Association of Night Eating Syndrome Risk with Sleep Disturbance in Medical Students: A Cross-Sectional Study
TL;DR
Sleep disturbance and NES risk are prevalent and cross-sectionally associated in Turkish medical students and poor sleep quality was the strongest independent correlate of night eating scores.
Abstract
Background: Medical students are vulnerable to sleep disturbance and nocturnal eating behaviors that may jointly increase psychiatric and metabolic morbidity. Comprehensive sleep-eating evaluations in Turkish medical students remain limited. To determine the prevalence of poor sleep quality and night eating syndrome (NES) risk and to examine their cross-sectional associations with insomnia, anxiety, emotional eating, restless legs symptoms, chronotype, and lifestyle factors. Methods: This cross-sectional study, conducted from March to December 2023, included 265 medical students who completed validated instruments assessing sleep quality (Pittsburgh Sleep Quality Index [PSQI]), night eating syndrome (NES) risk (a score of ≥25 indicating high risk), insomnia severity, daytime sleepiness (Epworth Sleepiness Scale [ESS]), anxiety, emotional eating, Restless Legs Syndrome (RLS) symptoms, and chronotype. Group comparisons, Spearman’s correlation analyses, and multiple linear regression were conducted. Results: Poor sleep quality (PSQI > 5) affected 50.2% of participants, and high NES risk affected 14.7%. PSQI correlated with NEQ (r = 0.301; p < 0.001), insomnia severity (r = 0.538), anxiety, and emotional eating. Mean NEQ was higher in the poor-sleep group (18.67 ± 8.00 vs. 15.38 ± 6.36; p < 0.001; Cohen’s d = 0.455). Females had higher PSQI and ESS scores, anxiety, and emotional eating scores; high NES risk was numerically more frequent in males (18.6% vs. 11.8%; p = 0.175). RLS symptoms (score > 0) were present in 34.7% (mild 22.6%; moderate 7.9%; severe 4.2%). In the primary multivariable model (PSQI and sex), PSQI was the strongest independent correlate of NEQ (standardized β = 0.325; p < 0.001); sensitivity analyses showed stable PSQI estimates after additional covariate adjustment, whereas sex coefficients were non-significant and unstable across models. A secondary model excluding PSQI identified alcohol use as a significant lifestyle correlate (standardized β = 0.167; p = 0.014). R2 = 0.148 for the full covariate model and R2 = 0.108 for the primary model. Conclusions: Sleep disturbance and NES risk are prevalent and cross-sectionally associated in Turkish medical students. Poor sleep quality was the strongest independent correlate of night eating scores. Integrated campus screening targeting sleep, eating behavior, anxiety, and comorbid sleep symptoms may facilitate prevention and referral.