Drug Discovery
Computer Screening Uncovers Compounds Against ALS

Neuroscience
Optic Nerve Regrown in Mice

Public Health
Stats Tool Puts Health Disparities on the Map

Statistical Method Detects Disease Outbreaks Without Background Population Data

Molecular Modulator of Synaptic Plasticity Revealed

Hot Spots for Genetic Recombination Different in Chimps and Humans

Proceedings of the HMS Faculty Council

New Appointments to Full Professor

Nominations Invited for Biostatistics Alum Award

Grants Offered in Women’s Health

Honors and Advances

Madras Receives Public Service Award

Students Bag Pharma in Second Year Show

Front Page

Stats Tool Puts Health Disparities on the Map

Census Tracts Reflect Socioeconomic Effects on Health Status

Money may not buy happiness, but it can buy health. From infancy to senility, evidence shows, the risk of disease and early death is directly proportional to people’s income, employment, and education. According to a new study, a person’s home address turns out to be a reasonably good stand-in for the powerful health effects of socioeconomic resources.

Your home address reveals much about your risk of disease and early death, say (clockwise from top left) Nancy Krieger, Jarvis Chen, S.V. Subramanian (top inset), David Rehkopf (bottom inset), Pamela Waterman, and Lorna Haughton. (Photo by Graham Ramsay; top inset photo courtesy of Harvard School of Public Health; bottom inset photo by Pamela Waterman)

People who live in neighborhoods with the fewest poor people are healthier by most measures, while people who live in areas with the most poor people are the sickest and most likely to die prematurely, say HSPH researchers who have developed a new tool to track such socioeconomic health disparities.

The tool matches health outcomes data collected by public health systems to geography—census tracts, in this case—for which the federal government collects economic data. The process uses a technique called geocoding, which links data coded to the same geographic location and may be as potent a prognosticator of health as someone’s genetic code.

“We’re taking two sources of data that otherwise weren’t talking, and getting much more of the story,” said Nancy Krieger, HSPH associate professor of society, human development, and health, who led the team that developed the method. “This is a fundamental piece of information. We need to know where the problems are and what needs to be addressed to reduce these disparities in health.”

Missing Data
Dozens of research studies in this country and others have documented the social gradient of health and wealth. The link has persisted for more than a century, even though the things that kill people and make them sick have changed. Socioeconomic inequality also seems to contribute mightily to the well-documented racial and ethnic disparities in health. Yet most public health surveillance systems in this country do not routinely monitor socioeconomic factors, even when they collect racial and ethnic information, Krieger said.

Geocoding units. The percentage of persons living below the poverty line in a census tract matched key health indicators better than block groups or ZIP codes, says the final report of the Public Health Disparities Geocoding Project. The census tract, block group, and ZIP code for Gordon Hall at 25 Shattuck St. are shown above. (Adapted by Rachel Meyer)

“This failure to include socioeconomic data severely impedes efforts to understand, routinely monitor, and address social disparities in health in the United States,” Krieger and her colleagues write in the February American Journal of Public Health. Their paper is the culmination of two years of rigorous searching for an easy, standardized method for public health departments to compile socioeconomic data, compare it on small or large scales, and track it over time.

Krieger and her colleagues tested and rejected several geographic levels, settling on the census tract. Of the wealth of census tract socioeconomic data—median household income, education, home price, occupation, unemployment, number of people in a household, and combinations thereof—the poverty level proved to be both a sensitive and simple measure that performed on a par with more complex economic indicators.

A Deadly Combination Of Poverty and Race Living in a poor neighborhood may be bad for someone’s health, but the most impoverished neighborhoods in Massachusetts are deadliest to blacks, HSPH researchers report in the February American Journal of Public Health. “If you are black and living in a high-poverty area, you are three times more likely to die compared to whites living in a low-poverty area and 1.5 times more likely to die than your white neighbor,” said first author S.V. Subramanian. Neighborhood environments matter much more for blacks than whites, Subramanian found. The census-tract distribution of poverty corresponds closely with the distribution of excess death among blacks. The results come from a multilevel statistical analysis of the same data set developed by Nancy Krieger, Subramanian, and their colleagues to enable routine monitoring of socioeconomic disparities in health. Subramanian wanted to learn how neighborhoods influence individual health. He plans to follow up by mapping and identifying neighborhoods that are unusually harmful or particularly protective for different racial and ethnic groups.

For the 1990 census data the researchers used to develop and validate their method, the poverty line was $12,647 for a family of two adults and two children. The poorest census tracts were those with more than 20 percent of people living below the poverty line, and the wealthiest census tracts had fewer than five percent of people living below the poverty line.

To show the feasibility and compelling results of the geocoding tool, Krieger and her colleagues drew on the public health surveillance systems of Massachusetts and Rhode Island. The researchers believe it is the first such statewide analysis of health across multiple outcomes, from birth to death, stratified by a consistent socioeconomic measure, as well as by race/ethnicity and gender.

Burden of Disease
People who live in neighborhoods with the highest poverty levels had more heart disease, cancer, diabetes, sexually transmitted diseases, childhood lead poisoning, low–birth weight infants, gunshot wounds, tuberculosis, homicides, and other common indicators of public health, Krieger and her colleagues found. The only exceptions were the proportionally higher rates of breast and prostate cancers in wealthier areas and consistent rates of colon cancer across the socioeconomic spectrum.

For almost all of the health measures, adjusting for the home address halved the excess risk observed among blacks and Hispanics compared with whites. But the absolute magnitude of the socioeconomic effect on racial and ethnic health disparities may be affected by a suspected census undercount of minorities, the researchers cautioned.

They speculated that a complex combination of three factors is at work in shaping the observed associations between people’s health and their neighborhoods: people in poor areas have poor health because poor individuals have poor health; a concentration of poverty creates or exacerbates harmful social interactions; and poor areas may have fewer public goods, such as supermarkets and health clinics, and more environmental pollution. The socioeconomic data give other researchers a tool to tease out the features of poverty that are especially harmful to health (see box below).

“Your residential location can play a critical role in making or breaking your health,” said S.V. Subramanian, HSPH assistant professor of society, human development, and health, a co-author on the paper.

The methods are freely available on the Public Health Disparities Geocoding Project website at www.hsph.harvard.edu/thegeocodingproject. In the next two years, Krieger and her colleagues will be holding workshops to train about 100 people from health departments and universities to apply the new tools.

“This failure to include socioeconomic data severely impedes efforts to understand, routinely monitor, and address social disparities in health in the United States.”

The Washington State Department of Health used the technique to produce the 2004 Supplement to the 2002 Health of Washington State. “We are trying to raise awareness that communities affect people and that health is not just individual choice,” Juliet VanEenwyk, the state epidemiologist who led the effort, wrote in an e-mail. “Policymakers in Washington are beginning to understand this, and our document contributes to the effort.”

“Everyone knows the general relationship between socioeconomic status and a variety of adverse health outcomes,” said Bruce Cohen, acting director of the Center for Health Information and Statistics at the Massachusetts Department of Public Health. “This tool helps us turn reams of data into something that is actionable. It refines our ability to identify high-risk populations and target programs. It has enormous implications for policy development.”

—Carol Cruzan Morton

top