Background:  HAART is a highly effective strategy for preventing mother-to-child HIV transmission (PMTCT), but may lead to lower birth-weight infants. Longitudinal effects of in utero HAART exposure on infant growth have not been reported.

Methods:  The Mashi and Mma Bana PMTCT intervention trials enrolled HIV-infected pregnant women (regardless of CD4) at 4 sites in southern Botswana. Infant weight and length were evaluated at birth and monthly for 6 months. Only breastfed, HIV-uninfected infants born ≥37 weeks and exposed in utero to at least 2 weeks of either HAART or zidovudine (ZDV) were included in the current analysis. Infants in the HAART-exposed group received ZDV for 1 month. Infants in the ZDV-exposed group received 6 months of ZDV-prophylaxis during breastfeeding. Gender-based weight-for-age, length-for-age, and weight-for-length z-scores were calculated using WHO Child Growth Standards. Mean z-scores were compared using the Student’s t-test and analysis of response profiles. 
Results:  We included 437 ZDV-exposed infants from Mashi, and 592 HAART-exposed infants from Mma Bana. Median maternal baseline CD4 counts were 393 cells/mm³ and 337 cells/mm³ respectively, and demographics were generally similar between cohorts. Median in utero ZDV exposure was 5.7 weeks (range 2.0 to 10.9 weeks), and median in utero HAART exposure was 12.1 weeks (range 2.6 to 22.3 weeks). Median birth weights were 3.1 kg in ZDV-exposed and 3.0 kg in HAART-exposed (P <0.001), with significantly lower mean weight-for-age, length-for-age, and weight-for-length z-scores among HAART-exposed infants (P <0.001, P = 0.02, and P = 0.007, respectively). By 3 months, median weight no longer differed by exposure group, and weight remained similar through 6 months. Mean weight-for-age differed over time by exposure group (P <0.001) (Figure 1). Length-for-age remained lower in the HAART-exposed group through 6 months, but weight-for-length improved significantly over time compared with ZDV-exposed infants (P <0.001) (Figures 2, 3). The proportions of infants with z-scores >2 standard deviations below the mean were not different between exposure groups.

Conclusions:  Exposure in utero to HAART was associated with lower birth weight than with ZDV exposure alone, but this difference resolved by 3 months. Length-for-age remained lower over time in the HAART-exposed group. Longer-term developmental comparisons are planned, but these results demonstrate a reassuring early correction of the birth weight reduction presumed to be associated with in utero HAART exposure. 

	Figure 1		Figure 2
					Figure 3