Overcoming Racial and Ethnic Disparities in Healthcare
Behind the Cancer Headlines®
February 28, 2005
Hidden biases in data and analytical methods may obscure important insights needed to close the gap in persistent U.S. racial and ethnic disparities in health, said David Williams, the Harold W. Cruse Collegiate Professor of Sociology at the University of Michigan, who delivered the inaugural Diversity and Public Health lecture at the Harvard School of Public Health (HSPH).
"If you think about it, we’ve done a lot – greater access to hospitals, Medicare and Medicaid, the Civil Rights revolution," said Williams. "One of the implications [of health disparities] is that what we have done to date hasn’t worked. We need to think about what we can do differently."
Williams spoke at the first lecture in a new series designed to bring distinguished minority scientists and scholars to HSPH to speak on important public health topics.
Covering a range of issues in health disparities research, Williams began with well-documented differences in health conditions among groups of Americans. The best-documented pattern of differences is between African Americans and whites, he said. For example, in Mississippi, where both whites and African Americans have the highest rates of heart disease in the country, the death rates for the sickest white women are lower than for the healthiest black women.
Even for experts, the existence or magnitude of disparities still may be hidden in the way data are measured, analyzed, and presented, Williams said.
For instance, age adjustments are a reasonable and widespread way to compare relative risks between populations, but they conceal a racial health gap in early life, he said. Age-specific mortality rates for blacks are at least 100 percent higher than whites from ages one to four and from ages 25 to 54. But this huge difference is hidden when policymakers talk about overall age-adjusted rates, which is 30 percent higher for blacks when compared to whites. Still high, but not as dramatic as the more specific analysis.
Infant mortality rates have dropped dramatically for both blacks and whites, but the gap in death rates between the two groups has widened tremendously. Today, a black infant is 2.5 times more likely to die than a white infant. In 1950, a black infant was 1.6 times more likely to die than a white infant.
Biology alone cannot explain the differences. "The whole issue of race and biology raises important opportunities for research," Williams said. "Biology is adaptive to environmental conditions. Not surprisingly, you will find biological differences in groups living in different conditions. It’s important to observe whether the differences are innate due to an underlying genetic endowment, or acquired due to the environment. I don’t think we’re looking at an either/or model. We need to understand how these issues come together to affect health."
For example, a simple biological model doesn’t seem to be an option for explaining underlying differences in hypertension. Researchers are seeking a possible gene underlying the higher rates of hypertension in U.S. blacks, he said, but one study showed low rates of hypertension among people of West African origin still living in Nigeria and Cameroon compared to high rates among those in Illinois. U.S. whites have higher rates of hypertension than blacks living in Africa, he said.
As health disparities researchers have moved away from biology, they have focused on socioeconomic status (SES) to explain health disparities, Williams said. SES is a powerful predictor of health, but race has an independent effect at every income level.
Residential segregation is also high at every income level and, in many U.S. cities, it is close to the legally mandated level of apartheid South Africa. Sixty-six percent of blacks in the U.S. would need to move in order to evenly distribute blacks and whites in neighborhoods.
Effective interventions are a crucial area of research, he said. The relatively better health of U.S. whites is still not optimal compared to people in two dozen countries spending less on health care. Researchers need to identify policies that improve health overall, as well as target interventions to address disparities.
Researchers also need to delve into resilience factors, such as religion, which provides a strikingly broad range of protection across a range of health outcomes, Williams said.
Harvard School of Public Health (http://www.hsph.harvard.edu)