Infant birth size and early infant pneumonia predict linear growth faltering among HIV-exposed uninfected infants in Kenya.

INTRODUCTION: Stunting remains a global health priority, particularly in sub-Saharan Africa. Identifying determinants of linear growth in HIV-exposed uninfected (HEU) infants can inform interventions to prevent stunting in this vulnerable population.

METHODS: HIV-infected mothers and their uninfected infants were followed monthly from pregnancy to 12 months postpartum in Nairobi, Kenya. Mixed-effects models estimated the change in length-for-age z-score (LAZ) from birth to 12 months by environmental, maternal, and infant characteristics. Multivariable models included factors univariately associated with LAZ.

RESULTS: Among 372 HEU infants, mean LAZ decreased from -0.54 (95% confidence interval [CI]: -0.67, -0.41) to -1.09 (95% CI: -1.23, -0.96) between 0-12 months. Declines in LAZ were associated with crowding (≥2 persons/room; adjusted difference in 0-12 month change [AD]: -0.46; 95% CI: -0.87, -0.05), use of a pit latrine versus a flush toilet (AD: -0.29; 95% CI: -0.57, -0.02), and early infant pneumonia (AD: -1.14; 95% CI: -1.99, -0.29). Infants with low birthweight (<2500g; AD 1.08: 95% CI: 0.40, 1.76) and birth stunting (AD: 1.11; 95% CI: 0.45, 1.78) experienced improved linear growth. By 12 months of age, 46 infants were stunted, of whom 11 (24%) were stunted at birth. Of the 34 infants stunted at birth, with an available 12-month LAZ, 68% were not stunted at 12 months.

CONCLUSIONS: Some low birthweight and birth stunted HEU infants had significant linear growth recovery. Early infant pneumonia and household environment predicted poor linear growth and may identify a subgroup of HEU infants for whom to provide growth-promoting interventions.