Simple and multivariate logistic regression for testing the association between Trp64Arg polymorphism and hyperuricemia.
| Model | Normouricemic (%) | Hyperuricemic (%) | OR (unadjusted) | 95% CI | p | OR (adjusted) | 95% CI | p* |
|---|---|---|---|---|---|---|---|---|
| Codominant | ||||||||
| T/T | 557 (88.2) | 46 (76.5) | 1.0 | 0.002 | 1.0 | < 0.0001 | ||
| T/A | 70 (11.2) | 10 (15.7) | 1.6 | 0.7–3.6 | 2.07 | 0.8–5.2 | ||
| A/A | 3 (0.6) | 5 (7.8) | 16.1 | 3.4–74.5 | 29.5 | 4.9–177.2 | ||
| Dominant | ||||||||
| T/T | 557 (88.2) | 46 (76.5) | 1.0 | 0.02 | 1.0 | 0.0084 | ||
| T/A, A/A | 73 (11.8) | 15 (23.5) | 2.3 | 1.1–4.6 | 3.15 | 1.39–7.15 | ||
| Recessive | ||||||||
| T/T-T/A | 627 (99.4) | 56 (92.2) | 1.0 | 0.001 | 1.0 | 0.0005 | ||
| A/A | 3 (0.6) | 5 (7.8) | 15.1 | 3.2–69.3 | 25.90 | 4.4–152.3 | ||
| Overdominant | ||||||||
| T/T-A/A | 560 (88.8) | 51 (84.3) | 1.0 | NS | 1.0 | NS | ||
| T/A | 70 (11.2) | 10 (15.7) | 1.4 | 0.66–3.27 | 1.78 | 0.72–4.44 | ||
| Log-additive | ||||||||
| 0, 1,2 | 630 (91.3) | 61 (8.7) | 2.5 | 1.4–4.4 | 0.002 | 3.35 | 1.73–6.50 | 0.0006 |
The codominant model compared heterozygous T/A and homozygous A/A genotypes to the homozygous for the most frequent allele T/T. The dominant model compared a combination of T/A-A/A genotypes to the homozygous T/T. The recessive model compared a combination of T/T-T/A genotypes to the homozygous A/A. The log-additive model is equivalent to calculating the odds ratio for the risk A allele. Data are number of subjects, percentage of subjects with each genotype for each group (normouricemic and hyperuricemic).
↵*p values adjusted for age, weight gain, triglyceride concentrations, uric acid, and insulin resistance.