Abstract
Background:People with severe mental illnesses (SMI), such as schizophrenia, experience an 10- to 20-year reduced life expectancy compared to the general population, primarily due to preventable physical causes. The contribution of neighbourhood-level characteristics, such as residing in areas with more green spaces, in influencing mortality risk in SMI remains uncertain. This thesis aims to address this knowledge gap.
Methods:
Study 1: I conducted a systematic review examining the association between area-level exposures and mortality in SMI.
Studies 2-4: Cohort data was extracted from the Clinical Records Interactive Search, a secondary mental healthcare case register containing de-identified patient records from the South London and Maudsley (SLaM) NHS Foundation Trust. SLaM is one of the largest European mental health providers covering a catchment area of 1.3 million residents. De-identified patient-level data linked to neighbourhood-level exposures at middle super output area-level (MSOA) (mean population: 7,000 people), were used.
In study 2, age and gender standardized mortality ratios (SMR) were calculated at the area level, visually mapped, and analysed for spatial autocorrelation using hot spot analyses. Area-level indicators assessed for associations with the SMRs included: (urban built environment) area deprivation, population density, green spaces and food retail outlets, and (urban social environment): social fragmentation, ethnic density and places of worship. Ecological regression analyses were also undertaken.
Studies 3 and 4 were cohort studies which used multilevel regression approaches to assess area-level associations with mortality in SMI. Study 3 assessed the overall availability of environmental features in the surrounding local area, whereas study 4 assessed the walking distance to the environmental features.
Results:
Study 1: Systematic review searches identified 3,353 abstracts. 7 studies from 4 countries were included in the synthesis - quality of the studies was moderately high. Consistent associations between residence in more deprived areas and increased risk of mortality outcomes in SMI, and evidence of an association between areas with higher proportions of racialised minority groups and a reduced risk of mortality in SMI, was noted. No studies investigated the effect of the built environment on mortality outcomes.
Study 2: SMRs ranged from 1.28 to 6.84 in people with SMI across 148 MSOAs, in the catchment. Hot spot analyses revealed significant clustering of area-level exposures, while no significant clustering was observed for SMRs. Ecological analyses suggested that residence in areas with higher levels of deprivation and convenience stores were associated with higher SMRs. However, this association was attenuated in the fully adjusted models.
Studies 3 and 4: followed over 25,000 people until death or until the end of the study (study 3, n = 22,689 and study 4, n = 22,564). In a cohort study design, weak associations were observed between residence in areas with higher proportions of more walkable areas, convenience stores and population density with lowered risk of all-cause and natural-cause mortality at the MSOA level. Associations were attenuated after adjusting for confounders.
Study 4: Weak associations were observed between residence in areas with increased distance to green spaces and fast-food outlets and higher risk of all-cause mortality. Associations between increased distance to green space and all-cause mortality persisted after accounting for individual-level covariates. However, the association between fast-food outlets and risk of mortality was attenuated.
Conclusions:
Excess mortality remains a health inequality in SMI, with some associations observed between neighbourhood-level indicators and mortality. Limitations include the absence of rural areas in the dataset and a lack of information on comorbid physical health. The thesis concludes with discussions on future clinical, policy, and practice implications based on the findings.
| Date of Award | 1 Mar 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Jayati Das-Munshi (Supervisor), Til Wykes (Supervisor) & Craig Morgan (Supervisor) |