LAUSR

Cognitive performance may be adversely affected during the menopause transition from hot flash-induced sleep fragmentation even without changes in sleep duration. We examined the effects of experimentally-induced sleep fragmentation without shortened sleep duration on daytime sleepiness and neurobehavioral performance in women in a high and low estradiol (E2) state. Seven pre-menopausal women (29.4 ± 3.8 years) participated in two 6-day inpatient studies repeated in a high-E2 (mid-to-late follicular phase) then low-E2 state (gonadotropin-releasing hormone agonist-induced E2 suppression - similar to levels during menopause) ~6 weeks apart. Sleep was uninterrupted on nights 1–2 [8-h time-in-bed (TIB)] and fragmented on nights 3–5 (9-h TIB) using an auditory stimulus delivered every 15 min that sustained wake for 2 minutes, producing 1-h total wake after sleep onset. Wakefulness was confirmed by event-markers during polysomnographically-recorded sleep episodes. Daytime subjective sleepiness (Karolinska Sleepiness Scale; KSS) and neurobehavioral performance (Psychomotor Vigilance Task; PVT) were assessed every 2–3 hours on study days 2–5. The effects of study day and E2 state on KSS scores and PVT measured reaction time (RT) and attentional failures (RT>500ms) were examined using linear mixed models. Participants reported feeling sleepier (+10%), had longer RTs (+22ms), and more attentional failures (+53%) after sleep fragmentation than after uninterrupted sleep (all p<0.001). While there was no main effect of E2 state, there was a differential effect of sleep fragmentation by E2 state on PVT, but not sleepiness, such that the increase in RT and attentional failures in response to sleep fragmentation was only observed in the high-E2 state (p<0.001). Eight hours of total sleep time may not be sufficient to maintain subjective sleepiness and PVT performance levels when sleep is not consolidated. These findings have important implications for understanding the role of sleep and E2-modulated cognitive impairment during the menopause transition. This work was supported by the NIH: 5R01 AG053838-02 (HJ) and K24-HL105664 (EBK).