Objectives: Women report worse sleep quality during and after the menopausal transition than in premenopause, but null or minimal alterations in sleep architecture have been found. Concentrations of cortisol seem to increase temporally during perimenopause. However, the associations between cortisol and sleep architecture in women around the menopausal transition have not been examined in detail to date.

Study design: 24-hour serum cortisol samples were taken at 20-minute intervals from 16 perimenopausal and 18 postmenopausal healthy women. Sleep architecture was measured with all-night polysomnography from 11:00 pm to 7:00 am.

Main outcome measures: Associations between day, night and diurnal cortisol measures and overnight sleep architecture were examined via correlation analyses and linear regression models adjusted for age, body mass index and depression symptom score.

Results: Higher cortisol levels correlated with lower total sleep time (day mean ρ=-0.36, night maximum ρ=-0.38, diurnal mean ρ=-0.37; p<0.05), sleep efficiency (day mean ρ=-0.36, night mean ρ=-0.35, night maximum ρ=-0.39, diurnal Area Under Curve -AUC ρ=-0.35; p<0.05), slow wave sleep (SWS) percentage (day maximum r=-0.38, night minimum r=-0.37, diurnal maximum r=-0.36, diurnal minimum r=-0.37; p<0.05), total slow wave activity (night minimum ρ=-0.35, p<0.05), number of REM (day mean r=-0.40, diurnal minimum r=-0.37, day AUC r=-0.37, diurnal AUC r=-0.42, p<0.05; night mean r=-0.44, diurnal mean r=-0.44, night AUC r=-0.41, p<0.01) and SWS periods (day mean ρ=-0.36, night mean ρ=-0.35, night minimum ρ=-0.40, diurnal mean ρ=-0.38, diurnal minimum ρ=-0.41, day AUC ρ=-0.36, diurnal AUC ρ=-0.39; p<0.05). Higher cortisol levels correlated with higher percentage of wake after sleep onset -WASO (day mean ρ=0.35, night mean ρ=0.36, night maximum ρ=0.43, diurnal mean ρ=0.40, night AUC ρ=0.39, p<0.05). Additionally, higher cortisol awakening response correlated with lower SWS percentage (r=-0.52, p=0.002) and longer latency to SWS (ρ=0.47, p=0.007). Most of the associations of higher cortisol levels with less REM and SWS periods, and with higher WASO percentage remained significant in adjusted linear regression models.

Conclusions: High day, night and diurnal cortisol levels seem to be associated with worse objective sleep in perimenopausal and postmenopausal women. Whether higher cortisol levels around the menopausal transition contribute to poorer sleep quality in menopause has to be studied.