Individual Submission Summary
Share...

Direct link:

Trajectories of grandmother childcare across early life in rural Pakistan

Fri, March 24, 3:30 to 4:15pm, Salt Palace Convention Center, Floor: 2, Meeting Room 250 D

Abstract

Background: Achieving optimal child development in early life (0-8 years) is a global public health priority. The majority of child health interventions focus on the mother-child relationship; however, across many contexts, grandmothers are key decision-makers on issues related to pregnancy, newborn health, and childcare. Given that grandmother caregiving roles are dynamic and evolving, understanding involvement across a child’s early years is crucial. This study sought to characterize trajectories of grandmother childcare from infancy to 6 years and assess predictors of trajectory group membership.

Methods: We used data from 916 families participating in the Bachpan Cohort, a birth cohort in rural Pakistan with an embedded perinatal depression trial. Grandmother childcare was quantified using two methods of maternally-reported caregiver roles: (1) At 3 months, 1, and 2 years, data came from semi-structured "Day-in-the-Life" interviews that quantified childcare responsibilities and individual roles by asking mothers to describe caregiving from their child’s perspective; (2) At 6 years, data came from an adapted child development module of the UNICEF Multiple Indicator Cluster Survey. At each time point, we summed the number of times grandmothers were reported. Then, we identified groups of children who shared common grandmother childcare trajectories across the 4 time points using censored normal models. We first identified the number of trajectories using Bayesian Information Criteria (BIC) and then determined the functional form (constant, linear, quadratic) of each trajectory group using BIC. To understand predictors of group membership, we conducted multinomial logistic regression models with baseline demographic and psychosocial factors (the referent was the Low group). For all models, we used sampling weights to account for over-sampled depressed mothers.

Results: Roughly half of mothers had no or low education, almost 1/3 were in their first pregnancy, and 1/4 had clinical depression. Roughly 70% of grandmothers were living with the index child and 2/3 were in joint or extended households. We found three grandmother childcare trajectories between 3 months and 6 years of age: (1) Persistently Low (32.1%), (2) Medium-Decreasing (33.7%), (3) High-Decreasing (34.2%). Overall, grandmother childcare decreased over time. Across all groups, grandmothers engaged in multiple domains of childcare including instrumental care (e.g. feeding, bathing), socioemotional care (playing, soothing, teaching cultural values), and providing support to the mother when she was unavailable.
Due to missing data on predictors, 862 families were included in the prediction modeling. Compared to the Persistently Low group, first-time motherhood, maternal depression during pregnancy, and grandmother-co-residence predicted Medium-Decreasing group membership. Greater maternal education, first-time motherhood, grandmother co-residence, and higher household assets predicted High-Decreasing group membership.

Discussion and future directions: From 3 months to 6 years of child age, 3 distinct grandmother childcare trajectories emerged in our sample of rural Pakistani families. First-time motherhood, maternal depression, and indicators of greater socioeconomic status predicted higher trajectory group membership. This may help inform intervention targeting to enhance grandmother childcare for mothers and children and help fortify early child development strategies in such settings. Ongoing work will also explore to what extent grandmother childcare trajectories are associated with child cognitive and socioemotional development.

Authors