W14. EXPLORING THE RELATIONSHIP BETWEEN GENETIC PREDISPOSITION TO MENTAL HEALTH DISORDERS AND THE SOCIAL ENVIRONMENT IN THE UK BIOBANK

• Published in: European neuropsychopharmacology Vol. 75; p. S111
• Main Authors: Vassos, Evangelos, Maxwell, Jessye, Herle, Moritz, Coleman, Jonathan, Breen, Gerome
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-10-2023
• ISSN: 0924-977X; 1873-7862
• DOI: 10.1016/j.euroneuro.2023.08.205

There is a higher prevalence of mental disorders in individuals born in urban settings, an observation that led to the proposition that urbanicity acts as an environmental risk factor for mental disorders. This effect may be confounded by a shared genetic aetiology and thus individuals born in cities may be at higher risk because their parents are predisposed to both mental disorders and urban living. In a previous study published in JAMA Psychiatry, we found evidence supporting the hypothesis of genetic selection of the environment we live in, which intersects the traditional gene-environment dichotomy. In this study, we dive deeper by exploring potential pathways explaining the association of polygenic scores with neighbourhood characteristics including urbanicity, Townsend and social deprivation and fragmentation. We test the mediating effects of behavioural phenotypes including risk-taking, psychotic experiences, education, neuroticism, unhappiness, and social isolation.
We analysed 131,632 UK Biobank participants who completed the Mental Health Questionnaire, by cross-referencing address coordinates with external demographic sources to derive neighbourhood measures of urbanicity, social fragmentation, and Townsend and social deprivation. Polygenic scores (PGS) for schizophrenia, bipolar disorder, major depression, ADHD, ASD, anorexia nervosa, PTSD, and panic disorder were generated on the latest GWAS summary statistics downloaded from the Psychiatric Genetics Consortium. We used a multivariate mediation framework to decompose the association of eight psychiatric disorder PGS with four neighbourhood characteristics into direct effects and indirect effects acting through behavioural phenotypes.
Higher schizophrenia, bipolar disorder, and ASD PGS were all associated with higher urbanicity, social fragmentation and Townsend deprivation (β=0.015 to 0.032), while indirect effects via behavioural mediators explained 20-59% of the total effects. Additionally, behavioural phenotypes largely mediated the association of ADHD, major depression, and PTSD PGS with increased Townsend deprivation. ADHD, major depression, and PTSD PGS were associated with higher, while bipolar disorder and anorexia nervosa were associated with lower social deprivation. Among the six behavioural phenotypes examined, risk-taking, psychotic experiences, and education were the principal mediators across the different psychiatric disorder PGS and neighbourhood characteristics.
We have previously shown genetic predispositions to a variety of mental disorders are associated with population density in the UK Biobank. In this study, we confirm the previous associations with a more elaborate measure of urbanicity and we expanded to other neighbourhood characteristics, including social fragmentation and Townsend and social deprivation. Furthermore, we investigated the role of behavioural characteristics and we found that risk taking, psychotic experiences, and education partially mediated the associations between psychiatric disorder PGS and neighbourhood characteristics. These results contribute to a deeper understanding of the underlying mechanisms that may link genetic predisposition for psychiatric disorders to environmental exposure. To interpret these findings, it is imperative to further explore the causal links between the behavioural phenotypes, as well as those between the behavioural phenotypes and psychiatric conditions.