The association between food intake timing and the risk of obstructive sleep apnoea: a cross-sectional analysis

Obesity is a key risk factor contributing to obstructive sleep apnoea (OSA) and its severity. Late or mistimed eating potentiates the development of obesity, with circadian misalignment as a putative mechanism compromising metabolism. OSA severity and poor sleep quality may also be associated with later meal timings. This cross-sectional analysis investigated the association between intake timing patterns and OSA risk. Data from adult participants in the National Health and Nutrition Examination Survey (NHANES) 2017–2020 cycle were analysed. OSA risk was computed as a composite score using the STOP-BA(N)G scoring system. Intake timing variables were derived from 24-hour dietary recalls, which include the eating window (total duration between the first to last intake), timing of highest energy intake (time in which the highest energy intake occurred), caloric midpoint (timing in which 50% of total energy intake was first attained), and the duration between the last intake to sleep onset. Circadian timings of first intake, last intake and caloric midpoint were derived relative to individual chronotype (time difference between original intake timings and the midpoint of sleep on work-free days). All intake timing variables were classified into tertiles (T1-3), whereby the first tertile (T1; reference group) corresponds to the shortest eating windows, or the earliest intake timings in chronological order. Weighted sample multivariate logistic regression was used to assess the associations of intake timing variables with OSA risk. Mediation analysis with daily energy intake as a potential mediator was performed. Results are presented as odds ratio with 95% confidence intervals. Within the weighted sample, 42.4% were identified with high OSA risk. Compared to shorter eating windows or earlier intake timings (T1), greater odds of high OSA risk was observed in the tertiles with longer eating windows (T1 vs. T2: 1.31 (1.10, 1.58); T1 vs. T3: 1.60 (1.33, 1.93)), later timing of highest energy intake (T1 vs. T3: 1.29 (1.06, 1.58)), later caloric midpoint (T1 vs. T3: 1.48 (1.18, 1.85)), later last timing of intake (T1 vs. T3: 1.39 (1.15, 1.67)), intake closer to sleep time (T1 vs. T3: 1.33 (1.06, 1.67)), and relative to individual chronotype (T1 vs. T2: 1.45 (1.13, 1.86); T3: 1.58 (1.26, 1.99)). Significant pure indirect effects through daily energy intake were observed across the eating window (T1 vs. T2: 0.06 (0.03, 0.09); T3: 0.13 (0.06, 0.20)), last timing of intake (T1 vs. T3: 0.12 (0.07, 0.18)), timing of highest energy intake (T1 vs. T3: 0.04 (0.01, 0.06)), caloric midpoint (T1 vs. T2: 0.02 (0.01, 0.03); T1 vs. T3: 0.03 (0.01, 0.05)), duration between last intake to sleep onset (T1 vs. T3: 0.09 (0.04, 0.14)), circadian timing of last intake (T1 vs. T2: 0.03 (0.01, 0.06)) and circadian timing of caloric midpoint (T1 vs. T3: 0.02 (0.00, 0.03)), although not all associations showed significant overall mediation effects ( p > 0.05). There were no significant associations observed across other intake timing variables or tertiles assessed. Individuals with longer eating windows and later intake timings had a greater odds of high OSA risk, likely influenced by higher energy intake across an extended daily window. These findings support the need for longitudinal studies assessing the applications of timed dietary interventions to potentially mitigate health risks in populations with/at risk of OSA.