Review

HIV-exposed, uninfected infants: new global challenges in

the era of paediatric HIV elimination
Ceri Evans, Christine E Jones, Andrew J Prendergast

The number of infants infected with HIV is declining with the rise in interventions for the elimination of paediatric Lancet Infect Dis 2016;
HIV infection, but the number of uninfected infants exposed to HIV through their HIV-infected mothers is 16: e92–107

increasing. Interest in the health outcomes of HIV-exposed, uninfected infants has grown in the past decade, with Published Online
March 31, 2016
several studies suggesting that these infants have increased mortality rates, increased infectious morbidity, and
http://dx.doi.org/10.1016/
impaired growth compared with HIV-unexposed infants. However, heterogeneous results might reflect the inherent S1473-3099(16)00055-4
challenges in studies of HIV-exposed, uninfected infants, which need large populations with appropriate, Blizard Institute, Queen Mary
contemporaneous comparison groups and repeated HIV testing throughout the period of breastfeeding. We review University of London, London,
the effects of HIV exposure on mortality, morbidity, and growth, discuss the immunological abnormalities identified UK (C Evans MBBCh,
A J Prendergast DPhil);
so far, and provide an overview of interventions that could be effective in this susceptible population. As the number
Zvitambo Institute for
of infants infected with HIV declines, the health needs of HIV-exposed, uninfected infants should be prioritised Maternal and Child Health
further, to ensure that post-2015 Sustainable Development Goals are achieved. Research, Harare, Zimbabwe
(C Evans, A J Prendergast);
Paediatric Infectious Diseases
Introduction few management options were available for HIV-infected
Research Group, Institute for
Mother-to-child transmission is one of the great tragedies infants at that time. However, despite many subsequent Infection and Immunity,
of the HIV epidemic, particularly in sub-Saharan Africa, studies, the precise effect of vertical HIV exposure on St George’s, University of
where 90% of new infections occur. Although the global child health remains unclear. Interpretation of published London, London, UK
(C E Jones PhD); and
target to eliminate paediatric HIV by 2015 has not been findings is complicated by several factors: (1) early studies
Department of International
achieved, there has been huge progress in the reduction of did not include HIV testing, making the effect of HIV Health, Johns Hopkins
the number of HIV-infected infants through expansion of exposure difficult to separate from infant HIV infection; Bloomberg School of Public
prevention of mother-to-child transmission (PMTCT) (2) follow-up testing of HIV-uninfected mothers and Health, Baltimore, MD, USA
(A J Prendergast)
programmes. In the pre-antiretroviral therapy (ART) era, infants to ensure accurate HIV status is often not available;
Correspondence to:
up to 40% of infants acquired HIV from their mothers,1 but (3) HIV-unexposed comparison groups have either not
Dr Andrew J Prendergast, Centre
with increasing availability of effective PMTCT been included, or are not demographically comparable; (4) for Genomics and Child Health,
interventions, transmission rates close to 1% are possible HIV-exposed, uninfected infants and HIV-unexposed Blizard Institute, Queen Mary
in breastfeeding populations in developing countries,2 and infants often differ in breastfeeding uptake and duration; University of London,
London E1 2AT, UK
less than 1% in non-breastfeeding populations.3 The and (5) the contributions of HIV and ART exposure are
[email protected]
existing public health approach to PMTCT is initiation of not distinguishable in contemporary studies. Despite
lifelong ART for all pregnant and breastfeeding women these limitations, several large studies suggest that HIV-
(so-called Option B+)4 to eliminate new HIV infections in exposed, uninfected infants have increased mortality and
children and keep their mothers alive.5 As PMTCT coverage morbidity in early life.
expands, therefore, the number of HIV-infected infants is
declining, but the number of HIV-exposed, uninfected Mortality in the pre-ART era
infants is increasing (figure). As progress towards The largest HIV-exposed, uninfected cohort (ZVITAMBO),8
elimination of paediatric HIV infection accelerates, HIV- which prospectively followed up 14 110 infants in
exposed, uninfected infants are increasingly being Zimbabwe before availability of ART, found three times
recognised as a group potentially susceptible to uncertain higher mortality in HIV-exposed infants compared with
health needs. In this Review, we summarise the evidence of HIV-unexposed infants up to 2 years of follow-up; this
mortality, infectious morbidity, immune function, and mortality risk was higher in the first year of life compared
growth of HIV-exposed, uninfected infants, and highlight with the second (table 1). Infants and mothers had regular
research gaps that need to be addressed to better define HIV testing, and sensitivity analyses showed that higher
interventions for this growing population. mortality was not due to unascertained postnatal HIV
transmission.8 A study from Uganda10 reported signi-
Mortality and morbidity ficantly higher mortality among HIV-exposed, uninfected
The observation that infants exposed to HIV, but not infants at 18 months of age, but not at 12 or 24 months. A
infected, might have increased susceptibility to infections Zambian study,11 which prospectively assessed infants
first emerged in 1992 when HIV-exposed, uninfected from 9 months to 3 years of age, showed three times
infants in Kenya were found to have a high incidence of higher mortality in HIV-exposed, uninfected children
measles.6 In 1993, HIV-exposed, uninfected infants in compared with HIV-unexposed infants (95% CI 0·7–14·0).
Zaire were noted to have a greater risk of persistent Statistical power in these studies could have been limited
diarrhoea than HIV-unexposed infants;7 HIV-exposed, by small sample sizes. Studies with longer follow-up
uninfected infants were highlighted as a priority group, as periods report mixed results. HIV-exposed, uninfected

www.thelancet.com/infection Vol 16 June 2016 e92

 Review

600 000 exposed to PMTCT and receiving co-trimoxazole

1 600 000 prophylaxis, but breastfeeding duration was significantly
500 000 shorter than in HIV-unexposed infants.
Number of pregnant women living with

1 400 000
Studies comparing mortality in middle-income and

Number of new HIV infections in

1 200 000 high-income countries are scarce, but a need to
400 000
determine the impact of HIV exposure in these settings

children in LMICs
1 000 000
HIV in LMICs

300 000 exists, particularly in the current ART era.

800 000

600 000 200 000 All-cause morbidity and hospitalisation

62%
400 000 47%
56% Before the availability of ART, HIV-exposed, uninfected
47%
31%
40%
100 000 infants in the ZVITAMBO trial9 had more hospitalisations
200 000
13%
22% during the neonatal period and more all-cause clinic visits
0 0 throughout infancy, compared with HIV-unexposed
2005 2006 2007 2008 2009 2010 2011 2012
infants, particularly for pneumonia and oral thrush. In the
Year
ART era, investigators of a large South African study20
Figure: Reduction in perinatally HIV-infected infants with increasing PMTCT coverage found no difference in late-onset sepsis, and a slightly
The number of pregnant women living with HIV in low-income and middle-income countries (white bars) declined lower incidence of early-onset sepsis, in HIV-exposed,
between 2005 and 2012, from 1·64 million to 1·48 million. In the same period, the proportion of HIV-infected uninfected infants during the neonatal period. However,
pregnant women receiving effective PMTCT interventions (green bars) increased substantially from 13% in 2005
to 62% in 2012, leading to a reduction in the number of HIV-infected infants (green line). The number of HIV-exposed, uninfected infants receiving single dose
HIV-exposed uninfected infants is, therefore, increasing. Data from WHO and UNAIDS. LMICs=low-income and nevirapine and 6 months of zidovudine for PMTCT in the
middle-income countries. PMTCT=prevention of mother-to-child transmission. Mashi trial16 in Botswana were more than two times more
likely to be admitted into hospital than HIV-unexposed
For the original data from WHO children in Rwanda had no increase in mortality up to infants up to 24 months. In infants surviving to 20 months
and UNAIDS see http://www. 5 years,12 but in The Gambia, uninfected children exposed of age in Malawi, no differences in hospital admissions
who.int/hiv/data/arvpmtct2014.
png and http://www.unaids.org/
to HIV-1 and HIV-2 who survived beyond 4 months of age were observed between HIV-exposed, uninfected infants
sites/default/files/media_asset/ had increased mortality up to a median age of 6 years and HIV-unexposed infants, although this study might
JC2571_AIDS_by_the_numbers_ compared with HIV-unexposed children.13 have been limited by survival bias.21 Results from one
en_1.pdf small study22 from South Africa showed that, whereas the
Mortality in the ART era frequency of infections in HIV-exposed, uninfected
Where infants were exposed to one drug (single dose infants and HIV-unexposed infants was similar,
nevirapine or zidovudine monotherapy) for PMTCT for a HIV-exposed, uninfected infants had a trend towards
short period of time, results have been heterogeneous. more hospital admissions (relative risk [RR] 2·74, 95% CI
Findings from a large South African study14 showed no 0·85–8·78). HIV-exposed, uninfected infants in
difference in 12-month mortality between HIV-exposed, Mozambique receiving co-trimoxazole prophylaxis, single
uninfected infants and unexposed groups. Most infants in dose nevirapine, and 4 weeks of zidovudine, had fewer
both groups were exclusively breastfed, which might have outpatient clinic attendances than HIV-unexposed infants
partly mitigated differences. Results from a smaller South at 12 months, and similar hospitalisation rates.18
African study,15 in which HIV-exposed, uninfected infants The most common infections in HIV-exposed,
were less likely than HIV-unexposed infants to breastfeed uninfected infants in sub-Saharan Africa, Latin America,
beyond 12 weeks, showed no difference in mortality up to and the Caribbean are skin infections, lower respiratory
36 weeks of age, although approximately one-third were tract infections, and oral thrush.9,23 Reports exist of
recruited from a wealthier area where formula feeding unusual infections in HIV-exposed, uninfected infants,
might have been safe. Mortality was four times higher in including invasive group A streptococcal disease,
HIV-exposed, uninfected infants exposed to single dose haemorrhagic varicella, and oral and invasive candida.9,24,25
nevirapine and 6 months of zidovudine for PMTCT in the HIV-exposed, uninfected infants seem to have more
Mashi trial16 in Botswana than in HIV-unexposed infants severe infections and a greater risk of treatment failure,
up to 24 months of age. In a large trial in Malawi, Tanzania, especially in pneumonia,22,24,26–28 and higher rates of
and Zambia,17 HIV-exposed, uninfected infants had 50% surgical complications.29
higher 12-month mortality despite single dose nevirapine
and co-trimoxazole prophylaxis. Researchers of a small Pneumonia
study from Mozambique,18 in which HIV-exposed infants Pneumonia was more common in HIV-exposed, uninfected
received co-trimoxazole prophylaxis, single dose infants up to 6 months of age in Zimbabwe; in sensitivity
nevirapine, and 4 weeks of zidovudine, found a trend analyses including only those surviving to the end of the
towards increased mortality in HIV-exposed, uninfected interval, HIV-exposed, uninfected infants had more than
infants (odds ratio [OR] 3·74, 95% CI 0·41–34·3). In a three times as many hospital admissions for pneumonia
small Ugandan study,19 mortality was almost 14 times higher during the neonatal period (incidence rate ratio 3·4, 95% CI
(13·7, 1·12–167·3) in HIV-exposed, uninfected infants 1·9–6·0).9 HIV-exposed, uninfected infants in South Africa

e93 www.thelancet.com/infection Vol 16 June 2016

 Review

had an increased incidence of invasive pneumococcal first-line antimicrobial regimens, or altered pathogen-
disease compared with HIV-unexposed infants, before and specific immune responses.
after introduction of pneumococcal conjugate vaccine.30,31
Several studies have described pneumonia caused by an Tuberculosis
unusually broad range of organisms in HIV-exposed, HIV-exposed, uninfected infants in developing countries
uninfected infants, including Staphylococcus aureus,27 could be at greater risk of tuberculosis. In a South African
Escherichia coli, Pseudomonas aeruginosa, other Gram- series,39 four of eight infants with congenital tuberculosis
negative bacteria,27,32 cytomegalovirus and other viruses,27,32,33 were HIV-exposed, uninfected infants; a further two were
and Pneumocystis jirovecii.24,27,34–38 Differences in outcome HIV-exposed, but did not undergo virological testing.
might therefore be partly due to a different range of Compared with HIV-unexposed infants in Uganda, HIV-
causative organisms, which cannot be treated by standard exposed, uninfected infants had 2·6 times increased odds

HIV- HIV- HIV testing Feeding Outcome Result Comments

exposed, unexposed,
uninfected uninfected
infants, n infants, n
Pre-ART
Zimbabwe8 3135 9210 Mothers and EBF at 3 months: HIV-exposed 12-month mortality; HR 3·9 (95% CI Very large number of HIV-exposed
infants tested at uninfected 10·5% vs 24-month mortality 3·15–4·78), HR 2·5 on uninfected and HIV-unexposed infants;
each visit HIV-unexposed uninfected 9·3% sensitivity analysis; comparable, contemporaneous and
(NS); predominantly breastfed HR 2·0 (1·2–3·5), HR 2·0 demographically similar comparison group;
at 3 months: HIV-exposed on sensitivity analysis most infants in both groups did not
uninfected 30·6% vs exclusively breastfeed; regular HIV testing
HIV-unexposed uninfected undertaken; sensitivity analysis supported
31·9% (NS); mixed primary findings
breastfeeding at 3 months:
HIV-exposed uninfected 58·9%
vs HIV-unexposed uninfected
58·9% (NS)9
Uganda10 269 3183 Infants tested 98% breastfed; duration or 12-month mortality; RR 1·08 (NS); Small number of HIV-exposed uninfected
several times; exclusivity not reported 18-month mortality; RR 1·16, p<0·05; infants; infants rested
mothers not 24-month mortality RR 1·29 (NS) regularly—HIV-exposed uninfected group
retested unlikely to include postnatally infected
infants; mothers not retested—unexposed
group possibly included HIV-exposed
uninfected or HIV-infected infants
Zambia11 260 127 Infants tested Not stated 9-month–3-year HR 3·2 (95% CI Small number of HIV-exposed uninfected
several times; mortality 0·7–14·0), p=0·13 and HIV-unexposed infants; not followed
mothers not from birth; period of highest morbidity and
tested mortality not captured (from birth to
6 months); HR not adjusted for
confounders; mothers not
tested—exposure status determined from
infant serology (HR might therefore be
underestimated because unexposed group
might have included HIV-exposed
uninfected or HIV-infected infants);
feeding modality not stated
Rwanda12 138 209 Mothers and Majority breastfed; median of 5-year mortality HR 0·4 (95% CI 0·1–1·6), Small number of HIV-exposed uninfected
infants tested 18 months breastfeeding; p=0·20 and HIV-unexposed uninfected infants; HR
several times exclusivity not reported not adjusted for confounders; long follow-
up, but without breakdown into different
time periods
The Gambia13 258 448 Infants retested Not stated Median 6-year HIV-1 HR 1·6 (95% CI Small number of HIV-exposed uninfected
(HIV-1 64, several times; mortality; median 0·81–3·3), p=0·18; infants; HIV-1 and HIV-2-exposed
HIV-2 194) mothers not 6-year mortality HIV-2 HR 1·2 (0·70–2·0), uninfected infants included; infants
retested dependent on p=0·5; without HIV test classified as unknown;
survival to 4 months; HIV-1 HR 2·7 (1·3–5·6); unclear how missing results were
median 6-year HIV-2 HR 2·0 (1·1–3·6); treated—possibility that HIV-exposed
mortality HIV-1 HR 3·6 (1·6–8·1); uninfected group included undiagnosed
conditional on HIV-2 HR 2·2 (1·1–4·4) postnatally infected infants; feeding
survival to modality not stated; more accurate HIV
18 months status when findings conditioned on
survival to 4 months or 18 months
(Table 1 continues on next page)

www.thelancet.com/infection Vol 16 June 2016 e94

 Review

HIV- HIV- HIV testing Feeding Outcome Result Comments

exposed, unexposed,
uninfected uninfected
infants, n infants, n
(Continued from previous page)
Single dose nevirapine for PMTCT
South Africa14 936 1155 Infants tested Never breastfed: HIV-infected 12-month mortality aHR 0·77 (95% CI Large number of HIV-exposed uninfected
monthly; mothers vs 1·5% HIV-uninfected 0·49–1·21), p=0·253 and HIV-unexposed uninfected infants;
mothers not mothers HAART available before the end of the
retested 6–8 weeks EBF: HIV-infected study (unclear how many mothers
mothers 81·4% vs received HAART); only those with
HIV-uninfected 92·9% available feeding data were included; very
3–4 months EBF: HIV-infected high rates of EBF; mothers not
mothers 61·8% vs retested—unexposed group possibly
HIV-uninfected 72·6% included HIV-exposed uninfected or
HIV-infected infants; data on missing test
results unavailable
South Africa15 543 218 Infants tested 12 weeks EBF: HIV-exposed 9-month mortality HR 0·7 (95% CI 0·3–1·5) 65% of those born to mothers who died
several times; uninfected 54·7% vs HIV- not included in analyses—most at-risk
mothers not unexposed uninfected 9·4% infants possibly excluded; small number of
retested HIV-unexposed uninfected infants;
mothers not retested—unexposed group
possibly included HIV-exposed uninfected
or HIV-infected infants
Single dose nevirapine and 6 months’ infant zidovudine for PMTCT
Botswana16 534 137 Infants tested Median breastfeeding duration: 6-month mortality; HIV-exposed uninfected HIV-infected mothers breastfed for less
several times; HIV-infected mothers 24-month mortality 3·6% vs HIV-unexposed time, but significantly more were
mothers not 5·8 months vs HIV-uninfected uninfected 0·8%, exclusively breastfeeding at 5 months;
tested mothers 9·0 months; p=0·01; small number of HIV-unexposed
5 months EBF: HIV-infected HIV-exposed uninfected uninfected infants; HAART available
mothers 17·5% vs HIV- 6·7% vs HIV-unexposed before the end of the study (unclear how
uninfected mothers 9·5% uninfected 1·6%, many mothers received HAART);
p=0·002 mothers not retested—unexposed group
possibly included HIV-exposed
uninfected or HIV-infected infants;
infants lost to follow-up censored at last
HIV negative test or last study visit
Co-trimoxazole prophylaxis (HIV-exposed, uninfected infants only) and single dose nevirapine for PMTCT
Malawi, 1573 131 Infants tested Not stated 12-month mortality HIV-exposed uninfected Large number of HIV-exposed uninfected
Tanzania, several times; 7·2% vs HIV-unexposed infants; infants lost to follow-up or died
Zambia17 mothers not uninfected 4·8% (no without recent testing not included;
retested; statistical test breastfeeding not stated; small number of
possibility of performed) HIV-unexposed uninfected infants; no
HIV-exposed statistical comparisons undertaken;
uninfected mothers not retested—unexposed group
group including possibly included HIV-exposed uninfected
postnatally or HIV-infected infants; possibility of
infected infants HIV-exposed uninfected group including
postnatally infected infants
Co-trimoxazole prophylaxis (HIV-exposed, uninfected infants only) and single nevirapine and 4 weeks’ zidovudine for PMTCT
Mozambique18 158 160 Not clear if Similar feeding until 6 months; 12-month mortality OR 3·74 (95% CI Not primary trial outcome; small number
postnatally HIV HIV-unexposed uninfected 0·41–34·26), p=0·195 of HIV-exposed uninfected and
infected included breastfed for longer; HIV-unexposed uninfected infants
in analysis 100% of HIV-unexposed
uninfected breastfed to
12 months
Co-trimoxazole prophylaxis (HIV-exposed, uninfected infants only)
Uganda19 186 389 Infants tested at Breastfeeding at 6 months: 24-month mortality 13·7 (95% CI Dedicated study clinic; recruited from
several 84·4% vs 99·7%, p<0·001; 1·12–167·3), p=0·04 4 months of age; period of highest
timepoints; breastfeeding at 12 months: morbidity and mortality not recorded
mothers not 28·7% vs 98·9%, p<0·0001; (birth–4 months); adjusted for age, malaria
retested breastfeeding at 24 months: prophylaxis, breastfeeding, and wealth
0% vs 24%, p<0·0001 index

ART=antiretroviral therapy. EBF=exclusively breastfeeding. aHR=adjusted hazard ratio. NS=not statistically significant. Predominantly breastfed=breast milk and non-milk liquids. Mixed breastfed=breast milk
plus non-human milk, solid foods, or both. RR=relative risk. HR=hazard ratio. PMTCT=prevention of mother-to-child transmission. HAART=highly-active antiretroviral therapy. OR=odds ratio.

Table 1: Studies comparing mortality in HIV-exposed uninfected infants and HIV-unexposed infants

e95 www.thelancet.com/infection Vol 16 June 2016

 Review

of tuberculosis infection, diagnosed by tuberculin skin risk of infection over the first year of life, with 13 times
testing or interferon-γ release assay.40 higher odds of invasive group B streptococcal disease
and four times higher odds of invasive pneumococcal
Diarrhoea disease compared with HIV-unexposed infants.44 HIV-
Studies from sub-Saharan Africa comparing the incidence exposed, uninfected infants born in the ART era had
of diarrhoea between HIV-exposed, uninfected and HIV- almost three times higher morbidity than those born
unexposed infants might partly be confounded by feeding before ART availability (adjusted hazard ratio [aHR]
practices, because of changes in breastfeeding 2·93, 95% CI 1·07–8·05), whether or not mothers
recommendations for HIV-infected mothers over time. A actually received ART. A Danish study45 reported
comparison between formula-fed HIV-exposed, un- increased hospital admissions in HIV-exposed, un-
infected infants and breastfed HIV-unexposed infants in infected children compared with HIV-unexposed
Uganda showed a six times higher risk of severe diarrhoea children in the first 4 years of life; however, this increase
in HIV-exposed, uninfected infants at 6–11 months of was related to haematological disorders rather than
age,19 whereas in Mozambique, HIV-exposed, uninfected infectious diseases.
infants had less diarrhoea and a similar frequency of
diarrhoea-related hospital admission as HIV-unexposed Opportunistic infections
infants, despite a shorter duration of breastfeeding.18 In P jirovecii pneumonia is an opportunistic infection of
Zaire, there was a higher frequency of persistent diarrhoea HIV-infected infants, but was reported in two 7-week-old
in HIV-exposed, uninfected infants compared to HIV- HIV-exposed, uninfected infants from Texas in 1997.34 In
unexposed infants, and results were not influenced by each case, infection was associated with a transient drop
feeding practices.7 After adjustment for feeding practices in CD4 cell count. Other cases of P jirovecii pneumonia in
in South Africa, no significant associations were recorded HIV-exposed, uninfected infants below 6 months of age
between HIV exposure and acute, persistent, or overall have been reported in France,42 South Africa,24,27,46 and the
diarrhoea to 6 months of age.14 Where feeding practices USA.38 In South Africa, over half of infants admitted to
were similar between HIV-exposed, uninfected and HIV- hospital with acute hypoxic pneumonia had P jirovecii
unexposed infants in Kenya, there was no difference in pneumonia detected using molecular techniques; HIV
frequency of diarrhoea,41 and in the ZVITAMBO trial, exposure and malnutrition were risk factors among HIV-
where almost all infants were breastfed, HIV-exposed, uninfected infants.37 Infants in this study were thoroughly
uninfected infants did not have more acute diarrhoea than investigated, meaning the burden of P jirovecii
HIV-unexposed infants.9 pneumonia in HIV-exposed, uninfected infants else-
where might be higher than previously recognised.
Morbidity in developed countries
Fewer studies have assessed morbidity in HIV-exposed, Causes of increased mortality and morbidity
uninfected infants in developed countries, where routine Despite heterogeneity in the scientific literature, the body
ART exposure, lower background infectious morbidity, of evidence from various settings suggests increased
and a scarcity of appropriate HIV-unexposed control morbidity and mortality among HIV-exposed, uninfected
groups from similar socioeconomic backgrounds make infants, and a risk of severe, unusual, and complicated
the effect of HIV exposure harder to establish. Despite infections, compared with HIV-unexposed infants. The
these limitations, findings from several studies suggest causes of increased morbidity and mortality are probably
increased infectious morbidity in HIV-exposed, un- multifactorial, and might be partly driven by adverse
infected infants. In a French study42 of 7638 HIV-exposed, environmental and socioeconomic conditions. However,
uninfected infants, a high proportion (699 of 7638 infants, findings are not driven purely by differences in maternal
Kaplan-Meier estimate of 9·3%, 95% CI 8·7–10·0) were care-taking capacity due to ill health, duration of
admitted to hospital for serious infections during the breastfeeding and quality of breast milk, or colonisation
first year of life; however, no HIV-unexposed control or vertical transmission of pathogens.16,20,47–49 Associations
group was assessed and there was potential for hospital between HIV-exposed, uninfected infant mortality and
admission bias. The most common bacterial infection severity of maternal disease, assessed either by maternal
was pneumonia caused by encapsulated organisms CD4 cell count, viral load, or mortality, have been reported
(Streptococcus pneumoniae and Haemophilus influenzae). in several studies (appendix).8,47,48,50–53 Incidence of
The risk was inversely associated with maternal CD4 cell respiratory tract infection has also been associated with
count for serious bacterial, but not viral, infections. In a severity of maternal HIV infection in several studies.9,23,54,55
Belgian cohort, HIV-exposed, uninfected infants had In Zimbabwe,9 infectious morbidity was high even at
almost 20 times higher incidence of invasive group B maternal CD4 cell counts of more than 500 cells per μL,
streptococcal disease compared with unexposed infants and only reduced to normal with CD4 more than 800 cells
in the first month of life (relative risk 19·6, 95% CI per μL. HIV-exposed, uninfected infants have
7·5–51·7), although the absolute number of cases was immunological abnormalities that could at least partly
small.43 HIV-exposed, uninfected infants remained at account for their poor health outcomes.

www.thelancet.com/infection Vol 16 June 2016 e96

 Review

Immune dysfunction analysis.18 Thymic dysfunction could account for high

An excellent overview of the immunology of HIV-exposed, levels of circulating double negative (CD4– CD8–) T cells
uninfected infants is provided in a review published in in HIV-exposed, uninfected infants.66,71 HIV-exposed,
2014.56 Here, we summarise the probable causes of immune uninfected infants have HIV-specific T-cell responses,72
dysfunction in HIV-exposed, uninfected infants and even in the context of maternal ART, which suggests that
discuss additional studies published in the past few years. exposure to HIV antigens takes places in utero.

Immune activation Humoral immunity and vaccine responses

Chronic immune activation is the hallmark of HIV At birth, HIV-exposed, uninfected infants in developing
infection, and is driven by multiple factors, including HIV countries and developed countries have lower con-
itself, chronic coinfections, and intestinal microbial centrations of specific antibody to vaccine-preventable
translocation.57 HIV-exposed, uninfected infants are infections compared with HIV-unexposed infants.73–77
therefore born to mothers who have high levels of pro- Lower concentrations of specific antibodies in
inflammatory cytokines and activated T cells and B cells HIV-infected mothers, compounded by reduced trans-
during pregnancy.58 Similarly, evidence exists of immune placental transfer of antibody from mother to infant, seem
activation in both innate and adaptive immune cells of to account for the paucity of antibody in HIV-exposed, un-
HIV-exposed, uninfected infants compared with HIV- infected, newborn infants,74,75,78–81 although no association
unexposed controls. In a study of Zimbabwean infants, a between maternal CD4 cell count or viral load and
more pro-inflammatory milieu (ie, raised C-reactive transplacental transfer of antibody exists.74
protein) was evident in HIV-exposed, uninfected infants, However, responses to primary vaccination in
compared with HIV-unexposed infants at 6 weeks of age, HIV-exposed, uninfected infants are similar (for diphtheria,
and was still evident at 6 months of age (Prendergast AJ, tetanus, H influenzae B, and BCG),69,74,77,82–84 or even
unpublished). HIV-exposed, uninfected infants born to increased (for pertussis, pneumococcal capsular
ART-treated women with undetectable viral loads had polysaccharide, and measles),74,76,77,82,85,86 compared with
significantly higher concentrations of pro-inflammatory HIV-unexposed infants (table 2; appendix). Additionally,
cytokines (interleukin 1β and interleukin 8) than HIV- HIV-exposed, uninfected infants have good serological
unexposed controls.59 By comparison with HIV-unexposed evidence of protection after combined measles, mumps,
infants, stimulated monocytes and dendritic cells secrete and rubella vaccination.95 Furthermore, the fold-increase of
higher concentrations of pro-inflammatory cytokines,60 antibody concentration from pre-vaccination to post-
dendritic cells have greater upregulation of the activation vaccination is higher among HIV-exposed, uninfected
markers CD80 and CD86,61 and natural killer (NK) cells infants compared with HIV-unexposed infants.74 An
have higher expression of the activation markers CD38 explanation for this is reduced interference of maternally
(bright) and CD69 in HIV-exposed, uninfected infants.62 acquired antibody with infant B-cell response to
T cells shift towards memory (CD45RA-) and away from vaccination, which has been reported for multiple vaccines.
naive (CD45RA+) phenotypes, have higher expression of Although the mechanism in human beings is unknown,
the activation markers CD154 and CD38, and demonstrate the mechanism in animal models seems to be due to
greater proliferation but reduced polyfunctionality (ie, cross-linking of the B-cell receptor with the regulatory
produce fewer different cytokines) upon stimulation.63–68 receptor FcγRIIB.96 When the concentration of pre-
Compared with HIV-unexposed controls, CD4 T cells vaccination antibody is increased in HIV-exposed,
isolated from HIV-exposed, uninfected infants have higher uninfected compared with HIV-unexposed infants, lower
expression of inflammatory response-related chemokine concentrations are observed after immunisation.77,86 The
receptors (in particular CCR3 and CCR8), and in-vitro functionality of vaccine-specific antibody is similar in
stimulation results in upregulation of CCR8 expression HIV-exposed, uninfected infants and unexposed infants.73,80
and tumour necrosis factor α.59,69 These findings suggest These factors support the hypothesis that differences in
an immune-priming role of HIV exposure; activated cells vaccine responses are driven by the concentration of
are more susceptible to HIV infection in vitro, and maternally derived antibody pre-vaccination, rather than
potentially other intracellular infections.59 differences in infant B-cell function.

T-cell number and T-cell function Birth, growth, and development

HIV-exposed, uninfected infants have lower absolute Long-term nutritional and neurodevelopmental outcomes
numbers of CD4 T cells, which might be due to more are influenced by environmental factors during the first
apoptosis.70 In Malawi, a cohort of HIV-exposed, uninfected 1000 days after conception—the period from conception to
infants had similar CD4 cell counts to HIV-infected infants 2 years of age. HIV exposure during this period could
at 2 weeks, but not 10, weeks of age.63 In Mozambique, the therefore crucially affect birth, growth, and development.
percentage of CD4 T cells was 3% lower, and CD8 In developing countries, HIV-exposed, uninfected
percentage 1·15 times higher in HIV-exposed, uninfected newborn babies might be more likely to be small for
than in HIV-unexposed infants in multivariate longitudinal gestational age or have low birthweight compared with

e97 www.thelancet.com/infection Vol 16 June 2016

 Review

HIV-unexposed newborn babies (table 3, appendix). Low particularly because of pneumonia.55 In HIV-exposed, For more on the 1000 days
birthweight among HIV-exposed, uninfected infants is uninfected infants, preterm birth, low birthweight, and principle see http://www.
thousanddays.org
associated with 2·5–12 times higher risk of mortality,48,51,102,103 being small for gestational age were each associated with

HIV-exposed, HIV- Main findings (HIV-exposed vs HIV-unexposed) Summary

uninfected unexposed,
infants, n uninfected
infants, n
Diphtheria
South Africa77 120 114 No difference in antibody concentration or proportion of infants protected ←→
Brazil87 19 112 No difference in antibody concentration or proportion of infants protected ←→
Tetanus
South Africa74 38 55 No difference in antibody concentration ←→
South Africa77 120 114 No difference in antibody concentration or proportion of infants protected ←→
South Africa82 27 28 No differences in antibody concentration post-primary DTP, but higher concentration post-booster response in ←→ ↑
HIV-exposed uninfected infants. No differences in proportion protected
South Africa83 118 60 No difference in antibody concentration or proportion of infants protected ←→
Brazil87 19 112 Lower concentrations of specific antibody in HIV-exposed uninfected infants, but no difference in seroprotection. ↓ ←→
Uganda88 Roughly 129 Roughly 1304 Reduced interferon γ, interleukin 5, and interleukin 13 responses to tetanus toxin in HIV-exposed uninfected infants ↓
Kenya89 8 (3 months) 8 (3 months) No differences in response to TT antigens at 3 months. At 12 months, lower CD4 cell interleukin 2 response and dual ←→ ↓
26 (12 months) 17 (12 months) interleukin 2 or tumour necrosis factor α response to TT antigens in HIV-exposed uninfected infants
Pertussis (PT and FHA)
South Africa74 38 55 Higher antibody concentration in HIV-exposed uninfected infants, higher fold increase pre-vaccination to ↑
post-vaccination
South Africa77 120 114 Higher PT antibody concentration in HIV-exposed uninfected infants, but similar FHA antibody concentration. ↑ ←→
Increased proportion of HIV-exposed uninfected infants with 4 times increase in PT antibody concentration, but no
difference between HIV-exposed uninfected and HIV-unexposed uninfected infants for FHA antibody
South Africa82 27 28 Higher antibody concentration in HIV-exposed uninfected infants post-primary and pre-booster vaccines, no difference ↑ ←→
post-booster. Higher proportion of seropositive HIV-exposed uninfected infants at 6 months
South Africa83 118 60 Higher antibody concentration in HIV-exposed uninfected infants at 24 weeks, but no difference at 14 and 52 weeks. ↑ ←→
Higher proportion of seropositive HIV-exposed uninfected infants at 24 weeks
Haemophilus influenzae type b (PRP)
Uganda (HIV-exposed 57 14 Higher antibody concentration in HIV-exposed uninfected infants at 48 weeks, but no difference in avidity ↑ ←→
uninfected) and USA
(HIV-unexposed
uninfected)73
South Africa74 38 55 No difference in antibody concentration after three vaccine doses at 16 weeks, but higher fold increase pre-vaccination ←→ ↑
to post-vaccination
South Africa77 120 114 No difference in antibody concentration or SBA GMT ←→
South Africa82 27 28 No difference in vaccine responses ←→
South Africa83 118 60 Higher antibody concentration in HIV-exposed uninfected infants at 14 and 24 weeks, but no difference at 52 weeks. ↑
Higher proportion of HIV-exposed uninfected infants with protective immunity
Denmark84 19 7 No difference in vaccine responses ←→
Pneumococcal capsular polysaccharide
South Africa74 38 55 Higher antibody concentration in HIV-exposed uninfected infants, higher fold increase pre-vaccination to post- ↑
vaccination
South Africa85 120 114 No differences in GMC for serotypes ST4, 9V, 14, 18C, 19F, higher GMC for ST 6B and 23F in HIV-exposed uninfected ↑ ←→
infants after three doses of PCV. No difference in seroprotection, except for ST 6B. Similar results in OPA
South Africa76 121 116 No differences in GMC for ST 6B, 9V, 14, 18C, 19F, 23F, higher GMC for ST 4 in HIV-exposed uninfected infants after ↑ ←→ ←→
2 doses of PCV. No difference in seroprotection, except for ST 4
Similar results for most ST in OPA
Hepatitis B vaccine
South Africa77 120 114 Lower antibody concentration in HIV-exposed uninfected than HIV-unexposed uninfected infants. No difference in ↓ ←→
seroprotection
South Africa82 27 28 No difference in vaccine responses ←→
South Africa83 118 60 Lower antibody concentrations in HIV-exposed uninfected infants at 14 and 52 weeks, no difference at 24 weeks. No ↓ ←→
differences in proportion protected
Brazil87 45 112 Increased proportion of HIV-exposed uninfected infants with high responses, similar proportion with very low ↑ ←→
responses.
(Table 2 continues on next page)

www.thelancet.com/infection Vol 16 June 2016 e98

 Review

HIV-exposed, HIV- Main findings (HIV-exposed vs HIV-unexposed) Summary

uninfected unexposed,
infants, n uninfected
infants, n
(Continued from previous page)
Measles
South Africa82 27 28 No difference in vaccine responses post-primary or post-booster measles vaccine ←→
South Africa86 116 112 Post-primary vaccine: greater vaccine responses; no difference in seroprotection rates
Post-booster vaccine: no difference in vaccine responses; no difference in seroprotection rates ↑ ←→ ↓
At 2 years of age: poorer vaccine responses in HIV-exposed uninfected infants; fewer HIV-exposed uninfected infants
with seroprotection
Oral polio vaccine
Zambia90 133 397 Lower mean antibody titre in HIV-exposed uninfected infants, but no difference when breastfeeding adjusted for ←→
Malawi63 13 21 Similar median anti-polio IgG ←→
BCG
South Africa68 46 46 No difference in BCG-specific T-cell proliferation at 6 weeks; higher frequency at 14 weeks in HIV-exposed uninfected ↑ ←→ ↓
infant; decreased cytokine polyfunctionality in HIV-exposed uninfected infants
South Africa91 118 60 In infants vaccinated with BCG at birth, poorer responses to antigens in HIV-exposed uninfected infants at 14 weeks, ↓ ←→
similar responses at 24 weeks, and comparable responses at 52 weeks. Significantly lower proportion of HIV-exposed
uninfected infants produced a positive interferon-γ response to PPD compared to HIV-unexposed infants at 14 weeks.
In infants vaccinated with BCG at 14 weeks, poorer responses in HIV-exposed uninfected infants at 24 weeks, similar
responses at 52 weeks
South Africa69 47 62 No difference in BCG-specific T-cell proliferation, intracellular cytokine expression, polyfunctional T cells, or secreted ←→
cytokines
Malawi63 13 21 2 weeks: similar interferon-γ response to heat-killed BCG and PPD; 10 weeks: similar proliferation in response to heat- ←→ ↑
killed BCG, poorer proliferation in response to PPD in HIV-exposed uninfected infants; 10 weeks: similar median
anti-Mycobacterium tuberculosis IgG
Uganda88 Roughly 136 Roughly 1370 No difference in responses to crude culture filtrate proteins of Mycobacterium tuberculosis ←→
Kenya89 16 (3 months) 9 (3 months) No differences in frequency of responders to PPD antigens at either timepoint; at 3 months, increased frequency of ←→ ↑
29 (12 months) 17 (12 months) interleukin 2 or tumour necrosis factor α dual positive PPD-specific CD4 T cells in HIV-exposed uninfected infants.
Alterations in memory T-cell subsets at 3 months of age in response to PPD stimulation
South Africa92 94 12 No significant differences in BCG-specific release of interferon γ, measured by ELISA ←→
South Africa93 25 23 No difference in BCG-specific CD4 T cell cytokine response or polyfunctional cells ←→
Brazil94 58 38 Delayed BCG responses in HIV-exposed uninfected infants—by 18·1–26·3 months HIV-exposed uninfected had similar ↓ ←→
responses to HIV-unexposed uninfected infants aged 7·0–8·7 months

DTP=diphtheria, tetanus toxoid, cellular pertussis vaccine. TT=tetanus toxin. PT=pertussis toxin. FHA=filamentous haemagglutinin. PRP=polyribosyl-ribitol phosphate. SBA=serum bactericidal assay.
GMT=geometric mean titre. GMC=geometric mean concentration. OPA=opsonophagocytic activity. ST=serotypes. PCV=pneumococcal conjugate vaccine. OPV=oral polio vaccine. PPD=purified protein derivative.
Better results in HIV-exposed uninfected infants (↑); poorer results in HIV-exposed uninfected infants (↓); similar results between HIV-exposed uninfected and HIV-unexposed uninfected infants (←→).

Table 2: Comparison of vaccination responses between HIV-exposed uninfected infants and HIV-unexposed infants

at least six times higher risk of neonatal mortality, and at Investigators of a large study from South Africa,106 where
least two times increased mortality up to 1 year of age.53 nevirapine was used for PMTCT and most infants were
Data on postnatal growth in HIV-exposed, uninfected exclusively breastfed, did not find a difference in weight-
infants are heterogeneous (appendix). Before the for-age Z scores in the first 24 months of life. In Zambia,
availability of ART (table 3), the largest study (ZVITAMBO) HIV-exposed, uninfected infants who were breastfed had
showed that HIV-exposed, uninfected infants had more better weight-for-age, but not length-for-age, Z scores to
stunting (impaired linear growth), wasting (impaired 15 months of age, compared with formula-fed infants.107
ponderal growth), and underweight than HIV-unexposed Large, well designed studies in the context of fully
infants throughout infancy (Evans C and Prendergast AJ, suppressive maternal ART are needed to determine the
unpublished). Length-for-age remained lower up to 12 effect on growth in the ART era.
months of age and 24 months of age in Zimbabwe (Evans Interest in the impact of impoverished environments
C and Prendergast AJ, unpublished) and Malawi,97 on poor linear growth is emerging; the cleaner
respectively (table 3). Stunting has been associated with environments of developed countries might mitigate
maternal disease severity in some studies,104,105 but not all growth concerns. However, results have been hetero-
(Evans C and Prendergast AJ, unpublished). Impaired geneous. HIV-exposed, uninfected infants in the UK
growth in HIV-exposed, uninfected infants was not seen in were of normal length and weight compared with HIV-
several, mostly smaller, studies, although many showed unexposed infants at 3 years of age.108 Compared with
non-significant trends towards poorer growth (appendix). formula-fed, HIV-unexposed controls, formula-fed

e99 www.thelancet.com/infection Vol 16 June 2016

 Review

HIV-exposed, HIV- Follow-up period Main findings (HIV-exposed vs HIV-unexposed)

uninfected unexposed,
infants, n comparison
group, n
Birthweight
Zimbabwe (Evans C and Roughly 3120 Roughly 9209 24 months Mean birthweight (SD): 2·94 kg (0·46) vs 3·00 kg (0·45), p<0·05†
Prendergast AJ, unpublished)* Percentage low birthweight: 13·9% (95% CI 12·8–15·0) vs 9·9% (9·3–10·5)†; underweight at birth
(OR 1·47 [1·31–1·64])
Malawi97 270 686 24 months Significant difference in mean birthweights†
Democratic Republic of Congo98 190 256 20 months Mean birthweight (SE): 2·95 kg (0·35) vs 3·04 kg (0·32)† (p value not stated)
Mean Z score at birth (SE): 0·20 (0·07) vs −0·47 (0·07) (p value or comparison not stated)
Stunting
Malawi97 270 686 24 months Lower mean LAZ to 24 months†
Democratic Republic of Congo98 190 256 20 months No difference in mean LAZ scores to 20 months
Kenya99 155 139 18 months Mean LAZ scores: −0·48 vs −0·19, p<0·05† at 1·5 months only, otherwise similar
Rwanda100 140 207 48 months Lower mean LAZ scores only at 6 months†, not throughout rest of 48 months follow-up
Zimbabwe (Evans C and Roughly 3120 Roughly 9209 24 months Significantly lower mean LAZ to 12 months†; stunting: birth OR 1·22 (95% CI 1·08–1·38)†, 6 weeks
Prendergast AJ, unpublished)* OR 1·25 (1·11–1·41)†, 6 months OR 1·23 (1·08–1·41)†, 12 months OR 1·23 (1·08–1·39)†, 18 months
NS, 24 months NS
Uganda101 200 400 4–6 months Adjusted OR for stunting at median age of 5·2 months: 2·23, p=0·005†
(cross-sectional)
Underweight
Malawi97 270 686 24 months Lower mean WAZ to 24 months†
Democratic Republic of Congo98 190 256 20 months No difference in mean WAZ scores to 20 months
Rwanda100 140 207 48 months No difference in mean WAZ scores to 48 months
Zimbabwe (Evans C and Roughly 3120 Roughly 9209 24 months Significantly lower mean WAZ to 12 months†; underweight: at birth (see above), 6 weeks OR 1·55
Prendergast AJ, unpublished)* (95% CI 1·30–1·84)†, 6 months OR 1·52 (1·26–1·84)†, 12 months OR 1·37 (1·15–1·63)†, 18 months
NS, 24 months NS
Uganda101 200 400 4–6 months Adjusted OR for underweight at median age of 5·2 months: 1·73, p=0·09
(cross-sectional)
Wasting
Democratic Republic of Congo98 190 256 20 months No difference in mean WLZ scores to 20 months
Kenya99 155 139 18 months Mean WLZ at 6 months: 0·45 vs 0·10, p<0·05‡
Mean WLZ at 18 months: 0·16 vs −0·73, p<0·05‡
Zimbabwe (Evans C and Roughly 3120 Roughly 9209 24 months Significantly lower mean WLZ to 12 months†; wasting: birth OR 1·22 (95% CI 1·10–1·35)†, 6 weeks
Prendergast AJ, unpublished)* OR 1·58 (1·21–2·05)†, 6 months OR 1·43 (1·08–1·90)†, 12 months OR 1·56 (1·22–2·00)†, 18 months
NS, 24 months NS
Malawi97 270 686 24 months Lower mean WLZ to 24 months†
Uganda101 200 400 4–6 months Adjusted OR for wasting at median age 5·2 months: 3·29, p=0·02†
(cross-sectional)

OR=odds ratio. LAZ=length for age Z score. NS=not significant. WAZ=weight for age Z-score. WLZ=weight for length Z-score. *Total numbers of infants in the ZVITAMBO cohort differed at each time point due to
deaths, loss to follow-up, or postnatal HIV infection (infants were censored at the last negative HIV test). †Significantly poorer outcome of HIV-exposed, uninfected infants compared with HIV-unexposed
infants. ‡Significantly poorer outcome of HIV-unexposed infants compared with HIV-exposed, uninfected infants.

Table 3: Studies comparing fetal and postnatal growth in HIV-exposed, uninfected infants and HIV-unexposed infants in developing countries before the availability of ART

HIV-exposed, uninfected infants in Italy had similar unlikely up to 2 years of age in HIV-exposed, uninfected
weight-for-length and weight-for-age Z scores, but lower infants. Authors of more studies from Malawi21 and
length-for-age Z scores at 18 and 24 months of age.109 In Colombia112 reported no significant developmental
the USA, ART-exposed HIV-exposed, uninfected newborn differences at 20 and 24 months of age, respectively.
babies had lower weight and length at birth than matched HIV-exposed, uninfected, Zimbabwean children aged
HIV-unexposed controls, but growth accelerated during 6–8 years did not have more cognitive impairment than
infancy, such that anthropometric measurements were HIV-unexposed controls (OR 1·32, 95% CI 0·69–2·52).113
similar by the age of 10 months.110 However, HIV-uninfected, Zambian children who had
The effect of HIV exposure on neurodevelopment been exposed to HIV in infancy had poorer mathematics,
remains uncertain. A systematic review111 identified mostly but not English, grades at 6–12 years of age than HIV-
studies with methodological shortcomings from high- unexposed children after adjustment for socioeconomic
income settings without control groups. The authors status.114 After adjustment for plausible confounders,
concluded that, once confounders (ie, maternal substance including carer education, HIV-exposed, uninfected
misuse) were accounted for, a developmental delay was children aged 2–12 years in Thailand and Cambodia had

www.thelancet.com/infection Vol 16 June 2016 e100

 Review

poorer verbal IQ, full-scale IQ, and Stanford-Binet Bead impaired tight junction function within 2–4 h of intestinal
memory scores compared with HIV-unexposed controls, epithelial cell exposure to HIV was reported in in-vitro
although differences were small and their clinical experiments, and this dysfunction was associated with
significance remains questionable.115 Results from a Thai bacterial translocation.125
study116 comparing HIV-exposed, uninfected with HIV-un-
exposed children at a mean age of 10·3 years, showed no Antiretroviral therapy
difference in neuroanatomy or brain integrity measured Prolonged exposure to ART during the developmentally
by MRI with diffusion tensor imaging, after adjustment sensitive period from conception to the end of breastfeeding
for socioeconomic status. could be associated with risks. A detailed discussion of
these issues is beyond the scope of this Review, but ART
Mechanisms underlying adverse outcomes in could be associated with preterm birth126 and mitochondrial
HIV-exposed, uninfected infants toxicity.127 Associations between ART and congenital
Multiple factors could contribute to the excess risk of abnormalities have been reported in some studies,128,129 but
morbidity, mortality, and impaired growth that has been not others.130,131 Although an association between antenatal
reported for HIV-exposed, uninfected infants. tenofovir and impaired postnatal growth was observed in
one US study,132 similar results have not been reported
Family environment from sub-Saharan Africa.131 A recent analysis from
Children in households affected by HIV are at a high risk Botswana showed that HIV-exposed, uninfected children
of orphanhood; over 17 million children are estimated to exposed to maternal combination ART had significantly
have lost at least one parent to AIDS since the beginning lower length-for-age and weight-for-age Z scores at
of the HIV epidemic.117 For each woman dying of AIDS in 24 months than those exposed to zidovudine monotherapy.133
Africa, an average of two children are orphaned.118
Maternal mortality is associated with poor infant outcomes Interventions for HIV-exposed, uninfected infants
regardless of the cause of death,119 because mothers who HIV-exposed, uninfected infants are a vulnerable
are sick might have impaired capacity to care for their population who could benefit from additional monitoring
offspring, and orphaned children might experience and interventions. Here, we review the evidence behind
extreme poverty and homelessness.120 However, the existing interventions and consider the potential benefits
offspring of mothers with advanced HIV infection could of additional management approaches.
be at risk of both maternal death and exacerbated immune
abnormalities,121 compounding the morbidity risk in HIV- Antiretroviral therapy
exposed, uninfected infants (appendix). Even in mothers Since lifelong ART for pregnant and breastfeeding women
who survive, HIV infection has far-reaching consequences is an essential component of the efforts to eliminate
for families. HIV-infected parents are at risk of poverty paediatric HIV infection, the benefits outweigh currently
because of an inability to work, which could affect food reported risks of antenatal ART exposure. Few studies
security and access to health care for the household. HIV- have compared health outcomes of HIV-exposed,
exposed, uninfected children might undertake caregiver uninfected infants who are either exposed or not exposed
roles for infected family members, or could be neglected to maternal ART, although in one study,134 an 81%
when other children in the family are HIV-infected. reduction in mortality to 10 years of age (aHR 0·19,
Furthermore, HIV-exposed, uninfected infants might be 95% CI 0·06–0·59) was reported in infants exposed to
brought up in an environment with greater exposure to maternal ART and receiving co-trimoxazole, compared
pathogens, including tuberculosis.122 with those without interventions. With expansion of
PMTCT Option B+, more women will conceive while on
Immune activation ART, and maternal virological suppression and CD4 cell
Immune activation and systemic inflammation could lead reconstitution before pregnancy should improve the
to infection susceptibility and impaired growth. Both health of HIV-exposed, uninfected infants. Longitudinal
immune events could be explained by in-utero and postnatal studies of well characterised cohorts in sub-Saharan Africa
exposure to HIV. Immune activation might result from are needed to establish whether maternal ART normalises
exposure to a pro-inflammatory fetal environment or from immune function, growth, and health in HIV-exposed,
exposure to co-infections (such as cytomegalovirus123) or uninfected infants, or whether additional interventions
HIV itself; HIV-exposed, uninfected infants have evidence are needed.
of HIV-specific T-cell responses,72 suggestive of in-utero
HIV exposure.124 Maternal bacterial translocation, which Co-infections
causes immune activation in HIV infection,58 could directly Whether HIV-exposed, uninfected infants would benefit
affect the developing immune system during fetal life, from additional interventions to reduce P jirovecii pneu-
although further studies are needed to explore this monia, tuberculosis, and vaccine-preventable infections is
hypothesis. Postnatal exposure to HIV through breast- unclear. WHO recommends co-trimoxazole prophylaxis for
feeding might disrupt the intestinal mucosal barrier; all HIV-exposed, uninfected infants from 4–6 weeks of age

e101 www.thelancet.com/infection Vol 16 June 2016

 Review

until cessation of breastfeeding,135 because of the ongoing Infant feeding

risk of HIV transmission and rapid disease progression in Early cessation of breastfeeding reduces HIV transmission,
infected infants. Although side-effects have been noted18 co- but increases morbidity and mortality from other causes in
trimoxazole is generally well tolerated in HIV-exposed, developing countries.14,51,148–153 In Zambia, reduction of the
uninfected infants, and safety data for antenatal136 and duration of breastfeeding increased infant mortality,47 even
postnatal137 co-trimoxazole support its use as a universal if weaning took place in the second year of life.51 In Uganda,
intervention. In a study from Mozambique,18 HIV-exposed, mortality was 6·2 times higher in HIV-exposed, uninfected
uninfected infants receiving co-trimoxazole had fewer infants breastfed for fewer than 6 months compared with
clinic attendances than HIV-unexposed infants not those breastfed for longer (95% CI 1·4–27·0).151 An
receiving co-trimoxazole (incidence ratio 0·79, 95% CI association between breastfeeding cessation and mortality
0·63–0·99), although in trial data from Malawi, Zambia, and morbidity among HIV-exposed, uninfected infants
and Tanzania,17 HIV-exposed, uninfected infants receiving from 6 to 15 months of age was found in a Malawian trial.152
co-trimoxazole had 12-month mortality of 7·2% compared Most, but not all, studies comparing breastfeeding with
with 4·8% in HIV-unexposed infants. In a non-randomised replacement feeding from birth report increased morbidity
comparison of infants in the Malawian BAN trial,138 co- or mortality in formula-fed infants.52,153–156 Compared with
trimoxazole was associated with short-term protection breastfed HIV-exposed, uninfected infants, formula-fed
from malaria, but had no effect on a combined endpoint of HIV-exposed, uninfected infants in South Africa were twice
severe illness or death up to 36 weeks of age. However, as likely to have an illness in the first 2 months of life
subsequent analysis found that co-trimoxazole was (OR 2·02, 95% CI 1·16–3·51).155 In a trial156 from Burkina
associated with reductions in pneumonia, serious febrile Faso, Kenya, and South Africa, HIV-exposed, uninfected
illness, diarrhoea, impaired growth, and malaria infants who were formula-fed from birth had six times as
throughout the first year of life.103 Although protection from many serious infections as those who were breastfed
malaria is commonly reported,103,138–140 not all studies have (adjusted OR 6·0, 95% CI 2·2–16·4). After adjustment for
shown reductions in diarrhoea and pneumonia in HIV- infant growth and maternal disease status, breastfeeding
exposed, uninfected infants receiving co-trimoxazole, and among HIV-exposed, uninfected infants in Kenya was
the risks of co-trimoxazole are argued to outweigh associated with 47% lower risk of pneumonia compared
benefits.141 To address this uncertainty, two randomised with those never breastfed (HR 0·53, 95% CI 0·39–0·73).154
trials of co-trimoxazole in HIV-exposed, uninfected infants However, causative organisms in HIV-exposed, uninfected
are underway in Botswana (ClinicalTrials.gov NCT01229761) infants with diarrhoea or pneumonia did not differ
and South Africa (Pan-African Clinical Trials Registry according to feeding strategy in an analysis from the Mashi
PACTR201311000621110). Interventions to prevent trial in Botswana.32
tuberculosis have been unsuccessful; isoniazid preventive Although replacement feeding is recommended in
therapy did not reduce tuberculosis infection in HIV- PMTCT programmes in developed countries, it is only
exposed, uninfected infants in South Africa.142 recommended in developing countries if it is acceptable,
feasible, affordable, sustainable, and safe.157 Breastmilk of
Vaccinations HIV-infected women has similar levels of innate immune
Vaccine responses seem to be broadly similar in factors and specific immunoglobulins as HIV-uninfected
HIV-exposed, uninfected infants and HIV-unexposed women.16,158 Breastfeeding in the context of Option B+,
infants, which highlights the importance of timely and where women are on lifelong ART, leads to low levels of
complete vaccination. However, there remains a paucity HIV transmission.
of clinical data on vaccine effectiveness in HIV-exposed,
uninfected infants.143,144 Vaccine strategies that might Summary of trial outcomes
mitigate the risk of infectious morbidity early in life As progress is made towards elimination of paediatric HIV
include antenatal immunisation, neonatal vaccination, infection, research interest in the health needs of
and earlier commencement of vaccine schedules. HIV-exposed, uninfected infants is growing. The
However, antenatal vaccination in the context of HIV population remains challenging to study because of the
might be associated with reduced maternal immuno- confounding effects of ART exposure and feeding
genicity and therefore reduced transplacental transfer of modality, difficulties in confirming HIV status, and a
antibody to the infant,145 especially in cases of increased scarcity of comparable, contemporaneous control groups.
maternal disease severity.146 Earlier infant vaccination However, the largest study, which followed over 3000 well
might be more efficacious; however, the feasibility of characterised HIV-exposed, uninfected infants before
delivering such a programme in developing countries has availability of ART in Zimbabwe, showed increased
not been assessed and is likely to be a barrier. In HIV- mortality,8 increased infectious morbidity,9 and impaired
prevalent settings, BCG should be delayed until infant growth (Evans C and Prendergast AJ, unpublished), in line
HIV status has been confirmed to avoid the vaccination with several other well designed, but smaller,
of HIV-infected infants, who are at risk of disseminated studies.10,16,17,30,31,97,159 A South African study done in the pre-
BCG disease.147 ART era with high exclusive breastfeeding rates found no

www.thelancet.com/infection Vol 16 June 2016 e102

 Review

Panel: Suggested areas for future research in HIV-exposed, uninfected infants Search strategy and selection criteria
Assess mortality and growth of HIV-exposed, uninfected infants in the ART era We searched PubMed using the terms (HIV or human
• Clarify deleterious effects of ART exposure immunodeficiency virus[MeSH Terms]) AND (expos* or
• Better understand whether poorer outcomes of HIV-exposed, uninfected infants persist in uninfected or maternal or affect*[MeSH Terms]) AND (child*
the current era, by comparison of HIV-exposed, uninfected infants exposed to ART, with or infant* or fetus or fetal or fet* or neonat*[MeSH Terms])
appropriate HIV-unexposed groups in developing and developed settings for English language papers between January, 1983, and
Establish the causes of infection susceptibility December, 2015. We reviewed reference lists of included
• Discern the effects of immunological abnormalities, maternal caretaking capabilities, studies and screened abstracts from relevant conference
exposure to opportunistic infections, and feeding methods proceedings. Articles were selected if the infants described
• Explore the relationship between inflammation and infection susceptibility or growth of were diagnosed as HIV-exposed and uninfected. A systematic
HIV-exposed, uninfected infants review was beyond the scope of this paper. Larger, better
quality, and newer studies with contemporaneous
Better characterise the immune function of HIV-exposed, uninfected infants comparison groups were preferentially selected, as were
• Assess ontogeny of the innate and adaptive immune system in HIV-exposed, uninfected systematic reviews or meta-analyses of previous data. Studies
infants with direct comparisons between HIV-exposed and
• Define functional immune defects in HIV-exposed, uninfected infants using modern HIV-unexposed groups were preferentially included, whereas
immunological techniques studies in which the HIV status of infants was not established
• Find out the effect of HIV exposure on immune activation beyond infancy and early with virological testing were excluded.
childhood
Evaluate the mechanisms of immune activation, inflammation, and T-cell abnormalities
• Explore the contributions of maternal or infant enteropathy, maternal or infant microbial prospective studies of appropriate cohorts are needed to
translocation, maternal CD4 cell count, exposure to HIV in utero or through breastfeeding, determine the relative importance of immune function,
maternal or infant co-infections, maternal or infant malnutrition, and exposure to ART feeding practices, and family environment in mediating
poor outcomes of HIV-exposed, uninfected infants
Assess the impact of immunological abnormalities on mortality or morbidity outcomes of
(panel). Further detailed immunological analyses using
HIV-exposed, uninfected infants
modern laboratory techniques are needed, particularly
• Correlate markers of immune dysfunction with clinical outcomes
in well characterised cohorts from sub-Saharan Africa.
• Better understand the timecourse of immune defects and interventions that might be
Reassuringly, HIV-exposed, uninfected infants mount
beneficial
robust responses to vaccination; timely vaccination is
• Establish the long-term impact of immune activation on health outcomes
vital to reduce the risk of vaccine-preventable infections.
Investigate the impact of co-trimoxazole HIV-exposed, uninfected infants should remain in
• Study the effect of co-trimoxazole on immune activation, HIV transmission, growth, and clinical follow-up after early infant diagnosis at 4–6
health of HIV-exposed, uninfected infants weeks of age, with repeat HIV testing after complete
Infection prophylaxis and vaccination strategies cessation of breastfeeding. Although breastfeeding
• Explore the role of additional antimicrobial prophylaxis to prevent co-infections, including increases the risk of HIV acquisition, transmission rates
cytomegalovirus, tuberculosis, pneumocystis pneumonia in the context of fully suppressive maternal ART are low
• Explore the protection provided by antenatal immunisation, neonatal vaccination, or (approximately 0·2% per month160) and prolonged
accelerated vaccination schedule to young infants, and the potential for modulation of breastfeeding reduces mortality in settings where
the infant’s own response to vaccination formula-feeding is unsafe.
Identification of new techniques to reduce the transmission of HIV during breastfeeding
Conclusion
• Breastfeeding could ameliorate some of the risks associated with HIV exposure, but HIV
Prevention of mother-to-child transmission of HIV is an
transmission is still possible
evolving success story, although several challenges
ART=antiretroviral therapy. remain in the quest to eliminate paediatric HIV.
Integration of PMTCT into broader maternal, newborn,
and child health services would meet the objective of the
differences in mortality14 or weight106 of HIV-exposed, improvement of child and maternal survival together.161
uninfected infants compared with HIV-unexposed infants, However, as the number of HIV-infected infants declines,
which suggests that uptake of exclusive breastfeeding the health needs of HIV-exposed, uninfected infants
could at least partly mitigate adverse outcomes. should be prioritised further, to ensure that post-2015
Reports from developed countries suggest that, despite Sustainable Development Goals are achieved.
maternal ART, HIV-exposed, uninfected infants remain Contributors
at increased risk of infections.42–44 Immunological abnor- CE undertook the literature review and wrote the original draft of the
malities have been reported in HIV-exposed, uninfected paper, which was reviewed and revised by CEJ and AJP.
infants from both developing and developed countries. Declaration of interests
These might persist despite maternal ART, although We declare no competing interests.

e103 www.thelancet.com/infection Vol 16 June 2016

 Review

Acknowledgments 20 Cutland CL, Schrag SJ, Zell ER, et al. Maternal HIV infection and
CE is funded by the National Institute for Health Research. AJP is funded vertical transmission of pathogenic bacteria. Pediatrics 2012;
by the Wellcome Trust (108065/Z/15/Z). 130: e581–90.
21 Landes M, van Lettow M, Chan AK, Mayuni I, Schouten EJ,
References Bedell RA. Mortality and health outcomes of HIV-exposed and
1 Datta P, Embree JE, Kreiss JK, et al. Mother-to-child transmission of unexposed children in a PMTCT cohort in Malawi. PLoS One 2012;
human immunodeficiency virus type 1: report from the Nairobi Study. 7: e47337.
J Infect Dis 1994; 170: 1134–40.
22 Slogrove A, Reikie B, Naidoo S, et al. HIV-exposed uninfected infants
2 Shapiro RL, Kitch D, Ogwu A, et al. HIV transmission and 24-month are at increased risk for severe infections in the first year of life.
survival in a randomized trial of HAART to prevent MTCT during J Trop Pediatr 2012; 58: 505–08.
pregnancy and breastfeeding in Botswana. AIDS 2013; 27: 1911–20.
23 Mussi-Pinhata MM, Freimanis L, Yamamoto AY, et al.
3 Forbes JC, Alimenti AM, Singer J, et al. A national review of vertical Infectious disease morbidity among young HIV-1-exposed but
HIV transmission. AIDS 2012; 26: 757–63. uninfected infants in Latin American and Caribbean countries:
4 WHO. Consolidated guidelines on the use of antiretroviral drugs for the National Institute of Child Health and Human Development
treating and preventing HIV infection: recommendations for a public International Site Development Initiative Perinatal Study.
health approach. Geneva: World Health Organization, 2013. http:// Pediatrics 2007; 119: e694–704.
apps.who.int/iris/bitstream/10665/85321/1/9789241505727_eng.pdf 24 Slogrove AL, Cotton MF, Esser MM. Severe infections in
(accessed July 28, 2015). HIV-exposed uninfected infants: clinical evidence of
5 UNAIDS. Global plan towards the elimination of new HIV infections immunodeficiency. J Trop Pediatr 2010; 56: 75–81.
among children by 2015 and keeping their mothers alive. Geneva: 25 Hoffmann KS, Green RJ. Candida perforation of the intestine in an
Joint United Nations Programme on HIV/AIDS (UNAIDS), 2011. HIV-exposed but uninfected infant. SAJCH 2014; 8: 119–21.
http://www.unaids.org/sites/default/files/media_asset/20110609_
26 Izadnegahdar R, Fox MP, Jeena P, Qazi SA, Thea DM.
JC2137_Global-Plan-Elimination-HIV-Children_en_1.pdf (accessed
Revisiting pneumonia and exposure status in infants born to
July 28, 2015).
HIV-infected mothers. Pediatr Infect Dis J 2014; 33: 70–72.
6 Embree JE, Datta P, Stackiw W, et al. Increased risk of early measles in
27 McNally LM, Jeena PM, Gajee K, et al. Effect of age, polymicrobial
infants of human immunodeficiency virus type 1-seropositive
disease, and maternal HIV status on treatment response and cause
mothers. J Infect Dis 1992; 165: 262–67.
of severe pneumonia in South African children: a prospective
7 Thea DM, St Louis ME, Atido U, et al. A prospective study of diarrhea descriptive study. Lancet 2007; 369: 1440–51.
and HIV-1 infection among 429 Zairian infants. N Engl J Med 1993;
28 Kelly MS, Wirth KE, Steenhoff AP, et al. Treatment failures and
329: 1696–702.
excess mortality among HIV-exposed, uninfected children with
8 Marinda E, Humphrey JH, Iliff PJ, et al. Child mortality according to pneumonia. J Pediatric Infect Dis Soc 2015; 4: e117–26.
maternal and infant HIV status in Zimbabwe. Pediatr Infect Dis J 2007;
29 Karpelowsky JS, Millar AJ, van der Graaf N, van Bogerijen G, Zar HJ.
26: 519–26.
Outcome of HIV-exposed uninfected children undergoing surgery.
9 Koyanagi A, Humphrey JH, Ntozini R, et al. Morbidity among human BMC Pediatr 2011; 11: 69.
immunodeficiency virus-exposed but uninfected, human
30 von Mollendorf C, Cohen C, de Gouveia L, et al. Risk factors for
immunodeficiency virus-infected, and human immunodeficiency
invasive pneumococcal disease among children less than 5 years of
virus-unexposed infants in Zimbabwe before availability of highly
age in a high HIV prevalence setting, South Africa, 2010 to 2012.
active antiretroviral therapy. Pediatr Infect Dis J 2011; 30: 45–51.
Pediatr Infect Dis J 2015; 34: 27–34.
10 Brahmbhatt H, Kigozi G, Wabwire-Mangen F, et al. Mortality in
31 von Mollendorf C, von Gottberg A, Tempia S, et al. Increased risk for
HIV-infected and uninfected children of HIV-infected and uninfected
and mortality from invasive pneumococcal disease in HIV-exposed
mothers in rural Uganda. J Acquir Immune Defic Syndr 2006;
but uninfected infants aged <1 year in South Africa, 2009–13.
41: 504–08.
Clin Infect Dis 2015; 60: 1346–56.
11 Sutcliffe CG, Scott S, Mugala N, et al. Survival from 9 months of age
32 Zash R, Shapiro, RL, Leidner, J, et al. The aetiology of diarrhoea,
among HIV-infected and uninfected Zambian children prior to the
pneumonia and respiratory colonization of HIV-exposed infants
availability of antiretroviral therapy. Clin Infect Dis 2008; 47: 837–44.
randomized to breast- or formula-feeding. Paediatr Int Child Health
12 Spira R, Lepage P, Msellati P, et al. Natural history of human 2015; published online June 8. DOI:10.1179/2046905515Y.0000000038.
immunodeficiency virus type 1 infection in children: a five-year
33 Hsiao NY, Zampoli M, Morrow B, Zar HJ, Hardie D.
prospective study in Rwanda. Mother-to-Child HIV-1 Transmission
Cytomegalovirus viraemia in HIV exposed and infected infants:
Study Group. Pediatrics 1999; 104: e56.
prevalence and clinical utility for diagnosing CMV pneumonia.
13 Schim van der Loeff MF, Hansmann A, Awasana AA, et al. Survival of J Clin Virol 2013; 58: 74–78.
HIV-1 and HIV-2 perinatally infected children in The Gambia. AIDS
34 Heresi GP, Caceres E, Atkins JT, Reuben J, Doyle M. Pneumocystis
2003; 17: 2389–94.
carinii pneumonia in infants who were exposed to human
14 Rollins NC, Ndirangu J, Bland RM, Coutsoudis A, Coovadia HM, immunodeficiency virus but were not infected: an exception to the
Newell ML. Exclusive breastfeeding, diarrhoeal morbidity and AIDS surveillance case definition. Clin Infect Dis 1997; 25: 739–40.
all-cause mortality in infants of HIV-infected and HIV uninfected
35 Jeena PM, Bobat B, Thula SA, Adhikari M. Children with
mothers: an intervention cohort study in KwaZulu Natal, South Africa.
Pneumocystis jiroveci pneumonia and acute hypoxaemic respiratory
PLoS One 2013; 8: e81307.
failure admitted to a PICU, Durban, South Africa. Arch Dis Child
15 Chopra M, Doherty T, Goga A, Jackson D, Persson LA. Survival of 2008; 93: 545.
infants in the context of prevention of mother to child HIV
36 Morrow BM, Hsaio NY, Zampoli M, Whitelaw A, Zar HJ.
transmission in South Africa. Acta Paediatr 2010; 99: 694–98.
Pneumocystis pneumonia in South African children with and
16 Shapiro RL, Lockman S, Kim S, et al. Infant morbidity, mortality, and without human immunodeficiency virus infection in the era of
breast milk immunologic profiles among breast-feeding HIV-infected highly active antiretroviral therapy. Pediatr Infect Dis J 2010;
and HIV-uninfected women in Botswana. J Infect Dis 2007; 196: 562–69. 29: 535–39.
17 Chilongozi D, Wang L, Brown L, et al. Morbidity and mortality among 37 Morrow BM, Samuel CM, Zampoli M, Whitelaw A, Zar HJ.
a cohort of human immunodeficiency virus type 1-infected and Pneumocystis pneumonia in South African children diagnosed by
uninfected pregnant women and their infants from Malawi, Zambia, molecular methods. BMC Res Notes 2014; 7: 26.
and Tanzania. Pediatr Infect Dis J 2008; 27: 808–14.
38 Paul ME, Chantry CJ, Read JS, et al. Morbidity and mortality during
18 Moraleda C, de Deus N, Serna-Bolea C, et al. Impact of HIV exposure the first two years of life among uninfected children born to human
on health outcomes in HIV-negative infants born to HIV-positive immunodeficiency virus type 1-infected women: the women and
mothers in Sub-Saharan Africa. J Acquir Immune Defic Syndr 2014; infants transmission study. Pediatr Infect Dis J 2005; 24: 46–56.
65: 182–89.
39 Bekker A, Du Preez K, Schaaf HS, Cotton MF, Hesseling AC.
19 Marquez C, Okiring J, Chamie G, et al. Increased morbidity in early High tuberculosis exposure among neonates in a high tuberculosis
childhood among HIV-exposed uninfected children in Uganda is and human immunodeficiency virus burden setting.
associated with breastfeeding duration. J Trop Pediatr 2014; 60: 434–41. Int J Tuberc Lung Dis 2012; 16: 1040–46.

www.thelancet.com/infection Vol 16 June 2016 e104

 Review

40 Carina M, Chamie G, Achan J, et al. Tuberculosis infection in early 61 Velilla PA, Montoya CJ, Hoyos A, Moreno ME, Chougnet C,
childhood in Uganda and the influence of HIV exposure. Rugeles MT. Effect of intrauterine HIV-1 exposure on the frequency
Conference on Retroviruses and Opportunistic Infections; and function of uninfected newborns’ dendritic cells. Clin Immunol
Boston, MA; March 3–6, 2014. 2008; 126: 243–50.
41 van Eijk A, Brooks, JT, Adcock, PM, et al. Diarrhea in children less 62 Slyker JA, Lohman-Payne B, John-Stewart GC, et al. The impact of
than two years of age with known HIV status in Kisumu, Kenya. HIV-1 infection and exposure on natural killer (NK) cell phenotype in
Int J Infect Dis 2010; 14: e220–25. Kenyan infants during the first year of life. Front Immunol 2012; 3: 399.
42 Taron-Brocard C, Le Chenadec J, Faye A, et al. Increased risk of 63 Miles DJ, Gadama L, Gumbi A, Nyalo F, Makanani B, Heyderman RS.
serious bacterial infections due to maternal immunosuppression in Human immunodeficiency virus (HIV) infection during pregnancy
HIV-exposed uninfected infants in a European country. induces CD4 T-cell differentiation and modulates responses to Bacille
Clin Infect Dis 2014; 59: 1332–45. Calmette-Guérin (BCG) vaccine in HIV-uninfected infants.
43 Epalza C, Goetghebuer T, Hainaut M, et al. High incidence of Immunology 2010; 129: 446–54.
invasive group B streptococcal infections in HIV-exposed 64 Rich KC, Siegel JN, Jennings C, Rydman RJ, Landay AL. Function and
uninfected infants. Pediatrics 2010; 126: e631–38. phenotype of immature CD4+ lymphocytes in healthy infants and
44 Adler C, Haelterman E, Barlow P, Marchant A, Levy J, early lymphocyte activation in uninfected infants of human
Goetghebuer T. Severe infections in HIV-exposed uninfected immunodeficiency virus-infected mothers. Clin Diagn Lab Immunol
infants born in a European country. PLoS One 2015; 10: e0135375. 1997; 4: 358–61.
45 Larsen E, Lunding S, Helleberg M, et al. Hospitalizations among 65 Ono E, Nunes dos Santos AM, de Menezes Succi RC, et al.
uninfected children exposed or unexposed to HIV—a nationwide Imbalance of naive and memory T lymphocytes with sustained high
cohort study. Conference on Retroviruses and Opportunistic cellular activation during the first year of life from uninfected children
Infections; Seattle, WA, Feb 23–26, 2015. born to HIV-1-infected mothers on HAART. Braz J Med Biol Res 2008;
46 McNally LM, Jeena PM, Lalloo U, et al. Probable mother to infant 41: 700–08.
transmission of Pneumocystis jiroveci from an HIV-infected woman 66 Clerici M, Saresella M, Colombo F, et al. T-lymphocyte maturation
to her HIV-uninfected infant. AIDS 2005; 19: 1548–49. abnormalities in uninfected newborns and children with vertical
47 Kuhn L, Kasonde P, Sinkala M, et al. Does severity of HIV disease exposure to HIV. Blood 2000; 96: 3866–71.
in HIV-infected mothers affect mortality and morbidity among their 67 Romano MF, Buffolano W, Bisogni R, et al. Increased CD154
uninfected infants? Clin Infect Dis 2005; 41: 1654–61. expression in uninfected infants born to HIV-positive mothers
48 Fox MP, Brooks DR, Kuhn L, et al. Role of breastfeeding cessation exposed to antiretroviral prophylaxis. Viral Immunol 2006; 19: 363–72.
in mediating the relationship between maternal HIV disease stage 68 Kidzeru EB, Hesseling AC, Passmore J-A, et al. In-utero exposure to
and increased child mortality among HIV-exposed uninfected maternal HIV infection alters T-cell immune responses to vaccination
children. Int J Epidemiol 2009; 38: 569–76. in HIV-uninfected infants. AIDS 2014; 28: 1421–30.
49 Gray KJ, Kafulafula G, Matemba M, Kamdolozi M, Membe G, 69 Jones CE, Hesseling AC, Tena-Coki NG, et al. The impact of HIV
French N. Group B Streptococcus and HIV infection in pregnant exposure and maternal Mycobacterium tuberculosis infection on infant
women, Malawi, 2008–2010. Emerg Infect Dis 2011; 17: 1932–35. immune responses to bacille Calmette-Guérin vaccination. AIDS 2015;
50 Chatterjee A, Bosch RJ, Hunter DJ, Fataki MR, Msamanga GI, 29: 155–65.
Fawzi WW. Maternal disease stage and child undernutrition in 70 Economides A, Schmid I, Anisman-Posner DJ, Plaeger S, Bryson YJ,
relation to mortality among children born to HIV-infected women Uittenbogaart CH. Apoptosis in cord blood T lymphocytes from
in Tanzania. J Acquir Immune Defic Syndr 2007; 46: 599–606. infants of human immunodeficiency virus-infected mothers.
51 Kuhn L, Sinkala M, Semrau K, et al. Elevations in mortality Clin Diagn Lab Immunol 1998; 5: 230–34.
associated with weaning persist into the second year of life among 71 Nielsen SD, Jeppesen DL, Kolte L, et al. Impaired progenitor cell
uninfected children born to HIV-infected mothers. Clin Infect Dis function in HIV-negative infants of HIV-positive mothers results in
2010; 50: 437–44. decreased thymic output and low CD4 counts. Blood 2001; 98: 398–404.
52 Newell ML, Coovadia H, Cortina-Borja M, et al. Mortality of infected 72 Liu AY, Lohman-Payne B, Chung MH, et al. Maternal plasma and
and uninfected infants born to HIV-infected mothers in Africa: breastmilk viral loads are associated with HIV-1-specific cellular
a pooled analysis. Lancet 2004; 364: 1236–43. immune responses among HIV-1-exposed, uninfected infants in
53 Slyker JA, Patterson J, Ambler G, et al. Correlates and outcomes of Kenya. Clin Exp Immunol 2015; 180: 509–19.
preterm birth, low birth weight, and small for gestational age in 73 Gaensbauer JT, Rakhola JT, Onyango-Makumbi C, et al. Impaired
HIV-exposed uninfected infants. BMC Pregnancy Childbirth 2014; Haemophilus influenzae type b transplacental antibody transmission
14: 7. and declining antibody avidity through the first year of life represent
54 Mussi-Pinhata MM, Motta F, Freimanis-Hance L, et al. Lower potential vulnerabilities for HIV-exposed but -uninfected infants.
respiratory tract infections among human immunodeficiency Clin Vaccine Immunol 2014; 21: 1661–67.
virus-exposed, uninfected infants. Int J Infect Dis 2010; 74 Jones CE, Naidoo S, De Beer C, Esser M, Kampmann B,
14 (suppl 3): e176–82. Hesseling AC. Maternal HIV infection and antibody responses against
55 Mwiru R, Spiegelman D, Hertzmark E, et al. Nutritional vaccine-preventable diseases in uninfected infants. JAMA 2011;
predictors of acute respiratory infections among children born 305: 576–84.
to HIV-infected women in Tanzania. J Trop Pediatr 2013; 75 Jones C, Pollock L, Barnett SM, Battersby A, Kampmann B.
59: 203–08. Specific antibodies against vaccine-preventable infections:
56 Afran L, Garcia Knight M, Nduati E, Urban BC, Heyderman RS, a mother-infant cohort study. BMJ Open 2013; 3: e002473.
Rowland-Jones SL. HIV-exposed uninfected children: a growing 76 Madhi SA, Izu A, Violari A, et al. Immunogenicity following the first
population with a vulnerable immune system? Clin Exp Immunol and second doses of 7-valent pneumococcal conjugate vaccine in
2014; 176: 11–22. HIV-infected and -uninfected infants. Vaccine 2013; 31: 777–83.
57 Klatt NR, Chomont N, Douek DC, Deeks SG. Immune activation 77 Simani OE, Izu A, Violari A, et al. Effect of HIV-1 exposure and
and HIV persistence: implications for curative approaches to HIV antiretroviral treatment strategies in HIV-infected children on
infection. Immunol Rev 2013; 254: 326–42. immunogenicity of vaccines during infancy. AIDS 2014; 28: 531–41.
58 Brenchley JM, Price DA, Schacker TW, et al. Microbial translocation 78 Cumberland P, Shulman CE, Maple PA, et al. Maternal HIV infection
is a cause of systemic immune activation in chronic HIV infection. and placental malaria reduce transplacental antibody transfer and
Nat Med 2006; 12: 1365–71. tetanus antibody levels in newborns in Kenya. J Infect Dis 2007;
59 Bunders MJ, van Hamme JL, Jansen MH, Boer K, Kootstra NA, 196: 550–57.
Kuijpers TW. Fetal exposure to HIV-1 alters chemokine receptor 79 de Moraes-Pinto MI, Almeida AC, Kenj G, et al. Placental transfer and
expression by CD4+T cells and increases susceptibility to HIV-1. maternally acquired neonatal IgG immunity in human
Sci Rep 2014; 4: 6690. immunodeficiency virus infection. J Infect Dis 1996; 173: 1077–84.
60 Reikie BA, Adams RC, Leligdowicz A, et al. Altered innate immune 80 de Moraes-Pinto MI, Verhoeff F, Chimsuku L, et al. Placental antibody
development in HIV-exposed uninfected infants. transfer: influence of maternal HIV infection and placental malaria.
J Acquir Immune Defic Syndr 2014; 66: 245–55. Arch Dis Child Fetal Neonatal Ed 1998; 79: F202–05.

e105 www.thelancet.com/infection Vol 16 June 2016

 Review

81 Le Doare K, Allen L, Kampmann B, et al. Anti-group B Streptococcus 102 Wei R, Msamanga GI, Spiegelman D, et al. Association between
antibody in infants born to mothers with human immunodeficiency low birth weight and infant mortality in children born to human
virus (HIV) infection. Vaccine 2015; 33: 621–27. immunodeficiency virus 1-infected mothers in Tanzania.
82 Reikie BA, Naidoo S, Ruck CE, et al. Antibody responses to vaccination Pediatr Infect Dis J 2004; 23: 530–35.
among South African HIV-exposed and unexposed uninfected infants 103 Kourtis AP, Wiener J, Kayira D, et al. Health outcomes of
during the first 2 years of life. Clin Vaccine Immunol 2013; 20: 33–38. HIV-exposed uninfected African infants. AIDS 2013; 27: 749–59.
83 Hesseling AC, Jones CE, Blakney AK, et al. Delayed BCG 104 Venkatesh KK, Lurie MN, Triche EW, et al. Growth of infants born
immunization does not alter antibody responses to EPI vaccines in to HIV-infected women in South Africa according to maternal and
HIV-exposed and HIV-unexposed South African infants. Vaccine infant characteristics. Trop Med Int Health 2010; 15: 1364–74.
(in press). 105 McGrath CJ, Nduati R, Richardson BA, et al. The prevalence of
84 Kolte L, Rosenfeldt V, Vang L, et al. Reduced thymic size but no stunting is high in HIV-1-exposed uninfected infants in Kenya.
evidence of impaired thymic function in uninfected children born to J Nutr 2012; 142: 757–63.
human immunodeficiency virus-infected mothers. Pediatr Infect Dis J 106 Patel D, Bland R, Coovadia H, Rollins N, Coutsoudis A,
2011; 30: 325–30. Newell ML. Breastfeeding, HIV status and weights in South
85 Madhi SA, Adrian P, Cotton MF, et al. Effect of HIV infection status African children: a comparison of HIV-exposed and unexposed
and anti-retroviral treatment on quantitative and qualitative antibody children. AIDS 2010; 24: 437–45.
responses to pneumococcal conjugate vaccine in infants. J Infect Dis 107 Arpadi S, Fawzy A, Aldrovandi GM, et al. Growth faltering due to
2010; 202: 355–61. breastfeeding cessation in uninfected children born to
86 Simani OE, Adrian PV, Violari A, et al. Effect of in-utero HIV exposure HIV-infected mothers in Zambia. Am J Clin Nutr 2009;
and antiretroviral treatment strategies on measles susceptibility and 90: 344–53.
immunogenicity of measles vaccine. AIDS 2013; 27: 1583–91. 108 Ross A, Raab GM, Mok J, Gilkison S, Hamilton B, Johnstone FD.
87 Abramczuk BM, Mazzola TN, Moreno YMF, et al. Impaired humoral Maternal HIV infection, drug use, and growth of uninfected
response to vaccines among HIV-exposed uninfected infants. children in their first 3 years. Arch Dis Child 1995; 73: 490–95.
Clin Vaccine Immunol 2011; 18: 1406–09. 109 Agostoni C, Zuccotti GV, Giovannini M, et al. Growth in the first
88 Elliott AM, Mawa PA, Webb EL, et al. Effects of maternal and infant two years of uninfected children born to HIV-1 seropositive
co-infections, and of maternal immunisation, on the infant response mothers. Arch Dis Child 1998; 79: 175–78.
to BCG and tetanus immunisation. Vaccine 2010; 29: 247–55. 110 Neri D, Somarriba GA, Schaefer NN, et al. Growth and body
89 Garcia-Knight MA, Nduati E, Hassan AS, et al. Altered memory T-Cell composition of uninfected children exposed to human
responses to Bacillus Calmette-Guerin and tetanus toxoid vaccination immunodeficiency virus: comparison with a contemporary cohort
and altered cytokine responses to polyclonal stimulation in and United States National Standards. J Pediatr 2013; 163: 249–54.
HIV-exposed uninfected Kenyan infants. PLoS One 2015; 10: e0143043. 111 Le Doaré K, Bland R, Newell ML. Neurodevelopment in children
90 Sanz-Ramos M, Manno D, Kapambwe M, et al. Reduced poliovirus born to HIV-infected mothers by infection and treatment status.
vaccine neutralising-antibody titres in infants with maternal Pediatrics 2012; 130: e1326–44.
HIV-exposure. Vaccine 2013; 31: 2042–49. 112 Gómez C, Archila ME, Rugeles C, Carrizosa J, Rugeles MT,
91 Hesseling AC, Jaspan HB, Black GF, Nene N, Walzl G. Cornejo JW. A prospective study of neurodevelopment of
Immunogenicity of BCG in HIV-exposed and non-exposed infants uninfected children born to human immunodeficiency virus type
following routine birth or delayed vaccination. Int J Tuberc Lung Dis 1 positive mothers. Rev Neurol 2009; 48: 287–91 (in Spanish).
2015; 19: 454–62. 113 Kandawasvika GQ, Kuona P, Chandiwana P, et al. The burden
92 Van Rie A, Madhi SA, Heera JR, et al. Gamma interferon production in and predictors of cognitive impairment among 6- to 8-year-old
response to Mycobacterium bovis BCG and Mycobacterium tuberculosis children infected and uninfected with HIV from Harare,
antigens in infants born to human immunodeficiency virus-infected Zimbabwe: a cross-sectional study. Child Neuropsychol 2015;
mothers. Clin Vaccine Immunol 2006; 13: 246–52. 21: 106–20.
93 Mansoor N, Scriba TJ, de Kock M, et al. HIV-1 infection in infants 114 Nicholson L, Chisenga M, Siame J, Kasonka L, Filteau S.
severely impairs the immune response induced by Bacille Growth and health outcomes at school age in HIV-exposed,
Calmette-Guérin vaccine. J Infect Dis 2009; 199: 982–90. uninfected Zambian children: follow-up of two cohorts studied in
94 Mazzola TN, da Silva MT, Abramczuk BM, et al. Impaired Bacillus infancy. BMC Pediatr 2015; 15: 66.
Calmette-Guérin cellular immune response in HIV-exposed, 115 Kerr SJ, Puthanakit T, Vibol U, et al. Neurodevelopmental
uninfected infants. AIDS 2011; 25: 2079–87. outcomes in HIV-exposed-uninfected children versus those not
95 Siberry GK, Patel K, Bellini WJ, et al. Immunity to measles, mumps, exposed to HIV. AIDS Care 2014; 26: 1327–35.
and rubella in US children with Perinatal HIV infection or perinatal 116 Jahanshad N, Couture MC, Prasitsuebsai W, et al. Brain imaging
HIV exposure without infection. Clin Infect Dis 2015; 61: 988–95. and neurodevelopment in HIV-uninfected Thai children born to
96 Kim D, Huey D, Oglesbee M, Niewiesk S. Insights into the regulatory HIV-infected mothers. Pediatr Infect Dis J 2015; 34: e211–16.
mechanism controlling the inhibition of vaccine-induced 117 UNAIDS. Global report: UNAIDS report on the global AIDS
seroconversion by maternal antibodies. Blood 2011; 117: 6143–51. epidemic 2013. Geneva: Joint United Nations Programme on
97 Henderson RA, Miotti PG, Saavedra JM, et al. Longitudinal growth HIV/AIDS (UNAIDS), 2013. http://www.unaids.org/sites/default/
during the first 2 years of life in children born to HIV-infected files/media_asset/UNAIDS_Global_Report_2013_en_1.pdf
mothers in Malawi, Africa. Pediatr AIDS HIV Infect 1996; 7: 91–97. (accessed July 28, 2015).
98 Bailey RC, Kamenga MC, Nsuami MJ, Nieburg P, St Louis ME. 118 Quinn TC. Global burden of the HIV pandemic. Lancet 1996;
Growth of children according to maternal and child HIV, 348: 99–106.
immunological and disease characteristics: a prospective cohort study 119 Ronsmans C, Chowdhury ME, Dasgupta SK, Ahmed A,
in Kinshasa, Democratic Republic of Congo. Int J Epidemiol 1999; Koblinsky M. Effect of parent’s death on child survival in rural
28: 532–40. Bangladesh: a cohort study. Lancet 2010; 375: 2024–31.
99 Sherry B, Embree JE, Mei Z, et al. Sociodemographic characteristics, 120 Joint United Nations Programme on HIV/AIDS, UNICEF, and
care, feeding practices, and growth of cohorts of children born to HIV- the United States Agency for International Development
1 seropositive and seronegative mothers in Nairobi, Kenya. (USAID). Children on the brink 2004: a joint report of new
Trop Med Int Health 2000; 5: 678–86. orphan estimates and a framework for action. New York, NY: Joint
100 Lepage P, Msellati P, Hitimana DG, et al. Growth of human United Nations Programme on HIV/AIDS, the United Nations
immunodeficiency type 1-infected and uninfected children: Children’s Fund, and the United States Agency for International
a prospective cohort study in Kigali, Rwanda, 1988 to 1993. Development, 2004. http://www.unicef.org/publications/cob_
Pediatr Infect Dis J 1996; 15: 479–85. layout6-013.pdf (accessed July 29, 2015).
101 Osterbauer B, Kapisi J, Bigira V, et al. Factors associated with malaria 121 Kakkar F, Lamarre V, Ducruet T, et al. Impact of maternal HIV-1
parasitaemia, malnutrition, and anaemia among HIV-exposed and viremia on lymphocyte subsets among HIV-exposed uninfected
unexposed Ugandan infants: a cross-sectional survey. Malar J 2012; infants: protective mechanism or immunodeficiency.
11: 432. BMC Infect Dis 2014; 14: 236.

www.thelancet.com/infection Vol 16 June 2016 e106

 Review

122 Gupta RK, Lawn SD, Bekker LG, Caldwell J, Kaplan R, Wood R. 143 Groome MJ, Page N, Cortese MM, et al. Effectiveness of monovalent
Impact of human immunodeficiency virus and CD4 count on human rotavirus vaccine against admission to hospital for acute
tuberculosis diagnosis: analysis of city-wide data from Cape Town, rotavirus diarrhoea in South African children: a case-control study.
South Africa. Int J Tuberc Lung Dis 2013; 17: 1014–22. Lancet Infect Dis 2014; 14: 1096–104.
123 Slyker JA, Rowland-Jones SL, Dong T, et al. Acute cytomegalovirus 144 Habib NA, Daltveit AK, Bergsjo P, Shao J, Oneko O, Lie RT.
infection is associated with increased frequencies of activated and Maternal HIV status and pregnancy outcomes in northeastern
apoptosis-vulnerable T cells in HIV-1-infected infants. J Virol 2012; Tanzania: a registry-based study. BJOG 2008; 115: 616–24.
86: 11373–79. 145 Nunes MC, Cutland CL, Dighero B, et al. Kinetics of
124 Schramm DB, Kuhn L, Gray GE, Tiemessen CT. In vivo effects of hemagglutination-inhibiting antibodies following maternal influenza
HIV-1 exposure in the presence and absence of single-dose nevirapine vaccination among mothers with and those without HIV infection and
on cellular plasma activation markers of infants born to their infants. J Infect Dis 2015; 212: 1976–87.
HIV-1-seropositive mothers. J Acquir Immune Defic Syndr 2006; 146 Baroncelli S, Galluzzo CM, Mancinelli S, et al. Antibodies against
42: 545–53. pneumococcal capsular polysaccharide in Malawian HIV-positive
125 Nazli A, Chan O, Dobson-Belaire WN, et al. Exposure to HIV-1 directly mothers and their HIV-exposed uninfected children. Infect Dis 2015;
impairs mucosal epithelial barrier integrity allowing microbial 48: 317–21.
translocation. PLoS Pathog 2010; 6: e1000852. 147 Hesseling AC, Johnson LF, Jaspan H, et al. Disseminated bacille
126 Townsend CL, Tookey PA, Newell ML, Cortina-Borja M. Calmette-Guérin disease in HIV-infected South African infants.
Antiretroviral therapy in pregnancy: balancing the risk of preterm Bull World Health Organ 2009; 87: 505–11.
delivery with prevention of mother-to-child HIV transmission. 148 Fawzy A, Arpadi S, Kankasa C, et al. Early weaning increases
Antivir Ther 2010; 15: 775–83. diarrhea morbidity and mortality among uninfected children born to
127 Barret B, Tardieu M, Rustin P, et al. Persistent mitochondrial HIV-infected mothers in Zambia. J Infect Dis 2011; 203: 1222–30.
dysfunction in HIV-1-exposed but uninfected infants: clinical 149 Kafulafula G, Hoover DR, Taha TE, et al. Frequency of gastroenteritis
screening in a large prospective cohort. AIDS 2003; 17: 1769–85. and gastroenteritis-associated mortality with early weaning in
128 Sibiude J, Mandelbrot L, Blanche S, et al. Association between prenatal HIV-1-uninfected children born to HIV-infected women in Malawi.
exposure to antiretroviral therapy and birth defects: an analysis of the J Acquir Immune Defic Syndr 2010; 53: 6–13.
ANRS CO1/CO11 French Perinatal Cohort Study. PLoS Med 2014; 150 Onyango-Makumbi C, Bagenda D, Mwatha A, et al. Early weaning of
11: e1001635. HIV-exposed uninfected infants and risk of serious gastroenteritis:
129 Sibiude J, Le Chenadec J, Bonnet D, et al. In utero exposure to findings from two perinatal HIV prevention trials in Kampala,
zidovudine and heart anomalies in the ANRS French perinatal cohort Uganda. J Acquir Immune Defic Syndr 2010; 53: 20–27.
and the nested PRIMEVA randomized trial. Clin Infect Dis 2015; 151 Homsy J, Moore D, Barasa A, et al. Breastfeeding, mother-to-child
61: 270–80. HIV transmission, and mortality among infants born to HIV-Infected
130 Thorne C, Newell ML. Safety of agents used to prevent mother-to-child women on highly active antiretroviral therapy in rural Uganda.
transmission of HIV: is there any cause for concern? Drug Saf 2007; J Acquir Immune Defic Syndr 2010; 53: 28–35.
30: 203–13. 152 Taha TE, Hoover DR, Chen S, et al. Effects of cessation of
131 Wang L, Kourtis AP, Ellington S, Legardy-Williams J, Bulterys M. breastfeeding in HIV-1-exposed, uninfected children in Malawi.
Safety of tenofovir during pregnancy for the mother and fetus: Clin Infect Dis 2011; 53: 388–95.
a systematic review. Clin Infect Dis 2013; 57: 1773–81. 153 Ásbjörnsdóttir KH, Slyker JA, Maleche-Obimbo E, et al. Breastfeeding
132 Siberry GK, Williams PL, Mendez H, et al. Safety of tenofovir use is associated with decreased risk of hospitalization among
during pregnancy: early growth outcomes in HIV-exposed uninfected HIV-exposed, uninfected Kenyan infants. J Hum Lact 2015; published
infants. AIDS 2012; 26: 1151–59. online Sept 30, 2015. DOI:10.1177/0890334415607854
133 Powis KM, Smeaton L, Hughes MD, et al. In-utero triple antiretroviral 154 Ásbjörnsdóttir KH, Slyker JA, Weiss NS, et al. Breastfeeding is
exposure associated with decreased growth among HIV-exposed associated with decreased pneumonia incidence among HIV-exposed,
uninfected infants in Botswana. AIDS 2016; 30: 211–20. uninfected Kenyan infants. AIDS 2013; 27: 2809–15.
134 Mermin J, Were W, Ekwaru JP, et al. Mortality in HIV-infected 155 Coutsoudis A, Pillay K, Spooner E, Coovadia HM, Pembrey L,
Ugandan adults receiving antiretroviral treatment and survival of their Newell ML. Morbidity in children born to women infected with
HIV-uninfected children: a prospective cohort study. Lancet 2008; human immunodeficiency virus in South Africa: does mode of
371: 752–59. feeding matter? Acta Paediatr 2003; 92: 890–95.
135 WHO. Guidelines on co-trimoxazole prophylaxis for HIV-related 156 Bork KA, Cournil A, Read JS, et al. Morbidity in relation to feeding
infections among children, adolescents and adults: recommendations mode in African HIV-exposed, uninfected infants during the first
for a public health approach. Geneva: World Health Organization, 6 mo of life: the Kesho Bora study. Am J Clin Nutr 2014; 100: 1559–68.
2006. http://www.who.int/hiv/pub/guidelines/ctxguidelines.pdf?ua=1 157 WHO. Guidelines on HIV and infant feeding. Principles and
(accessed July 29, 2015). recommendations for infant feeding in the context of HIV and a
136 Ford N, Shubber Z, Jao J, Abrams EJ, Frigati L, Mofenson L. Safety of summary of evidence. Geneva: World Health Organization, 2010.
cotrimoxazole in pregnancy: a systematic review and meta-analysis. http://apps.who.int/iris/bitstream/10665/44345/1/9789241599535_
J Acquir Immune Defic Syndr 2014; 66: 512–21. eng.pdf (accessed Dec 23, 2015).
137 Dryden-Peterson S, Jayeoba O, Hughes MD, et al. Cotrimoxazole 158 Pedersen SH, Wilkinson AL, Andreasen A, et al. Longitudinal analysis
prophylaxis and risk of severe anemia or severe neutropenia in of mature breastmilk and serum immune composition among mixed
HAART-exposed, HIV-uninfected infants. PLoS One 2013; 8: e74171. HIV-status mothers and their infants. Clin Nutr 2015; published online
138 Dow A, Kayira D, Hudgens M, et al. Effects of cotrimoxazole June 3. DOI:10.1016/j.clnu.2015.05.016.
prophylactic treatment on adverse health outcomes among 159 Muhangi L, Lule SA, Mpairwe H, et al. Maternal HIV infection and
HIV-exposed, uninfected infants. Pediatr Infect Dis J 2012; 31: 842–47. other factors associated with growth outcomes of HIV-uninfected
139 Davis NL, Barnett EJ, Miller WC, et al. Impact of daily cotrimoxazole infants in Entebbe, Uganda. Public Health Nutr 2013; 16: 1548–57.
on clinical malaria and asymptomatic parasitemias in HIV-exposed, 160 Rollins N, Mahy M, Becquet R, Kuhn L, Creek T, Mofenson L.
uninfected infants. Clin Infect Dis 2015; 61: 368–74. Estimates of peripartum and postnatal mother-to-child transmission
140 Sandison T, Homsy J, Arinaitwe E, et al. Protective efficacy of probabilities of HIV for use in Spectrum and other population-based
co-trimoxazole prophylaxis against malaria in HIV exposed children in models. Sex Transm Infect 2012; 88 (suppl 2): i44–51.
rural Uganda: a randomised clinical trial. BMJ 2011; 342: d1617. 161 UNICEF. The Double Dividend: Action to improve survival of HIV
141 Coutsoudis A, Pillay K, Spooner E, Coovadia HM, Pembrey L, ‘exposed’ children in the era of eMTT and renewed child survival
Newell ML. Routinely available cotrimoxazole prophylaxis and campaigns. Cape Town: United Nations Children’s Emergency Fund,
occurrence of respiratory and diarrhoeal morbidity in infants born to 2013. http://www.unicef.org/aids/files/Action_Framework_Final.pdf
HIV-infected mothers in South Africa. S Afr Med J 2005; 95: 339–45. (accessed July 29, 2015).
142 Madhi SA, Nachman S, Violari A, et al. Primary isoniazid prophylaxis
against tuberculosis in HIV-exposed children. N Engl J Med 2011;
365: 21–31.

e107 www.thelancet.com/infection Vol 16 June 2016