LAUSR

This study sought to advance understanding of the impact of habitual sleep patterns on glucose regulation in individuals at risk for type 2 diabetes (T2D). To achieve this aim, we examined associations between a comprehensive panel of sleep parameters and glucose metabolism marker among people with prediabetes (n = 19, age = 60.0y, male = 47.4%) using wearable technology. Briefly, participants underwent fasting plasma glucose (FPG) test and wore Fitbit Ionic band to assess their habitual sleep patterns. Sleep parameters were obtained for a median of 50 days per each participant such as total sleep duration, duration of each sleep stage per night, bed-time, wake-time, etc. To examine associations of sleep parameters with blood glucose levels, a least absolute shrinkage and selection operator (LASSO) regression was used to identify sleep parameters that predict FPG levels with the enhanced prediction accuracy. Mixed effects regression was also performed. In LASSO regression of FPG levels, wake-time (β = −0.013) and percentage of the rapid eye movement (REM) sleep duration out of the total sleep duration (REM%; β = −0.231) were found to inversely predict FPG levels among participants. Mixed effects regression model also showed that REM% is inversely associated with FPG levels (R2 = 0.61; β = −1.57, P = 0.058) after adjusting other covariates. In sum, people with prediabetes who have earlier wake-time and shorter REM proportion have shown higher FPG levels. Overall, these findings suggest that habitual sleep patterns may influence physiologic defect underlying dysglycemia and progression to T2D in individuals with prediabetes. Precision Health and Integrated Diagnostic (PHIND) Center at Stanford University.