University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Breastfeeding, mother-to-child HIV transmission, and mortality among infants born to HIV-infected women on highly active antiretroviral therapy in rural Uganda
Global Health Sciences Literature Digest
Published June 03, 2010
Journal Article

Homsy J, Moore D, Barasa A. Breastfeeding, mother-to-child HIV transmission, and mortality among infants born to HIV-infected women on highly active antiretroviral therapy in rural Uganda. J Acquir Immune Defic Syndr. 2010 Jan 1;53(1):28-35.

In Context

Recommendations for the prevention of postnatal transmission of HIV during breastfeeding in resource-constrained areas remains controversial. Studies have shown that in these countries, maternal use of highly active antiretroviral therapy (HAART) can reduce but does not eliminate late postnatal HIV transmission.(1, 2, 3, 4, 5)


To assess the rate of mother-to-child transmission (MTCT) of HIV and mortality based upon duration of breastfeeding among a cohort of women whose HIV-infection was being treated with HAART


Tororo, a rural district in Uganda

Study Design

Prospective cohort study


Infants born to HIV-infected women enrolled in a randomized controlled trial (RTC) of clinical and laboratory strategies for monitoring HAART between March 1, 2003 and January 1, 2007


Infant mortality, impact of breastfeeding on the probability of survival


HIV-infected women enrolled in the study were eligible for HAART if their CD4 counts were <250 cells/µL or if they met the World Health Organization (WHO) criteria for clinical stage III or IV disease. The women were visited at home weekly by lay health workers who monitored clinical status and provided HAART and cotrimoxazole. Weight and pregnancy status were monitored monthly. CD4 counts and viral loads were measured every three months.

All women in the study who delivered live-born infants between March 1, 2003 and January 1, 2007 were followed. Pregnant women were provided with PMTCT counseling and antiretroviral prophylaxis. Women were provided with 2 mg/kg of nevirapine (NVP) syrup to be administered to their infants within 72 hours of birth. After September 2005, women were instructed to administer zidovudine (AZT) at a dose of 4 mg/kg bid for seven days postpartum or 28 days if the mother initiated HAART less than one month before giving birth. All women were counseled to breastfeed their infants exclusively for three to six months followed by rapid weaning and safe replacement feeding. Infants were followed until July 1, 2007.

Testing of infants by HIV DNA polymerase chain reaction (PCR) was offered to all mothers at six weeks postpartum. Breastfed infants were tested quarterly until at least six weeks after weaning was completed. Children younger than 18 months were tested for antibodies and for HIV DNA/RNA by PCR, and children aged 18 months and older were antibody tested. Specimen collection for PCR testing required that the mother bring the infant to the Tororo clinic. Breastfeeding status type and date of cessation were recorded at the time of the regular home visits. Infant and maternal deaths were also recorded during these home visits.


There were 118 infants born to 102 women. Median follow-up time was 18.1 months (interquartile range [IQR] 9.3-26.0); 84% of the children living at the end of follow-up were followed for 12 months or more. Mothers' median time on HAART was 20.3 months (IQR 12.1-28.2). Ninety-two percent of the women breastfed exclusively for a median of 4 months (IQR 3.0-6.0) and weaned their infants at a median age of five months (IQR 3.0-7.0). There were 114 infants (97%) tested for HIV DNA or RNA PCR and none had positive test results. Four infants died before HIV testing: one had a negative antibody test four days after weaning at 19 days and the other three had positive antibody tests before weaning or had never been breastfed.

Eighty-three percent (98/118) of the PCR test results were final. Twenty-three (19%) of the infants died during follow-up and none were found to have HIV, although 12 infants and the infant with the negative antibody test did not have final negative results. Of the 23 deceased infants, three were never breastfed, 13 died after a median time of 6.7 weeks (IQR 3-14.5) after weaning, five died during the period that they were breastfed, and two died during the period that they were mixed-fed.

The probability of survival was significantly lower among weaned or never breastfed infants compared with infants continuing to breastfeed at follow-up. Breastfeeding was strongly and independently predictive of survival. In the Cox proportional model with duration of breastfeeding analyzed as a continuous variable, each additional month of breastfeeding reduced the infant mortality risk by 24% (adjusted hazard ration [HR] 0.71, 95% CI: 0.57-0.87). The risk of death associated with the duration of breastfeeding was also examined using categorical variables with categories ranging from four to seven months), the risk of death was six-fold higher for infants who breastfed for less than six months (adjusted HR 6.19, 95% CI: 1.41-27.0) compared to breastfeeding for more than six months.


Maternal HAART provided strong protection against ante- and postnatal HIV infection but did not reduce infant mortality, much of which was due to reduced duration of breastfeeding.

Quality rating

This was a good study based upon the Newcastle-Ottawa Quality Assessment scale. A strength of the study was the methods used to monitor and follow the mothers and infants and to obtain feeding histories. The study did not provide information on differences in the characteristics of mothers based on infant feeding practices, did not provide information on representativeness of the cohort, and did not have definitive test results for all infants.

Programmatic implications

This study provides additional support for recommendations for women to breastfeed their infants for six months. Efforts to diagnose and treat HIV infection in HAART-eligible women and to provide ART prophylaxis to all HIV-infected pregnant women should remain a high priority. Continued research into effective methods to reduce post-natal HIV transmission also should remain a high priority.


  1. Thomas T, Masaba R, Ndivo R, et al. Prevention of mother-to-child transmission of HIV-1 among breastfeeding mothers using HAART: The Kisumu Breastfeeding Study, Kisumu, Kenya, 2003-2007. Presented at: 15th Conference on Retroviruses and Opportunistic Infections; February 3-6, 2007; Boston, MA. Session 14, Abstract 45aLB.
  2. Kilewo C, Karlsson K, Ngarina M, et al. Prevention of mother to child transmission of HIV-1 through breastfeeding by treating mothers prophylactically with triple antiretroviral therapy in Dar es Salaam, Tanzania - the MITRA PLUS study. Presented at: 4th International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention, Sydney, Australia, July 22-25, 2007. Abstract TuAX101.
  3. Towne-Gold B, Ekouevi DK, Viho I, et al. Antiretroviral treatment and prevention of peripartum and postnatal HIV transmission in West Africa: evaluation of a two-tiered approach. PLoS Med. 2007;4:1362-73. Abstract not available.
  4. Ekouevi D, Coffie P, Benaud B, et al. Antiretroviral therapy in pregnant women with advanced HIV disease and pregnancy outcomes in Abidjan, Cote d'Ivoire. AIDS. 2008;22:1815-20.
  5. Taha T, Kumwenda J, Kafulafula G, et al. Maternal highly active antiretroviral treatment (HAART): does it improve child survival? Presented at: 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention; July 19-2, 2009; Cape Town, South Africa.