Table 3.
Association of microbial translocation markers and cirrhosis in 88 subjectsa
| OR | 95% CI | p-value | |
|---|---|---|---|
| Univariate | |||
| LPS ≥ 42 pg/mLb | 19.0 | 2.98 – 120.79 | 0.0018 |
| LBP (highest quartile) | 4.39 | 0.99 – 19.36 | 0.05 |
| sCD14 (highest quartile)b | 8.65 | 1.98 – 37.72 | 0.0041 |
| AAL ≥ 5-fold above controlb | 27.77 | 5.64 – 136.71 | <0.0001 |
| EndoCAb IgM (highest quartile) | 0.10 | 0.01 – 0.86 | 0.036 |
| CD4+ lymphocyte <350/mm3b | 7.02 | 1.36 – 36.31 | 0.02 |
| Multivariatec | |||
| LPS ≥ 42 pg/mL | 18.71 | 2.62 – 133.78 | 0.0035 |
| CD4+ lymphocyte <350/mm3 | 6.29 | 0.97 – 40.69 | 0.054 |
a
LPS-lipopolysaccharide; LBP – LPS binding protein; sCD14 – soluble CD14; EndoCAb IgM antibody to LPS core; AAL- Aleuria aurantia lectin. Cirrhosis was determined by liver biopsy and/or clinical events including esophageal varices, encephalopathy, or ascites.
b
Adjusted for age because of an association detected in univariate analysis
c
Shown is a model of cirrhosis, with values adjusted for CD4+ lymphocyte depletion. Similar results were obtained for models of cirrhosis and the other markers (sCD14, LBP, EndoCAb IgM, and AAL) but are not shown. Inclusion of more than one marker in the same model of cirrhosis did not improve the fit significantly.