Genes and diet in obesity-related type 2 diabetes

Abstract

The prevalence of obesity, defined as a body mass index (BMI) of ?30, and type 2 diabetes is rising rapidly worldwide. It is therefore important to study the underlying, etiologic mechanisms of obesity and type 2 diabetes to gain insight into their development, which could eventually lead to preventive strategies and we have studied genetic variation and dietary intake in relation to adiposity, obesity and type 2 diabetes. We also investigated the value of adding genetic variation to non-genetic risk factors for type 2 diabetes. We described the relation between single nucleotide polymorphisms (SNPs) of the protein tyrosine phosphatase nonreceptor type 1 (PTPN1) gene and macronutrient intake, measures of adiposity, including visceral fat, and cholesterol. We found no statistically significant associations for SNPs in the PTPN1 gene with any dietary phenotypes or measures of obesity. However, our results suggest that SNPs in the PTPN1 gene are associated with total plasma and LDL cholesterol levels. We confirmed some of the findings for 12 obesity loci associated with general adiposity in healthy Dutch women. Seven SNPs were associated (p<0.05) with weight, BMI, and waist circumference (unadjusted for BMI). They were in or near six loci: FTO, MC4R, KCTD15, MTCH2, NEGR1, and BDNF. Five of the SNPs were associated with dietary intake (p<0.05), which were in or near five loci: SH2B1 (particularly with increased fat), KCTD15 (particularly with carbohydrate intake), MTCH2, NEGR1, and BDNF. Our results suggested that these loci are not specifically associated with abdominal adiposity, but more generally with obesity and some of the SNPs were associated with macronutrient-specific food intake. We aimed to clarify the role of dietary patterns in the development of type 2 diabetes in overweight and obese individuals. Two main dietary patterns were defined: a Mediterranean-like pattern and a Snack pattern. Scoring on the Mediterranean-like pattern was not associated with type 2 diabetes risk, whereas a high score on the Snack pattern was associated with a higher risk of type 2 diabetes (HR Q4 versus Q1 (95%CI): 1.70 (1.31; 2.20), ptrend=<0.0001), particularly among less active individuals (less active: HR Q4 versus Q1 (95% CI): 2.14 (1.48; 3.09), ptrend=0.00004, more active: HR Q4 versus Q1 (95%CI): 1.35 (0.93; 1.97), ptrend=0.01; pinteraction=0.02). Our findings support the view that a dietary pattern quantitatively and qualitatively characterized by snacking increases the risk of type 2 diabetes in overweight and obese subjects, especially in physically less active subjects. At last, we investigated the added value of genetic risk factors in predicting type 2 diabetes according to the latest methodological standards of prediction research. Comparison of models 1a (only clinical variables) and 1b (clinical and genetic variables) resulted in a net reclassification improvement (NRI) of 4.6% (p=0.019). For models 2a (only clinical variables) and 2b (clinical and genetic variables), the NRI was 3.6% (p=0.031). In conclusion, adding genetic variants to phenotype-based risk models for type 2 diabetes improves the reclassification in type 2 diabetes case prediction. However, the discrimination was scarcely improved.