Seeking a contemporary understanding of factors that influence physical activity

Abstract

The Mozart opera, Abduction from the Seraglio, went far beyond the usual limits of tradition with its long, elaborate arias. Asked about the opera, Emperor Joseph II responded with the terse and famous comment, “Too many notes.” This supplement to the American Journal of Preventive Medicine contains articles from the Cooper Clinic Conference Innovative Approaches to Understanding and Influencing Physical Activity. The articles paint a picture of “a very complex causal web” of factors affecting physical activity behavior. As one ponders the boxes and arrows in logic models of how determinants, correlates, mediators, confounders, and moderators operate at intrapersonal, interpersonal, physical environment, and social/cultural environmental levels, should our comment be the analogous “too many boxes and arrows?” Curiously, physical activity behavior is not inherently complex. Up until a few hundred years ago, the study of physical activity determinants was not as complicated, because the means to accomplish daily tasks in a sedentary manner had not been invented yet. We can now study the choice to be active or inactive because of technology that created many sedentary options for accomplishing daily tasks. But technology also opened the door to a large number of new types of activities that were either unusual or impossible prior to good equipment, including bicycling, skiing, basketball, and roller-skating. We have not met anyone proposing to promote physical activity by going backwards in time, for example, by returning to manual labor on farms. The tack we are on is to put back into our lives a healthful and enjoyable subset of physical activities. In pursuing this tack, we are undergoing a paradigm shift in our approaches to promoting physical activity. On the public health practice side of this shift, public health is rebuilding its emphasis on healthy environments by including a new focus on active community environments. Unheard of ten years ago, partnerships between public health and urban planners, transportation planners, and park planners are growing steadily. On the research side of the paradigm shift, researchers are shifting to transdisciplinary models and multilevel research designs. The conference essentially proposed and sought to encourage, as part of this shift, a more structured approach to research. The approach involves: (1) more standardization in terminology; (2) a shift away from disciplinary-based conceptual models focusing on a subset of variables, and a shift toward trans-disciplinary, multilevel models better suited to the breadth of variables affecting activity; and (3) more attention to the mechanisms and details of how interventions influence physical activity behavior. A more structured and coherent approach is important. To an ever greater extent, a research study influences practice because it is included in evidence syntheses. A case in point are the evidence syntheses of the Task Force on Community Preventive Services, that have recently identified six recommended or strongly recommended interventions to increase physical activity behaviors. Of 253 reports retained for full review, 159 were excluded from the syntheses, often because of limitations in execution or design. More structure also benefits an important group of “customers” of this research: practitioners who seek to do competent and useful evaluations of community physical activity initiatives. These practitioners need coherence to the logic models guiding the evaluation, e.g., which variables are the important mediators to track? They need measurement tools that correspond to boxes in the logic models. As research in physical activity evolves, it needs to pay increasing attention to such “customers.” In this way, the impact of research results is leveraged. As this supplement takes stock of where we are and where we need to go, several points are worth considering. First, the articles point us in the strategic direction of a comprehensive approach to promoting physical activity. Given the breadth of possible determinants, we do not sense that a few key determinants of physical activity behavior have emerged, and that just a few community interventions could produce large effects on activity. Rather, it appears that large effects on activity are From the Physical Activity and Health Branch, CDC/NCCDPHP (Buchner), Atlanta, Georgia; and the Department of Urban and Regional Planning and the Center for the Study of Population, Florida State University (Miles), Tallahassee, Florida Address correspondence and reprint requests to: David M. Buchner, MD, MPH, Chief, Physical Activity and Health Branch, CDC/ NCCDPHP, 4770 Buford Highway NE, MS K-46, Atlanta, GA 303413717. E-mail: DBuchner@cdc.gov OR zdg4@cdc.gov.