Menopause and mitochondria: windows into estrogen effects on Alzheimer's disease risk and therapy.
BACKGROUND A randomized trial of postmenopausal women over age 65 reported increased risks of cognitive decline with 4 years of combined estrogen and progestin treatment. However, questions remain, including the effect of longer duration or of hormone therapy initiated at younger ages. METHODS The Nurses' Health Study is a prospective cohort begun in 1976, comprising 121,700 female nurses who report health information via biennial mailed questionnaires. This substudy includes 13,807 participants age 70 to 81 who completed two telephone cognitive assessments, 2 years apart, between 1995 and 2003. General cognition, verbal memory, category fluency, and attention were tested. Multiple linear regression was used to estimate adjusted mean declines and logistic regression to estimate adjusted risks of substantial decline in cognition (> or =2 SD of baseline performance) across hormone groups. APOE genotype was available in a subset of 3,667 participants. RESULTS Overall, little difference was found in mean cognitive decline between current hormone users and never users. However, for long-term users of estrogen alone or combined with progestin, increased risk of substantial decline was observed on most cognitive tests (relative risk [RR] = 1.25 to 1.72). Decline was particularly high among women initiating hormones at older ages compared with never users: for example, on our test of general cognition, RR of substantial decline was 1.74 (95% CI 1.08, 2.81) and mean difference in decline was -0.43 (95% CI -0.73, -0.12). No significant interactions between hormone use and APOE epsilon4 allele were observed. CONCLUSION Postmenopausal hormone therapy provides no appreciable cognitive benefits in older women.