Iron deficiency among low income Canadian toddlers: a cross-sectional feasibility study in a Community Health Centre and non-Community Health Centre sites

Table 3 Multivariable linear regression model for association between CHC status and serum ferritin

Predictor	β (log)	95% CI	Change in serum ferritin, %	Change in median serum ferritin		P
Predictor	β (log)	95% CI	Change in serum ferritin, %	μg/l	95% CI	P
CHC, yes	0.17	−0.10, −0.45	19.12	4.78	−2.46, 14.34	0.22
Age, mo	−0.0006	−0.01, 0.01	0.06	0.14	−0.15, 0.18	0.87
Female sex	0.08	0.003, 0.16	8.01	2.00	0.07, 4.25	0.06
Income
$0–29,999	−0.23	−0.41, − 0.04	−20.17	−5.04	−8.48, − 0.89	0.02*
$30,000-79,999	0.02	−0.11, 0.15	1.97	0.49	−2.69, 4.12	0.77
≥$80,000 (Reference)
Maternal education (Post-secondary, no)	0.08	−0.09, 0.24	7.93	1.98	02.10, 6.79	0.36
zBMI, unit	−0.07	− 0.10,-0.03	−6.49	−1.62	−2.49, 0.71	0.0006*
Breastfeeding duration, mo	−0.01	−0.02, − 0.01	− 1.46	−0.37	− 0.54, − 0.20	<.0001*
Milk consumption (> 2 cups daily)	− 0.11	−0.20, − 0.02	− 10.37	−2.59	−4.49, − 0.52	0.02*
Meat / fish consumption daily	0.01	− 0.04, 0.06	0.94	0.24	− 0.91, 1.44	0.69
CRP (mg/L)	0.07	0.05, 0.10	7.69	1.92	1.17, 2.70	< 0.0001*
Iron/ multivitamin supplementation	0.04	−0.16, 0.25	4.50	1.13	−3.80, 7.20	0.68

zBMI indicates body mass index z score; CI indicates confidence interval; CRP indicates C-reactive protein
Adjusted β estimates are reported to 2 decimal points for statistical precision. Negative values indicate a decrease in median serum ferritin, positive values indicate an increase in median serum ferritin. *indicates statistically significant effects

ISSN: 1471-2296