Child neurodevelopment in a resource constrained setting: a 8 year follow up of children born in a high HIV prevalence community in Zimbabwe

Abstract

To determine neurodevelopmental outcomes we conducted a cohort study prospectively from 2002 to 2011, among Zimbabwean children infected, exposed uninfected and uninfected unexposed with HIV whose mothers had participated in a national HIV prevention program in Harare. The Bayley Infant Neurodevelopmental Screener (BINS) and the McCarthy Scales of Children’s Abilities (MSCA) tools were used in infancy and at school age respectively. Ethical approved for the study was granted both in Zimbabwe and Norway. Results: Five hundred and ninety eight infants were assessed between the ages of 3 to 12 months, of whom 65 (11%) were HIV infected. The overall prevalence of high risk for Neurodevelopmental impairment (NDI) at any time was 9.4% (95% CI 7.1–11.1%) with any difference by gender. The high-risk category for NDI was double in infants infected with HIV compared to uninfected infants, ([OR] 2.1; 95% CI 1.0-4.3). After adjusting for other risk factors, small head circumference and lack of family income remained risk factors for NDI with an OR of 2.2 (1– 5) and 2.6 (1.0–6.4) respectively. To validate the MSCA, a 101 children aged 6-8 years old attending normal schools were assessed, of whom 60 were female. Sensitivity rates were low while the specificity rates were high. The number of children identified with cognitive impairment was 3%. The rural children scored significantly lower compared to their peers from urban areas. At school age, a total of 306 children from the primary cohort were available and agreed to participate. Of these, 32 were HIV infected, 121 HIV exposed uninfected and 153 HIV unexposed uninfected. Overall, 49 children (16%) (95% CI 12-20 %) had cognitive impairment. Children with HIV infection scored significantly lower than the HIV unexposed uninfected children in the perceptual performance domain, p value = 0.028.There was no difference in the prevalence of cognitive impairment by child HIV status. Cognitive impairment was significantly associated with parental loss, caregiver unemployment status and presence of under nutrition in univariate analysis. In the multivariate logistic regression model, caregiver unemployment status remained a risk factor for cognitive impairment after adjusting for other factors, with an odd ratio of 2.1 (95% CI 1.03-3.36) for all children. Conclusion: Lower socioeconomic status was associated with a high risk for NDI in infancy and cognitive impairment at 6 to 8 year of age in this study population. Children with HIV infection showed greater risk for NDI by age 3 months. In resource limited settings, strategies aimed at poverty alleviation and prevention of malnutrition should complement early HIV infant diagnosis and treatment of all children in order to optimize neurocognitive potential.