TABLE 4.
The association of individual and combinations of SNPs with type 2 diabetes in the Go-DARTS type 2 diabetes case-control study
| SNP | Nearest genes* | Type 2 diabetes OR per allele (95% CI) | P | Expected OR† = a × b (95% CI) |
|---|---|---|---|---|
| rs108893353 | ANGPTL3 | 0.97 (0.91–1.03) | 0.27 | |
| rs17216525 | NCAN, CLIP2, PBX4 | 0.89 (0.80–0.99) | 0.03 | |
| rs2954029 | TR1B1 | 1.01 (0.95–1.06) | 0.83 | |
| rs714052 | MLXIPL | 0.96 (0.89–1.05) | 0.38 | |
| rs7557067 | APOB | 1.00 (0.93–1.07) | 0.99 | |
| rs7679 | PLTP | 0.98 (0.91–1.06) | 0.64 | |
| rs7819412 | XKR6-AMAC1L2 | 0.97 (0.91–1.03) | 0.26 | |
| rs328 | LPL | 1.02 (0.93–1.12) | 0.69 | |
| rs3135506 | APOA5 | 1.04 (0.92–1.18) | 0.52 | |
| rs662799 | APOA5 | 1.00 (0.88–1.12) | 0.97 | |
| Allele score | 0.99 (0.96–1.01) | 0.26 | 1.10 (1.08–1.12) | |
| Q1 vs. Q2 | 0.95 (0.84–1.07) | 0.39 | 1.19 (1.09–1.30) | |
| Q1 vs. Q3 | 0.89 (0.79–1.01) | 0.08 | 1.32 (1.19–1.46) | |
| Q1 vs. Q4 | 0.90 (0.80–1.02) | 0.11 | 1.37 (1.24–1.53) | |
| Q1 vs. Q5 | 0.93 (0.82–1.06) | 0.27 | 1.52 (1.34–1.72) |
Q = quintile of weighted allele score.
*Nearest gene information reported as in Kathiresan et al. (31), except for rs328, which is from Kathiresan et al. (34), and for rs3135506 and rs662799, which are from Pennacchio et al. (32).
The sample sizes ranged from 10,378 to 10,574 for the individual SNP(s) vs. type 2 diabetes analyses. The number of individuals in the weighted allele score vs. type 2 diabetes analyses is 10,676 and ranged from 4,212 to 4,318 for quintiles of the allele score vs. type 2 diabetes analyses. The total number of individuals in the SNP(s) vs. type 2 diabetes analyses is greater than the corresponding SNP(s) vs. triglyceride analyses in Table 2 because the numbers in the latter were also restricted by the number of individuals with pretreatment triglyceride levels measured. The results for the SNP(s) vs. type 2 diabetes analyses were similar when we restricted the analyses to individuals with pretreatment triglycerides measured.
†The point estimate of the expected OR for the allele score vs. type 2 diabetes was calculated by multiplying (in Fig. 1a) the effect size of the allele score vs. the triglyceride association by (in Fig. 1b) the OR of the triglyceride vs. type 2 diabetes association. For example, the expected type 2 diabetes OR for Q1 vs. Q2 was calculated by multiplying the effect size of the allele score vs. triglyceride association for the Q1 vs. Q2 comparison (0.17 SDs) by the natural log of the relevant triglyceride vs. type 2 diabetes OR (i.e., natural log of 2.9). The exponent of the answer from this multiplication is the expected type 2 diabetes OR for Q1 vs. Q2.