Hey, Wait A Minute

Ass Backwards

First they told us to eat fiber. Now they’re telling us to forget it. What’s a colon to do? 

On April 20, the New England Journal of Medicine tied America’s bowels in a knot with two new studies declaring that high-fiber diets do nothing to prevent the development of colon cancer, the second leading cause of cancer deaths in this country. For 30 years the experts had been telling us that stuffing ourselves with cereal with the taste and texture of Kitty Litter was the key to keeping our colons from killing us. Now we were free to fling our bowls across the breakfast nook.

According to most press coverage, the two studies were both powerful enough to induce the vapors (New York Times: “Experts on colon cancer said the research … left them stunned”) and definitive (NBC: “The new studies are actually the first big ones to thoroughly test the fiber-colon cancer connection”). No less than Dr. Jerome Groopman of Harvard Medical School, wrote an op-ed in the New York Times about how he wished that a patient and friend who had died of colon cancer had lived to learn that his diet had had nothing to do with his disease. The friend, who was also a doctor, felt guilty for bringing on his condition by “eating salami and eggs” instead of “whole grains for breakfast.”

But don’t go on the Colon Smart All Salami diet just yet. It turns out the studies were limited and disputable. And they fail to make the case that there’s no connection between diet and colon cancer.

Both studies started with this thesis: Put people in their 60s who’ve already been diagnosed with colon polyps on high-fiber diets, and they won’t get colon cancer. But when the results came in, thesis had become anti-feces. In one study, led by Dr. David S. Alberts of the University of Arizona, 719 people in an intervention group were given a daily serving of high-fiber cereal, a sort of intestinal bran bomb. They had about the same number of polyps—benign colorectal growths that signal increased risk of colon cancer—at the end of three years as a control group who were given cereal with minimal fiber. In the other study, led by Dr. Arthur Schatzkin of the National Cancer Institute, 958 people in an intervention group extensively changed their diets, giving up chuck roasts for chickpeas and raising their fiber intake by 75 percent. At the end of four years their colons, too, were as efficient polyp factories as a control group who didn’t change their diet at all.

The idea that what goes in one end affects what happens at the other grew in part out of international studies of colorectal cancer rates. Incidence is high in prosperous, developed parts of the world where people eat a lot of meat, processed foods, and sugar—diets high in fat and low in fiber. If that sounds like dinner, then it’s probably no surprise that when it comes to cases of colorectal cancer, the United States leads the world. Rates are also high in Canada, Europe, and Australia. But incidence is low where diets are based on plant foods, from grains to roots to vegetables, such as in Africa, Asia, and parts of Latin America.

Underscoring the diet connection is what happens when people move from places with a low colon cancer incidence to places where it is high. Their colorectal cancer rates soar. Take a survey of Japanese-born women who moved to Hawaii. Within their lifetime their rates of colon cancer jumped almost fourfold and remained high in their children. Chinese men who moved from Shanghai to Los Angeles experienced a doubling of colon-cancer rates. (Yes, a lot of things change when people move to America, so there may be some other factor at work here. But the New England Journal of Medicine has yet to publish a study on the connection between colon cancer and the purchasing of minivans.)

There are also biological explanations for the belief that a healthy colon is a colon that moves with autobahn-like efficiency. Diets high in fiber are thought to eliminate carcinogens we eat, as well as binding with bile acids, diluting their possible damaging effects. The high-fiber eaters in the Alberts and Schatzkin studies each consumed about 30 grams of fiber a day, the federal government’s recommended level for adults. (For comparison purposes one cup of Raisin Bran contains 8 grams of fiber; an orange, 3.) Alas, the average American eats only about 13 grams a day. But in parts of the world with low incidence of colon cancer, people can consume 60 to 80 grams a day, says Dr. T. Colin Campbell of Cornell University. Pity the Ex-Lax marketing manager assigned to this territory.

Since America is No. 1 in colon cancer (there’s a catchy headline for USA Today!), let’s accept the idea that it’s connected to the American diet. But we still don’t know the exact mechanism. Is it the high consumption of meat and fat? The low consumption of vegetables and fruit? The high intake of sugar and refined carbohydrates? Or some combination of these? Since the Alberts study simply told its participants to keep eating American style but sprinkle some bran on top, it seems an unlikely way to prevent the disease.

That’s what Cornell’s Campbell believes. “The [Alberts] study should never have been done. Anyone could guess it wouldn’t work,” he says. “As soon as one begins to focus so narrowly on one nutrient or food, particularly in a population consuming a high-risk diet to begin with, one is setting themselves up for failure.”

Look at it this way: Colon cancer is not a deficiency disease like scurvy. The way to cure scurvy is simple—consume sufficient vitamin C. But the development of colon cancer is a complicated, possibly decadeslong series of cellular changes and not simply the result of a bran deficiency. A co-author of the Alberts study, Dr. Maria Elena Martinez, concedes as much, “Sometimes we’re guilty of looking for a magic bullet. Sometimes we find it, sometimes we don’t.”

The bran supplement was also the only fiber added to the participants’ diet. But dietary fiber is not just bran, it’s many different compounds, found in everything from grains to nuts to vegetables and fruits that have a variety of effects on the body. If foods high in fiber help prevent cancer, says Dr. Diane Birt of Iowa State University, “it may not be the fiber itself, but other constituents in high-fiber foods.”

This was the conclusion of a 1997 report by the World Cancer Research Fund, an international nonprofit group that focuses on the prevention of cancer through diet. It reviewed 4,500 studies on the connection between cancer and nutrition and found that cancer is “mostly a preventable disease,” with colon cancer being one of the most preventable. The famous studies by Dr. Denis Burkitt in the 1970s postulating that the low rate of colon cancer in Africa was due to a high-fiber diet led to the modern craze for high-fiber cereal. But the WCRF report found that while cereals may have some benefit, the key ingredient in reduction of colon cancer risk is vegetable consumption.

A less conceptual, but more practical problem with the Alberts study was that people found it to be, well, a pain in the ass. Even though the participants only added one serving of bran cereal to their normal diets, they found that eating the dose—13.5 grams—led to, in the authors’ clinical phrase, “gastrointestinal effects.” People who increase their fiber intake are usually told to do so gradually, to minimize the “effects.” It turned out that the people in the high-fiber half of the study were so uncompliant that the authors had to change the protocol to keep enough of them enrolled. As the researchers wrote, “Our experience underscores the difficulty of performing large-scale nutritional intervention trials.”

Oddly, the Schatzkin study in the same issue of the New England Journal of Medicine doesn’t indicate any difficulty in getting its participants, who were mostly in their 60s, to make much more dramatic changes in their diets. This study, at least, was not about adding a single component to an existing diet, but about changing overall eating habits. And according to their self-reported records, the intervention group cut fat consumption by a third, increased their daily servings of vegetables and fruits by about two-thirds, lowered red meat consumption 20 percent, and increased fiber by 75 percent.

Keeping track of the changes made in overall diet by hundreds of people over four years is daunting. Each year 10 percent of the participants received phone calls asking them what they had eaten the previous 24 hours, and once a year everyone filled out a questionnaire and a four-day food diary.

If a wholesale change in diet is the key to a beautiful colon, what happened to Schatkin’s participants? It turns out that the few laboratory measurements of compliance used by the authors indicate that as diarists, the study’s participants perhaps weren’t as accurate as Samuel Pepys. For example, they reported that they cut their caloric intake by 100 calories a day for four years. But they only lost, on average, 1.5 pounds, and the weight loss was greater in the first year of the trial than the fourth. Schatzkin explains that since the weight loss was less than would be expected, “it probably represents underreporting of intake.”

More telling was the level of the participants’ serum carotenoids, the measure of certain plant chemicals in the blood. There are many different carotenoids, the best-known being beta-carotene, the substance that makes carrots orange. Studies in recent years have found carotenoid levels to be an excellent marker of vegetable and fruit consumption. People who increase vegetable and fruit consumption show dramatic increases in carotenoid blood levels. A doubling of consumption can raise blood levels 20 percent to 30 percent. But the serum carotenoid levels of the intervention group in the Schatzkin study increased their level only about 5 percent, and again the amount of change was less the longer the study went on. “The fact that they didn’t change very much raises the question, Why not?” says Dr. Henry Thompson, of the AMC Cancer Research Center.

Schatzkin, chief of the NCI’s nutritional epidemiology branch, acknowledges that the 5 percent “was not a big change.” He says of the participants’ increased vegetable and fruit consumption that “[i]t is possible it was less than what people said.” Even accepting that the participants did change their diets as much as they say they did, the laboratory evidence shows that the changes were of such a small order as to make it unlikely they would end up with Third World colons.

Schatzkin himself says that his study does not sever the diet-and-colon-cancer connection, but it raises other questions. Maybe dietary changes need to take place earlier in life, before the development of the first polyp. Or maybe diet can stop the progression of polyp to cancer, but the dietary change has to be more radical. Or maybe the study was too short to show significant results. As Dr. John Potter, of the Fred Hutchinson Cancer Research Center and chairman of the WCRF study, says, “If you did a smoking cessation trial just for four years, even if it were very successful, you would not expect to see much reduction in the incidence of lung cancer in the intervention group versus the continuing smokers.”

In his Times op-ed, Jerome Groopman makes the excellent point that blaming people for their cancer is both counterproductive and cruel. But to use these two studies to tell people that diet has nothing to do with a disease like colon cancer misses an opportunity. Though people have been hectoring other people about what they should eat since someone ate the first cockroach, the study of the effects of diet on cancer is still a new field. AMC’s Thompson is doing research on people at elevated cancer risk, looking at whether a diet extremely high in vegetables and fruits reduces measurable cellular damage. So far, some short-term studies have been promising. “That’s why this is an exciting field,” he says. “People should know they have in their grasp the means to make changes that affect their health.”

Yes sir, that’s the good news. So, settle back into your nook, have a cup of coffee, and pass the cauliflower, please.