Examining the Relationship between Children’s Weight Status and the Different Patterns of Educational, Emotional, and Behavioral Problem Outcomes: A Latent Class Analysis

Background: Health and education are essential to general wellbeing. However, emotional, and behavioral problems or health related issues like obesity and overweightness among children might interfere in the attainment of quality education, and quality of life in general. As such, there has been an increasing interest in studying obesity among children and the interconnectedness in the relationships between educational, and behavioral and emotional problem outcomes. Emerging evidence suggests that childhood obesity might have consequential effects on both schooling and emotional or behavioral outcomes of children. The relationships between children’s weight status and educational outcomes, and emotional or behavioral problem outcomes have been studied in isolation from each other, employing mainly variable-oriented statistical approaches. These variable-oriented approaches focus solely on the aggregated relationship between variables, whilst ignoring the fact that there may be qualitatively significant differences among various groups of people. Purpose: This study seeks to employ a person-oriented approach, which focuses on sub-grouping of people in each population. A latent class analysis (LCA) was employed to identify groups of children, based on educational, emotional, and behavioral problem outcomes (Research question 1). Further, the study investigated the relationship between child’s weight-status and the identified groups, controlling for child and household-based factors (Research question 2). Methods: This cross-sectional study utilizes data from a 2018 survey conducted by the National Survey of Children’s Health (NSCH) with a sample size of n= 21,954. The majority of the children identified as white (69.0%) and male (52.3%). Five educational variables were used, school engagement (two items; children who care about doing well in school, and children who do all required homework), grade repetition, school absence, and school problems. Five emotional and behavioral problem variables included conduct problems, bullying (two items; children who are bullied, and those who bully others), depression, and anxiety. Results: Unconditional latent class models were estimated to address research question one. Five distinct groups were identified - thriving, educational and internalizing problems, multi-problems, low school engagement, conduct problems and bullying victimization- when using the full sample, and when separately examining the sample of children aged 10-17 years. Four similar groups except for educational and internalizing problems emerged for children aged 6-9 years. A multinomial logistic regression was performed to address research question two. Children with obesity compared to normal weight were more likely to be classified into the multi-problems, educational, and internalizing problems groups compared to the thriving group. Several child and household-based factors such as physical activity, ethnicity, and household income, to mention but a few predicted the different identified groups. This provided further an insight of the characteristics of the individuals assigned to each group. Conclusion: This study will contribute to extant literature on educational, emotional, and behavioral problem outcomes, and associated factors by focusing on person-oriented approaches, as an alternative view of studying data. Lastly, this research has the potential to provide theoretical knowledge and implications for policies related to children’s wellbeing.