Discrepancies in nutritional recommendations: the need for evidence based nutrition.

The widespread acceptance that 'evidence-based medicine' should determine all aspects of clinical practice leads to a consideration as to whether 'evidence-based nutrition' should be based on similar principles. Randomised controlled trials (RCT) are universally regarded as the gold standard by which to determine whether a drug is appropriate in a particular clinical situation. The evidence for some nutritional recommendations is indeed substantiated by RCT but in the case of some chronic diseases, notably cancers, where nutritional factors may operate as promoters or protectors many years before the onset of clinical disease, RCT may not be particularly appropriate. A range of experimental studies and descriptive epidemiological approaches may be regarded as sufficient to justify nutritional recommendations or dietary guidelines. Recommendations for the prevention and treatment of selected diseases will be considered in the context of their evidence-base.