CONCLUSION
This overview of observational epidemiologic studies has
shown that a high intake of fruit and vegetables rich in /3-car-
otene or high blood concentrations of /3-carotene are associated
with reduced risk of cancer at several common sites. The
association appears most consistent for lung and stomach can-
cer, whereas the association seems inconsistent for breast and prostate cancer. For colon cancer, an influence of /3-carotene, if
present, will be moderate. For other sites, eg, esophageal can- cer, data are promising but the number of studies and the power of these studies are limited. The associations may indeed be
ascribed to /3-carotene, because several plausible mechanisms
have been shown for cancer preventive properties of this com- pound. However, observational epidemiology cannot yet re-
solve whether other constituents from fruit and vegetables or
other factors may explain the findings from the case-control and cohort studies. The results of the ongoing intervention studies on cancer incidence will provide more answers, but these studies may need long follow-ups to be conclusive.