Gene-environment interaction in the development of obesity

The prevalence of obesity has risen sharply over the past five decades. Explanations for the obesity epidemic have focused on changes to the physical, technological, and cultural ‘environment’, creating sustained imbalances between energy consumed and energy expended – e.g., through the availability of cheap, energy-dense foods and the growth of sedentary leisure activities.

A large literature has examined the role of the physical environment upon obesity – e.g. the availability of fast-food outlets and accessibility to greenspace. Overall, the literature has not produced consistent findings, but more fundamentally, these aspects of the environment are not distributed randomly spatially. As such, observed associations may be confounded by other aspects of the environment – e.g., area deprivation. A growing literature has also examined whether associations between genetic markers (as indexed by polygenic scores, PGS) and obesity varies across the aspects of the physical environment. However, these studies may also be materially affects by unobserved confounding.

In this study, we link genetic and geographic data from Understanding Society to examine the causal effect of three aspects of the obesogenic environment (local availability of fast-food, health food stores, and sporting facilities) upon adiposity and to additionally explore whether this causal effect varies according to a person’s PGS for BMI. As – like previous studies – our data our observational, we use two causal inference methods to assess and correct for bias. First, we repeat analyses using ‘negative control’ environments (local availability of betting shops and future obesogenic environment among movers) that should not influence current adiposity, but may be confounded in the same way. Second, we perform ‘sensitivity analysis’ allowing us to assess how associations would change when adjusting for plausible levels of confounding bias.

The analyses form part of the lead author’s UKHLS Fellowship project and will be completed by the conference.