Abstract

BackgroundTo assess the effects of human immunodeficiency virus (HIV) infection on susceptibility to malaria, we compared the incidence rates of malaria by HIV type 1 (HIV-1) serostatus, baseline blood HIV-1 RNA concentration, and baseline CD4 cell count, over the course of a malaria season

MethodsWe followed a cohort of 349 adults in Malawi. For the 224 HIV-1–seropositive adults (64% of the cohort), we measured HIV-1 RNA concentration (n=187) and CD4 cell count (n=184) at baseline. Parasitemia was defined as presence of asexual parasites on a thick film of blood and was treated with sulfadoxine/pyrimethamine (SP), in accordance with national policy. Hazard ratios (HRs) of parasitemia were estimated using Cox regression. Demographics were adjusted for

ResultsHIV-1 seropositivity was associated with parasitemia (adjusted HR, 1.8 [95% confidence interval {CI}, 1.2–2.7] for a first parasitemia episode; adjusted HR, 2.5 [95% CI, 1.5–4.2] for a second parasitemia episode [>14 days after the first episode]; adjusted HR, 1.9 [95% CI, 1.4–2.6] for parasitemia overall). Treatment failure (parasitemia ⩽14 days after SP treatment) did not differ by HIV-1 serostatus (risk ratio, 1.3 [95% CI, 0.5–3.2]). HIV-1 RNA concentrations and CD4 cell counts were moderately but inconsistently associated with parasitemia. A high parasite density with fever was associated with HIV-1 seropositivity and low CD4 cell count

ConclusionHIV-infected adults in malaria-endemic areas are at increased risk for malaria. Where possible, additional malaria prevention efforts should be targeted at this population

Plasmodium falciparum malaria and HIV disease are endemic in many parts of the world. Adverse effects of either infection on the other could have serious implications [1–5 ]. The importance of understanding this interaction has increased as treatment and methods for control of both diseases continue to become increasingly available worldwide

HIV infection may increase the incidence and severity of malaria in adults by compromising acquired immunity to malaria, particularly in malaria-endemic settings. In pregnant women, HIV infection has been associated with increased prevalence and density of P. falciparum in peripheral and placental blood and with increased risk of malaria-related consequences in the individual and her child [6–9 ]. In Ugandan adults, HIV seropositivity and low CD4 cell count have been associated with increased incidence of clinical malaria [1, 4]. However, the incidence of recurrent clinical malaria (i.e., >28 days after the first episode) was not affected by CD4 cell count [1]

Important aspects of the effect of HIV infection on malaria have not been thoroughly examined. First, the relationship between blood HIV-1 RNA concentration and subsequent malaria has not been assessed. Understanding this relationship may be useful, since blood viral load is a direct marker of HIV disease severity and may serve as an important clinical marker in coinfected persons. Second, although clinical malaria is important, the incidence of any parasitemia may also be relevant, because asymptomatic parasitemia is common in malaria-endemic regions and because parasitemia is followed by a reversible increase in viral load in HIV-infected individuals [3]. Finally, CD4 cell count measured during a malaria episode or within a short window around the episode, as was done in a previous study [4], may not accurately represent an individual’s immune status, because the malaria infection itself may result in transient changes in CD4 cell counts. Using a single baseline measure of CD4 cell count with a short follow-up period would circumvent this problem

We conducted a prospective cohort study of adults in rural Malawi, to examine the effects of HIV infection on susceptibility to P. falciparum infection while addressing the issues highlighted above. Our objective was to estimate incidence rates of a first parasitemia episode and of a second parasitemia episode, as well as overall rates of parasitemia occurring through the malaria season. We also sought to estimate incidence rates of a first episode of symptomatic malaria and the incidence of recrudescent parasitemia. We assessed the impact of HIV-1 serostatus, baseline HIV-1 RNA concentration, and baseline CD4 cell count on these incidence measures

Subjects and Methods

DesignThe study was conducted on a tea estate in Thyolo District, Malawi. Using a prospective cohort design, we enrolled participants before the malaria season (between 1 October 2000 and 31 December 2000) and followed them through the end of the malaria season (30 June 2001), which roughly coincides with the rainy season. Ethics committees at the University of Malawi College of Medicine, the Liverpool School of Tropical Medicine, and the University of North Carolina approved the protocol. Informed consent for HIV testing and participation in the study was received from all enrollees

PopulationParticipants were adults recruited through an HIV screening process. HIV-1 seropositivity was determined by at least 2 rapid diagnostic assays (Determine, Abbott; Unigold, Trinity Biotech; Hemostrip HIV, Saliva Diagnostics). Antiretroviral treatment was not available. Adults with clinical AIDS–defining features [10], a recent history of tuberculosis, or a lack of intention to remain in the area for at least 9 months from enrollment were excluded. Enrollment was postponed for febrile (sublingual temperature, >37.4°C) individuals until they were adequately treated and asymptomatic at baseline. A random sample of adults who were HIV-1 seronegative at screening were frequency matched to HIV-1–seropositive participants by residential compound (∼2:1) and were also enrolled. We restricted our analyses to adults with known HIV-1 serostatus who were aparasitemic at enrollment and through the first 10 days of follow-up and had at least 1 follow-up visit beyond 10 days from enrollment

Schedule of visitsPersons enrolled in the study were requested to follow a schedule of visits, including a baseline visit (enrollment), routine visits every 8 weeks, interim visits when ill, and 4 treatment visits after administration of malaria chemotherapy. Parasitemia was treated with 1 dose of 3 tablets of sulfadoxine/pyrimethamine (SP) (Fansidar; Roche Isando) containing 25 mg of pyrimethamine and 500 mg of sulfadoxine, in accordance with national policy in Malawi

Data collectionAt baseline, we collected demographic data and measured CD4 cell counts (FACSCount; Becton Dickinson). At all visits, a medical history was taken, a physical examination was conducted, and a blood sample was collected for obtaining complete blood counts, detecting malaria parasites, and measuring HIV-1 RNA concentrations

P. falciparum parasites were identified using thick films. Slides were independently read at least twice. Additional readings were performed for discordant results. The mean reading served as the final parasite density. When at least half of the readings were 0, the slide was considered negative. For each reading, parasite density was calculated using the white blood cell (WBC) count measured at the same visit. For 69 persons, WBC counts were not available at the malaria visit, so the individual’s median WBC count was used instead

Plasma HIV-1 RNA concentration was quantified by polymerase chain reaction (Amplicor HIV Monitor 1.5 assay; Roche Molecular Systems) at the Malawi-Liverpool Wellcome Laboratories (Blantyre, Malawi). This assay had a lower detection limit of 400 copies/mL. Fever (self-reported or documented sublingual temperature, >37.4°C) was investigated by clinical and laboratory examination, including blood cultures, and was followed up with appropriate medical intervention when indicated

Exposures and outcomesWe examined the effect of HIV-1 serostatus and, among HIV-1–seropositive persons, the effect of baseline blood HIV-1 RNA concentrations and baseline CD4 cell counts on malaria incidence. We examined log HIV-1 RNA (referred to hereafter as HIV-1 RNA) concentrations using continuous terms as well as in categories: 2–3.999 log copies/mL (referent), 4–4.999 log copies/mL, and 5–6.999 log copies/mL. We examined CD4 cell count in 3 categories: <200 cells/mL, 200–399 cells/mL, and ⩾400 cells/mL (referent). We also examined CD4 cell count using linear terms for improved model fit. Linear representations of HIV-1 RNA concentration and CD4 cell count were justified, since both variables were approximately linear in the log-hazard

We calculated 3 different incidence rates of parasitemia, where parasitemia was defined as presence of parasites, regardless of fever or parasite density: (1) incidence rate of a first parasitemia episode from baseline; (2) incidence rate of a second parasitemia episode (an episode that occurred >14 days after the first episode); and (3) overall incidence rate of parasitemia throughout the study period, incorporating the first, second, and all subsequent episodes of parasitemia

We defined parasitemia episodes occurring within 14 days of the first episode as “recrudescent” episodes, in accordance with World Health Organization (WHO) 1996 criteria [11]. We also determined the incidence rate of an individual’s first episode of symptomatic malaria, defined as a parasite density ⩾2000 parasites/μL and presence of fever (self-reported or documented sublingual temperature, >37.4°C)

We chose to define first, second, and recrudescent parasitemia as described above after considering the limitations of our choices. We opted to use 14 days as an approximate period to distinguish recrudescence from reinfection, but we repeated the analyses using 28 days

Statistical methodsData were double-entered and validated using EpiInfo (version 6.04b; CDC and WHO). Analyses were conducted using Stata (version 7; Stata Corporation)

In all of our regression models, we accounted for potential clustering by residential compound by use of the Huber-White sandwich, or “robust,” variance estimator. We accounted for clustering by compound, because residents within a compound were likely to have similar exposure to Plasmodium-infected mosquitoes

We estimated incidence rates with robust 95% confidence intervals (CIs) by Poisson regression. To estimate incidence rate ratios of first and second parasitemia episodes, we used hazard ratios (HRs) and 95% CIs from Cox proportional-hazards models. We used the Efron method for ties

To estimate the overall incidence rate ratio of parasitemia, we used an extension of the Cox proportional-hazards model for analysis of multiple failure–time data to determine HRs and 95% CIs. Specifically, we used conditional risk set models for ordered events, where the time to an event is calculated from the previous event and the correlation of failure times within an individual is accounted for by adjusting the covariance matrix of the independent variables [12]

We examined potential effect-measure modification and confounding by sex, age, marital status, education, and employment status. We used a backward elimination procedure to remove variables from the full model on the basis of change of the beta coefficients for the factors of interest. No interaction terms were retained in the models. We calculated risk ratios (RRs) with 95% CIs to assess whether the cumulative incidence of recrudescent parasitemia occurring 3–14 days after the first episode [11] differed by HIV-1 serostatus

Results

ParticipantsHIV-1 serostatus was determined in 99% of the 2383 adults screened (figure 1). Of the 656 persons who tested positive for HIV-1, 367 (56%) were enrolled in the study. Of the 1710 persons who tested negative for HIV-1, a random sample of 202 (12%) were enrolled in the study. Thus, 569 (24%) of those screened were enrolled. Of the 569 adults, 440 (77%) were enrolled by 31 December 2000. Of these 440 adults, 410 (93%) were aparasitemic at baseline. Of the 410 adults, 349 (85%) had follow-up visits beyond 10 days after baseline

Figure 1

Derivation of the study cohort

Figure 1

Derivation of the study cohort

In the full study sample (n=349), 148 individuals (42%) were women, and the mean age was 32 years (range, 18–68 years). Most individuals were married, had a primary education, and were employed at the estate (table 1). The median follow-up time was 6.2 months (interquartile range [IQR], 4.2–7.2 months). Of the 349 individuals, 224 (64%) were HIV-1 seropositive, and 125 (36%) were HIV-1 seronegative

Table 1

Selected characteristics of the study population (n=349)

Table 1

Selected characteristics of the study population (n=349)

In analyses restricted to HIV-1–seropositive persons (n = 224), 41% were women. The distribution of sex, age, marital status, and education level in the HIV-1–seropositive population was similar to that of the HIV-1–seronegative population, although greater proportions of HIV-1–seropositive persons were employed. In HIV-1–seropositive persons with baseline HIV-1 RNA concentration data (n=187), the median HIV-1 RNA concentration was 85,422 copies/mL (IQR, 22,885–215,504 copies/mL), corresponding to a mean of 4.93 log copies/mL. In HIV-1–seropositive persons with baseline CD4 cell count data (n=184), the median CD4 cell count was 337 cells/mL (IQR, 223–483 cells/mL). Persons excluded because of a lack of HIV-1 RNA concentration and CD4 cell count measurements were similar to those included in the analyses, except that more men than women were missing HIV-1 RNA concentration (P = .01) and CD4 cell count (P = .09) measurements at baseline

Effect of HIV-1 serostatus on incidence of parasitemiaIn the full sample of 349 adults, 147 (42%) had parasitemia during the follow-up period. These 147 persons experienced a total of 254 parasitemia episodes: 147 (58%) first episodes, 35 (14%) recrudescent episodes, and 72 (28%) second and subsequent episodes

The number of first parasitemia episodes was 108 (48%) among 224 HIV-1–seropositive persons and 39 (31%) among 125 HIV-1–seronegative persons. The incidence rate of a first or subsequent episode was approximately twice as high in HIV-1–seropositive persons than in HIV-1–seronegative persons. Adjusted HRs were similar to unadjusted HRs (table 2)

Table 2

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, comparing HIV-1–seropositive persons (n=224) with HIV-1–seronegative persons (n=125)

Table 2

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, comparing HIV-1–seropositive persons (n=224) with HIV-1–seronegative persons (n=125)

When recrudescent episodes were excluded, HIV-1–seropositive persons had 172 episodes overall (mean, 1.6 episodes/person), and HIV-1–seronegative persons had 47 episodes overall (mean, 1.2 episodes/person). Among HIV-1–seropositive persons with a first episode of parasitemia and subsequent follow-up visits (n = 102), 38 (37%) experienced a second episode >14 days after the first episode, corresponding to an incidence rate of 5.8/1000 person-days (95% CI, 4.2–7.8/1000 person-days) (table 2). In contrast, among HIV-1–seronegative persons (n=35) with 1 parasitemia episode, 7 (20%) experienced a second episode (incidence rate, 2.6/1000 person-days [95% CI, 1.2–5.4/1000 person-days]), corresponding to an HR of ∼2.5, irrespective of adjustment. A similar relationship was observed when all episodes occurring during the study period were analyzed

The cumulative incidence of recrudescent episodes during the 14-day posttreatment interval was 0.19 (20/108; 95% CI, 0.12–0.27) in HIV-1–seropositive persons and 0.14 (5/35; 95% CI, 0.05–0.30) in HIV-1–seronegative persons, corresponding to a nonsignificant RR of 1.3 (95% CI, 0.5–3.2). Redefining recrudescence as an episode occurring within 28 days yielded similar results

Effect of baseline HIV-1 RNA concentration on incidence of parasitemiaIn analyses limited to HIV-1–seropositive persons with baseline HIV-1 RNA concentration measurements (n=187), we observed increased incidence rates of parasitemia in persons with baseline HIV-1 RNA concentrations ⩾4 log copies/mL (table 3). When a continuous representation of HIV-1 RNA concentration was used, the unadjusted HR for a first episode of parasitemia indicated a 20% relative increase in incidence for every 1-log increase in HIV-1 RNA concentration. Adjusted estimates were similar (table 3)

Table 3

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, by baseline log HIV-1 RNA concentration

Table 3

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, by baseline log HIV-1 RNA concentration

The association of HIV-1 RNA concentration with a second parasitemia episode was even stronger. A second episode, occurring a mean of 64 days after the first, was strongly and linearly associated with baseline HIV-1 RNA concentration (n = 101) (adjusted HR, 2.12 [95% CI, 1.14–3.92]) (table 3)

Through the malaria season, the relationship between HIV-1 RNA concentration and overall incidence of parasitemia was similar to that for a first episode and was weaker after adjustment (table 3). To assess whether sparsely populated groups of later episodes weakened this relationship, we conducted a sensitivity analysis limited to the first 4 episodes of parasitemia. Point estimates of the HR increased slightly but were less precise (unadjusted HR, 1.27 [95% CI, 0.92–1.74]; adjusted HR, 1.23 [95% CI, 0.92–1.63])

Effect of baseline CD4 cell count on incidence of parasitemiaA first parasitemia episode was inversely related to baseline CD4 cell count in HIV-1–seropositive persons (n=184) (table 4). This trend was demonstrated by an unadjusted HR of 1.6 (95% CI, 1.0–2.4) for persons with CD4 cell counts of 200–399 cells/mL (n=78) and 2.0 (95% CI, 1.1–3.8) for persons with CD4 cell counts <200 cells/mL (n=28), compared with persons with CD4 cell counts ⩾400 cells/mL (n=78). Adjustment for other variables had little effect on these estimates

Table 4

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, by baseline CD4 cell count

Table 4

Incidence rates (IRs) and hazard ratios (HRs) with 95% confidence intervals (CIs) of a first parasitemia episode and a second parasitemia episode and the overall rate of parasitemia, by baseline CD4 cell count

We observed a distinct lack of association between baseline CD4 cell count and a second parasitemia episode (n=84), whether CD4 cell count was represented categorically or continuously (table 4). Notably, for persons with CD4 cell counts <200 cells/mL, the incidence rate of a second parasitemia episode was comparable to that of a first parasitemia episode; for the other CD4 cell count groups, however, incidence rates of a second parasitemia episode increased over the corresponding rates of a first parasitemia episode. Thus, HIV-1 infection, regardless of CD4 cell count, conferred increased risk of a second parasitemia episode

Persons with CD4 cell counts <400 cells/mL experienced higher incidence rates of parasitemia during the study period, compared with those with CD4 cell counts ⩾400 cells/mL (table 4). Overall incidence rates by CD4 cell count were similar to incidence rates for a first parasitemia episode

Effect of the 3 factors of interest on incidence of a first episode of symptomatic malariaSixty-five (19%) of the 349 adults experienced symptomatic malaria. Fifty-three (24%) of 224 HIV-1–seropositive adults had symptomatic malaria, compared with 12 (10%) of 125 HIV-1–seronegative adults. The incidence rate of a first episode of symptomatic malaria was significantly higher in HIV-1–seropositive persons (unadjusted HR, 2.7 [95% CI, 1.5–5.0]) and remained unchanged after adjustment for sex and age (HR, 2.7 [95% CI, 1.2–6.1])

HIV-1–seropositive persons with HIV-1 RNA concentration measurements (n=187) experienced 42 first episodes of symptomatic malaria. HIV-1 RNA concentration did not appear to influence the incidence rate of symptomatic malaria (unadjusted HR, 0.94 [95% CI, 0.67–1.30]). Adjustment for sex, age, employment status, and marital status had little effect (HR, 1.08 [95% CI, 0.70–1.67])

Persons with baseline CD4 cell count measurements (n=184) experienced 32 first episodes of symptomatic malaria. The HR was 2.3 (95% CI, 1.0–5.5) for persons with CD4 cell counts <200 cells/mL and was 1.2 (95% CI, 0.5–2.7) for persons with CD4 cell counts of 200–399 cells/mL, indicating that the rate of a first episode of symptomatic malaria was associated with low CD4 cell counts at baseline. The HR after adjustment for sex, age, marital status, and employment was 2.4 (95% CI, 0.9–6.4) for persons with CD4 cell counts <200 cells/mL and was 1.2 (95% CI, 0.5–2.9) for persons with CD4 cell counts of 200–399 cells/mL

Discussion

HIV-seropositive persons, when compared with HIV-seronegative persons, had significantly higher incidence rates of a first P. falciparum malaria episode during the study period, regardless of whether symptoms were present. Among persons who became parasitemic, those who were HIV-seropositive had significantly higher incidence rates of a second parasitemia episode and significantly higher incidence rates of parasitemia through the malaria season. However, HIV-seropositive persons did not have a significantly greater incidence of recrudescent parasitemia

Our findings regarding HIV seropositivity and symptomatic malaria are consistent with those of a previous study that found an association of HIV serostatus and clinical malaria [4]. We extended this finding by demonstrating that persons who are HIV seropositive have significantly higher rates of a second parasitemia episode and of overall parasitemia through a malaria season in a region where malaria is highly endemic with stable transmission. This finding is important because parasitemia is associated with an increase in viral load [3]

The higher rates of malaria in HIV-seropositive persons may be explained by impaired immune responses to P. falciparum—in particular, CD4 cell activity—as a result of HIV infection [13–17 ]. The increased susceptibility of HIV-seropositive pregnant women to malaria has been attributed to depletion of CD4 cells [15] and impairment of cytokine functions, including protective interferon-γ responses [18]

The efficacy of SP, as measured by recrudescence within 14 days, was not substantially affected by HIV serostatus, suggesting that treatment of malaria was equally efficacious during the 2-week period. However, our estimates of the cumulative incidence of recrudescent malaria are imprecise. Also, without genotyping, we cannot confirm that these episodes were truly recrudescent. We used a somewhat stringent definition of “recrudescent,” since recurrence of parasitemia may occur even 28 days after treatment with SP [19], but our results were considerably similar when a 28-day cutoff was used. Others, however, have suggested that antimalarial therapy may be less efficacious in HIV-seropositive adults [20–23 ]. Further investigation of the efficacy of antimalarials in the context of HIV infection is warranted

The inverse association of CD4 cell count with rates of a first parasitemia episode and symptomatic malaria in HIV-seropositive persons observed in our study is consistent with results of previous studies of clinical malaria and CD4 cell count [1, 4]. Persons with greater immune suppression have a greater incidence of a first parasitemia episode, but a second parasitemia episode was not associated with baseline CD4 cell count [1]. The effect of immune suppression in our study, however, may have been underestimated because of the exclusion of persons with clinical AIDS. On the other hand, our study demonstrated an association of HIV-1 RNA concentration with a first parasitemia episode that became stronger with a second parasitemia episode. In general, HIV infection appears to confer consistent increased risk of malaria infection, and, in some incidence measures, high HIV RNA concentrations or low CD4 cell counts further increase the risk

We defined a new malaria episode as one occurring >14 days after the previous episode, on the basis of the WHO 1996 protocol for assessing malaria treatment failure. To assess whether using a short window of time, such as 14 days, may lead to misidentification of old episodes as new episodes [24], we repeated our analyses of the 3 factors of interest, using an aparasitemic period >28 days to define a second episode. The results were not substantially altered (data not shown)

Although the observed differences in incidence rates of parasitemia by HIV-1 serostatus, baseline HIV-1 RNA concentration, and baseline CD4 cell count may have been biased because of higher detection of parasitemia in HIV-seropositive persons who made more interim visits, the difference in visit type by HIV serostatus was not statistically significant. Further, the study population, on the whole, was a relatively secluded population with few alternate health-care facilities. Thus, although some HIV-seronegative persons may have visited the clinic less frequently, most of their illnesses would have been captured by our study

Our study may not have captured all relevant visits and episodes of parasitemia, since some participants did not report to the clinic on schedule. We assumed that persons were aparasitemic between visits, some of which were spaced months apart. Missed visits that did not differ by exposure group may have resulted in lower precision because of smaller sample sizes, but this would have led to an underestimate of the HR. Although unfeasible, daily measures of parasitemia would have provided the most accurate data on incidence rates of parasitemia and their variation by the 3 factors of interest

Our use of a broad malaria definition (i.e., presence of any parasitemia) allowed us to assess the effects of HIV serostatus and baseline markers of HIV infection on a common condition in malaria-endemic regions. Only approximately half of the parasitemia episodes occurred with fever or parasite densities >2000 parasites/μL

Other data that were not measured, such as clinically confirmed history of malaria (from the previous season), sickle-cell status, genetic factors, nonobvious pregnancy, insecticide-treated net use, and other malaria prevention efforts, may have influenced the rate of malaria observed in our study. Each of these factors may have influenced malaria incidence in some individuals and may have led to residual confounding in our measures of effect. However, this potential confounding would not substantially change our findings, since these factors were unlikely to be strongly associated with HIV infection

The evidence for higher rates of parasitemia and a first episode of symptomatic malaria is strong and consistent in adults infected with HIV. This justifies the need for intensified malaria prevention strategies in regions where coinfection with malaria and HIV is common. Intermittent presumptive antimalarial chemotherapy and the use of insecticide-treated bed nets may be combined with administration of antiretrovirals and counseling on HIV disease for HIV-seropositive persons who are positive for parasitemia by blood smear, regardless of fever or parasite density. Malaria prophylaxis may be considered for HIV-seropositive persons to prevent malarial infection and reduce malarial illness

References

1.
French
N
Nakiyingi
J
Lugada
E
Watera
C
Whitworth
JA
Gilks
CF
Increasing rates of malarial fever with deteriorating immune status in HIV‐1‐infected Ugandan adults
AIDS
 , 
2001
, vol. 
15
 (pg. 
899
-
906
)
2.
Hoffman
IF
Jere
CS
Taylor
TE
, et al.  . 
The effect of Plasmodium falciparum malaria on HIV‐1 RNA blood plasma concentration
AIDS
 , 
1999
, vol. 
13
 (pg. 
487
-
94
)
3.
Plowe
C
Kublin
J
Dzinjalamala
F
, et al.  . 
Sustained clinical efficacy of sulfadoxine‐pyrimethamine for uncomplicated falciparum malaria in Malawi after 10 years as first line treatment: five year prospective study
BMJ
 , 
2004
, vol. 
328
 (pg. 
545
-
8
)
4.
Whitworth
J
Morgan
D
Quigley
M
, et al.  . 
Effect of HIV‐1 and increasing immunosuppression on malaria parasitaemia and clinical episodes in adults in rural Uganda: a cohort study
Lancet
 , 
2000
, vol. 
356
 (pg. 
1051
-
6
)
5.
Rowland‐Jones
SL
Lohman
B
Interactions between malaria and HIV infection—an emerging public health problem?
Microbes Infect
 , 
2002
, vol. 
4
 (pg. 
1265
-
70
)
6.
ter Kuile
FO
Parise
ME
Verhoeff
FH
, et al.  . 
The burden of co‐infection with human immunodeficiency virus type 1 and malaria in pregnant women in sub‐Saharan Africa
Am J Trop Med Hyg
 , 
2004
, vol. 
71
 (pg. 
41
-
54
)
7.
Steketee
RW
Wirima
JJ
Bloland
PB
, et al.  . 
Impairment of a pregnant woman’s acquired ability to limit Plasmodium falciparum by infection with human immunodeficiency virus type‐1
Am J Trop Med Hyg
 , 
1996
, vol. 
55
 (pg. 
42
-
9
)
8.
van Eijk
AM
Ayisi
JG
ter Kuile
FO
, et al.  . 
Human immunodeficiency virus seropositivity and malaria as risk factors for third‐trimester anemia in asymptomatic pregnant women in western Kenya
Am J Trop Med Hyg
 , 
2001
, vol. 
65
 (pg. 
623
-
30
)
9.
Verhoeff
FH
Brabin
BJ
Chimsuku
L
Kazembe
P
Broadhead
RL
Malaria in pregnancy and its consequences for the infant in rural Malawi
Ann Trop Med Parasitol
 , 
1999
, vol. 
93
 
Suppl 1
(pg. 
S25
-
33
)
10.
World Health Organization
Acquired immune deficiency syndrome (AIDS): interim proposal for a WHO staging system for HIV‐1 infection and disease
Wkly Epidemiol Rec
 , 
1990
, vol. 
65
 (pg. 
221
-
8
)
11.
World Health Organization
Assessment of therapeutic efficacy of antimalarial drugs for uncomplicated falciparum malaria in areas with intense transmission [WHO/MAL/96.1077]
 , 
1996
Geneva
WHO
12.
Hosmer
DW
Lemeshow
S
Applied survival analysis: regression modeling of time to event data
 , 
1999
New York
John Wiley & Sons
13.
Troye‐Blomberg
M
Human T‐cell responses to blood stage antigens in Plasmodium falciparum malaria
Immunol Lett
 , 
1994
, vol. 
41
 (pg. 
103
-
7
)
14.
Winkler
S
Willheim
M
Baier
K
, et al.  . 
Frequency of cytokine‐producing T cells in patients of different age groups with Plasmodium falciparum malaria
J Infect Dis
 , 
1999
, vol. 
179
 (pg. 
209
-
16
)
15.
Brown
J
Greenwood
BM
Terry
RJ
Cellular mechanisms involved in recovery from acute malaria in Gambian children
Parasite Immunol
 , 
1986
, vol. 
8
 (pg. 
551
-
64
)
16.
Taylor‐Robinson
AW
Phillips
RS
Protective CD4+ T‐cell lines raised against Plasmodium chabaudi show characteristics of either Th1 or Th2 cells
Parasite Immunol
 , 
1993
, vol. 
15
 (pg. 
301
-
10
)
17.
Worku
S
Bjorkman
A
Troye‐Blomberg
M
Jemaneh
L
Farnert
A
Christensson
B
Lymphocyte activation and subset redistribution in the peripheral blood in acute malaria illness: distinct γδ+ T cell patterns in Plasmodium falciparum and P. vivax infections
Clin Exp Immunol
 , 
1997
, vol. 
108
 (pg. 
34
-
41
)
18.
Moore
JM
Ayisi
J
Nahlen
BL
Misore
A
Lal
AA
Udhayakumar
V
Immunity to placental malaria. II. Placental antigen‐specific cytokine responses are impaired in human immunodeficiency virus–infected women
J Infect Dis
 , 
2000
, vol. 
182
 (pg. 
960
-
4
)
19.
Stepniewska
K
Taylor
W
Mayxay
M
, et al.  . 
In vivo assessment of drug efficacy against Plasmodium falciparum malaria: duration of follow‐up
Antimicrob Agents Chemother
 , 
2004
, vol. 
48
 (pg. 
4271
-
80
)
20.
Parise
ME
Ayisi
JG
Nahlen
BL
, et al.  . 
Efficacy of sulfadoxine‐pyrimethamine for prevention of placental malaria in an area of Kenya with a high prevalence of malaria and human immunodeficiency virus infection
Am J Trop Med Hyg
 , 
1998
, vol. 
59
 (pg. 
813
-
22
)
21.
Shah
S
Smith
E
Obonyo
C
Bloland
P
Slutsker
L
Hamel
M
The effect of HIV‐infection on antimalarial treatment response: preliminary results of a 28‐day drug efficacy trial in HIV‐infected and HIV‐uninfected adults in Siaya, Kenya [abstract 318]
Program and abstracts of the 53rd Annual Meeting of the American Society of Tropical Medicine and Hygiene (Miami Beach)
 , 
2004
Northbrook, IL
American Journal of Tropical Medicine and Hygiene
pg. 
93
 
22.
van Eijk
AM
Ayisi
JG
ter Kuile
FO
, et al.  . 
Effectiveness of intermittent preventive treatment with sulphadoxine‐pyrimethamine for control of malaria in pregnancy in western Kenya: a hospital‐based study
Trop Med Int Health
 , 
2004
, vol. 
9
 (pg. 
351
-
60
)
23.
Van geertruyden
J
Mwananyanda
L
Chalwe
V
, et al.  . 
Higher risk of antimalarial treatment failure in HIV positive than in HIV negative individuals with clinical malaria [abstract 315]
Program and abstracts of the 53rd Annual Meeting of the American Society of Tropical Medicine and Hygiene (Miami Beach)
 , 
2004
Northbrook, IL
American Journal of Tropical Medicine and Hygiene
pg. 
92
 
24.
Talisuna
AO
Bloland
P
D’Alessandro
U
History, dynamics, and public health importance of malaria parasite resistance
Clin Microbiol Rev
 , 
2004
, vol. 
17
 (pg. 
235
-
54
)
Presented in part: 53rd meeting of the American Society for Tropical Medicine and Hygiene, Miami Beach, Florida, 7–11 November 2004 (abstract 456)
Potential conflicts of interest: none reported
Financial support: The Wellcome Trust (project grant 058390); Fogarty International Center, National Institutes of Health (grant AITRP D43 TWO1039-05s1)
Present affiliation: Fred Hutchinson Cancer Research Center, Seattle, Washington