Logistic regression models simultaneously considering AKT2 and GSK3B in PCOS risk
| Model . | Independent variables . | OR (95% CI) . | P value . |
|---|---|---|---|
| Independence | Carrier state (Y/N) for AKT2 risk haplotype* | 2.10 (1.2–3.7) | 0.010 |
| Carrier state (Y/N) for GSK3B risk haplotype* | 1.80 (1.0–3.1) | 0.036 | |
| Additive | Number of AKT2 or GSK3B risk haplotypes† | 1.58 (1.1–2.2) | 0.010 |
| Combined | Carrier state (Y/N) for both risk haplotypes in AKT2 and GSK3B‡ | 3.11 (1.4–6.9) | 0.0050 |
| Model . | Independent variables . | OR (95% CI) . | P value . |
|---|---|---|---|
| Independence | Carrier state (Y/N) for AKT2 risk haplotype* | 2.10 (1.2–3.7) | 0.010 |
| Carrier state (Y/N) for GSK3B risk haplotype* | 1.80 (1.0–3.1) | 0.036 | |
| Additive | Number of AKT2 or GSK3B risk haplotypes† | 1.58 (1.1–2.2) | 0.010 |
| Combined | Carrier state (Y/N) for both risk haplotypes in AKT2 and GSK3B‡ | 3.11 (1.4–6.9) | 0.0050 |
PCOS diagnosis was the dependent variable in all models. Each analysis also included age and BMI as independent variables.
Carriers are homozygous or heterozygous for risk haplotype (T-G-C-T in AKT2 and C-A-C-C-G-G-A-G-G in GSK3B).
Number of risk haplotypes per individual is 0–4.
Compares subjects who are carriers of at least one AKT2 risk haplotype and at least one GSK3B risk haplotype to all other subjects.