Abstract Introduction Sleep quality and circadian rhythm are both linked to emotional well-being: poorer sleep quality and an evening-type circadian rhythm each relate to higher negative affect and stress, and lower positive affect. However, most such studies examine only one pathway at a time and do not parcel out the unique associations of each path. Circadian rhythm is a related but separate construct in sleep research, but its unique contribution to affect and stress after controlling for shared variance with sleep quality remains understudied. Examining these constructs within a single analytical framework clarifies their unique associations. Methods We collected online survey responses from undergraduates using the Pittsburgh Sleep Quality Index (PSQI), Morningness-Eveningness Questionnaire (MEQ), Positive and Negative Affective Schedule (PANAS), and Perceived Stress Scale (PSS). The final sample included 720 participants (mean age = 21.1, 67% female). We conducted structural equation modeling (SEM) with sleep quality and circadian rhythm as predictors and positive affect (PA), negative affect (NA), and perceived stress (PS) as outcomes. We used item-to-construct balancing (using item factor loadings from separate confirmatory factor analyses CFA) to create item parcels as indicators for the latent variables PA (from the PANAS PA subscale), NA (from the PANAS NA subscale), PS (from PSS), and circadian rhythm (from MEQ). We used the PSQI subscales as indicators for the latent variable sleep quality. Results SEM analysis revealed that sleep quality was significantly associated with PA (β = -.300, p .001), NA (β = .431, p .001), and PS (β = .424, p .001). Circadian rhythm was significantly associated with PA (β = .168, p .001) and PS (β = -.132, p = .005) but not NA (β = -.061, p = .218). Analyses of CFA models demonstrating discriminant validity will also be included in the poster. Conclusion Poorer sleep quality uniquely predicted lower PA, higher NA, and higher PS. Although evening-type circadian rhythm predicted lower PA and higher PS, its associations were weaker after controlling for shared variance with sleep quality, and it was not associated with NA. Our findings suggest the relative importance of quality over chronotype differences. Support (if any)
Kaiqi Zhang (Fri,) studied this question.