A Machine Learning Approach for Predicting Early Childhood Obesity Among WIC Participating Families

Project: Research project

Project Details

Description

PROJECT SUMMARY Early childhood obesity continues to be a clinical and public health challenge in the United States; almost a third of children ages 2-to-5 years old are either overweight or obese. Children who are overweight during early childhood are at least 5 times more likely than normal weight children to be overweight/obese as adults. Non- Hispanic blacks (NHB) have double and Hispanics have triple the rates of obesity compared to non-Hispanic whites (NHW) and Asians (10.4%,15.6%, 5.2%, and 5.0%, respectively). Once established, obesity and its comorbid medical conditions are difficult to treat and usually persist throughout adulthood. Thus, preventing its development as early as possible in life is critical to reduce the health care costs associated with obesity- related chronic diseases such as type 2 diabetes and cardiovascular disease. Current evidence suggests that pregnancy, infancy and early childhood may be critical time periods in the development of overweight/obesity throughout the lifespan but many knowledge gaps remain. These gaps include understanding factors that contribute to high risk for excessive weight gain during gestation, infancy and early childhood which may also partially explain differential response to intervention strategies. Moreover, risk for the development of overweight and obesity during these critical developmental periods appear to be increased in children from racial and ethnic minority and low socioeconomic status backgrounds, but there is a paucity of studies in the literature that include these groups. To better understand how early life factors impact the onset of childhood obesity by age 5, this study will utilize data from the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) program in Miami Dade County (MDC), Florida. Specifically, we will examine the relationship between pregnancy (maternal weight, gestational diabetes, smoking, alcohol/substance use) and early life (initiation, exclusive and duration of breastfeeding, time of introduction to solid food) risk/ protective factors and child body mass index percentile and z score at age 5. Covariates include mother's race/ethnicity, education level and household income. Statistical models will be evaluated using machine learning-cross validation techniques. This approach has not been previously applied to the obesity science to examine pregnancy and early life obesity risk/protective factors. The analyses take advantage of a robust secondary data set that is arguably one of the most diverse in the country; the WIC program currently services almost half of the nation's mothers babies and young children (as does the local program for MDC) who are from predominantly low income and ethnically diverse backgrounds. The findings from this study will inform the development of obesity prevention interventions targeting pregnancy, prenatal and early childhood developmental periods.
StatusFinished
Effective start/end date8/24/188/23/20

Funding

  • National Institute of Diabetes and Digestive and Kidney Diseases: $38,124.00
  • National Institute of Diabetes and Digestive and Kidney Diseases: $38,616.00

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.