Longitudinal associations of hormone therapy and depressive symptoms in cognitively unimpaired older women

Abstract

Abstract Background Women are disproportionately affected by depressive disorders relative to men, and mid-life differences like menopause can exacerbate symptoms. Hormone therapy (HT) is a common treatment to ease menopause transitions, including vasomotor and depressive symptoms, that may occur due to menopause-related hormonal changes, but its effects on later-life depressive symptoms remain unclear. Our objective was to examine the association between self-reported post-menopausal HT use and longitudinal changes in depressive symptoms in cognitively unimpaired age-matched older men and women from the Harvard Aging Brain Study and affiliated studies, along with sub-analyses examining interactions with Alzheimer’s Disease (AD) risk factors and neuroimaging markers. Methods Linear mixed-effect models examined the relationship between HT use and changes in the Geriatric Depression Scale (GDS). Secondary analyses tested the moderating effects of apolipoprotein ε4 ( APOE ε4) status, white matter hyperintensities (WMH), amyloid-β (Aβ) status, medial temporal lobe tau, and hippocampal volume on the association between HT and GDS scores over time. Results In a sample of 585 individuals (58.3% women, Age = 71.6 (+ 7.72 SD)), 42% of women reported current or past HT use. 7% of women self-reported mild to moderate depressive symptoms at baseline (GDS = 11–19). Female HT users reported significantly greater depressive symptoms over time relative to age-matched men (β = 0.11; 95% CI:-0.22- -0.01; p = 0.03) even after adjusting for history of depressive episodes, but not different relatively to female non-users (β=-0.10; 95% CI:-0.21- 0.01; p = 0.07). WMH volume moderated the relationship between HT use and changing depressive symptoms, with female HT users reported greater depressive symptoms over time relative to men (β=-0.09; 95% CI:-0.18- -0.01; p = 0.03) in the setting of high WMH volume, but not relative to female non-users (β=-0.06; 95% CI:-0.15–0.04; p = 0.28). Moderating effects of Aβ, medial temporal lobe tau, and hippocampal volume were not significant. Conclusions In the overall group, female HT users and female non-users did not differ in changes in depression scores. However, female HT users exhibited subtly greater depressive symptoms over time relative to men and WMH moderated this relationship. This underscores the importance of investigating menopausal treatment factors that may affect depressive symptoms in older women and to understand their relationship with white matter lesions.

The authors' abstract, as published at the source. BMC Medicine, 2026 · DOI ↗