Chrono-Cuisine: Investigating Meal Timing Patterns and Cancer Susceptibility

Principal Investigator

Name
Mary Playdon

Degrees
Ph.D., M.P.H., B.Sc.

Institution
The University of Utah

Position Title
Associate Professor

Email
u6017378@umail.utah.edu

About this CDAS Project

Study
IDATA (Learn more about this study)

Project ID
IDATA-97

Initial CDAS Request Approval
Feb 6, 2026

Title
Chrono-Cuisine: Investigating Meal Timing Patterns and Cancer Susceptibility

Summary
SUMMARY. Regulating the timing of meals and snacks to re-align the body’s circadian clock and improve metabolic health is emerging as a promising approach for cancer prevention in early animal and small clinical studies. Yet, a major barrier to studying meal timing and cancer is that large population studies rarely measure meal timing, which makes it impossible to conduct epidemiological studies of meal timing and cancer risk on a large-scale and across individuals with different biological and environmental characteristics and varied meal timing practices. Although obesity and its related metabolic dysregulation are important risk factors for at least 13 cancer types, weight management is notoriously difficult in the long-term. Behavioral strategies are needed that can improve metabolic risk factors for cancer but that do not necessarily rely on weight loss. Herein, we propose to discover and then externally validate novel objective biomarkers of meal timing practices. Our central hypothesis is that meal timing is associated with perturbations in blood metabolomic profile. We will test our hypothesis with unique data from IDATA with validated measures of meal timing and sleep and longitudinal metabolomics data measured on the same metabolomics platform. In Aim 1, we will Identify biomarkers of meal timing patterns in the Interactive Diet and Activity Tracking in AARP (IDATA) Study (n=718). The proposed study will answer critical, outstanding questions about which meal timing practices are associated with cancer-relevant metabolic factors and facilitate large-scale investigations of meal timing and cancer risk at the population level. These meal timing biomarkers could also be used to assess response to meal timing interventions in clinical studies. Following successful completion of this project, we plan to apply the resulting biomarker profiles to study meal timing and cancer risk across international cohorts in the Consortium of Metabolomics Studies (COMETS). Epidemiological research stemming from our study findings will be vitally important prior to issuing public health guidance on meal timing for cancer prevention.

Aims

AIM 1. Identify biomarkers of meal timing in the Interactive Diet and Activity Tracking in AARP (IDATA) Study. We hypothesize that metabolites associated with meal timing will be involved in circadian regulated metabolism, and cancer risk biomarkers. We will identify metabolites associated with ONF duration (primary), late last eating episode, midpoint of fasting time, and number of eating episodes and compute integrated metabolite scores in IDATA (n=718) that collected six 24HR, two 4-day FR, two FFQs and two blood samples over 12-months.
Metabolomics and meal timing data will be cleaned, scaled, and transformed using standardized protocols. Aim 1 The longitudinal metabolite and meal timing measures will be averaged. Regression analyses will examine meal timing variables in relation to each metabolite, separately, adjusting for covariates. Next, machine learning (e.g., Elastic Net) will identify metabolite groups associated with meal timing. We will prioritize metabolites selected into elastic net models with temporal reliability (ICC>0.50).
Secondary analysis. We will test for effect modification by age, sex, race, education, geographic location, urban/rural status, mid-sleep time, and diet quality to (1) identify metabolites robust to participant characteristics; (2) determine if group (e.g., sex) specific meal timing metabolite profiles are warranted. We will compare models ± meal timing x effect modifier interaction terms with likelihood ratio tests.
Pre- and post-menopausal individuals will be included in secondary analyses. Missing data will be reported, excluded in primary analyses, and retained as missing categories in secondary analyses.
Alternative analysis. Rather than averaging repeated meal timing/metabolite variables, we can run mixed effects models to evaluate longitudinal associations. We could adjust for individual diet cancer risk factors instead of HEI. We could also impute missing covariate data with multiple imputation.
Power. In IDATA (n=718), we have 80% power to detect a small effect size of 0.03 at α<0.05.
We expect that shorter ONF duration, and late eating/midpoint of ONF time will be associated with blood metabolites involved in circadian regulated metabolism, including cancer risk biomarkers.

Collaborators

Erikka Loftfield NCI
Mary Playdon Mary Christine Playdon
Sara Salas Mary Christine Playdon