A multivariate look at TV viewing and physical activity in high school students

Ode J, Bohnert K, Pivarnik J, Anthony J
Departments of Kinesiology and Epidemiology
Michigan State University

BACKGROUND AND AIM: A May 2004 report on health behaviors in high school students, based on the 2003 Youth Risk Behavior Surveillance System data, observed that 44% of high school students watch more than 2 hours of television per day. To extend this epidemiological research, which treated physical activity behaviors one-by-one, we conducted a latent class analysis (LCA) of five physical activity behavior measures. This data will offer a multivariate perspective on television viewing in context with other physical activity behaviors in high school students.

METHODS: Data are from public use files; 13,184 students in grades 9-12 were sampled and answered standardized questions on recent physical activity behaviors: vigorous physical activity (VPA), strength training (ST), attending a physical education class (PE), participating on a sports team (SPRT), and watching more than 2 hours of television (TV).

RESULTS: LCA analysis, via LatentGold software, disclosed a four-class solution of best fit. There was a highly active class (34% prevalence) in which the estimated prevalence of VPA was 95%, ST was 83%, SPRT was 78%, and PE was 98%. By comparison, there was a highly active class with no PE (20% prevalence) that had similar probabilities of VPA (95%), SPRT (67%), and ST (83%), but was lower in PE (6%). There was a moderately active class (28% prevalence) in which the prevalence of VPA was 33%, ST was 21%, SPRT was 40%, and PE was 58%. In addition, there was a minimally active class (18% prevalence) in which an estimated 4% prevalence of VPA, 8% prevalence of ST, 19% prevalence of SPRT, and 6% prevalence of PE. The estimated prevalence of TV was similar within the 3 classes (39-51%).

DISCUSSION: Physical activity participation is evenly distributed between highly active, moderately active, and minimally active classes within a high school population. However, television viewing is similar between these three classes. It also appears that attending physical education class has little impact on the highly active classes. In order to improve physical activity levels, it will be important to understand high school students’ physical activity behaviors in a multivariate context.