Race vs. Place
Can an integrated neighborhood in Southwest Baltimore help overturn decades of race-based assumptions about the origins of health disparities?
From both terra firma and Google Earth's satellite view of the world, census tracts 1902 and 1903 are entirely bleak, drained of color and vitality. Street after street of this impoverished Southwest Baltimore neighborhood reeks of unrelenting hardscrabble existence, of teenage mothers, cigarettes and sodas in hand, walking the same cracked pavement as prostitutes and drug dealers, all within eyeshot of millionaires' homes and Baltimore's major league baseball and football stadiums.
"Half the people out here have guns," admits a 15-year-old of this area bordering on Washington Village, known to the locals as Pigtown for the 19th-century swine that used to be herded through its streets. Such is life when the only easily available resources to the local citizenry are hazardous to their health. Smiley & Nancy's liquor store, the Gilmore Pleasure Club ... no shortage of bars here. But there's a huge dearth of beneficial fare. The bullet-proofed corner store across the street sells high-priced Spaghetti O's by the score, but "fresh" food is limited to four lonely loaves of white bread and a shoebox-sized array of cold cuts and cheese—little surprise from a food store whose only outside advertisements are for cigarettes.
National data on health disparities led policymakers and scientists to seek medical solutions where none existed, says researcher Thomas LaVeist.
Even on a gorgeous Sunday afternoon, the pain of this place is hard to miss. It sings out from every nail slammed into the boarded row houses that, on some streets, outnumber inhabited homes. It glitters in the broken glass strewn across the blue bottom of a drained city wading pool, ignored by two youngsters standing in its midst tossing a football. Buildings or people, it's all the same; neglect, benign or otherwise, has—like rain on limestone—eroded the health of this neighborhood. And no one here, black, white or otherwise, is exempt.
That's not opinion. It's fact. And the researchers who have come to this neighborhood to investigate this long-suspected but never before proven notion—that poor health is not so much about race, as place—may eventually rewrite the way health disparities are viewed and treated in this country.
The traditional approach to health disparities always left investigator Thomas LaVeist, PhD, with a half empty feeling. (NIH defines health disparities as any "significant disparity in the overall rate of disease incidence, prevalence, morbidity, mortality or survival rates" in one population compared to the general population.) On one hand, LaVeist, director of the Center for Health Disparities Solutions, was grateful to have any data at all. It wasn't until the 1980s that the government began collecting large-scale health disparities data that included information on all ethnicities. That effort, the National Health Interview Survey (NHIS), gave researchers like LaVeist their first comprehensive glimpse of how different populations manifested illness.
But the NHIS survey also provided LaVeist with a dilemma. It created reams of data pointing out racial differences in health disparities, especially when comparing minorities to whites. But that's as far as the data went—in essence delivering a whole lot of "who" and precious little "why." For a researcher who had spent his whole career trying to unravel "why," the data was, in LaVeist's opinion, leading policymakers and scientists to seek medical explanations and solutions where none existed. "It's not that the data was faulty," says LaVeist, "but rather, the way we explained how the data came to be was faulty. It's perfectly accurate to look at the national statistics and say that African Americans have three times the death rate from a certain condition as do whites. But then you have to ask why that's so." For LaVeist, the William C. and Nancy F. Richardson Professor of Health Policy, his 20-plus years of research intimated that the effects of poverty and other social factors played a huge role in creating health disparities. His first papers suggested that by addressing socioeconomic factors as well as medical issues, differences in infant mortality rates between poor African Americans and other races could be reduced. Later he would document the overabundance of liquor stores that pockmark impoverished communities, and the lack of supermarkets—creating so-called "food deserts"—in such areas. Though these were mostly African-American communities, LaVeist was working to define a problem—poverty, lack of access to healthy resources—that knew no color.
It wasn't theory alone that suggested this approach, but personal experience. Early in his career, LaVeist worked for a time in a hospital in the same Baltimore community he would later survey. For all the talk and commonly held belief that blacks were sicker and more pathologic than whites, LaVeist's eyes told him that both groups were distressed. "I got to know that part of the city pretty well," he says. "I knew that there were a lot of sick white people in that area. There was a lot of obesity; if you just walked down
the street you'd see it. A lot of smoking. I started wondering, if they're all living in these same conditions... It looked pretty unhealthy here, and I wondered, if you studied it, what would you find?"
LaVeist knew that to move the health disparities discourse beyond a biological conceptualization of race, to consider instead the role of place, he had to create a methodologically sound survey that might yield such data. The Exploring Health Disparities in Integrated Communities (EHDIC) study is an effort to take race and standard socioeconomic differences such as income out of the health disparities picture and to investigate the following hypothesis: If people live in the equivalent of an urban war zone, is everyone a casualty?
The analogy isn’t so far-fetched. Roland J. Thorpe Jr., PhD, an assistant scientist who worked on the study with LaVeist and fellow Center colleagues Darrell Gaskin, PhD, and Tiffany Gary-Webb, PhD, often uses military examples to illustrate the indiscriminate effects of environment. “You have whites and blacks over in Iraq fighting a war. When they come back home, the rates of whites with PTSD and blacks with PTSD are the same. Because they’ve been in the same environment together,” says Thorpe.
“You can’t make someone who is black into someone white. But you can modify environments,” says health policy expert Sara Bleich.
LaVeist saw EHDIC as a prototype and a gamble: To propose research that might contradict long-held notions that environment, not biology, was the root cause of health disparities could have career-damaging consequences if the numbers didn’t pan out. “It’s risky to think outside the box. It truly is. Suppose I do the study and it shows disparities are no different than in the previously done national studies? How do I even get that published?” asks LaVeist. He chose to attack the problem with a meticulous, conservative methodology. His goal was to mimic NHIS’ data collection rigor to safeguard his results.
“It’s always been my practice to use very traditional research methods to study nontraditional, unconventional questions,” he says. “If I use the same techniques you are using, then you might not like my results, but to say my study is invalid, you’d have to say your own study is invalid.”
Originally, LaVeist thought he’d have to leave Baltimore to find a census tract that would meet his methodological requirements, but then the area sandwiched between Washington Village and the Union Square area popped up on the radar. With nearly equal levels of blacks and whites (51 percent and 44 percent), and nearly identical median incomes ($24,002) and high school graduation rates (just over 20 percent), the area was fertile territory for data mining.
In the summer of 2003, with funding from the National Center for Minority Health and Health Disparities and the Pfizer Corporation, surveyors canvassed the area. Each of the 1,489 respondents was asked identical health behavior and status questions as on the 2003 NHIS questionnaire, along with other psychosocial queries designed to measure potential stressors including perceived race and gender discrimination, access to social support, and feelings about medical care. Environment, including general conditions of homes and nearby vacant properties, was noted, as were physical factors such as blood pressure, body mass index, cigarette smoking and alcohol consumption.
For three years, LaVeist and his team crunched the numbers. In 2007 they published their first findings in the Journal of Urban Health. The results turned some of NHIS’ data inside out. When it came to rates of obesity, inactivity and drinking, blacks and whites in the survey area were the same. These findings contradicted the national data that adjusted for socioeconomic status; they showed blacks to have significantly higher odds of being obese and significantly lower odds of drinking.
Since then, additional studies have expanded on EHDIC’s original findings. In 2008, Roland Thorpe found that the difference in hypertension rates between blacks and whites examined in EHDIC was nearly a third lower than that measured by the National Health and Nutrition Examination Survey that adjusted for socioeconomic status. In a 2009 Journal of General Internal Medicine article, LaVeist, Thorpe and others used EHDIC to show that rates of diabetes in blacks and whites were comparable, whereas NHIS showed that African Americans were far more likely than whites to contract the disease. Obesity rates comparing African-American and Caucasian women in the EHDIC study are also being culled by Sara Bleich, PhD, an assistant professor in Health Policy and Management, to see if the trend toward shrinking racial disparities holds true.
LaVeist sees these studies as moving toward a critical mass that could tilt possible policy solutions away from medical investigations of inherent biological differences between races, a concept LaVeist vehemently rejects. He agrees with colleague Darrell Gaskin, who uses sickle cell anemia as an example of a disease that many view as an illness solely affecting African Americans. In fact, it’s a regional disease found in parts of the world where the sickle cell trait developed to ward off malaria. This includes Greece, South America and other areas outside of Africa. “There are whites with sickle cell disease,” notes LaVeist.
The medicalization of health disparities grew out of national studies completed prior to EHDIC, he says. When those studies had been adjusted for socioeconomic differences and found large racial disparities in health status, LaVeist says the medical conclusion was, “It must be genetic.”
“But what is that gene that produces these outcomes?” he asks. “The solution isn’t isolating some gene that’s somehow producing diabetes and heart disease and obesity and stroke and homicide … all of which to me seems unlikely. Maybe the solution is that we need to understand the social and perhaps the behavioral factors that are accounting for these differences.”
Josef Coresh, MD, PhD ’92, MHS ’92, a cardiovascular epidemiologist who has focused on genetic and non-genetic biomarkers of disease, sees LaVeist’s work as complementing biomarker studies. As an example, he cites extensive efforts to parse the causes of the up to four-fold higher rates of kidney failure among African Americans, a line of research begun more than 20 years ago by Paul Whelton, MD, and Dean Michael J. Klag, MD, MPH ’87. Linda Kao, PhD ’99, MHS ’97, an associate professor in Epidemiology, and others discovered that a mutation in the MYH9 gene made people twice as likely to develop non-diabetic kidney failure. African Americans were far more likely to have the mutation than whites (60 percent vs. 4 percent). Overall, this difference is estimated to account for 70 percent of the excess risk of non-diabetic kidney failure in African Americans.
While those findings might suggest medical interventions including genetic screening targeting African Americans, Coresh notes that a tremendous amount of kidney disease might still be related to neighborhood factors that interest LaVeist. These include lack of access to health care that could control hypertension and blood glucose levels. In fact, when it comes to diabetic-related kidney disease, Coresh says epidemiologic studies suggest that nearly all of the excess risk is accounted for by conditions directly related to how people live, and not their genes.
“I think it’s absolutely true that environment and particularly the impact of poverty and access to care are very important to study,” says Coresh of LaVeist’s work, adding that these environmental issues could also trigger disease among both the genetically susceptible and the general population.
LaVeist’s message and methodologies are being carried forth by the many colleagues he’s mentored. Some, such as Thorpe, note that LaVeist’s research dovetails with the past work of another Hopkins researcher, Marsha Lillie-Blanton, DrPH ’88, MHS ’82. In 1993, Lillie-Blanton found that crack cocaine, while far more available in poor African-American communities, was actually used almost equally, on a percentage basis, by blacks and whites within those blighted borders. So much for the common perception that crack was solely an African-American issue.
For EHDIC team members, their research is not just a single study but a prototype for similar efforts and a model for giving back to the communities being surveyed.
Given the growing pool of data, Sara Bleich says that focusing on race as the reason people suffer health disparities is, from a programmatic viewpoint, a nonstarter. “Maybe the difference we observe between blacks and whites regarding, say, hypertension, isn’t a race story but where they live. What that implies is something hugely different for public policy,” Bleich says. She notes that even if disparities turn out to have both biological and environmental components, it may make more sense to focus time and money on the latter. “You can’t make someone who is black into someone white. But you can modify environments. It’s very difficult to do, obviously, but EHDIC and similar studies create policy levers that didn’t exist before that are actually viable,” she says.
Tiffany Gary-Webb, who is now an associate professor at Columbia University researching the social determinants behind diabetes, said she was stunned by the data and its potential implications.
“I made Tom look at the data every-which-way because I had never seen a study where there was no disparity in diabetes between whites and African Americans. It goes to show that if you get people living in similar environments, they’ll have some of the same problems,” she says.
LaVeist isn’t shy about his hopes for EHDIC. He sees it as a prototype for similar studies in other integrated census tracts around the country, and not just in the results they generate. One component of the work is dedicated to giving back to the communities being surveyed, the opposite of parachuting into a community, collecting data and disappearing.
The EHDIC team, for example, published The Southwest Baltimore Community Health Report and mailed it to every active address in the community. EHDIC also helped establish a community center for health and wellness where, LaVeist says, “the programs don’t take a race approach. They take a community approach.”
To address gangs, drugs, smoking and health care issues raised in the EHDIC report, the Faith Center for Community Wellness and Advancement has become a one-stop shop, running everything from smoking cessation and heart-healthy programs to a makeshift gymnasium—all under one small roof. One afternoon this spring, while teenagers pound weights upstairs, 53-year-old Cecelia Battaglia talks about how the Faith Center’s programs were helping her and her family. Battaglia, who has a weakened heart from a childhood bout with rheumatic fever, says that despite being familiar with her condition, she learned “things I never knew before,” such as the need for regular exercise versus the occasional long walk. She’s also visited the Center’s farmers market to buy fresh produce, something not often seen in her neighborhood. “It was great. The kids were just going up and buying it and eating it,” Battaglia says. “When’s the last time you saw a kid buy a tangerine and eat it?”
LaVeist’s team will head next to Prince George’s County, Maryland, which has an annual median income of $86,000. They intend to compare health disparities between upper-income African Americans and whites. LaVeist wants to see whether his findings from lower-income areas hold, or if there are health stressors particular to African Americans who achieve the conventional notion of success, and whether they perceive their environment in the same manner as their white neighbors.
Ultimately, LaVeist hopes to use his meticulously collected scientific data to influence the decision-makers, be they on hospital boards or Capitol Hill.
“This is all about policy,” he says. “I’ve got this policy paper in my head that I want to write, but I don’t want to write it until I’ve done enough of these studies that the evidence is so overwhelming that it becomes difficult to debate. But eventually they’ll be a book or a series of policy papers that say, when it comes to health disparities, we have to look at social factors.”