The medium-term impact of a micronutrient powder intervention on anemia among young children in Rural China

BMC Public Health

Table 2 Dose–response effects of per 100 MNP sachets consumption on child anemic status (average treatment effects on the treated analysis)

		Hemoglobin concentration (g/L) (95% CI)	Anemia prevalence (1 = yes) (95% CI)
Variable		(1)	(2)
Panel A: Full sample (N = 1,069)
(1)	Effect of per 100 MNP sachet consumption	0.13 (-0.44 to 0.69)	-0.02* (-0.04 to 0.0015)
		(0.29)	(0.01)
(2)	Mean (SD) number of sachets consumed in the treatment group	238.13 (149.93)
(3)	Median number of sachets consumed in the treatment group	247.5
Panel B: Anemic sample at baseline (N = 524)
(4)	Effect of per 100 MNP sachet consumption	0.15 (-0.60 to 0.91)	-0.02 (-0.05 to 0.007)
		(0.39)	(0.01)
(5)	Mean (SD) number of sachets consumed in the treatment group	233.03 (148.43)
(6)	Median number of sachets consumed in the treatment group	242
Panel C: Moderate anemic sample at baseline (N = 210)
(7)	Effect of per 100 MNP sachet consumption	0.77 (-0.46 to 2.00)	-0.05** (-0.096 to -0.008)
		(0.63)	(0.02)
(8)	Mean (SD) number of sachets consumed in the treatment group	220.49 (144.33)
(9)	Median number of sachets consumed in the treatment group	221.5

Anemic is defined as Hb < 110 g/L when child was 6–59 months old, and Hb < 115 g/L when child was 60–65 months old; moderate anemic is defined as 70 g/L < Hb < 100 g/L when child was 6–59 months old, and 80 g/L < Hb < 110 g/L when child was 60–65 months old. Each column presents the results of one regression of the outcome variable (corresponding to the column title) on the treatment dummy variable. Controls include the baseline value of the outcome variable, child’s age, gender, and whether the child had a low birth weight. We adjusted for cohort and county fixed effects, and standard errors are clustered at village level in full samples. Data missing for 395 samples in regard to the number sachets consumed (244 samples in Cohort 1 and 151 samples in Cohort 2)
^*p < 0.10
^**p < 0.05

ISSN: 1471-2458