Sleep disturbances during perimenopause and later-life cognitive function

To evaluate whether self-reported sleep disturbances during perimenopause are associated with cognitive function decades later in postmenopausal women.

We analyzed 2,097 postmenopausal women without known dementia in the St Louis Baby Tooth-Later Life Health Study (2019‑2023). At one postmenopausal visit, participants retrospectively reported perimenopausal sleep, rated current sleep, and completed a cognitive battery. Sleep disturbances were assessed with the Neuro-QoL Sleep Disturbance scale and classified as none to slight, mild, or moderate to severe. Cognitive function was summarized as a composite Z-score from the TestMyBrain battery. Multiple linear regression estimated adjusted differences (SD units) in cognition across sleep disturbance categories, stratified by frequency of being woken by menopausal symptoms (ie, "never or rarely" vs. "sometimes to always").

Perimenopausal sleep disturbances were associated with lower cognitive scores compared with none to slight (mild: -0.07; 95% CI, -0.17 to 0.02; moderate to severe: -0.15; 95% CI, -0.29 to -0.02). No association was observed among women never or rarely woken by menopausal symptoms, whereas larger deficits were seen among those woken sometimes to always (mild: -0.17; 95% CI, -0.28 to -0.06; moderate to severe: -0.31; 95% CI, -0.47 to -0.16). Findings were unchanged after adjustment for hormone therapy or for concurrent sleep disturbances measured at the time of cognitive testing.

Perimenopausal sleep disturbances were linked to poorer cognitive function decades later, particularly in women frequently woken by menopausal symptoms. These results suggest that sleep and menopausal symptoms in midlife may be relevant markers of later-life cognitive health, and warrant confirmation in prospective studies.