Although HIV infection has been reported to have little effect on pregnancy outcome in the developed world,1-3 early studies from sub-Saharan Africa suggest that infants of HIV-infected mothers may be at increased risk of adverse pregnancy outcomes such as lower birth weight, prematurity, and perinatal and neonatal death.4-7 It is difficult to determine the relative contribution of HIV, however, because these studies often had relatively small sample sizes and limited HIV-specific data, including CD4 cell counts. Furthermore, these studies reported findings from populations affected by HIV types other than clade C, which is dominant in South Africa and has been reported to be more likely to result in mother-to-child transmission than clades A and B, which are found elsewhere in Africa.8-10
Low birth weight has been suggested as a surrogate for prematurity in babies of HIV-infected mothers.5,11 In Nairobi, HIV-infected women were 3 times more likely to deliver a low-birth-weight baby, especially in the presence of HIV-related symptoms.12 In Rwanda, birth weight was significantly lower in singleton babies of asymptomatic women than in babies born to uninfected women.13 In a recent study from Italy, fetal growth assessed by abdominal circumference on ultrasound in HIV-infected women was shown to be reduced in comparison to that in HIV-uninfected women, but there was no effect of the infant's infection status or of exposure to antiretroviral therapy (S. Fiore, PhD, personal communication, 2005). Paradoxically, in developed country settings, highly active antiretroviral therapy (HAART) begun before pregnancy or in early pregnancy has been associated with an increased risk of premature delivery and neonatal death.14
Maternal HIV infection, especially advanced disease, may also be associated with reduced fecundity, miscarriages, and stillbirths.15 Increased stillbirth rates have been reported from areas in which the epidemic has been present for a long time.4 A large study in Nairobi showed an independent association between HIV infection and intrauterine and intrapartum death after controlling for the presence of other sexually transmitted disease (STDs).12
To inform the debate on pregnancy outcome in high-prevalence HIV African settings and to provide a baseline against which more substantial maternal antiretroviral prophylaxis as well as HAART for women with low CD4 cell counts can be evaluated, especially in clade C-affected populations, we present data on pregnancy outcomes (spontaneous abortions, stillbirths, maternal and early neonatal deaths; birth weight; and HIV status) and estimates of the relative risk factors in HIV-infected and uninfected women enrolled in a prospective mother-to-child transmission study investigating the risk of HIV transmission associated with exclusive breast-feeding.
Between September 2001 and September 2004, pregnant women attending 7 rural, 1 semiurban, and 1 urban antenatal clinic in KwaZulu Natal were offered voluntary counseling and testing for HIV. Initially, all HIV-infected women and a systematic subsample of uninfected women were offered enrollment in the study; 1 in 10 women who tested HIV-negative were randomly and confidentially identified through a computer-based program from batches of test results identified only with an encrypted study number. All women had been advised of the study and had given signed written consent for this screening process. These women were offered enrollment when they were subsequently counseled after testing. From July 2003, all women attending an antenatal clinic were offered enrollment before HIV testing. After enrollment, maternal sociodemographic and health information was recorded and antenatal CD4 cell counts were determined in HIV-infected women (September 2001-March 2003; FACScan, Becton Dickinson and Company, Franklin Lakes, NJ; Epics XL, Beckman Coulter, Brea, CA). All pregnant women were routinely offered 5 mg of folate daily, 200 mg of ferrous fumarate from the first antenatal visit until delivery, and 200,000 IU of vitamin A after delivery, and all infected women and their babies were offered single-dose nevirapine.16 Gestational age at delivery was determined on the basis of last menstrual period (LMP) unless adjusted by the midwife on the basis of a clinical assessment. All women were counseled on options for infant feeding.17 At the time of the study, HAART was not available through the health services. The study was approved by the Biomedical Research Ethics Committee of the University of KwaZulu-Natal.
Delivery, birth weight, and nevirapine exposure information was taken from clinic registers and/or from the mother directly, and a dried blood spot sample (DBS) was collected by heel prick on all liveborn babies within 72 hours of birth. If babies were delivered at home, weight measured by the study team within 3 days of birth was used instead. Low birth weight was defined as <2500 g. Data on infant and maternal morbidity were collected at the routine 6-week clinic visit for immunizations. DBSs were also collected from all infants, and 2 independent measurements of maternal middle upper arm circumference (MUAC) were recorded; the average of these 2 measurements was used in analyses.
HIV status of the baby was determined by a quantitative HIV RNA assay (Nuclisens HIV-1 QT, Organon Teknika, Boxtel, The Netherlands; Nuclisens EasyQ HIV-1, Biomerieux, Boxtel, The Netherlands) with a sensitivity of 80 copies of HIV RNA per milliliter of blood (equivalent to 1600 copies of HIV RNA per 50-μL DBS).18 If HIV RNA was not detectable (<80 copies/mL of blood), the baby was considered to be uninfected with HIV. Babies were regarded as infected with HIV if the first sample and a second confirmatory sample had >4000 copies/mL.19 If the second sample had ≤4000 copies/mL, a third sample was obtained. Classification was then based on the results of the third sample (≤4000 copies/mL, HIV uninfected; >4000 copies/mL, HIV infected). For infants with a detectable HIV RNA viral load at 6 weeks, earlier stored samples were tested if available.
Data were captured using optical imaging recognition software (Teleform; Cardiff, San Diego, CA) in a Microsoft Structured Queries Language (SQL) (Seattle, WA) Server database using custom-written applications (Perlcom, Durban, South Africa). Analysis was based on the database created on October 10, 2005 (ACCHC_master_2005-10-10). Adverse pregnancy outcomes were analyzed in STATA (version 9.1; Stata Corporation, College Station TX), whereas analysis on mortality and transmission to 6 weeks postpartum was performed using R version 2.1.0 (R Development Core Team, a language and environment for statistical computing; R Foundation for Statistical Computing, Vienna, Austria). Differences between HIV-infected and uninfected women were assessed using the t test for continuous data with a normal distribution, Mann-Whitney test for nonnormal distributions, χ2 test for categoric variables, and Fisher exact test if numbers were small. Antenatal maternal deaths, spontaneous abortions (spontaneous expulsion of fetus at ≤24 weeks of gestation or <500 g), and stillbirths were combined into a single adverse pregnancy outcome variable for further analysis. The risk of adverse pregnancy outcome and associations between maternal or infant characteristics and low birth weight were assessed in multivariable logistic regression analyses. Mortality in the first 6 weeks of life among all live births was assessed in a Kaplan-Meier analysis, and the association with maternal and infant variables was quantified in a Cox regression analysis. The rate of mother-to-child transmission was based on the HIV test results on an infant sample collected at 6 ± 2 weeks.20
Between September 2001 and September 2004, 3465 women were enrolled, of whom 615 withdrew, mainly because they did not correctly fulfill the selection criterion of likely remaining in the area for long-term follow-up (n = 353), did not fulfill follow-up criteria (n = 71), or had indeterminate or missing HIV status information (n = 9); the proportion that withdrew was similar between HIV-infected and uninfected women and in urban, semiurban, and rural clinics. This left 2850 women with information about HIV status and pregnancy outcome (1401 uninfected with HIV, 1449 infected with HIV; Table 1). HIV-uninfected women were significantly younger than HIV-infected women, were more likely to be enrolled in the more rural clinics, and the current pregnancy was more likely to be their first. HIV-infected women did not differ from uninfected women with respect to highest level of attained education, type of toilet, or type of fuel used for cooking or just boiling water. HIV-infected women were more likely to have access to piped water at home, whereas HIV-uninfected women tended to have to use public stand pipes or river water. There was some evidence to suggest that HIV-infected women may be better off economically (eg, living in a brick house), but the significant difference was attributable to the higher percentage of missing data among infected women on this variable. HIV-infected women were significantly more likely to be the main income provider at home than uninfected women (10.5% vs. 5.8%; P < 0.001), who were more reliant on other household members resident at home or living away. The percentages of women delivering at home did not differ by HIV infection status (see Table 1).
Risk of Adverse Pregnancy Outcome
Six women died in the antenatal period, spontaneous abortions were reported by 17 women, and there were 104 stillbirths (see Table 1). Overall, HIV-infected women were 75% more likely to have an adverse pregnancy outcome (antenatal death, spontaneous abortion, or stillbirth) than uninfected women (81 vs. 46 women; see Table 1, Table 2). Maternal age, gravidity, and maximum level of education attained were not significantly associated with adverse pregnancy outcomes, but in multivariable analyses, HIV infection was independently associated with a 63% increased risk, urban enrollment with a 60% reduced risk, and a nonhospital delivery with a 13-fold increased risk (see Table 2). Limiting the outcome to stillbirth only, a nonhospital or clinic delivery increased the risk 11 times (odds ratio [OR] = 11.45, 95% confidence interval [CI]: 7.45 to 18.34; P < 0.001).
Risk of Adverse Pregnancy Outcome Among HIV-Infected Women
Among the 1449 HIV-infected women, associations between maternal and delivery characteristics and risk of adverse pregnancy outcomes were similar to those found for all women (see Table 2). Women living in an urban setting were less likely to have an adverse pregnancy outcome (adjusted odds ratio [AOR] = 0.39; P = 0.020), whereas those delivering outside a health facility were 13.6-fold more likely to have an adverse pregnancy outcome (P < 0.001).
Five hundred seventy-four (39.6%) of 1449 infected women had a CD4 count >500 cells/mL, 639 (44.1%) had counts between 200 and 500 cells/mL, and 160 (11%) had counts of 200 cells/μL or less. CD4 cell count information was not available for 76 (5.2%) HIV-infected women. In multivariable analysis, including maternal CD4 cell counts, although adjusted ORs on the other variables remained similar, women with a CD4 count <200 cells/mL were nearly twice as likely to have an adverse pregnancy outcome than those with a cell count of 500 cells/mL or more (AOR = 1.86; P = 0.127) and those with cell counts between 200 and 500 cells/mL had a similar risk as the reference category (AOR = 0.88; P = 0.682).The risk was increased 3-fold for women with missing information on CD4 cell count (AOR = 3.22; P = 0.025).
Risk of Having a Low-Birth-Weight Baby
Excluding 53 second liveborn twins, there were 2724 liveborn singleton babies with maternal characteristics similar to those described previously: 1346 boys and 1343 girls (35 newborns of unknown gender), 1368 of whom were born to HIV-infected mothers and 1356 of whom were born to uninfected mothers. Birth weight information was not available for 195 babies (107 born to uninfected mothers and 88 born to HIV-infected mothers). Among the 2529 babies with a known birth weight, maternal HIV infection was associated with reduced birth weight, with a median of 3100 g (range: 1150-4600 g) for HIV-infected mothers versus 3150 g (range: 800-5500 g) for uninfected mothers (P < 0.001) and a greater proportion weighed <2500 g (134 [10.6%] vs. 94 [7.9%] respectively; P < 0.010). In multivariable analyses on 2182 babies with information on all variables (3 excluded with missing gender, 4 with missing maternal age, and 340 with missing MUAC), maternal HIV infection was associated with an increased risk of low birth weight (AOR = 1.45; P = 0.02). Although boys were 20% less likely to be of low birth weight than girls, this did not reach statistical significance (AOR = 0.79; P = 0.13). Maternal nutritional status as indicated by MUAC was independently and significantly associated with the risk of having a low-birth-weight baby when treated as a continuous variable (AOR = 0.93; P < 0.001) and when categorized as less than or greater than the median of 27.6 cm (AOR for <27.6 cm = 1.77; P < 0.001) for this population. Location of maternal residence, education, maternal age, or any of the socioeconomic variables was not significantly associated with low birth weight in the multivariate analysis that included maternal HIV status, gender, and MUAC.
There were 340 women with missing information on MUAC, of whom 129 were uninfected and 211 were infected. In a subsequent analysis including these women (and a missing category for this variable), the AORs on all variables were virtually unchanged (maternal HIV, AOR = 1.49; gender, AOR = 0.75; and MUAC less than median, AOR = 1.80).
Mortality in the First 6 Weeks of Life
A total of 68 (of the 2724 livebirths) babies died in the first 6 weeks of life: 38 (2.7%, 95% CI: 2.0% to 4.0%) born to 1368 HIV-infected mothers and 30 (2.2%, 95% CI: 1.5% to 3.1%) born to 1356 uninfected mothers. Four babies who died were excluded from further analyses because of missing information on gender (all died soon after delivery without a birth weight record, and all had missing HIV infection status information). Overall, early deaths in babies were not significantly associated with maternal HIV infection (hazard ratio [HR] = 1.18; P = 0.52) or with male gender (HR = 0.7; P = 0.16). Babies born to mothers who had completed at least primary school (7 years) were significantly less likely to die than those born to mothers who had not (HR = 0.21, P < 0.001 for 7 years completed; HR = 0.32, P < 0.01 for 10 years completed; and HR = 0.26, P < 0.01 for 12 years or more completed). Babies born to women from urban settings were significantly less likely to die (HR = 0.34; P = 0.045) than those from semiurban settings, but mortality was not increased for those living in rural settings compared with the semiurban reference group. Babies weighing <2500 g were 8.3 times more likely to die than babies with a normal birth weight (P < 0.001), and this association remained strong after stratification by maternal HIV status (Fig. 1; log rank test = 60.0; P < 0.001).
For normal-birth-weight children, there was little difference in early mortality risk by maternal HIV status (see Fig. 1); by 6 weeks of age, an estimated 98.6% (95% CI: 97.9% to 99.3%) in the maternal HIV-infected group and 99.0% (95% CI: 98.4% to 99.6%) in the uninfected group were still alive (P = 0.16 for differences by maternal HIV status). The effect of low birth weight on mortality was substantial for babies in both groups, however; by 6 weeks of life, an estimated 7.7% (95% CI: 3.0% to 12.1%) of low-birth-weight infants born to HIV-infected mothers and 2.2% (95% CI: 0.0% to 5.1%) of low-birth-weight infants of uninfected mothers would have died (P = 0.08 for differences by maternal HIV).
Children for whom birth weight information was not available (n = 163 in the Kaplan-Meier analyses of Fig. 1) were likely to have died (HR = 40.3 in univariable analysis and 38.0 in multivariable analysis), but this was largely explained by the lack of data on babies born at home and dying soon afterward. There were 46 children born to HIV-infected mothers with missing birth weight information, of whom 32 were uninfected at 6 weeks, 8 were infected with HIV, and 6 had a missing HIV status.
In an analysis of babies born to HIV-infected women (Table 3), mortality in the first 6 weeks of life was not significantly associated with maternal CD4 cell counts (although the lack of statistical significance for the nearly trebling in risk for babies born to severely immunocompromised women may have been attributable to a limited number of events and lack of statistical power) or with infant HIV infection status (see Table 3). In multivariable analysis, the association with an urban enrollment clinic seen univariably was no longer statistically significant.
Risk of Mother-to-Child Transmission by 4 to 8 Weeks of Age
HIV test results between 4 and 8 weeks of age were available on 962 liveborn babies, of whom 141 (14.7%, 95% CI: 12.5% to 17.1%) were infected. Four hundred four HIV test results were missing because of stillbirths or spontaneous abortions (n = 81), mothers withdrawing from the study (n = 153, mainly antenatal), babies dying before sampling (n = 56), results of the second twin being excluded,21 loss to follow-up,3 and sample being collected outside the 4- to 8-week period (n = 79). The median age of babies at blood sampling was 43 days. If the 31 babies with unknown HIV status who died before 8 weeks of age are regarded as infected, the rate of mother-to-child transmission would have been 17.3% (95% CI: 15.0% to 19.8%, n = 172 of 993 babies).
The risk of HIV transmission was inversely associated with maternal CD4 cell count and birth weight: children born to mothers with a CD4 count <200 cells/mL were at highest risk, but the risk for children with CD4 counts from 200 to 500 cells/mL was still significantly increased when compared with those with values >500 cells/mL. Taking babies with a normal birth weight (>2500 g) as a reference, babies with a birth weight <2500 g were significantly more likely to be infected. Transmission was not associated with a baby's gender, maternal education, or location of residence; the association between increased transmission risk and a home delivery was statistically significant in univariable analysis but not in multivariable regression (Table 4).
In this largely rural population in KwaZulu Natal, South Africa, adverse pregnancy outcomes (maternal antenatal deaths, spontaneous abortions, and stillbirths) were significantly more common among HIV-infected than uninfected women, particularly among those with low CD4 cell counts; rural settings and home deliveries were independently associated with an increased risk. Furthermore, the risk of having a low-birth-weight baby was substantially and significantly increased with maternal HIV infection. Although low birth weight was associated with increased early mortality for infected and uninfected women, the risk of a baby dying in the first 6 weeks of life was not independently associated with maternal HIV infection. Almost 15% of 4- to 8-week-old infants were infected with HIV, which is a rate possibly higher than would be expected in a population in which single-dose nevirapine was routinely available for all infected women and their babies.20
These findings are consistent with other sub-Saharan Africa studies, primarily from urban areas, that HIV infection is linked to an adverse pregnancy outcome, especially for those women with CD4 counts <200 cells/mL.15 Substantially higher stillbirth rates in HIV-infected women have been reported4,15; our somewhat lower rate may have been related to lower background rates of syphilis,22 relatively better nutritional status of pregnant women,23 lower prevalence of advanced HIV disease,4-6,12 or HIV virus clade differences. Our stillbirth rates of 4.35% in infected women and half of that in uninfected women are lower than those found in a retrospective hospital-based study in Rwanda.24 Although this could be suggestive of possible underreporting in our study, the difference is more likely associated with study population characteristics, given the high level of surveillance implemented in the study.
We lacked reliable data on gestational age, but using birth weight as an indirect proxy, our findings are in line with suggestions regarding the risk of premature delivery in HIV-infected women with CD4 counts <200 cells/mL seen in the pretreatment era in developed countries.25,26 Maternal HIV infection has been associated with low birth weight in previous studies in urban Africa.5,11-13,27 We found that maternal HIV infection was associated with an approximate 45% increased risk of having a low-birth-weight baby, whereas maternal nutritional status as indicated by MUAC was significantly and independently associated with the risk of low birth weight. The median of 27.6 cm in this population compares with 27.1 cm in black African nonpregnant women aged 18 to 24 years in the United States.28 The lack of macronutrient malnutrition in this HIV-infected population might suggest that their adverse pregnancy outcomes were instead attributable to micronutrient deficiencies.23 Low birth weight was associated with an 8-fold increased risk of early mortality in babies of uninfected and infected mothers in rural and urban sites in this study. In Tanzania, low-birth-weight babies born to HIV-infected mothers were 5 times more likely to die than babies with a normal birth weight;27 birth weight in infants who were HIV uninfected at 6 weeks continued to influence survival, with those with a low birth weight 3 times more likely to die in the postneonatal period. Together, these findings once again highlight the need for adequate and appropriate antenatal care for all women and that for HIV-infected women, supportive care, including postnatal care (in addition to HIV-specific treatment) remains important to reduce the risks to a mother and her unborn baby.29
Paradoxically, maternal HIV infection was not significantly associated with early infant mortality, although the lack of statistical significance in our study may have been attributable to the limited number of events. A total of 2.5% of liveborn babies died in the first 6 weeks of life, with babies born to women who had at least completed primary school approximately 75% less likely to die than those of women without any education. This 6-week mortality rate is similar to the 1.8% recently reported from Harare, Zimbabwe30 and to rates reported in the breast-fed arm (1.0%) and the formula-fed arm (3.9%) of a randomized trial of infant feeding in Nairobi, Kenya.31
The rate of transmission of HIV by 8 weeks of age was 14.8%, increasing to 17.7% if all babies who died before that age were regarded as infected. This rate could be considered relatively high, given that nevirapine for the prevention of mother-to-child transmission was routinely available in this mostly rural population, where all HIV-infected women were identified before delivery.20 HIV infection rates at 6 weeks of age in this study were similar to those found in Blantyre, Malawi, where single-dose nevirapine was also routinely offered,32 whereas in the Zvitambo study in Harare,30 the rate was 23.8% in the absence of routine single-dose nevirapine prophylaxis. Part of the nevirapine effectiveness “fall-out” may be attributable to a failure to deliver the drug at all-only approximately 10% of all HIV-infected pregnant women globally are thought to receive any prevention of mother-to-child transmission (PMTCT) prophylaxis21; however, the modest reductions in transmission from programs in the region argue for more effective drug regimens to be used in programs at scale33 and for greater attention to the quality and not just the extent of PMTCT coverage. We found transmission to be significantly associated with low birth weight, which is likely to have been a proxy for premature delivery,26,32,34,35 even when allowing for maternal CD4 cell counts. Women with CD4 counts between 200 and 500 cells/mL and <200 cells/mL were 2 and 3 times more likely, respectively, to transmit than women with CD4 counts >500 cells/mL, similar to findings elsewhere.20
Our findings confirm the additional risks that HIV-infected women living in rural Africa, especially those with advanced disease, experience with respect to their own and their baby's survival. The programmatic implications are clear, although complex. High-risk HIV-infected mothers need to be identified during pregnancy and engaged in a continuum of care that includes comprehensive nutrition support, appropriate antenatal care, and HAART for life. Unless the health care systems in southern Africa are substantially strengthened, with a particular focus on HIV interventions, avoidable maternal and infant deaths are going to continue.36 Surveillance mechanisms to track adverse pregnancy outcomes, population-based vertical transmission rates, and infant mortality would help to monitor health system performance and identify possible gaps and “PMTCT fatigue”.37 Finally, preventing and optimizing the care of low-birth-weight babies, irrespective of HIV exposure, remains a challenge if improved overall child survival is to be achieved.38,39
The authors are grateful to all the women and children enrolled in the study and for the continued dedication of field, clinic, and data management staff at the Africa Centre, particularly Thembi Blose, Zanele Fakude, Cookie Govender, Nqobile Mkhwanazi, Londiwe Mthethwa, and Samukelisiwe Mtshali. They also thank Mario Cortina-Borja for preparing data for preliminary analyses presented at the International AIDS Society conference in July 2005; Jan van den Broeck and Colin Newell, who assisted with data management during the study and analyses; Dave Perlman and Kobus Herbst for their assistance in the design of the database; and Geoff Solarsh, Shuaib Kauchali, and Ameena Goga for assistance in the initial formative pilot work and training. In particular, the authors thank the Community Liaison Office of the Africa Centre and the Community Advisory Board for their guidance and feedback throughout the study. They are also grateful to the independent members of the Study Steering Committee and the Data Monitoring and Safety Committees. Members of the Study Steering Committee include Janet Darbyshire (Chair), Nono Simelela (South Africa National Department of Health), and Victoria Sithole (Community Advisory Board). Members of the Data Monitoring and Safety Committee include Cathy Wilfert (Chair, Pediatric AIDS Foundation), Carl Lombard (Statistician, MRC, South Africa), Ames Dhai (Department of Obstetrics and Gynaecology and the Biomedical Ethics Unit, University of KwaZulu-Natal) and Francis Crawley (Free University of Brussels).
1. Bucceri A, Luchini L, Rancilio L, et al. Pregnancy outcome among HIV positive and negative intravenous drug users. Eur J Obstet Gynecol Reprod Biol
2. Newell ML, Thorne C. Pregnancy and HIV infection in Europe. Acta Paediatr Suppl
3. van Benthem BH, de Vincenzi I, Delmas MC, et al. Pregnancies before and after HIV diagnosis in a European cohort of HIV-infected women. European Study on the Natural History of HIV Infection in Women. AIDS
4. Brocklehurst P, French R. The association between maternal HIV infection and perinatal outcome: a systematic review of the literature and meta-analysis. Br J Obstet Gynaecol
5. Temmerman M, Plummer FA, Mirza NB, et al. Infection with HIV as a risk factor for adverse obstetrical outcome. AIDS
6. Turner BJ, McKee LJ, Silverman NS, et al. Prenatal care and birth outcomes of a cohort of HIV-infected women. J Acquir Immune Defic Syndr Hum Retrovirol
7. Leroy V, Ladner J, Nyiraziraje M, et al. Effect of HIV-1 infection on pregnancy outcome in women in Kigali, Rwanda, 1992-1994. Pregnancy and HIV Study Group. AIDS
8. Blackard JT, Renjifo B, Fawzi W, et al. HIV-1 LTR subtype and perinatal transmission. Virology
9. Renjifo B, Fawzi W, Mwakagile D, et al. Differences in perinatal transmission among human immunodeficiency virus type 1 genotypes. J Hum Virol
10. Renjufi B, Gilbert P, Chaplin B, et al. Preferential in-utero transmission of HIV-1 subtype C as compared to HIV-1 subtype A or D. AIDS
11. Taha TE, Dallabetta GA, Canner JK, et al. The effect of human immunodeficiency virus infection on birthweight, and infant and child mortality in urban Malawi. Int J Epidemiol
12. Braddick MR, Kreiss JK, Embree JB, et al. Impact of maternal HIV infection on obstetrical and early neonatal outcome. AIDS
13. Bulterys M, Chao A, Munyemana S, et al. Maternal human immunodeficiency virus 1 infection and intrauterine growth: a prospective cohort study in Butare, Rwanda. Pediatr Infect Dis J
14. Thorne C, Patel D, Newell ML. Increased risk of adverse pregnancy outcomes in HIV-infected women treated with highly active antiretroviral therapy in Europe. AIDS
15. Langston C, Lewis DE, Hammill HA, et al. Excess intrauterine fetal demise associated with maternal human immunodeficiency virus infection. J Infect Dis
16. KwaZulu Natal Provincial Department of Health. Protocol for the Phased Implementation of a Comprehensive Package of Care for the Prevention of Mother to Child Transmission of HIV in Kwazulu Natal
. 4th ed. Pietermaritzburg, S. Africa, 2003.
17. WHO/UNICEF/UNAIDS. HIV and Infant Feeding Counselling: A Training Course
. WHO/SCH/CAH/00.4; UNICEF/PD/NUT/(J)00-3; UNAIDS/99.57Eed. Geneva, Switzerland: World Health Organization; 2000.
18. Cassol S, Gill MJ, Pilon R, et al. Quantification of human immunodeficiency virus type 1 RNA from dried plasma spots collected on filter paper. J Clin Microbiol
19. Brambilla D, Jennings C, Aldrovandi G, et al. Multicenter evaluation of use of dried blood and plasma spot specimens in quantitative assays for human immunodeficiency virus RNA: measurement, precision, and RNA stability. J Clin Microbiol
20. Leroy V, Sakarovitch C, Cortina-Borja M, et al. Is there a difference in the efficacy of peripartum antiretroviral regimens in reducing mother-to-child transmission of HIV in Africa? AIDS
21. UNAIDS/WHO. AIDS Epidemic Update
. UNAIDS/05.19Eed. Geneva, Switzerland: World Health Organization; 2005.
22. Mullick S, Beksinksa M, Msomi S. Treatment for syphilis in antenatal care: compliance with the three dose standard treatment regimen. Sex Transm Infect
23. Papathakis PC, Rollins NC, Brown KH, et al. Comparison of isotope dilution with bioimpedance spectroscopy and anthropometry for assessment of body composition in asymptomatic HIV-infected and HIV-uninfected breastfeeding mothers. Am J Clin Nutr
24. Rahlenbeck S, Hakizimana C. Deliveries at a district hospital in Rwanda, 1997-2000. Int J Gynaecol Obstet
25. O'Shea S, Newell ML, Dunn DT, et al. Maternal viral load, CD4 cell count and vertical transmission of HIV-1. J Med Virol
26. The European Collaborative Study. Maternal viral load and vertical transmission of HIV-1: an important factor but not the only one. The European Collaborative Study. AIDS
27. Wei R, Msamanga GI, Spiegelman D, et al. Association between low birth weight and infant mortality in children born to human immunodeficiency virus 1-infected mothers in Tanzania. Pediatr Infect Dis J
28. Centers for Disease Control and Prevention. US Anthropometric Reference Data US 1976-80. National Health and Nutrition Examination Survey
. Atlanta, GA, 2000.
29. Rosenfield A, Schwartz K. Improving the health of women in developing countries: the time is now. J Midwifery Womens Health
30. Iliff PJ, Piwoz EG, Tavengwa NV, et al. Early exclusive breastfeeding reduces the risk of postnatal HIV-1 transmission and increases HIV-free survival. AIDS
31. Mbori-Ngacha D, Nduati R, John G, et al. Morbidity and mortality in breastfed and formula-fed infants of HIV-1-infected women: a randomized clinical trial. JAMA
32. Chi BH, Wang L, Read JS, et al. Timing of maternal and neonatal dosing of nevirapine and the risk of mother-to-child transmission of HIV-1: HIVNET 024*. AIDS
33. World Health Organization. Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants in resource-limited settings. Towards universal access. Recommendations for a public health approach. Available at: http://www.who.int/hiv/pub/guidelines/en/index.html
. Accessed October 6, 2006.
34. Simonds RJ, Steketee R, Nesheim S, et al. Impact of zidovudine use on risk and risk factors for perinatal transmission of HIV. Perinatal AIDS Collaborative Transmission Studies. AIDS
35. Mandelbrot L, Mayaux MJ, Bongain A, 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
36. Freedman LP, Waldman RJ, de Pinho H, et al. Transforming health systems to improve the lives of women and children. Lancet
37. Rollins NC, Dedicoat M, Danaviah S, et al. Prevalence, incidence, and mother-to-child transmission of HIV-1 in rural South Africa. Lancet
38. Martines J, Paul VK, Bhutta ZA, et al. Neonatal survival: a call for action. Lancet
39. Travis P, Bennett S, Haines A, et al. Overcoming health-systems constraints to achieve the Millennium Development Goals. Lancet