SAT-207 The Effects of the Menstrual Cycle and Caloric Restriction on Sleep in Young Women

Background: Dynamic changes in reproductive hormone levels across the menstrual cycle have been hypothesized to disrupt normal sleep patterns. While some studies reported increased sleep disruption in the luteal phase, others did not. In addition, metabolic hormones including ghrelin and cortisol are linked to both sleep regulation and the reproductive axis. However, studies have focused on the responses of these hormones to sleep, rather than their effects on sleep. The purpose of this study was to evaluate the effects of hormonal changes across the menstrual cycle and decreased energy availability on Actigraphy-defined objective sleep measures in an in-home setting in young, regularly-cycling women. Methods: As a part of a larger study, reproductive hormones were collected from 10 healthy, regularly-cycling women aged 18-28 years over the course of two menstrual cycles. Daily Actigraphic sleep data (n=578 sleep episodes) and urinary concentrations of LH, estrone-3-glucuronide (E1G), and pregnanediol-3-glucuronide (PDG) were collected. Subjects completed two 5-day diet interventions (neutral versus decreased energy availability) during the early follicular phases (EFP) of separate cycles. Cycles were centered on day of ovulation and standardized to 14-day follicular and 14-day luteal phases. Sleep data were analyzed using linear mixed models by menstrual phase, diet interventions, and reproductive hormones, adjusting for weekday vs weekend. Results: Hormonal measurements confirmed ovulation in both cycles in all subjects (age 24.5 ± 2.5, BMI 22.2 ± 2.1). There was an effect of menstrual phase on sleep efficiency (SE, p=0.005), wake after sleep onset (WASO, p=0.04), number of awakenings per night (p=0.02), and sleep fragmentation index (SFI, p=0.06), consistent with increased sleep disruption in the late luteal phase (LLP). In comparison with the EFP, SE decreased by 3.3% (p=0.0002), WASO increased by 15 minutes (p=0.001), and number of awakenings increased by 3 (p=0.04) in the LLP. Decreased energy availability increased sleep disruption, as indicated by decreased SE (p<0.05), increased WASO (p=0.03), and increased SFI (p=0.07). With decreased energy availability, post-hoc analyses showed increased sleep disruption in both the late follicular phase (LFP) and LLP. E1G was associated with an increase in number of awakenings (p=0.007), such that an increase of E1G by 50 mcg/g Cr increased the number of awakenings by 2.4. PDG was associated with a trend toward increased SFI (p=0.07). Conclusions: Among healthy, young women, objective sleep varies across the menstrual cycle, marked by decreased SE, increased WASO, and increased number and duration of awakenings in the LLP. In addition, decreased energy availability in the EFP increases sleep disruption, with greatest effects in the LFP and LLP.