The study, published on Monday in the journal Nature, puts to rest a decades-long debate over the genetics that influence the risk of diabetes, which affects one in 10 people over the course of their lifetime.
And it has identified more than a dozen specific genes directly involved in the development of type 2 diabetes that might serve as potential drug targets.
“There was a whole furious debate that arose about this,” said Dr. Francis Collins, director of the National Institutes of Health, one of more than 300 scientists collaborating on the work.
Prior studies turned up more than 80 spots in the genome associated with the development of adult-onset diabetes, but most of these genetic errors were common, meaning they occurred frequently in the population, and they explained only a small fraction of disease risk.
These discoveries were based on genome-wide association studies or GWAS, which used gene chips that scan thousands of genes at a time. Researchers used these to scan DNA from large populations of individuals with a specific disease and compare them with DNA from similar groups of healthy people.
Critics, including geneticist Dr. David Goldstein at Columbia University, argued that such studies were a waste of resources because they only found common variants that explained just a small fraction of the risk for disease.
He said the really important drivers of common diseases such as diabetes and schizophrenia were more likely to be found in extremely rare genes, those occurring in individuals or in families, not those shared by large populations of people.
Goldstein “argued very persuasively that it was all about rare variants and we were all going down the wrong road looking at the common ones,” Collins said in a telephone interview.
The new study took a deeper look, using next-generation sequencing to search the entire genetic code of 2,657 people with and without diabetes to assess the contribution of both rare and common genes driving diabetes.
They also sequenced all of the protein-making genes in 12,940 people, and used statistical methods to estimate risk in another 111,548 people with less complete DNA data.
They found that, indeed, most of the genetic risk for type 2 diabetes is caused by common mistakes in the genetic code, with each mistake contributing only a small portion of an individual’s risk for developing the disease.
“What this study says quite definitively for diabetes is the vast majority of hereditary risk variants are in fact these common ones, and the rare ones, while they pop up here and there, are a much smaller contribution,” Collins said.
The study also turned up more than a dozen examples where variants alter the way proteins are made, suggesting that these gene variants have some direct impact on the development of type 2 diabetes.
“These represent promising avenues for efforts to design new ways to treat or prevent the disease,” said Mark McCarthy, a senior author of the study from Oxford University.
All of the data will be made publicly available online through the Accelerating Medicines Partnership, a public-private partnership between the NIH, the U.S. Food and Drug Administration, 10 drug companies and several nonprofits.
Goldstein said the work was “a careful, solid investigation” that does not change his view much overall, adding that it was time to “quit arguing.”
“What I care about now is finding the exact variants that infer risk, and understanding how they do so,” he said.