ABSTRACT
Objective
To explore the direct and indirect pathways from evening diet high in ultra-processed foods, caffeine and/or alcohol to sleep quality, including the roles of evening latency and individual factors.
Methods
The participants (n = 2050; 18–65 y) were part of a virtual population-based study. Sleep quality was assessed using a composite score incorporating sleep duration, nocturnal awakenings, sleep latency, insomnia and self-reported sleep quality, with higher scores indicating poorer sleep quality. A structural equation model was used to test the direct and indirect effects of an evening diet high in ultra-processed foods, caffeine and/or alcohol (sleep-disturbing diet), evening latency (time between the last eating event and sleep onset), screen time before bedtime, age, physical activity, marital status and body mass index (BMI) on sleep quality score.
Results
For the direct effect, an increase in the sleep quality score was observed for an evening sleep-disturbing diet (DE: 0.189; 95% CI 0.14; 0.24; p < 0.001), age (DE: 0.007; 95% CI 0.00; 0.01; p < 0.001), BMI (DE: 0.013; 95% CI 0.01; 0.02; p < 0.001) and screen use before bedtime (DE: 0.001; 95% CI 0.00; 0.00; p < 0.001). An evening latency ≤ 2 h showed a decrease in sleep quality scores (DE: −0.126; 95% CI −0.17; −0.08; p < 0.001), an effect attenuated by the negative association mediated by the indirect effect by evening sleep-disturbing diet (IE: 0.013; 95% CI: 0.00; 0.02; p < 0.001).
Conclusion
By demonstrating that a shorter evening latency—especially sleep-disturbing diet—along with longer screen exposure before bed and higher BMI, was negatively associated with sleep quality, our findings suggest that reducing such food intake, managing screen time, particularly close to bedtime, and maintaining a healthy BMI to enhance sleep quality.
Journal of Human Nutrition and Dietetics, Volume 38, Issue 6, December 2025. Read More