Monday June 28 2010
Effective genetic screening for type 2 diabetes is some way off
“Doctors are closer to building a DNA profile of people who are at risk from diabetes after pinpointing another set of genes associated with the disease,” reported The Daily Telegraph.
The study behind the news report combined the results from several genome-wide association studies, which compared the DNA of thousands of people with type 2 diabetes to that from people without the disease. In addition to confirming several genetic variants that the previous study had associated with the disease, the research identified 12 new variants, bringing the number of linked genes to 38.
This well-conducted research has generally been accurately reported by the press. It furthers our understanding of the genetic variations that may increase the risk of type 2 diabetes. These findings are important, but having all of these gene variants does not mean a person will definitely develop the disease. This advance in unlikely to have any immediate implication for screening or prevention of type 2 diabetes. Being overweight or obese is a known modifiable risk factor for this disease.
Where did the story come from?
The study was carried out by researchers from a variety of international academic and medical institutions, including the University of Oxford and the Wellcome Trust Sanger Institute in the UK. The research was financially supported by many organisations, including the Academy of Finland, the American Diabetes Association, manufacturers and national research councils. The study was published in the peer-reviewed journal Nature Genetics.
The news coverage of this research is clear, explaining that the discovery increases the understanding of the genetic risk factors for developing type 2 diabetes.
What kind of research was this?
This was a meta-analysis (statistical pooling) of data from several genome-wide association studies. The researchers report that combining these studies resulted in twice as much data as that analysed in an earlier study some of the researchers had published, which was covered by Behind the Headlines in March 2008.
What did the research involve?
Researchers combined the data from eight studies for a total of 8,130 patients with type 2 diabetes and 38,987 controls. Combining data from a number of different studies increases the power of the research to detect associations between genetic variants and disease. In the first stage of this particular study, data from individual studies were combined to determine how common particular genetic variations were in people with type 2 diabetes.
As is common with these studies, the researchers then confirmed their findings in a separate population of cases and controls, in a total of 34,412 people with diabetes and 59,925 people without the disease. They were particularly interested in looking at genetic variations in regions of the DNA not previously associated with type 2 diabetes.
The researchers then carried out further data analyses to attempt to explain why all genetic variants discovered so far only account for about 10% of the observed clustering of the disease in families. Further explanation of this may give better understanding of how these genes are related to the disease. These analyses included looking for additional variants near to the ones identified in the first stage of the study. They also grouped the population of people with diabetes by other known risk factors including BMI and age of diagnosis.
What were the basic results?
In the initial meta-analysis part of the study, the researchers identified those genetic regions that were associated with disease risk (in 8,130 patients and 38,987 controls). They then confirmed whether the significant variants from this analysis were also associated with disease in a second independent sample (34,412 cases and 59,925 controls).
In a final step, they combined these two samples to increase their study power and identified 14 different variants that were associated with type 2 diabetes. Of these, two were previously known associations, and 12 were new associations that have been uncovered by this meta-analysis.
How did the researchers interpret the results?
The researchers say they have expanded the number of genetic variants known to be associated with type 2 diabetes. They note that together with those discovered in other studies, there are now 38 confirmed genetic variants associated with this disease.
This is well-conducted, well-reported research that confirmed its initial findings in an independent population, and used accepted methods for this field of research. The researchers discuss possible biological explanations for how each of the newly identified variants might affect disease susceptibility, Points to note include:
- Thirty-eight genetic variants associated with type 2 diabetes have now been confirmed. However, there are likely to be others and a large proportion of heritability in type 2 diabetes remains unexplained. The researchers say that increasing the size of this type of meta-analysis further could detect many more variants. However, it is probable that this may still not account for most of the heritability of the disease. This leads them and other researchers to believe that uncommon genetic variants may be playing a greater role. This will be the subject of further study.
- As with all meta-analyses, bias can be introduced when selecting the studies to be included and a systematic approach is the best way of avoiding this. However, it is not clear from the publication whether these researchers carried out systematic searches or not.
- Importantly, having all of these gene variants does not mean a person will definitely develop the disease. A number of non-genetic factors have been associated with the risk of type 2 diabetes, including being overweight or obese.
Before these findings can be translated into technologies that could assist with the treatment or diagnosis of type 2 diabetes, more research will be needed. The researchers acknowledge that further mapping of the genes as well as ‘functional studies in humans and in animal models’ are needed to further characterise the variants involved in the risk of type 2 diabetes.