0965 Later Sleep Timing is Related to Worsened Glucose Tolerance in South Asian Youth Likely Secondary to Increased Visceral Fat: The Charisma Study

Abstract Introduction Adolescents have delayed and shorter sleep, which is associated with adverse metabolic consequences. We hypothesized that later sleep phase is associated with worse glucose tolerance in overweight youth. Methods The Charisma study enrolled youth aged 12-21y with BMI>=80%ile (or>=23kg/m2 if>=18y) from three ancestry groups: South Asian (SA), African American (AA), and White (W). Participants completed actigraphy (14d, ActiGraph GT3XP-BTLE), oral glucose tolerance tests, and Hologic DXA scans to measure visceral fat area (VF). We examined the association between mid-sleep time and 2-hour glucose using linear regression in the combined cohort, then examined whether ancestry and VF affected these associations in stratified analyses and interaction models. Results Participants (n=74, 38F) had median(IQR) age 19.1y(17.5, 20.9) and BMI-Z 1.1(0.64, 1.64). The ancestry groups (11 SA, 30 AA, 33 W) had similar demographics, except the SA group was older [SA: 20.7y(19.2, 21.6), AA: 18.4y(16.8, 20.0), W: 19.1y(17.3, 20.9), p=0.02], had higher VF [SA: 101.9cm2(95.2,121.4), AA: 74.9cm2(53.9,97.7), W: 91.0cm2(66.1,114.3), p=0.03] and later mid-sleep time vs W [SA: 5:50a(3:28a,7:23a), AA: 4:29a(3:34a, 5:45a), W:4:34a(3:53a, 5:10a), p=0.04]. In the full cohort, there was no association between mid-sleep time and 2-hour glucose unadjusted or adjusted for age, sex, BMIZ, ancestry, and sleep duration (p=0.4). In stratified analyses, only the SA group had higher 2-hour glucose with later mid-sleep time unadjusted(β=7.7, p=0.049) and adjusted for the same variables (β=13.1, p=0.04), but lost significance if adjusted for VF in place of BMIZ(β=11.1, p=0.06). In the overall cohort, there was a significant interaction between mid-sleep time and VF in predicting 2-hour glucose unadjusted (p=0.003) and adjusted for age, sex, group, and sleep duration (p=0.011), but no significant interactions by ancestry (p=0.3) or BMI (p=0.7) unadjusted. Conclusion In a combined cohort of overweight youth, glucose tolerance was not related to delayed sleep phase. However, SA youth with later sleep timing had worse glucose tolerance. SA youth exhibited higher VF area, and there was significant effect modification of VF on the relationship between sleep timing and glucose tolerance in the overall cohort. These results suggest that higher VF may influence susceptibility to the glycemic effects of delayed sleep phase. Support (if any) NIH R01DK115648; CTSA: CHOP: UL1TR001878, Hopkins: UL1TR001079