Living Apart, Breathing Apart: Associations between Racial Residential Segregation, Air Pollution Exposure, and Chronic Disease

Abstract

Racial residential segregation (RRS) is a key structural driver of health inequities. Racist and discriminatory government-sanctioned housing policies have produced and maintained the spatial separation of different race and ethnicity groups. This separation has shaped the inequitable distribution of resources, opportunities, and adverse physical and social exposures across neighborhoods. As a result, communities of color shoulder a disproportionate burden of exposures harmful to health, such as increased air pollution exposure. They also may be more vulnerable to the health effects of these exposures due to co-occurring harms associated with RRS, such as poverty, discrimination, and reduced access to health-promoting resources. A preponderance of evidence shows that air pollution is deleterious to human health. Air pollution causes cardiovascular disease (CVD) and is linked to premature death while a growing body of evidence suggests a link to depression. Within segregated communities of color, older adults may be especially susceptible to the cardiovascular effects of air pollution exposure and RRS, given the elevated risk of CVD with age. Older cis-females may be particularly susceptible to both CVD and depression due the role of the menopausal transition in accelerating CVD and depression risk, as well as striking gender inequities rooted in sexism. However, scant research exists on the joint associations of RRS and air pollution with CVD and depressive symptoms by race, ethnicity, and sex group. Rigorous epidemiologic research is critical to address the health consequences of RRS, reduce racial and gender health inequities among older adults, and promote environmental justice. This dissertation aimed to fill these gaps. Aim 1 involved investigations of associations of RRS with stroke, myocardial infarction (MI), and CVD mortality and assessment of whether exposure to long-term nitrogen dioxide (NO2) and particulate matter less than 2.5 ?m in diameter (PM2.5) modified these associations. I pooled data from non-Hispanic Black and non-Hispanic white participants in two large prospective cohort studies: Multi-Ethnic Study of Atherosclerosis (MESA) and REasons for Geographic and Racial Differences in Stroke (REGARDS). Racial residential segregation was measured using the Divergence and Isolation Indices. The Divergence Index indicates the extent to which the racial composition of a census tract deviates from that of a broader referent region. The Isolation Index captures the probability that a member of a given racial group shares a census tract with another member of the same group. All analyses were stratified by race, ethnicity and sex jointly. Higher Divergence was associated with greater CVD risk, particularly MI and CVD mortality, among non-Hispanic white females. Meanwhile, higher white Isolation was associated with lower CVD risk in this group. A stronger association between Divergence and stroke was observed among non-Hispanic Black males in MESA. However, estimates were closer to the null in REGARDS and pooled analyses. There was limited and inconsistent evidence that NO2 or PM2.5 meaningfully modified associations between RRS and CVD. Aim 2 involved examination of sex-specific associations of annual average NO2 and PM2.5 with depressive symptoms using longitudinal MESA data. Depressive symptoms were measured using the Center for Epidemiologic Studies Depression Scale (CES-D). Long-term NO2 and PM2.5 exposures were not significantly associated with depressive symptoms among older adults. Associations appeared slightly positive among females and null or slightly inverse among males. However, these sex-specific patterns were imprecise and should be interpreted cautiously.Aim 3 involved investigations of associations of RRS with depressive symptoms and effect modification by annual NO2 and PM2.5. I used longitudinal MESA data to assess these associations across eight race, ethnicity and sex strata. Some evidence of heterogeneity in associations of RRS with depressive symptoms was observed by race, ethnicity and sex group. Specifically, higher Divergence was associated with elevated depressive symptoms among non-Hispanic Black and Hispanic males only. However, effect sizes were relatively small in magnitude. Associations were null among all other race, ethnicity and sex groups. Although associations among Hispanic females became more negative at higher PM2.5 and NO2 levels, estimates were imprecise. Evidence of effect modification by air pollution was limited. Findings from this dissertation project suggest that RRS is a complex, multi-dimensional construct. Associations of RRS with CVD and depressive symptoms vary by segregation dimension, outcome, study setting, and race, ethnicity and sex group. Although air pollution is an established health risk, this dissertation found minimal evidence that NO2 or PM2.5 consistently modified associations of RRS with CVD or depressive symptoms. These findings underscore the need for context-specific approaches to addressing the health consequences of RRS that account for local patterns of segregation, environmental exposures, and resource distribution across different race, ethnicity and sex groups.