Relationship between Child Opportunity Index and Body Mass Index Z-score: A mixedeffects analysis with data from a lifestyle intervention with Hispanic children

Christian E Vazquez^1*Christian Wood¹Swasati Handique¹Yuanyuan Liang²Zenong Yin³Deborah Parra-Medina⁴

• ¹University of Texas at Arlington, Arlington, United States
• ²School of Medicine, University of Maryland, Baltimore, Maryland, United States
• ³University of Texas at San Antonio, San Antonio, Texas, United States
• ⁴University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States

Background: The Childhood Opportunity Index (COI) is a relatively new measure for assessing opportunity across education, health and environment, and socio-economic context. Research indicates that higher COI is associated with lower obesity risk; however, existing research offers no evidence for differences, or lack thereof, across racial/ethnic groups. The larger body of research on the relationship between neighborhood environments and obesity risk among Hispanic children with low-income between 5 to 11-year-olds is limited. The study aims to further explore the relationship between neighborhood opportunities, measured by the COI, and children’s body mass index z-scores (BMIz), adjusted for age and sex.Materials and Methods: The data are from a sample of Hispanic child-parent dyads (n=253) who participated in a one-year family lifestyle intervention. A linear mixed-effects model was fitted, with BMIz as the dependent variable, COI categorized into four levels, time, parent BMI, family income, adult education, child age, child sex, calories, language spoken in the household, physical activity, treatment group, and a time*treatment group interaction. Predicted probabilities were also produced.Results: After adjusting for covariates, children in the second (β=-.15, 95% CI=-0.27, -0.03), third (β=-.19, 95% CI=-0.31, -0.06), and fourth (β=-.15, p = 95% CI=-0.28, -0.02) quartiles of the COI quartiles had significantly lower BMIz compared to those in the first (lowest) COI quartile. Predicted probabilities show the different predictive margins of BMIz at each time point for each quartile compared to the first quartile.Conclusions: All the higher COI levels were linked to healthier weight status compared to the lowest COI level, though the pattern was not linear for any of the observed associations. Further investigation into the impact of different COI levels may be warranted to assess each quartile’s impact against each other, which was outside the scope of the current study. Results also provide evidence for potentially strengthening intervention supports for those at the lowest COI level, respective to those from all other COI levels.