Persistence of vaccine-induced measles antibody beyond age 12 months: a comparison of response to one and two doses of Edmonston-Zagreb measles vaccine among HIV-infected and uninfected children in Malawi.

BACKGROUND: Previously, we demonstrated that measles antibody prevalence was lower at age 12 months among children infected with human immunodeficiency virus (HIV) than uninfected children following measles vaccination (MV) at ages 6 and 9 months. Among HIV-uninfected children, measles antibody prevalence was lower among 1- than 2-dose MV recipients. Here, we report results through age 24 months. METHODS: Children born to HIV-infected mothers received MV at 6 and 9 months, and children of HIV-uninfected mothers were randomized to MV at 6 and 9 months or MV at 9 months. We followed children through age 24 months. The child's HIV status was determined and measles immunoglobulin G (IgG) level was measured by enzyme immunoassay (EIA) and by plaque reduction neutralization (PRN) on a subset. RESULTS: Among HIV-uninfected children, the difference in measles antibody prevalence at age 12 months between one- and two-dose recipients reported previously by EIA was shown to be smaller by PRN. By age 24 months, 84% and 87% of HIV-uninfected children receiving 1 or 2 doses, respectively, were seroprotected. Only 41% of 22 HIV-infected children were measles seroprotected at age 20 months. DISCUSSION: Measles seroprotection persisted through age 24 months among HIV-uninfected children who received 1 or 2 doses of MV. HIV-infected children demonstrated seroprotection through age 12 months, but this was not sustained.

Epidemiology of rotavirus infection in children in Blantyre, Malawi, 1997-2007.

Acute gastroenteritis caused by rotavirus infection is an important cause of morbidity and mortality among infants and young children in Africa. From 1997 through 2007, we enrolled 3740 children <5 years of age with acute gastroenteritis who received hospital care at the Queen Elizabeth Central Hospital in Blantyre, Malawi. Group A rotavirus was detected in fecal specimens by enzyme immunoassay. Rotavirus strains were characterized for VP7 (G) and VP4 (P) types with use of reverse-transcription polymerase chain reaction. Overall, rotavirus was detected in one-third of children. The median age of children with rotavirus gastroenteritis was 7.8 months, compared with 10.9 months for those without rotavirus in stool specimens (P > .001). Rotavirus circulated throughout the year, with the detection proportion greatest during the dry season (from May through October). A total of 15 single rotavirus strain types were detected during the study period, with genotypes P[8]G1, P[6]G8, P[4]G8, P[6]G1, P[8]G3, and P[6]G9 comprising 83% of all strains characterized. Serotype G12 was detected for the first time in Blantyre during the final 2 years of study. Zoonotic transmission and viral reassortment contributed to the rich diversity of strains identified. Current rotavirus vaccines have the potential to greatly reduce the rotavirus disease burden in Malawi, but they will be required to protect against a broad range of rotavirus serotypes in a young population with year-round rotavirus exposure.

Supplemental measles vaccine antibody response among HIV-infected and -uninfected children in Malawi after 1- and 2-dose primary measles vaccination schedules.

BACKGROUND: The long-term antibody response to measles vaccine (MV) administered at age 6 months with or without subsequent doses is not well documented. METHODS: Measles serum antibody responses were evaluated after a supplemental dose of measles vaccine (sMV) administered at a median age of 20 months among Malawian children who had previously received 2 doses of measles vaccine (MV) at ages 6 and 9 months (HIV-infected and random sample of HIV-uninfected) or 1 dose at age 9 months (random sample of HIV-uninfected). We compared measles antibody seropositivity between groups by enzyme linked immunoassay and seroprotection by plaque reduction neutralization geometric mean concentrations. RESULTS: Of 1756 children enrolled, 887 (50.5%) received a sMV dose following MV at 9 months of age and had specimens available after sMV receipt, including 401 HIV-uninfected children who received one MV dose at 9 months, 464 HIV-uninfected and 22 HIV-infected children who received two doses of MV at ages 6 and 9 months. Among HIV-uninfected children, protective levels of antibody were found post sMV in 90-99% through ages 24-36 months and were not affected by MV schedule. Geometric mean concentration levels of measles antibody were significantly increased post-sMV among those HIV-uninfected children previously non-responsive to vaccination. Among HIV-infected children, the proportion seroprotected increased initially but by 9 months post-sMV was no higher than pre-sMV. CONCLUSIONS: Our findings support early 2-dose MV to provide measles immunity for young infants without risk of interference with antibody responses to subsequent MV doses administered as part of SIAs.

Short postexposure prophylaxis in newborn babies to reduce mother-to-child transmission of HIV-1: NVAZ randomised clinical trial.

BACKGROUND: In sub-Saharan Africa, most women present late for delivery with unknown HIV status, which limits the use of intrapartum nevirapine to prevent mother-to-child transmission of HIV. We aimed to determine whether post-exposure prophylaxis of nevirapine plus zidovudine given to babies only reduced transmission of HIV more than did a regimen of nevirapine alone. METHODS: We randomly assigned 1119 babies of Malawian women with HIV-1 who presented late (ie, within 2 h of expected delivery) to either nevirapine alone or nevirapine and zidovudine. Both drugs were given immediately after birth: one dose of nevirapine (2 mg/kg weight) was given as a single dose; babies in the nevirapine plus zidovudine group also received zidovudine twice daily for 1 week (4 mg/kg weight). Infant HIV infection was determined at birth and at 6-8 weeks. Primary outcome was HIV infection in babies at 6-8 weeks in those not infected at birth. Analysis was by intention to treat. FINDINGS: The overall rate of mother-to-child transmission at 6-8 weeks was 15.3% in 484 babies who received nevirapine and zidovudine and 20.9% in 468 babies who received nevirapine only (p=0.03). At 6-8 weeks, in babies who were HIV negative at birth, 34 (7.7%) babies who had nevirapine and zidovudine and 51 (12.1%) who received nevirapine only were infected (p=0.03)-a protective efficacy of 36%. This finding remained after controlling for maternal viral load and other factors at baseline. Adverse events were mild and of similar frequency in the two groups. INTERPRETATION: Postexposure prophylaxis can offer protection against HIV infection to babies of women who missed opportunities to be counselled and tested before or during pregnancy. The nevirapine and zidovudine regimen is safe and easy to implement.

Status of carotenoids, vitamin A, and vitamin E in the mother-infant dyad and anthropometric status of infants in Malawi.

This prospective study was carried out during February 2000-April 2003 to characterize the relationship between the status of carotenoids, vitamin E, and retinol and anthropometric status in apparently healthy infants and their mothers in Blantyre, Malawi. Anthropometric status of infants and concentrations of carotenoids (alpha-carotene, beta-carotene, beta-cryptoxanthin, lutein, zeaxanthin, and lycopene), retinol, and alpha-tocopherol in plasma were measured in 173 infants at 12 months of age, and concentrations of carotenoids, retinol, and a-tocopherol in plasma were measured in their mothers two weeks postpartum. In multivariate analyses, concentrations of retinol, total carotenoids, non-provitamin A carotenoids, and alpha-tocopherol in infants were associated with under-weight (p = 0.05). Concentrations of a-tocopherol were associated with wasting (p = 0.04). Concentrations in mothers and infants were all correlated (correlation coefficients from 0.230 to 0.502, p < 0.003). The findings suggest that poor status of carotenoids, retinol, and alpha-tocopherol in infants is associated with their poor anthropometric status, and status of carotenoids, retinol, and alpha-tocopherol in mothers and infants has a low-to-moderate association in the mother-infant dyad.

Effect of HIV-1 antiretroviral prophylaxis on hepatic and hematological parameters of African infants.

OBJECTIVE: To measure hepatic and hematological parameters among neonates randomized to receive ultra-short antiretroviral regimens. DESIGN: As part of an on-going clinical trial in Malawi, infants born to women who received (early presenters) or did not receive (late presenters) standard intrapartum nevirapine (NVP) dosing were randomized to receive orally either single dose NVP alone or NVP plus zidovudine (twice daily for 1 week). An additional group of untreated infants (born to HIV-uninfected women) was enrolled as a control. METHODS: Laboratory measurements were performed at birth and repeated at 6 weeks of age. Serum alanine aminotransferase (ALT) was measured on approximately 200 infants consecutively enrolled and randomized at the start of the trial. Complete blood count (CBC) was performed on approximately 800 infants at birth and 600 infants at 6 weeks of age. ALT and CBC were also determined on approximately 200 control infants. RESULTS: At birth there were no differences in ALT values between the groups of children. At 6 weeks of age, ALT levels were significantly higher among the treated groups compared with control group (geometric mean of 11.5 U/l for controls and 16.2-19.1 U/l for treated groups; P < 0.0001). Hematological parameters did not differ between groups at birth. At 6 weeks of age, levels of hemoglobin, hematocrit, granulocytes, and platelets were significantly (P < 0.0001) lower among antiviral drug-treated groups compared with controls. These changes were consistent with grade 1 (mild) toxicity, and were more noticeable among HIV-infected infants. CONCLUSIONS: Hepatic and hematologic abnormalities associated with short-term neonatal antiretrovirals among African children are minimal.

Antenatal vitamin A supplementation increases birth weight and decreases anemia among infants born to human immunodeficiency virus-infected women in Malawi.

Vitamin A is essential for immunity and growth. A controlled clinical that involved 697 human immunodeficiency virus (HIV)-infected pregnant women was conducted to determine whether vitamin A prevents anemia, low birth weight, growth failure, HIV transmission, and mortality. Women received daily doses of iron and folate, either alone or combined with vitamin A (3 mg retinol equivalent), from 18-28 weeks' gestation until delivery. In the vitamin A and control groups, respectively, the mean (+/-SE) birth weights were 2895+/-31 g and 2805+/-32 g (P=.05), the proportions of low-birth-weight infants were 14.0% and 21.1% (P=.03), the proportions of anemic infants at 6 weeks postpartum were 23.4% and 40.6% (P<.001), and the respective cumulative proportions of infants who were HIV infected at 6 weeks and 24 months of age were 26.6% and 27.8% (P=.76) and 27.7% and 32.8% (P=.21). Receipt of vitamin A improved birth weight and neonatal growth and reduced anemia, but it did not affect perinatal HIV transmission.

Zinc homeostasis in Malawian children consuming a high-phytate, maize-based diet.

BACKGROUND: Zinc deficiency in children is an important public health concern in the developing world, and the consumption of predominantly cereal-based diets with a high phytate content may contribute to the risk. The gastrointestinal tract plays a central role in absorbing and conserving zinc, yet it has not been carefully studied in such children. OBJECTIVE: This study investigated zinc homeostasis in healthy, free-living Malawian children with habitually high-phytate diets to better understand the role of the gastrointestinal tract. DESIGN: We evaluated zinc homeostasis in 10 children aged 2-5 y who were consuming a maize-based diet (phytate:zinc molar ratio of 23:1). Zinc stable isotopes were administered orally and intravenously. The tracer and tracee were measured in urine and feces. RESULTS: Endogenous fecal zinc was high in comparison with results for this measure in previous studies. Typical correlations seen in subjects consuming a low-phytate diet between total absorbed zinc, the size of the exchangeable zinc pool, and endogenous fecal zinc were not observed. Fractional absorption of zinc was 0.24. CONCLUSIONS: Zinc homeostasis was perturbed, particularly by large, endogenous fecal zinc losses, in this vulnerable population. The effects of interventions to improve zinc status, including dietary phytate reduction, on zinc homeostasis merit further study.

Urea production and leucine oxidation in malnourished children with and without acute infection.

We compared the kinetics of urea production and leucine oxidation in severely malnourished Malawian children. We tested the hypotheses that the rate of urea production was directly proportional to the rate of leucine oxidation and that the relationship between the two is altered by acute infection. Thirty-six marasmic children, aged 12 to 60 months, were enrolled; 26 had acute infection and 10 did not. The rates of urea and CO(2) production were estimated using primed, constant, intravenous stable isotope-labeled tracer infusions followed by intermittent sampling of breath and blood. The rate of urea production was greater in infected children when compared to uninfected children (169 +/- 85 v 105 +/- 44 micromol urea x kg(-1) x h(-1), P <.02). For children with and without infection, the rates of leucine oxidation and urea production were directly correlated (r = 0.49 and r = 0.74, respectively; P <.01), but the slopes of the regression lines were different. In uninfected children the degree of wasting was correlated with the rates of urea production and leucine oxidation (r = 0.67 and r = 0.48, respectively; P <.05). These data suggest that the rates of leucine oxidation and urea production are both measures of nitrogen catabolism, that acute infection alters the relationship between the two, and that less nitrogen is lost as urea in children with more wasting.

HIV infection among paediatric in-patients in Blantyre, Malawi.

To investigate the impact of HIV infection on hospital admission and death we studied children admitted to paediatric medical and surgical wards in Blantyre, Malawi, in March 2000. Unselected children whose parents or guardians consented to HIV testing of the child were recruited and HIV infection was determined by serology, with confirmation in children aged 15 months or less by PCR. We assessed the prevalence of HIV infection by age, clinical diagnosis and outcome of admission. Of 1064 admissions, 991 were tested for HIV infection, and 187 (18.9%) were infected. HIV was most common in children aged less than six months, 53 of 166 (32%). Parents of HIV-infected children were better educated, and more likely to have died, than those of uninfected children. Clinical symptoms and signs were not adequately sensitive or specific to be used for diagnosis of HIV. HIV was common in children with malnutrition (prevalence 40%), lower respiratory tract infection (29%) and sepsis (28%), and less prevalent among children with malaria (11%) or surgical admissions (11%). Almost 30% of HIV-infected children died, compared with 8.9% of uninfected children, and HIV-infected children constituted over 40% of in-patient deaths.

Nevirapine and zidovudine at birth to reduce perinatal transmission of HIV in an African setting: a randomized controlled trial.

CONTEXT: Antenatal counseling and human immunodeficiency virus (HIV) testing are not universal in Africa; thus, women often present in labor with unknown HIV status without receiving the HIVNET 012 nevirapine (NVP) regimen (a single oral dose of NVP to the mother at the start of labor and to the infant within 72 hours of birth). OBJECTIVE: To determine risk of mother-to-child transmission of HIV when either standard use of NVP alone or in combination with zidovudine (ZDV) was administered to infants of women tested at delivery. DESIGN, SETTING, AND PARTICIPANTS: A randomized, open-label, phase 3 trial conducted between April 1, 2000, and March 15, 2003, at 6 clinics in Blantyre, Malawi, Africa. The trial included all infants born to 894 women who were HIV positive, received NVP intrapartum, and were previously antiretroviral treatment-naive. Infants were randomly assigned to NVP (n = 448) and NVP plus ZDV (n = 446). Infants were enrolled at birth, observed at 6 to 8 weeks, and followed up through 3 to 18 months. The HIV status of 90% of all infants was established at 6 to 8 weeks. INTERVENTION: Mothers received a 200-mg single oral dose of NVP intrapartum and infants received either 2-mg/kg oral dose of NVP or NVP (same dose) plus 4 mg/kg of ZDV twice per day for a week. MAIN OUTCOME MEASURES: HIV infection of infant at birth and 6 to 8 weeks, and adverse events. RESULTS: The mother-to-child transmission of HIV at birth was 8.1% (36/445) in infants administered NVP only and 10.1% (45/444) in those administered NVP plus ZDV (P =.30). A life table estimate of transmission at 6 to 8 weeks was 14.1% (95% confidence interval [CI], 10.7%-17.4%) in infants who received NVP and 16.3% (95% CI, 12.7%-19.8%) in those who received NVP plus ZDV (P =.36). For infants not infected at birth and retested at 6 to 8 weeks, transmission was 6.5% (23/353) in those who received NVP only and 6.9% (25/363) in those who received NVP plus ZDV (P =.88). Almost all infants (99%-100%) were breastfed at 1 week and 6 to 8 weeks. Grades 3 and 4 adverse events were comparable; 4.9% (22/448) and 5.4% (24/446) in infants receiving NVP only and NVP plus ZDV, respectively (P =.76). CONCLUSIONS: The frequency of mother-to-child HIV transmission at 6 to 8 weeks in our 2 study groups was comparable with that observed for other perinatal HIV intervention studies among breastfeeding women in Africa. The safety of the regimen containing neonatal ZDV was similar to that of a standard NVP regimen.

The effect of human immunodeficiency virus and breastfeeding on the nutritional status of African children.

BACKGROUND: The risk of HIV-1 infection is high among breast-fed children in sub-Saharan Africa. Monitoring the nutritional status can provide useful information to determine the effect of HIV infection and breast-feeding on child growth and development. We longitudinally assessed the nutritional status and determined its association with HIV infection and breast-feeding among Malawian children. METHODS: We analyzed data from 2 clinical trials to prevent mother-to-child transmission of HIV in Malawi. These trials were conducted during 2000-2003 before the current guidelines were implemented to breast-feed exclusively during the first 6 months and wean thereafter. The nutritional status of children was measured up to age 24 months, using z-scores. Age-specific differences in length-for-age (L/A), weight-for-age (W/A), and weight-for-length (W/L) were compared stratifying by gender and HIV infection status. Multivariable models examined the mean change in z-scores controlling for breast-feeding and other factors. RESULTS: In this analysis, 1589 children were included. Boys had significantly lower L/A scores and became stunted (z-score -<2 standard deviations) earlier than girls. HIV-infected children had significantly lower mean L/A and W/A z-scores than HIV-uninfected children and became stunted and underweight at an earlier age. In multivariable analysis not being breast-fed and being HIV infected were significantly (P < 0.001) associated with decreases in mean L/A, W/A, and W/L z-scores. CONCLUSIONS: This study shows the impact of infant HIV infection on growth and supports the critical importance of breast-feeding. Mother-to-child transmission of HIV programs should endeavor to preserve breast-feeding and find alternative measures to prevent postnatal HIV transmission.

The role of transplacental microtransfusions of maternal lymphocytes in HIV transmission to newborns.

BACKGROUND: Perinatal HIV transmission could occur via microtransfused maternal blood during delivery. If so, detecting maternal cells in umbilical cord blood should correlate with infection risk. OBJECTIVE: To develop sensitive assays for maternal DNA in infant's blood stored as dried blood spots (DBS) and examine the correlation between microtransfusion and perinatal HIV infection risk. METHODS: Blood-in-blood serial dilutions were prepared as DBS. Extracted DNA was amplified for unique minor-population sequences using 24 allele-specific polymerase chain reaction assays. Using newborns born to HIV+ mothers, paired mother-infant samples were similarly examined to identify unique maternal sequences targeted by allele-specific polymerase chain reaction of DNA extracted from cord blood DBS. Cord-blood PCR-negative infants were categorized as uninfected or perinatally infected by HIV PCR on samples collected 4-8 weeks after birth. RESULTS: Sequences from added cells were detected at less than 1: 1000 dilutions in 19 of 20 aliquots, and less than 1: 10 000 dilutions in seven of 20 aliquots; the median limit of detection (probit analysis) was one added genomic sequence in 9500 background sequences of amplifiable DNA. Maternal sequences were detected in cord-blood DBS of 50% of infected infants (N = 18) and 44% of uninfected infants (N = 43). Infection did not correlate with more frequent detection of maternal sequences. CONCLUSION: This semiquantitative assay reliably detected maternal DNA sequences in DBS at levels of less than 1: 1000 cells. Maternal sequences were frequently detected but did not correlate infection risk with detection or level of maternal DNA in umbilical cord blood. Therefore, we could not demonstrate that microtransfusions at parturition were responsible for perinatal HIV transmission.

Evaluation of the immune response to a 2-dose measles vaccination schedule administered at 6 and 9 months of age to HIV-infected and HIV-uninfected children in Malawi.

BACKGROUND: The World Health Organization recommends that infants at high risk for developing measles before 9 months of age, including human immunodeficiency virus (HIV)-infected infants, receive measles vaccination (MV) at 6 and 9 months of age. METHODS: Children born to HIV-infected mothers received MV at 6 and 9 months, and children of HIV-uninfected mothers were randomized to receive MV at 6 and 9 months, MV at 9 months, or routine MV without follow-up. Blood samples were obtained before and 3 months after each MV. Data were collected on adverse events for 21 days after each MV, at all clinic visits, on any hospitalization, and for subjects who died. HIV-infection status was determined by antibody assays and polymerase chain reaction; the presence of measles IgG was determined by EIA. RESULTS: Twenty-two hundred mother-infant pairs were enrolled. After the first and second doses of measles vaccine, respectively, the percentages of children who were measles seropositive were 59% (36 of 61) and 64% (29 of 45) among HIV-infected children, 68% (152 of 223) and 94% (189 of 202) among HIV-exposed but uninfected children, and 62% (288 of 467) and 92% (385 of 417) among HIV-unexposed children. Of 521 HIV-unexposed children vaccinated only at 9 months, 398 (76%) were measles seropositive at 12 months. No serious vaccine-related adverse events were identified. CONCLUSIONS: An early, 2-dose MV schedule was immunogenic, but a higher proportion of HIV-infected children remained susceptible to measles, compared with HIV-uninfected children (whether HIV exposed or HIV unexposed).

Late postnatal transmission of HIV-1 and associated factors.

BACKGROUND: The present study was undertaken to determine the risk and timing of late postnatal transmission (LPT) of human immunodeficiency virus type 1 (HIV-1). METHODS: Breast-fed infants previously enrolled in 2 trials of antiretroviral prophylaxis were monitored in Malawi. Kaplan-Meier and proportional hazard models assessed cumulative incidence and association of factors with LPT. RESULTS: Overall, 98 infants were HIV infected, and 1158 were uninfected. The cumulative risk of LPT at age 24 months was 9.68% (95% confidence interval, 7.80%-11.56%). The interval hazards at 1.5-6, 6-12, 12-18, and 18-24 months were 1.22%, 4.05%, 3.48%, and 1.27%, respectively. CONCLUSIONS: The risk of LPT beyond 6 months is substantial. Weaning at 6 months could prevent >85% of LPT.

Higher in utero and perinatal HIV infection risk in girls than boys.

OBJECTIVE: This study analyzed mother-to-child HIV transmission rates by sex and exposure time for babies born to HIV-infected, untreated African women. METHODS: Data were analyzed from 2 independent studies done in Malawi during the 1990s. Infections were established by polymerase chain reaction on blood samples. Odds ratios (ORs) for transmission were examined by period at risk: in utero (infected in umbilical cord blood), perinatal (infected in 1st postnatal blood > or =4 weeks), and postnatal (later postnatal infection). RESULTS: Among 1394 singleton births, girls were more likely to become infected than boys. For in utero transmission, the OR was 1.4 (95% CI: 0.9 to 2.2). For transmission during early life (umbilical cord blood not available) the OR was 2.7 (95% CI: 1.5 to 4.9). However, transmission risks in the perinatal and postnatal infection periods did not differ in boys and girls. Among 303 tested twin-birth pairs, girls were at higher risk than boys for in utero (OR: 2.6; 95% CI: 1.2 to 5.8) and perinatal (OR: 1.9; 95% CI: 1.0 to 3.7) infection. Recognized mother-to-child transmission risk factors did not explain the higher risk of infection in girls. CONCLUSIONS: Girls were at higher risk of early (in utero and perinatal) HIV infection than boys. It is proposed that minor histocompatibility reactions between maternal lymphocytes and infant Y chromosome-derived antigens reduce the risk of HIV transmission in boys.

The impact of breastfeeding on the health of HIV-positive mothers and their children in sub-Saharan Africa.

OBJECTIVE: We assessed the impact of breastfeeding by women infected with human immunodeficiency virus (HIV)-1 on their morbidity and risk of mortality and on the mortality of their children. METHODS: We analysed longitudinal data from two previous randomized clinical trials of mother-to-child transmission of HIV conducted between April 2000 and March 2003 in the Republic of Malawi, Africa. Mothers infected with HIV, and their newborns, were enrolled at the time of their child's birth; they then returned for follow-up visits when the child was aged 1 week, 6-8 weeks and then 3, 6, 9, 15, 18, 21 and 24 months. Patterns of breastfeeding (classified as exclusive, mixed or no breastfeeding), maternal morbidity and mortality, and mortality among their children were assessed at each visit. Descriptive and multivariate analyses were performed to determine the association between breastfeeding and maternal and infant outcomes. FINDINGS: A total of 2000 women infected with HIV were enrolled in the original studies. During the 2 years after birth, 44 (2.2%) mothers and 310 (15.5%) children died. (Multiple births were excluded.) The median duration of breastfeeding was 18 months (interquartile range (IQR)=9.0-22.5), exclusive breastfeeding 2 months (IQR=2-3) and mixed feeding 12 months (IQR=6-18). Breastfeeding patterns were not significantly associated with maternal mortality or morbidity after adjusting for maternal viral load and other covariates. Breastfeeding was associated with reduced mortality among infants and children: the adjusted hazard ratio for overall breastfeeding was 0.44 (95% confidence interval (CI)=0.28-0.70), for mixed feeding 0.45 (95% CI=0.28-0.71) and for exclusive breastfeeding 0.40 (95% CI=0.22-0.72). These protective effects were seen both in infants who were infected with HIV and those who were not. CONCLUSION: Breastfeeding by women infected with HIV was not associated with mortality or morbidity; it was associated with highly significant reductions in mortality among their children.