New genes linked to diabetes
“Scientists have found six new genes linked to type 2 diabetes”, reports The Guardian today. It goes on to say that the discovery will improve understanding of how the disease develops. The Times also covers the story, saying that the study has discovered an increased risk of diabetes with several gene variants, one of which has previously been associated with a reduction in the risk of prostate cancer.
The study behind these reports is a meta-analysis of three genome-wide association studies, the results of which have been replicated in other populations. It provides good evidence that there are other gene variants that can increase a person’s susceptibility to type 2 diabetes, a condition associated with increasing age and obesity that is characterised by the body becoming resistant to insulin.
More research is needed before these findings can be translated into tools for diagnosis or improved treatments. It should be understood that these gene variants increase susceptibility to the disease, but they do not cause it. There are a host of other factors, including environmental ones, at work in the development of type 2 diabetes.
Where did the story come from?
Dr Eleftheria Zeggini and colleagues from the Diabetes Genetics Replication and Meta-analysis (DIAGRAM) Consortium based at Oxford University, the University of Michigan, Massachusetts General Hospital and Harvard Medical School carried out this meta-analysis. The study was published in Nature Genetics, a peer-reviewed medical journal.
What kind of scientific study was this?
For this publication, the researchers carried out a meta-analysis of three previous genome-wide association studies that investigated the links between certain gene variants and type 2 diabetes. The three studies were the Diabetes Genetics Initiative (DGI), the Finland-United States Investigation of NIDDM Genetics (FUSION) and the Wellcome Trust Case Control Consortium (WTCCC). Based on this pooling, 10,128 people and over 2.2 million gene variants were available for analysis.
As the purpose of the meta-analysis was to identify previously unknown gene variants associated with type 2 diabetes, the researchers excluded variants (and variants near these genes) that have previously been associated with the disease.
Genome-wide association studies have a drawback in that, independently, they can have a limited ability to detect some small associations between variants and disease. By combining the three studies, the researchers addressed this limitation and their analysis had greater “power” (i.e. was more likely to pick up an assocation if there was one) to identify additional variants than the individual studies.
To confirm their findings, the researchers investigated the significant links identified in their first meta-analysis in an additional 20,000 people whose data was available from the original three studies. Those links that were statistically significant at this stage were then investigated further using the pooled results from 10 other studies (over 57,000 additional people).
What were the results of the study?
By combining the populations in previous studies the researchers identified six previously unknown gene variants that had a significant association with type 2 diabetes. The researchers say that further sequencing and mapping is needed to identify exactly where these gene variants are, though they give some indication of which genes they are near.
The strongest evidence for an association was with a variant in a non-coding region of a gene called JAZF1. The researchers say that another variant in this same gene has been associated with prostate cancer. Overall, people who had this particular variant were 1.1 times (95% CI 1.07 to 1.13) more likely to have type 2 diabetes. The other five variants were also significantly associated with an increased likelihood of having type 2 diabetes.
What interpretations did the researchers draw from these results?
The researchers conclude that they have detected at least six ”previously unknown loci with robust evidence for association” with type 2 diabetes. Additionally, they say their results show that there is value in using their meta-analytic approach to gain further insights into the inherited basis of type 2 diabetes.
What does the NHS Knowledge Service make of this study?
This is a well-conducted study that combined research using recognised methods in this field. By combining the results of other studies, the researchers have increased the power of existing data to detect previously unknown relationships between gene variants and type 2 diabetes. The researchers were able to show that the associations they felt were important had a similar pattern of association in separate populations.
Aspects of this study to consider include:
- For each of the new variants they identify, the researchers discuss a potential biological reason why the variant might have an effect on susceptibility to diabetes.
- As with all meta-analyses, some bias can be introduced in the selection of studies that are combined. which is why a systematic approach is the best. It is unclear from the research paper why the researchers chose the three studies that were used.
- The Times article focused predominantly on the ‘trade-off’ between type 2 diabetes and prostate cancer based on the presence of one gene variant, but the study did not investigate prostate cancer. It is not possible to draw conclusions about the relationship between any of the gene variants and prostate cancer on the basis of this study.
Importantly, even if these types of study were to identify all possible gene variants associated with type 2 diabetes, they do not mean that any one person with a particular gene variant will develop the disease. There are a host of other factors that are related to risk, including body mass index (BMI), cholesterol and blood pressure.
Further studies are needed before these findings are translated to technologies that could aid the treatment or diagnosis of disease. The researchers also say that more in-depth research and mapping is needed to identify exactly where the variants are situated.
