Andy · Frontiers in endocrinology 2024 · systematic review and meta-analysis · n=34 studies (27,593 participants)

Systematic review and meta-analysis of the effects of menopause hormone therapy on cognition.

Cited 76 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of randomized controlled trials

PubMed 38501109 · doi:10.3389/fendo.2024.1350318 · record verified 2026-08-29

What was done

A systematic review, random-effects meta-analysis, and multi-level meta-regression of 34 randomized controlled trials (14,914 treated participants and 12,679 placebo controls) evaluated the effects of menopausal hormone therapy (MHT) on cognitive function. Subgroup analyses assessed formulation (estrogen-only vs. estrogen-progestogen), timing of initiation (midlife/close to menopause vs. late-life), treatment duration, and menopause etiology (surgical vs. spontaneous).

What was found

MHT had no overall effect on pooled cognitive domain scores. In surgical menopause, treatment (mostly estrogen-only) improved global cognition compared to placebo (SMD = 1.575, 95% CI 0.228 to 2.921; P = 0.043). Estrogen therapy initiated in midlife or close to menopause onset improved verbal memory (SMD = 0.394, 95% CI 0.014 to 0.774; P = 0.046), whereas late-life initiation showed no effect. For spontaneous menopause, estrogen-progestogen therapy was associated with lower Mini-Mental State Examination (MMSE) scores compared to placebo, largely driven by late-life populations (SMD = -1.853, 95% CI -2.974 to -0.733; P = 0.030). In timing analyses, estrogen-progestogen showed no midlife effect but was associated with improved late-life verbal memory (P = 0.049). Treatment duration longer than one year was associated with worse visual memory compared to shorter durations.

Why it matters

These results provide clinical trial evidence supporting the "timing hypothesis," showing that cognitive effects of MHT depend heavily on formulation, surgical status, and when therapy is initiated rather than exerting uniform benefit or harm.

Limits

The abstract notes substantial variability in results across individual cognitive tests, heterogeneous study designs, and wide confidence intervals for key effect estimates. In addition, combined estrogen-progestogen data were skewed toward late-life trial participants, limiting direct comparisons for early-initiation combined regimens.

Cited by