Description
Obesity has reached severe levels both in the United States and abroad, and childhood obesity carries with it the added problem of an extensive latency period for problems to manifest. Physical activity interventions directed towards child obesity require objective data (usually by accelerometer) to ensure that the intervention had an impact on the target population. However, characteristics associated with compliance to accelerometers by children have not been well studied. This study's overall hypothesis is that parent's education and the child's body mass index (BMI) will be positively related to compliance, whereas the child's age and exercise per week will be negatively related to compliance. The study included 178 child participants and their primary caregivers from the MOVE/me Muevo study in the San Diego county region. A cross sectional study was conducted in 20XX with these participants to identify factors associated with child compliance with accelerometer use. Using binary logistic regression, the following variables were tested for their relation to compliance: child age, child BMI, parent education level, parent BMI, child gender, and child days of exercise per week. Main effect variables were chosen for inclusion into a final model by conducting a univariate analysis on variables and included those that were below the cutoff of p= 0.2. Those variables that were below the p-value set in the univariate model were included into the multivariate model. In the multivariate model, after adjusting for all other variables, child age, parent education, and parent BMI were significantly related to child compliance at an alpha level of 0.05. Increasing childhood age was associated with decreasing levels of child compliance (P-value= 0.029). Increasing levels of parent education was associated with increasing child compliance (P-value=0.0060). Increasing levels of parent BMI was associated with increased child compliance (P-value = 0.0248). These results, that child age, parent BMI, and parental education are related to compliance, suggest that there are factors not readily influenced by researchers that affect compliance. This could lead to changes in study design such as oversampling of groups with lower rates of compliance.
