Abstract MP90: Identifying Risk Profiles for Sedentary Behavior in Youth Using Recursive Partitioning Based on Individual, Familial, and Neighborhood Environment Factors
Background: Being sedentary is an established risk factor for obesity and related cardiometabolic complications, independently of physical activity (PA) levels. We used recursive partitioning analysis (RPA) to identify unique combinations of individual, familial, and neighborhood factors that increased the likelihood of being very sedentary.
Methods: Baseline data were collected in 2005-2008 for 512 Quebec youth (aged 8-10 years) with a history of parental obesity (QUALITY study). Sedentary behavior and PA were assessed by accelerometry (Actigraph). Children with ≥10h of valid wear time on at least 4 days were retained for analysis. Children were categorized as being very sedentary if they accumulated at least 300 minutes/day of ≤100 counts/min on average. Fifteen variables were submitted to the recursive partitioning process in order to identify sub-groups by likelihood of being very sedentary. MLR was used to estimate likelihood of being very sedentary across subgroups, controlling for the child’s age, sex, household income, and PA level. Indicator variables were used, retaining the lowest risk group as the reference.
Results: Data were complete for 445/512 participants. Six variables were retained to construct the classification tree. A total of 7 subgroups were identified, with proportions very sedentary equal to 5%, 22%, 31%, 26%, 33%, 40%, and 77%, respectively. The 7 risk subgroups, in order of increasing likelihood of being very sedentary, comprised children who: (1) met MVPA guidelines (engaged in at least 60 min/day of MVPA), (2) did not meet MVPA guidelines, but resided in lower poverty areas, (3) did not meet MVPA guidelines, resided in higher poverty areas, but with a higher park area ratio; (4) did not meet MVPA guidelines, resided in higher poverty areas, had a lower park area ratio, but did not have an obese father; (5) did not meet MVPA guidelines, resided in higher poverty areas, had a lower park area ratio, had an obese father, but lived in more urban areas; (6) same as subgroup (5) but living in less urban areas and not exceeding 2 hours/day of screentime on weekends; and finally (7) same as subgroup (6) but exceeding 2 hours per day of screentime on weekends. In multivariable logistic regressions, compared to subgroup 1, groups 2 to 7 were significantly more likely to be very sedentary, after controlling for age, sex and household income. However, after further controlling for minutes of MVPA, only children in group 7 remained significantly more likely to be categorized as very sedentary (OR: 7.3, 95% CI: 1.3-45.1).
Conclusion: The relationship between physical activity and sedentary behaviour is complex.
Specific combinations of factors appear particularly conducive to engaging in excessive sedentary behaviour, with weekend screen time possibly being the most salient individual contributor.
Author Disclosures: T.A. Barnett: None. G. Contreras: None. A. Van Hulst: None. M. Mathieu: None. M. Henderson: None.
- © 2016 by American Heart Association, Inc.