Exploring the relationship between residential natural environment exposure and prevalence of depression in the UK Biobank

Dissertation submitted for the MSc in Environment and Human Health, University of Exeter.

Abstract

Background

Depression is the leading cause of ill health and disability globally and rates are continuing to rise. Evidence suggests exposure to the natural environment is beneficial for human mental health and studies have shown living in an area with a greater proportion of green and/or blue spaces is protective against depression.

Aims

The present study aimed to explore the relationship between residential natural environment exposure and lifetime depression status in participants from the UK Biobank study. I first examined whether an association is present and secondly, whether this relationship is 1) mediated by participants’ physical activity level, social connectedness or the time they spend outdoors, and 2) modified by participants’ genetic predisposition to depression.

Methods

This large, cross-sectional study used data from the UK Biobank study collected from participants at time of recruitment. The prevalence of lifetime depression was measured using data from the self-report questionnaire in the UK Biobank and Hospital Episode Statistic data. Residential natural environment exposure was measured as the percentage of land classified as natural environment present within a 300m buffer of the participants’ home addresses using land cover map data. Individual genetic risk scores were provided by Dr Jessica Tyrrell from the University of Exeter. Iterative logistic regression models were fitted to assess the main association and investigate potential mediators and gene–environment interactions. All models were adjusted for potential confounders.

Findings

Of the 230,392 participants included in the main and mediation analyses, 15.2% had a history of depression. A protective effect of greater residential natural environment on depression was consistently observed. Individuals living in the highest category of natural environment (≥35% natural environment)  had 7% lower odds of lifetime depression compared to the lowest category (<3% natural environment) in the fully adjusted model (OR 0.932, 95% CI 0.90–0.97). Time spent outdoors and physical activity level, but not social connectedness, were associated with amount of residential natural environment. However, no evidence was found to suggest they mediate the relationship between residential natural environment and depression. The gene– environment interaction analyses used a population of 293,706 participants, of whom 14.4% had a history of depression. No evidence was found to suggest that genetic risk for depression modifies the relationship between residential natural environment and depression in this cohort.

Conclusions

The results of this study provide further evidence that more natural residential environments are associated with lower prevalence of depression. However, longitudinal studies are needed to confirm the direction of the observed associations. This finding has considerable public health implications and highlights the need to protect and expand the small pockets of nature within a neighbourhood. While a gene–environment interaction was not found, this possibility cannot be ruled out. As depression subgroups become better characterised and genetic risk scores for depression are refined through improved genome-wide analysis techniques, evidence for an interaction may be found.