ABSTRACT In 1981 it was hypothesized that a high dietary
intake of n-carotene might reduce human cancer rates. Since then,
several observational epidemiologic studies have addressed this
topic. The results of both case-control and cohort studies show a
remarkable consistency for the association of increased lung can-
cer risk with low amounts of dietary 3-carotene or low plasma /3-carotene concentrations. For stomach cancer, the evidence is
also consistent, although the number of studies is more modest. For breast and prostate cancer, the studies indicate no consistent
association of plasma or dietary n-carotene and reduced cancer risk. For colorectal cancer, the effect will be moderate, if existent.
For several other cancer sites, the numbers of cases in prospective
studies are often small, implying that only strong associations can
be detected. For some of these sites, results from retrospective studies are promising. The epidemiologic studies should be care-
fully interpreted because dietary habits may be misclassified and
smoking may reduce plasma a-carotene concentrations. Observa-
tional epidemiology cannot definitively resolve whether associa-
tions are indeed due to f3-carotene, or to other components of fruit
and vegetables that are rich in /3-carotene. However, overall results are promising and several plausible cancer preventive mechanisms have been reported for a-carotene. The ongoing human interven-
tion studies will provide more answers regarding cancer preven-
tion by 3-carotene but may need long follow-ups to be conclusive.