Abstract
Background
Low birth weight (LBW), anaemia and malaria are common in Papua New Guinean women.
Methods
To identify risk factors for LBW, anaemia and preterm delivery (PTD), pregnant women recruited into a cohort study in Madang, Papua New Guinea, were followed to delivery.
Results
Of 470 women enrolled, delivery data were available for 328 (69.7%). By microscopy, 34.4% (113/328) of women had malaria parasitaemia at enrolment and 12.5% (41/328) at delivery; at each time point, PCR detected sub-microscopic parasitaemia in substantially more. Most infections were with Plasmodium falciparum; the remainder being predominantly P. vivax. Anaemia and smoking were associated with lower birth weight, and LBW (16.7%; 51/305) and PTD (21.8%; 63/290) were common. Histopathologically diagnosed chronic placental malaria was associated with LBW (adjusted odds ratio [aOR] 3.3; p=0.048) and PTD (aOR 4.2; p=0.01). Lack of maternal education predisposed to PTD. Sub-microscopic parasitaemia at delivery appeared to increase the risk of LBW. Of the genetic polymorphisms, Southeast Asian ovalocytosis, α+-thalassaemia and complement receptor 1 (CR1) deficiency, a CR1 heterozygous genotype was associated with decreased risk of anaemia and substantial but non-significant effects were noted in other comparisons.
Conclusions
In coastal Papua New Guinea, malaria and anaemia are important causes of adverse pregnancy outcomes.
Keywords: Genetics, Low birth weight, Plasmodium, Pregnancy, Preterm delivery
Introduction
Approximately 125 million women living in malaria-endemic areas become pregnant each year.1 Pregnant women are more susceptible to malaria than non-pregnant adults partly due to the emergence of pregnancy-specific antigenic variants of Plasmodium falciparum that accumulate in the placenta.2 Further, primigravid women are more susceptible than multigravid women, who typically develop pregnancy-specific immunity, including antibodies to pregnancy-specific parasite variants that confer protection against placental malaria in subsequent pregnancies.3 The consequences of malaria during pregnancy include maternal deaths due to severe anaemia, and low birth weight babies (LBW; <2500 grams) who have increased perinatal and infant mortality.4 LBW results from preterm birth (<37 weeks' gestation) or intrauterine growth retardation.
Coastal Papua New Guinea (PNG) experiences some of the highest levels of malaria transmission outside of Africa.5 Malaria in pregnancy due to P. falciparum and P. vivax is common,6 and its greatest impact on birth weight is in primigravid women.6–8P. vivax placental infection has been reported,9 but P. falciparum is a more important cause of poor pregnancy outcomes.1 Prospective studies of malaria in pregnancy from the area are rare; the most recent was performed in the 1980s, when malaria prevalence was significantly higher than at present.7
Several red cell genetic abnormalities are common in coastal PNG, including α+-thalassaemia, Southeast Asian ovalocytosis (SAO) and complement receptor 1 (CR1) deficiency, and each has been associated with protection against severe malaria,10–12 although they do not appear to protect against uncomplicated malaria, parasitaemia or parasite density.13,14 Studies of SAO and α+-thalassaemia in pregnant PNG women did not show clear evidence of protection against placental malaria, but suggested a decrease in intensity of placental infection in SAO heterozygotes, particularly in first-time mothers.8,15,16
To determine the prevalence, risk factors and consequences of malaria during pregnancy in Madang province, PNG, we performed a longitudinal study of women followed from first antenatal visit to delivery. The associations between malaria infections diagnosed by microscopy, PCR and placental histology and pregnancy outcomes were examined, and potential associations between locally prevalent red blood cell genetic polymorphisms and pregnancy outcomes were investigated.
Materials and methods
Study site
The study took place between September 2005 and October 2007 at the Antenatal Clinic, Alexishafen Health Centre, Madang province, PNG, where malaria transmission is high and perennial.
Study population
Pregnant women >16 years of age were enrolled at their first antenatal visit. Eligibility criteria included evidence of fetal movement, no history of multiple births or delivery complications and the intention to deliver at the Health Centre. Eligible women were read a statement describing the study and gave written informed consent. As staffing was limited, a convenience sample was enrolled, and data were not available on women not enrolled. Women with haemoglobin (Hb) <5 g/dl were excluded and referred for appropriate care.
Study design
Ethics approval was granted by the PNG Medical Research Advisory Council, the Melbourne Health Human Research Ethics Committee and Alfred Health Human Research Ethics Committee. Study participation occurred in parallel with clinic attendance at first antenatal visit, 30–34 weeks' gestation, delivery and 6–8 weeks post-delivery. Due to missing data, only enrolment and delivery visits are analysed here. At enrolment, women received chloroquine (9 or 12 tablets, 150 mg base) and (when available) sulphadoxine pyrimethamine (500/25 mg, three tablets), followed by weekly chloroquine prophylaxis (two 150 mg tablets weekly), and ferrous sulphate 270 mg and folic acid 0.3 mg daily, according to local policy. Prophylaxis was not monitored.
Sample and data collection
At each visit, clinical, malaria exposure (bed net use and residence) and demographic data were collected onto prepared case report forms. Participants were asked about history of fever, headache and chills within the previous 7 days. Gestational age was estimated from fundal height measurements. Weight was measured using sliding weight scales, and mid-upper arm circumference (MUAC) was measured using cloth tape measures. Temperature was recorded using a digital thermometer. At delivery, infants were weighed using SECA baby scales, and gestational age was estimated from Ballard scores.
At enrolment and delivery, 5 ml of venous blood was collected. At delivery, intervillous placental blood was collected from an incision in the maternal side of the placenta, and a 1 cm square full thickness placental biopsy was collected into 10% neutral buffered formalin.
At all time points, thick and thin blood films were prepared for microscopy to quantify parasitemia and blood was collected for molecular diagnosis of Plasmodium spp. infection (including from placental blood). Hb concentration was measured using a HemoCue haemoglobinometer (Hemocue, Ängelholm, Sweden). Residual blood was separated, with plasma and cellular fractions frozen separately.
Laboratory procedures
Blood films were stained with Giemsa and examined by two independent microscopists. Discordant results were referred for a third reading to adjudicate species and density of infection. DNA was extracted from erythrocyte pellets of 200 μl maternal blood (QIAmp DNA Blood Minikit, Qiagen, Valencia, CA, USA). Molecular diagnosis of P. falciparum, P. vivax and P. ovale (P. malariae was not included in this assay) used a polymerase chain reaction/ligase detection reaction-fluorescent microsphere assay (LDR-FMA). Extracted DNA (2.5 μl) was used as a template for amplification of the multicopy Pf 18S small subunit ribosomal RNA gene.17 Samples were considered positive only if both species-specific primers generated a positive signal. Participants were typed for red blood cell polymorphisms: SAO, CR1 and α+-thalassaemia, using published methods.13 Placental biopsies were embedded in paraffin and processed routinely, and Giemsa-stained sections were examined by light microscopy to identify changes consistent with acute, chronic or past infection.18
Exposures and outcomes
We determined predictors of maternal Hb concentration and peripheral or placental parasitaemia (by light microscopy) at delivery, and of birth weight, LBW and preterm birth. Exposures of interest included other malariometric variables e.g., peripheral parasitaemia at enrolment (by light microscopy or PCR); Hb at enrolment; participant characteristics including age, gravidity, parity, smoking, MUAC, education, residence and bed net use; and host genetic polymorphisms (SAO, CR1 and α+-thalassaemia). Smoking and MUAC were included as exposures for infant, but not maternal, outcomes. Malariometric exposures and outcomes include all species types, however the majority (>85%) of infections detected by microscopy were P. falciparum.
Statistical analysis
Logistic and linear regression analyses were performed to identify risk factors for adverse maternal and infant outcomes at delivery. Variables were compared using Pearson's χ2 test for categorical variables and Student's t test for numerical variables. Changes in malariometrics within an individual woman between enrolment and delivery were quantified using McNemar's χ2 test and Student's paired t test. Univariable associations were determined between malariometric, demographic and host genetic variables and Hb, birth weight, LBW and preterm delivery. Predictors of infant and maternal outcomes were identified from adjusted regression models. Adjusted models included all exposure variables, with some exceptions due to collinearity. Age and parity showed collinearity with gravidity, which was retained because of well-established differences in risk of malaria during pregnancy between primigravidae and multigravidae. Due to limited statistical power, both the magnitude of associations and the p-value were considered in the interpretation of potential risk factors.
Results
Study cohort
Table 1 shows the baseline characteristics of the 470 pregnant women recruited, and of the 328 women (69.7%) with delivery information. Of 377 women (80.2%; 377/470) who returned at delivery, we excluded 6 women with multiple births and 43 who were not seen ≤3 days after delivery. There were no systematic differences between included women (n=328) and those who were excluded or lost to follow-up (n=142; all p>0.08), except that included women were slightly better educated (p=0.007; data not shown). Among women followed to delivery, the median age was 24 years (range 16–49), 127/328 (38.7%) were in their first pregnancy (primigravidae) and median (range) estimated gestational age was 25 (7–36) weeks. Seventy-five percent of women (234/321) used bed nets the night before enrolment.
Table 1.
Baseline characteristics of women enrolled in the cohort and of the sub-set who returned for a delivery visit
| Variable |
Enrolled (n=470)a,b | Delivered (n=328)a,b | |
|---|---|---|---|
| Demographic | |||
| Age (years) | 24 (21–28), 16–49 | 24 (21–28), 16–49 | |
| Gestational age (weeks) | 25.2 [4.28], 7–38 | 25.2 [4.17], 7–36 | |
| Gravidity | Primigravidae | 179 (38) | 127 (39) |
| Multigravidae | 292 (62) | 201 (61) | |
| MUAC (cm) | 22.5 [1.9], 12–30 | 22.4 [1.8], 12–30 | |
| Weight (kg) | 54.4 [6.1], 35–74 | 54.4 [5.9], 35–74 | |
| Mother's education | None | 60 (13) | 32 (10) |
| Primary (1–7) | 214 (46) | 149 (46) | |
| Secondary (8+) | 190 (41) | 142 (44) | |
| Mother's occupation | None | 310 (66) | 207 (63) |
| Self-employed | 114 (24) | 83 (25) | |
| Earns wages | 45 (10) | 36 (11) | |
| Smokes | 88 (19) | 63 (19) | |
| Malaria exposure | |||
| Owns bed net | 382 (81) | 268 (83) | |
| Used bed net last night | 335 (71) | 234 (75) | |
| Bed net last treated | Never | 328 (70) | 224 (75) |
| <6 months | 27 (6) | 19 (6) | |
| >6 months | 66 (14) | 52 (17) | |
| Residence | Town | 16 (3) | 9 (3) |
| Village | 452 (96) | 316 (97) | |
| Genetic | |||
| SAO | Normal | 410 (87) | 281 (86) |
| SAO | 59 (13) | 47 (14) | |
| CR1 | HHc | 39 (8) | 31 (9) |
| HLc | 197 (42) | 136 (41) | |
| LLc | 233 (49) | 161 (49) | |
| α-thalassaemia | Wildtype | 87 (18) | 63 (19) |
| Heterozygous | 185 (39) | 128 (39) | |
| Homozygous | 197 (42) | 137 (42) | |
| Clinical | |||
| Clinical historyd | 122 (26) | 82 (25) | |
| Palpable spleen | 86 (18) | 63 (20) | |
| Haemoglobin | |||
| Haemoglobin (g/dl) | 8.5 [1.4], 5.0–12.8 | 8.5 [1.4], 5.3–12.8 | |
| Severe anaemia (<8 g/dl) | 163 (35) | 112 (34) | |
Data are mean [SD], range; or median (25th–75th percentile), range; or n (%).
CR1: complement receptor-1; H: high; HH: high-high; HL: high-low; L: low; LL: low-low; MUAC: mid-upper arm circumference; SAO: Southeast Asian ovalocytosis.
a Enrolment: all women enrolled into study. Delivery: all women enrolled and included in final analyses. All p-values for the difference between study participants and excluded women >0.08, except for education (p=0.007).
b When totals do not add to sample size or percentages do not total 100%, missing values can be assumed.
c Abbreviations correspond to levels of expression on the cell surface.
d Clinical history of fever, headache or chills in the 7 days prior to enrolment visit.
At enrolment, 314/328 women (95.7%) were anaemic (Hb <11 g/dl) and 112/328 (34.1%) had severe anaemia (Hb <8 g/dl). Red cell polymorphisms associated with malaria in PNG were common: of 328 women, 47 (14.3%) had the SAO polymorphism; 297 (90.5%) had genotypes associated with intermediate or low CR1 expression on erythrocytes and 265 (80.7%) were heterozygous or homozygous for α+-thalassaemia (Table 1).
Prevalence of malaria infection
At enrolment, 113/328 women (34.4%) had peripheral parasitaemia by microscopy, which decreased to 41/298 (12.5%) at delivery (p<0.001). In peripheral blood, P. falciparum was detected in 105/113 (92.9%) parasitaemic women at enrolment, and 35/41 (85%) at delivery. Additionally, 39/44 (89%) positive placental smears showed infection with P. falciparum (Table 2). By placental histology, 154/242 women (63.6%) had past or present malaria infection (Table 2).
Table 2.
Malaria prevalence by microscopy and by PCR at enrolment and delivery
| Variable | Enrolment (n=328) | Delivery (n=298) | Change (95% CI)a | p-valueb |
|---|---|---|---|---|
| Mean [SD], range; or n (%) | Mean [SD], range; or n (%) | |||
| Peripheral blood microscopy | ||||
| All species | 113 (34) | 41 (14) | −22 (−29, −15) | <0.001 |
| P. falciparum | 105 (32) | 35 (12) | −21 (−28, −15) | <0.001 |
| P. vivax | 12 (4) | 7 (2) | −1.7 (−4.7, 1.4) | 0.22 |
| P. ovale | 0 (0) | 1 (0.3) | 0.3 (−0.7, 1.3) | 0.32 |
| Placental blood microscopy | ||||
| All species | ND | 44 (18) | NA | NS |
| P. falciparum | ND | 39 (16) | NA | NS |
| P. vivax | ND | 6 (2) | NA | NS |
| P. ovale | ND | 0 (0) | NA | NS |
| Placental histologyc | ||||
| No infection | ND | 88 (36) | NA | NS |
| Acute infection | ND | 93 (38) | NA | NS |
| Chronic infection | ND | 41 (17) | NA | NS |
| Past infection | ND | 20 (8) | NA | NS |
| Peripheral blood PCR | ||||
| All species | 217 (66) | 110 (47) | −18 (−28, 8.9) | <0.001 |
| P. falciparum | 160 (49) | 86 (37) | −12 (−21, 2.7) | 0.01 |
| P. vivax | 64 (20) | 28 (12) | −8.1 (−15, −1.0) | 0.02 |
| P. ovale | 38 (12) | 11 (4) | −6.1 (−11, −0.9) | 0.01 |
| Haemoglobin | ||||
| Haemoglobin (g/dl) | 8.5 [1.4], 5.3–12.8 | 9.2 [1.8], 4.2–14.8 | 0.69 (0.50, 0.89) | <0.001 |
| Severe anaemia (<8 g/dl) | 112 (34) | 68 (23) | −11 (−17, −4.3) | 0.001 |
Placental histology missing for 56 participants, PCR results at delivery missing for 65 participants.
NA: not applicable; ND: not done; NS: not significant.
a Between enrolment and delivery; discrepancies in percentage differences due to rounding.
b p-value derived from paired t tests for numeric variables, or McNemar's χ2 test for binary variables.
c Placental histology of placental biopsies by light microscopy: no infection; acute infection (parasites without pigments); chronic infection (parasites with monocyte and/or fibrin pigments); past infection (pigment without parasites).
More women were infected as determined by PCR than by microscopy (Table 2). Half the infections detected at enrolment, and two-thirds of infections at delivery, were sub-microscopic. Substantially more infections with P. vivax or P. ovale were detected by PCR than by microscopy, although P. falciparum still predominated.
Risk factors for malaria infection at delivery
We examined risk factors for blood slide parasitaemia at delivery (Table 3). On multivariate analysis, peripheral parasitaemia was more common in multigravid than primigravid women (OR 2.12 [0.95–4.73]; 0.07) and peripheral parasitaemia was less common in women who used a bed net (Table 3). No other factors, including host genetic polymorphisms, were significantly associated with risk of peripheral or placental parasitaemia (Table 3).
Table 3.
Risk factors for peripheral and placental parasitaemia at delivery as determined by light microscopy
| Peripheral parasitaemia (n=298) |
Placental parasitaemia (n=247) |
|||||||
|---|---|---|---|---|---|---|---|---|
| Variable | Normal (n=257) | Parasitaemic (n=41) | Univariable (OR [95% CI]; p) | Multivariable (OR [95% CI]; p) | Normal (n=203) | Parasitaemic (n=44) | Univariable | Multivariable |
| Age (years) | 247 (86) | 41 (14) | 0.91 [0.85, 0.98]; 0.01 | NDa | 196 (82) | 44 (18) | 1.02 [0.96, 1.08]; 0.51 | NDa |
| Gravidity | ||||||||
| Primigravidae | 103 (40) | 12 (29) | NA | NA | 80 (39) | 16 (36) | NA | NA |
| Multigravidae | 154 (60) | 29 (71) | 1.62 [0.79–3.31]; 0.19 | 2.12 [0.95–4.73]; 0.07 | 123 (61) | 28 (64) | 1.14 [0.58–2.24]; 0.71 | 1.32 [0.61–2.84]; 0.47 |
| Education | ||||||||
| None | 24 (9) | 2 (5) | NA | NA | 19 (9) | 3 (7) | NA | NA |
| Primary | 120 (47) | 15 (37) | 1.5 [0.32–6.99]; 0.61 | 1.49 [0.29–7.71]; 0.63 | 95 (48) | 18 (41) | 1.2 [0.32–4.48]; 0.79 | 0.95 [0.23–3.93]; 0.95 |
| Secondary | 109 (43) | 23 (57) | 2.53 [0.56–11.47]; 0.23 | 2.95 [0.58–14.99]; 0.19 | 85 (43) | 23 (52) | 1.71 [0.47–6.30]; 0.42 | 1.62 [0.39–6.70]; 0.51 |
| Used bed net | ||||||||
| No | 56 (23) | 14 (36) | NA | NA | 43 (22) | 13 (30) | NA | NA |
| Yes | 188 (77) | 25 (64) | 0.53 [0.26–1.10]; 0.08 | 0.48 [0.22–1.04]; 0.06 | 148 (78) | 30 (70) | 0.67 [0.32–1.40]; 0.29 | 0.71 [0.33–1.55]; 0.40 |
| Symptoms | ||||||||
| No | 190 (75) | 28 (68) | NA | NA | 149 (74) | 32 (73) | NA | NA |
| Yes | 63 (25) | 13 (32) | 1.4 [0.68–2.87]; 0.36 | 1.42 [0.64–3.19]; 0.39 | 53 (26) | 12 (27) | 1.05 [0.51–2.20]; 0.89 | 1.09 [0.49–2.44]; 0.83 |
| Enrolment Hb | 257 (86) | 41 (14) | 0.95 [0.74, 1.21]; 0.70 | 1.03 [0.77–1.36]; 0.86 | 203 (82) | 44 (18) | 1.06 [0.83, 1.34]; 0.65 | 1.05 [0.80–1.37]; 0.73 |
| Parasitaemia at enrolment | ||||||||
| No | 166 (65) | 25 (61) | NA | NA | 128 (63) | 32 (73) | NA | NA |
| Yes | 91 (35) | 16 (39) | 1.17 [0.59–2.3]; 0.65 | 1.36 [0.62–2.97]; 0.44 | 75 (37) | 12 (27) | 0.64 [0.31–1.39]; 0.23 | 0.54 [0.24–1.22]; 0.14 |
| SAO | ||||||||
| No | 218 (85) | 36 (88) | NA | NA | 177 (87) | 36 (82) | NA | NA |
| Yes | 39 (15) | 5 (12) | 0.78 [0.29–2.10]; 0.62 | 0.52 [0.16–1.65]; 0.27 | 26 (13) | 8 (18) | 1.51 [0.63–3.61]; 0.35 | 1.27 [0.48–3.33]; 0.62 |
| CR1 | ||||||||
| HH | 26 (10) | 4 (10) | NA | NA | 23 (11) | 5 (11) | NA | NA |
| HL | 107 (42) | 16 (39) | 0.97 [0.30–3.15]; 0.96 | 0.85 [0.25–2.9]; 0.80 | 85 (42) | 12 (27) | 0.65 [0.21–2.03]; 0.46 | 0.68 [0.21–2.20]; 0.51 |
| LL | 124 (48) | 21 (51) | 1.10 [0.35–3.48]; 0.87 | 1.01 [0.30–3.35]; 0.99 | 95 (47) | 27 (61) | 1.31 [0.45–3.76]; 0.62 | 1.36 [0.45–4.09]; 0.58 |
| α-thalassaemia | ||||||||
| Wildtype | 50 (19) | 8 (19) | NA | NA | 36 (18) | 7 (16) | NA | NA |
| Heterozygote | 101 (39) | 18 (44) | 1.11 [0.45–2.74]; 0.81 | 1.27 [0.45–3.46]; 0.64 | 76 (37) | 19 (43) | 1.30 [0.60–3.33]; 0.60 | 1.46 [0.53–4.01]; 0.46 |
| Homozygote | 106 (41) | 15 (37) | 0.88 [0.35–2.22]; 0.79 | 1.04 [0.37–2.89]; 0.94 | 91 (45) | 18 (41) | 1.02 [0.39–2.60]; 0.97 | 0.94 [0.34–2.62]; 0.91 |
Data are unadjusted and adjusted odds ratios [95% CI].
CR1: complement receptor-1; Hb: haemoglobin; HH: high-high cell surface expression; HL: high-low; LL: low-low; ND: not done; NA: not applicable (reference group); SAO: Southeast Asian ovalocytosis.
a Variables excluded due to collinearity.
Risk factors for adverse infant outcomes–birth weight, low birth weight and preterm birth
Mean (SD) birth weight was 2870 (480) grams and the median (interquartile range) gestational age was 38 (37–40) weeks. The prevalence of LBW and preterm birth were high (51/305; 16.7% and 63/290; 21.7%, respectively). Malaria parasitaemia diagnosed by light microscopy was not associated with birth weight or LBW, but features of chronic placental infection on placental histology were associated with LBW and with preterm birth. In adjusted analyses, women with chronic placental malaria had significantly increased odds of LBW and preterm birth (3.3- and 4.2-fold, respectively) compared to uninfected women. Higher Hb at enrolment was associated with higher birth weight (61 g per 1 g/dl increase in Hb; p=0.005), but Hb concentration was not significantly associated with LBW or preterm delivery.
Other factors that were associated with birth weight in unadjusted, but not adjusted analyses included positive associations between MUAC and birth weight (p=0.022), and negative associations between birth weight and history of recent febrile symptoms (p=0.001) or parasitaemia at enrolment (p=0.034) (Table 4). In adjusted analyses, factors that were independently associated with birth weight included gravidity (p<0.001), smoking (p=0.005) and SAO genotype. Multigravid women had babies that were on average 325 g heavier than primigravidae (p<0.001), while smokers' babies were 197 g lighter than non-smokers' (Table 4).
Table 4.
Factors associated with birth weight
| Variable | Unadjusted | Adjusteda |
|---|---|---|
| Age (years) | 8.4 [−1.5, 18]; 0.095 | NDb |
| Gravidity | ||
| Primigravidae | NA | NA |
| Multigravidae | 304 [198, 410]; <0.001 | 325 [211, 439]; <0.001 |
| MUAC (cm) | 34 [4.9, 64]; 0.022 | 18 [−15, 50]; 0.288 |
| Smokes | ||
| No | NA | NA |
| Yes | −140 [−274, −6.4]; 0.04 | −197 [−333, −61]; 0.005 |
| Education | ||
| None | NA | NA |
| Primary | 54 [−146, 254]; 0.59 | 100 [−104, 305]; 0.34 |
| Secondary | 2.5 [−198, 203]; 0.98 | 79 [−127, 286]; 0.45 |
| Used bed net | ||
| No | NA | NA |
| Yes | −72 [−201, 56]; 0.27 | −40 [−165, 84]; 0.52 |
| Clinical history | ||
| No | NA | NA |
| Yes | −221 [−345, −96]; 0.001 | −104 [−230, 22]; 0.10 |
| Enrolment Hb (g/dl) | 61 [22, 100]; 0.002 | 61 [19, 102]; 0.005 |
| Anaemia | ||
| No | NA | NDb |
| Yes (Hb <8 g/dl) | −101 [−215, 13]; 0.083 | |
| Parasitaemia (LM) at enrolment | ||
| No | NA | NA |
| Yes | −121 [−234, −9.1]; 0.034 | −30 [−149, 89]; 0.62 |
| SAO | ||
| Normal | NA | NA |
| SAO | −132 [−285, 22]; 0.093 | −129 [−287, 29]; 0.11 |
| CR1 | ||
| HH | NA | NA |
| HL | −179 [−370, 12]; 0.067 | −168 [−350, 15]; 0.071 |
| LL | −68 [−256, 120]; 0.48 | −88 [−265, 90]; 0.33 |
| α-thalassaemia | ||
| Wildtype | NA | NA |
| Heterozygote | −61 [−212, 91]; 0.43 | −12 [−159, 135]; 0.88 |
| Homozygote | 22 [−128, 171]; 0.78 | 47 [−100, 194]; 0.53 |
Data are mean differences in birth weight (grams) [95% CI]; p-value.
CR1: complement receptor-1; Hb: haemoglobin; HH: high-high cell surface expression; HL: high-low; LL: low-low; LM: light microscopy; MUAC: mid-upper arm circumference; ND: not done; NA: not applicable (reference group); SAO: Southeast Asian ovalocytosis.
a Adjusted model: n=271.
b Variables excluded due to collinearity. Missing data for age (11), smoking (1), MUAC (10), education (4), residence (3), bed net use (15), clinical history (4). Multivariable model includes all risk factors except age.
Other participant characteristics associated with LBW and preterm birth were examined (Table 5). On multivariate analyses, multigravidae had significantly reduced odds of LBW (OR 0.28; p<0.001) and preterm birth (OR 0.38; p=0.0087) compared to primigravid women. Preterm birth was significantly less common in women who had had either primary (OR 0.22; p=0.0072) or secondary education (OR 0.13; p<0.001). Large magnitudes of effect with adverse infant outcomes were observed for CR1 and α+-thalassaemia genotypes but these were not statistically significant; high-low (HL) and low-low (LL) CR1 genotypes reduced the odds of preterm birth by almost 50%, but had no effect on the odds of LBW compared to the normal high-high (HH) genotype. Heterozygous and homozygous α+-thalassaemia genotypes reduced odds of both preterm and LBW by up to 45%, compared to the wildtype (p>0.15). Women with SAO had increased odds of LBW (2.33-fold; p=0.07) compared to women of normal genotype (Table 5).
Table 5.
Risk factors for low birth weight and preterm delivery
| Risk factors for low birth weight (LBW;
<2500 grams), n=305 |
Risk factors for preterm birth (<37
weeks gestation), n=290 |
|||||||
|---|---|---|---|---|---|---|---|---|
| Variable | Normal (n=254) | LBW (n=51) | Univariable | Multivariable | Normal (n=227) | Preterm (n=63) | Univariable | Multivariable |
| Age (years) | 245 (83) | 29 (17) | 0.96 [0.91, 1.02]; 0.23 | NDa | 218 (78) | 61 (22) | 0.97 [0.92, 1.03]; 0.31 | NDa |
| Gravidity | ||||||||
| Primigravidae | 87 (34) | 30 (59) | NA | NA | 79 (35) | 32 (51) | NA | NA |
| Multigravidae | 167 (66) | 21 (41) | 0.36 [0.20–0.67]; 0.0013 | 0.28 [0.13–0.57]; 0.001 | 148 (65) | 31 (49) | 0.52 [0.29–0.91]; 0.02 | 0.38 [0.19–0.79]; 0.0087 |
| MUAC (cm) | 246 (83) | 50 (17) | 0.88 [0.74, 1.0]; 0.12 | 0.91 [0.74–1.1]; 0.36 | 220 (78) | 61 (22) | 0.93 [0.80, 1.1]; 0.38 | 1.0 [0.83–1.2]; 0.84 |
| Smokes | ||||||||
| No | 202 (80) | 40 (78) | NA | NA | 183 (81) | 48 (76) | NA | NA |
| Yes | 51 (20) | 11 (22) | 1.1 [0.52–2.27]; 0.82 | 1.4 [0.62–3.35]; 0.40 | 43 (19) | 15 (24) | 1.3 [0.68–2.6]; 0.40 | 1.1 [0.50–2.62]; 0.75 |
| Education | ||||||||
| None | 22 (9) | 5 (10) | NA | NA | 15 (7) | 11 (18) | NA | NA |
| Primary | 119 (48) | 22 (43) | 0.81 [0.29–2.4]; 0.71 | 0.48 [0.13–1.7]; 0.26 | 102 (45) | 31 (50) | 0.41 [0.17–0.99]; 0.05 | 0.22 [0.07–0.66]; 0.0072 |
| Secondary | 109 (44) | 24 (47) | 0.97 [0.33–2.8]; 0.95 | 0.57 [0.16–2.0]; 0.38 | 107 (48) | 20 (32) | 0.25 [0.10–0.63]; 0.0033 | 0.13 [0.04–0.42]; <0.001 |
| Used bed net | ||||||||
| No | 62 (26) | 12 (24) | NA | NA | 52 (24) | 15 (25) | NA | NA |
| Yes | 179 (74) | 37 (75) | 1.1 [0.52–2.2); 0.86 | 1.2 [0.53–2.7]; 0.66 | 162 (76) | 46 (75) | 0.98 [0.50–1.9]; 0.96 | 0.78 [0.36–1.7]; 0.53 |
| Clinical history | ||||||||
| No | 191 (76) | NA | NA | NA | 175 (77) | 40 (66) | NA | – |
| Yes | 59 (24) | 16 (31) | 1.5 [0.76–2.9]; 0.24 | 0.90 [0.42–1.9]; 0.79 | 51 (23) | 21 (34) | 1.8 [0.98–3.3]; 0.06 | 1.5 [0.70–3.1]; 0.30 |
| Enrolment Hb (g/dl) | 254 (83) | 51 (17) | 0.8 [0.68, 1.1]; 0.16 | 0.83 [0.63–1.08]; 0.16 | 227 (78) | 63 (22) | 0.86 [0.70, 1.1]; 0.16 | 0.85 [0.65–1.1]; 0.22 |
| Anaemia | ||||||||
| No | 168 (66) | 34 (67) | NA | NDa | ND | ND | ND | ND |
| Yes (Hb <8) g/dl) | 86 (34) | 17 (33) | 0.98 [0.52, 1.8] 0.94 | ND | ND | ND | ND | |
| Enrolment parasitaemia | ||||||||
| No | 207 (86) | 44 (90) | NA | NA | 154 (68) | 36 (57) | NA | NA |
| Yes | 33 (14) | 5 (10) | 1.3 [0.71–2.4]; 0.37 | 1.0 [0.50–2.2]; 0.91 | 73 (32) | 27 (43) | 1.6 [0.89–2.8]; 0.12 | 1.3 [0.65–2.7]; 0.44 |
| SAO | ||||||||
| Normal | 221 (87) | 40 (78) | NA | NA | 194 (85) | 55 (87) | NA | NA |
| SAO | 33 (13) | 11 (22) | 1.8 [0.86–3.94]; 0.12 | 2.33 [0.94–5.75]; 0.07 | 33 (15) | 8 (13) | 0.85 [0.37–1.96]; 0.71 | 1.10 [0.40–2.99]; 0.85 |
| CR1 | ||||||||
| HH | 25 (10) | 5 (10) | NA | NA | 20 (9) | 9 (14) | NA | NA |
| HL | 103 (41) | 21 (41) | 1.02 [0.35–2.97]; 0.97 | 1.08 [0.35–3.35]; 0.89 | 91 (40) | 26 (41) | 0.63 [0.26–1.56]; 0.32 | 0.56 [0.21–1.55]; 0.27 |
| LL | 126 (50) | 25 (49) | 0.99 [0.35–2.84]; 0.99 | 0.89 [0.29–2.72]; 0.84 | 116 (51) | 28 (44) | 0.54 [0.22–1.30]; 0.17 | 0.54 [0.20–1.43]; 0.21 |
| α-thalassaemia | ||||||||
| Wildtype | 46 (18) | 12 (23) | NA | NA | 43 (19) | 11 (17) | NA | NA |
| Heterozygote | 97 (38) | 21 (41) | 0.83 [0.38–1.83]; 0.64 | 0.66 [0.27–1.58]; 0.35 | 89 (39) | 26 (41) | 1.14 [0.52–2.52]; 0.74 | 0.93 [0.38–2.23]; 0.86 |
| Homozygote | 111 (44) | 18 (35) | 0.62 [0.28–1.40]; 0.25 | 0.55 [0.22–1.34]; 0.29 | 95 (42) | 26 (41) | 1.07 [0.48–2.36]; 0.87 | 0.63 [0.25–1.56]; 0.32 |
| Placental histology | Normal (n=200) | LBW (n=37) | Univariable | Multivariable | Normal (n=178) | Preterm (n=49) | Univariable | Multivariable |
| No infection | 75 (37) | 10 (27) | NA | NA | 74 (42) | 10 (20) | NA | NA |
| Acute | 82 (41) | 9 (24) | 0.82 [0.32, 2.1]; 0.69 | 0.66 [0.22, 2.0]; 0.47 | 65 (36) | 20 (41) | 2.3 [0.99, 5.2]; 0.052 | 2.08 [0.75, 5.8]; 0.16 |
| Chronic | 28 (14) | 13 (35) | 3.5 [1.4, 8.8]; 0.01 | 3.3 [1.0, 10.6]; 0.048 | 24 (13) | 15 (31) | 4.6 [1.8, 11.6]; 0.001 | 4.2 [1.3, 13.4]; 0.01 |
| Past infection | 15 (7) | 5 (13) | 2.5 [0.7, 8.4]; 0.14 | 1.5 [0.37, 6.1]; 0.58 | 15 (8) | 4 (8) | 2.0 [0.55, 7.1]; 0.30 | 1.34 [0.30, 6.1]; 0.70 |
Data are unadjusted and adjusted odds ratios [95% CI].
CR1: complement receptor-1; Hb: haemoglobin; HH: high-high cell surface expression; HL: high-low; LBW: low birth weight; LL: low-low; LM: light microscopy; MUAC: mid-upper arm circumference; NA: not applicable (reference group); ND: not done; SAO: Southeast Asian ovalocytosis.
a Variables excluded due to collinearity.
Risk factors for maternal haemoglobin and anaemia at delivery
In univariate analyses, both homozygous and heterozygous α+-thalassaemia genotypes were associated with decreases in mean Hb at delivery (0.79 g/dl and 0.74 g/dl) compared to women with normal genotype (Table 6). In adjusted analyses, the CR1 HL genotype was associated with protection against anaemia (OR 0.36; p=0.048), but not with Hb levels (p=0.50). Smoking was associated with a borderline increase in risk of anaemia in adjusted analyses (OR 2.0; p=0.075), while parasitaemia at enrolment was associated with a borderline increase in delivery Hb (0.42 g/dl; CI −0.04 to 0.88; p=0.073). There were non-significant but substantial associations between SAO and protection from anaemia (p=0.057), and between CR1 LL homozygosity and protection from anaemia (p=0.081).
Table 6.
Risk factors for severe anaemia and associations with maternal haemoglobin at delivery
| Risk factors for severe anaemia at
delivery |
Associations with maternal haemoglobin at
delivery |
|||||
|---|---|---|---|---|---|---|
| Variable | Not anaemic (n=230) | Anaemic (n=68) | Unadjusted | Adjusted | Unadjusted | Adjusted |
| Age (years) | 221 (100) | 67 (100) | 0.97 (0.92, 1.0); 0.27 | NDa | 0.01 [−0.03, 0.05]; 0.65 | NDa |
| Gravidity | ||||||
| Primigravidae | 89 (39) | 26 (38) | NA | NA | NA | NA |
| Multigravidae | 141 (61) | 42 (62) | 1.0 (0.58, 1.8); 0.94 | 0.82 [0.41, 1.6]; 0.57 | −0.17 [−0.58, 0.25]; 0.43 | 0.08 [−0.36, 0.52]; 0.72 |
| MUAC (cm) | 223 (100) | 65 (100) | 0.87 (0.74, 1.0); 0.077 | 0.91 [0.75, 1.1]; 0.36 | 0.06 [−0.05, 0.18]; 0.27 | −0.05 [−0.17, 0.08]; 0.46 |
| Smokes | ||||||
| No | 189 (82) | 51 (75) | NA | NA | NA | NA |
| Yes | 40 (18) | 17 (25) | 1.6 (0.83, 3.0); 0.17 | 2.0 [0.93, 4.4]; 0.075 | −0.43 [−0.95, 0.08]; 0.1 | −0.38 [−0.91, 0.15]; 0.162 |
| Education | ||||||
| None | 21 (9) | 7 (10) | NA | NA | NA | NA |
| Primary | 107 (48) | 28 (41) | 0.79 (0.30, 2.0); 0.62 | 1.2 [0.38, 3.7]; 0.76 | 0.33 [−0.40, 1.06]; 0.37 | −0.04 [−0.78, 0.71]; 0.93 |
| Secondary | 97 (43) | 33 (49) | 1.0 (0.40, 2.6); 0.97 | 1.5 [0.46, 4.7]; 0.52 | 0.18 [−0.56, 0.91]; 0.64 | −0.23 [−0.99, 0.53]; 0.56 |
| Used bed net | ||||||
| No | 53 (24) | 17 (27) | NA | NA | NA | NA |
| Yes | 167 (76) | 47 (73) | 0.9 (0.46, 1.7); 0.69 | 0.94 [0.43, 2.0]; 0.87 | −0.12 [−0.60, 0.36]; 0.63 | −0.09 [−0.57, 0.39]; 0.70 |
| Clinical history | ||||||
| No | 169 (74) | 50 (75) | NA | NA | NA | NA |
| Yes | 58 (26) | 17 (25) | 0.99 (0.53, 1.8); 0.98 | 0.65 [0.30, 1.4]; 0.27 | −0.09 [−0.56, 0.38]; 0.71 | 0.20 [−0.28, 0.68]; 0.40 |
| Enrolment Hb (g/dl) | 230 (100) | 68 (100) | 0.54 (0.42, 0.69); <0.001 | 0.55 [0.42, 0.73]; <0.001 | 0.54 [0.41, 0.68]; <0.001 | 0.56 [0.39, 0.72]; <0.001 |
| Enrolment anaemia | ||||||
| No | 169 (73) | 29 (43) | NA | NDa | NA | NDa |
| Yes (Hb <8 g/dl) | 61 (27) | 39 (57) | 3.7 (2.1, 6.5); <0.001 | −1.2 [−1.6, −0.77]; <0.001 | ||
| Enrolment parasitaemia | ||||||
| No | 148 (64) | 44 (65) | NA | NA | NA | NA |
| Yes | 82 (36) | 24 (35) | 0.98 (0.56, 1.7); 0.96 | 0.59 [0.29, 1.2]; 0.15 | −0.05 [−0.47, 0.37]; 0.82 | 0.42 [−0.04, 0.88]; 0.073 |
| SAO | ||||||
| Normal | 193 (84) | 62 (91) | NA | NA | NA | NA |
| SAO | 37 (16) | 6 (9) | 0.5 (0.20, 1.2); 0.14 | 0.28 [0.08, 1.0]; 0.057 | 0.24 [−0.33, 0.82]; 0.41 | 0.28 [−0.32, 0.89]; 0.35 |
| CR1 | ||||||
| HH | 20 (9) | 11 (16) | NA | NA | NA | NA |
| HL | 96 (42) | 24 (35) | 0.45 (0.19, 1.1); 0.073 | 0.36 [0.13, 0.99]; 0.048 | 0.34 [−0.36, 1.0]; 0.34 | 0.23 [−0.45, 0.91]; 0.50 |
| LL | 114 (50) | 33 (49) | 0.53 (0.23, 1.2); 0.13 | 0.44 [0.17, 1.1]; 0.081 | 0.49 [−0.20, 1.18]; 0.17 | 0.42 [−0.24, 1.07]; 0.21 |
| α−thalassaemia | ||||||
| Wildtype | 50 (22) | 7 (10) | NA | NA | NA | NA |
| Heterozygote | 86 (37) | 33 (48) | 2.7 (1.1, 6.6); 0.026 | 2.1 [0.80, 5.6]; 0.13 | −0.79 [−1.3, −0.23]; 0.006 | −0.45 [−1.02, 0.11]; 0.11 |
| Homozygote | 94 (41) | 28 (41) | 2.1 (0.87, 5.2); 0.099 | 1.5 [0.56, 4.1]; 0.42 | −0.74 [−1.3, −0.19]; 0.009 | −0.39 [−0.96, 0.17]; 0.17 |
Data are number (%), unadjusted or adjusted odds ratio (95% CI), or adjusted or unadjusted coefficient [95% CI].
CR1: complement receptor-1; Hb: haemoglobin; HH: high-high cell surface expression; HL: high-low; LL: low-low; LM: light microscopy; MUAC: mid-upper arm circumference; NA: not applicable (reference group); ND: not done; SAO: Southeast Asian ovalocytosis.
a Variables excluded due to collinearity. n of adjusted model=263. Missing data for age (10; 1 anaemic), smoking (1; 0 anaemic), mid-upper arm circumference (10; 3 anaemic), education (5; 0 anaemic), residence (3; 0 anaemic), bed net use (14; 4 anaemic), clinical history (4; 1 anaemic). Thirty women missing haemoglobin at delivery.
The contribution of sub-microscopic infections to adverse pregnancy outcomes
Sub-microscopic infections were seen in approximately one-third of women at enrolment and delivery. We compared our study endpoints of anaemia, birth weight and LBW in women with PCR-only infection and uninfected women at enrolment and delivery. Mean birth weight was not significantly different between women with sub-microscopic infection at delivery (2814±473 g) and uninfected women (2878±401 g; p=0.35), but there was a greater proportion of LBW deliveries among women with sub-microscopic infection (20/85; 24%) than in uninfected women (14/110, 12.7%; OR 2.41; CI 0.99–5.89; p=0.054; Table 7).
Table 7.
Association between submicroscopic Plasmodium infection, low birth weight and birth weight
| Variable | Low birth weight (n,%) | Unadjusted | Adjusted | Mean [SD] birth weight | Unadjusted |
|---|---|---|---|---|---|
| Peripheral parasitaemia at delivery (PCR) (n=195) | |||||
| Negative | 14/110; 12.7% | NA | NA | 2878 [401] | NA |
| Positive | 20/85; 23.5% | 2.1 (0.99, 4.5); 0.052 | 2.4 (0.99, 5.9); 0.054 | 2814 [473] | −64 [−188, 59]; 0.31 |
Data are unadjusted or adjusted odds ratio (95% CI), or adjusted or unadjusted coefficient [95% CI]. Adjusted for gravidity, smoking, mid-upper arm circumference, education, residence, bed net use, clinical history, peripheral infection at enrolment (light microscopy), haemoglobin at enrolment, SAO, CR1 and alpha thalassaemia.
CR1: complement receptor-1; NA: not applicable (reference group); SAO: Southeast Asian ovalocytosis.
Discussion
We described clinical characteristics, malaria prevalence and prevalence of common red blood cell genetic polymorphisms in a cohort of pregnant women living in coastal PNG, and performed multivariable linear and logistic regressions to identify risk factors for malaria infection at delivery and for poor pregnancy outcomes including lower maternal Hb, decreased birth weight and increased risk of LBW and preterm delivery. This is the first reported study outside of Africa to show associations between placental infection and adverse birth outcomes.
Malaria is highly endemic in coastal PNG. At enrolment, one-third of women were parasitaemic by microscopy and two-thirds by PCR. Both P. falciparum and P. vivax infections are common in school-age PNG children,19 but P. falciparum was responsible for >85% of microscopic infections and three-quarters of sub-microscopic infections in pregnant women. Previous observations in PNG demonstrate the earlier acquisition of immunity to P. vivax than to P. falciparum due to high force of infection.20 Whether this extends to protection from asymptomatic parasitaemia is unknown. There also appears to be a greater increase in pregnancy-associated susceptibility to P. falciparum than to P. vivax.21
The great majority of women were anaemic (Hb <11 g/dl) and 35% had severe anaemia (defined here as Hb <8 g/dl). Apart from malaria, common causes of anaemia in pregnancy in PNG include iron and/or folate deficiency.7,22 Alpha thalassaemia was common, and may also contribute to anaemia. The increase in Hb levels between enrolment and delivery may reflect combined effects of malaria prophylaxis and iron and folate supplementation.
The only factor associated with a decreased risk of peripheral parasitaemia at delivery was increasing age, which may reflect the age dependent acquisition of immunity. Surprisingly, there was no significant gravidity-dependent decrease in risk of peripheral or placental parasitaemia at delivery,4 which has been attributed largely to the acquisition of antibodies to pregnancy-associated P. falciparum parasite strains.2 Previous studies from the area do show gravidity-dependent susceptibility to malaria,23 and the gravidity-associated acquisition of antibodies to multiple different placental-binding P. falciparum isolates.24
Malaria prophylaxis was not highly effective, with parasitaemia detected in peripheral and/or placental blood at delivery in 24% of women. Participants received curative doses of chloroquine and (usually) sulphadoxine-pyrimethamine and unsupervised weekly chloroquine prophylaxis. Chloroquine resistance in P. falciparum is common in PNG, with chloroquine resistant P. vivax also well described.25 PNG recently introduced intermittent preventive treatment in pregnancy with sulphadoxine pyrimethamine, which should be more effective in controlling P. falciparum in pregnancy.
On placental histology, almost two-thirds of women had evidence of current or previous malaria, predominantly active infections (in which parasites were detected, frequently at low densities). Chronic infection was associated with LBW and preterm delivery. In African women, chronic placental infection (especially massive chronic intervillositis, the accumulation of large numbers of white cells in the maternal blood spaces of the placenta) has been particularly associated with growth restriction, while acute infection with high parasitaemia was associated with preterm delivery.26 Our observations suggest that the pathogenesis of these complications may be similar in PNG and Africa.
Gravidity, smoking and low maternal Hb were associated with significant reductions in birth weight in adjusted analyses. Although mild anaemia may not decrease birth weight,27 our cohort is notable for the severity of anaemia detected at enrolment, one-third of women had Hb <8 g/dl, as did 23% of women at delivery. Febrile symptoms were associated with lower birth weight in unadjusted analyses, but there was no association between febrile symptoms and malaria as determined by microscopy or PCR. The reductions in birth weight in first pregnancies translated into a high rate of LBW and preterm birth delivery in primigravid women, possibly due to their lack of pregnancy-specific immunity,2 or the lower general risk of LBW associated with increased gravidity.
Most participants reported using bed nets that were not impregnated with insecticides, and although bed net users were less commonly parasitaemic at delivery, this did not translate into decreases in LBW or preterm delivery. The decrease in peripheral blood parasite prevalence at delivery associated with bed net use is consistent with insecticide treated bed net studies from Africa,28 but bed net use and ownership might be a proxy for higher socio-economic status, better housing and decreased vector exposure, rather than directly mediating decreased malaria prevalence.
Previous studies performed in the region showed a similar prevalence of malaria infection, LBW and anaemia.6,7 The prevalence of preterm delivery was 21.7%, whereas it was estimated at 6–10% in other studies,6,7 probably due to differences in sensitivity of dating techniques used. We used Ballard scores, whereas fundal height supplemented by Dubowitz assessments, and Dubowitz assessments supplemented by cutaneous assessments were used in previous studies.6,7 Although a comparison suggests similar performance of Dubowitz and Ballard scores in Malawi,29 differences in staff training and interpretation may affect prevalence. Early ultrasound dating, the gold standard for estimating gestation was not available; in recent ultrasound studies in the same province (Unger et al., submitted), preterm delivery was observed in 10.6% of women receiving similar care.
Host genetics were associated with Hb levels and pregnancy outcomes, but not with malaria risk. Women with α+-thalassaemia had decreased Hb at delivery, while CR1 and SAO polymorphisms were associated with increased maternal Hb at delivery. SAO was associated with a non-significant increase in risk of LBW, and decrease in mean birth weight (but not preterm birth). Previous studies in Madang have not shown a significant relationship between SAO and pregnancy outcome,8,15,16 and in the absence of a clear mechanism by which it might affect fetal growth it seems unlikely to be a true effect. CR1 and α+-thalassaemia polymorphisms were not significantly associated with birth weight, or with risk of LBW and preterm birth. The α+-thalassaemia observations were consistent with an earlier study,15 while this is the first study of the effect of CR1 genotype on pregnancy malaria and pregnancy outcomes.
In this cohort, one-third of women had sub-microscopic malaria infection at first presentation, and a similar number at delivery. Between half and two-thirds of all infections were sub-microscopic. Women with sub-microscopic infection at delivery had increased rates of LBW, which approached significance; there was no significant impact on birth weight. Further studies on larger sample sets are indicated. The importance of sub-microscopic infections for pregnancy outcome is not fully resolved,30 and further studies are required to determine whether PCR can identify women at risk of adverse outcomes.
This study had some limitations. First, a convenience sample of women living close to the health centre was recruited, malaria or anaemia may be greater problems in more poorly served communities and we do not have information on non-enrolled women. Second, delivery data were only available on 70% of enrolled women, as many delivered at home. Third, the sample size was modest, which limited power to detect significant associations between maternal characteristics and malaria or pregnancy outcome. Few statistically significant associations were observed, and large magnitudes of effect for some non-significant associations may indicate that lack of statistical power could be concealing true associations. Nevertheless, we did demonstrate important associations between birth weight, risk of LBW or preterm delivery and factors such as smoking, Hb levels and education, and showed that histological diagnosis of chronic placental malaria is an important risk factor for LBW and preterm delivery, as well as confirming expected associations with gravidity.
Conclusions
Up to two-thirds of pregnant women living in an area of high malaria transmission of coastal PNG had malaria infection at the first antenatal visit and almost one-half at delivery. Of these, women with chronic placental infection had increased rates of LBW and preterm delivery; sub-microscopic infection may increase the risk of LBW delivery. Red blood cell polymorphisms that are common in PNG do not seem to explain the high rates of anaemia detected; malaria and possibly iron deficiency may instead be responsible. Strategies to better prevent malaria, and to improve maternal Hb, may decrease the burden of adverse pregnancy outcomes in settings like PNG.
Acknowledgments
Authors' contributions: SJR, IM, JGB, CLK and PMS conceived the study; SJR and DIS designed the study protocol; FB and DIS recruited patients; AU and CC carried out laboratory studies; KAM and SJR drafted the manuscript; FJF, IM, JGB and CLK critically revised the manuscript for intellectual content. All authors read and approved the final manuscript. DIS and SJR are guarantors of the paper.
Acknowledgements: We thank the staff of the Alexishafen Health Centre for their enthusiastic cooperation with the study, staff of the PNG Institute of Medical Research (PNG IMR) for assistance with microscopy, data entry and study administration, and the pregnant women for their participation.
Funding: This work was supported by AusAID (grant to PNG IMR), the National Health and Medical Research Council of Australia (project grants to JGB and SJR, Senior Fellowships to JGB and IM); Australian Research Council (Future Fellowship to FJF); Wellcome Trust (Senior Fellowship to SJR); and Veterans Affairs Research Service (to CLK). DIS was supported by a grant to the PNG Institute of Medical Research from the Bill & Melinda Gates Foundation's Global Health Program [# 34678]. The Burnet Institute and the Walter and Eliza Hall Institute are supported by the NHMRC Infrastructure for Research Institutes Support Scheme and Victorian State Government Operational Infrastructure Support.
Competing interests: None declared.
Ethical approval: Approval was granted by the PNG Medical Research Advisory Council, the Melbourne Health Human Research Ethics Committee and Alfred Health Human Research Ethics Committee. The procedures followed were in accordance with the ethical standards of the Helsinki Declaration (1964, amended most recently in 2008) of the World Medical Association.31
References
- 1. Dellicour S, Tatem AJ, Guerra CA, et al. Quantifying the number of pregnancies at risk of malaria in 2007: a demographic study. PLoS Med 2010;7:e1000221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Rogerson SJ, Hviid L, Duffy PE, et al. Malaria in pregnancy: pathogenesis and immunity. Lancet Infect Dis 2007;7:105–17. [DOI] [PubMed] [Google Scholar]
- 3. Fried M, Nosten F, Brockman A, et al. Maternal antibodies block malaria. Nature 1998;395:851–2. [DOI] [PubMed] [Google Scholar]
- 4. Desai M, ter Kuile FO, Nosten F, et al. Epidemiology and burden of malaria in pregnancy. Lancet Infect Dis 2007;7:93–104. [DOI] [PubMed] [Google Scholar]
- 5. WHO. World Malaria Report 2014: Papua New Guinea. Geneva: World Health Organization; 2014. [Google Scholar]
- 6. Allen SJ, Raiko A, O'Donnell A, et al. Causes of preterm delivery and intrauterine growth retardation in a malaria endemic region of Papua New Guinea. Arch Dis Child Fetal Neonatal Ed 1998;79:F135–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Brabin BJ, Ginny M, Sapau J, et al. Consequences of maternal anaemia on outcome of pregnancy in a malaria endemic area in Papua New Guinea. Ann Trop Med Parasitol 1990;84:11–24. [DOI] [PubMed] [Google Scholar]
- 8. Benet A, Khong TY, Ura A, et al. Placental malaria in women with South-East Asian ovalocytosis. Am J Trop Med Hyg 2006;75:597–604. [PubMed] [Google Scholar]
- 9. Mayor A, Bardaji A, Felger I, et al. Placental infection with Plasmodium vivax: a histopathological and molecular study. J Infect Dis 2012;206:1904–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Genton B, al-Yaman F, Mgone CS, et al. Ovalocytosis and cerebral malaria. Nature 1995;378:564–5. [DOI] [PubMed] [Google Scholar]
- 11. Allen SJ, O'Donnell A, Alexander ND, et al. Prevention of cerebral malaria in children in Papua New Guinea by southeast Asian ovalocytosis band 3. Am J Trop Med Hyg 1999;60:1056–60. [DOI] [PubMed] [Google Scholar]
- 12. Cockburn IA, Mackinnon MJ, O'Donnell A, et al. A human complement receptor 1 polymorphism that reduces Plasmodium falciparum rosetting confers protection against severe malaria. Proc Natl Acad Sci USA 2004;101:272–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lin E, Tavul L, Michon P, et al. Minimal association of common red blood cell polymorphisms with Plasmodium falciparum infection and uncomplicated malaria in Papua New Guinean school children. Am J Trop Med Hyg 2010;83:828–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Fowkes FJ, Michon P, Pilling L, et al. Host erythrocyte polymorphisms and exposure to Plasmodium falciparum in Papua New Guinea. Malar J 2008;7:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. O'Donnell A, Raiko A, Clegg JB, et al. Alpha+ -thalassaemia and pregnancy in a malaria endemic region of Papua New Guinea. Br J Haematol 2006;135:235–41. [DOI] [PubMed] [Google Scholar]
- 16. O'Donnell A, Raiko A, Clegg JB, et al. Southeast Asian ovalocytosis and pregnancy in a malaria-endemic region of Papua New Guinea. Am J Trop Med Hyg 2007;76:631–3. [PubMed] [Google Scholar]
- 17. McNamara DT, Kasehagen LJ, Grimberg BT, et al. Diagnosing infection levels of four human malaria parasite species by a polymerase chain reaction/ligase detection reaction fluorescent microsphere-based assay. Am J Trop Med Hyg 2006;74:413–21. [PMC free article] [PubMed] [Google Scholar]
- 18. Ismail MR, Ordi J, Menendez C, et al. Placental pathology in malaria: a histological, immunohistochemical, and quantitative study. Hum Pathol 2000;31:85–93. [DOI] [PubMed] [Google Scholar]
- 19. Michon P, Cole-Tobian JL, Dabod E, et al. The risk of malarial infections and disease in Papua New Guinean children. Am J Trop Med Hyg 2007;76:997–1008. [PMC free article] [PubMed] [Google Scholar]
- 20. Koepfli C, Colborn KL, Kiniboro B, et al. A high force of Plasmodium vivax blood-stage infection drives the rapid acquisition of immunity in Papua New Guinean children. PLoS Negl Trop Dis 2013;7:e2403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Martinez-Espinosa FE, Daniel-Ribeiro CT, Alecrim WD. Malaria during pregnancy in a reference centre from the Brazilian Amazon: unexpected increase in the frequency of Plasmodium falciparum infections. Mem Inst Oswaldo Cruz 2004;99:19–21. [DOI] [PubMed] [Google Scholar]
- 22. Amoa AB, Lavu E, Ray U, et al. The aetiology of severe anaemia among antenatal patients of the Port Moresby General Hospital. PNG Med J 2003;46:143–51. [PubMed] [Google Scholar]
- 23. Rijken MJ, McGready R, Boel ME, et al. Malaria in pregnancy in the Asia-Pacific region. Lancet Infect Dis 2012;12:75–88. [DOI] [PubMed] [Google Scholar]
- 24. Hommel M, Elliott SR, Soma V, et al. Evaluation of the antigenic diversity of placenta-binding Plasmodium falciparum variants and the antibody repertoire among pregnant women. Infect Immun 2010;78:1963–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Karunajeewa HA, Mueller I, Senn M, et al. A trial of combination antimalarial therapies in children from Papua New Guinea. N Engl J Med 2008;359:2545–57. [DOI] [PubMed] [Google Scholar]
- 26. Menendez C, Ordi J, Ismail MR, et al. The impact of placental malaria on gestational age and birth weight. J Infect Dis 2000;181:1740–5. [DOI] [PubMed] [Google Scholar]
- 27. Kozuki N, Lee AC, Katz J. Moderate to severe, but not mild, maternal anemia is associated with increased risk of small-for-gestational-age outcomes. J Nut 2012;142:358–62. [DOI] [PubMed] [Google Scholar]
- 28. Gamble C, Ekwaru PJ, Garner P, ter Kuile FO. Insecticide-treated nets for the prevention of malaria in pregnancy: a systematic review of randomised controlled trials. PLoS Med 2007;4:e107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Verhoeff FH, Milligan P, Brabin BJ, et al. Gestational age assessment by nurses in a developing country using the Ballard method, external criteria only. Ann Trop Paediatr 1997;17:333–42. [DOI] [PubMed] [Google Scholar]
- 30. Mayor A, Moro L, Aguilar R, et al. How hidden can malaria be in pregnant women? Diagnosis by microscopy, placental histology, polymerase chain reaction and detection of histidine-rich protein 2 in plasma. Clin Infect Dis 2012;54:1561–8. [DOI] [PubMed] [Google Scholar]
- 31. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 2000;284:3043–5. [PubMed] [Google Scholar]