Effect of periodic vitamin A supplementation on mortality and morbidity of human immunodeficiency virus–infectedchildren in Uganda: A controlled clinical trial

Objective Weinvestigated whether vitamin A supplementation would decrease mortalityand morbidity rates in children infected with the humanimmunodeficiency virus(HIV). Methods Weconducted a randomized, double-blind, placebo-controlled clinical trialat Mulago Hospital, a large hospital that serves the urban andsemiurban populations of Kampala, Uganda. One hundred eighty-oneHIV-infected children were enrolled at 6 mo and randomized to receivevitamin A supplementation, 60 mg retinol equivalent, or placebo every 3mo from ages 15 to 36 mo. Morbidity was assessed through a 7-dmorbidity history every 3 mo, and vital events were measured. Childrenreceived daily trimethoprim-sulfamethoxazole prophylactictherapy. Results Afterage 15 mo, children were followed for a median of 17.8 mo(interquartile range = 11.1 to 21.0 mo). The trial was stoppedwhen there was a new policy to implement a program of masssupplementation of vitamin A in the country. Mortality rates among 87children in the vitamin A group and 94 children in the control groupwere 20.6% and 32.9%, respectively, yielding a relativerisk of 0.54 (95% confidence interval, 0.30 to 0.98;P = 0.044) after adjusting for baselineweight-for-height Z score. Children who receivedvitamin A had lower modified point prevalences of persistent cough(odds ratio, 0.47; 95% confidence interval, 0.23 to 0.96;P = 0.038) and chronic diarrhea (oddsratio, 0.48; 95% confidence interval, 0.19 to 1.18;P = 0.11) and a shorter duration of eardischarge (P = 0.03). Vitamin Asupplementation had no significant effect on modified point prevalencesof fever, ear discharge, bloody stools, orhospitalizations. Conclusions Vitamin Asupplementation decreases mortality rate in HIV-infected children andshould be considered in the care for these children in developingcountries.