High fruit and vegetable intake linked to protection against pancreatic cancer

20/09/05 In one of the largest studies of its kind, UCSF researchers have found that eating lots of fruits and vegetables -- particularly vegetables -- is associated with about a 50 percent reduction in the risk of developing pancreatic cancer. Pancreatic cancer is difficult to diagnose and remains largely untreatable. It kills about 30,000 people in the U.S. each year and has a five-year survival under four percent.

The vegetables most strongly associated with increased protection were onions, garlic, beans, yellow vegetables (such as carrots, yams, sweet potatoes, corn and yellow squash), dark leafy vegetables and cruciferous vegetables. Light-green vegetables, tomatoes and tomato products showed weaker protective benefits.

Fruits were found to be protective but significantly less so than vegetables, with citrus fruits and citrus juices most protective.

The 50 percent reduced risk was associated with eating at last five servings per day of the protective vegetables or vegetables and fruit, compared to those who ate two servings a day or less. And eating nine servings per day of vegetables and fruit combined also was associated with about a 50 percent reduced pancreatic cancer risk compared with eating less than five servings per day. A serving is considered to be about a half cup of cooked vegetables, two cups of leafy salad or one medium-sized piece of fruit.

The study was based on in-person interviews of 2,233 San Francisco Bay Area residents: 532 pancreatic cancer patients and more than 1,700 randomly selected "controls." Control group participants did not have pancreatic cancer but were of a similar age distribution and similar male to female ratio as the pancreatic cancer patients. Participants were asked about their fruit and vegetable consumption for the one-year period prior to the interview, as well as other questions about diet, smoking, occupation and other factors.

The study findings regarding fruit and vegetable consumption are being published in the September issue of the journal Cancer, Epidemiology, Biomarkers and Prevention. "Pancreatic cancer is not nearly as common as breast or lung cancer, but its diagnosis and treatment are particularly difficult," said Elizabeth A. Holly, PhD, UCSF professor of epidemiology and biostatistics and senior author of the study. "Finding strong confirmation that simple life choices can provide significant protection from pancreatic cancer may be one of the most practical ways to reduce the incidence of this dreadful disease."

The results are considered particularly meaningful because of the study's size, the quantitative nature of the food questionnaire and the statistical significance of the findings, according to the study team. The likelihood that chance alone accounts for the discovered correlations between diet and cancer is less than one in a thousand for many vegetable categories, and well within the limits considered statistically significant for most of the vegetables studied, the scientists reported.

The research approach used by the UCSF team is called a "case-control" study because it involves a comparison of pancreatic cancer cases with a control group without pancreatic cancer. The main limitation of this approach is that many of the cancer patients identified at the outset die before they can be interviewed, as was the case with the UCSF study, the researchers reported.

Still, the alternative research approach -- a "prospective" study that inquires about diet and then follows people long enough to determine which of them contracts the disease -- is less feasible because it requires a very large sample size and a very long follow-up period, since pancreatic cancer is relatively rare, Holly said.

The authors note that a few prospective cohort studies have published results looking at fruit and vegetable consumption in relation to pancreatic cancer risk, and have observed no association or possibly a slight suggestion of benefit. This may be due, in part, to the reduced number of cancer cases available in such a study design, Holly suggested.

The researchers previously reported on other findings from this in-person interview-based study, namely: allergies, obesity, clinical symptoms of pancreatic cancer and environmental factors in relation to pancreatic cancer; also, genetic factors in relation to pancreatic cancer and cigarette use. Other factors still to be examined in this study are the relationship between pancreatic cancer an alcohol, detailed analyses on cigarette use, diet, other diseases and conditions.

The research was supported by the National Cancer Institute. This paper was partially supported by the Rombauer Pancreatic Research Fund.