Association between testosterone and cognitive performance in postmenopausal women: a systematic review of observational studies

This review was conducted to explore the association between endogenous testosterone blood concentrations and cognitive performance among community dwelling postmenopausal women. We searched Ovid MEDLINE, EMBASE, PsycINFO and Web of Science databases for observational studies with at least 100 postmenopausal participants. The results were categorized by study design, reporting of total or free testosterone and risk of bias assessments, narratively. Ten of the 26 articles retrieved for full-text review met the inclusion criteria, six provided cross-sectional data, seven provided longitudinal data and one provided case-control data. Cognitive performance tests differed between studies. Eight studies measured testosterone by immunoassay, one by liquid chromatography-tandem mass spectrometry (LC-MS/MS) and one did not specify their methodology. Eleven different cognitive domains were tested by 37 different instruments. Irrespective of the study design, the findings were inconsistent and inconclusive. Both positive and inverse associations were reported for each of global cognition and immediate and delayed verbal recall. The majority of studies reported no association between total or free testosterone and cognitive performance. Although this review did not demonstrate an association between testosterone and cognitive performance in postmenopausal women, the findings should be considered inconclusive due to the imprecision of testosterone measurement and the methodological heterogeneity of the included studies.