Comparisons of individual- and area-level socioeconomic status as proxies for individual-level measures: evidence from the Mortality Disparities in American Communities study

Table 5 Associations between individual-, census tract-, and county-level indicators of socioeconomic status with all-cause mortality over 7-year follow-up, Mortality Disparities in American Communities study

	Individual-level characteristic		Census tract-level characteristic		County-level characteristic
	OR	95% CI	OR	95% CI	OR	95% CI
Household income
Above US median	0.318	(0.315–0.321)	0.782	(0.776–0.789)	0.803	(0.796–0.810)
Poverty level
≤ 100% FPL	1.271	(1.255–1.287)	1.005	(1.004–1.005)	1.013	(1.012–1.013)
Education
Has high school degree	0.395	(0.391–0.399)	0.838	(0.831–0.844)	0.895	(0.887–0.902)
Has college degree	0.480	(0.475–0.486)	0.994	(0.993–0.994)	0.989	(0.989–0.989)
Employment/occupation
Is unemployed^a	0.356	(0.344–0.368)	1.092	(1.084–1.101)	1.112	(1.103–1.121)
Employed in blue-collar industry^b	0.284	(0.282–0.286)	1.006	(1.006–1.006)	1.015	(1.014–1.015)
Other characteristics
Owns home	0.799	(0.791–0.806)	0.999	(0.999–1.000)	1.002	(1.002–1.003)
Foreign born	0.536	(0.527–0.544)	0.867	(0.859–0.875)	0.859	(0.852–0.867)

For household income, education, and owns home, having a value of 1 was considered low risk for health outcomes, while having a value of 0 was considered high risk; in contrast, for poverty, employment/occupation, and foreign born, having a value of 1 was considered high risk, while having a value of 0 was considered low risk. Each cell is a different logistic regression model. Models included a control variable for county population. All p < .0001
OR odds ratio, CI confidence interval, FPL federal poverty level
^aUnemployment was measured among participants who reported being in the workforce (e.g., excluding retirees)
^bEmployment in a blue-collar industry was measured among participants who reported being employed (i.e., excluding unemployed participants)

ISSN: 1478-7954