Read your latest personalised notifications
No account yet? Start here
Don't miss out
Ok, got it
Prof. Brian Ference
Both obesity, as measured by body mass index (BMI), and polygenic predisposition are strong risk factors for the development of diabetes. A Late-Breaking Science presentation today by Professor Brian A. Ference (University of Cambridge, UK and University of Milan, Italy) sought to answer two important questions: whether combining BMI with a polygenic risk score (PGS) can identify people at the highest risk of developing diabetes and whether increased BMI has a cumulative effect or threshold effect on the risk of developing diabetes.
This study included 445,765 participants from the UK Biobank, of whom 31,298 developed diabetes after age 25 years. A PGS composed of 2,137,820 variants for diabetes was constructed and this was used to divide the population into quintiles to quantify the effect of increasing polygenic predisposition on risk of diabetes. A stepwise increase in the risk of diabetes with increasing PGS was observed such that participants in the highest as compared with the lowest PGS quintile had a hazard ratio (HR) for diabetes of 2.99 (95% confidence interval [CI] 2.90–3.14; p<0.001). There was also a stepwise increase in the risk of diabetes with increasing BMI. Participants in the highest BMI quintile (mean BMI 35 kg/m2) as compared with the lowest (mean BMI 22 kg/m2) had a HR for diabetes of 11.42 (95% CI 10.81–12.07; p<0.001).
The investigators then evaluated how much the risk of diabetes varied at each level of the PGS depending on differences in BMI. The effects of PGS and BMI on the risk of diabetes were independent and additive. Within each quintile of PGS, the risk of diabetes varied by more than 10fold depending on differences in BMI. As a result, participants in the lowest PGS quintile with high BMI had a much greater risk of diabetes than participants in the highest PGS quintile with low BMI.
“These findings indicate that BMI is a much more powerful risk factor for diabetes than polygenic predisposition, but a PGS can moderately improve estimates of lifetime risk of diabetes at all levels of BMI,” said Prof. Ference.
Next, the investigators sought to determine if elevated BMI has a cumulative or threshold effect on the risk of developing diabetes. To conduct this analysis, the investigators constructed a genetic score for BMI composed of 255 variants independently associated with BMI at the genome-wide level of significance to assess the effect of lifetime exposure to increased BMI on the risk of diabetes. They found that the effect of a one unit increase in lifetime exposure to BMI on the risk of diabetes in Mendelian randomisation analyses was approximately the same as a one unit increase in BMI measured in middle life in observational analyses, suggesting that lifetime exposure to increased BMI does not have a cumulative effect on the risk of developing diabetes.
Prof. Ference explained, “BMI appears to have a threshold effect rather than a cumulative effect on the risk of diabetes – the BMI level at which a person develops insulin resistance and hyperglycaemia is their diabetes threshold.” He continued, “The findings indicate that most cases of diabetes could be avoided, or reversed, by keeping BMI below the cut-off which triggers insulin resistance in each individual.”
“Both BMI and blood glucose level should be assessed regularly to prevent diabetes,” Prof. Ference concluded. “Furthermore, efforts to lose weight are critical when a person starts to develop hyperglycaemia as it may be possible to reverse diabetes by losing weight before permanent damage occurs.”
Our mission: To reduce the burden of cardiovascular disease.
© 2020 European Society of Cardiology. All rights reserved.