Characteristics of 335 children by study cohort and treatment arm

Values are mean (standard deviation) or n (%) unless noted. P values are for the overall group comparison using one-way ANOVA, Kruskall-Wallis or 2 x 3 ��2 test.*Information on water use was not available in the cohort of HIV-infected children for 1 child in the vitamin A group, 3 children in the vitamin A+zinc group, and 1 child in the multiple micronutrient group. In the cohort of HIV-uninfected children born to HIV-infected mothers, information was missing for 2 children in the vitamin A group and 2 children in the multiple micronutrient group. In the cohort of HIV-uninfected children born to HIV-uninfected mothers, information was missing for 4 children in the vitamin A group, and 2 children in the multiple micronutrient group.?As a Z-score of NCHS standards[15]��Hemoglobin values were not available in the cohort of HIV-infected children for 5 children in each of the three treatment groups. In the cohort of HIV-uninfected children born to HIV-infected mothers, hemoglobin determinations were not available for 26 children who received vitamin A, 28 children who received vitamin A+zinc, and 20 children who received multiple micronutrients. In the cohort of HIV-uninfected children born to HIV-uninfected mothers, hemoglobin measurements were not available for 34 children in the vitamin A group, 32 children in the vitamin A+zinc group, and 33 children in the multiple micronutrient group.?Days study drug was not given were determined during weekly home visits.\\Excludes days after withdrawal of study subjects for which home visits were not permitted, lost to follow-up, or death.