Time of day and eating behaviors are associated with the composition and function of the human gastrointestinal microbiota

Background: Preclinical research has shown that the gastrointestinal microbiota exhibits circadian rhythms and that the timing of food consumption can affect the composition and function of gut microbes. However, there is a dearth of knowledge on these relations in humans. Objective: We aimed to determine whether human gastrointestinal microbes and bacterial metabolites were associated with time of day or behavioral factors, including eating frequency, percentage of energy consumed early in the day, and overnight-fast duration. Design: We analyzed 77 fecal samples collected from 28 healthy men and women. Fecal DNA was extracted and sequenced to determine the relative abundances of bacterial operational taxonomic units (OTUs). Gas chromatography-mass spectroscopy was used to assess short-chain fatty acid concentrations. Eating frequency, percentage of energy consumed before 1400, and overnight-fast duration were determined from dietary records. Data were analyzed by linear mixed models or generalized linear mixed models, which controlled for fiber intake, sex, age, body mass index, and repeated sampling within each participant. Each OTU and metabolite were tested as the outcome in a separate model. Results: Acetate, propionate, and butyrate concentrations decreased throughout the day (P = 0.006, 0.04, and 0.002, respectively). Thirty-five percent of bacterial OTUs were associated with time. In addition, relations were observed between gut microbes and eating behaviors, including eating frequency, early energy consumption, and overnight-fast duration. Conclusions: These results indicate that the human gastrointestinal microbiota composition and function vary throughout the day, which may be related to the circadian biology of the human body, the microbial community itself, or human eating behaviors. Behavioral factors, including timing of eating and overnight-fast duration, were also predictive of bacterial abundances. Longitudinal intervention studies are needed to determine causality of these biological and behavioral relations. This trial was registered at clinicaltrials.gov as NCT01925560.