Transdermal testosterone improves verbal learning and memory in postmenopausal women not on estrogen therapy: A randomized placebo-controlled trial over 26 weeks

Background: There is biological evidence that testosterone (T) is neuro-protective (1). We have previously reported in 2 separate open-label studies that transdermal T improves verbal learning and memory in postmenopausal women (PM) (2,3). Objective: To determine the effects of transdermal T on cognitive performance in healthy PM women in a double-blind randomized placebo-controlled (RCT) single center study. Methods: 96 healthy PM women, aged 55-65 years were screened and 92 randomly allocated to treatment with a daily dose of transdermal T gel 0.22g per day (LibiGel®), or identical placebo (Pl), for 26 weeks. Women were excluded if they had used any form of systemic hormone therapy in the prior 12 months or had any characteristic that may impair cognitive performance. Cognitive performance was assessed at 0, 12 and 26 weeks using a computerized program, CogState, which delivers a different, randomly allocated test to each participant for each cognitive domain on each occasion. Wellbeing was assessed using the Psychological General Wellbeing Index (PGWB). The analysis was intention-to-treat. Modeling was performed for each cognitive domain using linear regression, with the outcome variable being the 26 week score and the independent variables, age, treatment group and baseline score. A bootstrapping approach with 5000 repetitions was used to address the issue of outlying values. Results: 90 women, mean age 61 years, provided data for analysis. There were no between-group differences for CogState, PGWB score or total T (mean 0.5 ± SD 0.3 nmol/L) at baseline. At week 26, the T group showed a significant improvement in performance for the International Shopping list task (improved verbal learning and memory) compared with the Pl group (mean baseline score both groups 33.7, T group increase in score over Pl group, 1.57, adjusted for individual age and baseline score, p=0.037). Serum total T increased in the T group, mean 1.9 ± 1.2 nmol/L and was unchanged with Pl therapy 0.4 ± 0.2 at week 26. There were no significant differences for any other cognitive test or PGWB total score or domain. Conclusion: This study shows improvement in short term verbal learning and memory with T therapy in PM women, consistent with our earlier studies. Together these studies provide compelling evidence for the conduct of large scale clinical studies to further investigate the use of T to prevent memory decline in PM women.