The complex association between sleep duration and mental and physical health: are we missing circadian preferences?
Non-linear associations between sleep duration and health outcomes are well-documented, but the role of circadian timing in these relationships remains scantly explored. This study examined linear and non-linear associations between self-reported sleep duration and mental and physical health-related quality of life in Italian university students, accounting for the moderating effect of bedtime.
A total of 1,234 students (mean age 23.3 ± 2.5 years; 87.3% women) responded to an online survey that included standardized measures of sleep quality, sleep timing, and health-related quality of life (HRQoL). Multiple regression models including quadratic and cubic terms and interactions were used to investigate linear and non-linear associations between sleep and circadian timing preferences and mental and physical HRQoL.
Later bedtimes were independently associated with lower mental HRQoL (β = -0.10, p < .001). In early sleepers, mental HRQoL showed an inverted U-shaped association with sleep duration, with poorer outcomes at short and long durations. For physical HRQoL, a U-shaped quadratic effect of sleep duration was observed (β = -0.10, p = .007), independent of bedtime.
These preliminary findings support the integration of circadian variables into sleep-health models using accessible self-report measures.
A total of 1,234 students (mean age 23.3 ± 2.5 years; 87.3% women) responded to an online survey that included standardized measures of sleep quality, sleep timing, and health-related quality of life (HRQoL). Multiple regression models including quadratic and cubic terms and interactions were used to investigate linear and non-linear associations between sleep and circadian timing preferences and mental and physical HRQoL.
Later bedtimes were independently associated with lower mental HRQoL (β = -0.10, p < .001). In early sleepers, mental HRQoL showed an inverted U-shaped association with sleep duration, with poorer outcomes at short and long durations. For physical HRQoL, a U-shaped quadratic effect of sleep duration was observed (β = -0.10, p = .007), independent of bedtime.
These preliminary findings support the integration of circadian variables into sleep-health models using accessible self-report measures.