Abstract
The epidemiologic evidence that carotenoids are involved in cancer etiology is evaluated. Low intake of vegetables and fruits and carotenoids is consistently associated with an increased risk of lung cancer in both prospective and retrospective studies. In addition, low levels of serum or plasma β-carotene are consistently associated with the subsequent development of lung cancer. The simplest explanation is that β-carotene is indeed protective. Since retinol is not related in a similar manner to lung cancer risk, β-carotene seems to play a role that does not require its conversion into vitamin A. However, the importance of other carotenoids, other constituents of vegetables and fruits, and other nutrients whose levels in the blood are partially correlated with those of β-carotene has not been adequately explored. In addition, smoking, a powerful risk factor for lung cancer, is associated with reduced intake of carotenoids and lowered blood levels of β-carotene and has not always been adequately controlled in these analyses. Prospective and retrospective studies suggest that carotenoids may reduce the risk of certain other cancers; however, too few studies have looked at these sites to examine the consistency of the evidence. Although clinical trials of the efficacy of β-carotene in cancer prevention are underway, it is still necessary and prudent to continue well-designed prospective and retrospective studies of the carotenoid hypothesis.