Table 4 Regression models for correlation between weighted and Box–Cox-transformed values of transaminases and weighted and/or Box–Cox-transformed (if required) inflammation markers (†).

From: Longitudinal relationship of liver injury with inflammation biomarkers in COVID-19 hospitalized patients using a joint modeling approach

Cor(AST, CRP) ~ s(Age) + s(Time)*Exitus

Cor(AST, D-Dimer) ~ Sex + s(Age) + s(Time)* Exitus

Cor(AST, IL-6) ~ s(Age) + s(Time)*Exitus

Cor(AST, Lymphocytes) ~ Sex + s(Age) + s(Time)*Exitus

Cor(ALT, CRP) ~ Sex + s(Age) + s(Time)*Exitus

Cor(ALT, D-Dimer) ~ Sex + s(Age) + s(Time)* Exitus

Cor(ALT, IL-6) ~ s(Age) + s(Time)*Exitus

Cor(ALT, Lymphocytes) ~ Exitus + s(Time)*Exitus

Cor(GGT, CRP) ~ Sex + s(Age) + s(Time)*Exitus

Cor(GGT, D-Dimer) ~ Exitus + Antiviral Treatment + s(Age) + s(Time)* Exitus

Cor(GGT, IL-6) ~ s(Time)*Exitus

Cor(GGT, Lymphocytes) ~ Sex + s(Age) + s(Time)*Exitus

Cor(ALP, CRP) ~ Antiviral Treatment + s(Age) + s(Time)*Exitus

Cor(ALP, D-Dimer) ~ s(Age) + s(Time)* Exitus

Cor(ALP, IL-6) ~ s(Time)*Exitus

Cor(ALP, Lymphocytes) ~ Sex + s(Age) + s(Time)*Exitus

  1. s() indicates the smooth effect of the covariate. * indicates association. For sex, the reference level is “Male”. For Exitus, the reference level is “Living”. For antiviral treatment, the reference level is “No”. The centered smooth effects of s(Time) and s(Age) in the correlation are depicted in Figs. 3 and 4, respectively. (†): From the general model type 2 every regression model in this table only includes the covariates that were found to be statistically significant (p < 0.05).
Back to article page