Developing a Food Exposure and Urine Sampling Strategy for Dietary Exposure Biomarker Validation in Free-Living Individuals

Abstract

Scope: Dietary choices modulate the risk of chronic diseases and improving diet is a central component of public health strategies. Food-derived metabolites present in urine could provide objective biomarkers of dietary exposure. To assist biomarker validation, this work aims to develop a food intervention strategy mimicking a typical annual diet over a short period of time and assesses urine sampling protocols potentially suitable for future deployment of biomarker technology in free-living populations.

Methods and results: Six different menu plans comprehensively represent a typical UK annual diet that is split into two dietary experimental periods. Free-living adult participants (n = 15 and n = 36, respectively) are provided with all their food, as a series of menu plans, over a period of three consecutive days. Multiple spot urine samples are collected and stored at home.

Conclusion: A successful food exposure strategy is established following a conventional UK eating pattern, which is suitable for biomarker validation in free-living individuals. The urine sampling procedure is acceptable for volunteers and delivered samples suitable for biomarker quantification. The study design provides scope for validation of existing biomarker candidates and potentially for discovery of new biomarker leads, and should help inform the future deployment of biomarker technology for habitual dietary exposure measurement.

Keywords: biomarkers; dietary exposure; free-living; metabolomics; urine sampling.

Figure 1

MAIN study experimental period. Participants were asked to consume a low polyphenol meal (provided or of the participants own choice) in the evening prior to starting each 3‐day menu plan (pre‐day). All foods and drinks for four meals each day for three consecutive days were provided. These foods/drinks were prepared, cooked, and consumed by the participant at home. Participants were asked to collect all urines as separate “spot urines” in specific time‐windows including () first morning void, FMV; (○), fasting; (●) and post‐prandial urines (post‐breakfast, post‐lunch, and post‐dinner) for the duration of the experimental period.

Figure 2

Timings of post‐prandial urine sample collections for 15 participants following menu plans 1–3 over two weeks (expressed as a percent of collections at each time‐point).

Figure 3

Principal component‐linear discriminant analysis (PC‐LDA) scores plots of flow infusion electrospray ionization‐high resolution mass spectrometry (FIE‐HRMS) fingerprints of each urine sample collected on the experimental day and the FMV and fasting urines collected the next morning for 15 participants following menu plan 1 A) positive ionization mode; B) negative ionization mode. PC‐Discriminant Function (PC‐DF) Eigenvalues (T _w values) are given in brackets; x, baseline first morning void; ∆, baseline fasting; , post‐breakfast; *, post‐lunch; ○, bed‐time/post‐dinner; ∇,overnight◇, next day first morning void; +, next day fasting.