Linking housing, socio-demographic, environmental and mental health data at scale

Abstract

Mental disorders are a growing problem worldwide, putting pressure on healthcare systems and wider society. Anxiety and depression are estimated to cost the global economy US$1 trillion per year, yet only 2% of global median government healthcare expenditure goes towards mental health. There is growing evidence linking housing, socioeconomic status and local environmental conditions with mental health inequalities. The aim of this paper is to link several open-access datasets at the local area level (N = 32,844) for England to clinical mental health metrics and describe initial statistical findings. Two mental health metrics were used: Small Area Mental Health Index (SAMHI) and diagnosed depression prevalence. To demonstrate the utility of the longitudinal mental health data, changes in depression prevalence were investigated over two study periods (2011–19, i.e. austerity; and 2019–22, i.e. COVID-19). These data were linked to housing data (energy efficiency, floor area, year built, type and tenure) from Energy Performance Certificates (EPCs); socio-demographic data (age, sex, income and education deprivation, household size) from administrative records; and local environment data (winter temperature, air pollution and access to green space). The linked dataset provides a useful resource with which to investigate the social and environmental determinants of mental health. PRACTICE RELEVANCE Initial observations of the data revealed a non-linear relationship between home energy efficiency (EPC band) and the mental health metrics, with depression prevalence higher in local areas where the mode EPC bands were C and D, compared with B and E. Researchers can further investigate this relationship using the dataset through robust statistical analysis, adjusting for confounding variables. National and local governments may use the dataset to help allocate resources to prevent and treat mental health conditions. Practitioners can map and interrogate the data to describe their local areas and make preliminary conclusions about the relationships between the built environment and mental health. This preliminary analysis of the data demonstrated a gradient in SAMHI and depression prevalence with income and employment deprivation at the local area level.