Health Disparities among the U.S. Elderly

This dissertation examines health disparities related to race/ethnicity and socioeconomic status among the U.S. elderly taking into account empirical challenges including: (1) self-selection and unobserved factors; (2) subjective and objective measures of disease; and (3) attrition bias. Using the 2006 Health and Retirement Study, we find that current estimates of racial/ethnic disparities in awareness of chronic disease are sensitive to self-selection and unobserved factors. Contrary to prior studies reporting that African-Americans are more aware of having hypertension than non-Latino whites, we do not find this conclusion to be true after controlling for self-selection and disease severity. Likewise, prior studies show mixed evidence of racial/ethnic disparities in awareness of diabetes, but after accounting for selection, we find that African-Americans and Latinos are less aware of having diabetes compared to non-Latino whites. Using the National Health and Nutrition Examination Survey from 1999-2008, we find that education is not associated with self-reported measures of diabetes and hypertension, and it is positively associated with self-reported high cholesterol. However, there is a strong negative association between education and diabetes and hypertension when we use objective measures. There is no association between education and objective measures of high cholesterol. When we account for the possibility of shared, unmeasured determinants of disease prevalence and diagnosis that are correlated with education, we find that education is negatively associated with having undiagnosed hypertension and diabetes. In addition, we find that trends in cardiovascular disease (CVD) risk factors, including (1) high blood glucose, (2) high blood pressure, (3) high cholesterol, and (4) smoking, improved over the past two decades among individuals with diabetes, but racial/ethnic and education-related disparities have emerged in some areas. Finally, we estimate income-related health inequality, measured by the concentration index with an unbalanced panel dataset, explicitly accounting for attrition due to mortality, using the RAND version of HRS. We find that income-related health inequality improves as people age, but this improvement primarily comes from the high mortality rate among poorer individuals. A balanced panel dataset analysis generates misleading results of progressive health deterioration, while an unbalanced panel dataset analysis brings out this regressive deterioration.

racial differences

69542