Social
Expansion of 4-year colleges in the United States associated with lower later-life blood pressure with similar relationships at all quantiles of blood pressure Amanda Irish* Amanda Irish M. Maria Glymour Fei Jiang Anusha Vable
Introduction: More schooling predicts lower mean blood pressure in later life, however impacts along the entire blood pressure distribution are unclear. Increases in education may have differential impacts for people who would otherwise have high versus low blood pressure; for example, educational benefits in the high-risk right tail of the blood pressure distribution could disproportionately reduce stroke and heart attack risk compared to benefits at the mean or low-risk left tail. We evaluate the effect of the increase in 2- and 4- year colleges from the 1940s to the 1980s on the systolic blood pressure distribution, and test for differential relationships in race-by-gender subgroups.
Methods: Using Health and Retirement Study data (N = 6,145), we estimated the association of the number of 2- and 4-year colleges per county per year, standardized to the number of 18- to 22-year-olds from 1940 to 1982, with measured systolic blood pressure. We used linear regression to evaluate mean effects and conditional quantile regression to evaluate distributional effects, and adjusted for individual- and state-level covariates. Interaction terms evaluated differential associations by the combination of race (Black vs White) and gender (male vs female).
Results: In pooled analysis, the number of 2-year colleges in the county was not associated with the mean or any quantile of systolic blood pressure. However, the number of the 4-year colleges in the county was associated with lower mean systolic blood pressure (b= -0.1 mmHg (95% CI: -0.19, -0.015), which was consistent across its distribution: there was no evidence of differential impact. We also did not find meaningful differences in the effect of education across race-by-gender subgroups.
Conclusion: Geographic accessibility to 4-year colleges was associated with lower systolic blood pressure, with little distributional or subgroup variation.
