Personalizing Diabetes Prevention: Is it Time to Focus on the Intervention?

Abstract

T here are approximately 86 million adults with prediabetes in the United States, 1 and an estimated 70 % of these individuals will develop type 2 diabetes during their lifetime. Several clinical trials have shown a remarkable reduction of diabetes incidence with intensive lifestyle interventions, 2 but these behavioral changes are difficult to sustain. Less intensive trials testing the real-world effectiveness of behavioral interventions in various settings have found about half the effect on weight loss as the intensive trials. 3 As we enter this age of personalized medicine, could knowing one's genetic risk motivate patients to make and sustain effective behavioral changes?In this issue of JGIM, Voils and colleagues tested whether the addition of genetic risk results to standard diabetes risk estimates with brief lifestyle counseling would be more effective than standard diabetes risk estimates with lifestyle counseling alone for weight loss and behavioral outcomes among veterans. 4 In this well-designed randomized controlled trial conducted in a primary care setting, investigators analyzed three single nucleotide polymorphisms (SNPs) associated with diabetes risk to calculate the genetic risk for the intervention group and categorized their genetic risk as low, moderate, or high. They found that weight, insulin resistance, physical activity and perceived lifetime risk of diabetes did not differ between groups at 3 or 6 months. Daily caloric intake and fat intake were lower in the genetic-risk-informed group at 3 months, but there was no difference at 6 months. This result did not vary among those with high, moderate or low genetic risk in the intervention arm.A prior trial performed in another primary care setting found similarly unimpressive effects of providing genetic risk information from 36 diabetes risk SNPs to participants enrolled in a diabetes prevention program on short-term outcomes of motivation, program attendance or weight loss. 5 Other trials testing the effect of communication of genetic risk of disease on risk-reducing behaviors, such as smoking cessation or physical activity, have also been mostly negative. 6 Why was this the case for the present study? One explanation may be that the trial was underpowered to detect clinically significant effects. The sample size for this study was determined based on the weight loss difference observed between groups from an intensive diabetes prevention trial that required 300 people to be randomized per group. It is possible that a more modest weight loss outcome based on a less intensive intervention such as the proposed method would have required a larger sample size. While 601 participants were randomized, only 534 (89 %) attended the risk counseling session and only 84 % completed the primary endpoint evaluation at 3 months, which further diminished the study's power. Yet the real question is whether the knowledge of high (or moderate/high) genetic risk will motivate individuals to adopt healthier behaviors. The power to detect an interaction by gen...