What the T Can Tell Us about Health Inequalities
A pictorial essay
Sandro Galea is dean of Boston University’s School of Public Health and a professor of epidemiology. Photo courtesy of Sandro Galea.
There are many lenses through which to examine the health of a community. For example, Massachusetts has about 315 doctors per 100,000 people—more than 10 percent higher than the next closest state (Maryland). This is in large part because of the remarkable density of physicians and trainees in Boston. Massachusetts also spends more on health care per person than any other state ($9,278 in 2009—36 percent above the national average of $6,815) and has the lowest percentage of residents without health insurance (4.4 percent).
All of this might suggest that Boston would be a tremendously healthy city, a paragon of urban health. And in many ways, it is. We have some of the highest life expectancy of any US city (about 81 years). But like many US cities, Boston also has some extraordinary disparities, both in health indicators and in the drivers of those indicators within its borders.
One way we can see those indicators—in Boston and other cities, such as New York and Chicago—is through a simple device: public transportation. In Boston, that means the T (short for MBTA).
Suppose we are riding the T and stopping at various stops: What does health look like at these stops, and what do the drivers of health look like? To illustrate these questions, below are a few representative T stops that can summarize what we see throughout the city.
We can start with a core health indicator: premature death rates per 1,000. The T map below shows that the premature death rate in the area around the Arlington stop, for example, is more than 50 percent lower than at Dudley Square station.
This is linked, in no small part, to health indicators like violence, with the map below showing how the homicide rate around the Dudley Square stop is eight times the rate around Arlington, and the homicide rate at the Mattapan stop is almost six times the rate at Fenway.
These differences extend well beyond health indicators among adults to other core indicators such as low birth weight, a marker for a substantial burden of poor health and disability later in life. In this case, the health differences are in the order of 25 percent when comparing Arlington and Fenway to Mattapan and Dudley.
It is not then surprising that there are similar differences, of similar magnitudes, for adult noncommunicable disease indicators such as diabetes, as shown in the map below.
These health indicators are inexorably linked to a broad range of social indicators that are unevenly distributed across the city. Poverty is a frequently used summary indicator of socioeconomic position, well established as a marker of a broad range of other adversities. It is not then surprising, given the maps above, that poverty rates in some parts of Boston are four to eight times higher than those surrounding the healthier stops on the T.
Other measures of socioeconomic position, such as education, track accordingly, with a graduate education being three or more times more common around the Arlington and Fenway stops than it is around Dudley Square, Mattapan, or Maverick stations.
And, these differences are associated with commensurately poor health behaviors, such as physical activity, which is substantially lower on the Red Line at Mattapan than it is on the Green Line at Fenway, for example. The question of whether socioeconomic differences in health are attributed in part, or at all, to differences in health behavior is, in and of itself, a difficult and complex question, and I would refer the reader to conflicting papers on the topic.
Inured as we are to inequalities in health, we might well shrug off these health differences as ones between far-apart worlds. But are they? In fact, the geographic space we are talking about here is remarkably small. We are dealing with geographic differences of roughly four miles, or about an hour’s walk. In many respects, it is remarkable that areas so close to one another should have such dramatically different health indicators—“health worlds apart” that are simply down the street from one another.
Rounding this out perhaps brings us back to the fundamental condition of Boston that we discussed earlier—the incredible density of physicians and hospitals throughout the city. It is therefore not surprising that none of the T stops we are discussing are particularly far from medical facilities. Clearly, medical centers differ in terms of populations served and variations in availability of care, but as the map below shows (and as can be verified through much more thorough analysis), there are negligible differences in physical distance of these neighborhoods to quality medical care.
This, then, tells a story of a city richly characterized by top-of-the-line medical resources and overall health indicators that are enviably good, but that has, within it, substantial heterogeneity in those same health indicators, associated in large part with variation in the fundamental socioeconomic circumstances that produce health in populations. The challenge to public health is apparent and vivid—how do we contribute to the generation of knowledge that can bridge these health gaps, and to the creation of conditions that produce health not just for some, but for all, across a city like Boston?
Author, Sandro Galea is the dean of the School of Public Health. He can be reached at sgalea@bu.edu.
A version of this article originally appeared on the BU School of Public Health website.