Clustering of obesity-related risk behaviors among families with preschool children using a socioecological approach: Cross-sectional study

Abstract

Background: Limited attention has been given to assessing home environments of parents with preschool-aged children using a socioecological approach to better understand potential influencers of obesity risk. Objective: The purpose of this cross-sectional study was to examine the clustering of obesity-related risk behaviors among mothers with preschool children. Methods: Mothers with preschool-aged children (ages 2 to 5 years) who participated in the online Home Obesogenic Measure of Environments (HOMES) survey were examined in clustering of four healthy recommended behaviors (ie, mother's fruit and vegetable intake ≥5 per day, sedentary screen time <4 hours per day, sugar-sweetened beverage intake <1 time/day, and increased physical activity level). Frequencies and percents of the clustering variables were conducted along with Spearman rank order correlations to determine significant associations. Ward's method with squared Euclidean distances were performed for the cluster analysis using the four standardized continuous variables. Identification of total cluster number was determined by visually inspecting the dendogram. Sociodemographic, intrapersonal, social environment, and home physical environment characteristic differences between cluster groups were further examined by independent t tests and chi-square analysis to validate findings. Results: Of the 496 participants (72.6%, 360/496 white; age mean 32.36, SD 5.68 years), only a third (37.1%, 184/496) consumed five or more servings of fruits/vegetables daily, had low sedentary screen time of <4 hours/day, and reported moderate to high levels of physical activity (34.1%, 169/496). More than half (57.7%, 286/496) consumed <1 sugar-sweetened beverage serving daily. A positive correlation (r=.34, P <4 hours/day) and low sugar-sweetened beverage intake (<1 serving/day) were found. Ward's hierarchical analysis revealed a two-cluster solution: less healthy/inactive moms (n=280) and health conscious/active moms (n=216). Health conscious/active moms were significantly (P