0327 WEEKNIGHT VERSUS WEEKEND DIFFERENCES IN TOTAL SLEEP TIME ACROSS SLEEP GROUPS

Abstract Introduction: Americans exhibit discrepant sleep on weekends compared to weeknights. However, research on this phenomenon has been limited to healthy sleepers. This study utilized multilevel modeling to characterize differences in weeknight-weekend total sleep time (TST) among four sleep groups: 1) insomniacs (INS), 2) noncomplaining poor sleepers (NP), 3) complaining good sleepers (CG), and 4) healthy sleepers. Methods: We recruited community-dwelling adults ages 20 to 98. This study analyzed 14 days of sleep diary data from 633 participants classified into 4 sleep groups based on quantitative sleep criteria and subjective sleep complaint. Data was modeled according to a two-level repeated-measures structure with time points (level 1) nested within individuals (level 2) predicting TST. To examine intraindividual variability in TST, we estimated a random-intercept model with no predictors (Model 1). Next, dummy-coded day-type (weekday=0, weekend=1) was entered as a level-1 predictor (Model 2). Lastly, significant demographic covariates and dummy-coded sleep groups were entered at level 2 to examine weeknight-weekend differences in TST among sleep groups (Model 3). Results: Model 1 revealed that 61% of the variance in TST was attributable to intraindividual differences across time. When day-type was added in Model 2, it accounted for 6.3% of the intraindividual variance in TST (p < .001). Across participants, the average TST on weeknights was 6.9 hours, and TST during weekend nights was 0.32 hours longer than during weeknights (p < .001). For Model 3, compared to healthy sleepers, CG’s average TST was 0.31 hours shorter (p < .01) and NP’s average TST was 0.26 hours shorter (p < .01). When comparing INS to healthy sleepers, there was a significant interaction of day-type by sleep group. Compared to healthy sleepers, INS’s TST was shorter, and they exhibited less weekend-weeknight discrepancy (p < .01). Conclusion: Compared to healthy sleepers, CG, NP, and INS exhibited shorter TSTs, and INS exhibited less weeknight-weekend discrepancy in TST. Consideration of day-type and sleep group is important for better characterization of sleep behavior. Future studies will examine the effect of day-type and sleep group on other sleep parameters. Support (If Any): Research supported by NIA grants AG12136 and AG14738.