Iron-containing micronutrient powder provided to children with moderate-to-severe malnutrition increases hemoglobin concentrations but not the risk of infectious morbidity: a randomized, double-blind, placebo-controlled, noninferiority safety trial.

BACKGROUND A link between the provision of iron and infectious morbidity has been suggested, particularly in children with malnutrition. Two meta-analyses concluded that iron is not harmful, but malnourished children were underrepresented in most available studies. OBJECTIVE This study evaluated the effect of iron-containing micronutrient powder (iron MNP) on infectious morbidities when provided to children with moderate-to-severe malnutrition and anemia. DESIGN A randomized, double-blind, placebo-controlled, noninferiority safety trial using a 2-mo course of daily iron MNP or placebo powder (PP) was conducted in 268 Bangladeshi children aged 12-24 mo with moderate-to-severe malnutrition (weight-for-age z score ≤ -2) and a hemoglobin concentration between 70 and 110 g/L. The primary endpoint was a composite of diarrhea, dysentery, and lower respiratory tract infection episodes (DDL) recorded through home visits every 2 d during the intervention and then weekly for 4 mo. The noninferiority margin was 1.2. Secondary endpoints included hemoglobin and anthropometric changes at 2 and 6 mo. All deaths and hospitalizations were documented. To capture seasonal variation, the study was repeated in the winter and summer with 2 distinct groups. An intention-to-treat analysis of recurrent events was performed by using the univariate Anderson-Gill model. RESULTS The baseline characteristics of the subjects were similar. Analysis of phase-aggregated DDL data showed that iron MNP was not inferior to PP (relative risk: 0.81; 95% CI: 0.62, 1.04) and improved hemoglobin concentrations (P < 0.0001). We recorded no deaths, and hospitalizations were rare. CONCLUSION Iron MNP is safe and efficacious when provided to children aged 12-24 mo with moderate-to-severe malnutrition and anemia. This trial is registered at clinicaltrials.gov as NCT00530374.