Table 2.
Cox regression analyses of serum Ang-2 levels predicting all-cause mortality
| All (N = 313) | Male (n = 206) | Female (n = 107) | ||||
|---|---|---|---|---|---|---|
| Analyses | HR (95% CI) | P-value | HR (95% CI) | P-value | HR (95% CI) | P-value |
| Univariate Cox regression | ||||||
| Continuous Ang-2 | 1.002 (1.001–1.004) | 0.005 | 1.002 (1.001–1.004) | 0.009 | 1.003 (0.998–1.008) | 0.221 |
| Binary Ang-2 | 1.934 (1.412–2.649) | <0.0001 | 2.475 (1.660–3.689) | <0.0001 | 1.273 (0.742–2.185) | 0.381 |
| Multivariable Cox regression | ||||||
| Model A | 1.756 (1.274–2.419) | 0.001 | 2.467 (1.624–3.747) | <0.0001 | 0.885 (0.503–1.555) | 0.670 |
| Model B | 2.613 (1.698–4.023) | <0.0001 | 2.937 (1.660–5.196) | 0.0002 | 1.990 (0.946–4.186) | 0.070 |
| Model C | 2.245 (1.443–3.493) | 0.0003 | 3.294 (1.768–6.138) | 0.0002 | 1.084 (0.476–2.467) | 0.847 |
Binary Ang-2 was divided according to optimal cut-off values of Ang-2 concentrations (111.0 pmol/L). Multivariable Cox regression analyses were performed in three models. Model A was adjusted for age, comorbidity and smoking; model B was adjusted for dialysis vintage, serum creatinine, hemoglobin, CRP, serum albumin, ferritin, transferrin, iPTH, serum calcium, serum phosphorus, LDL and Kt/V and model C was adjusted for the above risk factors (model A + model B) plus ultrafiltration volume.