The association of physical activity, sedentary behaviors, and body mass index classification in a cross-sectional analysis: are the effects homogenous?
1 Department of Health Policy & Management, School of Rural Public Health, Texas A&M Health Science Center, College Station, Texas, USA
2 Department of Health Promotion & Behavior, College of Public Health, The University of Georgia, Athens, Georgia, USA
3 Department of Environmental & Occupational Health, School of Rural Public Health, Texas A&M Health Science Center, College Station, Texas, USA
4 Department of Social & Behavioral Health, School of Rural Public Health, Texas A&M Health Science Center, College Station, Texas, USA
BMC Public Health 2011, 11:926 doi:10.1186/1471-2458-11-926Published: 14 December 2011
While much is known about the benefits of physical activity (PA) and the consequence of sedentary behaviors relative to body mass index (BMI), little is known about the homogeneity of these effects across individuals. The goal of this study was to determine if PA and sedentary behaviors have the same effect on individuals of all BMI classifications.
Data from a community health assessment were analyzed and a sample was selected to include respondents who self-reported a chronic disease associated with obesity (n = 2,840). Descriptive statistics were used to describe the association between selected independent variables and BMI. Simultaneous quantile regression was used to identify the degree of homogeneity in the effect of demographic independent variables, minutes per week of moderate PA, and hours per day spent watching television on BMI classification. In studies using simultaneous quantile regression, the word "effect" is used to describe association, not causation.
Minutes per week of moderate PA had a significant effect on lower BMI, but only when respondents were at least classified as obese-class I (β = -0.001, p = 0.006). The change in effect of moderate PA in lower BMI increased significantly when respondents were classified as obese-class II versus obese-class I (F = 4.54, p = 0.033). Hours per day spent watching television had a significant effect on higher BMI, but only when the respondent was at least classified as overweight (β = 0.87, p < 0.001). The change in effect of watching television on higher BMI increased significantly when respondents were classified as obese-class I versus overweight (F = 5.57, p = 0.018).
PA and watching television were more related to BMI for obese individuals than those who were just overweight. Customized interventions for specific BMI classifications should be developed to maximize public health benefits.