Munch on fruits and vegetables, health officials implore--they will reduce the risk of chronic ills such as heart disease, high blood pressure and cancer. But large, long-term studies have mostly failed to corroborate initial signs of the cancer-fighting powers of produce. Recent studies suggest that any protective effect, especially for breast cancer, is weak or confined to subgroups of people.

The 1970s and 1980s saw hundreds of reports documenting dips in cancer rates of relatively voracious veggie eaters. In most of these studies, researchers asked people newly diagnosed with cancer to reconstruct their recent diets and compared the results with those of healthy volunteers. Such "case control" studies can be misleading, however. People in the cancer group may overestimate their consumption of foods they perceive as harmful, and volunteers tend to be more health-conscious than the general population. Nevertheless, the findings persuaded the National Cancer Institute to launch the national 5-a-Day Program in 1991 to take advantage of the apparent effect, encouraging people to eat five to nine servings of fruits and vegetables daily.

Stronger evidence can emerge from clinical trials and prospective, or cohort, studies, in which researchers interview large groups of healthy people about their diet and follow them over time. Results have begun coming in during the past five years. In a surprise, the findings failed to turn up protective effects against colon, lung and breast cancer. But negative results may simply mean the investigators were not looking in the right places. After all, diet is notoriously variable and difficult to quantify. (Try to catalogue what you ate in the past month.) And at least in the case of breast cancer, researchers might have misclassified fruit and vegetable consumption by using different questionnaires for different groups.

Work published last November seems to strengthen the case against a general protective effect. A questionnaire-based study that combined cohorts of American doctors and nurses found no change in cancer incidence based on fruit and vegetable consumption, although it did turn up a 25 percent decrease in cardiovascular disease among the heaviest consumers. The outcomes, reported by Harvard School of Public Health scientists in the Journal of the National Cancer Institute, suggest that any protective effect for cancer must be small, says Lawrence H. Kushi, associate director for etiology and prevention research at Kaiser Permanente in Oakland, Calif.

That conclusion gained further support in relation to breast cancer in January, when researchers published results from the European Prospective Investigation into Cancer and Nutrition (EPIC) study, the biggest of its kind yet. A team from the University Medical Center of Utrecht in the Netherlands examined five years' worth of data from 286,000 women aged 25 to 70. Participants spanned eight countries from northern to southern Europe, with foods ranging from sauerkraut to eggplant parmesan. The researchers, publishing in the Journal of the American Medical Association, found no difference in breast cancer rates based on total consumption of fruits and vegetables or on intake of specific vegetable groups, such as leafy greens or roots. A protective effect might still lurk in the data, however. For one thing, none of the studies has probed childhood diet. In the case of breast cancer, young girls may pass through a critical period during which fruits and vegetables could block the disease's early stages. Another factor could be genetic differences in anticancer enzymes, such as those that scavenge harmful free radicals, highly reactive compounds that damage DNA. Some studies have found differences in cancer rates in people who have specific gene variants and consume particular vegetables, notes Christine B. Ambrosone, an epidemiologist at Roswell Park Cancer Institute in Buffalo, N.Y. I have a little statue of broccoli on my desk, she says. I don't think these cohort findings are the final word.

Despite the uncertainty, experts agree the five-a-day recommendation still holds as part of overall healthy eating habits. So don't ditch the fruit salad and crudités just yet.

JR Minkel is a frequent contributor.