Overnight Melatonin Concentration and Sleep Quality Are Associated With Clinical Features of Polycystic Ovary Syndrome

Introduction: Women with polycystic ovary syndrome (PCOS) have an increased incidence of sleep disturbances compared to healthy women. Circulating melatonin (MEL) is elevated in women with PCOS, thought to reflect an increased daytime and blunted peak in overnight MEL, consistent with an altered circadian rhythm. Whether circadian disruptions coincide with sleep disturbances in women with PCOS or their symptom severity is unclear. Objective: To determine whether altered MEL production coincides with reduced sleep quality in women with PCOS and to examine whether there is a relationship between MEL production, sleep disturbances and the diagnostic features of PCOS. Methods: Women with PCOS (n=22) and controls (n=12) were recruited prospectively. PCOS was defined based on the 2018 International Guideline. Controls exhibited no more than 1 diagnostic feature of PCOS. Women underwent a reproductive history, clinical exam, and transvaginal ultrasound. Fasting blood samples were obtained to measure reproductive hormones. Urine samples were collected in the evening and upon awakening on 1-2 days and assayed for urinary 6-sulfatoxymelatonin as a proxy for daytime and overnight MEL production, respectively. The night:day (N:D) MEL ratio was determined to assess the rhythm of MEL production. Sleep quality and duration were assessed using the Pittsburgh Sleep Quality Index (PSQI) and via overnight wrist actigraphy. Differences between measures of urinary MEL and sleep quality were analyzed using two tailed t-tests. Associations between diagnostic features of PCOS and sleep-related measures were computed using Pearson partial correlations after adjusting for BMI. Results: No differences were detected in overnight MEL, daytime MEL, or the N:D ratio in women with PCOS versus controls. PCOS group experienced reduced weekend sleep efficiency vs. controls (81.18% vs. 87.76% p<0.05), albeit no differences were detected in PSQI scores, sleep duration or total sleep efficiency determined via wrist actigraphy between groups. Longer menstrual cycle length correlated with poor sleep quality as defined by PSQI (rho=0.3662, p<0.05) and FNPO was positively associated with overnight MEL (rho=0.3586, p<0.05). Conclusions: Day and night MEL production and sleep quality did not differ between women with PCOS and controls despite weekend sleep efficiency being reduced in women with PCOS. Diagnostic features of PCOS were associated with MEL production and sleep disturbances suggesting that women with a more severe clinical presentation of PCOS may be more likely to experience altered MEL production or sleep disturbances. Further studies with a larger sample size are needed to understand the link between degree of symptomology in PCOS, MEL production, and sleep disturbances.