Study Design
Validation of a 24-hour timing grid to assess weekday and weekend sleep in the Cancer Prevention Study-3 Charlie Zhong* Charlie Zhong Matthew Masters Sidney M Donzella Alpa V Patel Marjorie McCullough
Sleep misalignment (difference in weekday vs weekend sleep patterns) is increasingly being implicated in higher risk of mortality and morbidity. Several prospective cohorts have assessed sleep in their populations, but typically assess overall sleep or utilize clinic-based instruments (e.g., Pittsburgh Sleep Quality Index). To better understand sleep misalignment, we validated a 24-hour grid-based sleep assessment tool. In 2015, participants of the American Cancer Society Cancer Prevention Study-3 (CPS-3) indicated on a grid consisting of one-hour blocks, when they were eating or sleeping, encompassing all 24-hours of the weekday and weekend. A subset of 170 participants were also enrolled in the Activity Validation SubStudy (AVSS), a validation study of activity patterns and provided a Daysimeter – a device designed to capture light exposure during the day and sleep at night. We utilized the method of triads to calculate validity coefficients (VCs) and bootstrapped 95% CI for sleep duration between the grid, device, and sleep diaries; and correlation coefficient (Kendall’s Tau) for chronotype between grid and questionnaire. Among 134 participants with data for at least 3 weekdays and 1 weekend, average weekday sleep duration was longer on the grid (459 minutes) than for device or diary (409 and 405 min, respectively). Similar patterns were seen for the weekends (517 min grid, 441 min device, 433 min diary). Weekday grid sleep duration exhibited good validity with a VC of 0.68 (95% CI 0.54, 0.79) compared to weekend (VC 0.55, 95% CI 0.39, 0.69). On average, midpoints (clock time between going to sleep and waking) from the grid were 15.3 min later than diary on weekdays, and 10.7 min later on weekends. The correlation coefficient for chronotype was 0.52. The grid over-estimated sleep duration compared to device and sleep diary due to the coarser nature of the hour-long grids, but still provided a succinct and valid method to assess weekday and weekend sleep patterns.
