Profiles of Family-Based Risk Experiences Predict Early Developmental Outcomes for Children in Low-Income, Rural Communities

• In Event: 3-011 - Multi-Faceted Risk Exposure and Outcomes Through Early Adolescence: Transdisciplinary Findings from the Family Life Project

Integrative Statement

Family socio-economic status likely influences early development via child experiences within the family; however, the multitude of potential pathways complicates the study of this process. Cumulative risk approaches additively composite the presence or absence of different types of exposures, but this treats all experiences as equivalent and does not allow for interactions between experiences. Latent profile analyses (LPA) apply a person-oriented approach to the simultaneous study of multiple exposures without the constraints of cumulative risk models. The current paper adopts an LPA approach to examine unique combinations of seven distinct family-based experiences during the first three years of life as predictors of early developmental outcomes.
Interviews, questionnaires, observations, and assessments of parents and children in the Family Life Project - one of the largest (n=1,292) and most comprehensive longitudinal studies of children from low-income, rural American communities - were collected at 2, 6, 15, 24, and 36 months of age. Seven domains of early experience were examined: (a) Sensitive Caregiving, (b) Negative and Controlling Parenting, (c) Corporal Punishment, (d) Aggressive Adult Conflict, (e) Learning and Cognitive Stimulation, (f) Maternal Internalizing Symptoms, and (g) Household Disorganization. Analyses were weighted to accommodate the complex sampling design, and full information maximum was used to account for missing data.
A four-class model was selected based on model fit, parsimony, and interpretability (see Figure 1). Profile 1 – labeled “Average Exposure” (35% of the sample) - exhibited largely average/moderate levels of exposures across domains. Profile 2 - labeled “Positive Exposure” (38%) - exhibited more optimal levels of exposures across domains. Profile 3 - labeled “Problematic Adult Exposure” (9%) - was uniquely characterized by very high maternal internalizing and aggressive adult conflict. Profile 4 - labeled “Problematic Parenting Exposure” (19%) – was uniquely characterized by very low maternal sensitivity and stimulation and very high maternal negativity and household disorganization.
Income-to-needs ratios, maternal age, education, and ethnicity differentiated profiles and were included as controls in subsequent analyses of differences in children’s social, emotional, and cognitive outcomes at three years across the family profiles (also Table 1). First, children in the “Positive Exposure” profile had significantly better developmental outcomes across domains than children in all other profiles. Next, children in the “Problematic Parenting Exposure” profile had greater behavior problems, less prosocial behavior, fewer socio-emotional competencies, lower cognitive and executive functioning scores, and fewer language skills compared to children in the “Positive Exposure” and “Average Exposure” profiles. Additionally, children in the “Problematic Parenting Exposure” profile had lower prosocial behavior, emotional competencies, and cognitive and executive functioning scores compared to children in the “Problematic Adult Exposure” profile. Finally, children in the “Problematic Adult Exposure” profile had more problematic developmental outcomes than children with “Positive Exposure” and “Average Exposure”.
Within a population of families from low-income, rural American communities, the current analyses identifies homogeneous subsets of early family experience clusters that meaningfully predict developmental outcomes. The benefits and limitations of this approach, along with its relevance to the other papers in this symposium, will be discussed.

Authors

• W. Roger Mills-Koonce, University of North Carolina at Chapel Hill
  Presenting Author

• Nissa Towe-Goodman, University of North Carolina at Chapel Hill
  Non-Presenting Author

• Margaret Swingler, University of North Carolina at Chapel Hill
  Non-Presenting Author