Different temporal relationship between sex hormones and sleep status in midlife women: a longitudinal cohort study

Fluctuation in sex hormones and the occurrence of sleep disturbance are 2 major health challenges among midlife women. However, the temporal relationship between them remains unclear.

This study included 2488 females (mean ± SD age, 49.0 ± 2.70 years) with an average follow-up of 6.95 years. We constructed a composite score by summing items related to sleep problems to reflect the comprehensive sleep status of the participants in the Study of Women's Health Across the Nation. Cross-lagged path analysis was used to examine the temporal relationship between sex hormones and sleep status. Sensitivity analyses were conducted in nonoverweight and overweight groups and adjusted for vasomotor symptoms in the main model.

In this study, we aimed to examine the temporal relationship between sex hormones and sleep status in midlife women using cross-lagged path analysis.

The primary outcomes included results of the cross-lagged path analysis between sex hormones and sleep status.

After adjusting for age, race, income, menopausal status, body mass index, hormone therapy use, smoking, and drinking, the cross-lagged path coefficients from baseline follicle-stimulating hormone (FSH) and estradiol (E2) to follow-up sleep status were 0.054 (P = .017) and -0.054 (P = .016), respectively. The path coefficient from baseline sleep to follow-up dehydroepiandrosterone sulfate (DHAS) was 0.042 (P = .017). The path coefficients between testosterone and sleep were not statistically significant. In the nonoverweight group, the patterns of the temporal relationship between sex hormones and sleep were the same as the total sample, and the point estimates were larger. However, the temporal relationships in the overweight group were nonsignificant. After adjustment for vasomotor symptoms in the main model, results were basically consistent.

Given the temporal relationship between sex hormones and sleep, our findings will provide scientific perspectives to benefit health management in the transition of menopause.

This study used a longitudinal theoretical model to distinguish the temporal relationship between sex hormones and sleep status in midlife women. Limitations include limited causal evidence in observational studies, unknown confounders, and careful extrapolation.

There were distinct patterns in the unidirectional temporal relationship between (1) FSH, E2, and DHAS and (2) sleep. Changes in FSH and E2 occurred earlier than the change of sleep, while the change of DHAS was later. In contrast, there was no temporal relationship between testosterone and sleep.