacco consumption broadly contributes to hypertension and adverse lipid profiles.⁴⁸ Alcohol further influences cardiovascular risk in a dose- and pattern-dependent approaches, with heavy use contributing to cardiovascular failure risk despite some favorable changes seen with moderate intake.^49,50 The evidence indicates that the use of AAS, polypharmacy, and licit substance usage may exert a significant cardiovascular risk,^{14,15,36,51-53} but there is a significant lack of concrete evidence on how polypharmacy among AAS consumers affects cardiovascular health, highlighting a crucial gap in current scientific research.⁵⁴

Despite substantial evidence for individual substance risk, limited research has comprehensively evaluated how AAS-based co-substance (polypharmacy, licit) use affects IPED consumers’ cardiovascular trajectories.⁵⁵ Most existing studies rely on small clinical samples, focus on single compounds, or do not account for the complexity of polypharmacy and co-substance use,¹⁹ and no large, international dataset has systematically examined how demographic characteristics and patterns of IPED polypharmacy, ROA, and licit substance use patterns are associated with self-reported cardiovascular-related concerns. Given these gaps, this research aimed to (i) describe consumer characteristics and patterns of ROA and co-substance use; (ii) examine how substance use patterns, ROA, and age are associated with self-reported negative cardiovascular-related concerns, including the relative contribution of individual substances and their combinations.

Methods

Sample and Procedure

The study employed a cross-sectional design, using data from the 2024 Global Drug Survey (GDS2024), an international, large, online survey aimed at describing global patterns of drug use. The self-report survey is anonymous and provided on-line and distributed throughout the world across a variety of recruitment efforts including mainstream media (eg, The Sydney Morning Herald), social media (eg, X [formerly Twitter], YouTube), as well as partnerships with online content creators and influencers (eg, Vigorous Steve) who promoted participation in the survey.

Comprehensive sociodemographic data, substance use behaviors and related health information were obtained through the questionnaire. Anatomical specifics differentiated between AAS and corticosteroids to reduce the risk of misclassification. The full questionnaire collects detailed information on substance use behaviors, patterns and consequences.

The GDS2024 was available from January 10 to April 30, 2024, and completion time varied from about 15 to 60 minutes depending on participants’ substance use profiles. Documentation of survey development, recruitment process and validation procedures had been described in many previous publications.^56,57 Ethical approval was obtained from The University of Queensland’s Human Research Ethics Committee (2023/HE001927). Individuals were eligible to participate if they were male, aged 16 years or older and had used AAS within the past 12 months. The survey was completely anonymous; no names, IP addresses, dates of birth, or other identifying details were collected. Reporting follows the STROBE guidelines for observational research.

Measures

The present study utilized demographic information related to age, gender, country of residence, ethnicity, and highest education. Further, questions were asked about participants’ AAS usage, including whether they have used AAS within the past 12 months, and ROA (eg, injectable, oral). In addition, the questionnaire assessed IPED polypharmacy, and licit polysubstance use, capturing participants’ use of alcohol, tobacco, E-cigarette, HGH, insulin, and clenbuterol.

The GDS employs a structured, self-completed questionnaire, and its validity and reliability for assessing substance use behaviors and related concerns have been demonstrated in multiple peer-reviewed publications.^56-58 The survey also collected data on self-reported physical concerns including perceived cardiovascular-related effects. Participants were asked about perceived cardiovascular-related concerns of AAS use, including items such as “physical concern: negative impact on the heart.” These were presented as binary (Yes/No) response items and have been presented comprehensively previously.^59,60

Analysis

All analyses were conducted using RStudio (v2024.12.0), and Python (v3.12) was used for data visualization. Demographic factors and co-substance use patterns were summarized as means, standard deviations, or percentage using descriptive statistics. Chi-square (χ²) analyses were used to compare the differences across substances for substance-use patterns, and between involving co-substance comorbidities prevalence, enabling evaluation of statistically significant relationship between categorical data. Factors associated with self-reported negative cardiovascular-related concerns were explored using binary multivariable logistic regression. The first model was the null model (M₀), which included only the intercept. In the complete adjusted model (M₁), substance-use variables were presented simultaneously: oral AAS, injectable AAS, alcohol, tobacco, E-cigarette, HGH insulin, and clenbuterol, together with age. The fit of the model was assessed by AIC, BIC, and pseudo-R². Odds ratios (OR) with 95% confidence intervals were calculated to determine the association of each factor with self-reported cardiovascular-related concerns.

To further explore associations between combinations of substances, pairwise logistic regression models were conducted. Each model evaluated the association between a specific substance and self-reported cardiovascular-related concerns while adjusting for concurrent use of other substances. This approach enabled the determination of a dominant effect-producing substance and revealed compound amplification of cardiovascular-related concerns with implications for common trends in IPED polypharmacy and licit polysubstance use. The model’s performance will be assessed and verified with the assessment metrics provided in the comparative performance plot, which includes area under the curve (AUC), accuracy, sensitivity, specificity, precision, F1-score, balanced accuracy, Kappa, positive predictive value (PPV), negative predictive value (NPV), and Brier score.

To visualize model results, we generated forest plots for odds ratios and confidence intervals, and heatmaps and bar charts to show the prevalence of cardiovascular-related concerns by age and substance. A trend stability plot assessed the robustness of predictor effects across model specifications. Model performance was compared using bar charts for 5 model variants.

Results

Descriptive Statistics

Demographic Characteristics

The study included 10 120 participants (M_age = 32.90 years, SD = 11.30) from 119 countries. From this cohort, our final analytic sample comprised 1167 male AAS/IPED consumers who had used AAS within the past 12 months, representing respondents from 69 countries (M_age = 31.44 years, SD = 9.89). The majority of participants self-identified as White (n = 918, 78.4%), followed by Hispanic/Latinx (n = 82, 7%). Education levels were generally high, and over 22.1% reported college certificate/diploma qualifications (Table 1).

Table 1.

Demographic Characteristics of Male AAS Consumers (N = 1167).

Characteristic	Categories	N	%
Ethnicity	White	916	78.4
	Black/African American	4	0.3
	Asian	39	3.3
	Hispanic/Latinx	82	7
	Aboriginal/Torres Strait Islander/Māori	4	0.3
	Native American	6	0.5
	Mixed	67	5.7
	Other	28	2.4
	Missing	22	1.9
Education	No formal schooling	7	0.6
	Primary school	63	5.4
	Lower secondary school/School or Intermediate certificate	82	7
	Technical or trade certificate	136	11.6
	Higher secondary school/HSC/VCE/Leaving certificate	208	17.8
	College certificate/Diploma	258	22.1
	Undergraduate degree	243	20.8
	Postgraduate degree	167	14.3
	Missing	4	0.3

Abbreviation: AAS, anabolic-androgenic steroids.

IPED Polypharmacy and Licit Polysubstance Use Patterns

To explore the co-occurrence of AAS use with co-substance use and self-reported cardiovascular-related concerns, associations between ROA and different forms of co-substance use were evaluated (see Table 2). Substance use was common in the sample, and alcohol, tobacco, and E-cigarette consumption were commonly reported. Significant associations were found between both oral (Alcohol: χ²(1) = 5.88, P = .015; Tobacco: χ²(1) = 4.70, P = .032) and injectable AAS consumers (Alcohol: χ²(1) = 5.48, P = .019; Tobacco: χ²(1) = 14.85, P < .001) in relation to these substances. Although E-cigarette use was significantly associated only with injectable use (χ²(1) = 9.31, P = .002), it was still commonly reported by oral consumers as well.

Table 2.

Prevalence of Co-Substance Use and Its Association with ROA among Past 12-Month AAS Consumers (N = 1167).

Variable	Level	Oral AAS (n = 546)n (%)	χ²(1)	P-value	Injectable AAS (n = 787)n (%)	χ²(1)	P-value
Alcohol	Yes	417 (76.4)	5.88	.015	610 (77.5)	5.48	.019
Alcohol	No	129 (23.6)	5.88	.015	177 (22.5)	5.48	.019
Tobacco	Yes	211 (38.6)	4.7	.032	300 (38.1)	14.85	<.001
Tobacco	No	335 (61.4)	4.7	.032	487 (61.9)	14.85	<.001
E-cigarette	Yes	212 (38.8)	1.23	.266	295 (37.5)	9.31	.002
E-cigarette	No	334 (61.2)	1.23	.266	492 (62.5)	9.31	.002
HGH	Yes	187 (34.2)	121.31	<.001	560 (71.2)	106.28	<.001
HGH	No	359 (65.8)	121.31	<.001	227 (28.8)	106.28	<.001
Insulin	Yes	63 (11.5)	27.5	<.001	704 (89.5)	38.1	<.001
Insulin	No	483 (88.5)	27.5	<.001	83 (10.5)	38.1	<.001
Clenbuterol	Yes	155 (28.4)	87.99	<.001	191 (24.3)	81.79	<.001
Clenbuterol	No	391 (71.6)	87.99	<.001	596 (75.7)	81.79	<.001
Cardiovascular-related concerns	Yes	290 (53.1)	134.97	<.001	394 (50.1)	216.77	<.001
Cardiovascular-related concerns	No	256 (46.9)	134.97	<.001	393 (49.9)	216.77	<.001

Abbreviations: AAS, anabolic-androgenic steroids; ROA, route of administration; HGH, human growth hormone.

Percentages are calculated within each ROA group: oral AAS consumers, n = 546; injectable AAS consumers, n = 787.

Use of HGH and insulin was substantially higher among injectable AAS consumers, with large and highly significant associations observed (HGH: χ²(1) = 106.28, P < .001; insulin: χ²(1) = 38.10, P < .001). Clenbuterol use was also significantly associated with both oral AAS use (χ²(1) = 87.99, P < .001) and injectable AAS use (χ²(1) = 81.79, P < .001) and was reported by 28.4% of oral AAS consumers and 24.3% of injectable AAS consumers.

Finally, there was a strong and statistically significant association between self-reported cardiovascular-related concerns with both oral (χ²(1) = 134.97, P < .001) and injectable (χ²(1) = 216.77, P < .001) AAS use.

IPED Polypharmacy, Licit Polysubstance Use, Age Patterns in Cardiovascular-Related Concerns

To characterize differences in cardiovascular-related concerns by substance type and age group, a multidimensional visualization integrating prevalence data, age-stratified heatmaps, and group-level patterns were developed (see Figure 1). There was considerable variation in cardiovascular-related concerns depending on the type of exposure. Pharmacological substances, such as injectable AAS, oral AAS, HGH, insulin, and clenbuterol, showed significantly higher reported rates of negative cardiovascular-related concerns, all consistently exceeding 50%. By contrast, licit substances (ie, E-cigarette, tobacco, and alcohol) were associated with much lower negative cardiovascular-related concerns, generally between 30% and 38%. This reflects a clear absolute risk gap, with polypharmacy showing negative cardiovascular-related concerns that are roughly 15% to 25% higher than licit substances.

Figure 1.

Prevalence of negative cardiovascular-related concerns across substance use, and age groups.

The age-stratified heatmap further uncovers key statistical patterns. Young adults (ages 16-24 years) exhibited some of the highest levels of self-reported cardiovascular-related concerns, particularly for injectable AAS, insulin, and clenbuterol, with prevalence exceeding 60% and approaching 70% to 75% for insulin and clenbuterol. The most striking example was insulin use in the 16 to 19 age group, where cardiovascular-related concerns peaked at 75%, a significant outlier compared to other age groups. In contrast, people aged 35 to 39 consistently had the lowest negative cardiovascular-related concerns, forming a distinct low point on the age-distribution curve. However, E-cigarette use, and to a lesser extent clenbuterol, showed limited age-related variability in the prevalence of self-reported negative cardiac impact (approximately 28%-43%), whereas tobacco and alcohol use exhibited greater age-dependent variation, with prevalence decreasing to around 23% to 29% in mid-adulthood and increasing again to approximately 37% to 40% in older age groups.

These patterns are further confirmed by the age-group summary plot, which shows that the youngest (16-19 years) and oldest (50+ years) consumers have the highest overall rates of negative cardiovascular-related concerns, at 46.1% and 47.1%, respectively. This supports the presence of a meaningful bimodal trend, with middle-aged adults (especially those aged 35-39, at 32%) forming a contrasting low point.

IPED Polypharmacy, Licit Polysubstance Use, Qualification Patterns in Cardiovascular-Related Concerns

A multidimensional visualization characterizing cross-classification of cardiovascular-related concerns by substance type and qualification was developed (see Figure 2), integrating prevalence data, qualification-stratified heatmaps, and group-level patterns. Substance exposure type was associated with very different cardiovascular concerns. Reported rates of cardiovascular-related concerns associated with the use of pharmacological substances (ie, injectable AAS, oral AAS, HGH, insulin and clenbuterol) were consistently high overall and generally over 50% per qualification level. On the contrary, the usage of licit substances (ie, alcohol, tobacco and E-cigarettes) was associated with lower prevalence compared to illicit substances through various studies ranging from an estimate of just over 30% to around 48%. The implication is a large absolute difference in cardiovascular concerns visibility or experience between IPED use and licit substance use (older pharmacological substances showed around 10%-25% elevated rates overall).

Figure 2.

Prevalence of negative cardiovascular-related concerns across substance use and qualification.

The qualification-stratified heatmap further illustrates clear qualification levels in cardiovascular-related concerns. No formal schooling level also had among the highest self-reported prevalence (oral AAS, alcohol and tobacco reached values of 100%), however, these represent subgroups with few individuals. Usage of insulin and clenbuterol was especially pronounced as a concern in subjects with higher degrees, with this prevalence reaching around 76.9% for postgraduate individuals and 71.0% to 71.1% in undergraduates. In contrast, AAS consumers with mid-level education (eg, technical/trade and college/diploma) generally exhibited more moderate and stable prevalence for most substances, indicating a potential attenuation of concerns in these groups. Licit substances were less variable across qualification levels, though alcohol and tobacco did show some variation with lower prevalence in higher secondary and college/diploma groups.

Overall, further patterns were supported by a summary plot of self-reported cardiovascular concerns by qualification level; those without formal schooling had the highest overall prevalence of self-reported cardiovascular morbidity (74.8%) and postgraduate students (53.4%) and undergraduates (51.7%). In contrast, the prevalence among college/diploma educated was very low (39.4%) and represented a clear minimum in the distribution. This indicates a non-linear association between qualification levels and cardiovascular-related concerns; with both the lowest as well as the highest levels associated with increased prevalence, while intermediate levels seem to convey relatively decreased concerns.

Associations of IPED Polypharmacy and Licit Substances on Cardiovascular-Related Concerns

A multivariable binary logistic regression analysis was conducted to examine the independent associations between IPED use, licit substances, and self-reported cardiovascular-related concerns among AAS consumers. The full model (M₁), which included age, ROA, IPED polypharmacy, and licit substance use, significantly improved model fit compared to the null model (Δχ² = 289.34, P < .001), explaining a moderate proportion of variance (McFadden R² = 0.190; Nagelkerke R² = 0.301).

After adjustment for all covariates, injectable AAS use emerged as the strongest predictor of cardiovascular-related concerns (OR = 10.57, 95% CI: 6.56-17.02, P < .001), indicating more than a tenfold increase in odds. Oral AAS use was also significantly associated with increased odds (OR = 1.92, 95% CI: 1.47-2.59, P < .001), while clenbuterol use showed a smaller but statistically significant association (OR = 1.44, 95% CI: 1.01-2.05, P = .044).

In contrast, age and other substances (HGH, insulin, alcohol, tobacco, and e-cigarette use) were not significantly associated with cardiovascular-related concerns after adjustment (all P > .05) (Figure 3a; Table 3).

Figure 3.

Odds ratios for negative cardiovascular-related concerns by substance and substance pairings: (a) single substance effects; (b) substance combinations; (c) age effect in combination models; (d) oral AAS; and (e) injectable AAS.

Table 3.

Multivariable Binary Logistic Regression Predicting Cardiovascular-Related Concerns Among AAS Consumers.

Model	Estimate	SD	OR	z	Wald statistic	P
M₀ (Intercept)	−0.587	0.061	0.556	−9.609	92.339	<.001
M₁ (Intercept)	−2.268	0.353	0.104	−6.420	41.219	<.001
Age	−0.010	0.007	0.990	−1.400	1.960	.161
Injectable AAS	2.358	0.243	10.571	9.713	94.344	<.001
Oral AAS	0.652	0.153	1.920	4.257	18.118	<.001
HGH	−0.250	0.182	0.778	−1.374	1.888	.169
Insulin	0.155	0.270	1.168	0.575	0.331	.565
Clenbuterol	0.363	0.180	1.437	2.017	4.070	.044
Alcohol	−0.155	0.173	0.856	−0.901	0.811	.368
Tobacco	−0.150	0.161	0.860	−0.934	0.873	.350
E-cigarette	−0.103	0.162	0.902	−0.636	0.404	.525

Abbreviations: OR, odds ratio; SD, standard deviation.

Cardiovascular-related concerns coded as 1, OR <1 indicates reduced likelihood of reporting negative cardiovascular-related concerns and does not necessarily indicate a true reduction in risk. All Wald tests conducted with df = 1.

To evaluate how different substance combinations relate to self-reported cardiovascular-related concerns, pairwise logistic regression models were used, assessing the effect of each substance while controlling for the concurrent use of another substance, age, and ROA.

In the pairwise combination models (see Figure 3b and Table S2), several two-substance combinations were associated with higher odds of reporting cardiovascular-related concerns. The highest odds ratios were observed for combinations involving clenbuterol and/or insulin, particularly clenbuterol + insulin (OR = 2.66, 95% CI = 1.53-4.71, P = .0006) and HGH + insulin (OR = 2.66, 95% CI = 1.76-4.69, P < .0001). Other clenbuterol-based pairs, such as clenbuterol + tobacco (OR = 2.34, 95% CI = 1.48-3.72, P = .0003), clenbuterol + E-cigarette (OR = 2.10, 95% CI = 1.32-3.37, P = .0018), and clenbuterol + alcohol (OR = 2.00, 95% CI = 1.42-2.84, P = .0001), also showed significantly elevated concerns, as did combinations such as insulin + alcohol (OR = 2.08, 95% CI = 1.22-3.56, P = .0074) and HGH + alcohol (OR = 1.65, 95% CI = 1.19-2.29, P = .0027). While some pairs of alcohol and tobacco substances (eg, tobacco + E-cigarette, alcohol + E-cigarette, or alcohol + tobacco) had odds ratios below 1.0 with P < .05, indicating a lower relative odds of reporting cardiovascular-related concerns within specific pairings. Age effects within the combination models were essentially null across all pairs (ORs = 1.00, 95% CIs = 0.99-1.01, all P > .20; see Figure 3c, Table S3), indicating that the excess concern associated with IPED polypharmacy patterns is largely age independent. When oral and injectable AAS effects were decomposed within the combination models (see Figure 3d and e, and Tables S4 and S5), oral AAS showed a stable, moderate elevation in concern across all combinations (ORs = 1.9-2.1, 95% CIs = 1.45-2.7, P < .00001), whereas injectable AAS consistently produced very large odds ratios of approximately 10 to 11 (95% CIs = 6.7-17.3, P < 10⁻²²) irrespective of co-substance.

Model Performance and Predictive Validity

To evaluate the validity and predictive balance of the combination models illustrated in the forest plots, model performance was assessed across 11 evaluation metrics (see Figure 4 and Table S6). The base combination model demonstrated very high sensitivity (0.93), indicating strong ability to identify individuals reporting negative cardiovascular-related concerns; however, this was accompanied by low specificity (0.19), reduced discrimination (AUC = 0.61), and weaker calibration (Brier score = 0.21), reflecting a trade-off toward false-positive classification. Adding age to the combination model did not alter any performance metric, confirming that age contributes minimal predictive value and does not meaningfully modify substance-related cardiac concern.

Figure 4.

Comparison of 5 predictive models across 11 performance metrics.

In contrast, explicit inclusion of ROA substantially improved model performance. The Combo + Oral AAS model showed higher discrimination (AUC = 0.71), improved specificity (0.45), greater balanced accuracy (0.62), and stronger agreement (κ = 0.26), while maintaining acceptable sensitivity (0.80). Similarly, the Combo + Injectable AAS model further improved discrimination (AUC = 0.74), specificity (0.51), balanced accuracy (0.63), and calibration (Brier score = 0.18), approaching the performance of the single-substance model. Collectively, these findings indicate that while unadjusted combination models prioritize sensitivity, ROA, particularly injectable AAS, restores predictive balance and supports the robustness of the observed concern estimates, whereas age provides no meaningful enhancement to model validity.

Trend Stability and Spearman’s Rank Correlation (ρ) Analysis

The stability of trends was analyzed through leave-one-out models with Firth penalized logistic regression to confirm the robustness of multivariable logistic regression results. Spearman’s rank correlation coefficient (ρ) is a non-parametric measure which indicates the consistency between predictor OR patterns in different model settings. Spearman’s ρ assesses the concordance of the rank order and sign of OR estimates when model structure changes, as opposed to focusing on numeric equality. Low ρ values show high sensitivity of the model and higher values correspond to consistent correlations (without any covariates) (see Figure 5 and Table S7).

Figure 5.

Trend stability of predictors across sensitivity models.

Trend stability for the use of injectable AAS was quasi-perfect in sensitivity analysis, exhibiting higher ORs under all 3 model options. The pattern has a high ρ, indicative of its strong cardiac association independent of the model. Maintained consistency of this effect suggests that it can be considered a strong and independent risk factor. The crest of the OR magnitude curve for oral AAS use also varied greatly when injectable AAS usage was omitted from the model; however, direction parameters were unchanged. If the ρ is high but not 1 (ie, ρ < 0.95), there could be some confounding or overlapping exposure patterns reflected by a lower but non-perfect ρ, not by relationship instability.

Clenbuterol positively correlated to all models with a high ρ value and a narrow range of OR values. Insulin was more variable, with ORs close to one in some of the models. The pattern shows moderate ρ and a context-dependent relationship. The OR estimates for HGH alcohol, tobacco, and E-cigarette use were around or <1 in all moderately heterogeneous models. Such significant but weak connections are indicated by the large ρ values. Age was not associated with OR in any of the models, suggesting very good temporal trend stability (although not a clinically relevant one). The Spearman’s ρ-based sensitivity analysis of trend stability shows that the main study findings, particularly the significant cross-link between injectable AAS consumers and cardiac concerns, are insensitive to model choice and sparse data bias.

Discussion

This study found that negative cardiovascular-related concerns among a large international sample of men who use AAS is driven primarily by injectable AAS, oral AAS, and clenbuterol, which were the only substances independently associated with negative cardiovascular-related concerns. Pairwise analysis showed that cardiovascular-related concerns in AAS consumers is driven predominantly by injectable AAS, with oral AAS and clenbuterol acting as additional independent contributors. Furthermore, certain IPED polypharmacy patterns, particularly combinations that include clenbuterol and/or insulin together with AAS, substantially amplify the likelihood of negative cardiovascular-related concerns. These findings indicate that negative cardiovascular-related concerns are shaped not only by individual substances but also by synergistic interactions between them.

Considering the prevalence of co-substance use, evidence showed consumers often combine multiple substances such as alcohol, tobacco, HGH, insulin, and clenbuterol,^6,61 to enhance effects, combat side effects, or for recreational purposes.⁶ This aligns with extensive evidence showing that AAS use rarely occurs in isolation; rather, individuals commonly engage in complex IPED polypharmacy with licit substances use.^62-64 This reflects a distinct enhancement typology rather than recreational substance use alone. Within this context, cardiovascular concerns appear to be driven primarily by pharmacological burden and biological interaction, rather than by lifestyle behaviors.

Moreover, IPED polypharmacy frequently occurs alongside AAS use and is associated with greater self-reported cardiovascular-related concerns. Insulin use alone is associated with a 52% increase in cardiovascular risk,^65,66 while HGH shows mixed profiles, improvements in lipid and blood pressure markers have been reported,⁶⁷ yet without consistent reductions in cardiovascular mortality.⁶⁸ Clenbuterol stands out as a particularly hazardous component of enhancement-oriented polypharmacy. Commonly combined with AAS during cutting phases to accelerate fat loss and improve muscular definition,⁶ it carries well-documented cardiotoxic risks, including tachyarrhythmias and acute cardiovascular instability.^40,59,69 Additional reports describe structural myocardial injury, supraventricular tachycardia, hypokalemia, and acute cardiac-electrolyte disturbances.^70-72 These findings imply that AAS and frequently co-administered substances like insulin and clenbuterol exacerbate cardiovascular-related concerns among AAS consumers. The combined, or synergistic toxicities of these substances may underlie consumers’ polypharmacy-related additional cardiovascular-related concerns, underscoring the importance of examining combination versus single compound approaches to exposure analysis.

In contrast to enhancement-oriented polypharmacy, the role of licit substances such as alcohol, E-cigarettes, and tobacco warrants separate consideration. Although licit substance use was prevalent among AAS consumers in this sample, these substances were associated with increased cardiovascular-related concerns⁷³; however, their independent effects were attenuated after accounting for AAS/IPED use. This suggests that licit substances may contribute to cardiovascular-related concerns,^47,74 their influence is likely secondary to the more salient effects of AAS use and enhancement-oriented polypharmacy. This finding should not be interpreted as indicating cardiovascular safety; but rather as reflecting differences in how risk is perceived and contextualized within this population. AAS consumers who engage in licit substance use may represent more recreational or socially oriented typologies, where substance use is associated with social interaction, stress relief, or lifestyle behaviors rather than targeted physiological enhancement.⁷⁵ As such, cardiovascular concerns in this group may be less directly attributed to substance use and more diffuse in nature.^76,77

Perception-related factors may further explain these findings. Widespread public health messaging and media coverage regarding the harms of alcohol and tobacco may increase general awareness of cardiovascular risk,^78,79 potentially elevating concern among AAS consumers. Conversely, familiarity with these substances and their normalization in social settings may reduce the perceived immediacy or severity of cardiovascular consequences compared to AAS or polypharmacy use.⁶ In this context, cardiovascular-related concerns likely reflect a balance between awareness, personal experience, and perceived controllability of risk.

Regarding AAS use, the observed association with increased cardiovascular-related concerns may be interpreted considering existing knowledge regarding their potential adverse effects.^15,53 While the present study does not assess clinical cardiovascular outcomes, prior evidence linking AAS exposure to adverse cardiac effects may contribute to heightened risk perception among consumers.^21,51,52,80 Exposure to such information may shape consumers’ awareness and contribute to increased reporting of cardiovascular concerns. Additionally, personal experiences such as palpitations, fluctuations in blood pressure, or reduced exercise tolerance may further reinforce these perceptions.^14,15,81

With regard to IPED polypharmacy and cardiac-related concerns, there remains a substantial gap in the literature on how combinations of AAS and co-administered substances shape cardiac concerns among AAS consumers. The concurrent use of multiple substances⁸² may increase consumers’ awareness of potential additive or synergistic effects, particularly within communities where information sharing and experiential knowledge are common.^13,14,83 In the absence of clear clinical evidence, this awareness may amplify concern, especially among individuals who have previously experienced adverse symptoms or are exposed to discussions emphasizing the risks of combining substances.

Implications

This study carries important clinical, public health, and policy implications. The findings indicate that AAS use is associated with greater self-reported cardiovascular-related concerns, highlighting the importance of routine screening for AAS use in clinical settings and careful consideration of cardiovascular health. Prevention and harm reduction must work in tandem. While cessation should remain a prevention goal, abrupt discontinuation is rarely feasible for consumers; structured cessation and recovery approaches, including appropriate post-cycle therapy, are necessary to reduce withdrawal-related and endocrine risks.⁸⁴

From a harm reduction perspective, accessible and routine cardiovascular monitoring is essential. Simple measures such as regular blood pressure and heart rate checks provide quick, low-cost indicators of strain, while echocardiography offers a more precise tool to detect left ventricular hypertrophy and other early signs of cardiac remodeling. Importantly, presenting consumers with objective clinical data can prompt meaningful behavior change, consistent with evidence from other biometric-informed interventions.^{85, 86}

Intervention approaches may differ by level of perceived concern: individuals with low concern may require targeted education to improve concern awareness, whereas those with higher concern may benefit from structured clinical engagement, including monitoring and reassurance.

At a population level, these results highlight the need for targeted education and health promotion that clearly communicates the potential cardiac concerns of AAS and polypharmacy. Improving access to knowledgeable healthcare providers, reducing stigma, and supporting early diagnostic pathways can help prevent serious cardiovascular events. Further research using standardized, reliable measures of AAS and IPED polypharmacy and polysubstance use, alongside prospective monitoring of cardiac concerns, is needed to strengthen the evidence base and inform effective interventions.

Limitations

The present study is not without its limitations, which should be considered when interpreting these findings. Recall bias, underreporting and response error may have arisen because of relying on self-reports. Additionally, self-reported cardiovascular-related concerns may be influenced by cognitive bias (eg, the Dunning-Kruger effect), leading to underestimation due to poor insight or overestimation with greater awareness. Thus, reported concerns may reflect differences in perception and health literacy rather than true cardiovascular status.⁸⁷ Similarly, while there were observed correlations between substance use and cardiac measures, the cross-sectional nature of this study precludes inferences about temporal ordering or cause-effect relationships. In addition, the outcome measure does not constitute a validated clinical assessment of cardiac health, is based on subjective perceptions, and does not identify specific cardiac conditions; therefore, validation by healthcare professionals is recommended. Information on dosage, duration and patterns of use, stacking practices, and prescription status was not available; these factors are likely to moderate risk and warrant dedicated modeling in future research. Despite the international scope of this study, country-level regulatory, healthcare, and cultural factors that may influence substance-use patterns and risk perception were not fully addressed. Finally, the recruitment process may have over-represented digitally connected individuals while under-sampling those who are more marginalized or disengaged from health services.

Conclusion

This international study highlights that negative cardiovascular-related concerns among AAS consumers is mainly influenced by ROA as well as IPED polypharmacy (eg, clenbuterol), rather than by licit substances (eg, alcohol). Injectable AAS, oral AAS, and clenbuterol each independently contribute to negative cardiovascular-related concerns, with injectable AAS showing by far the most pronounced effect. Clenbuterol has a substantial independent effect and it increases concern when combined with other substances. These results offer that cardiovascular-related concerns are associated with not only isolated substances (particularly injectable AAS), but also their interactions in IPED polypharmacy, ROA, and licit substance. These findings highlight the necessity of rising awareness regarding cardiac concerns linked to injectable AAS polypharmacy, especially concerning high-risk substances like clenbuterol, to help mitigate preventable cardiovascular-related concerns among AAS consumers.

Supplemental Material

sj-docx-1-saj-10.1177_29767342261460762 – Supplemental material for How Substance Use Patterns Shape Perceived Cardiac Concern Among An International Sample of Men Who Use Anabolic-Androgenic Steroids

Supplemental material, sj-docx-1-saj-10.1177_29767342261460762 for How Substance Use Patterns Shape Perceived Cardiac Concern Among An International Sample of Men Who Use Anabolic-Androgenic Steroids by Bahareh Ahmadinejad, Benjamin Bonenti, Kimberley Akrigg, Jason Ferris, Adam Winstock, Lauren Ball, Jason M. Nagata, Kyle T. Ganson and Timothy Piatkowski in Substance Use & Addiction Journal

Footnotes

Acknowledgements

We would like to thank all the participants of the Global Drug Survey who donated their time to completing the survey. We would also like to thank A/Prof Monica Barratt and Dr Cheneal Puljevic, for their intellectual contribution to the conceptualization and methodology of the GDS.

ORCID iDs

Bahareh Ahmadinejad

Benjamin Bonenti

Timothy Piatkowski

Ethical Considerations

Ethical approval was obtained from The University of Queensland’s Human Research Ethics Committee (2023/HE001927).

Consent to Participate

All participants provided informed consent via the survey checkbox.

Author Contributions

BA: Conceptualization, Formal analysis, Data curation, Visualization, Writing – original draft. BB: Writing – review & editing. KA: Writing – review & editing. JF: Methodology, Writing – review & editing. AW: Methodology, Writing – review & editing. LB: Writing – review & editing. JMN: Writing – review & editing. KTG: Writing – review & editing. TP: Conceptualization, Methodology, Supervision, Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: TP is supported by a National Health and Medical Research Council Investigator Grant (2041822).

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Adam Winstock is the founder and owner of Global Drug Survey (GDS) Ltd, an independent data exchange hub. Jason Ferris is a member of the GDS Core Research Team.

Data Availability Statement

Data available from the corresponding author on reasonable request.*

Supplemental Material

Supplemental material for this article is available online.

References

Abrahin

OSC

de Sousa

Santos

. Prevalence of the use of anabolic-androgenic steroids in Brazil: a systematic review. Subst Use Misuse. 2014;49(9):1156-1162. doi:10.3109/10826084.2014.903750

Urban

Bodenburg

Gilkison

, et al. Testosterone administration to elderly men increases skeletal muscle strength and protein synthesis. Am J Physiol Endocrinol Metab. 1995;269(5):E820-E826. doi:10.1152/ajpendo.1995.269.5.E820

Spungen

Grimm

Strakhan

Pizzolato

Bauman

. Treatment with an anabolic agent is associated with improvement in respiratory function in persons with tetraplegia: a pilot study. Mt Sinai J Med. 1999;66(3):201-205.

Anawalt

. Detection of anabolic androgenic steroid use by elite athletes and by members of the general public. Mol Cell Endocrinol. 2018;464:21-27. doi:10.1016/j.mce.2017.09.027

Magnolini

Falcato

Cremonesi

Schori

Bruggmann

. Fake anabolic androgenic steroids on the black market-a systematic review and meta-analysis on qualitative and quantitative analytical results found within the literature. BMC Public Health. 2022;22(1):1371.

Sagoe

McVeigh

Bjørnebekk

Essilfie

M-S

Andreassen

Pallesen

. Polypharmacy among anabolic-androgenic steroid users: a descriptive metasynthesis. Subst Abuse Treat Prev Policy. 2015;10(1):12.

Sagoe

Molde

Andreassen

Torsheim

Pallesen

. The global epidemiology of anabolic-androgenic steroid use: a meta-analysis and meta-regression analysis. Ann Epidemiol. 2014;24(5):383-398.

Piatkowski

Whiteside

Robertson

Henning

Lau

Dunn

. What is the prevalence of anabolic-androgenic steroid use among women? A systematic review. Addiction. 2024;119(12):2088-2100. doi:10.1111/add.16643

Sagoe

Andreassen

Pallesen

. The aetiology and trajectory of anabolic-androgenic steroid use initiation: a systematic review and synthesis of qualitative research. Subst Abuse Treat Prev Policy. 2014;9(1):27.

10.

Scarth

Havnes

Jørstad

Bjørnebekk

. Psychological traits associated with anabolic androgenic steroid use and dependence: an exploratory cross-sectional study among female athletes. BMC Womens Health. 2025;25(1):214.

11.

Piacentino

Kotzalidis

Del Casale

, et al. Anabolic-androgenic steroid use and psychopathology in athletes. A systematic review. Curr Neuropharmacol. 2015;13(1):101-121. doi:10.2174/1570159X13666141210222725

12.

D’Andrea

Caso

Salerno

, et al. Left ventricular early myocardial dysfunction after chronic misuse of anabolic androgenic steroids: a Doppler myocardial and strain imaging analysis. Br J Sports Med. 2007;41(3):149-155. https://doi.org/10.1136/bjsm.2006.030171

13.

Smit

Buijs

de Hon

den Heijer

de Ronde

. Positive and negative side effects of androgen abuse. The HAARLEM study: a one-year prospective cohort study in 100 men. Scand J Med Sci Sports. 2021;31(2):427-438. doi:10.1111/sms.13843

14.

Mingxing

Yanfei

. Adverse effects of anabolic androgenic steroid abuse in athletes and physically active individuals: a systematic review and meta-analysis. Subst Use Misuse. 2025;60(6):873-887. https://doi.org/10.1080/10826084.2025.2460986

15.

Alhusban

Alaaraj

Saimeh

, et al. Steroid-induced cardiomyopathy: insights from a systematic literature review and a case report. Clin Case Rep. 2025;13(3):e70171.

16.

Rasmussen

Schou

Madsen

, et al. Cardiac systolic dysfunction in past illicit users of anabolic androgenic steroids. Am Heart J. 2018;203:49-56. doi:10.1016/j.ahj.2018.06.010

17.

Baggish

Weiner

Kanayama

, et al. Long-term anabolic-androgenic steroid use is associated with left ventricular dysfunction. Circ Heart Fail. 2010;3(4):472-476. doi:10.1161/CIRCHEARTFAILURE.109.931063

18.

Horwitz

Andersen

Dalhoff

. Health consequences of androgenic anabolic steroid use. J Intern Med. 2019;285(3):333-340. doi:10.1111/joim.12850

19.

Lundholm

Frisell

Lichtenstein

Långström

. Anabolic androgenic steroids and violent offending: confounding by polysubstance abuse among 10 365 general population men. Addiction. 2015;110(1):100-108. doi:10.1111/add.12715

20.

Angell

Chester

Green

, et al. Anabolic steroid use and longitudinal, radial, and circumferential cardiac motion. Med Sci Sports Exerc. 2012;44(4):583-590.

21.

Buhl

Christensen

Hjortebjerg

, et al. Impact of androgenic anabolic steroids on cardiovascular health in men and women-PART OF THE FITNESS DOPING IN DENMARK (FIDO-DK) STUDY. medRxiv. 2024:2024.11. 18.24317516. doi:10.1101/2024.11.18.24317516

22.

Cox

Piatkowski

Dunn

. “Injecting yourself there is stigma around it; taking a few tablets is not too bad, is it?”: understanding perceptions and preferences of anabolic-androgenic steroid route of administration. Harm Reduct J. 2025;22(1):128.

23.

Van der Vies

. Pharmacokinetics of anabolic steroids. Wien Med Wochenschr. 1993;143(14-15):366-368. doi:10.12775/QS.2025.38.57738

24.

King

Sharp

Vukovich

, et al. Effect of oral androstenedione on serum testosterone and adaptations to resistance training in young men: a randomized controlled trial. JAMA. 1999;281(21):2020-2028.

25.

Lamb

. Anabolic steroids in athletics: how well do they work and how dangerous are they? Am J Sports Med. 1984;12(1):31-38. doi:10.1177/036354658401200105

26.

Halstead

Groah

Libin

Hamm

Priestley

. The effects of an anabolic agent on body composition and pulmonary function in tetraplegia: a pilot study. Spinal Cord. 2010;48(1):55-59.

27.

Lovejoy

Bray

Greeson

, et al. Oral anabolic steroid treatment, but not parenteral androgen treatment, decreases abdominal fat in obese, older men. Int J Obes Relat Metab Disord. 1995;19(9):614-624.

28.

Supasyndh

Satirapoj

Aramwit

, et al. Effect of oral anabolic steroid on muscle strength and muscle growth in hemodialysis patients. Clin J Am Soc Nephrol. 2013;8(2):271-279. doi:10.2215/CJN.00380112

29.

Kuipers

Wijnen

Hartgens

Willems

. Influence of anabolic steroids on body composition, blood pressure, lipid profile and liver functions in body builders. Int J Sports Med. 1991;12(04):413-418.

30.

Mielczarek

Sroczyńska

Sroczyński

Lachowska

Senior

Smandek

. Anabolic androgenic steroids intake and its impact on male reproductive system-systematic review. Quality in Sport. 2025;38:57738. doi:10.12775/QS.2025.38.57738

31.

Thompson

Cullinane

Sady

, et al. Contrasting effects of testosterone and stanozolol on serum lipoprotein levels. JAMA. 1989;261(8):1165-1168.

32.

Hartgens

Rietjens

Keizer

Kuipers

Wolffenbuttel

. Effects of androgenic-anabolic steroids on apolipoproteins and lipoprotein (a). Br J Sports Med. 2004;38(3):253-259.

33.

Griffiths

Murray

Mitchison

Mond

. Anabolic steroids: lots of muscle in the short-term, potentially devastating health consequences in the long-term. Drug Alcohol Rev. 2016;35(4):375-376. doi:10.12968/joan.2015.4.7.336

34.

Farley

Peregoudov

Waites

. Effects of testosterone enanthate in normal men: experience from a multicenter contraceptive efficacy study. Fertil Steril. 1996;65(3):626-636. doi:10.1016/S0015-0282(16)58166-7

35.

Yates

Perry

MacIndoe

Holman

Ellingrod

. Psychosexual effects of three doses of testosterone cycling in normal men. Biol Psychiatry. 1999;45(3):254-260. doi:10.1016/S0006-3223(98)00028-6

36.

Beezer

Al Hatrushi

Husband

Kurdi

Forsyth

. Polypharmacy definition and prevalence in heart failure: a systematic review. Heart Fail Rev. 2022;27(2):465-492.

37.

Martinez-Quintana

Saiz-Udaeta

Marrero-Negrin

Lopez-Mérida

Rodriguez-Gonzalez

Nieto-Lago

. Androgenic anabolic steroid, cocaine and amphetamine abuse and adverse cardiovascular effects. Int J Endocrinol Metab. 2013;11(4):e8755. doi:10.5812/ijem.8755

38.

Di Girolamo

Biasinutto

Mangogna

, et al. Metabolic consequences of anabolic steroids, insulin, and growth hormone abuse in recreational bodybuilders: implications for the World Anti-Doping Agency Passport. Sports Med Open. 2024;10(1):28.

39.

van Beek

Bruls

Vanweert

, et al. Effect of β2-agonist treatment on insulin-stimulated peripheral glucose disposal in healthy men in a randomised placebo-controlled trial. Nat Commun. 2023;14(1):173.

40.

Kumari

Pal

Sahu

Prabhakar

Tewari

. Adverse events of clenbuterol among athletes: a systematic review of case reports and case series. Int J Leg Med. 2023;137(4):1023-1037.

41.

Graham

Baker

Evans

, et al. Evidence for a decrease in cardiovascular risk factors following recombinant growth hormone administration in abstinent anabolic-androgenic steroid users. Growth Horm IGF Res. 2007;17(3):201-209. doi:10.1016/j.ghir.2007.01.010

42.

Graham

Baker

Evans

, et al. Short-term recombinant human growth hormone administration improves respiratory function in abstinent anabolic-androgenic steroid users. Growth Horm IGF Res. 2007;17(4):328-335. doi:10.1016/j.ghir.2007.04.003

43.

Corona

Rastrelli

Marchiani

, et al. Consequences of anabolic-androgenic steroid abuse in males; sexual and reproductive perspective. World J Mens Health. 2021;40(2):165. doi:10.5534/wjmh.210021

44.

Meader

King

Moe-Byrne

, et al. A systematic review on the clustering and co-occurrence of multiple risk behaviours. BMC Public Health. 2016;16(1):657.

45.

Hackshaw

Morris

Boniface

Tang

J-L

Milenković

. Low cigarette consumption and risk of coronary heart disease and stroke: meta-analysis of 141 cohort studies in 55 study reports. BMJ. 2018;360. doi:10.1136/bmj.j5855

46.

Chang

Corey

Rostron

Apelberg

. Systematic review of cigar smoking and all cause and smoking related mortality. BMC Public Health. 2015;15(1):390.

47.

Skotsimara

Antonopoulos

Oikonomou

, et al. Cardiovascular effects of electronic cigarettes: a systematic review and meta-analysis. Eur J Prev Cardiol. 2019;26(11):1219-1228. doi:10.1177/2047487319832975

48.

Shrivastava

Jha

Kumar

. Comparative study of tobacco chewer & smokers as a risk factor for cardiovascular disease. Indian J Appl Res. 2014;4(1):1-6.

49.

Djoussé

Gaziano

. Alcohol consumption and heart failure: a systematic review. Curr Atheroscler Rep. 2008;10(2):117-120.

50.

Brien

Ronksley

Turner

Mukamal

Ghali

. Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies. BMJ. 2011;342:d636. doi:10.1136/bmj.d636

51.

Pope Jr

Wood

Rogol

Nyberg

Bowers

Bhasin

. Adverse health consequences of performance-enhancing drugs: an Endocrine Society scientific statement. Endocr Rev. 2014;35(3):341-375. doi:10.1210/er.2013-1058

52.

Baggish

Weiner

Kanayama

, et al. Cardiovascular toxicity of illicit anabolic-androgenic steroid use. Circulation. 2017;135(21):1991-2002. doi:10.1161/CIRCULATIONAHA.116.026945

53.

Cavalcante

Romariz

Neto

Fante

Júnior

SMP

Bacca

. Anabolic-androgenic steroids on cardiac structure and function in resistance-trained athletes: a systematic review and meta-analysis. Int J Cardiol. 2025:133896. doi:10.1016/j.ijcard.2025.133896

54.

Jodłowski

Dubniewicz

Majewska

, et al. Health implications of anabolic-androgenic steroids: usage, risks, and treatment approaches—the literature review. Qual Sport. 2025;38:57914. doi:10.12775/QS.2025.38.57914

55.

Piatkowski

Robertson

Lamon

Dunn

. Gendered perspectives on women’s anabolic-androgenic steroid (AAS) usage practices. Harm Reduct J. 2023;20(1):56.

56.

Barratt

Ferris

Zahnow

Palamar

Maier

Winstock

. Moving on from representativeness: testing the utility of the Global Drug Survey. Subst Abuse Res Treat. 2017;11:1178221817716391. doi:10.1177/1178221817716391

57.

Winstock

Davies

Ferris

Maier

Barratt

. Using the Global Drug Survey for Harm Reduction. Monitoring Drug Use in the Digital Age: Studies in Web Surveys; 2022.

58.

Winstock

Barratt

Maier

, et al. Global Drug Survey 2019: Key findings report. 2020

59.

Bonenti

Puljević

Steve

, et al. Clenbuterol and the cost of cutting: a brief report comparing self-reported side effects of clenbuterol consumption to anabolic-androgenic steroid compounds. Perform Enhanc Health. 2025;13(4):100356. doi:10.1016/j.peh.2025.100356

60.

Grobler

Puljević

Bonenti

, et al. Associations between anabolic-androgenic steroid testing, healthcare access and undesirable effects among international consumers. Perform Enhanc Health. 2026;14(1):100390. doi:10.1016/j.peh.2025.100390

61.

Zuliani

Bernardini

Catapano

Campana

Cerioli

Spattini

. Effects of anabolic steroids, testosterone, and HGH on blood lipids and echocardiography parameters in body builders. Int J Sports Med. 1989;10(01):62-66.

62.

Barnett

Tenerowicz

Perry

. The Anabolic 500 survey: characteristics of male users versus nonusers of anabolic-androgenic steroids for strength training. Pharmacotherapy. 2011;31(8):757-766. doi:10.1592/phco.31.8.757

63.

Ganson

Testa

Jackson

Nagata

. Associations between concurrent substance use and anabolic-androgenic steroid use among adolescents. Subst Use Misuse. 2022;57(7):1160-1163. doi:10.1080/10826084.2022.2064510

64.

Dodge

Hoagland

. The use of anabolic androgenic steroids and polypharmacy: a review of the literature. Drug Alcohol Depend. 2011;114(2-3):100-109. doi:10.1016/j.drugalcdep.2010.11.011

65.

Khatib

Shao

Shi

Fonseca

. The association between baseline insulin treatment and cardiovascular events: a meta-analysis. J Endocr Soc. 2021;5(2):bvaa193.

66.

Robins

Rubins

Faas

, et al. Insulin resistance and cardiovascular events with low HDL cholesterol: the Veterans Affairs HDL Intervention Trial (VA-HIT). Diabetes Care. 2003;26(5):1513-1517. doi:10.2337/diacare.26.5.1513

67.

Maison

Griffin

Nicoue-Beglah

Haddad

Balkau

Chanson

. Impact of growth hormone (GH) treatment on cardiovascular risk factors in GH-deficient adults: a metaanalysis of blinded, randomized, placebo-controlled trials. J Clin Endocr Metab. 2004;89(5):2192-2199.

68.

Deodati

Ferroli

Cianfarani

. Association between growth hormone therapy and mortality, cancer and cardiovascular risk: systematic review and meta-analysis. Growth Horm IGF Res. 2014;24(4):105-111. doi:10.1016/j.ghir.2014.02.001

69.

Jessen

Solheim

Jacobson

, et al. Beta2-adrenergic agonist clenbuterol increases energy expenditure and fat oxidation, and induces mTOR phosphorylation in skeletal muscle of young healthy men. Drug Test Anal. 2020;12(5):610-618. doi:10.1002/dta.2755

70.

Daubert

Mabasa

Leung

Aaron

. Acute clenbuterol overdose resulting in supraventricular tachycardia and atrial fibrillation. J Med Toxicol. 2007;3(2):56-60. doi:10.1007/BF03160909

71.

Christopoulos

DeSimone

Anavekar

, editors. 30-Year-Old Man With Outside-of-Hospital Cardiac Arrest. Mayo Clinic Proceedings; 2018: Elsevier Limited.

72.

Kamalakkannan

Petrilli

George

, et al. Clenbuterol increases lean muscle mass but not endurance in patients with chronic heart failure. J Heart Lung Transplant. 2008;27(4):457-461. doi:10.1016/j.healun.2008.01.013

73.

Magna

Polisena

, et al. The hidden dangers: e-cigarettes, heated tobacco, and their impact on oxidative stress and atherosclerosis—a systematic review and narrative synthesis of the evidence. Antioxidants. 2024;13(11):1395.

74.

Corrao

Rubbiati

Bagnardi

Zambon

Poikolainen

. Alcohol and coronary heart disease: a meta-analysis. Addiction. 2000;95(10):1505-1523. doi:10.1046/j.1360-0443.2000.951015056.x

75.

Christiansen

Vinther

Liokaftos

. Outline of a typology of men’s use of anabolic androgenic steroids in fitness and strength training environments. Drugs Educ Prev Policy. 2017;24(3):295-305. doi:10.1080/09687637.2016.1231173

76.

Lassi

Imam

Dean

Bhutta

. Preconception care: caffeine, smoking, alcohol, drugs and other environmental chemical/radiation exposure. Reprod Health. 2014;11(Suppl 3):S6.

77.

Cho

Thrasher

Swayampakala

, et al. Does adding information on toxic constituents to cigarette pack warnings increase smokers’ perceptions about the health risks of smoking? A longitudinal study in Australia, Canada, Mexico, and the United States. Health Educ Behav. 2018;45(1):32-42. doi:10.1177/1090198117709884

78.

Boateng

Wekesah

Browne

, et al. Knowledge and awareness of and perception towards cardiovascular disease risk in sub-Saharan Africa: a systematic review. PloS One. 2017;12(12):e0189264. doi:10.1371/journal.pone.0189264

79.

Stead

Angus

Langley

, et al. Mass media to communicate public health messages in six health topic areas: a systematic review and other reviews of the evidence. Public Health Res. 2019;7(8). doi:10.3310/phr07080

80.

Anawalt

. Diagnosis and management of anabolic androgenic steroid use. J Clin Endocrinol Metab. 2019;104(7):2490-2500. doi:10.1210/jc.2018-01882

81.

Maior

Simão

de Salles

Alexander

Rhea

Nascimento

. Acute cardiovascular response in anabolic androgenic steroid users performing maximal treadmill exercise testing. J Strength Cond Res. 2010;24(6):1688-1695. doi:10.1519/JSC.0b013e3181dc46c9

82.

Nascimento

Corrêa

Lemos

Lima

de Brito Amaral

. Anabolic-androgenic steroids at supraphysiological doses: cardiovascular impacts and pathophysiological mechanisms. J Steroid Biochem Mol Biol. 2026:106938. doi:10.1016/j.jsbmb.2026.106938

83.

Proietti

Raparelli

Olshansky

Lip

. Polypharmacy and major adverse events in atrial fibrillation: observations from the AFFIRM trial. Clin Res Cardiol. 2016;105(5):412-420.

84.

Easton

Grant

Spurling

Magnolini

Piatkowski

. Looking for the best way to come off steroids safely? Exploring post-cycle therapy, cessation, and recovery discourse and practice in Australian steroid consumer forums. Drug Alcohol Rev. 2025;44(7):1951-1960. doi:10.1111/dar.70042

85.

Piatkowski

Coomber

Francis

, et al. Anabolic-androgenic steroid testing as a tool for consumer engagement and harm reduction: a sequential explanatory mixed-method study. Harm Reduct J. 2025;22(1):114.

86.

Piatkowski

Coomber

Francis

, et al. The world’s first anabolic-androgenic steroid testing trial: a two-phase pilot combining chemical analysis, results dissemination and community feedback. Addiction. 2025;120(7):1366-1377. doi:10.1111/add.70009

87.

Mazor

Fleming

. The Dunning-Kruger effect revisited. Nat Hum Behav. 2021;5(6):677-678.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.03 MB