Adverse Pregnancy Outcomes in an Area Where Multidrug-Resistant Plasmodium vivax and Plasmodium falciparum Infections Are Endemic

Abstract

Background

Plasmodium falciparum infection exerts a considerable burden on pregnant women, but less is known about the adverse consequences of Plasmodium vivax infection.

Methods

In Papua, Indonesia, where multiple drug resistance to both species has emerged, we conducted a cross-sectional hospital-based study to quantify the risks and consequences of maternal malaria.

Results

From April 2004 through December 2006, 3046 pregnant women were enrolled in the study. The prevalence of parasitemia at delivery was 16.8% (432 of 2570 women had infections), with 152 (35.2%) of these 432 infections being associated with fever. The majority of infections were attributable to P. falciparum (250 [57.9%]); 146 (33.8%) of the infections were attributable to P. vivax, and 36 (8.3%) were coinfections with both species. At delivery, P. falciparum infection was associated with severe anemia (hemoglobin concentration, <7 g/dL; odds ratio [OR], 2.8; 95% confidence interval [95% CI], 2.0–4.0) and a 192 g (95% CI, 119–265) reduction in mean birth weight (P < .001). P. vivax infection was associated with an increased risk of moderate anemia (hemoglobin concentration, 7–11 g/dL; OR, 1.8; 95% CI, 1.2–2.9; P = .01) and a 108 g (95% CI, 17.5–199) reduction in mean birth weight (P < .019). Parasitemia was associated with preterm delivery (OR, 1.5; 95% CI, 1.1–2.0; P = .02) and stillbirth (OR, 2.3; 95% CI, 1.3–4.1; P = .007) but was not associated with these outcomes after controlling for the presence of fever and severe anemia, suggesting that malaria increases the risk of preterm delivery and stillbirth through fever and contribution to severe anemia rather than through parasitemia per se.

Conclusions

These observations highlight the need for novel, safe, and effective treatment and prevention strategies against both multidrug-resistant P. falciparum and multidrug-resistant P. vivax infections in pregnant women in areas of mixed endemicity.

Pregnant women are particularly vulnerable to malaria; they have an increased risk of infection and a greater risk of severe disease, compared with their nonpregnant counterparts [1, 2]. In sub-Saharan Africa, even asymptomatic infections increase the risk of low birth weight, preterm delivery, and perinatal death [3, 4]. In Asia and South America, where the level of Plasmodium falciparum transmission is less than that in Africa, women have little acquired immunity, and malaria infections are more frequently symptomatic, with an associated increased risk of maternal and/or fetal death [5, 6]. In these regions, Plasmodium vivax frequently coexists with P. falciparum, often accounting for more than one-half of malaria infections [7]. Although P. vivax infection has received less attention than P. falciparum infection, it is clearly an important contributor to both maternal anemia and low birth weight [8-10].

The burden of malaria during pregnancy has been complicated by the emergence of antimalarial drug resistance. Whereas chloroquine resistance in P. falciparum emerged in Southeast Asia >40 years ago, drug resistance in P. vivax has been slower to evolve. The first chloroquine-resistant isolates of P. vivax were reported from Papua, Indonesia, and Papua New Guinea in 1991 and have subsequently spread throughout the Indonesian archipelago [11-15]. In southern Papua, multidrug-resistant strains of both P. falciparum and P. vivax are now highly prevalent, with 65% of patients experiencing treatment failure within 28 days after chloroquine monotherapy for P. vivax and 48% experiencing treatment failure after treatment with chloroquine plus sulfadoxine-pyrimethamine for P. falciparum infection [16]. Unsupervised oral quinine is associated with 80% treatment failure, likely because of poor adherence to 7 days of therapy. Despite these limitations, chloroquine and quinine remained the standard treatments for P. vivax and P. falciparum malaria in pregnant women until 2006. The aim of the current study was to quantify the prevalence of P. vivax and P. falciparum infections in pregnant women in an area with high levels of drug resistance to both parasites and to describe the associated health-related consequences.

METHODS

Study site

The study was performed at Rumah Sakit Mitra Masyarakat (RSMM) Hospital, Timika, in the southern part of Papua, Indonesia. RSMM is the only hospital in this district, servicing an area of 21,522 km² with >180,000 people living in 85 villages within 12 subdistricts. The area is largely forested, with both coastal and mountainous areas. Malaria transmission is restricted to the lowland area, where it is associated with the 3 following mosquito vectors: Anopheles koliensis, Anopheles farauti, and Anopheles punctulatus. The annual incidence of malaria in the region is 938 infections per 1000 person-years, with a ratio of P. falciparum to P. vivax infections of 57:43 and with high rates of chloroquine resistance in both species [16].

Study population

In this region, the maternal mortality ratio is 1145 deaths per 100,000 live births, with infant mortality reaching 68 deaths per 1000 live births [17]. Each year, >3000 women become pregnant, ~1200 (~40%) of whom deliver at RSMM Hospital [18, 19]. Less than 40% of pregnant women attend antenatal care visits during their pregnancies. Neither intermittent presumptive treatment nor insecticide-treated nets programs are targeted specifically toward pregnant women in this region, and there is no routine HIV testing.

Because of economic migration, the ethnic origin of the local population is diverse, with highland Papuans, lowland Papuans, and non-Papuans all residing in the region. In view of the high number of infections in nonimmune patients, local protocols recommend that all patients with parasitemia at any level be given antimalarial therapy.

Data collection

All pregnant women admitted to RSMM Hospital from April 2004 through December 2006 were eligible for inclusion in this observational study, providing that they gave informed consent. All women received standard care according to hospital protocols.

Pregnant women were interviewed by a research nurse using a standardized questionnaire. Examinations of mothers and neonates were performed by the attending clinician or research clinician. Assessment of the gestational age using the New Ballard score [20] was performed by the research nurse. Information regarding history of febrile illness and treatment received during the current pregnancy was collected by interview and review of clinical notes.

Venous blood samples (5 mL) were obtained from pregnant women for complete blood counts and malaria smears. Malaria smears were checked daily until the time of hospital discharge. Patients who were found to be positive for malaria were treated according to RSMM Hospital protocol. Before March 2006, P. falciparum infection was treated with sulfadoxine-pyrimethamine and chloroquine or a 7-day course of quinine plus clindamycin therapy, and P. vivax infection was treated with chloroquine. Since March 2006, both P. falciparum infection and P. vivax infection have been treated with a 7-day course of quinine plus clindamycin during the first trimester and dihydroartemisinin-piperaquine during the second and third trimesters [21]. From April 2004 through May 2005, women with severe malaria were randomized to receive quinine or artesunate as part of a multicenter trial [22]; thereafter, intravenous artesunate was recommended for all cases of severe malaria.

Parasite counts were determined from the number of parasites per 200 WBCs on Giemsa-stained thick films, and peripheral parasite load was calculated using the recorded WBC count. Slide findings were considered to be negative only after review of 200 high-power fields. A thin smear was also examined to confirm parasite species and was used for quantification if the parasite load was >200 parasites per 200 WBC. Hemoglobin (Hb) concentration was determined by electronic counter (Coulter JT).

Malaria was defined as the presence of peripheral asexual parasitemia, irrespective of clinical signs or species. Maternal anemia was categorized as either moderate (Hb concentration, 7–11 g/dL) [23] or severe (Hb concentration, <7 g/dL) [24], and neonatal adverse outcomes were categorized according to World Health Organization criteria [23]. At admission to the hospital, women were defined as having a fever if they had a fever within the 24 h before hospital admission or an axillary temperature >37.5°C at hospital admission. Although routine antenatal surveillance was not available, the women who had a history of a febrile illness during their current pregnancy were considered to have a history of possible malaria.

Statistical analysis

Data from the questionnaire and laboratory findings were entered into EpiData software, version 3.02 (EpiData), and analyzed using SPSS, version 13.0 (SPSS). Normally distributed data were compared using Student's t test. Data that did not conform to a normal distribution were compared by the Mann-Whitney U test. Categorical data were compared by the χ² test with Yates correction or by Fisher's exact test with ORs and 95% CIs. Parasitemia and fever were assessed using a receiver operator curve, and the pyrogenic threshold was defined from the maximum value of Youden's Index (sensitivity plus specificity minus 1). A multiple logistic regression model was used to determine adjusted ORs for risk factors for adverse outcomes. Any variables found to be statistically significantly associated with the dependent variable in a univariate analysis were entered into the equation, and the model was constructed using a forward stepwise analysis of the Wald statistic, with P < .05 as the cutoff for statistical significance and inclusion of the predictor variable.

Ethical approval

Ethical approval for this study was obtained from the ethics committees of the National Institute of Health Research and Development, Ministry of Health, Indonesia, and Menzies School of Health Research, Darwin, Australia.

RESULTS

From April 2004 through December 2006, 3744 pregnant women were admitted to the maternity ward at RSMM Hospital; 3046 (81.4%) of these women were enrolled in the study. Of these 3046 women, 2518 (82.7%) were admitted to the hospital for delivery, 307 (10.1%) were admitted to the hospital for the treatment of malaria, and 221 (7.3%) were admitted to the hospital for other medical reasons. Seventy-five pregnant women (2.5%) were admitted to the hospital more than once during the observation period, the majority of whom (69 [92%] of 75 women) were admitted to the hospital twice. Of the 3046 women, 331 (10.9%) were discharged from the hospital with an ongoing pregnancy, 2601 women (85.4%) delivered, and pregnancy was aborted in 114 women (3.7%) (figure 1). The remainder of the analysis was restricted to women at the time of delivery, with baseline characteristics and outcomes shown in table 1.

Figure 1.

Flow of pregnant women with or without Plasmodium parasitemia (due to Plasmodium falciparum, Plasmodium vivax, Plasmodium ovale, and/or Plasmodium malariae) throughout the study.

Table 1. Maternal characteristics at delivery and pregnancy outcomes.

	Maternal malaria at delivery
Variable	Positive (n = 432)	Negative (n = 2138)	P
Maternal characteristic
Age, mean years (95% CI)	24.5 (23.9–25.1)	25.9 (25.7–26.1)	<.001
Gravidity, median no. of pregnancies (IQR)	2 (1–3)	2 (1–4)	<.001a
Primigravid	160/432 (37.0)	638/2135 (29.9)	.004
Multigravid	255/432 (59.0)	1415/2135 (66.3)	.005
Grand multigravidb	17/432 (3.9)	82/2135 (3.8)	NS
Ethnicity
Papuan	342 (79.2)	1360 (63.6)	<.001
Non-Papuan	89 (20.6)	774 (36.2)
Hemoglobin concentration, mean g/dL ± SD	8.9 ± 2.0	9.8 ± 1.9	<.001
Severe anemia	67/429 (15.6)	178/2127 (8.4)	<.001
Pregnancy outcome
Delivered
Singleton	418 (96.7)	2077 (97.1)	.9
Twins	13 (3.0)	61 (2.9)
Gestational age, median weeks (range)	38 (28–43)	38.6 (28–43)	<.001a
Preterm delivery	60/396 (15.1)	217/2000 (10.8)	.02
Birth weight,c mean g ± SD	2896 ± 567.9	3060 ± 509.2	<.001
Low birth weightc	70/387 (18.1)	212/2015 (10.5)	<.001
Stillbirth	17 (3.9)	38 (1.8)	.008
Early neonatal deaths	12 (2.8)	27 (1.3)	.03

NOTE. Data are no. (%) or proportion (%) of women, unless otherwise indicated. IQR, interquartile range; NS, not statistically significant.

^aDetermined by Mann-Whitney U test.

^bFifth or subsequent pregnancy.

^cLive singleton births.

Maternal Malaria

Malaria smear results were recorded for 2570 (98.9%) of 2601 pregnant women at delivery. The overall prevalence of parasitemia was 16.8% (432 of 2570 women had parasites found), with P. falciparum accounting for 250 infections (57.9%) and P. vivax accounting for 146 infections (33.8%); 15 infections (3.8%) were mixed infections (P. falciparum and P. vivax), 19 (4.4%) were caused by Plasmodium malariae, and 2 (0.5%) were caused by Plasmodium ovale. A history of possible malaria infection was reported by 388 (14.9%) of 2598 women, 76 (19.6%) of whom experienced >1 febrile episode and 356 (91.8%) of whom received empirical antimalarial treatment.

Of the 432 women with peripheral parasitemia who delivered a neonate, 152 (35.2%) were febrile or had a fever during the preceding 24 h. This proportion was significantly higher among women with P. falciparum infection (105 [42.0%] of 250 women) than among women with P. vivax infection (35 [23.9%] of 146; OR, 2.3; 95% CI, 1.4–3.7; P < .001). Women with P. falciparum infection also presented with higher parasite loads (geometric mean parasite load, 1622 parasites/μL; 95% CI, 1187–2215 parasites/μL) than did those infected with P. vivax (geometric mean parasite load, 602 parasites/μL; 95% CI, 414–875 parasites/μL; P < .001) (table 2). Using Youden's index, the pyrogenic density for P. falciparum was 2953 parasites/μL, and that for P. vivax was 632 parasites/μL. At delivery, 15 (0.6%) of 2601 women presented with ⩾1 modified World Health Organization criteria for severe malaria, and 14 (93.3%) of these 15 women had severe anemia.

Table 2. Geometric mean parasite load in mothers with Plasmodium falciparum or Plasmodium vivax infection.

	Parasite load, geometric mean parasites/μL (95% CI)
Mothers	P. falciparum	P. vivax
All	1622 (1187–2215)	602 (414–876)
Afebrile	956 (653–1399)	478 (316–723)
Febrile	3339 (2032–5487)	1336 (579–3083)

Three pregnant women died. One multigravid woman (at 38 weeks of gestation with fetal death in utero) had P. vivax infection and clinical features of severe bacterial sepsis. Two women (1 of whom was primigravid and 1 of whom was multigravid), both in their first trimester, had P. falciparum infection with multiorgan failure.

A total of 348 women received documented treatment for malaria at the time of delivery; 194 (55.7%) of these women received quinine with or without clindamycin, 92 (26.4%) received dihydroartemisinin-piperaquine, 12 (3.4%) received intravenous artesunate plus dihydroartemisinin-piperaquine, 43 (12.4%) received chloroquine plus sulfadoxine-pyrimethamine, and 7 (2.0%) received oral artesunate only. An additional 84 (19.4%) of the 432 women with peripheral parasitemia discharged themselves from the hospital before the administration of antimalarial medication.

Risk Factors for Maternal Malaria

Parasitemia at hospital admission was present in 160 (20.0%) of 798 primigravid women, compared with 272 (15.4%) of 1770 multigravid women (P = .004). The magnitude of the risk of malaria associated with primigravid women was similar for both P. falciparum and P. vivax infections (table 3). Other univariate risk factors are shown in table 3. In a multivariate model, the 4 following factors were found to be independently associated with parasitemia: primigravidity (OR, 1.3; 95% CI, 1.03–1.7), Papuan ethnicity (OR, 2.3; 95% CI, 1.8–2.9), fever or history of fever during the 24 h before hospital admission (OR, 10.8; 95% CI, 7.8–15.0), and history of possible malaria infection during the current pregnancy (OR, 2.6; 95% CI, 2.0–3.5).

Table 3. Univariate risk factors for parasitemia at hospital admission.

	Parasitemia, OR (95% CI)
Risk factor	Any	Plasmodium falciparuma	Plasmodium vivaxa
Primigravid	1.4 (1.1–1.7)b	1.4 (1.0–1.8)c	1.4 (1.03–2.0)c
Age ⩽16 years	2.4 (1.4–4.0)b	2.4 (1.3–4.6)c	2.9 (1.4–6.1)c
Papuan ethnicity	2.2 (1.7–2.8)d	3.4 (2.4–4.9)d	1.2 (0.8–1.7)e
Fever or history of fever in previous 24 h	14.3 (10.6–19.3)d	19.1 (13.6–26.8)d	8.3 (5.3–12.9)d
History of possible malaria infection during pregnancy	4.7 (3.7–6.0)d	5.7 (4.3–7.6)d	3.5 (2.4–5.1)d

^aRisks compared with women without parasitemia.

^bP < .005.

^cP < .05.

^dP < .001.

^eP = .37.

Maternal Anemia

The mean Hb concentration at delivery was 9.7 g/dL (95% CI, 9.6–9.8 g/dL), with moderate anemia noted in 1691 (66.1%) of 2558 women and severe anemia noted in 245 (9.6%) of 2558 women. Mothers infected with P. falciparum had a mean Hb concentration that was 1.1 g/dL (95% CI, 1.0–1.4 g/dL) lower than the mean Hb concentration in mothers who did not have parasitemia, and they had an increased risk of severe anemia, compared with women who did not have parasitemia (OR, 2.8; 95% CI, 2.0–4.0; P < .001). The difference in Hb concentration between women who did and did not have P. vivax infection was more modest (0.4 g/dL; 95% CI, 0.1–0.7 g/dL), and there was an increased risk of moderate anemia (but not of severe anemia) among women with P. vivax parasitemia, compared with women who did not have P. vivax parasitemia (OR, 1.8; 95% CI, 1.2–2.9; P = .01). Severe anemia (Hb concentration, <7g/dL) was more prevalent among multigravid women than among primigravid women (prevalence of severe anemia, 10.5% [185 of 1764 multigravid women] vs. 7.6% [60 of 792 primigravid women]; OR, 1.4; 95% CI, 1.04–1.96; P = .02). Other risk factors are presented in table 4. The risk of anemia was also apparent in women with asymptomatic malaria (OR, 1.7; 95% CI, 1.2–2.5; P = .004).

Table 4. Risk factors for severe maternal anemia and low birth weight.

Risk factor	Univariate analysis, OR (95% CI)	Multivariate analysis, adjusted OR (95% CI)
Maternal severe anemia
Malaria
No	1	1
Plasmodium falciparum	2.8 (1.9–4.0)a	1.9 (1.3–2.9)a
Plasmodium vivax	0.6 (0.3–1.4)b	NS
Papuan ethnicity	3.4 (2.4–4.9)a	2.7 (1.9–3.8)a
Fever or history of fever in previous 24 h	2.4 (1.7–3.14)a	1.7 (1.1–2.7)c
Age ⩽16 years	2.0 (1.1–3.9)c	2.0 (1.0–4.0)c
Multigravid	1.4 (1.0–2.0)c	1.6 (1.2–2.3)b
Low birth weight
Malaria
No	1	1
Any parasitemia	1.9 (1.4–2.6)a	1.5 (1.04–2.2)c
P. falciparum	1.9 (1.2–2.7)c	…
P. vivax	1.9 (1.2–3.1)c	…
Severe anemia	3.3 (2.4–4.5)a	3.5 (2.3–5.3)a
Papuan ethnicity	1.5 (1.2–2.0)d	NS
Primigravid	1.7 (1.3–2.2)a	1.7 (1.3–2.4)a
Prematurity	22.8 (16.7–31.2)a	22.0 (15.9–30.5)a

NOTE. Not statistically significant.

^aP < .001.

^bP= .005.

^cP < .05.

^dP= .01.

Pregnancy Outcomes

Of the 2601 women who reached the end of pregnancy during their hospital admission, 2505 (96.3%) had live births (including 67 twins), 55 (2.1%) had stillbirths, and 41 (1.6%) delivered neonates who died.

Preterm delivery

Gestational age was assessed for 2419 (93.0%) of 2601 women who delivered, with a prevalence of preterm delivery of 11.5% (279 women). The risk of preterm delivery (<37 weeks of gestation) was 15.2% (60 of 396 women) among women with maternal malaria, compared with 10.9% (217 of 2000) among women who did not have malaria (OR, 1.5; 95% CI, 1.1–2.0; P = .02). However, the risk of preterm delivery while infected with malaria was not apparent after controlling for fever (adjusted OR, 2.0; 95% CI, 1.3–2.8; P = .001), severe anemia (adjusted OR, 2.0; 95% CI, 1.3–2.8; P < .001), and primigravidity (adjusted OR, 1.5; 95% CI, 1.2–2.0; P = .002).

Low birth weight

The mean birth weight (±SD) of live singleton neonates in this study was 3024±537 g, compared with 2196±648 g for twin neonates (P < .001). Subsequent analysis is restricted to live singleton births. Low birth weight was observed in 212 (10.5%) of 2015 neonates of mothers who did not have malaria, compared with 70 (18.1%) of 387 neonates of mothers with malaria (OR, 1.9; 95% CI, 1.4–2.6; P < .001), resulting in a population-attributable risk of 9%. Maternal malaria was a significant independent risk factor for low birth weight (adjusted OR, 1.5; 95% CI, 1.04–2.2; P = .03) in multivariate analysis (table 4). The risk of low birth weight was also apparent among women with asymptomatic malaria (OR, 1.5; 95% CI, 1.1–2.1; P = .008).

The mean birth weight of neonates delivered from mothers with P. falciparum infection at delivery was 192 g (95% CI, 119–265 g) lower than that of neonates of mothers who did not have such infection. Although the decrease in birth weight associated with P. vivax infection was more modest (108 g; 95% CI, 17.5–199 g; P = .019), both P. vivax and P. falciparum infections were associated with a similar risk of low birth weight (17.9% [39 of 218 neonates] and 18.5% [25 of 135 neonates], respectively) (table 4).

Perinatal deaths

Stillbirths occurred in 55 (2.1%) of 2601 women, and early neonatal deaths occurred for 41 (1.6%) of 2601 women. Mothers with P. falciparum parasitemia had an increased risk of having a stillbirth (OR, 2.3; 95% CI, 1.2–4.2; P = .008) and early neonatal death, compared with women who were aparasitemic (OR, 2.2; 95% CI, 1.1–4.6; P = .03). After controlling for fever (adjusted OR, 2.0; 95% CI, 1.3–2.8; P < .001) and severe anemia (adjusted OR, 2.0; 95% CI, 1.4–2.8; P < .001), the difference in such risk was no longer statistically significant.

DISCUSSION

In Asia, P. falciparum and P. vivax infections are usually equally prevalent, although P. vivax infection is frequently assumed to be benign, and its associated morbidity is often ignored [7, 25]. To investigate the relative impact of both species on maternal malaria, we conducted a study in an area in Papua, Indonesia, where health services are limited and high levels of multidrug resistance exist in both P. falciparum and P. vivax [16]. Our results revealed that more than one-half of all women receiving empirical antimalarial therapy during pregnancy had subsequent patent parasitemia at delivery. Because rates of reinfection were low (1–3 infective bites per year), many of these infections reflect recrudescence of drug-resistant malaria and relapses of P. vivax infection. We were unable to observe women throughout their pregnancies; thus, our analysis focused on the adverse effects associated with peripheral parasitemia at delivery. At delivery, 17% of women had peripheral parasitemia. P. vivax infection accounted for 34% of infections in our study, compared with 17% in a study from Thailand [5] and 16% in a study from Papua New Guinea in the 1980s [26]—a likely consequence of the high prevalence of drug-resistant strains of P. vivax [16] and the paucity of antenatal health care in Papua. The higher transmission rate found in our study, compared with the transmission rate found in the study in Thailand, may provide enough immunity to suppress both symptoms and parasitemia and result in persistent infections going undetected and untreated. Although P. vivax had a lower pyrogenic density (632 parasites/μL) than P. falciparum (2953 parasites/μL), only 24% of women infected with P. vivax presented with symptoms, compared with 42% of women with P. falciparum infection who presented with symptoms, making P. vivax–infected women less likely to seek treatment. When either species was associated with fever, the risk of anemia and premature delivery almost doubled. Asymptomatic infections were also associated with poor outcome, but this was apparent only for low birth weight and severe maternal anemia.

In Papua, both P. falciparum and P. vivax malaria susceptibilities were higher in first-time mothers than in multigravid women. The modest difference in susceptibility contrasts with the marked effect of parity in areas of high P. falciparum transmission [3]. Similar to studies from Thailand and India [5, 9] (but not Africa or Papua New Guinea [3, 24, 27]), multigravid women were more anemic than primigravid women in our study. Less exposure to malaria and failure to develop protective immunity during the first pregnancy provide a plausible explanation, with iron deficiency, micronutrient deficiency, helminth infection, and HIV infection likely to be significant cofactors [24, 28-31].

In areas of high transmission, the effect of P. falciparum infection on birth weight is associated with chronic placental infection and inflammation rather than with peripheral parasitemia [32-34]. However, in areas of low transmission, placental changes are less common, and reduction in birth weight can occur even with a single episode of malaria [5, 8, 35]. In our study, the most important determinant of low birth weight was premature delivery, with 22% of premature deliveries being associated with malaria. In full-term deliveries, both P. falciparum and P. vivax parasitemia were associated with a similar risk of low birth weight. Low birth weight of the neonate after full-term delivery was unlikely to have been a result of infection at the time of delivery but, instead, may have been a result of repeated episodes or sustained parasitemia during late pregnancy. We did not continue to observe neonates after discharge from the hospital and, therefore, could not quantify the overall burden of maternal malaria on infant mortality. However, because birth weight is inversely correlated with the risk of death during the first year of life [6, 36], the low birth weight observed among neonates of mothers with P. falciparum and P. vivax infections in Papua is likely to make a significant contribution to the high infant mortality present in the region.

Our study also demonstrated an association between maternal malaria and both prematurity and stillbirth, the major determinants of which were concomitant fever and severe anemia. After controlling for these factors, the effect of parasitemia was no longer apparent. Fever from any cause has been shown to be independently associated with premature labor [6], suggesting that malaria exerts an adverse effect on the outcome of pregnancy through an inflammatory response and exacerbation of anemia rather than through peripheral parasitemia per se. Indeed, the success of intense serial antenatal malaria screening and prompt antimalarial treatment at the Thai-Burmese border is likely to have resulted both from an absolute reduction in the number of parasitemic episodes and from the detection of parasitemia before the onset of fever [5].

In conclusion, our study demonstrated a significant burden of maternal malaria in Papua that was associated with both P. vivax and P. falciparum infection. In areas where intense antenatal screening cannot be sustained, a strategy of both opportunistic and routine screening of pregnant women presenting to clinics, irrespective of the presence of symptoms, and the prompt administration of effective antimalarial therapy will be a key element of any successful antenatal intervention. The additional efficacy of insecticide-treated bed nets and intermittent presumptive therapy in regions of mixed endemicity is largely unknown, and the investigation of these infection-control measures warrants priority.