Investigating biomarkers of fruit and vegetable intake in humans

Owen, Elliot Joseph (2024) Investigating biomarkers of fruit and vegetable intake in humans. Doctoral thesis (PhD), Manchester Metropolitan University.

File will be available on: 27 March 2026. Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (5MB)

Abstract

Background: Low fruit and vegetable (F&V) intake has been identified as a risk factor in the development of non-communicable diseases. However, traditional self-reported F&V intake assessments are inherently subject to biases. This thesis aimed to investigate objective methods of F&V intake through dietary biomarkers to ameliorate dietary assessment accuracy and negate self-report associated measurement error. Methods: A literature review appraised a range of F&V intake biomarker candidates and identification techniques. Standardised criteria were sought to evaluate dietary biomarkers. Analysis of the UK National Diet and Nutrition Survey (NDNS) was conducted to identify a concise number of F&Vs that could predict total F&V intake, serving as targets for biomarker discovery. Targeted and untargeted metabolomic techniques were explored to assess individual F&V biomarkers. A multidose intervention with an NDNS prediction model component used untargeted metabolomics to identify and validate novel intake biomarkers. Results: The literature review identified no dose-dependent total F&V intake biomarkers and proposed utilising nutritional metabolomics for biomarker discovery. Validation criteria covering biological plausibility, dose-response, time-response, robustness, reliability, stability, analytical performance, and reproducibility were outlined. A total F&V intake prediction model was derived from NDNS data, comprised of tomatoes, apples, carrots, bananas, pears, strawberries, and onions, with ~95% of model residuals within the limits of agreement. A pilot study used targeted urinary sulforaphane assessments as a broccoli intake biomarker, reporting significant differences from baseline but not between 80 g and 160 g servings, exhibiting limited utility as a biomarker. An untargeted metabolomic approach putatively identified four urinary biomarkers of onion intake that were elevated following consumption, demonstrating a dose-response between 40 g and 80 g or 160 g servings. A model combining these biomarker responses predicted corresponding intakes of 23.9 ± 5.8 g, 86.7 ± 13.3 g, and 121 ± 18.0 g, respectively. Biomarkers can provide objective measurements of individual F&V intakes, comprise a wider panel measuring total F&V intake, or calibrate self-report measurements from epidemiological studies. Conclusions: This thesis provides a thorough examination of objective F&V intake biomarkers. The application of this research represents an opportunity to improve the accuracy of dietary assessments, elucidate diet-disease relationships, and assess public health policy interventions.