Predictors of TB disease in HIV-exposed children from Southern Africa.

BACKGROUND: P1041 was a randomised, placebo-controlled isoniazid prophylaxis trial in South Africa. We studied predictors for TB in HIV-exposed children participating in the P1041 trial.METHODS: We included data from entry until Week 108. Predictors considered were type of housing, overcrowding, age, sex, ethnicity, tobacco exposure, weight-for-age percentile Z-score (WAZ), CD4%, viral load (VL), antiretroviral therapy (ART) and number of household smokers.RESULTS: Of 543 HIV-positive (HIV+) and 808 HIV-exposed uninfected (HEU) infants at entry, median age was 96 days (interquartile range: 92-105). Of 1,351 caregivers, 125 (9%) had a smoking history, and 62/1,351 reported current smoking. In 594/1,351 (44%) households, there was at least one smoker. Smoking caregivers consumed 1-5 cigarettes daily. In the HIV+ cohort, significant baseline TB predictors after adjusting covariates were as follows: WAZ (adjusted hazard ratio [aHR] 0.76, P = 0.002) and log10 HIV RNA copies/ml (aHR 1.50, P = 0.009). Higher CD4% (aHR 0.88, P = 0.002) and ART (aHR 0.50, P = 0.006) were protective. In the HEU cohort, smoking exposure was associated with reduced TB-free survival on univariate analysis, but not after adjustment in the multivariate model.CONCLUSION: Low WAZ and high VL were strong predictors of TB disease or death. Rising CD4 percentage and being on ART were protective in the HIV+ cohort.

Tuberculosis in HIV-infected South African children with complicated severe acute malnutrition.

SETTING: Academic tertiary referral hospital in Durban, South Africa. OBJECTIVE: To describe the incidence and diagnostic challenges of tuberculosis (TB) in human immunodeficiency virus (HIV) infected children with severe acute malnutrition (SAM). DESIGN: Post-hoc analysis of a randomised controlled trial that enrolled antiretroviral therapy naïve, HIV-infected children with SAM. Trial records and hospital laboratory results were explored for clinical diagnoses and bacteriologically confirmed cases of TB. Negative binomial regression was used to explore associations with confirmed cases of TB, excluding cases where the clinical diagnosis was not supported by microbiological confirmation. RESULTS: Of 82 children enrolled in the study, 21 (25.6%) were diagnosed with TB, with bacteriological confirmation in 8 cases. Sputum sampling (as opposed to gastric washings) was associated with an increased risk of subsequent diagnosis of TB (adjusted relative risk [aRR] 1.134, 95%CI 1.02-1.26). Culture-proven bacterial infection during admission was associated with a reduced risk of TB (aRR 0.856, 95%CI 0.748-0.979), which may reflect false-negative microbiological tests secondary to empiric broad-spectrum antibiotics. CONCLUSION: TB is common in HIV-infected children with SAM. While microbiological confirmation of the diagnosis is feasible, empiric treatment remains common, possibly influenced by suboptimal testing and false-negative TB diagnostics. Rigorous microbiological TB investigation should be integrated into the programmatic management of HIV and SAM.

Use of antiretrovirals in HIV-infected children in a tuberculosis prevention trial: IMPAACT P1041.

SETTING: International Maternal Pediatric Adolescent AIDS Clinical Trials (IMPAACT) P1041, a tuberculosis (TB) prevention trial conducted among children enrolled from 2004 to 2008 during South Africa's roll-out of combination antiretroviral therapy (ART). OBJECTIVE: To estimate TB incidence and mortality and the effect of ART. DESIGN: Children were pre-screened to exclude TB disease and exposure, actively screened 3-monthly for TB exposure and symptoms, and provided post-exposure isoniazid prophylaxis therapy (IPT). TB diagnoses were definite, probable, or possible, and mortality all-cause. Testing was at the 5% significance level. RESULTS: In 539 children (aged 3-4 months) followed up for a median of 74 weeks (interquartile range [IQR] 48-116), incidence/100 person-years (py) was 10.67 (95%CI 8.47-13.26) for any TB and 2.89 (95%CI 1.85-4.31) for definite/probable TB. Any TB incidence was respectively 9.39, 13.59, and 9.83/100 py before, <180 days after, and 180 days after ART initiation. Adjusted analysis showed a non-significant increase in any TB (HR 1.32, 95%CI 0.71-2.52, P = 0.38) and a significant reduction in mortality (HR 0.39, 95%CI 0.17-0.82, P = 0.017) following ART initiation. CONCLUSIONS: ART reduced mortality but not TB incidence in human immunodeficiency virus (HIV) infected children in IMPAACT P1041, possibly attributable to active screening for TB exposure and symptoms with post-exposure IPT. Research into this as a strategy for TB prevention in high HIV-TB burden settings may be warranted.

Source case identification in HIV-exposed infants and tuberculosis diagnosis in an isoniazid prevention study.

BACKGROUND: Identifying source cases of children exposed to tuberculosis (TB) is challenging. We examined the time-point of obtaining contact information of TB source cases in human immunodeficiency virus (HIV) infected and HIV-exposed uninfected (HEU) children in a randomised, placebo-controlled trial of pre-exposure to isoniazid prophylaxis. METHODS: A total of 543 HIV-infected and 808 HEU infants without TB exposure aged 3-4 months were enrolled between 2004 and 2008. At 3-monthly follow-up, infants were evaluated for TB and care givers were asked about new TB exposure. RESULTS: In total, 128 cases of TB disease and 40 deaths were recorded among 19% (105/543) of the HIV-infected and 8% (63/808) of the HEU children; 229 TB contact occasions were reported in 205/1351 (15%) children, of which 83% (189/229) were in the household. Of the 189 household contacts, 108 (53%) underwent microbiological evaluations; 81% (87/108) were positive. HIV-infected and HEU infants had similar frequencies of TB contact: in 48% of infants with definite TB, 58% with probable TB and 43% with possible TB. Of 128 children diagnosed with TB, a TB contact was identified for 59. Of these, 29/59 (49%) were identified at or after the child's TB diagnosis. CONCLUSION: TB source cases are often identified at or after a child's TB diagnosis. More effort is required for earlier detection.

Loss to follow-up among infants in a study of isoniazid prophylaxis (P1041) in South Africa.

OBJECTIVE: To assess risk factors for loss to follow-up (LFU) from the IMPAACT P1041 study, an isoniazid (INH) prophylaxis study conducted in southern Africa. DESIGN: Infants in two cohorts, human immunodeficiency virus-infected (HIV+) and HIV-exposed but non-infected (HIV-), were randomized to INH or placebo for 96 weeks. LFU was evaluated at week 96. RESULTS: Of 1351 infants, 12.9% were LFU (10.4% HIV+, 14.7% HIV-); 65% of the HIV+ cohort was asymptomatic. Among HIV+ infants, large household size (>6 vs. <4 members, P = 0.035) and presence of an elder (≥55 years, P = 0.05) were associated with better retention. Although attenuated in adjusted analysis, these associations held among HIV- infants. Among HIV- infants, having a younger mother increased the risk (P = 0.008) and maternal history of TB reduced the risk of LFU, the latter by nearly 70% (P = 0.048 univariate, 0.09 adjusted). LFU was largely due to inability to contact the participant (58% HIV+, 30% HIV-), and inability to attend the clinic and withdrawal of consent (HIV-). CONCLUSIONS: Household support was an important factor in participant retention, particularly for the non-HIV-infected cohort, as young maternal age was a risk factor for LFU. Retaining study participants from this mobile population can be challenging and may warrant additional support.

Expression of HLA-G1 at the placental interface of HIV-1 infected pregnant women and vertical transmission of HIV.

OBJECTIVES: The ability of Human Leucocyte Antigen-G (HLA-G) to inhibit the cytolytic effect to immunocompetent cell types, suggests that HLA-G has an immunomodulatory role. In view of this concept the objective of the study was to assess whether the Major Histocompatibility Complex -coded molecule HLA-G mRNA is a risk factor at the placental barrier in HIV-1 positive pregnant women. DESIGN: Placental HLA-G1 levels in HIV-1 infected mothers and viral loads in both mothers and their babies were performed on fifty-five participants. METHODS: Synthesis of complementary deoxyribose nucleic acid (cDNA) was performed using ribose nucleic acid (RNA) extracted from placental tissue samples. Amplification of cDNA using specifically designed primers complementary to the full length HLA-G1 isoform was quantified using real time-polymerase chain reaction (RT-PCR). Viral load assays (Amplicor Version 1.5, Roche Diagnostics) were performed on all plasma samples. RESULTS: HLA-G1 primers detected the full length isoform HLA-G1 PCR product at 86.5 °C. Logistic regression calculations indicated that the risk of babies becoming infected increased by 1.3 with every 1 unit increase in HLA-G1 expression. Female babies were 3.7 times more likely to become infected than male. There was a positive correlation between mothers' log viral load and transmission of infection to the baby (p = 0.047; 95%CI 1.029-11.499). CONCLUSION: Maternal viral load was a strong predictor of viral transmission. Placental HLA-G1 expression was up-regulated 3.95 times more in placentas of HIV-1 infected mothers with infected babies when compared to uninfected babies.

HIV-Associated Tuberculosis in the Newborn and Young Infant.

Each year, approximately 250 000 women die during pregnancy, delivery, or postpartum. Maternal mortality rates due to tuberculosis (TB) and HIV in Sub-Saharan Africa now supersede obstetric-related causes of mortality. The majority of cases occur in population-dense regions of Africa and Asia where TB is endemic. The vertical transmission rate of tuberculosis is 15%, the overall vertical transmission rate of HIV in resource-limited settings with mono- or dual-ARV therapy varies from 1.9% to 10.7%. If the millennium development goals are to be achieved, both HIV and TB must be prevented. The essential aspect of TB prevention and detection in the newborn is the maternal history and a positive HIV status in the mother. Perinatal outcomes are guarded even with treatment of both diseases. Exclusive breast feeding is recommended. The community and social impact are crippling. The social issues aggravate the prognosis of these two diseases.

The impact of human immunodeficency virus 1 on largyngeal airway obstruction in children.

Children with laryngeal airway obstruction (LAO) require admission to the intensive care unit (ICU). The unresolved ethical dilemma of ICU access for HIV infected children in resource poor settings requires further scientific data to help guide triaging. Of 38 children with LAO, 19 had HIV infection. The mortality, need for supportive management, duration of intubation, intermittent positive pressure ventilation, and ICU and hospitalisation stay were similar in the HIV infected group compared to the HIV uninfected group. Episodes of laryngotracheobronchitis were equally distributed between both groups (31.6% v 31.3%), while oropharyngeal/laryngeal candidiasis (26.3%), tuberculosis (15.8%), and benign lymphoid hyperplasia (15.8%) were commonly seen in the HIV infected group.

Growth in early childhood in a cohort of children born to HIV-1-infected women from Durban, South Africa.

This study describes growth in a cohort of black South African children born to HIV-1-infected women in Durban. Children born to HIV-1-seropositive women were followed up from birth to early childhood. At birth and at each visit, growth parameters were measured. Mean Z-scores were calculated for weight-for-length, weight-for-age and length-for-age and, if they were low, the children were regarded as wasted, malnourished or stunted, respectively. At the end of the study, there were 48 infected and 93 uninfected children. There were no significant differences between the two groups at birth. Thereafter, the infected group was found to have early and sustained low mean Z-scores for length-for-age and weight-for-age but not for weight-for-length. The means reached significance at ages 3, 6 and 12 months for length and at 3, 6 and 9 months for weight. Infected children who died early had more severe stunting, wasting and malnutrition than infected children who survived. Infected children born to HIV-positive women have early and sustained stunting and are malnourished but not wasted. Children with rapidly progressive disease have both stunting and wasting and are more severely affected. Early nutritional intervention might help prevent early progression or death in HIV-infected children, particularly in developing countries without access to anti-retroviral therapy in state hospitals.

Mortality in a cohort of children born to HIV-1 infected women from Durban, South Africa.

OBJECTIVES: To describe mortality in a cohort of infants with vertically transmitted HIV-1 infection. PATIENTS AND METHODS: Children of HIV-1 infected women were followed up from birth and a record was made at each visit of growth, development and all illnesses. Details surrounding death were obtained from hospital records. RESULTS: The final cohort comprised 48 infected and 93 uninfected children; there were 25 deaths, 17 of which (35%) were regarded as being HIV-related. The mean age at death of HIV-related cases was 10.1 months (range 1-48 months), with 83% of HIV-related deaths occurring before the age of 10 months. The commonest diagnoses at the time of death were diarrhoea, pneumonia, failure to thrive and severe thrush. These findings, together with neurological abnormalities, often presaged rapid deterioration and death. CONCLUSIONS: Mortality among children with vertically acquired HIV infection is high in the first year of life. Death in these subjects was due to the common causes of morbidity and mortality among all children in developing countries. A combination of diarrhoea, pneumonia, failure to thrive, and neurological abnormalities should alert one to the possibility of rapidly progressive disease and death.

The early natural history of vertically transmitted HIV-1 infection in African children from Durban, South Africa.

Forty-eight children with vertically transmitted HIV-1 infection and 93 uninfected infants were followed up at regular intervals from birth for a mean of 26 months. They were examined physically, growth and development were assessed and illnesses recorded. Seventy per cent of infected infants were symptomatic by 6 months. Relative risks in the infected infants were highest for lymphadenopathy (4.56; CI 2.7-7.7), failure to thrive (4.48; 2.57-7.81), and neurological abnormalities (3.32; 1.9-5.58). The most frequent findings were diarrhoea (78%), pneumonia (76%) and lymphadenopathy (70%). Thrush and pneumonia occurred early but declined over time, whereas diarrhoea and neurological abnormalities occurred later and increased in frequency. A diagnosis of AIDS was made in 44% of infected infants by 12 months of age. Mortality in infected infants was 35.4%, and 76% of deaths occurred within the 1st year. About two-thirds of HIV-infected infants survived into early childhood. In South African children with vertically acquired HIV-1 infection the onset of disease is early and deterioration to AIDS and death are rapid. Infected infants can be easily recognized clinically, the majority by 6 months of age.

PIP: 48 children with vertically acquired HIV-1 infection and 93 uninfected infants were followed up at regular intervals from birth for a mean period of 26 months. They were examined physically, had their growth and development assessed, and illnesses recorded. 70% of HIV-infected infants were symptomatic by age 6 months. Relative risks of specific signs of disease in the infected infants were highest for lymphadenopathy, failure to thrive, and neurological abnormalities. The most frequent findings were diarrhea (78%), pneumonia (76%), and lymphadenopathy (70%). Thrush and pneumonia occurred early, but declined over time, while diarrhea and neurological abnormalities occurred later and increased in frequency. AIDS was diagnosed in 44% of infected infants by age 12 months. HIV mortality in infected infants was 35.4%, with 76% of deaths occurring during the first year of life. About two-thirds of HIV-infected infants survived into early childhood.

Breastfeeding by HIV-1-infected women and outcome in their infants: a cohort study from Durban, South Africa.

BACKGROUND: Women in developing countries have the difficult choice of balancing the risk of transmitting HIV through breast milk against the substantial benefits of breastfeeding. It is not known, however, whether the benefits of breastfeeding are the same when the mother is HIV-infected. Therefore, we examined the impact of breastfeeding on infections, growth and mortality in the infants of HIV-1-infected women. METHODS: Infants of HIV-1-positive women were followed from birth and at each visit they were examined, growth parameters were recorded and notes were made of feeding method, and of current and interim illnesses. RESULTS: Of the 43 HIV-infected and 90 non-infected infants for whom feeding data were available, 36 infants (27%) were exclusively breastfed, 76 (57%) received mixed feeding, and 21 (16%) received formula only. The HIV transmission rate was 39% in those exclusively breastfed, 24% in those fed exclusively on formula and 32% in those receiving mixed feeding [relative risk (RR), 7.39; 95% confidence interval (CI), 1.67-32.6 between the exclusive breast and formula only groups]. There was a stepwise increase in the transmission rate with duration of exclusive breastfeeding of 1, 2 and 3 months (45%, 64%, and 75%, respectively). Of the infected infants, seven (50%) exclusively breastfed, 13 (51%) of those on mixed feeds and none on formula only developed AIDS; exclusively breastfed infants had a slower rate of progression to AIDS (mean age, 7.5 months versus 5.0 months, P = 0.2242) than those on mixed feeds. Mortality (which occurred in the infected infants only) was 19% in the exclusively breastfed infants; 13% in those on mixed feeds and 0% in those exclusively formula-fed. The frequency of failure to thrive and episodes of diarrhoea and pneumonia were not significantly different between the three groups in both the infected and non-infected infants. CONCLUSIONS: Exclusive breastfeeding by HIV-infected women does not appear to protect their infants against common childhood illnesses and failure to thrive, nor does it significantly delay progression to AIDS. The implication of the trend towards differential mortality rates according to feeding groups is uncertain and requires further investigation.

PIP: To determine whether breast feeding confers the same protective effects on child health and survival when the mother is infected with HIV, a prospective study was conducted of a cohort of 133 infants born in Durban, KwaZulu/Natal, South Africa, in 1990-93 and followed for up to 24 months. 36 infants (27%) were exclusively breast-fed, 76 (57%) received mixed feeds, and 21 (16%) received formula only. By 15 months of age, 43 infants were antibody-positive or had died from an HIV-related cause. The HIV transmission rate was 39% in those exclusively breast-fed, 24% in those fed formula only, and 32% in infants receiving mixed feeds. The relative risk of HIV transmission in exclusively breast-fed compared with entirely formula-fed infants was 7.39 (95% confidence interval, 1.67-32.6). The HIV transmission rate was 45% after 1 month of breast feeding, 64% after 2 months, and 75% after 3 months. Among HIV-infected infants, seven (50%) of those exclusively breast-fed, 13 (51%) of those on mixed feeds, and none on formula developed AIDS. However, exclusively breast-fed babies had, on average, a slightly slower rate of progression to AIDS (7.5 months) than those receiving both breast milk and formula (5.0 months). Mortality was 19% in exclusively breast-fed infants and 13% in those on mixed feeds; no infants in the formula-fed only group died. The frequencies of diarrhea, pneumonia, and failure to thrive did not differ between infected or non-infected infants in the three groups. These unexpected findings suggest that exclusive breast feeding by HIV-infected women does not protect infants against childhood illnesses or significantly delay progression to AIDS.

Lymphocyte subset changes between 3 and 15 months of age in infants born to HIV-seropositive women in South Africa.

The evolution of T-lymphocyte subsets during infancy in perinatally HIV-infected African babies has not been previously described. In a hospital-based cohort study, T-lymphocyte subset changes were investigated in 72 South African black children born to HIV seropositive mothers. Sixteen (22.2%; children were classified as infected and 56 (77.8%) as uninfected by 18 months of age. Four (25%) of the infected infants died before the age of 9 months from HIV-related disease. The CD4 and CD8 T-lymphocyte subsets, expressed in absolute numbers, as percentages, percentiles or as ratios, were clear indicators of HIV infection at all ages between 3 and 15 months. The most marked changes were a decreased percentage of CD4 cells and an increase in percentage of CD8 cells in the infected group. In the 4 infected infants who died, CD8 count and CD4:CD8 ratio clearly predicted poor clinical outcome at 3 months. Taken together, both CD4:CD8 ratio and CD4 percentage are reliable markers of HIV infection in an African paediatric population; however, a raised CD8 lymphocyte count rather than a CD4 count is a more specific prognostic marker of disease progression in HIV infected children.

The relationship between maternal-infant antibody levels and vertical transmission of HIV-1 infection.

This study assesses the predictive value of the ratio of HIV-1 antibodies in the newborn at birth to that in the mother for perinatally transmitted infection confirmed subsequently by age 18 months. The ratio of HIV-1 (EIA) antibody levels in the baby at birth to that in the seropositive mother after the first trimester (sequenstration index SI) was available in 114 of a perinatal cohort of 137 infants. We related this ratio to the HIV infection status of the children by 18 months, HIV-1 DNA PCR and HIV-specific IgA antibody detection at birth, between 3 and 6 months, and morbidity and mortality. Thirty-five of the 137 (26 per cent) children were diagnosed as infected by 18 months. The mean (SD) HIV SI was 1.57 (0.88) in 29 infected and 0.83 (0.42) in 85 uninfected infants (P < 0.0001). Sensitivity and specificity of a threshold SI of 1.27 (mean +/- 2 SD of uninfected group) for the prediction of perinatal HIV-1 infection were 41 and 98 per cent, respectively. The reason for the higher SI in the infected babies is the combination of lower antibody titres in the transmitting mothers with raised levels in the infected babies. A similar analysis of antibody ratios showed no statistical differences for measles and tetanus (P > 0.1) between HIV infected and uninfected groups. There was a tendency to increased morbidity (Pearson's correlation coefficient r = 0.31) and more severe disease in those with higher HIV-1 SI. Three of 17 (18 per cent) peripheral blood samples from infected children at birth were PCR positive; all had SI's above the threshold. Overall sensitivity and specificity of PCR were 85 per cent each. Eleven of the 29 infected children were HIV-1 specific IgA positive at birth; six (64 per cent) of these had an SI > 1.27. This simple SI of HIV-1 EIA antibodies at birth is comparable to elaborate techniques in its power to predict perinatally acquired infection. It may be a cheap, reliable and rapid screening test for vertically transmitted HIV-1 infection.

HIV-1 specific immunoglobulin A antibodies as an effective marker of perinatal infection in developing countries.

HIV-1 specific IgA antibody testing using commercially available reagents was evaluated at birth to 15 months in a group of infants born to HIV-seropositive South African women. Following IgG depletion of serum samples, 33/35 (94 per cent) of the infected infants and 3/99 (3 per cent) of the uninfected infants showed positive IgA reactivity. Sensitivity at birth was 24 per cent and improved with age; 82 per cent at 3 months, 87 per cent at 6 months and 94 per cent at 12 months. The overall positive and negative predictive values were 92 and 98 per cent, respectively. An evaluation of IgA and PCR in a subsample of infants indicated a better sensitivity of PCR within 3 months of birth, but IgA detection offered a higher overall sensitivity (87 v. 83 per cent) and specificity (91 v. 85 per cent). No significant difference in IgA level was observed between transmitting mothers and non-transmitting mothers. A moderate correlation existed between IgA level in the infant and the cumulative morbidity score, however a stronger association was observed between high IgA levels in the infected infant and rapid disease progression. The viral specific IgA assay is a simple, reliable and cost-effective diagnostic and prognostic test for perinatal HIV infection in developing countries.

Age-related pattern of immunoglobulins G, A and M in children born to HIV-seropositive women.

In a cohort of 56 children born to HIV-seropositive African women, 19 met the criteria for HIV-infected children and 37 remained antibody-negative at 18 months of age. Blood samples taken at birth and 3-monthly until 18 months of age were processed and analysed by laser nephelometry for serum immunoglobulin (IgG, IgA and IgM) levels. In the infected group of children. higher levels of IgG were observed during their 1st 18 months of life reaching statistical significance at 3, 6, 15 and 18 months. Higher levels of IgA at 3 months and at 15 and 18 months, and higher levels of IgM at 3 months and 18 months later were statistically significant. All four infected children who died before the age of 6 months showed signs of hypergammaglobulinaemia (IgG and IgA) by 3 months of age. In this study the earliest and most common immunological abnormality was hypergammaglobulinaemia and infected infants with higher morbidity and mortality had more evident immunoglobulin abnormalities than infected children who survived. However, the immunological abnormalities in this small cohort did not precede the onset of severe symptoms and cannot therefore be used to predict clinical outcome.

Use of HIV-1 specific immunoglobulin G3 as a serological marker of vertical transmission.

The objective of the study was to indicate HIV infection in infants. The patients were part of a longitudinal cohort of 43 infants born to HIV seropositive mothers. A modified Genelavia EIA primarily directed against HIV envelope proteins was used. An alkaline phosphatase labelled IgG3 conjugate was substituted in place of the kit conjugate. HIV specific IgG3 clearance was optimal at 6 months, whilst HIV total antibody was reliable only from age 12 months onwards. At 6 months no detectable IgG3 were found in 91 per cent of uninfected infants where more of these infants had reduced their total HIV antibody titres at the same period. We confirm that HIV specific IgG3 measurement is a reliable and cost effective means of identifying HIV infected infants from 6 months of age onwards.

Determinants of mother-to-child transmission of human immunodeficiency virus type 1 infection in a cohort from Durban, South Africa.

OBJECTIVES: To determine the vertical transmission rate of HIV-1 infection and to assess the influence of maternal risk factors on transmission in infants born to HIV-1-infected black women in Durban. DESIGN: A prospective, hospital-based cohort study conducted at King Edward VIII hospital, Durban. HIV-1-seropositive women were enrolled into the study, and their infants were followed up at regular intervals from birth to early childhood. The infection status of the children was classified and the transmission rate was computed according to the recommendations of the workshop held in Ghent, Belgium (1992). RESULTS: The final cohort of 181 infants were classified as 48 infected, 93 not infected and 40 indeterminate. Clearance of maternal antibodies was achieved by 12 months of age in virtually all infants who became seronegative. The intermediate transmission rate was 34% (95% confidence interval, 26 to 42). Deliveries by cesarean section had significantly lower transmission (relative risk, 0.46; 95% confidence interval 0.23 to 0.91). Women with lower hemoglobin concentrations during pregnancy (< 10 g/dl) had an increased risk of transmission (relative risk, 1.99; 95% confidence interval, 1.18 to 3.34). Advanced maternal age, multiparity, positive syphilis serology, duration of ruptured membranes, preterm delivery and breast-feeding were not associated with an increased risk of transmission. CONCLUSIONS: This study, the first from South Africa, has confirmed that the rate of vertical transmission of HIV-1 is as high as that reported from most African cohorts. Cesarean sections were protective against transmission, whereas low hemoglobin values values were associated with an increased risk of transmission. Twelve months could be used as the cutoff age for teh diagnosis of vertical infection using antibody tests.