Retinol, beta-carotene and cancer.

Evidence from epidemiological dietary studies does not demonstrate an association between retinol intake and the risk of cancer in developed countries. There is, however, an inverse association between beta-carotene intake and the risk of cancer. Evidence from prospective biochemical epidemiological studies demonstrates that low serum retinol is associated with a high risk of cancer but only within about the first three years, indicating that it is probably a metabolic consequence of cancer. A low serum beta-carotene level is also associated with a high risk of cancer, but here the association persists for many years before the diagnosis of cancer, indicating that it probably also precedes its development. The dietary and serum studies are therefore consistent in showing a long-term inverse association between beta-carotene and the risk of cancer. This may be due to beta-carotene affecting the risk directly, or may reflect an indirect association with some other component of vegetables or with a non-vegetable component of diet that is itself related indirectly to vegetable consumption. Whatever the explanation, the association represents an interesting epidemiological clue to the relationship between diet and cancer.