May 2, 2006
Published in: JAMA, The Journal of the American Medical Association, v. 295, no. 17, May 3, 2006, p. 2037-2045
Posted on RAND.org on April 30, 2006
CONTEXT: The United States spends considerably more money on health care than the United Kingdom, but whether that translates to better health outcomes is unknown. OBJECTIVE: To assess the relative heath status of older individuals in England and the United States, especially how their health status varies by important indicators of socioeconomic position. DESIGN, SETTING, AND PARTICIPANTS: We analyzed representative samples of residents aged 55 to 64 years from both countries using 2002 data from the US Health and Retirement Survey (n = 4386) and the English Longitudinal Study of Aging (n = 3681), which were designed to have directly comparable measures of health, income, and education. This analysis is supplemented by samples of those aged 40 to 70 years from the 1999-2002 waves of National Health and Nutrition Examination Survey (n = 2097) and the 2003 wave of the Health Survey for England (n = 5526). These surveys contain extensive and comparable biological disease markers on respondents, which are used to determine whether differential propensities to report illness can explain these health differences. To ensure that health differences are not solely due to health issues in the black or Latino populations in the United States, the analysis is limited to non-Hispanic whites in both countries. MAIN OUTCOME MEASURE: Self-reported prevalence rates of several chronic diseases related to diabetes and heart disease, adjusted for age and health behavior risk factors, were compared between the 2 countries and across education and income classes within each country. RESULTS: The US population in late middle age is less healthy than the equivalent British population for diabetes, hypertension, heart disease, myocardial infarction, stroke, lung disease, and cancer. Within each country, there exists a pronounced negative socioeconomic status (SES) gradient with self-reported disease so that health disparities are largest at the bottom of the education or income variants of the SES hierarchy. This conclusion is generally robust to control for a standard set of behavioral risk factors, including smoking, overweight, obesity, and alcohol drinking, which explain very little of these health differences. These differences between countries or across SES groups within each country are not due to biases in self-reported disease because biological markers of disease exhibit exactly the same patterns. To illustrate, among those aged 55 to 64 years, diabetes prevalence is twice as high in the United States and only one fifth of this difference can be explained by a common set of risk factors. Similarly, among middle-aged adults, mean levels of C-reactive protein are 20% higher in the United States compared with England and mean high-density lipoprotein cholesterol levels are 14% lower. These differences are not solely driven by the bottom of the SES distribution. In many diseases, the top of the SES distribution is less healthy in the United States as well. CONCLUSION: Based on self-reported illnesses and biological markers of disease, US residents are much less healthy than their English counterparts and these differences exist at all points of the SES distribution.