Working with the island's health department, Dr. Friedman has genotyped 3,300 people - most of the population - expecting to find that the few residents to remain thin are descended from a small group of Europeans who arrived in the 19th century. But "we didn't find any genes that could predict who was obese and who wasn't," he says. "So, if you don't even have that as a starting point, there's no where else to go."
John Speakman, a zoologist at Scotland's University of Aberdeen, says there never was anywhere else to go. "The thrifty-gene hypothesis is just a story," he says. "As far as I'm concerned, it's dead and we should discontinue thinking about it. But revolutions in science don't happen overnight."
Given the weight of evidence against the theory, a post-mortem has already begun, raising uncomfortable questions about why, with no real proof, it has been so popular for so long.
Jennifer Poudrier, an associate professor of sociology at the University of Saskatchewan, is one of many scholars to argue that the theory allows society to curl up with the notion that biology shoulders most of the blame for the ill health of native people. She dismisses it as "a colonial lens put on aboriginal history," promoting the myth that indigenous people all have the same genes that make their diabetes "a special problem" beyond the reach of public-health initiatives.
"It makes people feel deflated and defeated," says Prof. Poudrier, a Métis who counsels native women on body image. "It detracts from a focus on the social context."
That context, she notes, includes conditions on Saskatchewan reserves where fast food, such as Kentucky Fried Chicken and pop, sell for a small fraction of the cost of milk and ingredients for a salad. Yet "diabetes is not seen as a social issue, but a biomedical one."
Jeff Reading, director of the Centre for Aboriginal Health Research at the University of Victoria, calls the theory "potentially damaging" because "it discounts the social circumstances of poverty - it suggests that you'd be okay if you didn't have faulty genes.
"But saying your genes are bad is not that helpful; it's just an excuse not to do anything about it."
He says indigenous groups in New Zealand, northern Finland and Canada have very different genetic histories, yet all suffer from similar social and health problems. "It's not about race. It's more about colonization."
Fresh veg just turnips and spuds
Sandy Lake was a ramshackle collection of government-issue, two-bedroom cottages when Stewart Harris arrived in 1990, some so dilapidated that "you wouldn't park your car" in them. Unemployment hovered at 80 per cent, running water was a novelty and potatoes and turnips were the only produce available.
Posted to the reserve as the region's medical director, Dr. Harris was the first to call attention to its rampant diabetes. He and Bernard Zinman, an expert at Mount Sinai Hospital in Toronto, formed a partnership with the community to set up nutrition and exercise programs as they tried to gauge the scope of the problem.
Once they had the shocking numbers in hand, they contacted Dr. Hegele, suspecting that genes, not just environmental forces, played a role. Four years later, they pinpointed the first diabetes-risk gene ever identified in an aboriginal population.
In their 1999 report, the researchers wrote that the variant seemed to fit Dr. Neel's theory of a thrifty gene - especially given the Oji-Cree once lived as hunters and survived hungry winters.
"But even then," Dr. Hegele says, "we knew the variant couldn't be the sole factor underlying diabetes. It wasn't even present in the majority of diabetic Oji-Cree." Only two of five band members carried it. Neither did it have any bearing on diabetes in populations outside Sandy Lake. "In those early days, I speculated that perhaps we would find additional variants that would fill the gap," Dr. Hegele says.