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

Table 3. Determinants of mortality (Cox proportional hazards analysis).

univariate multivariate
model 1 model 2
hepcidin [10 ng/ml] 1.024 (0.951; 1.102) 1.513 (1.136; 2.015) 1.491 (1.095; 2.029)
EPO [log (U/L)] 1.497 (0.867; 2.585) 2.612 (0.972; 7.020) 2.466 (0.884; 6.877)
hepcidin * logEPO -- 0.846 (0.759; 0.944) 0.858 (0.763; 0.9763)
age [10 yrs] 1.917 (1.413; 2.602) 2.225 (1.399; 3.538) 1.982 (1.292; 3.042)
gender, male 1.982 (1.089; 3.601) 2.030 (0.873; 4.721) 2.03, p = 0.10
type 2 diabetes 3.715 (1.015; 13.60) 0.78, p = 0.8 --
GFR [10 ml/min/173m2] 0.718 (0.624; 0.826) 0.96, p = 0.7 --
proteinuria [log(mg/day)] 1.487 (1.229; 1.799) 1.538 (1.227; 1.929) 1.625 (1.305; 2.023)
hemoglobin [g/L] 0.829 (0.725; 0.949) 0.99, p = 0.9 --
CRP [log (mg/dl)] 1.509 (1.194; 1.908) 1.592 (1.163; 2.179) 1.566 (1.133; 2.165)
albumin [(g/dl)] 0.425 (0.276; 0.655) 0.74, p = 0.5 --
history of CVD 3.663 (2.041; 6.571) 1.63, p = 0.2 --
hypertension 1.216 (0.566; 2.611) -- --
hyperlipidemia 1.055 (0.598; 1.864) -- --
ferritin [log(μg/L)] 0.973 (0.715; 1.324) -- --

Legend: data are hazard ratios (HR) (95% confidence interval, CI), displayed bolded if p<0.05; multivariate Cox models were built with backwards selection (model 1: pexclusion>0.10, model 2: pexclusion>0.05), accounting for variables significant (p<0.05) in univariate associations (HRs and p-values are displayed before variables left the model)