Early Childhood Learning Activities Buffer Adverse Effects of HIV Exposure on Infant Cognitive Development: A Longitudinal Study

Abstract

Food insecurity (FI) and HIV exposure have been hypothesized to deleteriously affect infant cognitive development (ICD). Yet few studies have examined these relationships, and to our knowledge, none have done so longitudinally. Further, strategies to improve ICD in low‐resource settings are limited. Therefore, our objectives were to determine if (1) FI and HIV‐exposure had independent or joint effects on ICD at 3, 9, or 12 months postpartum; (2) assess longitudinally, whether FI and HIV exposure were predictive of ICD in the first year of life; (3) examine the extent to which early childhood learning activities (ECLA) may serve as a buffer against the deleterious effects of HIV exposure on ICD.Data for this study were drawn from infants born to mothers enrolled in a longitudinal pregnancy cohort study in western Kenya (n=224). At 3, 9, and 12‐month postpartum clinic visits, many covariates were assessed, including ICD, food insecurity (Individual Food Insecurity Access Scale), ECLA, maternal depression (Center for Epidemiological Studies‐Depression scale), perceived stress, social support, and socio‐demography. ICD was assessed using the Malawi Developmental Assessment Tool (MDAT). ECLA included parental reports of singing, counting, reading and telling stories to their infants.Following descriptive and bivariate analyses, we built multivariate models by including all covariates significant at p<0.2 in the bivariate model. We then examined whether effects of FI and ECLA varied by HIV status by including interaction terms between these factors. Generalized estimating equations were fit assessing associations over time, adjusting for covariates and including interaction terms between FI and ECLA with HIV.In bivariate models, maternal FI (β=−2.03; p≤0.05), infant HIV exposure (β=−1.43, p≤0.05), and ECLA (β=1.25; p≤0.001) at 9 months postpartum were associated with ICD at 12 months. In multivariable cross‐sectional analyses, FI was not significantly associated with ICD at any time point (Table 1). However, HIV exposure was negatively associated with ICD at 3 and 12 months postpartum (β=−1.17, 95% CI: −2.15, −0.17; β=−1.47, 95%CI: −2.79, −0.13). The interaction of FI and infant HIV exposure was not significantly associated with cognitive development scores at any time point. However, ECLA was positively associated with infant cognitive development at 12 months postpartum (β=0.93; 95% CI: 0.27, 1.59). In longitudinal GEE models, HIV exposure was associated with lower ICD scores (β=−2.78; 95% CI: −4.69, −0.86) while ECLA predicted higher ICD scores (β=0.90; 95% CI: 0.51, 1.29). The interaction term between HIV and ECLA (β=0.69; 95% CI: 0.15, 1.24) suggests that ECLA can buffer the negative effects of HIV exposure on infant cognitive developmentThese data suggest that ECLA can improve ICD and mitigate some of the adverse effects of HIV exposure. Future research should explore the pathways through which ECLA impacts ICD within the context of HIVSupport or Funding InformationSupported by NIH‐NIMH K01 MH098902 Multivariable Repeated Cross‐sectional and Longitudinal models showing the Relationships between Food Insecurity, HIV exposure and Infant cognitive development among 224 Kenyan infants ICD (3 months) ICD (9 months) ICD (12 months) ICD (Longitudinal GEE Models) Coef. (CI) Coef. (CI) Coef. (CI) Coef. (CI) Coef. (CI) Lagged FI (Ref.: LFU) Most Food Insecure −0.02 (−0.45, 0.49) −0.35 (−0.98, 0.28) −0.99 (−2.73, 0.74) −0.11 (−1.06, 0.85) −0.03 (−0.37, 0.31) HIV Exposed (Ref. HUU) −1.17 (−2.15, −1.19)* −0.11 (−0.45,0.23) −1.47(−2.79, −0.13)* −0.59 (−1.33, 0.14) −2.78 (−4.69, −0.86)** Lagged ECLA (0–7) 0.07 (−0.33, 0.46) 0.93 (0.27, 1.59)** 1.23 (0.90, 1.56)**** 0.90 (0.51, 1.29)*** HIV × FI −0.41 (−2.34, 1.52) 0.15 (−1.20, 1.51) 1.11 (−1.22, 3.44) 0.03 (−1.19, 1.26) HIV × ECLA XXX XXX 0.69 (0.15, 1.24)* Notes: ICD=Infant cognitive development; CI: confidence intervals; Coef. = Adjusted Coefficients; LFU= least food insecure; HUU: HIV Unexposed Uninfected; HIV × FI =interaction between HIV exposure and Food insecurity; HIV × ECLA=interaction between HIV exposure and early childhood learning activities; p≤0.05; p≤0.01; * p≤0.0001. Lagged FI = FI examined at a previous time point; Lagged ECLA=ECLA examined at a previous lime point; XXX=non‐significant interaction terms