In Utero HIV Exposure and the Early Nutritional Environment Influence Infant Neurodevelopment: Findings from an Evidenced Review and Meta-Analysis

Abstract

The developing brain is especially vulnerable to infection and suboptimal nutrition during the pre- and early postnatal periods. Exposure to maternal human immunodeficiency virus (HIV) infection and antiretroviral therapies (ART) in utero and during breastfeeding can adversely influence infant (neuro) developmental trajectories. How early life nutrition may be optimised to improve neurodevelopmental outcomes for infants who are HIV-exposed has not been well characterised. We conducted an up-to-date evidence review and meta-analysis on the influence of HIV exposure in utero and during breastfeeding, and early life nutrition, on infant neurodevelopmental outcomes before age three. We report that exposure to maternal HIV infection may adversely influence expressive language development, in particular, and these effects may be detectable within the first three years of life. Further, while male infants may be especially vulnerable to HIV exposure, few studies overall reported sex-comparisons, and whether there are sex-dependent effects of HIV exposure on neurodevelopment remains a critical knowledge gap to fill. Lastly, early life nutrition interventions, including daily maternal multivitamin supplementation during the perinatal period, may improve neurodevelopmental outcomes for infants who are HIV-exposed. Our findings suggest that the early nutritional environment may be leveraged to improve early neurodevelopmental trajectories in infants who have been exposed to HIV in utero. A clear understanding of how this environment should be optimised is key for developing targeted nutrition interventions during critical developmental periods in order to mitigate adverse outcomes later in life and should be a priority of future research.

Keywords: HIV; breastfeeding; neurodevelopment; nutrition.

Figure 1

Adapted Graphical Overview for Evidence Reviews (GOfER) diagram [34] of studies reporting on neurodevelopmental outcomes in infants in relation to HIV-exposure status. A plus sign (+) beside the timeline bar indicates that assessments were performed for infants beyond 36 months of age. HUU = HIV-unexposed, uninfected; HEU = HIV-exposed, uninfected; HEI = HIV-exposed, infected; d = days; w = weeks; m = months; y = years. ^a Two studies report on data from the Mother and Infant Health Study (MIHS) cohort [48,52]. ^b Development was assessed for all children at enrolment (range 6–18 m) and after 3 months. A third assessment was done for 25 children after 6 months of enrolment. ^c Infants under one year of age were eligible for recruitment. Neurodevelopment was assessed at baseline (prior to initiating ART for infants who were HEI) and again three and six months later. Breakdown of infant ages at baseline, second, and third assessments was not provided. ^d A breakdown according to infant HIV status for the 311 infants who had neurodevelopmental assessments was not available. All were HIV-exposed. One hundred and thirty-nine HEI and 519 HEU assessments were used in analysis for both cognitive and motor outcomes (repeat measures for infants were included). Data from this infant cohort are reported in another study under review [50].

Figure 2

Adapted Graphical Overview for Evidence Reviews (GOfER) diagram [34] of studies reporting on early life nutritional factors and neurodevelopmental outcomes in infants affected by exposure to maternal HIV infection. Detailed descriptions of the study treatments or interventions are presented elsewhere (Supplementary Table S3). A plus sign (+) beside the timeline bar indicates that assessments were performed for infants beyond 36 months of age. HUU = HIV-unexposed, uninfected; HEU = HIV-exposed, uninfected; HEI = HIV-exposed, infected; SOC = standard-of-care; ICYF = infant and young child feeding; WASH = water, sanitation, and hygiene; Vit A = Vitamin A intervention; MVI = Multivitamin intervention; d = days; w = weeks; m = months. ^a There were 12 infants included in the analyses whose HIV status was unknown. ^b Reports on two infant cohorts who were enrolled in two separate micronutrient trials in Tanzania. One of these cohorts is reported on in another study under review [57]. ^c All infants were HIV-exposed. Authors did not report cohort breakdown according to infant HIV status but did report that infant HIV status did not modify relationships between multivitamin supplementation and neurodevelopmental outcomes.

Figure 3

Quality assessment of articles according to study design using the (A) Quality Appraisal Tool for Case Series (18-item checklist, case series [23], the (B) Newcastle–Ottawa Quality Assessment Scale (cohort studies [22]), and the (C) Cochrane Collaboration’s Tool for Assessing Risk of Bias (randomized controlled trials [24]). ^a This sub-study was a secondary endpoint of a larger randomised controlled trial. This sample includes children who attended only one of the three research clinics and 12% of the infants who attended follow up at 15 months overall (71.9% of the total number of infants who were randomized). Demographic characteristics across the placebo and multivitamin sub-study groups remained similar, so risk of selection bias (internal) was assessed as low. ^b While authors report treatment compliance for the randomized arms, it is not clear that authors considered treatment compliance for the population of mothers whose infants had at least one neurodevelopmental assessment and were included in this analysis (327 of 1078 assigned to treatment arms).

Figure 4

Random-effects meta-analysis for studies reporting on BSID-III sub-scales for infants who are HEU compared to HUU. Studies that reported means and standard deviations for scaled BSID-III scores were included. Where longitudinal data were available, analyses used the oldest data reported to capture any persistent impacts of HEU on neurodevelopment [49]. Data from da Silva et al. (2017) [41] were a cross-section of four separate groups of infants (aged 4, 8, 12, and 18 m). For one study, cognitive data analysed were composite scores, as scaled scores for this domain were not available [48]. Methodological quality assessments for these studies are reported in Figure 3. A summary effect estimate is only included for the cognitive sub-scale, as this is the only assessment that included more than five studies [30]. Data are presented in forest plots as Hedge’s g (95% CI) in ascending order according to infant age at assessment. HUU = HIV-unexposed, uninfected; HEU = HIV-exposed, uninfected; m = months; BSID-III = Bayley Scales of Infant Development 3rd edition [28]; CI = confidence interval.

Figure 5

Summary of the results from randomised controls trials investigating the influence of early-life nutrition related interventions on neurodevelopmental outcomes in the first 24 m of life in infants exposed to HIV. Full descriptions of the study treatments or interventions are presented elsewhere (Supplementary Table S3). Chanda et al. (2020) [52] report positive effects of a combined IYCF + WASH intervention on motor, language, and behavioural outcomes at 24 months in a group of 300 infants who are HEU. McGrath (2006) [51] reports positive effects of daily maternal multivitamin supplementation from 12–27 weeks’ gestation to 18 months postpartum on motor developmental outcomes in 327 infants exposed to maternal HIV infant (breakdown of infant HIV status not provided) at six months of age. Methodological quality assessments for these studies are reported in Figure 3. Data are presented in forest plots as Hedge’s g (95% CI) in ascending order according to infant age at assessment. SOC = standard-of-care; ICYF = Infant and young child feeding; WASH = water, sanitation, and hygiene; VHW = village health worker; MDAT = Malawi Developmental Assessment Tool; CDI = Communicative Development Inventories; d = days; w = weeks; m = months; PDI = psychomotor development index; MDI = mental development index; CI = confidence interval; HEU = HIV-exposed, uninfected.