This has been exemplified by programs that reduce television viewing time and decrease BMI in children (Robinson, 1999).

  • Given the significant shortage at present of experimental evidence to guide programs and policies, and the fact that many societal variables of interest have not been well addressed in controlled experimental studies as moderating or mediating factors, obesity prevention will require an evidence-based public health approach that continues to draw on RCTs, quasi-experiments, and observational studies as important sources of information (Victora et al., 2004).

  • Given that obesity is a serious health risk, preventive actions should be taken even if there is as-yet-incomplete scientific evidence on the interventions to address specific causes and correlates of obesity. However, there is an obligation to accumulate appropriate evidence not only to justify a course of action but to assess whether it has made a difference.

  • Finally, for interventions that have minimal potential risk and require few resources, formative and process evaluations may be sufficient to provide a “preponderance” of evidence (Robinson et al., 1998). As described in Appendix C, the committee conducted a thorough bibliographic search of the relevant scientific databases and benefited from the expertise of academic, industry, government, and nonprofit sector experts during its deliberations. In examining the literature, the committee focused on studies that examined weight and body composition outcomes, but it also broadened its scope to include studies that looked at changes in physical activity (or sedentary behavior) levels and in dietary intake patterns.

In examining the evidence on obesity-related prevention interventions, the committee considered the methodologies used by individual studies. Evaluating such studies involves characterizing the appropriateness of their designs for measuring target outcomes (e.g., increasing physical activity) as well as assessing the quality and generalizability of the study execution. The committee also considered the strength of the overall body of available evidence. Other factors considered by the committee included the feasibility of implementing the recommended actions, the opportunities for making changes, and the past success of parallel public health and social change efforts. Where trends of social, dietary, and other factors and health outcomes ran in parallel, the committee believes these trends merit further study and concern while acknowledging the possible occurrence of confounding.

It is also important to note that the committee focused on areas for improvement rather than on specific products, mechanisms for distribution, or industries. For example, the report emphasizes the nutritional evaluation of the contents of vending machines in schools rather than the re-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement