Testosterone is the primary human endogenous androgen. German researchers developed the first synthetic testosterone in 1935, aiming to treat depressive disorders.¹ The Cold War stimulated the use of anabolic androgenic steroids (AAS), with the goal of enhancing sports performance, and the first report of doping with testosterone analogs in the Olympics dates back to 1954, involving Russian Olympic weightlifters.² Since the 1980s, AAS usage has expanded beyond elite athletes, and its use for aesthetic purposes has increased in the general population.³ Some studies suggest that the lifetime prevalence of AAS use may be between 1% and 5% of the global male population.³

Testosterone levels in men decrease with aging, and this reduction has been associated with increased atherosclerotic processes and cardiovascular risk.^4–6 In response, numerous studies have been conducted to evaluate the cardiovascular benefits of testosterone replacement therapy in hypogonadal patients. However, the results are conflicting. Meta-analyses show that testosterone replacement therapy has not been able to prevent major cardiovascular outcomes, such as death and myocardial infarction in men with hypogonadism.^7,8

Since 2010, several studies have been published, revealing the possible cardiovascular adverse effects of testosterone replacement therapy. Basaria et al. published a series of studies confirming a higher incidence of acute myocardial infarction in elderly individuals who received exogenous testosterone gel administration, as compared to those who received a placebo.^9,10 Recently, Lincoff et al.¹¹ published a prospective study where 5,246 hypogonadal men with high cardiovascular risk or established cardiovascular disease were randomized to receive testosterone gel replacement or placebo. After a mean follow-up of 33 months, there was no increase in the number of major cardiac events. However, there was an increased incidence of atrial fibrillation, thrombosis, and worsening renal function in patients who received testosterone replacement. These studies were criticized for including a population already at high cardiovascular risk and for their low reproducibility.

So far, the cardiovascular safety of testosterone use remains controversial. This uncertainty is justified by the heterogeneity of studies regarding population, preparations, and studied dosages. While little is known about the cardiovascular effects of therapeutic testosterone replacement in men with hypogonadism, uncertainties are even greater when it comes to the cardiovascular adverse effects of testosterone abuse and its analogs for aesthetic and performance purposes.

Most of these studies involved few individuals in retrospective analyses. Windfeld-Mathiasen et al.¹² recently published a cohort study where doping-sanctioned men (1,189 individuals) and control individuals (59,450 individuals) were followed for a mean period of 11 years. In this study, the relative risk of death was 2.81 times higher in individuals with a history of doping punishment, with cancer and cardiovascular disease being the main causes of non-violent death in this group.

In the last decade, some reports have been published regarding the direct toxicity of testosterone on the myocardium.^13–16 Baggish et al.¹⁶ evaluated 140 young bodybuilders, 86 of whom were users of anabolic-androgenic steroids (AAS) for at least two years, and observed differences in echocardiographic measurements when comparing data from users and non-users of AAS. However, few of the studied patients showed echocardiographic parameters outside the normal range.

Indeed, despite its value and availability in clinical practice, two-dimensional echocardiography presents important limitations, especially regarding inter-observer variability in the assessment of global and segmental ventricular function.¹⁷ Speckle tracking technique is relatively operator-independent, with a much lower inter-observer variability rate than traditional two-dimensional echocardiography.¹⁸ In this context, this advanced echocardiography technique may assist in the early identification of myocardial pathologies induced by the use of AAS.

Our hypothesis is that the use of advanced echocardiographic techniques, such as speckle tracking, may be able to identify cardiotoxicity in AAS users whose routine echocardiographic analyses are normal.

The volunteers had similar age and anthropometric characteristics, as summarized in Table 1. All volunteers denied hypertension and smoking, and were unaware of any comorbidities. Non-users of anabolic steroids (NUAS) denied regular use of medications and hormones, while UAS denied the use of any medications other than AAS. All participants had used AAS intramuscularly for more than two years.

According to the analyzed echocardiographic parameters (Table 2), UAS showed higher left ventricular (LV) mass and left atrial volume compared to NUAS. Although UAS had a lower left ventricular ejection fraction (LVEF), none of them had LVEF < 50%.

However, despite LVEF > 50%, all UAS, when compared to NUAS, showed reduced GLS of the left ventricle (GLS-LV) and longitudinal strain of the right ventricle (Table 3). All UAS (n=14) had GLS-LV < 18%, 11 of which (78%) showed reduced GLS-LV (< 16%), indicating LV dysfunction (Central Illustration).

Regarding myocardial work, UAS showed lower GCW, higher global deconstructive work, lower global efficiency work, and lower myocardial torsion (Table 3).

The use of AAS for aesthetic and performance enhancement purposes is increasing worldwide. In Brazil, it is estimated that about 3% of the population has used or currently uses these substances.²² This study demonstrates that even with normal conventional echocardiographic parameters, AAS users exhibit systolic dysfunction that can be identified using advanced techniques, such as speckle tracking and myocardial work analysis.

GLS of the left ventricle has proven to be an independent predictor of adverse outcomes that are better than LVEF.^18,21 All AAS users in this sample presented below normal GLS-LV values, 78% of whom actually showed systolic dysfunction despite normal LVEF.²³

Baggish et al.²⁴ detected reduced GLS of the left ventricle in AAS users. However, this study mostly included patients with LVEF < 50%, where GLS-LV analysis adds little to the detection of ventricular dysfunction. D’Andrea et al.²⁵ analyzed echocardiograms of 45 elite bodybuilders, of whom 20 reported prolonged AAS use (users), 25 practiced bodybuilding without AAS (non-users), and 25 were controls. In AAS users, as compared to non-users, the longitudinal strain was notably reduced, indicating subclinical systolic dysfunction. However, in this study, AAS users had higher blood pressure than non-users, which may have influenced the results, since increases in systolic blood pressure are known to reduce GLS-LV.²³ In our study, none of the included individuals were hypertensive. Furthermore, we performed myocardial work analysis, confirming the presence of subclinical systolic dysfunction in AAS users.

It is important to note that we found no previous study using both myocardial work analysis and speckle tracking echocardiography to analyze and detect early systolic dysfunction in bodybuilders using AAS. Myocardial work analysis is a recent technique that incorporates LV afterload data into GLS analysis. Besides correlating with myocardial oxygen consumption, myocardial work allows analysis of LV efficiency, proving to be a highly effective technique for ventricular performance analysis.²⁶ The lower values of GCW, myocardial efficiency, and myocardial torsion, as well as the higher global deconstructive work found in AAS users, support ventricular dysfunction related to steroid abuse in the studied population.

Although with a different objective than the present article, Grandperrin et al.²⁷ analyzed the myocardial work of 24 asymptomatic strength athletes using steroids, 22 athletes diagnosed with hypertrophic cardiomyopathy, and 20 healthy control athletes. Their findings did not indicate a difference in constructive and deconstructive work when comparing the UAS and NUAS groups, which differs from our study. These points of divergence may be related to variations in the studied population, as well as in the characteristics of the type and duration of steroid use. However, regarding global efficiency, the group of steroid-using athletes showed lower myocardial efficiency when compared to the controls, corroborating our findings.

Angell et al.²⁸ used cardiac magnetic resonance imaging and speckle-tracking echocardiography to document the impact of AAS abuse on the cardiac structure and function of bodybuilders. In their results, UAS, when compared to NUAS, showed a greater mean LV wall thickness, LV end-diastolic volume, and LV mass, even after covariate analysis, which aligns with our data. Additionally, the analysis of the GLS of the LV also supported our findings, showing a reduction in the group of athletes using anabolic steroids, indicating systolic dysfunction.

The indiscriminate use of AAS is a public health problem in our country and worldwide. It is not uncommon for some professionals to present studies where relatively low doses of AAS were used for a short period of time, using the absence of pathological variation in two-dimensional echocardiograms to justify the safety of this practice. In one of these studies, Chung et al.²⁹ compared the two-dimensional echocardiograms of 30 healthy men who received testosterone (200 mg/week, intramuscular), nandrolone decanoate (200 mg/week, intramuscular), or placebo for 4 weeks, and found no adverse or beneficial effects from these medications. The analysis of our data, together with data from other studies that examined advanced echocardiographic techniques,^16,23,24 emphasizes the need for these techniques to identify early myocardial injury caused by steroids. We highlight that these findings imply a worse prognosis³⁰ and may assist physicians in the difficult task of convincing AAS users to discontinue this practice.³¹

The presence of normal conventional echocardiographic parameters on two-dimensional echocardiography does not exclude the presence of subclinical ventricular dysfunction in UAS. The use of specific techniques, such as speckle tracking and myocardial work, not only provides greater accuracy, but also allows for the early identification of myocardial injury. This directly impacts the prognosis of these patients, providing clinicians with an important tool to persuade this population to discontinue the use of these substances. Future studies should analyze the effects of different steroids and dosages, in an attempt to impove knowledge on the subject and raise awareness about the harmful effects of this increasingly common and trivialized practice.