Babies of South African mothers taking HIV therapy have worse health in early months – unless they are vaccinated and consistently breastfed

Image: HTeam/

A South African study has found that, in general, babies born to HIV-positive mothers who are on antiretroviral therapy do not have worse health outcomes than the children of HIV-negative mothers.

However, it did find that the HIV-positive mothers’ babies had more hospitalisations – especially for lower respiratory tract infections and diarrhoeal disease – in the period from eight days to three months after birth.

Encouragingly, though, this health disadvantage disappeared in babies who received early and complete breastfeeding, and who also received the full course of recommended vaccines.


Antiretroviral therapy (ART) for pregnant women with HIV has stopped hundreds of thousands of babies acquiring HIV and has also, of course, saved the lives of millions of women.



Abnormal bowel movements, characterised by loose, watery or frequent stools, three or more times a day.

virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

exclusive breastfeeding

Feeding an infant only breast milk, with no other liquids or solids, for the first six months of life.


Any lung infection that causes inflammation. The infecting organism may be bacteria (such as Streptococcus pneumoniae), a virus (such as influenza), a fungus (such as Pneumocystis pneumonia or PCP) or something else. The disease is sometimes characterised by where the infection was acquired: in the community, in hospital or in a nursing home.

primary endpoint

The main result that is measured at the end of a clinical study to see if a given treatment worked (e.g., proportion of participants with viral suppression). The choice of primary endpoint is decided before the study begins.

In South Africa alone, 95% of pregnant women with HIV now receive ART (compared with 61% of all people with HIV) and as a result the number of children who acquire HIV per year has declined by 80%, from 70,000 to 13,000.

While these are considerable achievements, they have also had the effect of re-focusing attention on the children of HIV-positive mothers who do not acquire HIV. Previous studies had found that HIV-negative babies born to mothers with HIV generally had poorer health than the babies of negative mothers.

A range of causes were suggested. These included: a poorer immune environment in the womb; fewer disease-fighting antibodies transferred to the baby, either because the mother had fewer, or because she avoided breastfeeding; infections being transmitted to the baby, even if HIV was not; and socioeconomic factors that might be both causes and results of HIV in the mother, such as poverty or lack of sanitation.

However, most of these studies predated modern ART policies for women and were of mothers who were either untreated or who only started ART with low CD4 counts.

The study

This study, therefore, was done to see if HIV in the mother still led to poorer health in their babies, including in women who received ART for a substantial proportion of their pregnancy.

Dr Stanzi Le Roux and her colleagues from the University of Cape Town therefore recruited pregnant women attending the Gugulethu Midwife Obstetric Unit in Cape Town into two cohorts, one for positive and one for negative women.  This clinic sees a population with high infant mortality (2.3% of live births in 2013) and high HIV prevalence among mothers (30%). The clinical care package offered to babies includes vaccination against TB (BCG) at birth; rotavirus (the most common cause of diarrhoea in infants) at six and 14 weeks of age; and pneumococcus at six, ten and 14 weeks. Breastfeeding is promoted as the infant feeding method of choice for HIV-positive mothers who are virally suppressed on ART.

The study compared two cohorts totalling 1087 mother-baby pairs: 587 mothers with HIV and 500 HIV-negative mothers.

The 587 HIV-positive mothers all initiated ART during pregnancy; mothers already taking ART were not included, in order to isolate the effects of ART if started specifically during pregnancy.

To be retained in the cohorts, mothers had to be breastfeeding at their first postnatal clinic visit. This was scheduled for seven days after birth, but a time window of three months was allowed, so some mothers presented quite a long time after giving birth. At this first visit, 13% of the HIV-positive mothers and 8% of the HIV-negative ones were not breastfeeding.

There was another visit at six weeks and then quarterly for the first year, with a final follow-up visit when the babies were 18 months old. At the nine-month visit, retention in the study was 70%.

Participant details

The average age of the women was 28 and was almost the same in both cohorts. There were significant socio-economic differences between the HIV-positive and HIV-negative mothers, underlining the structural aspects of HIV vulnerability. Only a quarter of the HIV-positive mothers had completed secondary education versus 45% of the HIV-negative ones. A quarter of HIV-positive mothers declared ‘risky’ levels of alcohol use during and after pregnancy, versus 7% of the HIV-negative ones; 43% of the positive mother smoked after giving birth, versus 27% of the negative ones; and 22% versus 8% had experienced intimate partner violence. Twenty-seven per cent versus 41% had a flush toilet inside their home. Levels of poverty were roughly similar in both cohorts, however.

Half the HIV-positive mothers started ART with a CD4 count of over 350 and on average they started ART 22 weeks into their pregnancy. Three-quarters (76%) of mothers had a viral load below 50 copies/ml at delivery, with an average viral load at this time of 42 copies/ml.

"Babies who received all their vaccinations and were optimally breastfed were not more likely to be hospitalised than babies of HIV-negative mothers."

Babies born to HIV-positive mothers were slightly more likely to be more than two weeks premature (12% versus 9%) and with a low birthweight, though the percentage of very underweight babies was the same.

Mothers generally started breastfeeding within an hour of the baby being born, but HIV-positive mothers were slightly less likely to (87% versus 94%). However, it is notable that in a study where “optimal breastfeeding” was defined as nothing but breastfeeding for the first five months of life, the average duration of exclusive breastfeeding was 1.5 months in both positive and negative mothers. There was a big difference in the duration of any breastfeeding: on average, HIV-positive mothers had stopped it entirely after four months and HIV-negative after nine months.

However, more of the HIV-positive mothers were in the minority who did give optimal breastfeeding: 17% versus 10%. Two hundred and seventy-seven (60%) of the babies of the HIV-positive mothers received what is described as "complete and timely" vaccination.


During follow-up 12 babies (2%) tested positive for HIV. Two babies of the HIV-positive mothers almost certainly caught it in the womb, while eight caught it after birth, presumably through breastmilk and likely from mothers who were not virally suppressed. Two women in the HIV-negative group tested HIV positive after delivery and passed it on to their babies.

There were 14 infant deaths, with similar numbers in the two arms.

The primary endpoint (measure) of the study was the difference in the number of hospital admissions between babies born to HIV-positive and HIV-negative mothers, and the difference in the number specifically admitted for infectious illness. Fifty-six per cent of all time spent in hospital, and 94% of admissions in the first week of life, was for non-infectious neonatal illness, and there was no difference between the cohorts. 

After the first week, 74% of hospital admissions were for infections. The most common cause of hospital admissions after the first week was lower respiratory tract infections (43% of all infection-related admissions). This primarily means bacterial pneumonia, though there were some cases of neonatal TB, including one death (in a baby of an HIV-negative mother). The second most common cause was diarrhoea (24% of infection-related admissions); it may be significant that this was only seen from three months after birth in the babies of HIV-negative mothers, as breastfeeding tailed off.

There were no significance differences in outcomes between the babies of positive and negative mothers – except in the period from eight days to three months after birth. The overall risk of admission due to infectious illness was 40% greater in the babies of positive mothers, which was marginally statistically significant – but this was entirely due to the risk being 250% (3.5 times) greater in the eight days to three months period.

During this period there were 43 admissions to hospital of the positive mothers’ babies, 34 of them due to infections, compared with 14 admissions of the negative mothers’ babies, nine of them due to infections.

About 4% of all babies of HIV-positive mothers were hospitalised during this period due to lower respiratory tract (LRTI) infections and diarrhoea, whereas only 1.6% of the HIV-negative mothers’ babies were hospitalised during this period for LRTIs, and none for diarrhoea.

Diarrhoea was common throughout the study, and equally so regardless of the mothers’ HIV status, but was usually not severe enough to require hospitalisation. The difference was that cases severe enough to need hospitalisation were more common in the HIV-positive mothers’ babies. LRTIs were less common. They were more common in the babies of HIV-positive mothers, but usually needed hospitalisation in babies of positive and negative mothers alike.

After adjusting for factors including running water, flush toilets and postnatal depression (which of itself doubled the risk of hospitalisation during the eight days to three month period), three factors were associated with the greater risk of hospitalisation for the babies of positive mothers.

The first was the mother’s state of health at the time of starting ART, and how soon they started it. The babies of mothers who started ART with CD4 counts over 350, viral loads below 10,000, and before 24 weeks of pregnancy, were not significantly more likely to be hospitalised than babies of negative mothers.

Of even more influence was whether the babies received all their vaccinations, and whether they were optimally breastfed for those three months. Both of these needed to happen to reduce the relative risk of hospitalisation, compared with the risk to the HIV-negative mothers’ babies, to zero.

Having complete vaccines but suboptimal breastfeeding did not significantly reduce the risk. Optimal breastfeeding but incomplete vaccines reduced the risk a little, so that it became statistically non-significant, but there were only 14 mother-baby pairs in this category.


The findings of this study are hopeful, in that the researchers found that the babies born to mothers with HIV who are given timely antiretroviral therapy, and who do not acquire HIV themselves, are at most time points as healthy as the children of HIV-negative mothers. There was no difference in mortality.

Although there was a difference in hospitalisation rates between a week and three months after birth, this difference was eradicated by vaccinating against the most lethal infant infections, and by exclusive breastfeeding during those three months.

Breastfeeding remains a confusing issue in that the risk-benefit ratio of breastfeeding versus not breastfeeding is different in different parts of the world. In lower-income settings, where access to sanitation is not guaranteed, breastfeeding, in addition to taking ART and having your child vaccinated, is the best way to guard your baby’s health, as supported by the World Health Organization.

Dr Le Roux comments: “The increased risk of dying due to pneumonia or diarrhoea in the absence of breastfeeding outweighs the risks of breastmilk associated HIV transmission in most resource-limited settings, including South Africa – as our data shows.” She pointed out that HIV incidence in the breastfed babies was below 2%, compared with 15% in babies breastfed by mothers not given ART, and 30% in babies of recently infected mothers.

However, 12 babies in this study did acquire HIV from their mothers, ten almost certainly through breastmilk, and eight of those from mothers given ART. Although it is likely that the mothers did not have suppressed viral loads throughout the postnatal period, cases of transmission via breast milk from virally suppressed mothers have been reported. For this reason we cannot say that 'U=U' in the case of breastfeeding, and guidelines for higher-income countries still recommend bottle feeding.

It is important for women in higher-income countries with access to clean water to be guided by their care provider and midwife, as long as they are up to date with the most current evidence of benefits of breastfeeding.