Perfect Timing: Associations Between Dietary Timing, Eating Intervals and Metabolic Health

Abstract

Abstract Background: Metabolic disorders are on the rise resulting in an increased need for novel interventions to combat this growing concern. One such intervention is time-restricted eating, which consolidates caloric intake to a shortened eating duration, and has consistently demonstrated improvement in metabolic health. Additional interventions could harness circadian clocks, which help regulate daily rhythms of hormone release and maintain metabolic homeostasis. Circadian rhythms are influenced by timing of food intake and therefore meal timing may also impact metabolic health. Objective: To examine whether timing of eating was associated with metabolic health independently of eating duration. Methods: Data are from the National Health and Nutrition Examination Survey (NHANES), a nationally representative U.S. survey and physical exam. We analyzed two non-consecutive dietary recalls, fasting glucose and insulin from 10,575 adults (>18) over four cycles (2005–2012). We calculated eating interval duration as the time between first and last eating occasion (>10kcal) and formulated three groups: <10h, 10-13h, >13h. We then created 6 subgroups based on eating duration start time (before or after 0830 h) to analyze associations of eating interval timing. Linear regression analyses controlling for demographics were computed to determine whether eating interval duration and eating interval timing were associated with fasting glucose and estimated insulin resistance using HOMA-IR. Results: Fasting glucose did not differ significantly among eating interval groups. Adjusted mean HOMA-IR increased with shorter eating interval duration (2.23, 2.20, 1.97 for <10h, 10-13h, >13h; p<0.05). When eating start time occurred before 0830 h compared to start time after 0830 h, mean adjusted HOMA-IR decreased across all eating interval subgroups (1.96 vs 2.28 for <10h, 2.13 vs 2.28 for 10-13h, 1.94 vs 2.13 for >13h; p <0.05). Adjusted mean fasting glucose had a similar pattern (96.3 mg/dL vs 98.7 mg/dL for <10h, 97.3 mg/dL vs 98.4 mg/dL for 10-13h, and 97.3 mg/dL vs 99.3 mg/dL for >13h; p < 0.05). In regression models where eating duration and timing were both entered as continuous variables, only timing was significantly associated with fasting glucose and HOMA-IR. Later time was associated with higher glucose and HOMA values (p<.001) Conclusion: Shorter eating durations, i.e. time-restricted eating, was associated with worse metabolic outcomes, except when paired with an earlier start time. All subgroups with an early eating start time had better metabolic outcomes regardless of eating duration. These findings suggest that timing is more strongly associated with metabolic measures than duration and supports early eating strategies.