Table 2.
Results of quantitative synthesis involved the association between MMP-3 5A/6A polymorphisms and RA
| Comparison model | Overall | Caucasians | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Sample size (case/control) | OR (95% CI) | P h | P | P r | Sample size (case/control) | OR (95% CI) | P h | P | |
| 6A vs 5A | 2902/2478 | 1.19(0.91-1.56)* | 0.001 | 0.203 | 0.461 | 2340/2176 | 1.23(0.89-1.68)* | 0.000 | 0.207 |
| 5A/6A vs 5A/5A | 257/275 | 1.31(0.89-1.92) | 0.106 | 0.174 | 0.824 | 200/243 | 1.37(0.68-2.76)* | 0.058 | 0.373 |
| 6A/6A vs 5A/5A | 401/292 | 1.78(0.68-4.61)* | 0.003 | 0.238 | 0.835 | 169/169 | 2.11(0.66-6.67)* | 0.002 | 0.209 |
| Dominant model | 593/477 | 1.46(0.71-3.00)* | 0.015 | 0.299 | 0.766 | 312/326 | 1.61(0.68-3.80)* | 0.008 | 0.274 |
| Recessive model | 593/477 | 1.48(0.88-2.47)* | 0.023 | 0.141 | 0.257 | 312/326 | 1.69(0.84-3.40)* | 0.021 | 0.142 |
Dominant model: 6A/6A +5A/6A versus 5A/5A, Recessive model: 6A/6A versus 5A/5A +5A/6A, OR: odds ratio, CI: confidence interval, Ph: P -value of Q-test for heterogeneity test, P: P -value of Z-test for significance of the pooled OR, Pr: P -value of Egger's regression test for publication bias, *Significant heterogeneity: the random-effect model was chosen to summarize the result.