Postmenopausal hormone treatment alters neural pathways but does not improve verbal cognitive function.

Abstract

Cognitive outcomes in trials of postmenopausal hormone treatment have been inconsistent. Differing outcomes may be attributed to hormone formulation, treatment duration and timing, and differential cognitive domain effects. We previously demonstrated treatment benefits on visual cognitive function. In the present study, we describe the effects of hormone treatment on verbal outcomes in the same women, seeking to understand the effects of prior versus current hormone treatment on verbal function. This is a cross-sectional evaluation of 57 women (38 hormone users [25 prior long-term users and 13 current users] and 19 never-users). Hormone users took identical formulations of estrogen or estrogen + progestin (0.625 mg/d conjugated equine estrogens with or without medroxyprogesterone acetate) for at least 10 years, beginning within 2 years of menopause. Women were evaluated with tests of verbal function and functional magnetic resonance imaging (fMRI) of a verbal discrimination task. All women scored similarly on assessments of verbal function (Hopkins Verbal Learning Test and a verbal discrimination task performed during the fMRI scanning session); however, women ever treated with hormones had more left inferior frontal (T = 3.72; P < 0.001) and right prefrontal cortex (T = 3.53; P < 0.001) activation during the verbal task. Hormone-treated women performed slightly worse on the verbal discrimination task (mean accuracy 81.72 ± 11.57 ever-treated, 85.30 ± 5.87 never-treated, P = 0.14), took longer to respond (mean reaction time 1.10 ± 0.17 s ever-treated, 1.02 ± 0.11 never-treated, P = 0.03), and remembered fewer previously viewed words (mean accuracy 62.21 ± 8.73 ever-treated, 65.45 ± 7.49 never-treated, P = 0.18). Increased posterior cingulate activity was associated with longer response times (R = 0.323, P = 0.015) and worse delayed verbal recall (R = -0.328, P = 0.048), suggesting that increased activation was associated with less efficient cognitive processing. We did not detect between group differences in activation in the left prefrontal cortex, superior frontal cortex, thalamus, or occipital/parietal junction. Although current and past hormone treatment was associated with differences in neural pathways used during verbal discrimination, verbal function was not higher than never-users.