Circadian Dietary Patterns and Weight Trajectories in Adults Using the Daily24 Mobile Application

Introduction: It is unclear whether the timing of meals relative to sleep impacts cardiometabolic risk factors (e.g., weight gain). This cohort study aimed to detect and describe the association of the timing of caloric intake and dietary patterns (e.g., skipping breakfast) with weight trajectory over 2 years among participants who used the Daily24 mobile application to record their timing of eating and sleeping. Hypothesis: We hypothesized that skipping breakfast or consuming 50% or more of daily calories in the latter half of the day is associated with greater weight gain over 2 years. Methods: We conducted a secondary data analysis from a multi-site, electronic health record-based cohort study of adults from 3 healthcare systems in PCORnet’s PaTH Clinical Research Network. Of the 1017 participants enrolled in the study, N=407 patients downloaded and used the Daily24 mobile application for at least 2 weekdays and 1 weekend day over 6 months. We calculated the median of each participant’s daily eating and sleeping intervals. Participants’ weights were extracted from electronic health records available for up to 10 years prior to until 10 months after baseline. Weight trajectories were calculated over 2 years of this data. Participants completed online surveys about demographics and behavioral characteristics at baseline and after 4 months. We constructed multivariate linear regression models with weight trajectory as the dependent variable and breakfast skipping (i.e., not eating anything within 3 hours of waking), having a medium or large meal within 3 hours of sleep, ≥50% caloric consumption in the latter half of the day (after 12 pm), and ≥90% caloric consumption after 7 pm as independent variables. Additional independent variables included age, sex, race, days of application use, and body mass index (BMI) category. Results: Mean (SD) baseline age was 50.8 (15.0) years, 75.7% had a BMI≥25 kg/m 2 , 78.1% were women, and 84.0% were white. The average number of recorded Daily24 application days was 55.9 (53.5), significantly greater than the 3-day minimum inclusion criterion. The mean of participants’ median intervals from first to last meal was 11.2 (2.0) hours. Although not statistically significant, consuming 50% of daily calories in the latter half of the day and skipping breakfast were associated with an increasing weight trajectory (0.03 and 0.25 kg/year; 95% CI -0.39 to 0.45 and -0.32 to 0.82, respectively). Conclusions: Timing of eating (e.g., skipping breakfast or eating most calories later in the day) may be associated with very small weight increases over time. Our app-based approach may not have had enough observations to establish this association. Further research is needed to understand the clinical importance of circadian dietary patterns related to cardiometabolic health.