do is hint at associations to be investigated more rigorously. Inprospective cohort studies, large groups of people—thecohorts—are followed for years and interviewed regularly to see if patterns emerge among those who do and do not get cancer. Though these too sufferfrom biases, the evidence is considered stronger than for retrospective epidemiology.The largest prospective study on diet and health has found so far that eating fruits and vegetables has, at most,a very weak effect on cancer prevention. There are suggestions ofpossiblebenefits with a few cancers, but nothing that has lived up to the earlier hopes.
We were told to eat ourfiber, and when Nancy went shopping she would bring home breakfast cereals that tasted like pieces of the cardboard box. Intuitively it made sense. You could imagine all that fiber scouring clean your intestines on its way through the digestive tract. Fiber was alsosaid to nurture a mix ofbacteria that reduces the risk ofcolon cancer.The case for fiber may be a little stronger than for other foods, butthe evidence has been controversial. One big prospective study found an association while another did not.
This all might be less ambiguous if foods could be subjected to the same kind of rigorous trials used totest new drugs. A large group of people is randomly assigned to an experimental group, which receives the treatment, or a control group, which does not. In the end the results are compared. But these studies are rare in cancer nutrition research. It is hard enough to force people to arbitrarily eat or not eat a certain food. And to clinch the case, the enforcement would have to last for the decades it can take a cancer to develop. When a four-year controlled trial was carried out with a low-fat diet that was high in fiber as well as fruits and vegetables it foundno evidence of a reduction incolorectalpolyps, which are precursors to colon cancer. Another randomized trial of about equal duration found that a fibrous diet hadno effect on therecurrence of breast cancer.
Reading these less than ringing endorsements, I was reminded of the biochemistBruce Ames, who has reported thatbrussels sprouts, cabbage, broccoli, cauliflower, and other staples of the farmer’s market containnaturally occurring carcinogens, built-in pesticides like the ones that might have killed the poor
Edmontosaurus
. People apparently don’t ingest these foods in quantities that would cause apublic health problem—or maybe we have acquired a natural resistance. But how did the superstition arise that plants have the opposite effect, conferring us with the power to beat backcancer? Fruits and vegetables evolved to promote their own propagation. Then people started eating them.
There was nothing very rigid about Nancy’s dietary pursuits. We both loved steaks and hamburgers, but we tried to moderate ourconsumption. Here the science sounds a little more persuasive. If the epidemiology can be believed,eating a lot ofred meat every single day might have raised our chances of getting colorectal cancer during the next decade by as much as athird—from 1.28 percent to 1.71 percent. But given those odds, cooking a giant steak on the weekend seemed worth the trade-off. For penance we would have fish sometimes. Knowing that it is rich inomega-3 fatty acids may have made the salmon and halibut we grilled more satisfying. But any firm connection betweenfish, fish oils, and colon cancer prevention has remained elusive.
Fruits, vegetables, fibers, fish—if nothing else, loading up on these foods promised to reduce one’s intake ofmammalian fat. Yet even that has come under challenge as a serious cancerrisk, and it is possible thatsugar may pose a greater danger by increasing bloodinsulin levels and stimulating the growth of tumors. In the end, it probably doesn’t matter so much what you eat as how much.Obesity—like old age, sunlight,radioisotopes, and cigarettes—has joined the short list of unambiguous instigators of