Skip to main content
. 2018 Jun 19;38(12):2219–2227. doi: 10.1111/liv.13890

Table 3.

Univariate and multivariate linear regression analyses

Dependent variable Independent variables Univariate analysis Multivariate analysis R 2
Beta 95% CI (lower, upper) P Beta 95% CI (lower, upper) P
IFN‐gamma Child‐Pugh score −0.8 (−1.5, −0.07) .032 −0.8 (−1.4, −0.1) .030
Bacterial DNA 3.2 (0.8, 5.5) .009 1.3 (−1.0, 3.9) .249
LPS 4.9 (0.9, 8.8) .016 2.9 (−0.7, 7.4) .102
Leukocytes 5 × 10−4 (5 × 10−5, 1 × 10−3) .029 4 × 10−4 (6 × 10−6, 9 × 10−4) .050 0.19
IL10 Bacterial DNA 8.8 (3.3, 14.3) .002 5.5 (−0.4, 11.4) .065
LPS 16.6 (7.6, 25.6) .001 12.7 (2.9, 22.4) .012 0.21
IL6 NSBB 15.3 (1.0, 29.5) .036 15.7 (2.2, 29.2) .023
Leukocytes 3 × 10−3 (4 × 10−5, 6 × 10−3) .047 2 × 10−3 (6 × 10−4, 5 × 10−3) .118 0.19

INF, interferon; LPS, gram‐negative bacterial endotoxin; NO, nitric oxide; NSBB, non‐selective beta‐blocker; TNF, tumour necrosis factor.

Univariate linear regression analyses were calculated to predict NO, TNF‐alpha, IFN‐gamma, IL10 or IL6 based on NSBB, bacterial DNA, LPS and variables shown in Table 1. Only significant variables in the univariate analysis were included in the multivariate analysis and are shown in the table.