Trajectories of Cognitive Recovery Between 2-5 Years In A South African Birth Cohort

The brain is especially susceptible to adverse exposures in early life and sociodemographic, physical, and psychosocial variables during gestation and the early years may have lasting impact on neurocognitive function. Little is known about the longitudinal trajectories of neurocognitive outcomes through the pre-school years in low and middle income country contexts.

Methods Latent class growth analysis (LCGA) was conducted to identify longitudinal cognitive trajectory classes for 728 children across 2, 3.5 and 5 years as part of the Drakenstein Child Health Study. Cognition was measured using Bayley Scales of Infant Development III at 2 years, and non-verbal index scores on the Kauffman Assessment-battery for Children II at 3.5 and 5 years; scores were standardized before running the LCGA.

Results Two classes were identified, the first demonstrated stable performance over time (N= 667, 91%). The second class initially demonstrated low cognitive performance at 2 years but recovered by 5 years (N = 61, 8.4%). Univariate penalized logistic regression models were conducted to test a series of potential protective determinates of cognitive recovery (i.e., latent class 2): breastfeeding, food diversity, parenting-styles, maternal education, employment, household income, birthweight and growth at 2-years. Greater birth weight was significantly predictive of membership to the cognitive recovery class (OR=1.34, CI=1.04-1.70, p=0.022).

Discussion This study highlights the importance of birthweight, known to be a sensitive outcome measure of health and risk exposures during gestation. Here we show that it is also a predictor for a resilient cognitive trajectory in the preschool years.