Mother-to-child transmission (MTCT) of HIV can occur in utero, intra-partum and post-partum via breast feeding. Transmission occurs in 13–45 % of infants in the absence of any intervention [1–3]. The contributions of each of these routes to overall transmission have not been quantified exactly but it appears that early in utero infection is relatively infrequent and that most MTCT occurs late in pregnancy or at the time of delivery . Risk factors identified to date for MTCT include maternal virus load, CD4 cell counts, micronutrient status, duration and complications of labour and delivery including prolonged rupture of membranes, histologic chorioamnionitis, breastfeeding and shared HLA class I or II alleles between mother and infant and other genetic factors (reviewed in ).
In developed countries, the use of highly active antiretroviral (ARV) therapy combined with elective Caesarian delivery and avoidance of breastfeeding, have reduced MTCT to less than 2% . In developing countries short-course ARV therapy administered in the intra-partum period has been reported to reduce MTCT by 37–50% in breastfeeding African populations [7–10]. Currently, in most developing countries, the HIVNET 012 protocol that offers single-dose nevirapine to mother and infant in the intra-partum and immediate postnatal period, respectively, is the most common intervention being implemented in an effort to reduce peri-partum MTCT. Thus, the determination of the relative contributions of the three routes or modes of MTCT and the identification of risk factors associated with specific routes of MTCT are increasingly important for the design and interpretation of interventions intended to prevent intra- and early post-partum MTCT.
Polymerase chain reaction (PCR) techniques for the amplification of HIV DNA or RNA are currently employed in the diagnosis of HIV-1 infection particularly in infants under the age of 2 years. However, until recently, the absence of standardized PCR protocols for the detection of HIV strains in different regions of the world may have contributed to the global variation in the reported rates of MTCT and in the determination of the timing of the infection [1,11].
The development of a prototype DNA PCR kit by Roche Diagnostics, that incorporates universal primers SSK145 and SKCC1B for the detection of all the Group M HIV-1 viruses, (Version 1.5, Roche Diagnostics Incorporation, Alameda, California, USA) offers the possibility of accurate and timely detection in early infancy. Such early detection is critically important in the evaluation of ARV trials and other interventions aimed at reducing MTCT, as well as for designing early treatment strategies for infected infants.
We have previously evaluated the prototype Roche kit using whole blood from immediate post-partum mothers with enzyme-linked immunosorbent assay (ELISA)-confirmed HIV status . The prototype Roche DNA PCR assay had a 100% sensitivity and 100% specificity, for the detection of HIV-1 DNA. However, it remained to be established in our setting, if this assay would be equally sensitive in infants with possibly low copies of HIV DNA. The main objective of the present study was to determine the timing of MTCT and the impact of such transmission on mortality in the first 6 months of life, (in the absence of ARV therapy) in a cohort of almost 1000 babies born to HIV-1-positive mothers.
Materials and methods
Whole blood in ethylenediaminetetra-acetic acid (EDTA) was obtained from infants born to the first 1000 HIV-1-positive mothers enrolled in the ZVITAMBO clinical trial, which was designed to assess the impact of immediate post-partum vitamin A supplementation on vertical and horizontal transmission of HIV-1 as well as infant morbidity and mortality. The ZVITAMBO study was a randomized, placebo- controlled trial that enrolled 14 110 mother–infant pairs in Harare, from November 1997 to January 2000. Mother–infant pairs were randomized to one of the four treatment groups: (1) mother received 400 000 IU vitamin A and baby received 50 000 IU vitamin; (2) mother received vitamin A and baby received placebo; (3) baby received vitamin A and mother received placebo; and (4) both mother and baby received placebo. Mothers were enrolled within 90 h of delivery and maternal HIV status was determined by ELISA and confirmed by Western blot. The majority (> 99%) of the mothers were still breastfeeding by 6 months postnatally. Infant whole blood specimens in EDTA were collected at delivery, 6 weeks, 3 and 6 months. Infant specimens were processed and cells pelleted following the manufacturer's instructions and the cell pellets stored at −80°C (Roche Diagnostics Systems) at the Harare Central Hospital Laboratory which is 20 km from the PCR center at the University of Zimbabwe Medical School. At the PCR center, DNA extraction and reagents preparation for the amplification reaction were performed in two separate laboratories, housed in the Departments of Immunology and Paediatrics (both equipped with laminar flow hoods). Amplification and detection were conducted in a third dedicated laboratory equipped with an external-extractor flow hood. Each of the three PCR laboratories is self-contained and movement by the Roche certified scientists (L.S.Z. and M.M.; Roche Amplicor Academy, Luzern, Switzerland, 1999), was unidirectional.
DNA extraction, amplification and detection were performed and results were interpreted according to the manufacturer's instructions (Roche Diagnostic Corporation, Indianapolis, Indianna, USA) as we previously described .
The timing of infant infection was determined using the following working definitions adapted from those of Bryson et al. , Bertolli et al.  and Zijenah et al. . Briefly, infants found to be positive within 96 h of birth, and with one or more positive DNA PCR results at an older age (6 weeks, 3 or 6 months), were considered to have evidence of intra-uterine infection. Infants who were positive at birth but died or missed all subsequent visits, were considered to have acquired the infection in utero. Infants with negative PCR results from the sample obtained within 96 h of birth but who became positive by 6 weeks of age confirmed by a positive PCR result at an older age were considered to have evidence of intra-partum or early post-partum infection. Infants with negative results at birth and at 6 weeks of age but who subsequently became DNA PCR positive were considered to have late post-partum infection.
Comparisons of individual characteristics across infection status groups were made with Fisher's exact tests for the dichotomous variables, and Wilcoxon rank-sum tests for the continuous variables. Because of the interval-censored nature of the HIV-1 infection determinations, transmission risks were estimated using Turnbull's method  using SAS software (version 8 Proc Reliability; SAS Institute, Cary, North Carolina, USA). Forward stepwise logistic regression models were used to analyze risk factors for infection status and timing, as were Cox proportional hazards models for mortality, with P < 0.20 as the model entry criterion. The main mortality analyses used an age timeline beginning at 8 weeks of age, to give some leeway for the scheduled 6-week visits.
Of the 1000 infants born to HIV-1-positive mothers, four (0.4%) were not able to give any specimens at any of the four visits and two had indeterminate PCR results. In addition, seven more infants did not have a baseline PCR result, but were PCR positive at 6 weeks, thus we could not determine whether they had become infected in utero or thereafter. These thirteen infants were therefore excluded from the analyses of timing of infection that compared intra-uterine (IU)-transmission to later transmission, and the mortality analyses relating to that group. For the Turnbull-estimated transmission risks, however, only the first six were excluded, leaving 994 observations.
PCR detection of infant HIV DNA
Timing of infection was analyzed in the 994 infants who had results available for at least one of the scheduled visits during the 6-month follow-up period. The Turnbull-estimated overall 6-month transmission rate based on 249 infants who became infected was 30.7%. In our cohort, 89 [9.4%; 95% confidence interval (CI), 7.7–11.5] were PCR positive for DNA at birth suggesting IU infection. Of the 89 infants with evidence of IU, 75 (84.3%) had PCR-positive results at older ages, whereas 14 (15.7%) infants either died or did not have specimens at an older age. One hundred and four infants out of the 249 HIV-1-positive infants were PCR negative at birth but became PCR positive at 6 weeks with confirmation of these results at an older (3 and 6 months) age. These infants were considered to have been infected intra- and early post-partum (IP/ePP), and the absolute IP/ePP Turnbull-estimated transmission rate was 16.0% (95% CI, 10.8–21.2). Twenty one infants who had negative PCR results at birth and 6 weeks, and had a first positive PCR result either at 3 or 6 months were classified as having evidence of late post-partum (LPP) infection with confirmation of their results at an older age. The absolute Turnbull-estimated LPP transmission rate was 5.3% (95% CI, 1.6–12.2).
There were 35 infants who became infected but whose last negative and first positive samples were too far apart in time to determine whether they were infected IU, IP/ePP or LPP. For example, an infant with a negative test result at birth and a positive test at 6 months could have been infected either IP/ePP or LPP. The information on these infants, however, did contribute to overall transmission rate estimation.
Baseline characteristics of infants stratified by timing of MTCT
The baseline characteristics of (1) IU-infected infants versus all other infants; (2) the IU-infected infants versus IP/ePP-infected infants; and (3) the IP/ePP versus uninfected infants, who all survived up to 8 weeks were compared (Table 1). The mean age, parity, education, income, arm circumference, time of rupture of membranes, mode of delivery and immediate post-partum CD4+ cell counts were not statistically different among mothers who transmitted HIV infection IU (IU transmitters) compared to those who did not transmit the infection to their infants IU (IU non-transmitters) (Table 1). The IU infants had, not unexpectedly, lower birth weight (P = 0.03) and were more likely to die in the first 8 weeks of life compared with infants born uninfected (P = 0.01). A forward stepwise logistic regression analysis was carried out to identify risk factors for IU infection compared to no IU infection, with birth weight and proportion of mothers with CD4+ cell count < 200 × 106 cells/l variables forced into the models. IU transmitters had slightly higher CD4+ cell counts: the odds ratio of transmitting IU was 0.8 among mothers with CD4+ cell counts < 200 × 106 cells/l. (P = 0.62; 95% CI, 0.36–1.85).
Among HIV-infected infants who survived to 8 weeks, birth weight and gestational age were similar for those infected IU and IP/ePP (Table 1). Among women whose HIV-infected infants lived to at least 8 weeks, there were no statistically significant differences in any socio-economic or obstetric characteristic between the IU and IP/ePP transmitters. However, IP/ePP transmitters had significantly more advanced HIV disease compared to IU transmitters: their mean CD4 cell count was 105 × 106 cells/l lower, and over twice as many had less than 200 × 106 cells/l. In a forward stepwise logistic regression analysis to identify risk factors for IP/ePP (versus IU transmission), maternal CD4 < 200 × 106 cells/l increased the odds of being in the IP/ePP group by a factor of 4.6 (P = 0.01; 95% CI, 1.5–14.0), after adjustment for birth weight, maternal education, and maternal arm circumference.
We also compared this group of IP/ePP transmitters to those whose children were alive at 8 weeks of age without evidence of HIV infection. Smaller maternal arm circumference was a risk factor for having a IP/ePP-infected infant (P = 0.006), as was vaginal delivery (P = 0.003). The odds of being uninfected (versus IP/ePP infected) was 4.0 (95% CI, 1.3–20) times greater for the combined Caesarian section categories than for vaginal delivery. Mothers of infants who were HIV negative and alive at 8 weeks had approximately the same CD4 cell distribution as those of the IU transmitter group, with less advanced disease than mothers of IP/ePP-infected infants. In a stepwise logistic model, after adjustment for birth weight, parity, maternal arm circumference, and delivery mode, mothers with CD4 cell counts < 200 × 106 cells/l had a 3.2-fold (P < 0.001; 95% CI, 1.6–6.2) increased risk of infecting their children IP/ePP.
Mortality among HIV-1-negative and HIV-1-positive infants born to HIV-positive women
The crude risk of mortality in the first 6 months of life amongst infants born to HIV-1-positive mothers was 11.8% (116 of 987). Sixty-nine percent (80 of 116) of the children who died were PCR positive, whereas 31% (36 of 116) were HIV negative at the last sampling point before they died (P < 0.001 for crude mortality difference). To avoid bias in comparing mortality among babies infected IU versus IP/ePP, analysis was restricted to those alive at 8 weeks (the end of the 6-week visit period). Between 8 weeks and 6 months of age, there were 35 deaths among the 83 IU-infected infants and 30 deaths among the 103 IP/ePP infected infants, with the Kaplan–Meier estimated mortality risks of 42 and 29% respectively (P = 0.08 log-rank test). The difference between the curves increased steadily over time, with the IU group attaining a risk of 25% mortality 32 days before the IP/ePP group did (103 versus 135 days, Fig. 1). Causes of death were similar between the IU- and IP/ePP-infected infants, with acute respiratory infections/pneumonia being the major causes of death (data not shown). Investigation of the factors related to mortality among this cohort of 186 HIV-infected infants from day 56 to 6 months of age was carried out via Cox proportional hazards regression models. In the final model, timing of infection (IU versus IP/ePP), birthweight, and maternal CD4 were important factors in predicting infant death by 6 months (Table 2).
Our study provides data on timing of MTCT, factors associated with this timing as well as impact on infant mortality in a large cohort of infected mothers and their infants followed up to 6 months in the absence of ARV therapy.
Most previous studies designed to determine the contributions of intra-uterine, intra- and post-partum MTCT have suffered from a number of methodological difficulties, the most important of which has been the lack of a standardized assay suitable for detection of all HIV-1 subtypes.
Although several studies have identified risk factors associated with overall MTCT, few have been large enough to distinguish risk factors for IU and IP/ePP infections. In addition there are no data on mortality rates of IU-infected compared to IP/ePP-infected infants in the first 6 months.
The overall 6-month MTCT rate of 30.7% in our study is within the range of reported rates of 20 to 45% in African women who breastfeed and are not taking ARV drugs (reviewed in ). Our estimated absolute IU transmission rate of 9.4% (approximately 30% of total MTCT) is comparable with that reported by others [14,17,18]. Our estimated absolute IP/ePP rate of 16.0% (approximately 53% of total MTCT) is similar to that reported in the Kigali study . Interestingly, a combined intra-partum and early post-partum transmission rate of 65% has been reported in a breastfeeding population followed up to 18 months . Bertolli and colleagues defined the intra-partum/early post-partum transmission period as negative HIV-1 DNA PCR result in the first 2 days of life and a first DNA PCR-positive result at 3 to 5 months. In our study, IP/ePP transmission was defined as a negative PCR result within 96 h of birth and a first DNA positive at 6 weeks of age, whereas a first positive DNA PCR at 3 or 6 months was defined as LPP transmission. Of course, our estimate for absolute LPP transmission rate of 5.3%, does not take into account infections which may occur after 6 months. The estimated rate of LPP from an international multicentre pooled analysis of children from Butare, Kigali, Ivory Coast and Kenya, followed up to 2 years is about 5% .
Several studies have reported that advanced maternal disease as defined by low CD4 cell counts is a risk factor for overall MTCT (reviewed in ). However, this effect may largely be attributable to the fact that the majority of MTCT in these studies, occurred IP/ePP. Our study demonstrates that advanced maternal disease, as defined by low CD4 cell counts, is not a risk factor for IU transmission, but it is a risk factor for IP/ePP transmission. This finding is similar to the observations of Kuhn and colleagues in New York who found no evidence of association of severity of maternal disease (low CD4 cell counts and clinical status) with IU transmission, but with IP transmission  and Mock and colleagues in Bangkok  who reported that a low percentage of CD4 cells was associated with IP but not IU transmission. Our findings are based on a larger sample size (249 infected infants versus 49 infected infants in the Bangkok study). Mock and colleagues also reported, however, strong relationships between maternal viral load (also an indicator of disease progression) and both IU and IP transmission. We did not have HIV viral load data for the mothers, which would have been a useful adjunct to our CD4 cell count analysis. The association of maternal disease with IP but not IU transmission observed in our study and those of others is interesting. Low maternal CD4 cell counts have been reported to be associated with increased shedding of HIV-infected cells from the cervix and vagina [22,23] thus the infant may be exposed to more viral particles at delivery than IU, with a consequent increased risk of IP transmission.
In agreement with our earlier observations  and those of others [24–26], the mortality rate was higher among HIV-infected infants than uninfected infants.
The mean birth weight of infected infants was significantly lower than that of the uninfected infants, in agreement with observations reported by others [21,24–28]. In addition low birth weight was associated with IU transmission, in agreement with the Bangkok study , but different from that of Kuhn and colleagues in New York who reported that infants with low birth weight had a significantly higher proportion of IP infection when compared with IU-infected infants . In our cohort, IU-infected infants, who had lower birth weight when compared to IP/ePP-infected infants, were more likely to die by 8 weeks. Finally, timing of infection (IU versus IP/ePP), birthweight and maternal CD4 cell counts were all important factors in predicting infant death by 6 months. In Zimbabwe breastfeeding is universal and the mother is also the primary healthcare giver. Advanced maternal disease (defined here by low CD4 cell counts) is likely to result in inadequate nutrition and care for the infant and thereby contribute to increased early mortality.
Our findings on the different contributions of the three modes of MTCT may help to explain the different efficacies of published estimates for the efficacy of ARV drugs administered at different stages of pregnancy: 67% reduction in PACTG 076 , 50% in Thai non-breastfeeding populations  and 37–50% in African breastfeeding populations [7–10]. Our results showing, that of all the transmission that occurs in the first 6 months of infant life, half occurs during the IP/ePP period, emphasize that interventions aimed at reducing this mode of transmission should continue to be of high priority. We have shown, however, that the IU transmission which is not being addressed by the HIVNET 012 intervention is also substantial (9.4%), thus there is also a need to develop other short-course ARV programs targeted at reducing IU transmission as well. It is interesting to note that in the Perinatal AIDS Collaborative Transmission Study, conducted in the USA, the IU transmission rate of women who received zidovudine during pregnancy (according to the PACTG 076 protocol ), did not differ significantly from that of women who did not . Although the impact of selected short-course ARV programs to reduce IU transmission has not been fully determined, the possibility that some of these programs may reduce IU transmission, can not be ruled out. Thus, further research is needed.
The exact contribution of ePP/LPP transmission is difficult to quantify, as the period of ePP transmission is not clearly defined. Several studies conducted in Africa suggest that one-third to one-half of postnatal transmission is due to breastfeeding [31–35], emphasizing the need to also develop appropriate interventions to reduce this transmission. The simplest strategy to prevent this mode of transmission would be complete avoidance of breastfeeding. However, in most developing countries, this is not feasible, mainly because of cost of replacement feeding, culture, poor sanitation conditions and poverty, which are likely to result in higher mortality rates from other childhood diseases if breastfeeding is avoided completely. The recent study conducted in South Africa, which reported that exclusive breastfeeding is associated with a significant reduction in postnatal transmission, is encouraging and requires further investigation . Studies are also required to assess other interventions such as postnatal ARV prophylaxis of HIV-exposed infants and early weaning.
The data we have presented here on the relative contributions of IU and IP/ePP routes of MTCT and the mode of transmission-related risk factors may help in the design of interventions aimed at reducing the risk of MTCT. These data will also serve as a historical comparison in the efficacy assessment and interpretation of results obtained from the HIVNET 012 protocol currently being implemented throughout Africa, in an effort to reduce IP/ePP MTCT.
We are grateful to all the ZVITAMBO Study Group and the participants for their support and co-operation.
Sponsorship: The ZVITAMBO project was supported by the Canadian International Development Agency (R/C Project 690/M3688), United States Agency for International Development (USAID) (cooperative agreement number HRN-A-00-97-00015-00 between Johns Hopkins University and the Office of Health and Nutrition – USAID) and a grant from the Bill and Melinda Gates Foundation, Seattle WA. Additional funding was received from the Rockefeller Foundation (New York, NY) and BASF (Ludwigshafen Germany). This is a collaborative project of The University of Zimbabwe, The Harare City Health Department, the Johns Hopkins Bloomberg School of Public Health, and the Montreal General Hospital Research Institute (McGill University).
1. The Working Group on Mother-to-Child Transmission of HIV. Rates of mother to child transmission of HIV in Africa, America and Europe: results from 13 perinatal studies. J Acquir Immune Defic Hum Retrovir
2. Mofenson LM. Mother-to child HIV-1 transmission: timing and determinants. Obstetr Gynecol Clin North Am
3. Fowler MG. Update:transmission of HIV-1 from mother to child. Curr Opin Obstet Gynecol
4. Kourtis AP, Bulterys M, Nesheim SR, Lee FK. Understanding the timing of HIV transmission from mother to infant. JAMA
5. Willoughby A. Mother-to-child transmission of HIV.
In: Essex M, Mboupe S, Kanki P, Marlink R, Tlou SD, Holmes M (editors). AIDS in Africa
, 2nd ed. New York: Kluwer Academic/Plenum Publisher; 2002, pp. 251–262.
6. Mofenson LM. Munderi P. Safety of antiretroviral prophylaxis of perinatal transmission for HIV-infected pregnant women and their infants. J Acquir Immune Defic Syndr
, 2002, 30
7. Dabis F, Msellati P, Meda N, Welffens-Ekra C, You B, Manigart O, et al
. Six month efficacy, tolerance, and acceptability of a short regimen of oral zidovudine to reduce vertical transmission of HIV in breastfed children in Cote d'Ivoire and Burkina Faso: a double-blind placebo-controlled multicentre trial. Lancet
8. Wiktor SZ, Ekpini E, Karon JM, Nkengason H, Maurice C, Severin ST, et al
. Short course of oral zidovudine for prevention of mother-to-child transmission of HIV-1 in Abidjan, Cote d'Ivoire: a randomized trial. Lancet
9. 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 randomized trial. Lancet
10. The Petra study team. Efficacy of three short-course regimens of zidovudine and lamividune in preventing early and late transmission of HIV-1 from mother to child in Tanzania, South Africa, and Uganda (Petra study): a randomized, double-blind, placebo-controlled trial. Lancet
11. Dabis F, Msellati P, Dunn D, Lepage P, Newell ML, Peckham C, et al
. Estimating the rate of mother to child transmission of HIV. Report of a workshop on methodological issues in Ghent (Belgium). 17-20 February. AIDS
12. Zijenah LS, Humphrey J, Nathoo K, Malaba L, Zvandasara P, Mahomva A, et al
. Evaluation of the prototype Roche DNA amplification kit incorporating the new SSK 145/SKCC1B primers in the detection of HIV in Zimbabwe. J Clin Microbiol
13. Bryson YJ, Luzuriaga K, Sullivan JL, Wara DW. Proposed definition for in utero versus intrapartum transmission of HIV-1. N Engl J Med
14. Bertolli J, St Louis ME, Simmonds RJ, Nieburg P, Kamenga M, Brown C, et al
. Estimating the timing of mother to child transmission of human immunodeficiency virus in a breastfeeding population in Kinshasa, Zaire. J Infect Dis
15. Zijenah, L, Mbizvo MT, Kasule J, Nathoo K, Munjoma M, Mahomed K, et al
. Mortality in the first two years among infants born to HIV-infected women in Harare, Zimbabwe. J Infect Dis
16. Turnbull BW. The empirical distribution function with arbitrarily grouped, censored and truncated data. J R Statist Soc B
17. Simonon A, Lepage P, Karita E, Hitima DG, Dabis F, Msellati P, et al
. An assessment of the timing of mother-to-child transmission of human immunodeficiency virus type 1 by means of polymerase chain reaction. J Acquir Immune Defic Syndr
18. Dunn DT, Brandt CD, Krivine A, Cassol SA, Roques P, Borkowsky W, et al
. The sensitivity of HIV-1 DNA polymerase chain reaction in the neonatal period and the relative contributions of intra-uterine and intrapartum transmission. AIDS
19. Leroy V, Newell M-L, Dabis F, Peckam C, Van de Perre P, Bultreys M, et al
. International multicentre pooled analysis of late postnatal mother-to-child transmission of HIV-1 infection. Lancet
20. Kuhn L, Stesketee RW, Weedon J, Abrams EJ, Lambert G, Bamji M, et al
. Distinct risk factors for intrauterine and intrapartum human immunodeficiency virus transmission and consequences for disease progression in infected children. J Infect Dis
21. Mock PA, Shaffer N, Bhadrakom C, Siriwasin W, Chotpitayasunondh T, Chearskul S, et al
. Maternal viral load and timing of mother-to-child HIV transmission, Bangkok, Thailand. AIDS
22. Mostad SB, Overbaugh J, DeVange DM, Welch MJ, Chohan B, Mandaliya K, et al
. Hormonal contraception, vitamin deficiency, and other risk factors for shedding of HIV-1 infected cells from the cervix and vagina. Lancet
23. John GC, Nduati RW, Mbori-Ngacha D, Overbauch J, Welch M, Richardson BA, et al
. Genital shedding of human immunodeficiency type 1 DNA during pregnancy: association with immunosuppression, abnormal cervical or vaginal discharge, and severe vitamin A deficiency. J infect Dis
24. Lepage P, Van de Perre P, Msellati P, Hitiman DG, Simonon A, Van Goethem C, et al
. Mother-to-child transmission of human immunodeficiency virus type 1 (HIV-1) and its determinants: a cohort study in Kigali, Rwanda. Am J Epidemiol
25. Taha TE, Kumwenda NI, Broadhead RL, Hoover DR, Graham SM, Van Der Hoven L, et al
. Mortality after the first year of life among human immunodeficiency virus type 1-infected children. Pediatr Infect Dis
, 1999, 18
26. Lallemant M, Lallemant-le-Coeur S, Cheynier D, Nzingoula S, Jourdain G, Sinet M, et al
. Mother-to-child transmission of HIV-1 and infant survival in Brazzaville, Congo. AIDS
27. Ryder RW, Nsa W, Hassig SE, Behets F, Rayfield M, Ekungola B, et al
. Perinatal transmission of the human immunodficiency virus type 1 to infants of seropositive women in Zaire. N Engl J Med
28. Kuhn L, Abrams EJ, Matheson PB, Thomas PA, Lambert G, Bamji M, et al
. Timing of maternal-infant HIV transmission: associations between intrapartum factors and early polymerase chain reactions results. AIDS
29. Connor E, Sperling R, Gelber R, Kiselev P, Scott G, O'Sullivan MJ, et al
. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med
, 1994, 331
30. Shaffer N, Chauchoowong R, Mock PA, Bhadrakom C, Siriwasin W, Young NL, et al
. Short-course zidovudine for perinatal HIV-1 transmission in Bangkok, Thailand: a randomized controlled trial. Lancet
31. Van der 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
32. Ekpini ER, Wiktor SZ, Satten GA, Adjorlolo-Johnson GT, Sibailly TS, Ou C-Y, et al
. Late-postnatal mother-to-child transmission of HIV-1 in Abijan, Cote d'Ivoire. Lancet
33. Wiktor SZ, Ekpini ER, Nduati RW. Prevention of mother-to-child transmission of HIV-1 in Africa. AIDS
(Suppl B): S79–S87.
34. Miotti PG, Taha TE, Kumwenda NI, Broadhead R, Mtimavalye LA, Van der Hoeven L, et al
. HIV transmission through breastfeeding: a study in Malawi. JAMA
35. Nduati R, John G, Mbori-Ngacha D, Mbori-Ngacha D, Mwatha A, Ndinya-Achola J, et al
. Effect of breastfeeding and formula feeding on transmission of HIV: a randomized clinical trial. JAMA
36. Coutsidis A, Pillay K, Spooner E, Kuhn L, Coovadia HM, for the South African vitamin A Study Group. Influence of infant-feeding patterns on early mother-to-child transmission of HIV-1 in Durban, South Africa: a prospective cohort study. Lancet
Members of the ZVITAMBO Study Group, in addition to the named authors are: Henry Chidawanyika, Agnes Mahomva, Florence Majo, Edmore Marinda, Michael Mbizvo, Mary Ndhlovu, Robert Ntozini, Ellen Piwoz, Lidia Propper, Philipa Rambanepasi, Andrea Ruff, Naume Tavengwa, Claire Zunguza.