TABLE 2

Associations of four SNPs with proinsulin/insulin ratio at fasting state (Proins0/Ins0), during 0–30 min (ProinsAUC0–30/InsAUC0–30), 30–120 min (ProinsAUC30–120/InsAUC30–120), and 0–120 min (ProinsAUC0–120/InsAUC0–120) of an OGTT in nondiabetic subjects

Gene SNPAlleles MAF (%)Proins0/Ins0
ProinsAUC0–30/InsAUC0–30
ProinsAUC30–120/InsAUC30–120
ProinsAUC0–120/InsAUC0–120
Effect size B (SE)P values*P valuesEffect size B (SE)P values*P valuesEffect size B (SE)P values*P valuesEffect size B (SE)P values*P values
TCF7L2 rs7903146C/T 17.71.20 (1.22)0.0420.0210.75 (0.42)0.0026.0E-040.55 (0.44)0.0051.1E-030.57 (0.43)0.0040.001
SLC30A8 rs13266634C/T 39.11.59 (0.96)0.0060.0030.73 (0.33)1.9E-051.2E-050.64 (0.35)1.1E-044.2E-050.64 (0.34)8.2E-052.8E-05
HHEX rs1111875C/T 46.90.74 (0.94)0.3650.6220.80 (0.32)9.7E-066.5E-060.69 (0.34)0.0020.0020.71 (0.33)0.0016.6E-04
CDKAL1 rs7754840G/C 37.0−0.39 (0.98)0.3130.7750.32 (0.34)3.1E-040.0010.36 (0.35)0.0030.0090.35 (0.35)0.0020.005
  • Effect size shown is B-coefficient (SE) per copy of the type 2 diabetes risk allele and was calculated using untransformed variables adjusted for age by linear regression. P values were calculated using log-transformed variables (because of their skewed distribution) by linear regression. P values are adjusted for age;

  • *P values are adjusted for age, BMI, and Matsuda ISI. In the entire cohort, means ± SE of examined parameters and the number of subjects with available data were as follows: Proins0/Ins0 36.3 ± 0.67 (n = 2,712), ProinsAUC0–30/InsAUC0–30 12.5 ± 0.23 (n = 2,697), ProinsAUC30–120/InsAUC30–120 14.1 ± 0.24 (n = 2,693), ProinsAUC0–120/InsAUC0–120 13.8 ± 0.24 (n = 2,692). P values significant after correction for multiple testing (P < 6.9 × 10−4) are in bold. Risk alleles are underlined. Results for the additive model are presented.