Table 3.
Logistic regression analyses used to find the best genetic model of inheritance that describes the effect of the genotypes
| SNP | Model | Genotype | OR (95 % CI) | AIC |
|---|---|---|---|---|
| rs2672826 A>G COL23A1 |
Co-dominant | G/G G/A A/A |
1.00 3.94 (1.28–12.10)* 0.72 (0.10–5.25) |
206.4 |
| Dominant | G/G G/A–A/A |
1.00 2.86 (1.10–7.49)* |
206.6 | |
| Recessive | G/G–G/A A/A |
1.00 0.59 (0.08–4.33) |
211.6 | |
| rs12589592 G>A FBLN5 |
Co-dominant | G/G G/A A/A |
1.00 2.26 (1.15–4.46)* 13.00 (1.61–104.81)* |
202.7 |
| Dominant | G/G G/A–A/A |
1.00 2.71 (1.40–5.25)* |
204.8 | |
| Recessive | G/G–G/A A/A |
1.00 8.67 (1.10–68.06)* |
206.4 | |
| rs2158836 A>G LAMB1 |
Co-dominant | G/G G/A A/A |
1.00 1.27 (0.64–2.51) 18.43 (2.35–144.63)* |
199.3 |
| Dominant | G/G G/A–A/A |
1.00 1.89 (0.98–3.62)* |
210.1 | |
| Recessive | G/G–G/A A/A |
1.00 16.36 (2.14–124.9)* |
197.8 | |
| rs7323378 T>C POSTN |
Co-dominant | T/T T/C C/C |
1.00 4.01 (1.85–8.67) 8.25 (2.85–23.88)* |
192.6 |
| Dominant | T/T T/C–C/C |
1.00 4.88 (2.35–10.16)* |
192.7 | |
| Recessive | T/T–T/C C/C |
1.00 3.54 (1.37–9.14)* |
203.6 | |
| rs17516906 A>G FN1 |
Co-dominant | G/G A/G A/A |
1.00 0.00 (NA) 0.00 (NA) |
409.0 |
| Dominant | G/G A/G–A/A |
1.00 0.00 (NA) |
412.8 | |
| Recessive | G/G–A/G A/A |
1.00 2.81 (1.40–5.63)* |
410.7 |
Odds ratio, 95 % confidence interval, Akaike information criterion (AIC), and P values were retrieved from logistic regression analyses with the co-dominant, dominant, and recessive models. The best-fitting model has the lowest AIC, and this value is at least 2 lower than the other models. P values <0.05 are considered significant (*)
OR odds ratio, CI confidence interval, NA not available