Skip to main content
. 2015 Apr 20;10(4):e0123072. doi: 10.1371/journal.pone.0123072

Table 2. Determinants of log-hepcidin-level (linear regression analyses).

univariate multivariate
model 1 model 2
age [10 yrs] -0.02 (-0.09; 0.05) 0.002, p = 0.5 --
gender, male 0.24 (0.08; 0.41) -0.11, p = 0.13 0.21 (0.04; 0.37)
type 2 diabetes -0.09 (-0.37; 0.19) -0.03, p = 0.7 --
EPO [log (U/L)] -0.02 (-0.39; -0.09) -0.007 (-0.04; -0.01), p = 0.065 -0.01 (-0.02; 0.00)
GFR [10 ml/min/1.73 m2] -0.04 (-0.07; -0.004) -0.04 (-0.06; -0.01) -0.03 (-0.06; -0.01)
proteinuria [log(mg/day)] 0.05 (-0.007; 0.11) 0.02, p = 0.4 --
hemoglobin [g/L] -0.005 (-0.05; 0.04) -0.01, p = 0.8 --
ferritin [log(μg/L)] 0.48 (0.41; 0.55) 0.46 (0.39; 0.53) 0.45 (0.38; 0.52)
CRP [log(mg/dl)] 0.03 (-0.03; 0.10) -- 0.02, p = 0.6
albumin [(g/dL)2] 0.002 (-0.02; 0.02) -- -0.002, p = 0.7
history of CVD -0.13 (-0.31; 0.05) -- -0.05, p = 0.5
hypertension 0.19 (-0.02; 0.41) -- 0.18 (0.02; 0.35)
hyperlipidemia 0.21 (0.05; 0.39) -- 0.13 (0.00; 0.25)

Legend: data are beta-coefficients (95% CI), displayed bolded if p<0.05; multivariate linear regression models were built stepwise (model 1: basic patient characteristics, model 2: additional markers of inflammation and clinical variables), while eliminating variables with p>0.1 within each model (beta-coefficients and p-values are displayed before variables left the model); abbreviations: CI, confidence interval; EPO, endogenous erythropoietin; CVD, cardiovascular disease; GFR, glomerular filtration rate; CRP, C-reactive protein;