To tease out the genetic component of obesity, Christman and Rotimi will be looking for shared single nucleotide polymorphisms (SNPs) among obese versus non-obese African Americans. It's a potentially daunting feat: there are an estimated 10 million common SNPs known to exist, and comparing each and every variation against that of every subject would be a computational nightmare. Therefore, the team will zero in on 20 specific genes associated with the diseases it plans to study, and will map the SNPs in those particular genomic regions. Eventually, researchers will compare their results with samples from the Framingham group, and they also hope to do similar comparisons with other ethnic and geographic populations.

"The main payoff of genetics is not being able to test people and predict who will become overweight," says Christman. "It's in understanding the pathways that influence human obesity, so if we understand what the genes are, the future holds promise for making somewhat more intelligent drug targets for obesity."

Troy Duster, a sociologist at New York University, offers a note of caution. Duster, author of Backdoor to Eugenics, warns that the study of the genetics of a specific race can be easily misunderstood. While the Howard University study is aiming to determine the reasons for the disproportionately high incidence of obesity in African Americas, he warns that the public can misinterpret it as, for example, a hunt for a "black obesity gene."

Down the line, Christman and Rotimi plan to perform similar genetic scans on other populations. The long-term goal, they say, is to treat patients according to the individual's genetic profile. Their current work will make this goal that much more attainable.