Padrões crononutricionais e suas associações com desfechos metabólicos: resultados do National Health and Nutrition Examination Survey 2003–2018

Abstract

Introduction: Fasting duration has been widely investigated in relation to metabolism, weight loss, and chronic diseases. However, population-based studies describing its distribution and variation across sociodemographic subgroups remain limited. In addition, the timing and regularity of food intake influence circadian rhythms and metabolism and have been associated with various health outcomes such as obesity and metabolic syndrome. Long-term population based evidence integrating fasting patterns and eating timing remains scarce. Objectives: : To describe fasting patterns over a 16-year period among adults and elderly according to sociodemographic characteristics (manuscript 1). The second objective was to evaluate the association between chrononutrition patterns, defined by the timing of the first and last eating episodes, and metabolic outcomes in adults and elderly (manuscript 2). Materials and Methods: A total of 32,053 (manuscript 1) and 7,895 participants (manuscript 2) from eight cycles of the National Health and Nutrition Examination Survey (NHANES) (2003-2018) were analyzed. Information on fasting duration and eating timing was obtained from two 24-hour dietary recalls. The first eating episode was defined as the first intake of energy-containing foods or beverages recorded after 05:00, and the last eating episode corresponded to the final caloric intake of the day. Fasting patterns were defined as the interval between the last and first eating episodes, including fasting duration, fasting midpoint, and night fasting duration (18:00 6:00). Additional chrononutrition variables included eating duration, eating midpoint, timing of the first and last eating episodes, number of eating occasions, and total energy intake (manuscript 1). For manuscript 2, a chrononutrition pattern was defined based on the population median timing of the first and last eating episodes, categorized into earlier and later intake patterns. Associations were assessed using linear, logistic, and multinomial regression models adjusted for potential confounders in both manuscripts. In manuscript 2, Poisson regression was used to examine the association between chrononutrition patterns and metabolic outcomes, including fasting glucose, triglycerides, HDL cholesterol, obesity, abdominal obesity, and metabolic syndrome. Results: Fasting patterns differed across sociodemographic subgroups (p < 0.001), with later patterns more common among younger adults and individuals with lower educational attainment and income, and earlier patterns among elderly individuals. Differences in fasting duration and night fasting duration (18:00-6:00) were also observed across subgroups (p < 0.001), and findings from stratified analyses were consistent with those from the overall sample. Over the 16-year period, modest changes in fasting patterns were observed, including slight increases in night fasting duration (18:00–6:00) (p = 0.026), earlier timing of the last eating episode (p = 0.048), fewer eating occasions (p = 0.028), and lower total energy intake (p = 0.003) (manuscript 1). In manuscript 2, later eating patterns were associated with a higher prevalence of elevated triglycerides (PR: 1.23; 95% CI: 1.05-1.46) and elevated fasting glucose (PR: 1.13; 95% CI: 1.03-1.24; PR: 1.16; 95% CI: 1.05-1.28), compared with earlier eating patterns. Conclusion: Later fasting patterns were more common among younger adults and individuals with lower educational attainment and income, whereas earlier patterns were more frequent among elderly. Over time, only modest changes were observed in fasting patterns and other chrononutrition variables (manuscript 1). Additionally, later eating patterns were associated with worse metabolic outcomes (manuscript 2). These findings contribute to the understanding of fasting patterns across population subgroups and highlight the role of eating timing as a relevant factor in the understanding and potential prevention of metabolic disorders at the population level.