Background: HIV infection is associated with an increased risk of malaria among adults, which can be reduced with cotrimoxazole (CTX) prophylaxis. However, data are limited on the protective effect of CTX prophylaxis on malaria risk in children, the group that suffers the greatest burden of this disease. We evaluated the efficacy of CTX for prevention of malaria among HIV-infected children in Kampala, Uganda, where antifolate resistance to malaria is common.

Methods:  Between October 2005 and September 2006 we enrolled a cohort of 300 HIV-infected children aged 1 to 10 years in the Children with HIV and Malaria Project (CHAMP) in Kampala, Uganda. All children were prescribed CTX prophylaxis, and 88% reported using insecticide-treated bed-nets. Parasite prevalence was measured at enrollment and every 3 months. Thick blood smears were done for all children who presented with a new episode of fever. A concurrent cohort of 561 healthy children aged 1 to 10 years was followed simultaneously, using identical assessment schedules for malaria; these children were not prescribed CTX prophylaxis, and only 12% used insecticide-treated bed-nets. 

Results:  None of the 300 HIV-infected children and 20% of children recruited from the community (p <0.0001) had a positive malaria blood smear at enrollment. Between October 2005 and September 2006, only 9 episodes of malaria were diagnosed among HIV-infected children (incidence = 0.07/person-year) compared to 440 episodes among children recruited from the community (incidence = 0.90/person-year, p <0.0001) for an incidence rate ratio of 0.08 (0.04 to 0.15). The prevalence of the dhfr/dhps quintuple mutation, which has been associated with antifolate resistance, was common in samples both from patients taking CTX (100%) and those not taking CTX (75%). Compared with no intervention, the risk of malaria among HIV-infected children receiving CTX alone was decreased by 35% (IRR 0.65, 0.27 to 1.57) and the risk in those receiving CTX and using insecticide-treated bed-nets was decreased by 97% (IRR 0.03, 0.01 to 0.11). Malaria accounted for only 4% of all fever episodes in the HIV-infected cohort compared to 33% in the community-based cohort (p <0.0001).  

Conclusions:  In a malaria-endemic area with high-level antifolate resistance, CTX prophylaxis and insecticide-treated bed-nets were associated with dramatic reductions in malaria among HIV-infected children compared to healthy controls. Fevers among HIV-infected children receiving CTX and insecticide-treated bed-nets are unlikely due to malaria, and other etiologies should be sought.