Cancer Researchers to Harvard Nurses Study: Not So Fast!

American Institute for Cancer Research
Tuesday, 31 October 2000

Link between Fruit, Vegetable Consumption and Lowered Risk for Colon Cancer Remains Clear, Convincing

The American Institute for Cancer Research (AICR) today cautioned the public not to abandon diets that have been shown to protect against cancer merely because a recent study failed to find a protective effect. The Institute reminded the public of the "convincing and continually growing" scientific evidence that diets high in fruits and vegetables lower the risk for colon cancer.

AICR was responding to the latest findings of Harvard's Nurses Health Study and the Health Professionals Follow-up Study, published in the November 1 issue of the Journal of the National Cancer Institute. The surveys of 136,092 men and women reported no link between consumption of fruits and vegetables and colon cancer risk.

"The public needs to keep in mind that this finding contradicts the bulk of available evidence on the link between high consumption of fruits and vegetables and lower risks of cancer," said John Potter, M.D., Ph.D., head of the Cancer Prevention Research Program at Seattle's Fred Hutchinson Cancer Research Center. People may be misled by the tremendous amount of publicity this particular ongoing study often receives, he said. He stressed the need to keep these new results in a larger context.

"The data show that the subjects in these studies were consuming very few fruits and vegetables," said Ritva Butrum, Ph.D., Vice President for Research at AICR. "Five to nine servings a day are recommended for protection against cancer, but most of the respondees consumed far less. In fact, fewer than 2% of those in the Nurses' cohort reported eating more than four servings of vegetables a day, while only 3% of those in the Health Professionals' cohort said they consumed more than three servings of vegetables a day.

"This indicates that the studies' variability - in this case, the statistical range between those respondees who consume the most and the least amount of fruits and vegetables - is small. Statistically speaking, when variability is low, the association with high or low consumption gets harder to see," Butrum said.

Dr. Potter noted an additional consideration. "A greater issue is whether the analysis should actually control for total energy intake," he said. "Table 1 of the report makes it clear that those who consumed more fruits and vegetables consumed more food in general. A greater calorie consumption may have blunted the association one would expect to see with fruits and vegetables alone."

For these and other reasons related to their methodology, Dr. Potter said, the Nurses Study and Health Professionals Study do not and cannot make a definitive determination about the presence or absence of a protective effect.

"It would be a mistake to interpret these results as anything but what they are: a single set of findings on an important topic that's attracting more and more scientific attention each day."

Study's Limitations Noted

Because the Nurses and Health Professionals study is held to be the "gold standard" of nutrition research by some members of the media, the public is largely unaware of the important limitations in the study's design.

Cancer experts agree that no single study can prove or disprove any association. This is because each kind of study is specifically designed to examine very specific aspects of the diet-cancer connection. Dependable scientific consensus only emerges once different studies of different design performed by different investigators have been examined.

The Nurses and Health Professionals study, for example, is a cohort study. It tracks subjects by asking them to fill out a biennial questionnaire about their diet and lifestyle. No control group exists, and the limitations of self-reporting as a means of assessment is a hotly debated scientific topic.

"It is vital not to start over-interpreting recent studies merely because they are recent," said Dr. Potter. "We should wait and see how these data compare to the considerable body of research that is mounting across the globe," Dr. Potter said.

"The Nurses Study should contribute to ongoing scientific discourse," he added, "not replace it."

Dr. Potter is a Professor of Epidemiology at the University of Washington and recognized as a global leader in his field. He has published extensively on the causes of colorectal cancer, the role of plant foods in lowering cancer risk, and the interaction of environmental and genetic factors in determining cancer risk. He is Senior Editor for the scientific journal, Cancer, Epidemiology, Biomarkers and Prevention.

For more information, or to contact American Institute for Cancer Research, see their website at: www.aicr.org