Multiscale and Spatiotemporal Dynamics of Socioeconomic and Environmental Effects on Mental Illness Mortality
Loading...
Date
2024-01-10
Authors
Song, Insang
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oregon
Abstract
Mental illness is a pressing global and national public health concern, necessitating the identification of risk factors to develop effective prevention measures. In this dissertation, I attempt to fill two research gaps by revealing the spatial and/or temporal disparity in the impacts of unemployment and greenspace on mental illness mortality with spatiotemporal modeling and a causal analysis across three spatial scales.In Chapter 2, the association between mental illness and substance use mortality and unemployment was examined using Bayesian spatiotemporal hierarchical models. The findings revealed heightened positive effects in rural Appalachian and Midwestern counties. Overall mild effects were observed during the Great Recession period. The patterns could be attributed to local contexts such as the availability of healthcare supply and relative deprivation. Chapter 3 challenges the assumption of a spatially constant effect of greenspace exposure on mental illness mortality, using census tract-level data from Oregon and Washington. Results indicated that the impact of greenspace exposure on mental illness mortality varies across census tracts, with protective effects more likely in areas between Seattle and Portland. Protective effects were more likely observed in areas between Seattle and Portland. The contrast between urban and rural areas was explained through factors such as patient preference and differential availability and accessibility to greenspaces. Chapter 4 shed light on the spatial differences in the causal effects of greenspace exposure on mental illness mortality using data from the State of Washington. Dichotomized treatment settings and propensity score matching methods were leveraged to examine the spatial disparity in causal effects of greenspace exposure to mental illness mortality. The results elucidated that the causal effect differed significantly across regions within Washington state, emphasizing that spatial heterogeneity is a critical element when examining the causal effects of greenspace exposure on mental illness mortality.
By highlighting the spatial and/or temporal disparity in socioeconomic and physical environment factors’ effects, this dissertation provides new perspectives to spatiotemporal mental health research and suggests a transition from disease mapping to effect mapping. This transition offers evidence to devise locally-focused measures that consider the spatial disparities of associative and causal effects.
Description
Keywords
Bayesian analysis, Causal inference, Mental illness, Spatial disparity, Spatiotemporal analysis