Diet and cancer prevention: the roles of observation and experimentation

Abstract
Chemoprevention is defined as the use of pharmacological agents (including nutrients) to impede, arrest or reverse carcinogenesis and has been used to test the efficacy of numerous nutrients and other dietary factors. Observational epidemiology and experimentation by randomized controlled trials (RCTs) have been used to evaluate dietary factors in cancer chemoprevention; however, consistency in findings has been elusive. The study of diet and cancer associations in humans poses a variety of challenges owing to the complexity in measuring dietary intake as well as the multifactorial nature of the cancer endpoints. In several circles, RCTs are viewed as being more credible than observational studies. When discrepant results between observational studies and RCTs are reported, careful consideration needs to be given to the details of each set of studies before accepting RCTs as valid and those of observational studies as biased. Considerations in the interpretation of results from RCTs include the background diet of the study population; the dose, duration and timing of the intervention; the precise form of the nutrient or dietary factor tested; and compliance with the intervention.