A cluster-analytical approach toward physical activity and other health related behaviors

Purpose: The purpose of our study was to investigate, on the basis of cluster analysis, whether healthy behavior clusters in a healthy or unhealthy way, whether physical activity can discriminate among those clusters, and whether the groups identified could be characterized by demographic variables. Methods: The relationships between physical activity and other health related behaviors were studied in three age groups, ages 16-25, 35-45, and 50-65. Almost 2400 subjects from a medium-sized Belgian town were interviewed at home to assess their level of physical activity in the past year and to obtain information on other healthy behaviors. Results: The two-cluster solution was found to be most reliable in all three age groups. Smoking discriminated most among the clusters. Cluster analysis divided each age group into a healthy and a more unhealthy subgroup, in which the healthy subgroup is larger than the unhealthy one. Sex differences were found in all three age groups, with more men in the unhealthy cluster. Socio-economic differences are especially strong in the two youngest samples. Conclusions: The hypothesis that leisure time physical activity had discriminating power between healthy and unhealthy subgroups was not confirmed. Physical activity seems to be a unique factor in contrast to other health related behaviors. In the future, specific intervention directed at the promotion of physical activity in the population is necessary.