In Sub-Saharan Africa, an estimated 560 000 children acquired HIV and 450 000 died from AIDS-related complications in 2004 . This huge burden of pediatric HIV infection is fuelled by the high HIV prevalence among pregnant women, ranging between 8 and 42% across several countries . The HIV epidemic in Sub-Saharan Africa has been superimposed on an existing syphilis endemic which affects 4–15% of pregnant women . In Tanzania, maternal syphilis was responsible for ∼50% of stillbirths, ∼25% of preterm low birth weight (LBW) deliveries and 17% of adverse pregnancy outcomes . It is estimated that 492 000 infants in Sub-Saharan Africa die annually from congenital syphilis .
Because HIV and syphilis are predominantly transmitted through sexual contact, co-infections during pregnancy are common; in several studies, 8–11% of pregnant women were found to be co-infected [6–8]. Co-infection may have additive or synergistic effects on the risk of adverse pregnancy outcome. For example, placental pathological changes associated with syphilis infection, such as chronic inflammation, villous atrophy and focal necrosis [9,10] may damage the maternal–fetal endothelial vascular interface thereby impairing fetal growth and/or increasing HIV mother-to-child transmission (MTCT). Syphilis may also facilitate HIV MTCT by increasing the risk of preterm delivery, a known risk factor for HIV MTCT [11,12].
Two previous studies [7,13] have noted an association between maternal syphilis and the risk of HIV MTCT. HIV MTCT was more common in mothers with active syphilis (100%; 4/4) than in those with a history of treated syphilis (21%, 3/14; P = 0.01) or no syphilis (14%; 5/35; P = 0.0015), but the small sample size precluded adjusting for confounders . In secondary analysis of a study assessing the effects of breastfeeding on HIV-MTCT , maternal syphilis was a significant predictor of HIV MTCT in multivariate analysis [adjusted hazard ratio, 1.78; 95% confidence interval (CI), 1.03–3.07]. However, the study neither adjusted for maternal HIV viral load, an important predictor of HIV MTCT , nor reported the effect of syphilis on in utero and intrapartum HIV MTCT. In contrast to these two studies, no association between maternal syphilis and HIV MTCT was found in other studies [14,15]. This may be due to the misclassification of maternal syphilis status since only the non-treponemal test [such as the rapid reagin (RPR) test] or the presence of genital ulcers was used to define syphilis infection in these latter studies.
We conducted secondary data analysis of a prospective cohort study investigating the effect of placental malaria on HIV MTCT, to assess whether maternal syphilis infection was associated with in utero and intrapartum/postpartum HIV MTCT after controlling for HIV viral load, LBW and other confounders.
Participants were enrolled between December 2000 and June 2004 at Queen Elizabeth Central Hospital (QECH) in Blantyre district, Malawi. QECH is the main tertiary referral hospital in southern Malawi and the district hospital for Blantyre.
Details of the enrollment process for the parent study have been described elsewhere . Briefly, pregnant women admitted in the antenatal ward, before onset of labor or in the latent phase of labor, were screened for eligibility, after undergoing routine medical and obstetric assessment. Women were excluded if they were in the active phase of labor, were participating in other research studies, lived outside Blantyre district, were less than 15 years old or had hypertension, multiple gestations or altered consciousness. Informed consent was sought from eligible women, and venous blood samples were collected from consenting women after HIV pre-test counseling. Blood was tested for HIV, malaria and syphilis and hemoglobin concentration was measured. If women were infected with HIV, CD4 cell counts and peripheral HIV-1 viral load assays were also performed. Malaria-infected women were treated with oral sulphadoxine–pyrimethamine or quinine, in accordance with the Malawi Standard Treatment Guidelines (MSTG). Because the median time from enrollment to delivery was 4.5 days and syphilis testing was performed in batches, often after delivery, syphilis-infected women and their infants were treated postnatally with intramuscular benzanthine penicillin, in accordance with the MSTG. Some HIV-infected women left the antenatal ward after enrollment and never returned to the hospital to deliver (Fig. 1).
Women received post-test HIV counseling, and demographic and medical information were obtained using a standardized questionnaire. In addition, women underwent physical medical examination, which included measurement of weight and height. After delivery, placental tissue biopsies were collected for examination for malaria infection, while rolled sections of placental membranes and transverse sections of the umbilical cord were collected for assessment of chorioamnionitis. These samples were fixed in neutral buffered formalin, processed routinely and stained with Gurr's modified Giemsa (placental tissue; Merck BDH, Poole, UK) or hematoxylin and eosin (H&E) (membranes and cord).
Newborn infants were weighed and blood samples were drawn by heel prick, within 48 h of birth, for HIV testing. HIV-infected women and their infants received nevirapine (Roxane Laboratories Inc, Columbus, Ohio, USA) according to the HIVNET 012 protocol .
After discharge, HIV-infected mothers and their infants were followed up at 6 and 12 weeks postpartum, when heel-prick blood samples were collected from infants for repeat HIV testing. Information was collected regarding breast-feeding and maternal and infant morbidity using a standardized questionnaire.
The study was approved by the Malawi College of Medicine Research and Ethics Committee and by institutional review boards at the University of Michigan and the University of North Carolina at Chapel Hill.
Maternal HIV status was determined with the Determine™ HIV-1/2 Rapid Test (Abbott Laboratories, Illinois, USA) and the SeroCard™ HIV-1/2 Rapid Test (Trinity Biotech Plc, Co Wicklow, Ireland). Discordant results were resolved by the HIVSPOT HIV-1/2 Rapid Test (Genelabs Diagnostics, Singapore). Maternal HIV RNA assays were performed on stored plasma samples, using the Roche Amplicor HIV-1 Monitor® Test, version 1.5 (Roche Diagnostics, Branchburg, New Jersey, USA). Infant HIV status was determined with a real-time PCR against the HIV-1 long terminal repeat . The tests were performed in duplicate and discordant results warranted a third test which was the tie-breaker.
Syphilis testing was done using the RPR test (Omega Diagnostics, Alloa, Scotland), and if positive, confirmed by the Treponema pallidum hemagglutination assay (TPHA) (Omega Diagnostics, Alloa, Scotland). T. pallidum PCR was performed according to the methods described by Centurion-Lara et al.  on DNA extracted from unfixed, frozen placental biopsies and umbilical cord blood samples using QIAamp DNA mini-kits (Qiagen, Hilden, Germany). Hemoglobin concentrations were measured using a Hemocue hemoglobinometer (HemoCue AB, Ängelholm, Sweden). T-cell subset counts were performed by FACScount (Becton Dickinson, San Jose, California, USA).
Malaria infection was determined by histopathological examination of placental tissue sections, as described before . Chorioamnionitis was assessed by histopathological examination of placental, membrane and umbilical cord sections according to the methods described previously . Finally, among syphilis-infected mothers, histopathological examination of placental, membrane and umbilical cord sections was performed to detect abnormalities that have previously been associated with syphilis infection [9,10].
Infant HIV infection
Infants who tested positive for HIV DNA within 48 h of birth were considered to have been infected in utero (IU) . Infants who tested HIV negative at birth but tested positive at 6 and/or 12 weeks postnatally were considered to have been infected intrapartum (IP) or postpartum (PP) .
Maternal syphilis infection
Women were defined as syphilis-infected if they tested positive by both RPR and TPHA.
Definition was based on placental histopathology ; active malaria was defined as the presence of P. falciparum parasites with or without malaria pigment in fibrin and/or maternal leukocytes. Past malaria was defined as the presence of malaria pigment in fibrin or leukocytes in the absence of malaria parasites.
Axillary temperature of > 37.5°C at the time of enrollment or a history of fever within the previous week.
A history of breast swelling and pain or the presence breast inflammation on physical examination during the postnatal period.
Low body mass index (BMI)
BMI < 19.8 kg/m2 which corresponded to the lowest quartile of BMI distribution in the study population.
Low birth weight (LBW)
Birth weight of less than 2500 g.
Data were entered into Microsoft Access and analyzed in SAS, version 8.2 (SAS Institute Inc., Cary, North Carolina, USA). Crude plasma HIV-1 viral loads were log10 transformed to approximate normal distribution. In univariate and multivariate analysis log10 HIV viral load was categorized into quartiles, while CD4 cell count was categorized into the following groups; ≤ 200, 201–500 and > 500. Student's t test was used to compare means of normally distributed continuous variables, while the Wilcoxon rank-sum test was used to compare continuous variables that were not normally distributed. For categorical variables, we used the χ2 test or Fisher's exact test to compare proportions between groups.
In univariate analyses, we computed relative risks (RR) and 95% CI using simple logistic regression to identify potential predictors of IU or IP/PP HIV MTCT. Variables that were significant at a P value ≤ 0.15 in univariate analysis were entered in a multiple logistic regression model. In all the final multivariate models, any factor with a P value ≤ 0.05 was considered statistically significant. The variables LBW and 'HIV viral load’ were included in the final multivariate models a priori since they are known predictors of HIV MTCT. We used the formulae (RR−1)/RR to calculate the attributable risk proportion and Psyphilis(RR−1)/[1+Psyphilis(RR−1)] (Psyphilis = prevalence of syphilis) to calculate the Population Attributable Risk, as described by Rothman .
The study profile is summarized in Fig. 1. From December 2000 to June 2004, 3821 pregnant women were enrolled and screened for HIV, of whom 1155 (30.2%) were infected. Syphilis results were available for 3791 women, of whom 198 (5.2%) were infected. Results for both HIV and syphilis tests were available from 3787 women; of these 92 (2.4%) were co-infected. The prevalence of syphilis was higher in HIV-1 infected (HIV+) women (8.0%, 92/1147) than in HIV-1 uninfected (HIV−) women (4.0%, 106/2640; P < 0.0001).
Of the 1155 enrolled HIV+ women, 884 (76.5%) delivered at QECH. Table 1 shows the characteristics of HIV+ women who delivered at QECH and those who did not. Women who delivered at QECH were more likely than those who did not to be primigravidae, to have post-primary education and to make frequent antenatal visits. At the time of enrollment, they had a higher median hemoglobin concentration and a lower prevalence of recent fever than those who delivered elsewhere (Table 1). In addition, a higher proportion of delivering women had good quality housing and they were less likely to have peripheral malaria and syphilis than those delivering elsewhere, although these differences were not statistically significant. However, the two groups were similar for age, marital status and CD4 cell count levels.
Of the 884 HIV+ women who delivered at QECH, 29 had twins or triplets who were excluded from analysis. The remaining 855 women delivered singleton infants, of whom 48 were stillborn or died within 48 h of birth. Blood samples were not available from these infants. There was no significant difference in the prevalence of syphilis between mothers of non-surviving infants (6.4%, 3/47) and those of surviving infants (7.5%, 60/804; P = 0.99, Fisher's exact test). Of the 807 surviving infants, 649 (80.4%) were delivered vaginally, while 20 (2.5%) and 138 (17.1%) were delivered by elective and emergency cesarean sections, respectively.
Mother-to-child transmission of HIV
Of the 807 live singleton infants, 751 had HIV results from blood samples collected within 48 h of birth. Of these, 65 tested positive indicating an IU HIV MTCT rate of 8.7% (95% CI, 6.6–10.7%). Of the 686 infants who were HIV negative in the first 48 h of birth, 507 infants had HIV tests at 6 and/or 12 weeks postpartum, of whom 89 (17.6%; 95% CI, 14.2–20.9%) tested positive. These infants were classified as having been infected IP/PP.
Overall, 28.1% (211/751) of the infants who had HIV tests at birth were lost to follow-up. These infants were more commonly HIV+ at birth (15.2%, 32/211 versus 6.1%, 33/540; P < 0.0001) and of LBW (28.4%, 60/211 versus 17.1%, 92/538; P = 0.0005) compared to those who completed follow up. Their mothers tended to be less well educated, of lower gravidity, and to have lower BMI (data not shown). Maternal syphilis prevalence, HIV viral load and CD4 cell count did not differ between groups (data not shown).
Maternal syphilis and HIV MTCT
As shown in Table 2, maternal syphilis infection was associated with a significantly increased risk of IU HIV MTCT (RR, 2.62; 95% CI, 1.46–4.71). LBW and log10 maternal HIV viral load were also significantly associated with increased risk of IU HIV MTCT, while maternal CD4 cell count, mode of delivery, acute chorioamnionitis and history of sexually transmitted infection (STI) were not. In contrast, while the risk of IP/PP HIV-1 MTCT was higher in syphilis-infected than uninfected mothers, the difference was not statistically significant (RR, 1.62; 95% CI, 0.87–3.02); P = 0.13). Low CD4 cell count, high log10 HIV-1 viral load, recent fever at enrollment and LBW were significantly associated with an increased risk of IP/PP HIV-1 MTCT, while the mode of delivery and acute chorioamnionitis were not.
As shown in Table 3, after adjusting for maternal log10 HIV-1 viral load and LBW, maternal syphilis remained significantly associated with an increased risk of IU HIV-1 MTCT [adjusted relative risk (ARR), 2.77; 95% CI, 1.40–5.46]. Similarly, after adjusting for log10 maternal HIV-1 viral load, LBW, recent fever and postpartum breast infections, maternal syphilis infection became strongly associated with an increased risk of IP/PP HIV-1 MTCT (ARR, 2.74; 95% CI, 1.58–4.74). Self-reported history of STI and acute chorioamnionitis were not significant predictors of IU or IP/PP MTCT in the final multivariate model and were dropped.
Syphilis-associated placental pathology
Because treponemes could promote MTCT by infecting the placenta and causing pathological damage, we examined for T. pallidum DNA and syphilis-associated pathology in the placenta and cord. PCR amplifications of T. pallidum DNA were performed on placental and umbilical cord DNA samples from 44 infants born to HIV-syphilis co-infected women. Of these, 16 (36.4%) had PCR-amplifiable T. pallidum DNA in either the placenta or cord. No significant association between the PCR reactivity and MTCT was found.
Many placental and cord samples demonstrated histopathological changes previously associated with syphilis [9,10]. Of 56 infants born to syphilis-infected mothers, 20 (35.7%) had at least one histopathological abnormality. The commonest abnormalities included villous enlargement, hypercellular villitis and fetal vasculopathy (Table 4). In cases where infant HIV and placental histopathological data were available, IU HIV MTCT was somewhat more common when any placental pathology was detected (31.6%, 6/19) than when absent (16.1%, 5/31; P = 0.29, exact test). Similarly, among those successfully followed up, IP/PP HIV MTCT occurred more frequently when any placental abnormality was detected (62.5%, 5/8) than when absent (10.5%, 2/19; P = 0.01, exact test).
This study was conducted in a setting in Sub-Saharan Africa where ∼30% pregnant women were HIV-infected and ∼8% of the HIV-infected women had serological evidence of syphilis infection. We found that, after adjusting for confounders including HIV viral load and LBW, maternal syphilis infection was associated with an increased the risk of both IU and IP/PP HIV-1 MTCT by ∼2.7-fold. Assuming a causal relationship between syphilis and HIV MTCT, and by calculating the attributable risk , we estimate that ∼65% of IU and also ∼65% of IP/PP HIV MTCT among HIV-syphilis co-infected women was attributable to syphilis. In addition, by calculating the population attributable risk (PAR) , we estimate that maternal syphilis infection accounted for 4.9% (1 in 20) of all IU and a similar rate of IPP HIV MTCT.
Our observation of an association between maternal syphilis and HIV MTCT is consistent with two previous studies [7,13] but inconsistent with two others [14,15]. The definition of syphilis infection was based on positive RPR results in one of the conflicting studies  and the presence of genital ulcers in the other . Without a confirmatory treponemal test, RPR has a high false positive rate, especially in pregnant women . Genital ulcers are an insensitive indicator of syphilis infection, leading to false negative findings. Thus, non-differential misclassification bias may explain the negative findings of these studies. By utilizing a larger sample than the previous studies and controlling for important potential confounders, our study provides additional evidence of the association between maternal syphilis and HIV MTCT.
The mechanism by which maternal syphilis may be associated with increased risk of HIV MTCT, independent of HIV viral load and LBW, is unclear. However, it is possible that T. pallidum or host immunological response to the organism may compromise structural or functional integrity of the maternal–fetal barrier, thus facilitating HIV MTCT. Histopathological abnormalities were common in placental and umbilical cord samples from infants born to HIV-syphilis co-infected women. Although the abnormalities were not syphilis-specific, IU and IP/PP MTCT tended to occur more frequently when at least one placental abnormality was present than when absent. However, the study did not adequately assess the association between syphilis-associated placental histopathological changes and HIV MTCT due to missing data and because the study was not designed for this purpose. Thus, to unravel the mechanism by which syphilis increases HIV MTCT, future studies should be designed to focus on the assessment of this association.
In this study, syphilis-infected women were more likely than their un-infected counterparts to report a history of STI (data not shown). It may be argued that positive syphilis serology is simply a marker of other STI which are known to increase HIV viral concentration in the birth canal (through local inflammation), thereby increasing the risk of intrapartum HIV MTCT [25,26]. There are two reasons why this may be unlikely. First, many of the syphilis-infected women had placental lesions characteristic of syphilis infections and/or PCR evidence of T. pallidum infection in the placenta or cord suggesting that treponemes could have directly affected the integrity of the placental barrier. Second, we found no significant association between HIV MTCT and either acute chorioamnionitis or STI. Thus, our data suggest that syphilis may increase the risk of HIV MTCT directly.
One potential limitation of our study is selection bias due to loss to follow-up. First, ∼23.5% of the enrolled women did not deliver at the hospital. These women were of poorer socio-economic status and were in poorer state of health those who delivered at the hospital (Table 1). Therefore, the rate of HIV MTCT may have been higher in women lost-to-follow-up than those who delivered at the hospital. In the absence of HIV MTCT data from women lost-to-follow-up, it is difficult to estimate the magnitude of selection bias in our study.
Second, the HIV status of ∼6.9% of singleton live births could not be assessed at birth due to missing results. However, the unavailability of HIV results was neither related to syphilis status nor any known predictors of HIV MTCT (data not shown). Thus, it is unlikely the missing HIV results could have introduced significant selection bias in this study.
Third, ∼28.1% of infants who had HIV test results at birth were lost to follow-up. These infants had a higher prevalence of LBW and HIV infection at birth than those who completed follow-up. Thus, the overall rate of IP/PP HIV MTCT may have been higher among infants lost-to-follow-up. However, the two groups of women had a similar prevalence of syphilis, mean CD4 cell count and HIV viral load, and other characteristics. Thus, it is unlikely that the rates of IP/PP HIV MTCT in these two groups of women were significantly different.
Our study may have underestimated the true impact of syphilis on IU HIV-1 MTCT. Because pregnant women were enrolled in the late third trimester, we could have missed an association between maternal syphilis infection and fetal HIV infection which resulted in abortions and very premature deliveries. Similarly, because stillborn infants were not tested for HIV, we could have missed an association between maternal syphilis and IU HIV infection which led to stillbirths.
The prevalence of syphilis among our study subjects (8%) falls within the range of 4–15% reported previously in Sub-Saharan Africa , and the rate of IU HIV MTCT (8.7%) is comparable with findings in recent studies [27–29]. However, the rate of IP/PP MTCT (17.6%) is higher than the rates reported in those studies [27–29], perhaps because, in our study, CD4 cell counts were more frequently ≤ 200 cells/μl (25.5% of participants), LBW was common (20.3%) and some women (35.4%) took nevirapine less than 2 h before delivery. Thus, our population may have been at higher risk for HIV MTCT than women in the previous studies. Nevertheless, based on the range of syphilis prevalence in pregnancy reported in Sub-Saharan Africa and the attributable risk of syphilis to IU and IP/PP MTCT found in this study, we estimate, by calculating of PAR , that maternal syphilis may account for 2.5–8.2% of IU and IP/PP HIV MTCT cases in Sub-Saharan Africa.
The Malawi Ministry of Health's policy is to screen pregnant women for syphilis at first antenatal clinic attendance and institute prompt treatment to infected women. Our finding of undiagnosed and untreated syphilis-infected women in the late third trimester suggests shortfalls in policy implementation, as also found in countries in Tanzania . Thus, the finding of an association between syphilis and HIV MTCT could provide additional rationale for health systems in Sub-Saharan Africa to identify and overcome bottlenecks in syphilis screening and treatment programs for antenatal women.
Our study findings emphasize the need for integration of HIV and syphilis services at antenatal clinics. Currently, financial resources for prevention of mother-to-child transmission of HIV (PMTCT) in resource-poor countries are increasing through several donors including the Global Fund to fight AIDS Tuberculosis and Malaria which will enable countries to scale-up HIV interventions, including antenatal HIV screening. The cost of syphilis screening and treatment is estimated at US$4–19 per disability adjusted life year averted , a figure which would decrease if the effect of syphilis on HIV MTCT is considered. Thus, integration of syphilis management in PMTCT programs would add little cost, but have a major impact on child survival.
In conclusion, our study found that maternal syphilis is associated with IU and IP/PP HIV-1 MTCT. However, the mechanism for the increased risk of HIV-1 MTCT remains unclear. Integration of syphilis management in PMTCT programs may improve infant survival in areas where syphilis and HIV are co-endemic.
We thank the following members of staff from the Malawi-Liverpool Wellcome Clinical Research Programme; Mrs Njiragoma and Mrs Munthali, who were in-charge of the nursing team collecting research data, Mr Mkundika (deceased) and Mr Kanjala who processed laboratory specimens, Dr S. White and Mr P. Malange who led the data management team and Mr Visopo Harawa who performed DNA extraction. We also thank Sheila Lukehart (University of Washington) and Lola Stamm (University of North Carolina) for their advice and support in performing T. pallidum PCR and Chin-Yih Ou (Centers for Disease Control and Prevention) for his technical assistance on performing infant HIV-1 DNA PCR.
Sponsorship: Supported by NIH grant # AI 49084, the NIH-FIC grant # 5 D43 TW00908 and the Center for AIDS Research at the University of North Carolina. SJR is a Wellcome Trust Senior Overseas Fellow in Biomedical Sciences. Pregnant women in this study were given nevirapine provided by Global Strategies for HIV Prevention.
2. Detailed Listing of Estimates of HIV-1 & HIV-2 Seroprevalence, by Residence and Risk Factor, for Developing Countries: Circa 2003.
In U.S. Bureau of Census, 2004; 2004: Website: http://www.census.gov/ipc/www/hiv3afri.pdf
, Accessed on April 05, 2005.
3. Schulz KF, Cates W Jr, O'Mara PR. Pregnancy loss, infant death, and suffering: legacy of syphilis and gonorrhoea in Africa. Genitourin Med 1987; 63:320–325.
4. Watson-Jones D, Gumodoka B, Weiss H, Changalucha J, Todd J, Mugeye K, et al
. Syphilis in pregnancy in Tanzania. I. Impact of maternal syphilis on outcome of pregnancy. J Infect Dis 2002; 186:940–947.
5. Schmid G. Economic and programmatic aspects of congenital syphilis prevention. Bull World Health Org 2004; 82:402–409.
6. Temmerman M, Ali FM, Ndinya-Achola J, Moses S, Plummer FA, Piot P. Rapid increase of both HIV-1 infection and syphilis among pregnant women in Nairobi, Kenya. AIDS 1992; 6:1181–1185.
7. Coutsoudis A, Pillay K, Spooner E, Kuhn L, Coovadia HM. Influence of infant-feeding patterns on early mother-to-child transmission of HIV-1 in Durban, South Africa: a prospective cohort study. South African Vitamin A Study Group. Lancet 1999; 354:471–476.
8. Mbopi Keou FX, Mbu R, Mauclere P, Andela A, Tetanye E, Leke R, et al
. Antenatal HIV prevalence in Yaounde, Cameroon. Int J STD AIDS 1998; 9:400–402.
9. Mbopi Keou FX, Mbu R, Mauclere P, Andela A, Tetanye E, Leke R, et al
. Placental histopathology of congenital syphilis. Obstet Gynecol 2002; 100:126–133.
10. Qureshi F, Jacques SM, Reyes MP. Placental histopathology in syphilis. Hum Pathol 1993; 24:779–784.
11. Maternal viral load and vertical transmission of HIV-1: an important factor but not the only one
. The European Collaborative Study. AIDS
12. Risk factors for mother-to-child transmission of HIV-1. European Collaborative Study
13. Lee MJ, Hallmark RJ, Frenkel LM, Del Priore G. Maternal syphilis and vertical perinatal transmission of human immunodeficiency virus type-1 infection. Int J Gynaecol Obstet 1998; 63:247–252.
14. St Louis ME, Kamenga M, Brown C, Nelson AM, Manzila T, Batter V, et al
. Risk for perinatal HIV-1 transmission according to maternal immunologic, virologic, and placental factors. JAMA 1993; 269:2853–2859.
15. Van de Perre P, Simonon A, Msellati P, Hitimana DG, Vaira D, Bazubagira A, et al
. Postnatal transmission of human immunodeficiency virus type 1 from mother to infant. A prospective cohort study in Kigali, Rwanda. N Engl J Med 1991; 325:593–598.
16. Mwapasa V, Rogerson SJ, Molyneux ME, Abrams ET, Kamwendo DD, Lema VM, et al
. The effect of Plasmodium falciparum malaria on peripheral and placental HIV-1 RNA concentrations in pregnant Malawian women. AIDS 2004; 18:1051–1059.
17. Guay LA, Musoke P, Fleming T, Bagenda D, Allen M, Nakabiito C, et al
. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: HIVNET 012 randomised trial. Lancet 1999; 354:795–802.
18. Luo W, Yang H, Rathbun K, Pau CP, Ou CY. Detection of human immunodeficiency virus type 1 DNA in dried blood spots by a duplex real-time PCR assay. J Clin Microbiol 2005; 43:1851–1857.
19. Centurion-Lara A, Castro C, Shaffer JM, Van Voorhis WC, Marra CM, Lukehart SA. Detection of Treponema pallidum
by a sensitive reverse transcriptase PCR. J Clin Microbiol 1997; 35:1348–1352.
20. Rogerson SJ, Pollina E, Getachew A, Tadesse E, Lema VM, Molyneux ME. Placental monocyte infiltrates in response to Plasmodium falciparum malaria infection and their association with adverse pregnancy outcomes. Am J Trop Med Hyg 2003; 68:115–119.
21. Abrams ET, Milner DA Jr, Kwiek J, Mwapasa V, Kamwendo DD, Zeng D, et al
. Risk factors and mechanisms of preterm delivery in Malawi. Am J Reprod Immunol 2004; 52:174–183.
22. Bryson YJ, Luzuriaga K, Sullivan JL, Wara DW. Proposed definitions for in utero versus intrapartum transmission of HIV-1. N Engl J Med 1992; 327:1246–1247.
23. Rothman K. Modern Epidemiology. 2nd edn. London: Lippincott Williams & Wilkins; 1998.
24. Coutsoudis A, Pillay K, Spooner E, Kuhn L, Coovadia HM. Randomized trial testing the effect of vitamin A supplementation on pregnancy outcomes and early mother-to-child HIV-1 transmission in Durban, South Africa. South African Vitamin A Study Group. AIDS 1999; 13:1517–1524.
25. John GC, Nduati RW, Mbori-Ngacha DA, Richardson BA, Panteleeff D, Mwatha A, et al
. Correlates of mother-to-child human immunodeficiency virus type 1 (HIV-1) transmission: association with maternal plasma HIV-1 RNA load, genital HIV-1 DNA shedding, and breast infections. J Infect Dis 2001; 183:206–212.
26. Mandelbrot L, Mayaux MJ, Bongain A, Berrebi A, Moudoub-Jeanpetit Y, Benifla JL, et al
. Obstetric factors and mother-to-child transmission of human immunodeficiency virus type 1: the French perinatal cohorts. SEROGEST French Pediatric HIV Infection Study Group. Am J Obstet Gynecol 1996; 175:661–667.
27. Moodley D, Moodley J, Coovadia H, Gray G, McIntyre J, Hofmyer J, et al
. A multicenter randomized controlled trial of nevirapine versus a combination of zidovudine and lamivudine to reduce intrapartum and early postpartum mother-to-child transmission of human immunodeficiency virus type 1. J Infect Dis 2003; 187:725–735.
28. Jackson JB, Musoke P, Fleming T, Guay LA, Bagenda D, Allen M, et al
. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: 18-month follow-up of the HIVNET 012 randomised trial. Lancet 2003; 362:859–868.
29. Taha TE, Kumwenda NI, Hoover DR, Fiscus SA, Kafulafula G, Nkhoma C, et al
. Nevirapine and zidovudine at birth to reduce perinatal transmission of HIV in an African setting: a randomized controlled trial. JAMA 2004; 292:202–209.
30. Watson-Jones D, Oliff M, Terris-Prestholt F, Changalucha J, Gumodoka B, Mayaud P, et al
. Antenatal syphilis screening in sub-Saharan Africa: lessons learned from Tanzania. Trop Med Int Health 2005; 10:934–943.