759 The association between maternal-child interaction, maternal depression and early child development: an observational sub-study in rural Zimbabwe

Abstract

<h3>Aims</h3> Maternal responsiveness, defined as the mother’s ability to perceive their infant’s behaviour and respond appropriately, may impact early child development (ECD). The quality of maternal responsiveness may be reduced by maternal depression. This sub-study of the Sanitation Hygiene Infant Nutrition Efficacy (SHINE) trial in rural Zimbabwe aims to evaluate the relationship between maternal-child interaction, maternal depression and ECD using validated tools in the low-middle income setting. <h3>Methods</h3> Pregnant women from two rural districts of Zimbabwe were enrolled in a cluster randomised trial between 2012 and 2015 and followed until children were 18 months old; a subgroup had ECD assessments at 24 months. Maternal depression was measured using the Edinburgh Postnatal Depression Scale (EPDS) at baseline and 24 months postpartum. A cut-off of EPDS ≥12 at either time point was used to indicate depression. Maternal-child interaction was measured using the Observed Mother-Child Interaction (OMCI) tool during a 5-minute observed mother and child interaction at 24 months of age. ECD was assessed at 24 months of age, using the Malawi Developmental Assessment Tool (MDAT; assessing motor, cognitive, language and social development) and the MacArthur-Bates Communicative Development Inventory (CDI) (assessing vocabulary and grammar). Generalised estimating equations were used to estimate effect size, while accounting for within-cluster correlation. Possible confounding factors were assessed first in bivariate analyses and included in final models if p&lt;0.2 or difference &gt;0.25 SD. Multinomial and ordinal regression models with robust variance estimation and Somers’ D for medians were used to compare baseline factors between groups depending on the distribution. Primary results were adjusted for study nurse, age of child, calendar month of birth, sex, randomised arm, home environment factors, highest maternal education, maternal age, wealth and birth weight. <h3>Results</h3> A total of 1996 children were recruited to the sub-study and assessed for ECD outcomes at 24 months of age; 540 completed full OMCI assessment and were included in analysis. Baseline characteristics between mothers and infants with and without full OMCI assessments were similar in terms of household characteristics, sanitation, water and hygiene. 26 of 539 mothers (4.8%) met the depression criteria. A significant positive association between OMCI score and ECD score was observed for MDAT total, MDAT fine motor, MDAT gross motor, MDAT language and Mac-Arthur-Bates CDI domains. There was greatest evidence for effect in the MDAT language score, where for 10 units change in total OMCI score, there were 2.3 units (95% CI 1.7 to 3.0) increase in MDAT language score (out of 36 points), equivalent to 0.53 SD. In the MacArthur-Bates CDI score, for 10 units change in total OMCI score there was 8.7 units (95% CI 5.8 to 11.6) or 0.44 SD increase in MacArthur-Bates CDI score (out of 99 points). There was no statistically significant association found between maternal depression and maternal-child interaction or early child development. <h3>Conclusion</h3> Within a sample of rural Zimbabwean infants and mothers undergoing live assessment using the OMCI tool, a significant association was found between quality of maternal-child interaction and ECD outcomes at 24 months of age.