Results from the multivariable regression analysis on hepcidin-25 levels.<sup>a</sup>
收藏资源简介:
aRegression analyses were performed with natural logarithm of hepcidin-25 as dependent variable. Potential determinants of hepcidin-25 were selected using a backward multivariable linear regression model with a p-value <0.15 used as a cut-off level in which all patient, treatment and laboratory characteristics as listed in table 1 were entered. bThe regression coefficient (B) denotes a natural logarithm. Positive values indicate an increase in hepcidin-25 and negative values a decrease with one unit increase of the determinant. cResults of conversion of the regression coefficient (B) from natural logarithm to a percentage of change: for each increase in the determinant with one unit, hepcidin-25 changed with the percentage indicated in this column. Positive values indicate an increase in hepcidin-25 and negative values a decrease. R2 for multivariable regression model = 0.49. Further adjustment for participating center did not change the results (data not shown).



