Skip to Main Content

COVID-19 is an emerging, rapidly evolving situation.

What people with cancer should know:

Get the latest public health information from CDC:

Get the latest research information from NIH:

Principal Investigator
Erikka Loftfield
Ph.D., M.P.H.
National Cancer Institute
Position Title
Research Fellow (transitioning to Stadtman Tenure Track Investigator)
About this CDAS Project
IDATA (Learn more about this study)
Project ID
Initial CDAS Request Approval
Dec 11, 2020
Dietary biomarker development in the Interactive Diet and Activity Tracking in AARP Study
Metabolomics has tremendous potential to identify biomarkers that could improve dietary assessment, clarify contributions of diet to cancer and other chronic diseases, and identify underlying mechanisms. An increasing number of cross-sectional studies have explored the associations of dietary patterns, food group, food items, and food components, typically estimated using food frequency questionnaires (FFQs), with metabolites using an untargeted metabolomics approach. These studies have identified potential novel biomarkers of exposure (e.g. trigonelline for coffee) and response (e.g., 2-hydroxy-3-methylbutyric acid for alcohol) to habitual dietary intake as well as strong associations between prediagnostic levels of diet-related metabolites and disease outcomes. However, prior studies have not addressed important sources of measurement error in either dietary intake or metabolite data. Therefore, we propose a rigorous dietary biomarker discovery study that 1) combines data from multiple dietary assessment methods, using the NCI method, to account for measurement error and estimate usual dietary intake and 2) uses samples collected from the same individual six months apart to account for temporal variability in metabolite levels. To identify candidate biomarkers of a priori dietary factors including, dietary pattern metrics (e.g., HEI for diet quality and NOVA for ultra-processed diet), food groups (e.g., meat, fish and grains), food items (e.g., coffee), and food components (e.g., fiber and macronutrients), we will conduct an untargeted metabolomics analysis of blood and urine samples within the iData Study. In brief, we will include all participants who provided baseline and six-month serum, 24-hour urine, and first morning void (FMV) urine samples (n~800). We will use the Metabolon® untargeted platform to measure more than 1000 small molecules (<1 kilodalton), which are collectively referred to as “metabolites” but include both parent compounds and their downstream metabolic products. This platform employs ultra-high-performance liquid chromatography (UPLC) with tandem mass spectrometry (MS/MS) to separate compounds and measure their spectral peaks; it covers a broad range of metabolites and metabolic pathways including endogenously derived amino acids, carbohydrates, lipids, cofactors and vitamins, intermediates of energy metabolism as well as xenobiotics derived from exogenous sources such as food or drugs. We will use LASSO regression to identify metabolites in serum and urine samples that are associated with a given dietary factor. We will also calculate intraclass correlation coefficients (ICC) to estimate the intra-individual variability over six months for serum and urinary metabolites. These six-month ICCs will be used to estimate the number of cases required for nested case-control studies within established cohorts with biobanked blood or urine samples and to inform biospecimen collections in future prospective cohort studies. We will also conduct analyses exploring metabolite associations with the oral microbiota as well as potential diet-metabolite-microbiome interactions. Finally, this complete case metabolomics analysis of serum and urine samples will be a valuable resource for the broader scientific community that can be used to study other exposures (e.g., obesity and physical activity) that were assessed as a part of the iData Study.

In this study, we aim to:

• Identify robust biomarkers of usual diet using serum, 24-hour urine, and spot urine (FMV) collected six months apart from the participants in the iDATA study who have detailed dietary data collected by FFQ, 24-hour dietary recall, and food records;
• Determine six-month intra-individual variability across all measured metabolites to better estimate sample size for epidemiological studies with and without serial biospecimen collections;
• Ascertain whether metabolite measures in spot urine, which is more easily collected in large population-based studies, are reflective of those in 24-hour urine;
• Evaluate associations of serum and urinary metabolites with the oral microbiota.

The proposed analyses are expected to discover novel biomarkers of usual diet that will inform future targeted analyses of candidate dietary biomarkers and advance our understanding of the role of diet in cancer etiology.


Erikka Loftfield (National Cancer Institute)
Rashmi Sinha (National Cancer Institute)
Neal Freedman (National Cancer Institute)
Josh Sampson (National Cancer Institute)
Emily Vogtmann (National Cancer Institute)