Table 3.
Influence of cFGF23 and iFGF23 on hemoglobin after adjusting for biomarkers of iron metabolism, EPO concentrations and inflammation biomarkers.
| Dependent Variable | ||||||
|---|---|---|---|---|---|---|
| Hemoglobin (g/dL) | ||||||
| * cFGF23 (RU/mL) | ** iFGF23 (pg/mL) | |||||
| Model | Beta | p | R2 | Beta | p | R2 |
| Base model | −0.28 (−0.44/−0.11) | 0.001 | 0.49 | −1.04 (−2.1/0.03) | 0.057 | 0.45 |
| Iron model | −0.15 (−0.34/0.04) | 0.126 | 0.51 | −0.89 (−1.99/0.22) | 0.115 | 0.51 |
| EPO model | −0.22 (−0.38/−0.05) | 0.012 | 0.51 | −0.83 (−1.89/0.22) | 0.121 | 0.48 |
| Inflammation model | −0.28 (−0.45/−0.11) | 0.001 | 0.49 | −1.01 (−2.09/0.08) | 0.068 | 0.45 |
Regression coefficients (beta) and confidence intervals with hemoglobin as dependent variable for respective markers from regression models adjusting for albumin, gender, eGFR, and presence of Diabetes mellitus type II (Base model; only adjustment variables: R2 = 0.45); Base model plus serum iron, ferritin, transferrin, transferrin saturation and hepcidin (Iron model; only adjustment variables: R2 = 0.50); Base model plus EPO (EPO model; only adjustment variables: R2=0.48); Base model plus CRP and IL-6 (Inflammation model; only adjustment variables: R2 = 0.45). R2 values refer to adjustment variables plus the respective marker. * indicates that calculations were performed with binary-log-transformed marker values such that beta quantifies the effect of doubling the marker (otherwise beta quantifies the effect of one additional unit). ** indicates that calculations were performed with double log-transform, i.e., binary log of the binary log of iFGF23, such that beta quantifies the effect of squaring the value of iFGF23. cFGF23: c-terminal Fibroblast Growth Factor 23; iFGF23: intact Fibroblast Growth Factor 23; EPO: erythropoietin; CRP: C-reactive protein; IL-6: Interleukin 6; eGFR: estimated glomerular filtration rate (CKD-EPI formula).