Associations of socioeconomic position with stroke

Maxwell Pistilli

doi:10.17918/00000385

Problem: Stroke was responsible for 37.6 deaths per 100,000 people in the United States (US) in 2017, making it the 5th leading cause of death in the US. Stroke risk is higher in people of lower socioeconomic position (SEP), but estimates of this difference in risk vary greatly by study, are affected by limitations in data and study design, and do not emphasize the different domains of SEP. The American Heart Association issued a statement specifically calling for investigators to, "broaden the focus to incorporate... the social determinants of health..." into stroke research, and here we will do so in a series of US population based aims. Methods: Studies were based on the US population, first an ecologic analysis of the association between area-level stroke death rates and area-level SEP (aim 1), then two studies based on the population based Reasons for Geographic and Racial Differences in Stroke (REGARDS) study assessing the relationship between individual incident stroke and individual- and area-level SEP (aims 2 and 3). The SEP domains income/wealth, education, and occupation were isolated to identify their relative associations with stroke. Area-level variables were created at the county level by incorporating multiple census items specific to their respective domains. Aim 1 used geographically weighted regression to describe and incorporate spatial heterogeneity, aim 2 used proportional hazard regression to find optimal models, and aim 3 assessed the accuracy of a validated stroke risk model by stratified by SEP. Results: Aim 1 identified significant spatial heterogeneity in stroke risk, but with unclear implications for intervention. It also identified area-level income/wealth as the most relevant domain to area-level stroke death rates. Aim 2 found that area-level income was not associated with individual incident stroke, but that individual income and occupation, and area-level occupation, were all significantly associated with incident stroke even after adjustment for clinical risk factors. It also found a significant multi-level interaction within occupation. Aim 3 found substantial lack of accuracy in low risk individuals moderated by SEP, moderate lack of accuracy in medium risk individuals by SEP, and general overestimation that did not vary by SEP among high risk individuals. There was no association of area-level SEP with stroke risk accuracy. Conclusions: This thesis supports the use of socioeconomic position as an important risk factor for stroke. Socioeconomic characteristics of both the individual and of the county in which they reside are associated with stroke, even after adjustment for traditional clinical risk factors. An individual's income appears to be the domain of their SEP most relevant to their likelihood of stroke, and there is spatial heterogeneity in this association, but it is unclear whether the average income and wealth, or education, or occupation, of their county is the most relevant area-level domain. This research used a population-based cohort that is representative of the United States, and thus serves a reference for future studies, which will be required to identify mechanisms of the association.

Associations of socioeconomic position with stroke

Files and links (1)

Abstract

Metrics

Details

Associations of socioeconomic position with stroke

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media