Albertine Donker

Chapter 7 240 Table 3. Results of linear regression models for serum hepcidin concentrations (nmol/L) adjusted for age and time of blood sampling and stratified by sex Males 95% CI Variable Beta a Lower limit Upper limit R 2 , % b p Hb, g/dL 0.018 -0.086 0.121 18.4 0.737 Reticulocytes, x10^9/L 0.001 -0.005 0.007 18.7 0.668 MCV, fL 0.002 -0.022 0.026 18.4 0.879 Ferritin, µg/L d 1.313 0.946 1.679 38.8 0.000 Iron, µmol/L 0.005 -0.011 0.021 17.6 0.517 TIBC, µmol/L -0.009 -0.018 0.001 20.3 0.071 TSAT, % 0.006 -0.004 0.015 18.2 0.233 sTfR, mg/L -0.216 -0.526 0.094 19.3 0.170 ALT, IU/Ll d 0.522 -0.024 1.067 20.2 0.061 CRP, mg/L d,e 0.229 0.106 0.352 25.3 0.000 BMI f Normal weight ref ref ref Underweight 0.484 -0.087 1.056 20.5 0.096 Overweight 0.196 -0.050 0.442 0.118 Adipose 0.254 -0.194 0.702 0.264 Adjustment for age was done with age as a continuous variable; adjustment for sampling time was done with sampling time as a categorical variable (7.30 AM-12 PM, 12-3 PM, 3-6 PM). a Beta expresses the change in the dependent variable -log-transformed serum hepcidin- that is associated with a 1-unit change in the independent variable. Independent variables marked with d were log-transformed as well; the interpretation of the regression coefficients for these variables is as follows: a 1% change in the independent variable corresponds to a beta % change in serum hepcidin b For males multivariate linear regression model with log-transformed hepcidin as dependent variable and age (continuous variable) and sampling time as independent variables showed: R 2 18.9 %, with ß: -0.048, CI -0.065– -0.030 ( p 0.000) ß: 0.254, CI 0.081– 0.426 ( p 0.004), ß: 0.496, CI 0.255–0.737