Baseline Non-Comparability in Anabolic–Androgenic Steroid Research: An Overview of Systematic Reviews and Implications for Human Performance Enhancement and Health

Abstract Purpose Comparisons between users and non-users of anabolic–androgenic steroids (AAS) dominate the literature on associated health risks, often assuming AAS exposure as the primary distinguishing factor. However, AAS use in performance-enhancing contexts frequently co-occurs with baseline medical and psychosocial differences, polypharmacy, and concomitant substance use, potentially compromising causal interpretation of adverse outcomes. Methods This overview of systematic reviews was conducted in accordance with the Cochrane Handbook and the PRIOR guideline. Electronic databases and supplementary sources were searched to identify systematic reviews comparing AAS users and non-users. Outcomes focused on baseline medical and psychosocial characteristics, polypharmacy, concomitant substance use, and their implications for between-group comparability. Methodological quality and risk of bias were assessed using AMSTAR-2 and ROBIS, and primary study overlap was quantified using the Corrected Covered Area (CCA). Results Fourteen systematic reviews published between 2013 and 2025 were included. Baseline psychosocial differences, including body image dissatisfaction and psychiatric vulnerabilities, and polypharmacy were commonly reported among AAS users; however, no review explicitly evaluated their impact on group comparability or causal attribution. Polypharmacy and clustered substance use were prevalent in performance-enhancing populations. Methodological quality was predominantly low or critically low, while primary study overlap was minimal (CCA 1.3%). Conclusions The synthesis literature on AAS does not adequately account for baseline non-comparability and co-exposures, limiting causal inference and making the attribution of cardiovascular, psychiatric, reproductive, and metabolic outcomes to AAS alone uncertain. Greater caution is warranted when interpreting causal claims. These findings have implications for research design, clinical assessment, harm reduction strategies, and health monitoring in AAS users.