Subcortical Brain Volumes and Neurocognitive Function in Children With Perinatal HIV Exposure: A Population-Based Cohort Study in South Africa

Abstract

Background. Children who are HIV-exposed and uninfected (HEU) are at risk for early neurodevelopmental impairment. Smaller basal ganglia nuclei have been reported in neonates who are HEU compared to HIV-unexposed (HU); however, neuroimaging studies outside infancy are scarce. We examined subcortical brain structures and associations with neurocognition in children who are HEU. Methods. This neuroimaging study was nested within the Drakenstein Child Health Study birth cohort in South Africa. We compared (T1-weighted) magnetic resonance imaging–derived subcortical brain volumes between children who were HEU (n = 70) and HU (n = 92) at age 2–3 years using linear regression. Brain volumes were correlated with neurodevelopmental outcomes measured with the Bayley Scales of Infant and Toddler Development III. Results. Compared to HU children, on average children who were HEU had 3% lower subcortical grey matter volumes. Analyses of individual structures found smaller volume of the putamen nucleus in the basal ganglia (-5% difference, P = .016) and the hippocampus (-3% difference, P = .044), which held on adjustment for potential confounders (P < .05). Maternal viremia and lower CD4 count in pregnancy were associated with smaller child putamen volumes. Children who were HEU had lower language scores than HU; putamen and hippocampus volumes were positively correlated with language outcomes. Conclusions. Overall, children who are HEU had a pattern of smaller subcortical volumes in the basal ganglia and hippocampal regions compared to HU children, which correlated with language function. Findings suggest that optimizing maternal perinatal HIV care is important for child brain development. Further studies are needed to investigate underlying mechanisms and long-term outcomes.