On Sunday, April 22, The New York Times led its front page with a story revealing that infant death rates were rising in the South, particularly among African Americans. To sociologist David Williams, that was hardly news.
Williams is one of the world's leading scholars on racial disparities in health. Now the Norman Professor of Public Health at the Harvard School of Public Health, he has wrestled intellectually and morally with such dismal statistics for more than a quarter century. Yet having authored scores of journal articles and book chapters, he's still left with more questions than answers.
Williams was a community health educator at the Battle Creek Adventist Hospital in the early 1980s. Since those days as a front-line advocate, his research has demonstrated the persistent ill health suffered by those who face daily poverty, violence, neighborhood segregation, joblessness, and substandard schools. In the United States, the new policy thrust toward eliminating racial and ethnic health disparities -- reflected in a surge of legislation among cities, states, and the federal government, as well as foundation backing -- owes much to Williams' work at the University of Michigan and at Yale.
"Health disparities," as they are popularly termed -- the gap in physical and mental well-being between the most privileged people in society and the most disadvantaged -- launched the modern public health movement in the nineteenth century. Today in the United States, these gaps endure, especially along racial lines. According to a Kaiser Family Foundation report published in January, the infant mortality rate -- one of the most sensitive indicators of population health -- is more than twice as high among African Americans as among whites. Death from heart disease is highest among African Americans, as is mortality from breast, lung, and colorectal cancer. Even sudden infant death syndrome strikes black babies at higher rates than it does infants from any other groups.
Teasing apart the causes of the disparities and finding solutions remain elusive. Williams has analyzed what he calls the "pathogenic" effects of residential segregation. He has shown how differences in conventional "risk behaviors" -- eating right, exercising regularly, not smoking, etc. -- don't explain the abiding health discrepancies between whites and minorities. And he has forcefully questioned whether health inequities can ever be reversed without fundamental social change.
Indeed, it was his early frustration at recruiting nearby African Americans for his hypertension and weight management programs -- over the century, the demographics around Battle Creek Adventist Hospital had changed from affluent white to poor black -- that persuaded Williams that larger forces were at work. "Making progress in improving health was not simply a matter of education," he said. "Social context mattered."
Tall, trim and bearded, Williams speaks passionately about his subject in the lilting, musical cadence of his native home, the Caribbean island of St. Lucia. He talked recently in his office with the Prospect.
MD: What do you know now about health disparities that you didn't know 25 years ago?
DW: My dissertation at the University of Michigan used the Tecumseh Community Health Study. It's one of those old classic epidemiologic studies where, in 1958, researchers at the University of Michigan went to a small town south of Ann Arbor, called Tecumseh. They enrolled every man, woman, and child into this study, about 10,000 people. Physiological measures, data on individuals. Then they followed them over time.
For my dissertation, I looked at socioeconomic inequalities in health, both chronic disease and mortality. Now, Tecumseh is a place where you would not expect to find socioeconomic inequalities in health. First, Tecumseh was 99 percent white: There were no minorities. Tecumseh had been home to the Tecumseh Products Company -- which was, at one point, this country's largest producer of compressor units for air conditioners and refrigerators. It was a town in a rural agricultural environment, but with a strong manufacturing base and relatively high salary levels. And the Tecumseh Products Company had been very good to that town -- it built a school, it built a lovely hospital.
So here is a community with good access to care. No urban poor. No rural poor. No minority poor. And the average blue collar worker makes a good living. It was a place where I would not expect to find social inequalities in health. And yet I found social inequalities in health. Persons of higher levels of income and education did better than those who didn't. It was quite striking for me, and it pointed to the power of socioeconomic status.
What statistic that relates to health disparities do you consider to be the most appalling? Is it the black/white gap in life expectancy? Premature mortality? Infant mortality?
The single example of health inequality that's most dramatic to me is what occurs when we look at racial-ethnic differences and, at the same time, at measures of socioeconomic status. For multiple indicators of health, the most advantaged or the best-off African Americans are doing more poorly than the worst-off whites.
A good example to make that concrete: African American women. According to national data, African American women aged 20 years and older with a college degree or more education have higher rates of low birth weight infants, and higher rates of infants who die at birth, than do white women who are high school dropouts. Remember: These are the African American women who are doing the best.
Why is it that the best-off African American women are doing worse than the least-off white women? It exists for infant mortality, birth outcome, obesity, hypertension.
So while socioeconomic status is important in health, there is a residual effect of racism?
There is an added burden of race that is evident in the health statistics. It's not just about economic status.
You talk about the gap between the best-off African Americans and the worst-off whites. Were there times in past or recent history when their health status converged?
If you look at racial/ethnic differences in health in the United States over the last 50 years, we haven't made much progress. Black-white differences in health have narrowed and widened as black-white differences in socioeconomic circumstances have narrowed and widened.
During the decade of 1968 to 1978 -- the ten years after the gains of the civil rights movement, and the Great Society, and anti-poverty programs -- we had a narrowing of the black-white income gap and a narrowing of the black-white gap in health. The health of blacks improved more rapidly; both on an absolute basis and a relative basis, blacks made more progress. During the decade of the 1980s, when the black-white gap in income worsened compared to what it was in the '70s, we went in the opposite direction: The health of blacks worsened absolutely and relative to that of whites during the decade of the '80s.
Today, the black infant death rate is 2.4 times higher than that of whites. That's a good example, in terms of the relative differences: It was 1.6 times higher in 1950. Now, in fairness, the absolute black-white gap in infant mortality is about half today what it was in 1950. It's the relative gap that is worse.
It's difficult to make a strict causal connection, and we don't know the time lag, but the patterning is quite striking. When economic circumstances improve, health improves. When economic circumstances worsen, health worsens.
You've written that differences in physical activity, smoking, and diet don't fully explain racial disparities in death rates.
Many times in this country, we focus on individual risk factors. Those individual risk behaviors for which we have the best evidence are physical exercise, cigarette smoking, alcohol abuse, dietary patterns, and obesity. They make a contribution, but they don't account for most of the social inequalities in health. So we need to look at the broader range of factors playing a role. Even with those individual behaviors, we need to think about the ways in which the larger social environment makes it easier or more difficult for individuals to lead a healthy lifestyle.
There is a lot of good research indicating that if someone lives in a safe neighborhood -- a neighborhood that has good sidewalks, walking paths, facilities for physical exercise -- they're much more likely to exercise regularly than someone who lives in a neighborhood that lacks those amenities. If individuals live in a neighborhood where there are supermarkets where they can get fresh fruits and vegetables, their daily intake of fresh fruits and vegetables is much higher than if they live in a neighborhood where there is no major supermarket, or there are corner grocery stores which either sell few fruits or vegetables or what they have is very expensive.
Or take the dramatic progress we've made in cigarette smoking over time. One of the larger determinants in declines in cigarette smoking is level of education. People who have a college degree are much more likely to quit smoking than those who have a high school education or less -- four times more likely to quit smoking over time, for the last several decades. The college degree is not simply a matter of having more education; over 90 percent of all people who smoke in the United States believe that smoking is bad for their health. So it's not simply a matter of knowledge -- it's about providing individuals with the resources to act on knowledge, to make changes, and to develop healthier alternatives. And that is clearly linked to larger conditions of economic opportunity, of working conditions, of living conditions.
The bottom line is: I'm not opposed to the notion of individual choice. What our responsibility is as a society is to create those conditions that make it easier for individuals to make the right healthy choices. In the final analysis, people have to choose. But what can we do to facilitate that choice?
You started writing early on about what you called the "pathogenic" effects of residential segregation. Does that seem to you to be the most far-reaching determinant of poor health?
Looking at the health of the African American population in general, yes. The larger issues are the living and working conditions that residential segregation creates. David Cutler, at Harvard, did a study years ago looking at the impact of segregation on young African Americans transitioning into adulthood. He argued that if you could eliminate the effects of segregation, which are so pervasive, it would completely eliminate black-white differences in education, income, unemployment, and reduce black-white differences in single motherhood by two-thirds.
You've also written about "fundamental" versus "distal" factors in health and disease.
I like to think of it as Maslow's hierarchy of needs. The needs for survival and food and a place to sleep are foundational. When individuals are struggling with the foundational needs of survival, they are not concerned about the more distal and probabilistic needs of reducing risks for health and quality of life -- they're trying to survive. As you meet individuals' basic needs, they can now turn their attention to the higher order needs of improving quality of life and protecting health and planning for the future. If you're worried about tomorrow, you're not worried about ten years from now.
If you were America's health disparities "czar," backed by unlimited funding, what would you do first?
I would take actions that would improve educational opportunities, employment opportunities, improve the quality of residential environments, develop environments that make it easier for individuals to make the right choices in terms of health.
I would want to see medical care that is accessible to all, and medical care that is more proactive and preventive in orientation. Medical care as currently practiced does not play a large role as a determinant of health in the United States -- but it could. Cuba is a good example. They spend a tiny, a miniscule, fraction of what we spend on health care in this country, and have better outcomes than we do on a range of health status indicators. And they don't have the high tech or all the latest procedures and diagnostic tests. What they do have is a greater emphasis on public health.
If we can emphasize procedures in medicine that are more preventive, we can actually achieve more out of our health care dollars. Some research suggests that health care quality in the United States is poor for all persons. There is a gap in health care quality by race and by socioeconomic status -- but even for the advantaged, we're not doing well.
In 1979, the U.S. Surgeon General stated that medical care explains only 10 percent of the variation in health status. Is that still true?
That is a widely used statistic in policy-making circles. The Canadians estimated it at 20 percent -- that's the highest estimate I've seen. Medical care is not the major determinant of health.
And yet you've written that medical care may be crucially important to vulnerable populations. Why?
Because in the context of great vulnerability, when facing lots of deficits, medical care can become an important health-enhancing resource. Especially medical care that takes a holistic view of the individual.
An example of that is prenatal care. For the average middle-class woman who is doing well in terms of her nutrition and her lifestyle in taking care of the baby, prenatal care doesn't add much. But it can add a lot for a poor woman who was already in poor nutrition -- it could provide dietary supplements and, by monitoring, identify a potential risk.
Americans have an individualistic and moralistic view of health. They think it's the individual's responsibility to protect health, and the individual's fault when things go wrong. Is it still an uphill battle to get policymakers to understand the societal underpinnings of health?
It still is a challenge. If you look at national data, you find that most adult Americans are unaware that racial/ethnic disparities exist. Most whites, most Latinos -- and most African Americans -- were unaware that there were black-white differences in infant mortality and life expectancy. And there is more awareness of racial/ethnic disparities than of socioeconomic disparities.
The point is: If we don't perceive that a problem exists, we don't mobilize to do anything about it. We need to raise awareness that we have inequalities in health -- and that they are harming not just the health of the individual but the society more broadly. The reason we haven't made more progress is not only our lack of knowledge. It's also our individualistic focus. It leads us to invest less than do some other societies in what's in the collective good. Think of it as a deeply-rooted American ideology of rugged individualism that pays scant attention to the needs of the larger collective.
Should we even be investing money in individual behavior interventions when the larger surroundings undermine them? For example, is there a role for exercise programs in church basements or classes about healthy eating, in neighborhoods where it's too dangerous to walk around outside or there are no quality supermarkets?
We need to work at every level. We need to work at an individual level, we need to work at the larger society level. There are many people who are unaware of the importance of physical exercise and the difference it can make in the quality of life. We want to share that information with them -- we want to empower them by giving them choices. What we're trying to do is level the playing field. Setting up an aerobic exercise program in the church basement: OK, the neighborhood may not be safe to walk, but you build in group support and camaraderie and get them involved. I think that's a great thing.
We also need to empower communities and community organizations that they can make a difference and that they can have some control. People need to have a sense that they can make a difference in their lives, that they do can matter. But we can't stop there. We have to think of how can we shape the larger environment.
Some scientists have tried to reduce racial differences in health to genetic differences. There's a lot of talk lately about studying "gene-environment interactions." Do we really need to know about genes to understand health disparities?
Yes and no. You certainly don't need to understand the genetic interactions produced in health disparities in order to develop interventions to address and eliminate them. There are many examples in science -- take the classic example of the outbreak of infectious disease caused by the Broad Street pump. They didn't have microscopes to understand cholera. But they took action to fix the problem. As a society, there are interventions we can make to improve the quality of life that will ultimately improve health, even though we don't understand all the underlying biological mechanisms and processes.
At the same time, as a scientist, I certainly think there is a legitimate place for studying how social adversity gets translated into the biology of everyday life. Probably my biggest concern is that we have the appropriate balance of research.
Among advocates who work on health disparities, you sometimes hear: "We already know everything we need to know about healthy disparities -- we just have to take action." From your point of view, if health disparities research ceased now, would we know everything we need to know in order to take action? Or are there still questions that need answering?
I really think there are important things we don't know.
On the one hand, we know a lot about the magnitude of disparities and trends in disparities. We know which policy areas should be priorities -- like education, living conditions, housing, work environments. At that level, the answers are clear.
What we honestly don't know: What are the levels at which we should intervene -- state, federal, local? We don't know which strategies are likely to have the biggest payoff. We don't know which specific interventions would be the most successful for society.
There are some areas where we have good data, like early childhood interventions. We know that investment in early childhood has a good payoff, and that it's cost-effective. But in many other areas, the data just aren't there. Even in areas of society where we have made interventions, we have frequently not assessed the health impact. We haven't assessed the health impact of non-health policy, such as the earned income tax credit, the Social Security system, living wage ordinances. What impact do these have on the health of communities? I would argue that, given what we know about social determinants of health, the factors that are the largest drivers of the health of the population are outside of what we traditionally define as the health care sector.
From a scientific point of view, there will be a lot of work for a long time. But from a societal point of view, a policy point of view, we know enough.