repeated, well-conducted observational studies can provide sufficient support for policy recommendations. We also were receptive to evidence that a policy would be likely to affect a determinant of childhood obesity even if not studied for its direct influence on obesity. That is, we recommend policy changes expected to increase physical activity or promote more healthy eating in children because such intermediate outcomes are themselves associated with prevention of childhood obesity.
Our recommendations are predicated on the belief, supported by evidence, that a change in the target of a policy will produce the desired change in obesity or its behavioral risk factors (e.g., a policy resulting in reduced consumption of soft drinks would contribute to obesity prevention). This belief reflects an informed judgment that considers the quality of the studies and the results they generate.
Another important consideration is the possibility that a policy intended to reduce childhood obesity could have unintended adverse consequences. For example, policies to promote physical activity among children must take into account the potential for increased risk of injury. Because many important developmental processes occur during gestation and early childhood, it is essential to be as certain as possible that none of the recommended policy changes could have irreversible adverse effects during these critical periods. Lacking direct evidence in most cases, we relied on more subjective consideration of the plausibility of such consequences. We also considered ways to mitigate potential adverse consequences, such as environmental design to reduce playground injuries among young children.
In the report we note the current state of the scientific evidence pertinent to the recommended policies, including both uncertainties and areas in which research is needed. Ultimately, our recommendations take full account of the scientific data but also reflect our collective best judgment. We have attempted to be consistent in our approach to combining scientific evidence and expert judgment, holding policy options to the same standards. The urgency of addressing obesity in young children requires taking action rather than waiting for more conclusive research—applying the “best available” rather than the “best possible” evidence. Research in this context includes studies of more effective means of implementing or enforcing policy changes to reduce childhood obesity. Some of our policy recommendations (or our unwillingness to make recommendations) fall on the cusp of the scientific evidence, barely or almost sufficient, and those are issues for which more research could well tip the balance in one direction or another.