Table 2.
Association of PPARG and HHEX–IDE in type 1 diabetes families
| Gene region | SNP | Families, n | Allele or genotype | Transmitteda, n (%) | Not transmitteda, n (%) | RR (95% CI) | p value | pcombined valued |
|---|---|---|---|---|---|---|---|---|
| PPARG | rs1801282 | 3,312 | G | 862 (45.0) | 988 (55.0) | 0.87 (0.80–0.95) | 0.003b | 1.0 × 10−4 |
| C/C | 3,139 (77.4) | 9,176 (75.4) | 1.00 (reference) | |||||
| C/G | 853 (21.0) | 2,789 (22.9) | 0.86 (0.78–0.95) | |||||
| G/G | 65 (1.6) | 206 (1.7) | 0.84 (0.62–1.15) | 0.010c | ||||
| HHEX–IDE | rs1111875 | 3,229 | A | 1,728 (49.8) | 1,742 (50.2) | 0.99 (0.93–1.06) | 0.812b | 0.017 |
| G/G | 1,283 (34.8) | 3,855 (34.9) | 1.00 (reference) | |||||
| G/A | 1,788 (48.6) | 5,324 (48.2) | 1.01 (0.92–1.10) | |||||
| A/A | 611 (16.6) | 1,867 (16.9) | 0.98 (0.85–1.12) | 0.880c |
aNote that the frequencies of genotypes in the cases and pseudo-controls are given, as well as the transmission counts from the transmission disequilibrium test analysis
bThe p value for association from the transmission disequilibrium test, which assumes a multiplicative model
cThe p value for association from the 2 df conditional logistic regression model, which does not assume a specific mode of inheritance
dThe p value obtained by combining the results of the case–control and family data set (please refer to the Methods for further details)