Social determinants of health in older age: gene-environment interactions across the life course

Thesis

This thesis investigates how social and environmental factors shape health outcomes in later life and how these effects vary according to genetic risks, conceptualized through gene–environment interactions (𝐺 × 𝐸). Health in older age reflects a dynamic life- course process influenced by biological, social, and environmental exposures, with growing evidence that most conditions arise from the interplay of genetic risks and contextual environments. Leveraging advances in genome-wide association studies and polygenic scoring, this project applies quasi-experimental designs to examine how socioeconomic status and life stressors interact with genetic liability to influence cognitive function and mental health in older adults in the United Kingdom. The three essays use data from the English Longitudinal Study of Ageing (ELSA). The first applies group-based trajectory modelling, identifying distinct patterns of memory and depressive symptoms. Results indicate cognitive stability, modest reductions in depressive symptoms, and strong socioeconomic gradients, with disadvantages concentrated among older, female, and socioeconomically deprived groups. The second employs a regression discontinuity design exploiting the 1947 UK school reform to estimate the causal effect of education on late-life cognition and its interaction with genetic risks. Findings suggest that while additional schooling benefits cognition, it may amplify the detrimental influence of the APOE4 allele. The last essay examines mental health consequences of the COVID-19 pandemic, showing that men with higher polygenic scores for major depressive disorder experienced greater increases in depressive symptoms, underscoring differential vulnerability to environmental shocks. Taken together, the thesis provides novel evidence of the heterogeneity of cognitive and mental health trajectories in later life and substantive understanding of gene– environment interplay across the life course. By integrating polygenic scores with quasi- experimental designs, the analyses highlight how social exposures—such as education and the COVID-19 pandemic—interact with genetic risks to shape ageing outcomes.

Authors: Liu, Y

• DOI: 10.5287/ora-pyr8y7qe9
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: gene-environment interaction; social determinants of health
• Subjects: sociogenomics