Social jetlag and obstructive sleep apnea severity: A retrospective study based on polysomnography

Social jetlag, representing the misalignment between endogenous circadian rhythms and socially imposed sleep schedules, has been associated with various adverse health outcomes. However, its potential relationship with obstructive sleep apnea (OSA) severity, as quantified by the apnea-hypopnea index (AHI), remains unclear.

This retrospective study analyzed data from participants with OSA (AHI ≥5 events/hour) who completed sleep questionnaires and underwent polysomnography at our sleep center. Exclusions included non-OSA, prior OSA treatment, shift work, sleep-affecting medication use, or incomplete data. Social jetlag was defined as the actual difference between mid-sleep on work days and free days, and categorized as minimal (<1 h), moderate (1 to <2 h), and severe (≥2 h). Multiple linear regression assessed the association between social jetlag and AHI, adjusting for age, gender, body mass index (BMI), history of smoking and alcohol, hypertension, hyperlipidemia, and diabetes. Mediation analysis evaluated the Epworth Sleepiness Scale (ESS) as a mediator, with subgroup analyses and generalized additive models (GAM) exploring effect modification and non-linearity.

Among 2383 analyzed participants, severe social jetlag was independently associated with higher AHI (β = 6.90; 95 % CI: 2.18 to 11.61; p = 0.004; p-trend = 0.014), unlike the moderate group (p = 0.389). ESS mediated 19.4 % of this effect (indirect effect β = 0.46; 95 % CI: 0.22 to 0.72; p < 0.001). In a matched cohort analysis stratified by social jetlag severity (n = 637), AHI and ESS rose with social jetlag severity (AHI: 34.9 vs. 39.4 vs. 45.5 events/hour, p < 0.001; ESS: 10.1 vs. 11.2 vs. 11.8, p = 0.015), while habitual sleep duration decreased (7.0 vs. 6.8 vs. 6.6 h; p = 0.013). Stronger associations were observed in males (β = 2.94; 95 % CI: 1.34 to 4.54; p < 0.001; p-interaction <0.001) and smokers (β = 3.31; 95 % CI: 0.90 to 5.71; p = 0.007; p-interaction = 0.003), with significant effects in BMI 24-28 kg/m², non-hypertensive, and non-diabetic subgroups (all p < 0.05). GAM analysis revealed a gradual, non-linear AHI increase with social jetlag below 2 h, accelerating beyond 2 h.

Severe Social jetlag independently increases AHI in individuals with OSA, partially mediated by daytime sleepiness, with pronounced effects in males and smokers. These findings suggest circadian misalignment as a potentially modifiable factor influencing OSA severity.