Pathways to obesity

: investigating the role of modifiable lifestyle factors through intermediate phenotypes

Abstract

The causal pathway between modifiable lifestyle factors and obesity is complex. The growing obesity epidemic impacting across the lifecourse is a major public health concern in many countries, therefore understanding the causes and consequences of adiposity is important. In this thesis I investigate the role of the metabolome and methylome in the relationship between dietary behaviour and obesity.

Establishing causality in observational studies is challenging due to unmeasured confounders and potential for reverse causation. I use Mendelian randomization (MR), two-sample MR and longitudinal analysis to infer causality in the relationships between diet, the methylome, the metabolome and body mass index (BMI). A good understanding of causality in these relationships is important to address the question of how the major public health problem of obesity should be tackled.

Dietary behaviour is a complex trait, and hence few studies have identified genetic variants associated with diet. I performed a GWAS of macronutrient intake in UK Biobank, with the aim of identifying genetic variants that could be used to generate a robust genetic instrument for dietary behaviour for use in MR.

Whilst studies have demonstrated the effect of adiposity on metabolic signatures from early adulthood onwards, there is a lack of published data exploring the relationship between adiposity and the metabolome in childhood. Using the Avon Longitudinal Study of Parents and Children (ALSPAC), I observed strong evidence of associations between BMI and several metabolite measures in childhood and adolescence in children, showing that the ability of BMI to influence the metabolome starts in childhood. Many BMI-associated metabolites are also associated with dietary behaviour, so it is likely that the metabolome plays a key role when trying to understand the relationship between dietary behaviour and BMI.

Several associations have been observed between BMI and methylation, mostly in adults. I investigated the relationship between BMI and methylation in childhood and adolescence and explored their relationship with dietary behaviour. My observations corroborated the prevailing evidence that DNA methylation occurs as a consequence (rather than a cause) of BMI.

Date of Award	25 Sep 2018
Original language	English
Awarding Institution	The University of Bristol
Supervisor	Tom R Gaunt (Supervisor) & Caroline L Relton (Supervisor)