Non-alcoholic fatty liver disease and impaired proinsulin conversion as newly identified predictors of the long-term non-response to a lifestyle intervention for diabetes prevention: Results from the TULIP study.
Aims/hypothesis Lifestyle intervention is effective to prevent type 2 diabetes. However, a considerable long-term non-response occurs to a standard lifestyle intervention. We investigated which risk phenotypes at baseline and their changes during the lifestyle intervention predict long-term glycaemic non-response to the intervention. Methods Of 300 participants at high risk for type 2 diabetes who participated in a 24 month lifestyle intervention with diet modification and increased physical activity, 190 participants could be re-examined after 8.7 +/- 1.6 years. All individuals underwent a five-point 75 g OGTT and measurements of body fat compartments and liver fat content with MRI and spectroscopy at baseline, 9 and 24 months during the lifestyle intervention, and at long-term follow-up. Fasting proinsulin to insulin conversion (PI/I ratio) and insulin sensitivity and secretion were calculated from the OGTT. Non-response to lifestyle intervention was defined as no decrease in glycaemia, i.e. no decrease in AUC for glucose at 0-120 min during OGTT (AUCglucose(0-120) (min)). Results Before the lifestyle intervention, 56% of participants had normal glucose regulation and 44% individuals had impaired fasting glucose and/or impaired glucose tolerance. At long-term follow-up, 11% had developed diabetes. Multivariable regression analysis with adjustment for age, sex, BMI and change in BMI during the lifestyle intervention revealed that baseline insulin secretion and insulin sensitivity, as well as change in insulin sensitivity during the lifestyle intervention, predicted long-term glycaemic control after 9 years. In addition, increased hepatic lipid content as well as impaired fasting proinsulin conversion at baseline were newly detected phenotypes that independently predicted long-term glycaemic control. Conclusions/interpretation Increased hepatic lipid content and impaired proinsulin conversion are new predictors, independent of change in body weight, for non-response to lifestyle intervention in addition to the confirmed factors, impaired insulin secretion and insulin sensitivity.