Table 4.
Multivariate logistic regression analysis on the incidence of abnormal serum ALT level considering CYP2B6 SNVs, haplotypes, estimated phenotypes and non-genetic factors (age, sex).
| Model | Variables | Increased serum ALT levels | ||
|---|---|---|---|---|
| Coefficient B (SE) | Exp ß | P | ||
| SNVs, non-genetic factorsa | Constant | − 24.39 (40,192) | 0 | 1 |
| g.-82T>C (rs34223104) | 22.33 (40,192) | 4,996,498,070 | 1 | |
| g.15631G>T (rs3745274) | 0.27 (1.37) | 1.31 | 0.84 | |
| Model 1 | g.18053A>G (rs2279343) | 1 (1.39) | 0.14 | 0.16 |
| g.25505C>T (rs3211371) | − 1.39 (0.89) | 0.24 | 0.11 | |
| Sex | − 0.47 (0.70) | 0.95 | 0.94 | |
| Age | 0.02 (0.10) | 1.02 | 0.80 | |
| Haplotypes, non-genetic factorsb,d | Constant | 25.60 (79,461) | 1.31E+11 | 1 |
| CYP2B6*4 (g.-82T/15631G/18053G/25505C) | − 0.69 (1.09) | 0.49 | 0.76 | |
| CYP2B6*5 (g.-82T/15631G/18053A/25505T) | − 1.22 (0.86) | 0.29 | 0.16 | |
| Model 2 (corrected P ≤ 0.01) | CYP2B6*6 (g.-82T/15631T/18053G/25505T) | − 1.97 (0.74) | 0.13 | 0.008 |
| CYP2B6*22 (g.-82C/15631G/18053A/25505C) | − 23.31 (79,461) | 7.49E−11 | 1.00 | |
| Sex | 0.21 (0.70) | 1.23 | 0.76 | |
| Age | 0.64 (0.71) | 1.89 | 0.37 | |
| Estimated phenotype, non-genetic factorsc | Constant | 0.60 (0.75) | 1.83 | 0.42 |
| Poor/intermediate | − 1.56 (0.67) | 0.21 | 0.02 | |
| Model 3 (corrected P ≤ 0.02) | Sex | 0.11 (0.65) | 1.12 | 0.86 |
| Age | 0.60 (0.68) | 1.82 | 0.38 | |
Significant values are in bold.
aAIC: 37.71, BIC: 50.66 for model 1, bAIC: 40.89, BIC: 53.85 for model 2, cAIC: 25.59, BIC: 32.99 for model 3; dPost-hoc power: 94.1%.