Table 2.
haplotype frequency, SE, 95% CL | conditional logistic regression | |||||||
---|---|---|---|---|---|---|---|---|
controls | cases | OR* (95% CI) | p | |||||
SNP-221-224-225-227 | ||||||||
A-A-G-A | 0.016 | 0.005 | 0.007–0.026 | 0.018 | 0.005 | 0.008–0.028 | - | ns |
A-G-A-G | 0.738 | 0.016 | 0.705–0.771 | 0.685 | 0.018 | 0.650–0.719 | - | ns |
G-A-A-A | 0.034 | 0.007 | 0.021–0.048 | 0.049 | 0.008 | 0.033–0.065 | - | ns |
G-A-A-G | 0.049 | 0.008 | 0.033–0.065 | 0.065 | 0.009 | 0.047–0.084 | - | ns |
G-A-G-A | 0.085 | 0.011 | 0.064–0.106 | 0.105 | 0.012 | 0.082–0.128 | 1.52 (1.01–2.28) | 0.047 |
G-G-A-G | 0.064 | 0.009 | 0.046–0.083 | 0.063 | 0.009 | 0.045–0.081 | - | ns |
SNP-349-350-352-353-355-367-372 | ||||||||
A-A-A-A-A-A-A | 0.237 | 0.016 | 0.205–0.268 | 0.283 | 0.017 | 0.250–0.317 | 1.38 (1.02–1.86) | 0.035 |
A-A-A-A-A-G-A | 0.036 | 0.007 | 0.022–0.049 | 0.022 | 0.007 | 0.011– 0.033 | - | ns |
A-A-A-C-G-A-A | 0.412 | 0.019 | 0.375–0.449 | 0.314 | 0.018 | 0.279–0.349 | - | ns |
A-A-A-C-G-G-A | 0.135 | 0.013 | 0.110–0.160 | 0.157 | 0.014 | 0.130–0.184 | - | ns |
A-A-A-C-G-G-G | - | - | - | 0.013 | 0.004 | 0.004–0.021 | - | ns |
A-A-G-A-A-A-A | 0.021 | 0.005 | 0.010–0.032 | 0.018 | 0.005 | 0.008–0.028 | - | ns |
A-G-G-A-A-A-A | 0.020 | 0.005 | 0.010–0.031 | 0.019 | 0.005 | 0.009–0.029 | - | ns |
A-G-G-A-A-G-A | 0.010 | 0.003 | 0.003–0.017 | 0.013 | 0.004 | 0.004–0.021 | - | ns |
G-G-G-A-A-G-A | - | - | - | 0.019 | 0.005 | 0.009–0.029 | - | ns |
G-G-G-A-A-A-A | 0.073 | 0.009 | 0.053–0.092 | 0.073 | 0.010 | 0.053–0.092 | - | ns |
G-G-G-A-A-G-G | - | - | - | 0.026 | 0.006 | 0.014–0.039 | 10.5 (2.54–43.6) | 0.0012 |
G-G-G-C-A-A-A | 0.018 | 0.005 | 0.008–0.028 | 0.020 | 0.005 | 0.009–0.030 | - | ns |
Odds ratios (OR) and lower and upper 95% confidence intervals (CI) were estimated by conditional logistic regression in haplotype trend regression models (HTR) adjusted for baseline age and for time-dependent CD4+ T-cell covariates. (ns)- not significant at the 5% threshold level to enter or to stay in the stepwise selection logistic models.
-Test of the distal haplotype was further adjusted for the effects of HLA -B. Note that the frequencies do not add to 100% because only haplotypes with frequencies greater than 1% were modeled; the remaining haplotypes were grouped together and included in the HTR models as a single term. (−) -not observed at the frequency cutoff of 1%.