Table 2. Factors associated with children’s characteristic hepcidin levels.

The multivariate analysis includes all the variables with a significant influence based on forward selection with P < 0.05. Prior iron, iron supplementation over the previous 7 days; NS, not significant.

Predictor	Positive/negative association	F ratio*	P	R2
Univariate analysis
Iron indices
Log ferritin (also affected by inflammation)	Positive	37.0	<0.0001	17.1
Transferrin	Negative	15.2	<0.0001	13.4
UIBC	Negative	19.9	<0.0001	13.2
sTfR	Negative	22.0	<0.0001	13.0
Serum iron	–	4.4	0.04	–
TSAT	–	0.0	NS	–
Prior iron	–	0.1	NS	–
Hb	–	1.3	NS	–
MCV	Positive	15.2	<0.0001	8.1
MCH	Positive	21.0	<0.0001	11.4
Inflammatory markers
Log CRP	Positive	111.5	<0.0001	14.2
Log AAGP	Positive	45.4	<0.0001	6.5
Total WBC	Positive	5.3	<0.02	1.7
Granulocytes	Positive	39.1	<0.0001	3.9
Lymphocytes	Positive	6.0	0.014	1.7
Granulocyte-to- lymphocyte ratio	Positive	48.2	<0.0001
Fecal inflammatory marker
Log calprotectin	–	0.03	NS	–
Multivariate analysis
Log ferritin	Positive	10.0	0.0015	26.6
UIBC†	Negative	11.3	0.0013
Log CRP	Positive	46.6	<0.0001
Granulocyte-to-lymphocyte ratio	Positive	14.9	<0.0001
Stratified analysis of CRP as predictor of hepcidin‡
CRP range
All (n = 370)	Positive	111.5	<0.0001	14.2
<10 (n = 322)	Positive	65.2	<0.0001	10.9
<5 (n = 258)	Positive	35.6	<0.0001	8.9
<3 (n = 189)	Positive	23.9	<0.0001	7.2
<1 (n = 60)	Positive	14.4	0.0002	3.6

*Adjusting for individual subject (discrete variable) and week (continuous variable). The F ratios and P values shown are from the same multivariate model, including the four variables shown as well as subject and week.

†UIBC and sTfR were mutually replaceable in the multivariate analysis.

‡Log hepcidin against log CRP with adjustment for individual subject (discrete variable) and study week (continuous variable).