Abstract
BACKGROUND AND AIMS: Promising associations have been demonstrated between delayed last eating occasion and cardiorespiratory fitness in adults with heart failure (HF), however, it is unknown if time of eating is associated with clinical endpoints such as mortality. This study aimed to examine associations between time of eating variables and all-cause and cardiovascular mortality in the National Health and Nutrition Examination Survey (NHANES). METHODS AND RESULTS: Participants self-disclosed HF diagnosis. Two dietary recalls were obtained and categorical variables were created based on mean time of first eating occasion (8:31 AM), last eating occasion (7:33 PM) and eating window (11.02 h). Mortality was obtained through linkage to the National Death Index. Covariate-adjusted Cox proportional hazard regression models were created examining the association between time of eating and mortality. Participants (n = 991) were 68 (95 % CI 67-69) years of age, 52.6 (95 % CI 49.0-56.3)% men and had a body mass index of 32.5 (95 % CI 31.8-33.2) kg/m(2) with follow up time of 68.9 (95 % CI 64.8-72.9) person-months. When models were adjusted for time of eating variables and all other covariates, extending the eating window beyond 11.02 h was associated with decreased risk of cardiovascular (HR 0.36 [95 % CI 0.16-0.81]), but not all-cause mortality. Time of first and last eating occasions were not associated with mortality. CONCLUSIONS: In adults with HF, an extended eating window is associated with reduced risk for cardiovascular mortality. Randomized controlled trials should examine if extending the eating window can improve prognostic indicators such as cardiorespiratory fitness in this population.