Association of Father Involvement and Neighborhood Quality with Kindergartners' Physical Activity: A Multilevel Structural Equation Model

Abstract

Purpose. Examine the effects of father-child involvement and neighborhood characteristics with young children's physical activity (PA) within a multilevel framework. Design. Cross-sectional analysis of the Early Childhood Longitudinal Study—Kindergarten Cohort 1998. Setting. Nationally representative sample. Subjects. Data were available for 10,694 kindergartners (5–6 years; 5240 girls) living in 1053 neighborhoods. Measures. Parental report of child's PA level, father characteristics (e.g., time spent with child, age, education, socioeconomic status, hours worked), family time spent doing sports/activities together, and neighborhood quality (e.g., safety, presence of crime violence, garbage). Child weight status, motor skills, ethnicity, and television viewing were used as covariates. Analysis. Multilevel structural equation modeling with children nested within neighborhoods. Results. At the child level father-child time and family time doing sports together were positively associated with children's PA. At the neighborhood level parental perception of a neighborhood's safety for children to play outside fully mediated the effect of neighborhood quality on children's PA. Overall 19.1% and 7.6% of the variance in PA was explained at the child and neighborhood levels, respectively. Conclusions. Family-based interventions for PA should consider father-child time, with this contributing to a child's overall PA level. Further, neighborhood quality is an important predictor of PA only to the extent by which parents perceive it to be unsafe for their child to play outdoors.