Growth in late infancy among HIV-exposed children in urban Haiti is associated with participation in a clinic-based infant feeding support intervention

Abstract

The integration of nutrition support for infants of HIV-infected mothers is a recognized need; however, the evidence for effective programmatic solutions is weak. The objective of our study was to implement and evaluate a new infant feeding support intervention for HIV-exposed, uninfected, non-breast-fed infants 6-12 mo of age attending the Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes (GHESKIO) pediatric clinic in Port-au-Prince, Haiti. The 24-wk intervention included a lipid-based nutrient supplement, education, promotion of existing clinical services, and social support. We compared growth outcomes among intervention participants (n = 73) at start (wk 0) and end (wk 24) of intervention to a historical control group of HIV-exposed infants seen at the GHESKIO in the year prior to the intervention who would have met the intervention entrance criteria (n = 294). The intervention and historical control groups did not differ significantly at age 6 mo (wk 0). At age 12 mo (wk 24), the intervention group had a lower prevalence of underweight and stunting than the historical control group (weight-for-age Z-score < -2 SD: 6.8 vs. 20.8%, P = 0.007; length-for-age Z-score < -2 SD: 9.6 vs. 21.2%, P = 0.029). Wasting tended to be lower in the intervention group than the historical control (weight-for-length Z-score < -2 SD: 2.9 vs. 8.9%, P = 0.11). Implementation of the intervention was associated with reduced risk of growth faltering in HIV-exposed uninfected children from 6 to 12 mo of age. This is a promising intervention model that can be adapted and scaled-up to other HIV care contexts.

Trial registration: ClinicalTrials.gov NCT01434238.

FIGURE 1

Prevalence of HIV-exposed children being underweight (A), stunted (B), or wasted (C) in intervention and historical control groups at 6 and 12 mo of age. Intervention group, n = 73. Historical control at age 6 mo, n = 138 and at age 12 mo, n = 219. Age range of children at age 6 mo was 5.5–6.76 mo and at 12 mo was 11.0–12.25 mo. *Groups differ, P < 0.05.

FIGURE 2

LAZ by age predicted using unadjusted multilevel models for change in children in the intervention and historical control groups. *Group × age, P < 0.001. LAZ, length-for-age Z-score.