AIMS: To examine whether genetic variants that predispose individuals to type 2 diabetes (T2D) could predict the development of diabetes after gestational diabetes mellitus (GDM). METHODS: 13 SNPs (FTO rs8050136, CDKAL1 rs7754840 and rs7756992, CDKN2A/2B rs10811661, HHEX rs1111875, IGF2BP2 rs1470579 and rs4402960, SLC30A8 rs13266634, TCF7L2 rs7903146, PPARG rs1801282, GCK rs1799884, HNF1A rs1169288, and KCNJ11 rs5219) were genotyped in 793 women with GDM after a median follow-up of 57 months. RESULTS: After adjustment for age and ethnicity, the TCF7L2 rs7903146 and the FTO rs8050136 variants significantly predicted postpartum diabetes; hazard ratio (95% confidence interval 1.29 (1.01-1.66) and 1.36 (1.06-1.74), respectively (additive model) versus 1.45 (1.01-2.08) and 1.56 (1.06-2.29) (dominant model)). Adjusting for BMI attenuated the effect of the FTO variant, suggesting that the effect was mediated through its effect on BMI. Combining all risk alleles to a weighted risk score was significantly associated with the risk of postpartum diabetes (hazard ratio 1.11, 95% confidence interval 1.05-1.18, p=0.00016 after adjustment for age and ethnicity). CONCLUSIONS: The TCF7L2 rs7903146 and FTO rs8050136 polymorphisms, and particularly a weighted risk score of T2D risk alleles, predict diabetes after GDM. Further studies in other populations are needed to confirm our results.
AIMS: To examine whether genetic variants that predispose individuals to type 2 diabetes (T2D) could predict the development of diabetes after gestational diabetes mellitus (GDM). METHODS: 13 SNPs (FTOrs8050136, CDKAL1rs7754840 and rs7756992, CDKN2A/2B rs10811661, HHEXrs1111875, IGF2BP2rs1470579 and rs4402960, SLC30A8rs13266634, TCF7L2rs7903146, PPARGrs1801282, GCKrs1799884, HNF1Ars1169288, and KCNJ11rs5219) were genotyped in 793 women with GDM after a median follow-up of 57 months. RESULTS: After adjustment for age and ethnicity, the TCF7L2rs7903146 and the FTOrs8050136 variants significantly predicted postpartum diabetes; hazard ratio (95% confidence interval 1.29 (1.01-1.66) and 1.36 (1.06-1.74), respectively (additive model) versus 1.45 (1.01-2.08) and 1.56 (1.06-2.29) (dominant model)). Adjusting for BMI attenuated the effect of the FTO variant, suggesting that the effect was mediated through its effect on BMI. Combining all risk alleles to a weighted risk score was significantly associated with the risk of postpartum diabetes (hazard ratio 1.11, 95% confidence interval 1.05-1.18, p=0.00016 after adjustment for age and ethnicity). CONCLUSIONS: The TCF7L2rs7903146 and FTOrs8050136 polymorphisms, and particularly a weighted risk score of T2D risk alleles, predict diabetes after GDM. Further studies in other populations are needed to confirm our results.