Diet and cancer: timing makes a difference.

Diet and Cancer: Timing Makes a Difference

Breast cancer, the second leading cause of female cancer deaths, strikes about one in 10 U.S. women at some time in their lives. Though animal studies have linked high-fat diets with development of this disease, human data have proved less than conclusive. New research suggests that while animal fats may indeed increase breast cancer risk, other foods may decrease it. The study's provocative results also suggest that a woman's breast-disease status when she eats a potentially cancer-fostering or cancer-protecting food has an influence on whether that food will affect her future risk.

Between 1983 and 1985, researchers with the Vancouver Center of the Canadian National Breast Screening Study administered mammograms or physical breast exams to more than 9,300 women aged 40 to 59. Of these, 75 percent volunteered to fill out a questionnaire on risk factors.

The new study focuses on some 450 risk-surveyed women whose first-visit exams revealed a "lump" and led to its total removal. After excluding women with either cancer or no brease disease, T. Gregory Hislop and his colleagues at the Cancer Control Agency of British Columbia had a study group of 403 women with some form of benign brease disease.

From tissue analyses, the researchers divided the women into three groups: those with proliferative brease disease, characterized by abnormal, rapidly dividing cells; those with "severe atypias," a subset of the proliferative group that appeared far more cancer-like; and those with nonproliferative breast disease. Of these benign diseases, only the two proliferative types are known to increase a woman's risk of breast cancer.

Hislop's team matched each of the 403 patients with one or more volunteers of the same age who showed no evidence of breast disease in the screening. They then compared the women's diets. In the February AMERICAN JOURNAL OF EPIDEMIOLOGY, the researchers report signs of an apparent dietary influence over just two types of breast disease -- the ones with established links to an increased risk of breast cancer.

Vitamin A supplements and frequent consumption of green vegetables--often rich in vitamin A and related anticancer carotenoids (SN: 11/4/89, p. 294) -- appeared to protect women against the more general proliferative breast disease, apparently lowering a woman's cancer risk by more than half. Even heavy consumption of fat did not appear to influence the development of this abnormality, though it did increase the risk of developing severe atypias--thereby tripling these patients' breast cancer risk. However, eating vitamin A or green vegetables did nothing to limit severe atypias and their associated cancer risk.

Hislop says the more general proliferative category of breast disease tends to develop into severe atypias before maturing into a cancer. He concludes, therefore, that "vitamin A may play a role in the earlier stages of [precancerous] disease, and dietary fat later on."

"The design of this study is nice," says Meir Stampfer at Brigham and Women's Hospital in Boston. He notes that administering a questionnaire prior to cancer development -- in this case during the benign phase--limits diet-recall errors, a common concern in similar studies. However, he adds, the new data do "little to support the fat hypothesis" because the dietary-fat links in women with severe atypias "are far from statistically significant."

Norman F. Boyd at the Ontario Cancer Institute's Princess Margaret Hospital in Toronto offers another view. He notes the small number of women with severe atypias (32) makes it nearly impossible to achieve statistical significance in analyzing the risks associated with each of the study's four levels of meat-fat consumption. Boyd points out, however, that a related analysis in which the team shows an increasing risk with increasing fat consumption is statistically significant -- and "is generally more sensitive than the test applied to individual [consumption] categories]." Still, he cautions, it's too early to draw sweeping conclusions about diet and this cancer.

Hislop concurs, saying, "We view this as a hypothesis-generating study, not a hypothesis-testing one."