The research, sponsored by the National Cancer Institute (NCI) in Bethesda, Md., lends additional weight to numerous earlier findings linking pancreatic cancer with cigarette smoking (smokers have a 70% increased risk) and a history of cancer in the immediate family (family members have a 30% increased risk).

But it also sheds some light on several less-understood areas including new findings involving obesity and caloric intake that study author Debra T. Silverman, MD, describes as "more speculative but very important, because they provide clues to be followed up in additional studies."

In this study, Silverman, an epidemiologist and senior investigator at the NCI, summarizes a series of five of her earlier papers. Those papers were published during the 1990s and were based on analysis of interviews with 481 pancreatic cancer patients and 2,099 people without cancer. Researchers at centers in Atlanta, Detroit, and New Jersey conducted the interviews between 1986 and 1989, mostly in the subjects? homes. The average time between diagnosis and interview was seven weeks.

Silverman believes the following findings offer "important leads" for further research:

• people who were obese and who also ate a higher-calorie diet than the median had a 70% or greater pancreatic cancer risk. (The type of calories did not matter.)
• those who were severely overweight or obese had 50% greater pancreatic cancer risk.
• people who ate cruciferous vegetables more than four times a week had a 50-60% decrease in risk compared to those who ate less than 1 ? servings per week. Cruciferous vegetables include broccoli, cauliflower, cabbage, brussels sprouts, kale, turnips, and mustard greens. They contain compounds thought to stimulate enzymes that detoxify carcinogens in the body.
• heavy drinkers (57 or more drinks per week), particularly African Americans, had greatly increased risk.
• people with diabetes that had been identified 10 or more years before the cancer diagnosis had a 50% increased pancreatic cancer risk.

Marji McCullough, ScD, senior epidemiologist and registered dietitian with the American Cancer Society, considers the interaction of higher body fat, high caloric intake, and higher risk of pancreatic cancer the paper''s most interesting finding. "The downside of the study is they didn?t measure physical activity," says McCullough. "Future studies should definitely measure physical activity."

"Absolutely," agrees Silverman. "Those are the two other components of energy balance ? activity level and metabolism ? and I think it?s very important to pursue that in future research."

Directly interviewing large numbers of patients was an important element of this study, says Silverman. Most research relies on interviews with next-of-kin, a method less reliable than first-person accounts.

But McCullough cautions that even with direct interviews, these studies are subject to biases that can muddy the findings. "For example, people with cancer may remember their behavior and diet differently from those without cancer," she says. "But these results support ACS recommendations to eat more vegetables, particularly cruciferous vegetables, to stay within your healthy weight range, and to stop smoking."

Early detection of pancreatic cancer is rare, and the median lifespan after diagnosis is six months. The ACS estimates 29,200 new cases of the disease will be diagnosed in the U.S. this year, and 28,900 people are expected to die of pancreatic cancer.