0826 Chronotype, unhealthy lifestyle, and diabetes risk in middle-aged US women: A prospective cohort study in Nurses’ Health Study II

Abstract

Abstract Introduction Chronotype refers to one’s inclination to earlier or later sleeping times. Late chronotype has been linked to increased diabetes risk, which may be mediated through adherence to unhealthy lifestyle. This study aimed to evaluate the interplay among chronotype, lifestyle factors, and diabetes risk. Methods In 2009, 63,594 diabetes-free women (mean age=54 yrs) from the Nurses’ Health Study II self-reported their chronotype using a validated question from the Morningness-Eveningness Questionnaire and were prospectively followed for incident type 2 diabetes (T2DM) until June 2017. Unhealthy lifestyle was defined by having < 4 of the following 6 healthy lifestyle factors: high diet quality (AHEI≥49), physical activity (≥3.5 MET-hours/week), moderate alcohol intake (≥5 and < 15 g/day), normal body mass index (≥18.5 and < 25 kg/m2), non-current smoking, and 7-9 hours of sleep. T2DM was self-reported and confirmed using a validated supplementary questionnaire. We used modified Poisson regression models to estimate the cross-sectional associations between chronotype and unhealthy lifestyle. Cox proportional hazards models were used to estimate the adjusted hazard ratios (aHR) for incident diabetes risk according to chronotype and assess the proportion of the association explained by lifestyle factors (mediation analyses). Results Participants reporting a “definitely-evening” type were 55% (95% CI: 1.51-1.60) more likely to have an unhealthy lifestyle compared with participants reporting a “definitely-morning” type. We documented 1,921 T2DM cases over 468,555 person-years of follow-up. With increasing eveningness, the risk of T2DM became significantly higher (Ptrend< 0.0001). Compared with the “definitely-morning” type, the aHR (95% CI) for T2DM was 1.10 (0.97-1.24) for “more-morning”, 1.01 (0.81-1.27) for “intermediate”, 1.26 (1.11-1.42) for “more-evening”, and 1.65 (1.44-1.90) for “definitely-evening” chronotypes, after adjusting for sociodemographic factors, shift-work, depression and family history of diabetes. Additional adjustment of unhealthy lifestyle explained 56.8% (95% CI: 39.4%-72.6%) of the association [aHR (95% CI) for “definitely-evening” versus “definitely-morning”: 1.27 (1.10-1.46)]. Conclusion Women with a late chronotype are more likely to engage in unhealthy lifestyle behaviors. Late chronotype is prospectively associated with increased diabetes risk, which is partly explained by unhealthy lifestyle behaviors. Our results suggest that lifestyle interventions may potentially reduce the adverse impact of late chronotype on diabetes risk. Support (if any) R01HL155395, K01HL143034, U01CA176726