individuals are directly exposed on a day-to-day basis. A distinction also is often made between “upstream” and “downstream” approaches: the former address factors far removed from the individual’s control, whereas the latter reach individuals on a one-to-one basis. Upstream approaches are also more likely than downstream approaches to reduce disparities for socially and economically disadvantaged populations because they improve access to opportunities for healthful eating and physical activity (Kumanyika et al., 2008). Upstream approaches often target sectors outside of public health and health care, such as education, agriculture, transportation, urban planning, and industry, that influence eating and activity patterns.

To develop a framework for evaluating childhood obesity efforts, a prior IOM committee developed a series of logic models that are helpful in understanding pathways to obesity prevention at different levels and in different sectors. The overall model developed by that committee, which is relevant to obesity prevention for populations of all ages, is reproduced as Figure 2-2 to provide a conceptual basis for thinking about prevention strategies.

At the far right of this logic model, the ultimate outcomes related to obesity prevention (“Health Outcomes”) are shown to result from achievements related to four different types of intermediate outcomes—structural, institutional, systemic, and envi-

FIGURE 2-2 Evaluation framework for childhood obesity prevention policies and interventions.

SOURCE: IOM, 2007.