Association Between Malaria and CD4 Cell Count Decline Among Persons With HIV

Abstract

To the Editor: Recent information that episodes of Plasmodium falciparum malaria are associated with an average increase of 0.25 log in HIV viral load for approximately 2 months1 supports previous findings of an association between malaria and viral load.2 It is not known whether this level and duration of increase in viral load affect HIV disease progression,3 however, because CD4 cell count measurements, markers of HIV disease progression, were not conducted at follow-up. To examine the potential long-term effects of clinical malaria on HIV disease, we examined data collected during a 2-year cohort study in Uganda.4,5 Participants with HIV infection provided blood for CD4 cell count testing at baseline, at 5 months just before they began cotrimoxazole prophylaxis, and 1.5 years later. Participants were visited at their homes weekly. If they reported fever, we took a blood smear, and if the smear was positive for malaria, we brought treatment to the participant's home within 24 hours. Of 449 HIV-infected persons with available CD4 cell count data for at least 2 time points and no malaria parasitemia at the time of blood draws, baseline CD4 cell counts were <200 cells/μL for 24%, 200 to 500 cells/μL for 45%, and >500 cells/μL for 31%. There were 1024 slides from fever episodes during follow-up, 112 (11%) of which were associated with parasitemia in 92 participants (22.3 episodes of malaria per 100 person-years). The mean time from malaria to the subsequent CD4 cell count was 123 days (range: 7-553 days). Malaria was associated with a more rapid decline in CD4 cell count. Adjusting for sex, age, and baseline CD4 cell count, the mean difference in CD4 cell decline per each additional malaria episode was 40.5 cells/μL per year (95% confidence interval [CI]: 13.1 to 68.0; P = 0.0038). Compared with people with no malaria episodes, the mean difference in annual CD4 cell count decline for persons with 1 episode was 5.4 (95% CI: −63 to 74) cells/μL; for persons with 2 episodes, it was 84 (95% CI: 11 to 157) cells/μL; and for persons with ≥3 episodes, it was 142 (95% CI: −26 to 311) cells/μL. The association between malaria and CD4 cell count decline was seen across all baseline CD4 strata (P = 0.18 for effect modification of baseline CD4 cell count on the relation between malaria and CD4 cell count decline). Although cotrimoxazole prophylaxis was associated with a decreased incidence of malaria, there was no interaction between prophylaxis, malaria, and CD4 cell count decline. A possible reason for the association between malaria and CD4 cell count decline could be that people with more advanced disease at baseline were more likely to have malaria during follow-up, because CD4 cell count has previously been associated with incidence of malaria.6 We controlled for baseline CD4 cell count, however. In our study, people with malaria were treated promptly; yet, there was still an association with CD4 cell count decline. It may be that the immune reaction to malaria persists beyond the period of acute illness and the 8 to 9 weeks of increased HIV viral load. It would be useful for additional studies to examine the effect of malaria on direct outcomes of HIV disease progression, such as morbidity and mortality. The findings that malaria is associated with increased viral load and a more rapid decrease in CD4 cell count support the importance of malaria prevention among persons with HIV, including the routine use of cotrimoxazole prophylaxis and insecticide-treated bed nets. Jonathan Mermin John R. Lule John P. Ekwaru Centers for Disease Control and Prevention-Uganda, Global AIDS Program, Centers for Disease Control and Prevention, Uganda Virus Research Institute, Entebbe, Uganda