To test for associations between long-term anabolic-androgenic steroid (AAS) use and cognitive functioning, and establish a candidate neuronal basis by assessing the associations between cognitive performance and brain morphology both in users and nonusers.
Eighty four previous or current AAS-users and 69 non-AAS-using male weightlifters aged 19-75 years (mean 32.6, SD 8.8) underwent MRI of the brain and a comprehensive neuropsychological test battery. Performance on fine motor speed, speed of processing, learning and memory, working memory, executive functioning, and problem solving was compared between the groups, and between AAS users with short versus long AAS exposure. Associations between cognitive scores and regional cortical thickness and arealization defined using Free Surfer were tested using linear models.
Relative to nonexposed, AAS-exposed weightlifters performed significantly worse on several cognitive domains, independent of age, education, verbal IQ, and exposure to classical drugs of abuse. Strongest effects were observed for speed of processing (ηp2 = .07), working memory (ηp2 = .08) and problem solving (ηp2 = .09). Longer duration of AAS-use was associated with poorer memory function (ηp2 = .11). Within AAS users, individuals with better memory and working memory performance had with thicker frontoparietal cortex and larger medial frontal surface area, respectively.
Prolonged high-dose AAS use is associated with poorer cognitive function across multiple domains, and the observed regional associations between cortical brain morphometry and memory and working memory performance may suggest differential brain-based mechanisms. The public, health care professionals, and policymakers should be aware that use of AAS in large doses potentially could lead to poorer brain health and cognition.