Background: TCF7L2 is the strongest locus linked to type 2 diabetes that has been identified thus far, and rs7903146 is the most significantly associated variant. Few intervention studies have shown that it has negative effects on metabolic improvement after lifestyle programs.
Objective: Our objective was to assess the effects of this variant on lifestyle intervention-induced changes in glucose values and metabolic variables at 1- and 4-y follow-ups.
Design: The rs7903146 variant was genotyped in 335 nondiabetic, dysmetabolic participants in a randomized lifestyle intervention trial.
Results: Subjects with the unfavorable TT genotype showed higher values of fasting glucose and lower homeostasis model assessment of beta cell function at baseline. Lifestyle modifications were successful in the amelioration of metabolic traits in all genetic subgroups after 1 y. At 4-y follow-up most of the metabolic benefits had disappeared. In a multiple regression model, values for glucose and homeostasis model assessment of beta cell function at 4 y were significantly associated with the T allele (for glucose and homeostasis model assessments, respectively: beta = 6.6; 95% CI: 2.5, 10.7; P = 0.001; and beta = -0.37; 95% CI: -0.54, -0.20; P < 0.001) but not with intervention. There was no interaction between genotype and intervention. After 1 y, impaired fasting glucose and diabetes incidence were inversely associated with intervention. After 4 y, the presence of a T allele was associated with impaired fasting glucose (odds ratio: 3.04; 95% CI: 1.53, 6.04; P = 0.001) and diabetes (odds ratio: 2.63; 95% CI: 1.00, 6.96; P = 0.05) but not with intervention.
Conclusions: Lifestyle modifications improved the metabolic pattern in all genetic subgroups. At the end of the trial, however, weight gain occurred, and carriers of the T allele developed first hyperglycemia and decreased insulin secretion, which suggests the need for different "after-care" preventive approaches tailored to each genotype's metabolic risk.