Uncovering the role of food processing and adiposity in the rising incidence of upper-aerodigestive tract cancer

Abstract

In this thesis, I explored the associations between ultra-processed food (UPF) intake, adiposity and upper-aerodigestive tract cancers using multiple analytical methodologies to understand whether the rising incidence of upper-aerodigestive tract cancers could be attributed to UPF consumption.

In Chapter 1, I contextualised the work conducted in this thesis by providing an overview of the epidemiological landscape of upper-aerodigestive tract cancers, as well as a summary of the evidence on the health impacts of UPFs. Chapter 2 describes the key methods and datasets used in this thesis.

In Chapter 3, I reexamined the association between UPF intake and upper-aerodigestive tract cancers in a large prospective cohort study using conventional epidemiological methods. My findings were consistent with previously published findings obtained using the same dataset, suggesting UPF consumption was positively associated with head and neck cancer (HNC) and oesophageal adenocarcinoma (OAC). However, through the use of a negative control analysis, I uncovered bias in the association between UPF consumption and cancer risk.

Because the negative control analysis did not prove unequivocally that the main associations observed between UPF consumption and both HNC and OAC are completely biased, I further explored the mediating role of adiposity in these associations. In Chapter 4, I estimated the mediated effect of UPF consumption on upper-aerodigestive tract cancer risk via adiposity using a counterfactual framework, under the assumption that the observed associations between UPF consumption, adiposity and upper-aerodigestive tract cancers were causal. I found that adiposity did not play a major role in the associations between UPF consumption and HNC/OAC risk.

In Chapter 5, I reassessed the link between adiposity and HNC, while accounting for smoking behaviour, using a two-sample Mendelian randomization (MR) framework. I showed that conventional observational findings of an association between adiposity and HNC were likely explained by residual confounding.

In Chapter 6, I conducted a genome-wide association analysis to provide clues regarding the biology involved in food choices and the mechanisms underlying the association observed between UPF consumption and upper-aerodigestive tract cancer risk. I found eight independent genetic variants associated with the consumption of foods by their degree of processing.

Finally, in Chapter 7, I provided an overview of the main findings of the thesis, contextualised them within the broader body of literature, and reflected on how the different chapters connect to one another. I also discussed the overarching strengths, limitations, implications and future directions of this thesis as a whole.

I concluded that although this thesis provided limited support of an association between UPF consumption and upper-aerodigestive tract cancer risk, there is still value in strengthening existing health policies, particularly those aimed at reducing population inequalities and addressing the social determinants of food choice and consumption.

Date of Award	20 Jan 2026
Original language	English
Awarding Institution	University of Bristol
Supervisor	Maria C Borges (Supervisor), Rebecca Richmond (Supervisor), Caroline L Relton (Supervisor), George Davey Smith (Supervisor) & Tom G Richardson (Supervisor)