Skip to main content
ARS Home » Plains Area » Houston, Texas » Children's Nutrition Research Center » Research » Publications at this Location » Publication #405238
Research Project: Preventing the Development of Childhood Obesity

Location: Children's Nutrition Research Center

Title: Using machine learning to evaluate the impact of economic and family factors on obesity and physical activity outcomes in an online intervention among African American girls

Author
item MUSAAD, SALMA - Children'S Nutrition Research Center (CNRC)
item HAZAN, HANANEL - Tufts University
item CALLENDER, CHISHINGA - Children'S Nutrition Research Center (CNRC)
item PUYAU, MAURICE - Children'S Nutrition Research Center (CNRC)
item Thompson, Deborah - Debbe

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 6/15/2023
Publication Date: 8/1/2023
Citation: Musaad, S., Hazan, H., Callender, C., Puyau, M., Thompson, D.J. 2023. [abstract]. Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity (AIM-AHEAD) Annual Meeting. August 1-2, 2023; Bethesda, MD. Poster Presentation.

Interpretive Summary:

Technical Abstract: African American (AA) girls are at higher risk for obesity compared to non-Hispanic White girls. Behavioral factors are not well understood and partially determined by socioeconomic and environmental factors. Obesity-prevention interventions yield small or null effects and are inconsistent across population subgroups. There is an unmet need to identify the most effective components and strategies for different populations. In this study, we examined the socioeconomic factors that have the strongest relationship with child weight and physical outcomes among AA girls in an online obesity-prevention trial. Using Technology to Prevent Obesity among African American Girls (Butterfly Girls) was an 8-week online, 3-armed, obesity prevention trial promoting healthy diet and physical activity to 8–10-year-old AA girls. The primary outcome was excess weight gain. Secondary outcomes included child diet and physical activity (PA). Data was examined using decision tree methods. Severe obesity at baseline and change in steps/day were key predictors of whether girls will gain excess weight. Socioeconomic status was key predictor of change in steps/day at 6 months since baseline. Findings suggest that obesity prevention trials need to consider child weight status and socioeconomic status before randomizing, even when studying a single race/ethnicity or low-income population, with implications for study design.