Abstract MP29: Identifying Risk Profiles for Childhood Obesity Using Recursive Partitioning: Complex Associations with Individual, Familial, and Neighborhood Environment Factors
Objective: To identify unique combinations of individual, familial and neighborhood factors in relation to obesity in children using recursive partitioning, and to examine whether specific profiles of these factors predict obesity at follow-up.
Methods: Data include 512 participants from the first two waves of QUALITY, an ongoing study on the natural history of obesity in 630 Quebec youth aged 8-10 years at baseline with a parental history of obesity. Children were considered obese if their BMI was ≥95th CDC age- and sex-specific percentile. Residential neighborhoods were characterized using in person neighborhood audits conducted by trained observers and data from a Geographic Information System for 500 m network buffers around participants’ residential address. Eleven variables were submitted to the recursive partitioning process, based on evidence for associations with childhood obesity: 2 individual variables (sugar-sweetened beverage intake, meeting PA guidelines), 4 familial variables (number of BMI defined obese parents, number of parents with abdominal obesity, parental education, household income), and 5 neighborhood environment characteristics (disadvantage, prestige, and presence of ≥1 park, fast food restaurant, and convenience store). A classification tree was identified following a series of binary splits. Multivariable linear regression models were subsequently used to examine associations between the categorical variable that represents the recursive partitioning subgroups and BMI percentile while controlling for age, sex, puberty, and parental education. The lowest risk subgroup was the reference category; the remaining subgroups were identified using 6 indicator variables. In addition, associations between subgroup membership and BMI percentile at follow-up were examined while adjusting for BMI percentile at baseline.
Results: Recursive partitioning yielded 7 subgroups with prevalence of obesity equal to 8%, 14%, 26%, 28%, 41%, 61%, and 63%, respectively. The 2 highest risk subgroups comprised children not meeting PA guidelines, with ≥1 obese (BMI) parent, with 2 abdominally obese parents, living in a disadvantaged neighborhood with no parks and, among those with the same characteristics, with access to park(s) but also living in close proximity to at least one convenience store. After adjustment for baseline obesity, the likelihood of obesity was more than 12-fold for both subgroups compared to children with no obese parents. Although group membership was strongly associated with BMI at baseline, it did not systematically predict change in BMI.
Conclusion: Obesogenic environments are characterized by multiple individual, familial, and neighborhood factors that jointly relate to child obesity in complex ways. Alternate subgroup definitions may better predict change in obesity.
Author Disclosures: T.A. Barnett: None. A. Van Hulst: None. M. Roy-Gagnon: None. L. Gauvin: None. M. Henderson: None.
- © 2015 by American Heart Association, Inc.