Sleep Characteristics and Mental Health Among Kuwait University Students: Sensitivity Analyses for Item Overlap

Sleep is a potentially modifiable correlate of student mental health, but associations may be inflated by content overlap between sleep and mental health measures. We examined four self-reported sleep characteristics in relation to depressive symptoms and well-being, with sensitivity analyses reducing item overlap.

In this cross-sectional study, questionnaire data were collected from 389 Kuwait University students recruited by convenience from students who were concurrently participating in a campus-based physical-assessment study. Depressive symptoms used a nine-item PHQ-9-based measure (items 3, 9 modified); well-being used five WHO-5 items with modified frequency anchors. Associations used multivariable linear regression with sequential covariate adjustment.

For PHQ-9-based scores, independent associations were found for better sleep quality (B = -1.22, 95% CI -1.84 to -0.60; β = -0.20), more frequent disturbance (1.11, 0.49 to 1.73; β = 0.18), longer sleep duration (-0.48, -0.74 to -0.21; β = -0.17), and longer latency (0.54, 0.08 to 0.99; β = 0.11). Quality (4.01, 1.72 to 6.30; β = 0.19) and duration (1.35, 0.37 to 2.33; β = 0.14) were associated with well-being. No statistically significant sleep-by-sex interactions were detected; latency was not robust to item-3 removal.

These findings support considering sleep health within university mental health initiatives; longitudinal studies are needed to establish directionality and intervention effects.