Reproductive hormone levels, androgen receptor CAG repeat length and their longitudinal relationships with decline in cognitive subdomains in men: The European Male Ageing Study.

• Sex steroid levels are proposed to be a modifiable risk factor for cognitive ageing • We tested this hypothesis in a large-scale study of middle-aged and older men • Steroid levels did not affect cognition after adjusting for health and lifestyle • Androgen CAG repeat length did not alter hormone-cognition relationships It has been proposed that endogenous sex hormone levels may present a modifiable risk factor for cognitive decline. However, the evidence for effects of sex steroids on cognitive ageing is conflicting. We therefore investigated associations between endogenous hormone levels, androgen receptor CAG repeat length, and cognitive domains including visuoconstructional abilities, visual memory, and processing speed in a large-scale longitudinal study of middle-aged and older men. Men aged 40-79 years from the European Male Ageing Study (EMAS) underwent cognitive assessments and measurements of hormone levels at baseline and follow-up (mean = 4.4 years, SD ± 0.3 years). Hormone levels measured included total and calculated free testosterone and estradiol, dihydrotestosterone, luteinizing hormone, follicle-stimulating hormone, dehydroepiandrosterone sulphate and sex hormone-binding globulin. Cognitive function was assessed using the Rey-Osterrieth Complex Figure Copy and Recall, the Camden Topographical Recognition Memory and the Digit Symbol Substitution Test. Multivariate linear regressions were used to examine associations between baseline and change hormone levels, androgen receptor CAG repeat length, and cognitive decline. Statistical analyses included 1,827 and 1,423 participants for models investigating relationships of cognition with hormone levels and CAG repeat length, respectively. In age-adjusted models, we found a significant association of higher baseline free testosterone (β=-0.001, p =0.005) and dihydrotestosterone levels (β=-0.065, p =0.003) with greater decline on Rey-Osterrieth Complex Figure Recall over time. However, these effects were no longer significant following adjustment for centre, health, and lifestyle factors. No relationships were observed between any other baseline hormone levels, change in hormone levels, or androgen receptor CAG repeat length with cognitive decline in the measured domains. In this large-scale prospective study there was no evidence for an association between endogenous sex hormone levels or CAG repeat length and cognitive ageing in men. These data suggest that sex steroid levels do not affect visuospatial function, visual memory, or processing speed in middle-aged and older men. [ABSTRACT FROM AUTHOR]