Inequalities in life expectancy among Americans of different races and locations are well known. Hoping to bring the geographic variation of such disparities into clearer focus, researchers used census data and national death reports to calculate the death rate for each racial group in a given U.S. county. They combined these race-county population units based on similarities in population density, per capita income and murder rate. They settled on eight categories: blacks were split into either high-risk urban, rural Southern poor or "middle America" subgroups; the other categories were Asian, western Native American, rural Appalachian and Mississippi Valley white as well as white middle America, which is by far the largest group with 214 million members.
Life expectancy (as of 2001) varied among groups by as much as 20.7 years--the gap separating the 3.4 million high-risk black men from the 5.6 million Asian women. (See image above for life expectancies of black men and women by region.) Compared to Asian men, the life expectancy of the same high-risk black group is still 15.4 years less. The gradations in expectancy have held constant since 1982, the researchers found. "If you take pretty well-defined subgroups of the population, the inequalities are enormous," observes team member Majid Ezzati of the Harvard School of Public Health.
The disparities in death rate were highest for those between the ages of 15 and 64, who are not the traditional focus of the US health system, the group notes in their paper published online September 12 in PLoS Medicine. Among the adults, differences in rates of injury were a significant cause of the split. Other contributors were cardiovascular disease and various noninfectious diseases such as diabetes and liver cirrhosis. Given the well known risk factors for these ailments, such as smoking, alcohol use and obesity, "we really should be focusing on getting interventions to the people who most need them," Ezzati says.
Geography is not the whole story, though. "You have to go beyond simply identifying the characteristics of individuals and think about why this distribution occurs," says epidemiologist George Kaplan of the University of Michigan. "And that may have to do with social and economic policies."