Maki · Menopause (New York, N.Y.) 2013 · narrative review · n=?

Critical window hypothesis of hormone therapy and cognition: a scientific update on clinical studies.

Cited 239 times in the scientific literature.

Level 5 - mechanism / opinion, no new human data

Narrative review synthesizing observational studies, clinical trials, and neuroimaging without systematic review methodology

PubMed 23715379 · doi:10.1097/GME.0b013e3182960cf8 · record verified 2026-08-28

What was done

This review examined clinical evidence regarding the critical window hypothesis of hormone therapy (HT), which posits that cognitive effects depend on timing initiation relative to age or the menopausal transition. The author synthesized observational studies evaluating HT timing and Alzheimer's disease (AD) risk, observational studies assessing cognitive test performance, randomized controlled trials of HT on verbal memory, and relevant neuroimaging investigations.

What was found

The abstract reports counts and directions of studies without numerical effect sizes or confidence intervals: - Three of three observational studies assessing timing of initiation supported the critical window hypothesis for AD risk. - Three of five observational studies evaluating timing supported the hypothesis for cognitive test performance. - Randomized trials of estrogen therapy in younger women supported the hypothesis. - Conjugated equine estrogens combined with medroxyprogesterone acetate increased cognitive risks regardless of timing. - Limited data exist on the cognitive effects of other combination HT formulations.

Why it matters

This synthesis provides a framework to explain discrepant trial and observational findings on postmenopausal hormone use and brain health, highlighting timing and formulation as key variables. However, HT remains indicated strictly for managing vasomotor symptoms, not for preventing dementia or treating cognitive decline.

Limits

As a narrative review, it lacks systematic search criteria, quality appraisal, and meta-analytic pooling. The abstract provides study counts rather than pooled quantitative estimates or effect sizes. Much of the timing data relies on observational designs susceptible to healthy-user bias, and a definitive randomized trial across lifespan windows was noted to be infeasible.

Cited by