Naturalistic Patterns of Sustained Attention Across Early Childhood: Developmental Profiles and Longitudinal Associations With Executive Functions

The capacity to sustain attention over time develops rapidly over early childhood and is associated with socioemotional and cognitive outcomes. However, sustained attention has largely been studied in either shared or independent contexts, leading to gaps in our understanding of how trade-offs between sustained attention to shared versus individual targets may predict later outcomes. We examined this question in a longitudinal sample of 1,290 children (49% female, 43% Black), living in predominately rural, low-income regions, using a naturalistic shared picture book reading task. Children's sustained attention to individual relative to shared targets during the book reading task was measured at 24 and 35 months. Using latent profile analysis, we identified four developmental profiles of children's attentional trajectories: three of the profiles differed in the extent that children' attention shifted toward more socially directed relative to individually directed attention; a fourth profile showed atypical decreases in both socially directed and individually directed attention across development. Importantly, heterogeneity in children's attentional profiles were associated with differences in executive functions at 48 months of age. Specifically, children who showed greater relative increases in socially directed attention had higher executive functions performance, whereas children with atypical decreases in attention showed substantial deficits in this domain. These findings reveal distinct longitudinal patterns of sustained attention in naturalistic contexts and show that heterogeneity in these patterns are robust predictors of subsequent executive functions. This person-centered approach provides novel insights into how quantitative and qualitative changes in attention may impact executive functions development and may help identify children at risk for nonnormative trajectories.