Clustering of Obesity-Related Risk Behaviors in Children and Their Mothers

Purpose

To examine the clustering and patterns of obesity-related behaviors in children and their mothers and the concordance between mother and child pairs.

Methods

Primary school-aged children and their mothers in Victoria, Australia, participated (data from 549 mothers, 352 children, and 304 mother/child pairs). Examination of behavior patterns included 1) assessment of the overlap in national physical activity, screen-time, and fruit and vegetable consumption guidelines being met; and 2) cluster analysis of positive (consumption of fruits and vegetables) and negative (consumption of energy dense food/drink) dietary behaviors, sedentary behavior/screen-time, and physical activity.

Results

Only partial overlap was observed between groups meeting national recommendations for sedentary behavior and consumption of fruit and vegetables and energy-dense food. Less than 40% of mothers and children were meeting sedentary behavior guidelines. In both mothers and children five clusters were identified. With the exception of a single cluster in children with high levels of physical activity, clusters of healthy and unhealthy behavior were concordant in mothers and their children (p < .0001), particularly those defined by sedentary behaviors and consumption of energy-dense food/drink.

Conclusions

Complex patterns of obesity-related behaviors exist in children and their mothers. The concordance of clusters between children and their mothers suggests that modeling of sedentary behavior and creation of a child’s eating environment by parents may be particularly important influences on children’s behavior.

Introduction

Sedentary behavior (e.g., television viewing, time spent sitting), poor diet, and a lack of physical activity have been shown to be independently associated with increased obesity in children, adolescents, and adults 1, 2, 3, 4, 5, 6, 7. Despite the fact that these behaviors operate through different mechanisms and have different determinants, their distributions are not random throughout the population, with clustering of health behaviors within individuals frequently observed 8, 9, 10, 11.

Simple correlations between health-related behaviors often are weak, as demonstrated (for example) by a Dutch study of almost 3000 adults in which correlation coefficients between fruit consumption and physical activity and smoking were 0.07 and 0.08, respectively (12). The low correlations between health behaviors can be partly explained by complex clustering patterns, which mean that behaviors are only correlated among certain subgroups of the population (12). Treating those exhibiting an individual health behavior as a homogenous group can obscure the true relationships between health behaviors and reduces the potential for a deeper understanding of behaviors and their outcomes.

Numerous authors 8, 9, 10, 11, 12, 13, 14, 15 have examined the patterns of health behaviors within populations of adults and adolescents and have included those behaviors relevant to obesity. Few, however, have focused solely on obesity-related behaviors, and fewer still have included the comprehensive range of obesity-related behaviors included in this study (16). We are not aware of any evidence regarding the clustering of obesity-related behaviors among children.

The purpose of this study was to assess, with the inclusion of behaviors related to both diet (fruit/vegetable consumption and energy dense food/drink consumption) and activity (physical activity and sedentary behavior), the clustering of obesity-related behaviors in children and their mothers. It might be expected that clustering patterns would be similar in children and their parents given the strong parental influence on childhood environments, and the importance of parental modeling on the obesity-related behaviors of children 13, 17, 18. Therefore an additional aim was to test the concordance of clustering patterns between mother-child pairs.