Volume 11, Number 9—September 2005
Research
Malaria Attributable to the HIV-1 Epidemic, Sub-Saharan Africa
, Brian G. Williams*, Sake J. de Vlas†, Eleanor Gouws‡, Charles F. Gilks*, Peter D. Ghys‡, and Bernard L. Nahlen*Table 1
HIV-1, malaria incidence and death rates. and their interactions*
| Parameter | Assumption |
|---|---|
| Malaria transmission intensity | Index >0 and <0.75 denotes low-intensity transmission and >0.75 denotes high-intensity transmission, except for southern Africa, where index >0.75 denotes unstable transmission (18,19) |
| Overall malaria incidence | Middle Africa, high-transmission areas: 1.4 per person per year in children <5 y, 0.59 per person per year at 5–14 y, 0.11 per person per year at >15 y (19) Middle Africa, low-transmission areas: 0.182 per person per year in children <15 y, 0.091 per person per year at >15 y (19) Southern Africa: 0.0294 per person per year as all-age average in areas with (unstable) malaria transmission; divided as twice the rate at >15 y compared to <14 y (19) |
| Relative malaria incidence urban/rural | 0.50 (20) |
| Malaria deaths | High-transmission areas: 0.8% of incident cases in children <5 y, 0.3% at >5 y; Low-transmission and unstable transmission areas: 0.8% of incident cases in all age groups (21). |
| Effect of HIV-1 on incidence of clinical malaria | >5 years in areas with high-intensity malaria transmission, and all age groups in areas with low-intensity or unstable malaria transmission: CD4 >500/μL RR = 1.2 CD4 200–499/μL RR = 3.0 CD4 <200/μL RR = 5.0† <5 years in high-transmission areas: no effect |
| Effect of HIV-1 on malaria case fatality rate | All malaria transmission intensities and age groups: CD4 >500/μL RR = 2.0 CD4 200–499/μL RR = 4.0 CD4 <200/μL RR = 10‡ |
| Survival after HIV-1 infection | Median 9 years, following a Weibull curve with shape parameter 2.28 (22) |
| CD4 decline over the course of HIV-1 infection | Linear from 825/μL at seroconversion to 20/μL at death of AIDS (23–26) |
*RR = relative risk associated with HIV-1 infection; y = years of age.
†From (9,10). Earlier studies did not consistently show these effects, but these were cross-sectional and/or hospital-based (1). Effects of HIV-1 in these studies may have been obscured by a lack of adjustment for prestudy treatment with antimalarial drugs (which might be more common in HIV-1 patients with recurrent fevers [27]) and by their inherent dependence on the relative survival of HIV-infected and HIV-uninfected participants, given the increased case fatality of malaria among HIV-infected patients (6). At the specified CD4-stratum-specific relative risks, the relative risk averaged over all HIV-infected people would be 2.1 in Madagascar and 2.5 in all other countries (see Methods, CD4 distributions among HIV-infected people).
‡At these CD4-stratum-specific relative risks, the relative risk averaged over all HIV-infected people would be 3.4 in Madagascar and 4.1 in all other countries (see Methods, CD4 distributions among HIV-infected people).
