The global epidemiology of anabolic-androgenic steroid use: a meta-analysis and meta-regression analysis

To estimate the global lifetime prevalence rate of anabolic-androgenic steroid (AAS) use and investigate moderators of the prevalence rate.

A meta-analysis and meta-regression analysis was performed using studies gathered from searches in PsycINFO, PubMed, ISI Web of Science, and Google Scholar among others. Included were 187 studies that provided original data on 271 lifetime prevalence rates. Studies were coded for publication year, region, sample type, age range, sample size, assessment method, and sampling method. Heterogeneity was assessed by the I(2) index and the Q-statistic. Random effect-size modeling was used. Subgroup comparisons were conducted using Bonferroni correction.

The global lifetime prevalence rate obtained was 3.3% (95% confidence interval [CI], 2.8-3.8; I(2) = 99.7, P < .001). The prevalence rate for males, 6.4% (95% CI, 5.3-7.7, I(2) = 99.2, P < .001), was significantly higher (Qbet = 100.1, P < .001) than the rate for females, 1.6% (95% CI, 1.3-1.9, I(2) = 96.8, P < .001). Sample type (athletes), assessment method (interviews only and interviews and questionnaires), sampling method, and male sample percentage were significant predictors of AAS use prevalence. There was no indication of publication bias.

Nonmedical AAS use is a serious widespread public health problem.