Share this article on:

Breastfeeding is associated with decreased pneumonia incidence among HIV-exposed, uninfected Kenyan infants

Ásbjörnsdóttir, Kristjana H.a; Slyker, Jennifer A.b; Weiss, Noel S.a; Mbori-Ngacha, Dorothyd; Maleche-Obimbo, Elizabethd; Wamalwa, Daltond; John-Stewart, Gracea,b,c,e

doi: 10.1097/01.aids.0000432540.59786.6d
Epidemiology and Social: CONCISE COMMUNICATION

Objective: HIV-exposed uninfected (HEU) infants have higher infectious disease morbidity and mortality than unexposed infants. We determined the incidence and risk factors for pneumonia, a leading cause of infant mortality worldwide, in a cohort of HEU infants. Identifying predictors of pneumonia among HEU infants may enable early identification of those at highest risk.

Design: A retrospective cohort of HEU infants participating in a Kenyan perinatal HIV study, enrolled between 1999 and 2002.

Methods: Infants were followed monthly from birth to 12 months. Incidence of pneumonia diagnosed at monthly study visits, sick-child visits or by means of averbal autopsy was estimated with a 14-day window for new episodes. Cox proportional hazards regression was used to identify predictors of first pneumonia occurrence.

Results: Among 388 HEU infants with 328 person-years of follow-up, the incidence of pneumonia was 900/1000 child-years [95% confidence interval (CI) 800–1000]. Maternal HIV viral load at 32 weeks’ gestation [hazard ratio 1.2 (1.0–1.5) per log10 difference] and being underweight (weight-for-age Z-score <−2) at the previous visit [hazard ratio 1.8 (1.1–2.8)] were associated with increased risk of pneumonia. Breastfed infants had a 47% lower risk of pneumonia than those never breastfed [hazard ratio 0.53 (0.39–0.73)], independent of infant growth, maternal viral load and maternal CD4%. Breastfeeding was also associated with a 74% lower risk of pneumonia-related hospitalization [hazard ratio 0.26 (0.13–0.53)].

Conclusions: The incidence of pneumonia in this cohort of HEU infants was high. Our observations suggest that maternal viral suppression and breastfeeding may reduce the burden of pneumonia among HEU infants.

aDepartment of Epidemiology

bDepartment of Global Health

cDepartment of Pediatrics, University of Washington, Seattle, Washington USA

dDepartment of Paediatrics & Child Health, University of Nairobi, Nairobi, Kenya

eDepartment of Medicine, University of Washington, Seattle, Washington USA.

Correspondence to Kristjana Ásbjörnsdóttir, Harborview Medical Center, 325 9th Ave, Box 359931, Seattle, WA 98104, USA. Tel: +1 650 799 6700; fax: +1 206 543 4313; e-mail:

Received 18 March, 2013

Revised 23 June, 2013

Accepted 27 June, 2013

Back to Top | Article Outline


As increasingly effective measures to prevent infant HIV infection are implemented, the population of HIV-exposed uninfected (HEU) infants is growing. HEU infants have higher mortality and morbidity than unexposed infants [1,2]; this may be due to increased exposure to infectious pathogens, impaired maternal care-giving, reduced breastfeeding, or compromised immune development [3].

Pneumonia is a leading cause of child mortality [4]. Annual incidence of pneumonia in the developing world among children under 5 is estimated at 29/100 [4,5], with about 10% requiring hospitalization [4]. Approximately 50% of childhood pneumonias are caused by Haemophilus influenzae or Streptococcus pneumoniae [4]. Viral and mixed viral–bacterial infections are also common causes of childhood pneumonia in developing countries [6,7]. HIV-infected children are at particularly high risk for pneumonia [6,8]; in a perinatal HIV cohort in Kenya, we previously found pneumonia to be the leading cause of mortality among both HEU [9] and HIV-infected infants [10].

Elucidation of cofactors for pneumonia in HEU may inform interventions to prevent pneumonia or identification of mother–infant pairs at highest risk. In this study, we describe the incidence and predictors for pneumonia in a cohort of Kenyan HEU infants.

Back to Top | Article Outline


Participant selection

Mother–infant pairs were enrolled in a study of HIV transmission in Nairobi, Kenya, between 1999 and 2002, details of which have been described elsewhere [9,11]. HEU infants were followed until 12 months of age. Singletons and first-born twins were eligible for this analysis if they were confirmed HIV-uninfected at birth. Infants were excluded if they died or were lost to follow-up (LTFU) without completing a study visit, or if they were HIV-infected in utero or during delivery.

Back to Top | Article Outline

Study procedures

The study was approved by the Institutional Review Board at the University of Washington and the Ethics and Research Committee at Kenyatta National Hospital. Sociodemographic information was collected at enrollment, and maternal CD4+ T-cell percentage (CD4%) and HIV viral load were measured at 32 weeks’ gestation. Mothers received antenatal short-course zidovudine for prevention of mother-to-child transmission (PMTCT), but no antiretroviral therapy (ART) during breastfeeding, as per contemporary Kenyan guidelines. Study visits occurred at birth, 2 weeks after delivery and monthly thereafter.

Back to Top | Article Outline

Clinical data collection

At scheduled visits, infants underwent detailed clinical examination by a physician. Symptoms, diagnoses, history of illness and hospitalization since the last visit, and current feeding practices were collected on a standardized form. At sick-child visits, examination and history were limited to features related to the infant's reporting complaint. For infants who died outside hospital, verbal autopsy was conducted to determine the most likely cause of death.

Back to Top | Article Outline


Pneumonia was diagnosed according to Integrated Management of Childhood Illness (IMCI) guidelines. Diagnoses made at study visits, sick-child visits, or recorded as cause of death were included in this analysis. To rule out data entry error, diagnoses recorded at study visits were checked for the presence of IMCI pneumonia symptoms. To avoid double-counting an unresolved episode, new diagnoses were not counted for 14 days following each diagnosis.

Back to Top | Article Outline

Statistical analysis

Analyses were conducted using Stata 11.2 (StataCorp, College Station, Texas, USA). Infant growth was assessed using the WHO Child Growth Standards Stata igrowup package.

Incidence of all pneumonia, in the whole cohort and within strata of predictors, was calculated with censoring at non-pneumonia death, LTFU, last negative HIV-DNA before first HIV-DNA detection in HIV-infected infants, or study exit at 12 months. Infants did not contribute person-time for 14 days following a pneumonia diagnosis. Incidence of pneumonia-related hospitalization (allowing repeat hospitalizations) and pneumonia-related mortality was calculated the same way.

Predictors of first pneumonia diagnosis were identified using Cox proportional hazards regression with censoring as above. All exposures were assessed at baseline except maternal viral load, CD4%, breastfeeding, and weight-for-age Z-score (WAZ). Viral load and CD4% were measured at 32 weeks’ gestation; breastfeeding (ever/never, current) and WAZ were assessed at each study visit and analyzed as time-varying covariates.

In order to address whether associations with breastfeeding were mediated through infant growth, breastfeeding and lagged WAZ were included in the same model. Other models of the association between breastfeeding and pneumonia were adjusted for maternal viral load or CD4% to assess confounding by advanced maternal disease.

Back to Top | Article Outline


Study population

Three hundred and eighty-eight infants were included in this analysis; 365 remained uninfected, whereas 23 acquired HIV at a median of 33 days [interquartile range (IQR) 29–180]. Eighty-five infants were LTFU at a median of 243 days (IQR 105–320). Mothers were mostly young (median 25 years, IQR 22–28), married (90%), lived in one-room houses (78%), and lacked an independent income (70%). At 32 weeks’ gestation, mothers’ median CD4+ cell count was 438 cells/μl (IQR 310–611), median CD4% was 24 (IQR 18–30), and median viral load was 4.7 log10 copies/ml (IQR 4.1–5.2). Seventy-six percent of mothers breastfed for a median of 273 days (IQR 123–364).

Back to Top | Article Outline

Pneumonia burden

Infants completed 328 person-years of follow-up. One hundred and seventy-seven infants were diagnosed with pneumonia (Fig. 1a) and experienced a total of 296 episodes. Pneumonia incidence was 900/1000 infant-years [95% confidence interval (CI) 800–1000] and did not differ significantly between 3-month intervals of age. Median age at first pneumonia was 127 days (IQR 60–211). Thirty-three infants were hospitalized a total of 41 times for pneumonia, corresponding to 110 hospitalizations/1000 infant-years (95% CI 80–160). Median age at first hospitalization was 150 days (IQR 63–227). Ten infants died with pneumonia at a median age of 116 days (IQR 70–149), for a total pneumonia-related mortality rate of 30/1000 infant-years (95% CI 16–57).



Back to Top | Article Outline

Predictors of first pneumonia

Infant pneumonia was not associated with maternal age, socioeconomic indicators, or maternal CD4% (Table 1). Each log10 difference in maternal viral load at 32 weeks’ gestation was associated with 20% increased pneumonia risk [hazard ratio 1.2 (1.0–1.5)]. Low birth weight and prematurity were not associated with risk of pneumonia; however, being moderately to severely underweight (WAZ <−2) at a study visit was associated with increased risk of pneumonia at the following visit [hazard ratio 1.8 (1.1–2.8)]. Higher continuous lagged WAZ was associated with decreased risk of pneumonia [hazard ratio 0.86 (0.76–0.99)].

Table 1

Table 1

Breastfed infants had a 47% lower risk of pneumonia than infants who were never breastfed [hazard ratio 0.53 (0.39–0.73); Fig. 1b]. When analyzed as a time-varying covariate, current breastfeeding was associated with a 54% reduced risk of pneumonia [hazard ratio 0.46 (0.34–0.63)]. The association between ever-breastfeeding and pneumonia was unaffected by adjustment for lagged WAZ [adjusted hazard ratio (aHR) 0.50 (0.37–0.70)] and by adjustment for maternal log10 viral load [aHR 0.51 (0.36–0.72)] or CD4% [aHR 0.52 (0.38–0.72)].

Breastfed infants had a 74% lower risk of pneumonia-related hospitalization [hazard ratio 0.26 (0.13–0.53); Fig. 1c] and a trend for lower risk of pneumonia-related mortality [hazard ratio 0.32 (0.09–1.1); Fig. 1d] compared to infants who were never breastfed.

An alternative analysis using a composite endpoint of first pneumonia or all-cause infant mortality revealed similar relationships and estimates (data not shown).

Back to Top | Article Outline


In this study, almost half of HEU infants (46%) experienced an episode of pneumonia during the first year of life. The overall incidence of pneumonia (90/100 infant-years) was substantially higher than the estimated worldwide under-5 incidence of 29/100 child-years [4,5]. Similarly, incidence of hospitalization (110/1000 child-years) was higher than that reported in population-based studies in rural Kenya, which observed hospitalization rates of 48/1000 person-years among infants [7] and 7/1000 person-years among children under 5 [12]. Pneumonia-related mortality rate in our study was 30/1000 child-years, which is also high compared to previous reports from Kenya [13]. These differences may be partly explained by the age of our cohort – rates of pneumonia are higher in the first year of life [7] – and by increased detection, access to clinical care, and referrals for hospitalization. However, it is likely that there are true differences between our cohort and the general population of Kenyan infants as a result of maternal HIV infection; high rates of infant hospitalization have been observed in previous HEU cohorts [14]. Many factors related to maternal HIV may increase infant susceptibility to pneumonia, including compromised immunity or increased household respiratory infections; in our cohort, breastfeeding avoidance was likely a key factor.

Breastfeeding was associated with lower risk of pneumonia among HEU infants in our cohort, and the association was neither mediated by infant growth nor confounded by maternal HIV disease severity. This observation is consistent with literature supporting the role of breastfeeding in the prevention of respiratory tract infections among children in general [15–17] and in reducing infectious morbidity among HEU children [14,18–20]. Breastfed infants have a larger thymus at 4 months of age than those not breastfed, have higher titers of neutralizing antibodies after vaccination, and may mount more effective immune responses [17]. IgA antibodies from breast milk have been suggested to protect against infection by neutralizing pathogens at mucosal surfaces [17,21].

Underweight infants were at 80% increased risk of pneumonia at the following visit compared to normal-weight infants, consistent with previous studies [22,23]. This may be due to increased susceptibility to infection due to malnutrition, or to increased risk of pneumonia following other acute illness [23], which is often accompanied by weight loss [23,24].

Several HEU studies have noted an association between maternal CD4+ cell count and infant hospitalization [1,14], mortality [14,25], and lower respiratory tract infection [26]. We observed an association between maternal HIV viral load and infant pneumonia, but not with CD4+ cell count, consistent with a South African study [27]. Maternal viral load could affect infant immunity to respiratory pathogens; higher viral loads during pregnancy have been associated with decreased transplacental antibody transfer [28].

Strengths of our study included detailed standardized characterization of serial morbidity and exposure history. Limitations include a lack of systematic confirmatory diagnostic techniques such as chest radiographs; specificity of pneumonia diagnosis in African settings is as low as 15–40% [29,30]. Our ability to identify predictors of pneumonia would be attenuated by misclassification of the outcome; therefore the associations we report may be underestimated. A second limitation is the potential for non-random loss to follow-up. Low birth weight [31] and prematurity [32] have been consistently found to be associated with risk of pneumonia among children in general, but we found no association. It is possible that these factors contributed to mortality before the first study visit and exclusion from this analysis. We do not have an HIV-unexposed comparison cohort and thus cannot estimate contribution of HIV exposure to incidence of pneumonia. Finally, the study was conducted prior to addition of Hib and pneumococcal vaccines to the Kenyan vaccination schedule. Although these would be expected to decrease incidence of pneumonia, approximately 50% of pneumonias are estimated to be caused by other pathogens [4,7] and the cofactors we identified likely remain relevant to HEU infants.

As the population of HEU infants increases, understanding the increased vulnerability of this population will be critical for optimizing child health. WHO 2010 guidelines recommend that HIV-infected mothers breastfeed with either extended maternal ART or infant prophylaxis [33]. Our study suggests that in addition to enhanced prevention of infant HIV and improved maternal health resulting from maternal ART, these guidelines may lead to decreased pneumonia among HEU infants, both through reducing maternal viral load and by allowing safe extended breastfeeding.

Back to Top | Article Outline


G.J.-S. conceived and acquired funding for the primary cohort study. G.J.-S., D.M.-N., and E.M.-O. developed the primary cohort and designed clinical protocols. E.M.-O. and D.W. collected clinical data and interpreted clinical results. K.H.A., J.A.S., N.S.W., and G.J.-S. participated in design and interpretation of the statistical analyses. The final manuscript was written by K.H.A. with comments and input from all authors. The authors wish to thank the women and infants who participated in this study, the CTL study peer counselors and the data, clinical and laboratory teams at the University of Nairobi and Kenyatta National Hospital, without whom this study would not have been possible. We thank the University of Nairobi and Kenyatta National Hospital for providing administrative and physical infrastructure for the study. We thank Julie Overbaugh's team at the Fred Hutchinson Cancer Research Center for HIV viral assays and the Kizazi working group [UW Global Center for Integrated Health of Women, Adolescents and Children (Global WACh)], particularly Ken Tapia and Lisa Cranmer, for their comments on the analysis and support during the preparation of this article.

Financial support: US National Institute of Child Health and Human Development (grant R01 HD-23412). K.H.A. was supported in part by NIH K24 HD054314 and by NIH Interdisciplinary Training in Cancer Research, T32 CA080416.

Back to Top | Article Outline

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline


1. Koyanagi A, Humphrey JH, Ntozini R, Nathoo K, Moulton LH, Iliff P, et al. Morbidity among human immunodeficiency virus-exposed but uninfected, human immunodeficiency virus-infected, and human immunodeficiency virus-unexposed infants in Zimbabwe before availability of highly active antiretroviral therapy. Pediatr Infect Dis J 2011; 30:45–51.
2. Slogrove A, Reikie B, Naidoo S, De Beer C, Ho K, Cotton M, et al. HIV-exposed uninfected infants are at increased risk for severe infections in the first year of life. J Trop Pediatr 2012; 58:505–508.
3. Filteau S. The HIV-exposed, uninfected African child. Trop Med Int Health 2009; 14:276–287.
4. Scott JA, Brooks WA, Peiris JS, Holtzman D, Mulholland EK. Pneumonia research to reduce childhood mortality in the developing world. J Clin Invest 2008; 118:1291–1300.
5. Rudan I, Tomaskovic L, Boschi-Pinto C, Campbell H. Global estimate of the incidence of clinical pneumonia among children under five years of age. Bull World Health Organ 2004; 82:895–903.
6. Cashat-Cruz M, Morales-Aguirre JJ, Mendoza-Azpiri M. Respiratory tract infections in children in developing countries. Semin Pediatr Infect Dis 2005; 16:84–92.
7. Berkley JA, Munywoki P, Ngama M, Kazungu S, Abwao J, Bett A, et al. Viral etiology of severe pneumonia among Kenyan infants and children. JAMA 2010; 303:2051–2057.
8. Graham SM, Coulter JB, Gilks CF. Pulmonary disease in HIV-infected African children. Int J Tuberc Lung Dis 2001; 5:12–23.
9. Gichuhi C, Obimbo E, Mbori-Ngacha D, Mwatha A, Otieno P, Farquhar C, et al. Predictors of mortality in HIV-1 exposed uninfected postneonatal infants at the Kenyatta National Hospital. Nairobi East Afr Med J 2005; 82:447–451.
10. Obimbo EM, Mbori-Ngacha DA, Ochieng JO, Richardson BA, Otieno PA, Bosire R, et al. Predictors of early mortality in a cohort of human immunodeficiency virus type 1-infected african children. Pediatr Infect Dis J 2004; 23:536–543.
11. Obimbo EM, Wamalwa D, Richardson B, Mbori-Ngacha D, Overbaugh J, Emery S, et al. Pediatric HIV-1 in Kenya: pattern and correlates of viral load and association with mortality. J Acquir Immune Defic Syndr 2009; 51:209–215.
12. Tornheim JA, Manya AS, Oyando N, Kabaka S, Breiman RF, Feikin DR. The epidemiology of hospitalized pneumonia in rural Kenya: the potential of surveillance data in setting public health priorities. Int J Infect Dis 2007; 11:536–543.
13. Ye Y, Zulu E, Mutisya M, Orindi B, Emina J, Kyobutungi C. Seasonal pattern of pneumonia mortality among under-five children in Nairobi's informal settlements. Am J Trop Med Hyg 2009; 81:770–775.
14. Kuhn L, Kasonde P, Sinkala M, Kankasa C, Semrau K, Scott N, et al. Does severity of HIV disease in HIV-infected mothers affect mortality and morbidity among their uninfected infants?. Clin Infect Dis 2005; 41:1654–1661.
15. Bachrach VR, Schwarz E, Bachrach LR. Breastfeeding and the risk of hospitalization for respiratory disease in infancy: a meta-analysis. Arch Pediatr Adolesc Med 2003; 157:237–243.
16. Mwiru RS, Spiegelman D, Duggan C, Peterson K, Liu E, Msamanga G, et al. Relationship of exclusive breast-feeding to infections and growth of Tanzanian children born to HIV-infected women. Public Health Nutr 2011; 14:1251–1258.
17. Hanson LA. Session 1: feeding and infant development breast-feeding and immune function. Proc Nutr Soc 2007; 66:384–396.
18. Fox MP, Brooks D, Kuhn L, Aldrovandi G, Sinkala M, Kankasa C, et al. Reduced mortality associated with breast-feeding-acquired HIV infection and breast-feeding among HIV-infected children in Zambia. J Acquir Immune Defic Syndr 2008; 48:90–96.
19. Kafulafula G, Hoover DR, Taha TE, Thigpen M, Li Q, Fowler MG, et al. Frequency of gastroenteritis and gastroenteritis-associated mortality with early weaning in HIV-1-uninfected children born to HIV-infected women in Malawi. J Acquir Immune Defic Syndr 2010; 53:6–13.
20. Vora N, Homsy J, Kakuru A, Arinaitwe E, Wanzira H, Sandison TG, et al. Breastfeeding and the risk of malaria in children born to HIV-infected and uninfected mothers in rural Uganda. J Acquir Immune Defic Syndr 2010; 55:253–261.
21. Oddy WH. A review of the effects of breastfeeding on respiratory infections, atopy, and childhood asthma. J Asthma 2004; 41:605–621.
22. Bejon P, Mohammed S, Mwangi I, Atkinson SH, Osier F, Peshu N, et al. Fraction of all hospital admissions and deaths attributable to malnutrition among children in rural Kenya. Am J Clin Nutr 2008; 88:1626–1631.
23. Schlaudecker EP, Steinhoff MC, Moore SR. Interactions of diarrhea, pneumonia, and malnutrition in childhood: recent evidence from developing countries. Curr Opin Infect Dis 2011; 24:496–502.
24. Bloss E, Wainaina F, Bailey RC. Prevalence and predictors of underweight, stunting, and wasting among children aged 5 and under in western Kenya. J Trop Pediatr 2004; 50:260–270.
25. Fox MP, Brooks DR, Kuhn L, Aldrovandi G, Sinkala M, Kankasa C, et al. Role of breastfeeding cessation in mediating the relationship between maternal HIV disease stage and increased child mortality among HIV-exposed uninfected children. Int J Epidemiol 2009; 38:569–576.
26. Mussi-Pinhata MM, Motta F, Freimanis-Hance L, de Souza R, Szyld E, Succi RC, et al. Lower respiratory tract infections among human immunodeficiency virus-exposed, uninfected infants. Int J Infect Dis 2010; 14 (Suppl 3):e176–182.
27. Venkatesh KK, de Bruyn G, Marinda E, Otwombe K, van Niekerk R, Urban M, et al. Morbidity and mortality among infants born to HIV-infected women in South Africa: implications for child health in resource-limited settings. J Trop Pediatr 2011; 57:109–119.
28. Farquhar C, Nduati R, Haigwood N, Sutton W, Mbori-Ngacha D, Richardson B, et al. High maternal HIV-1 viral load during pregnancy is associated with reduced placental transfer of measles IgG antibody. J Acquir Immune Defic Syndr 2005; 40:494–497.
29. Cutts FT, Zaman SM, Enwere G, Jaffar S, Levine OS, Okoko JB, et al. Efficacy of nine-valent pneumococcal conjugate vaccine against pneumonia and invasive pneumococcal disease in The Gambia: randomised, double-blind, placebo-controlled trial. Lancet 2005; 365:1139–1146.
30. Roca A, Sigauque B, Quinto L, Morais L, Berenguera A, Corachan M, et al. Estimating the vaccine-preventable burden of hospitalized pneumonia among young Mozambican children. Vaccine 2010; 28:4851–4857.
31. Kirkwood BR, Gove S, Rogers S, Lob-Levyt J, Arthur P, Campbell H. Potential interventions for the prevention of childhood pneumonia in developing countries: a systematic review. Bull World Health Organ 1995; 73:793–798.
32. Chen CH, Wen HJ, Chen PC, Lin SJ, Chiang TL, Hsieh IC, et al. Prenatal and postnatal risk factors for infantile pneumonia in a representative birth cohort. Epidemiol Infect 2012; 140:1277–1285.
33. WHOGuidelines on HIV and infant feeding: principles and recommendations for infant feeding in the context of HIV and a summary of evidence. Geneva:World Health Organization; 2010.

breastfeeding; HIV-exposed uninfected; infants; morbidity; pneumonia

© 2013 Lippincott Williams & Wilkins, Inc.