Abstract
OBJECTIVES: To examine associations of pituitary-ovarian hormone levels with cognition before and after different formulations of hormone therapy (HT) or placebo independent of treatment group. METHODS: Recently menopausal, healthy women were randomized to 0.45 mg/day oral conjugated equine estrogens (o-CEE, n = 109), 50 μg/day transdermal 17β (tE2, n = 107) or placebo pills and patches (n = 146); women on active treatment received oral 200 mg/day micronized progesterone for 12 days per month. Levels of estrone, 17β-estradiol, follicle stimulating hormone, luteinizing hormone, androstenedione, and testosterone were determined prior to and after 48 months of study participation. Neuropsychological testing was administered at baseline, and months 18, 36 and 48. Latent growth curve models controlling for education level, age, APOE allele status, waist circumference, and treatment examined the trajectories of each cognitive domain after accounting for the effect of hormone levels at baseline and months 18, 36 and 48. A linear multivariate mixed model examined the effect of changes in hormone levels on changes in trajectories of complex attention tasks with varying degrees of difficulty. RESULTS: All women were adherent to treatment at month 48. Higher baseline estrone levels were associated with poorer global cognition, auditory attention and working memory, visual attention, and executive function, but not working memory. Higher levels of baseline 17β-E2 were associated with poorer cognitive performance, with marginal significance at baseline in speeded language and mental flexibility (p = 0.013). Other hormone levels were not associated with cognition. Controlling for all treatments, hormone levels at baseline and at month 48 did not have any significant correlation with cognitive trajectories over time. SUMMARY: In healthy, recently menopausal women, baseline estrone levels were inversely associated with selected cognitive factors independent of two types of HT or placebo during 4 years of follow-up. Baseline levels of the other pituitary-ovarian hormones studied were not associated with baseline cognition, nor were changes in any hormones associated with changes in cognition during the study. The marginal association between estradiol levels and cognitive factors warrants further investigation. GOV NUMBERS: NCT00154180, NCT00623311.