Table 8.

Adjusted association of infant feeding practice, health and nutritional status in rural Shaanxi Province (n = 948).

Infant Feeding Practice	Hb Concentrations (g/L) a			Hb < 110 g/L (Yes = 1) b
	Coefficient	95% CI		Coefficient	95% CI
Ever breastfed	−0.71	−2.77	1.35	0.01	−0.07	0.09
Exclusive or predominant breastfeeding < 6 months	−0.81	−3.87	2.24	0.05	−0.07	0.18
Still breastfed ≥ 6 months	−5.13 **	−7.60	−2.66	0.13 **	0.03	0.22
Ever formula-fed	3.25 **	1.11	5.39	−0.12 **	−0.20	−0.03
Introduction of solid, semi-solid and soft food ≥ 6 months c	−0.31	−1.88	1.27	0.01	−0.06	0.07

^a Regression estimates from multiple linear models adjusted for gender, age, low birth weight, birth order, maternal age, maternal education, whether primary caregiver is mother, whether household received Social Security Support, whether infant had fever or diarrhea last month and county fixed effect. Clustering is at the village level. ^b Regression estimates from logit models adjusted for gender, age, low birth weight, birth order, maternal age, maternal education, whether primary caregiver is mother, whether the family received Social Security Support, whether infant had fever or diarrhea last month and county fixed effect. Clustering is at the village level. ^c To test the robustness of our analysis, we also ran the analysis on the effect of solid feeding behavior on Hb concentrations and anemia status. Instead of breaking the sample into three categories (less than six months; between six and eight months; greater than eight months) as we did in the manuscript proper, we also broke it into two categories (less than six months; greater than six months). The results (which are not shown for brevity) were substantially the same. ** p < 0.01 and * p < 0.05.