Abstract
OBJECTIVE: To investigate the association of timing, frequency, and food quality of night eating with all-cause, cancer, and diabetes mortality. METHODS: This study included 41,744 participants from the US National Health and Nutrition Examination Survey (2002-2018). Night eating information was collected by 24-h dietary recall and the exposures were timing, frequency, and food quality of night eating. Food quality was assessed by latent class analysis. The outcomes were all-cause, cancer, and diabetes mortality, which were identified by the National Death Index and the International Classification of Diseases 10th Revision. Adjusted hazard ratios [aHR] with 95% confidence intervals [CI] were computed by Cox regression. RESULTS: During a median follow-up of 8.7 years, 6066 deaths were documented, including 1381 from cancer and 206 from diabetes. Compared with no night eating (eating before 22:00), the later timing of night eating was associated with higher risk of all-cause and diabetes mortality (each P-trend <0.05) rather than cancer mortality, with the highest risk of eating being 00:00-1:00 (aHR 1.38, 95% CI 1.02-1.88) and being 23:00-00:00 (aHR 2.31, 95% CI 1.21-4.40), respectively. However, the increased risks were not observed for 22:00-23:00. Likewise, one time or over frequency of night eating was associated with higher all-cause and diabetes mortality (each P < 0.05). That risks were further observed in high-dietary-energy-density group of night eating (all-cause mortality: aHR 1.21 [95% CI 1.06-1.38]; diabetes mortality: aHR 1.97 [95% CI 1.13-3.45]), but not in low-dietary-energy-density group. Finally, correlation analysis found positive associations of night eating with glycohemoglobin, fasting glucose, and OGTT. CONCLUSIONS: Night eating was associated with increased all-cause, cancer and diabetes mortality; however, reduction of excess mortality risk was observed when eating before 23:00 or low-dietary-energy-density foods.