Severe biventricular cardiomyopathy in both current and former long-term users of anabolic–androgenic steroids

Abstract

Aims

This study aims to explore the cardiovascular effects of long-term anabolic–androgenic steroid (AAS) use in both current and former weightlifting AAS users and estimate the occurrence of severe reduced myocardial function and the impact of duration and amount of AAS.

Methods and results

In this cross-sectional study, 101 weightlifting AAS users with at least 1 year cumulative AAS use (mean 11 ± 7 accumulated years of AAS use) were compared with 71 non-using weightlifting controls (WLC) using clinical data and echocardiography. Sixty-nine were current, 30 former (>1 year since quitted), and 2 AAS users were not available for this classification. Anabolic–androgenic users had higher left ventricular mass index (LVMI) (106 ± 26 vs. 80 ± 15 g/m², P < 0.001), worse left ventricular ejection fraction (LVEF) (49 ±7 vs. 59 ± 5%, P < 0.001) and right ventricular global longitudinal strain (−17.3 ± 3.5 vs. −22.8 ± 2.0%, P < 0.001), and higher systolic blood pressure (141 ± 17 vs. 133 ± 11 mmHg, P < 0.001) compared with WLC. In current users, accumulated duration of AAS use was 12 ± 7 years and in former 9 ± 6 years (quitted 6 ± 6 years earlier). Compared with WLC, LVMI and LVEF were pathological in current and former users (P < 0.05) with equal distribution of severely reduced myocardial function (LVEF ≤40%) (11 vs. 10%, not significant (NS)). In current users, estimated lifetime AAS dose correlated with reduced LVEF and LVGLS, P < 0.05, but not with LVMI, P = 0.12. Regression analyses of the total population showed that the strongest determinant of reduced LVEF was not coexisting strength training or hypertension but history of AAS use (β −0.53, P < 0.001).

Conclusion

Long-term AAS users showed severely biventricular cardiomyopathy. The reduced systolic function was also found upon discontinued use.

Structured Graphical Abstract

Open in new tabDownload slide

Introduction

Anabolic–androgenic steroids (AAS) constitute a wide range of synthetic derivatives of the male sex hormone testosterone that are primarily used in an illicit manner to gain muscle mass for cosmetic or ergogenic purposes.^1,2 A meta-study in 2014 estimated the global lifetime prevalence of AAS use at 6.4% among men, and 1.6% among women, suggesting that non-medical AAS use is widespread.³

Several case reports and small-scale studies have described an association between high-dose AAS use and myocardial hypertrophy and hypertension.⁴ In addition, post-mortem studies have examined young male bodybuilders who suffered sudden cardiac death and found an association between AAS use and myocardial hypertrophy and fibrosis.^4–7 It was not until the 2000s that larger cross-sectional studies were published on the subject. Baggish et al.⁸ and later Fyksen et al.⁹ demonstrated an association between long-term AAS use and myocardial hypertrophy, moderately decreased left ventricular (LV) systolic function, and coronary atherosclerosis. Furthermore, Rasmussen et al.¹⁰ compared current and former AAS users and found moderately decreased LV systolic function in both groups, suggesting that the toxic cardiovascular effects are not fully reversible. The underlying mechanisms of how supraphysiological doses of AAS induce adverse cardiovascular effects are not fully understood. Existing studies often include small numbers of participants, with varying lifetime duration of AAS use and definitions of how to define AAS use, including current and former use. Additionally, only few studies have properly adjusted for the use of psychostimulants, which are prevalent among users of AAS¹¹ and associated with acute and long-term cardiovascular effects,^12–14 These factors may contribute to conflicting results, and as a result, an ongoing debate of what contributes to the adverse cardiovascular effects found among AAS users.

The relationship between blood pressure and AAS use has been debated. In the 2018 European Society of Cardiology (ESC)/European Society of Hypertension (ESH) clinical practice guidelines for the management of arterial hypertension, AAS abuse is listed as one of the causes for resistant hypertension.^9,15–17 In a 2022 position paper on the cardiovascular effects of doping substances, it is still unclear whether pathological myocardial hypertrophy is attributed to arterial hypertension or to the direct effect of AAS on the myocardium.¹⁸

In this large-scale cross-sectional study, we explored the long-term effects of supraphysiological AAS doses on myocardial hypertrophy and myocardial function and if a substantial proportion of the users develop severely reduced myocardial function, by comparing current and former AAS-using weightlifters with a non-using weightlifting control (WLC) cohort, employing echocardiographic imaging techniques. We hypothesized that AAS use is the strongest independent determinant of both increased myocardial hypertrophy and decreased function and that increased duration and lifetime doses of AAS is associated with worsened effects.

Methods

Study design and population

We included men involved in heavy strength training, who volunteered from the community to participate in this cross-sectional study from May 2017 to October 2019. We recruited participants through social media, targeted online fora, and posters and flyers distributed in select gyms in Oslo, Norway. The participants were compensated with a 500 NOK (∼$60) gift card for their participation. The study is part of a longitudinal study investigating the impact of high-dose AAS use on the brain, carried out at the Anabolic Androgenic Steroid Research Group at Oslo University Hospital, Norway (https://www.ous-research.no/anabolic-steroids/). The participants were recruited from this project, and the sample is partly overlapping with the one described in previous brain health publications.^16,19,20

A self-reported one repetition maximum (1RM) bench press of at least 100 kg (∼220 lbs) was required to enter the study. The participants were either current or former (defined as at least 1 year since cessation of AAS use) AAS users reporting at least 1 year of cumulative AAS use or men who had never been exposed to AAS or equivalent doping substances (WLC). In addition, WLC who were treated medically for hypertension were excluded. The study was conducted in accordance with the Declaration of Helsinki and received ethical approval from the Regional Committee for Medical and Health Research Ethics in South-Eastern Norway (REK) (2013/601). Participants received an informational brochure with a complete description of the study prior to participation, and written informed consent was obtained from all participants.

Screening instruments, clinical interview, and examination

Self-report questionnaire, and a semi-structured clinical interview, performed by two investigators (A.B. and L.E.H.) was used to assess relevant background and health information, including hours spent per week on strength and endurance training, weekly intake of alcohol (in units), history of smoking, and use of illicit drugs. In particular, current or former use of the psychostimulants cocaine and amphetamines was mapped with questions related to whether they had ever used the substances (yes/no) and a follow-up quantification question, which was later combined to a 0–4 scale, with the options: never used, less than once a month, monthly, weekly, and daily/almost daily. In addition, AAS users were interviewed about the history and nature of AAS use, including AAS substances used, and whether and when they had ceased using AAS. Estimated lifetime AAS dose was calculated as the lifetime average weekly dose reported and lifetime weeks of AAS exposure, in line with previous studies.^20,21 We obtained blood pressure measurements using an automatic sphygmomanometer, with a standard bladder cuff for most patients and larger cuffs for bigger arm circumferences.

Doping analysis

Urine samples were collected and analysed for external use of androgens using gas and liquid chromatography coupled to mass spectrometry at the World Anti-Doping Agency (WADA)-accredited Norwegian Doping Laboratory at Oslo University Hospital.²² The criteria used to determine external androgen use were (i) urine samples positive for synthetic testosterone compounds and (ii) a testosterone to epitestosterone ratio (T/E) > 15 equivalent to previous work.^22–25 To minimize the problem of some participants not disclosing their use of AAS, WLC with urine findings suggestive of AAS or image- and performance-enhancing drugs banned by the WADA²⁶ were excluded from further evaluation.

Transthoracic echocardiography

Echocardiography was performed using Vivid E95 (GE Vingmed Ultrasound, Horten, Norway). All echocardiographic measurements were obtained by one investigator (VMA) and analysed offline, blinded to AAS status, by another investigator (RA) using the software EchoPAC v203 (GE, Horten, Norway). Two-dimensional (2D) and Doppler echocardiographic measurements were performed according to current standards.²⁷ Left ventricular ejection fraction (LVEF) was calculated by modified Simpson’s biplane method (LVEF Simpson). Left ventricular systolic function was further assessed by global longitudinal strain (LVGLS) using speckle-tracking in the three apical views and defined as the average of peak longitudinal strains from a model with 16 LV segments. If two or more segments in any given view were deemed inappropriate for strain assessments, LVGLS was excluded. Diastolic function was assessed by transmitral pulsed Doppler in the apical four-chamber view by measuring peak mitral inflow velocities of early and late atrial filling (E-wave and A-wave). At the mitral annulus, peak early velocity (eʹ) was measured at both septal and lateral locations. Average eʹ was then calculated from these two measurements. We calculated left atrial end-systolic volume using the area length method in apical four- and two-chamber views. Maximal wall thickness (MWT) was assessed from parasternal short-axis view from all LV segments from base to the apex of the LV. Left ventricular mass was estimated from parasternal views using the formula provided by Devereaux et al.²⁸ Right ventricular (RV) linear dimensions were estimated from an RV-focused apical four-chamber view. The proximal RV outflow tract (RVOT) diameter was measured in the parasternal long-axis. Right ventricular systolic function was evaluated using two parameters: fractional area change (FAC) and right ventricular global longitudinal strain (RVGLS). Fractional area change was obtained in a RV-focused apical four-chamber view and calculated from the difference in end-systolic and end-diastolic area divided by the area in diastole and expressed as a percentage. Right ventricular global longitudinal strain was obtained using speckle-tracking.

Statistical analyses

Continuous data are presented as mean ± standard deviation (SD) or median (25th–75th percentile). Student`s t-test were used to compare means between two groups and ANOVA and Bonferroni post hoc test when comparing three groups. Proportions were compared by the Pearson’s χ² or Fischer’s exact test. Linear regression models were used to assess the strength of the relationship between the continuous variables LV mass indexed to body surface area (BSA) (LVMI), LVEF, LVGLS, and RVGLS and predictor variables known to impact cardiac function or structure such as age, use of AAS, weekly hours of resistance and endurance training, and systolic blood pressure (SBP). Results were presented as standardized beta coefficient (ß) (95% CI) and significance level (P value). Significant predictors (defined as P < 0.10) from univariate models with no adjustment for covariates were included in a full model, with the exception of age, which was forced in. Analyses were reran in the AAS population with the significant predictors and variables associated with the AAS use such as debut age, years of accumulated AAS use, and estimated lifetime dose.

Sensitivity analyses

To further evaluate possible differences between WLC, current AAS users, and former AAS users, similar linear regressions models were run with LVMI, LVEF, LVGLS, and RVGLS as dependent variables, group as fixed factor, age, and the significant predictors (bench press 1RM, SBP, and LVMI for functional measures) included as covariates in the model. Bonferroni post hoc tests were used to test for differences between the three groups. Also, since the use of stimulants is known to cause adverse cardiovascular effects,^12,13 additional sensitivity analyses were computed including ordinal measures of amphetamine and cocaine use as additional covariates in these models. These were run as separate sensitivity analyses as more data were missing on these variables, to preserve power in the main model. To test for possible effects of degree of AAS exposure, the regression model was rerun comparing WLC and subgroups of current AAS users that differ in their duration of AAS use (<5, 5–10, and >10 years of use). Lastly, to investigate potential long-lasting cardiovascular effects into middle age, univariate regression models were conducted, comparing the subgroup of WLC and former users aged 40 or older. In addition, the relations among between LVMI, LVEF, and LVGLS and estimated lifetime AAS dose were investigated with Pearson’s or Spearman’s correlations upon violation of normality and/or linearity among current AAS users. All statistical analyses were performed using SPSS version 25.0 (SPSS, Chicago, Illinois, USA).

Results

Clinical characteristics

Anabolic–androgenic steroid users vs. weightlifting controls

We included 172 participants, 101 were AAS users and 71 WLC. The accumulated duration of AAS use was 11 ± 7 years. Anabolic–androgenic steroid users had higher weight (P = 0.003), body mass index (BMI) (P = 0.002), and BSA (P = 0.02) compared with WLC (Table 1). Systolic blood pressure was higher in AAS users compared with WLC (P < 0.001), while diastolic blood pressure was similar (P = 0.12). None of the WLC were medically treated for hypertension nor did they test positive for AAS or had a T/E ratio above threshold. The frequencies of the specific AAS compounds found in the urine samples are summarized in Supplementary material online, Figure S1 along with a summary of the most commonly used compounds based on self-reports. In addition to AAS, 10% of users reported use of growth hormones and/or clenbuterol in current or last cycle, and 18% reported use of post-cycle supplement drugs such as selective oestrogen receptor modulators, aromatase inhibitors, and/or human chorionic gonadotropin. Post-cycle therapy is commonly used following an AAS cycle to prevent further conversion of excessive testosterone into oestrogens via the aromatase enzyme, to restore the hypothalamus–pituitary–gonadal axis and to recover endogenous testosterone production.

Current, former, and no anabolic–androgenic steroid use

Among 101 AAS users, 69 were current users, 30 were former users, and 2 users were unclassifiable with regard to current or former status and therefore not included in further subgroup analyses (Table 1). Mean time since AAS cessation among former users was 6 ± 6 years [median 3.75 (1–25)]. Former AAS users reported ∼3 years shorter accumulated duration of AAS use (9 ± 6 vs. 12 ± 6 years, P = 0.05), but estimated lifetime AAS dose did not significantly differ between former and current users (461 ± 416 vs. 716 ± 710 g, P = 0.07). Systolic blood pressure was also at the same high level in former and current users, significantly different from WLC. Doping analysis tests indicative of AAS use were seen in 81% (n = 53) of current users and in 17% (n = 3) of former users. The positive tests among former users involved elevated ratio between testosterone to epitestosterone (T/E ratio) and were consistent with reported medical use of testosterone replacement therapy.

Echocardiography

Anabolic–androgenic steroid users vs. weightlifting controls

Measurements of LV structure were consistent with AAS users displaying concentric hypertrophy, with markedly higher LVMI, MWT, and relative wall thickness (all P < 0.001) compared with WLC (Table 2, Figure 1A). Anabolic–androgenic steroid users had reduced LV function, by LVEF and LVGLS compared with WLC (Table 2). Of note, 36% of AAS users had LVEF between 41 and 49%, and 11% had LVEF below or equal to 40%. Right ventricular systolic function measured by RVGLS and FAC was impaired among AAS users compared with WLC.

Assessment of LV diastolic function according to the American Society of Echocardiography (ASE)/European Association of Cardiovascular Imaging (EACVI) guidelines²⁹ showed that only three AAS users (3%) were found to have grade II diastolic dysfunction and no users in grade I or III (Table 2). Like all the WLC, 97% (98/101) of the AAS user did not have diastolic dysfunction according to ASE/EACVI guidelines.

Current, former, and no anabolic–androgenic steroid use

Current AAS use was associated with greater pathology, whereas former users were somewhere in between current users and WLC (Table 2, Figure 1, Supplementary material online, Table S1). The proportion of current and former users by LVEF category (≤40, 41–49, and ≥50%) did not significantly differ (Table 2). Significant differences between former users and WLC were also seen in the subsample aged 40 or older, with higher mass, reduced LVEF, and RVGLS (see Supplementary material online, Table S2).

Severely reduced left ventricle ejection fraction among anabolic–androgenic steroid users

The distribution between severely reduced LVEF (LVEF ≤ 40%), moderate reduced LVEF (LVEF 41–49%), and subnormal or normal LVEF (LVEF ≥ 50%) is showed in Table 2 and illustrated in Figure 2. The 11% AAS users with severely reduced LVEF (LVEF ≤ 40%) had higher weight and BMI and higher weekly AAS dose compared with the AAS users with moderate reduced or subnormal/normal LVEF, but estimated lifetime AAS dose, strength training load, and bench press 1RM did not achieve significance (Table 3).

Linear regression analysis for determinants of biventricular cardiomyopathy in the total population

In the multivariate regression model, history of AAS use was the strongest independent determinant of higher LVMI, while strength training, SBP, and bench press 1RM contributed to a lesser degree to the model (see Supplementary material online, Table S3). History of AAS use was the only independent determinant of decreased LVEF and of reduced RV function (see Supplementary material online, Tables S4 and S6). Left ventricular mass index was identified as the strongest independent determinant of reduced LV function, closely followed by history of AAS use (see Supplementary material online, Table S5). Sensitivity analyses revealed that the main effect of group remained significant for the two main measures of LV function, LVEF and LVGLS, when amphetamine and cocaine were included as additional covariates in these models (see Supplementary material online, Tables S7 and S8).

Impact of duration of use on key echocardiographic parameters

Sensitivity analyses among current AAS users divided into subgroups based on duration of AAS use showed main effects of group on all echocardiographic parameters (see Supplementary material online, Table S9). The post hoc comparisons indicated that all significant differences applied to differences between the WLC and subgroups of AAS users, whereas no significant differences were found between subgroups of users varying in different lengths of AAS use. Briefly, for LVMI and RVGLS, all AAS subgroups were significantly different compared with the WLC, with higher mass or reduced GLS, respectively. For LVEF and LVGLS, significant differences were also found, although not after shorter exposure of AAS (<5 years) (see Supplementary material online, Table S9). Lastly, Spearman’s ‘Rho’ correlations, including only current AAS users (n = 69), showed significant relations between estimated lifetime AAS dose and LVEF (r = −30.7, P = 0.01, n = 67) and LVGLS (r = 30.7, P = 0.01, n = 50) but no relation to LVMI (r = 19.0, P = 0.12, n = 68) (Figure 3).

Discussion

This study represents the largest cardiovascular study to date, comparing AAS users to WLC, enabling subgroup analyses to be performed on AAS users with adverse pathology. There are several important aspects to consider. First, our study showed that long-term AAS use was strongly associated with left ventricular hypertrophy (LVH), reduced LV and RV systolic function, and impaired diastolic function indicating that long-term AAS use is associated with biventricular cardiomyopathy. Second, increased estimated lifetime AAS dose and duration of AAS use was associated with increased myocardial pathology. Third, history of AAS use was the strongest independent determinant of increased LVH and decreased LV systolic function and outperformed psychostimulants, SBP, and strength training. Fourth, former AAS users had reduced LV and RV systolic function and sustained increased SBP, comparable with that of current AAS users. Finally, 11% of the AAS users had severely reduced LVEF <40%.

Cardiac structure and function

Using sensitive echocardiographic tools, we were able to demonstrate that long-term AAS users when compared with WLC exhibited cardiovascular changes compatible with a biventricular cardiomyopathy. Anabolic–androgenic steroids users had thicker LV walls and a higher LVMI, reflecting the structural hypertrophic changes that are associated with AAS, consistent with findings in past studies.^8,10 The rather moderate increase in LV wall thickness in our WLC can be explained by the blood pressure response and cardiac output during weightlifting, as heavy-resistance training increases blood pressure up to 480/350 mmHg.³⁰ The moderate LVH is also found among other strength athlete cohorts, while LV systolic function generally remains unchanged.³¹

Left ventricular systolic function, as assessed by both LVEF and LVGLS, was significantly impaired in AAS users compared with WLC, with substantial absolute differences, also consistent with past studies.^8,10 The findings could not be explained by concurrent psychostimulant use, albeit we cannot rule out that co-current use would present an additional risk for acute cardiovascular events. Importantly, our study featuring a large population provided a unique opportunity to examine previously unexplored subgroups of interest. Specifically, we observed a high occurrence of moderately and severely impaired LV systolic function (defined as LVEF 41–49 and ≤40%, respectively) among both current and former AAS users. A study combining two extensive community cohorts found a 5-year mortality rate of 66%, with 63% of deaths attributed to cardiovascular disease, following diagnosis of heart failure among patients with a reduced left ventricle ejection fraction (defined as LVEF <50%).³² Our study adds important knowledge regarding the substantial risk of developing advanced heart failure upon long-term use of AAS, suggesting that AAS users may be at high risk of cardiac death.

Right ventricular function was worse in AAS users compared with WLC, as demonstrated by a lower FAC and RVGLS, and is consistent with the findings of D`Andrea et al.³³ On RV linear dimensions and RVOT, both AAS users and WLC had values either in the upper range or higher according to ASE/EACVI guidelines.³⁴ However, RV linear dimensions nor RVOT differed between AAS users and WLC, which may reflect chronic maladaptive cardiac remodelling caused by weightlifting. Right ventricular outflow tract is a frequent focus of ventricular tachycardia both in the general population and in athletes³⁵ and is therefore of particular interest. In endurance athletes, lifetime exercise duration has been demonstrated to be associated with larger RVOT diameter.³⁶

Our findings of myocardial hypertrophy in former users fall somewhere in between current users and WLC, suggesting a partial regression of hypertrophy following discontinuation. In contrast, reduced left and right systolic function measured with LVEF and RVGLS was at the same level in former and current AAS users. Subsample findings indicate that these differences persist into middle age. Considering that the study sample comprises former AAS users, which on average ceased several years ago, the persistently impaired biventricular myocardial function and increased SBP is worth noting. Our finding of reduced LV systolic function in former AAS users is in contrast with recent studies that have found such impairment selectively in current AAS users.^8,10,37 In our study, nearly half of the AAS users, including both current and former users, had LVEF below 50%, and 11% LVEF below 40%, suggestive of moderate or severe LV systolic dysfunction in both current and former users. Our sample differs from previous studies in terms of the duration of AAS usage, showing an accumulated use of 12 years among current users and 9 years among former users. These durations notably exceed the corresponding durations observed in the aforementioned studies. Hence, it is possible that the low LVEF of past users of AAS is a consequence of long-term use.

Among current AAS users reporting <5, 5–10, and >10 years of accumulative use, we found comparable levels of myocardial pathology in subsequent subgroup analyses. For LVMI and RVGLS, significant differences were observed between AAS users with <5 years of use compared with WLC. While it is possible to infer that LVH and RV systolic dysfunctions may serve as early indications of cardiovascular toxicity associated with AAS usage, it is important to note that the limited sample size within the former subgroup hampers a definitive conclusion. Significant and more pronounced outcomes were observed with prolonged AAS usage among current users, indicating a progressive decline in myocardial function over time. These findings were supported by the observed correlation between decreased LVEF and LVGLS and the increasing estimated lifetime AAS dosage. Nevertheless, the findings point to long-term complications of use, suggesting persistent increased risk of adverse cardiovascular events, and need to be further examined in large-scale longitudinal samples.

Determinants of cardiac structure and function

The results of our univariate and multivariate linear regression analyses found that history of AAS use was the strongest determinant of LVH measured by LVMI and of LV and RV systolic function measured by LVEF and RVGLS, respectively. History of AAS use outperformed both SBP and strength training per week. It is not entirely clear whether the drug-mediated cardiovascular effects of AAS are secondary to that of hypertensive effects of AAS or if AAS affects both blood pressure and cardiac parameters, although our linear regression analysis suggests that the latter may be true. Further longitudinal studies are necessary to evaluate susceptibility, progression, and independent predictors of LVH and decreased LV and RV systolic function.

Clinical implications

This study showed that long-term use of AAS was associated with pathology affecting both ventricles and thus compatible with a biventricular cardiomyopathy. The biventricular changes may be early manifestations of AAS use. We showed that history of AAS use outperformed hypertension and strength training, as the biggest contributor to cardiovascular pathology in users of AAS, in the largest cardiovascular study to explore determinants of pathology in AAS users. We recruited AAS users with a broad sociodemographic background from the general population and not a subset of elite athletes. For that reason, clinicians who encounter long-term users of AAS should be aware that elevated blood pressure, LVH, and impaired LV and RV systolic function are not uncommon findings. Our study reiterates the critical need for longitudinal studies. It has been demonstrated that regression of LVH in hypertensive patients is associated with lower likelihood of cardiovascular morbidity and mortality.³⁸ A similar effect could occur in AAS patients upon cessation of substance use, but our study indicates that long-term former AAS use still shows cardiovascular pathology several years after cessation of AAS use.

Limitations

This study has several limitations. First, it was conducted as a single-centre, cross-sectional study, which does not allow for definitive conclusions regarding cardiovascular risk nor does it allow for claims of causality. Second, the exercise history and history and nature of AAS use was self-reported and therefore subject to recall bias. The reported strength training per week did not take into account the lifetime training burden, yielding possible underestimates of athletic training on myocardial hypertrophy and function. Third, there are potential threats to the study’s generalizability. Although we recruited participants from the community, selection bias could still arise. Our study population consisted mainly of white Scandinavians, and therefore, our results might not cover the full ethnic spectrum of AAS users. Conversely, this yields in a high degree of racial homogeneity, meaning race will not interfere as a confounding factor in explaining cardiovascular differences between our two main cohorts of AAS users and WLC. Finally, further evidence of an AAS-induced cardiomyopathy and cardiovascular risk would require a longitudinal follow-up.

Conclusions

Long-term AAS users exhibited myocardial hypertrophy, impaired LV and RV systolic function, and impaired diastolic function indicating a biventricular cardiomyopathy, with high occurrence of severely reduced ejection fraction. Importantly, substantial impairment of systolic function, as assessed by LVEF, and pathologically increased SBP were also seen in former users years after AAS discontinuation, which may indicate permanent changes in function. History of AAS use was the strongest independent determinant of both LVH and reduced LV and RV systolic function. Long-term high-dose AAS use could prove to be a major public health concern, but there is a need for large longitudinal studies examining occurrence of adverse cardiovascular events related to amount and duration of AAS use.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology.

Acknowledgements

The authors are grateful for the valuable contributions made by all the study participants.

Author contribution

Authorship: R.A., A.B., T.E., K.H.H., and V.M.A. contributed to the conception or design of the work. R.A., A.B., L.E.H., I.R.H., T.E., K.H.H., and V.M.A. contributed to the acquisition, analysis, or interpretation of data for the work. R.A., A.B., and V.M.A. drafted the manuscript. R.A., A.B., L.E.H., I.R.H., T.E., K.H.H., and V.M.A. critically revised the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Funding

This research was funded by the Helse Sør-Øst RHF [grant nos. 2016049, 2017025, 2018075, and 2020088 (to A.B.)] and Norges Forskningsråd (#309762) Precision Health Center for optimized cardiac care (ProCardio).

Data availability

The data underlying this article cannot be shared publicly due to privacy for the individuals that participated in the study. The data will be shared on reasonable request to the corresponding author.

References

Author notes