Role of intervention trials in research on nutrition and cancer.

Randomized intervention trials of manipulated dietary differences are needed to identify not only specific causal associations between diet and disease but also the safety of the tested dietary modifications and whether they can lower community-wide rates of disease. Randomized trials of manipulated dietary changes are also used to identify biological markers of exposure and indicators of possible mechanisms of disease. The latter include measurements of changes in levels of metabolites and changes in the frequency of occurrence of either predisposing or predisease conditions. This group of studies is stimulated to a great extent by a demand for short-term surrogate endpoint measurements and objective measures of compliance that will lower the cost of studies and improve their credibility. Another series of dietary trials focuses on interventions to bring about long-term changes in dietary behavior at both the individual and community levels. All trials are designed to be cost efficient and to limit the exposure of average and low-risk individuals to interventions that may have long-term adverse effects. This conservative approach leads to some confusion about which interventions are being tested for use in clinical settings to lower risk in high-risk individuals and which interventions are being tested because they have the potential to lower the overall rates of disease in communities.