Antioxidants and cancer.

"The Real Power of VITAMINS," trumpeted the April 6 edition of TIME magazine. "New Research Shows They May Help Fight CANCER, HEART DISEASE, and the Ravages of AGING."

Pretty exciting stuff.

Much of that excitement involves antioxidants like vitamins E and C and betacarotene. Yet last November, in a proposed regulation, the Food and Drug Administration said that there was not "significant scientific agreement" that antioxidants could prevent cancer.

So what's the scoop? What's the real evidence, pro and con?

Our conclusion: It may be sensible--even smart--to take antioxidants. But the evidence--while tantalizing--is not conclusive.

It's about time.

Platitudes like "if you eat a balanced diet, you don't need vitamins" are finally starting to go out of fashion. As TIME magazine put it, "attitudes [towards vitamins] have been shifting over the past two decades."

A growing number of scientists are no longer asking whether people get enough vitamins to prevent scurvy or rickets. They want to know if vitamins can help prevent cancer, heart disease, cataracts, and other chronic diseases.

And the evidence--especially on antioxidants--is exciting. All the more so because they appear to be safe.

But most press reports don't tell the whole story. They rarely report the holes in the evidence: which studies show little or no benefit and, more importantly, which studies have yet to be done.

That's partly because much of the research is so new. Few reporters know which researchers to call for criticism, and few researchers know the evidence well enough to critique it.

And there's another factor: The vitamin industry is working behind the scenes to get its message out.

FULL-COURT PRESS.

Last February, the New York Academy of Sciences held a conference entitled "Beyond Deficiency: New Views on the Function and Health Effects of Vitamins." The program included breakfast and luncheon briefings for the press and Congressional staff--rarely a part of scientific meetings. Afterwards, scientists and the media were blanketed with lavish press packets.

Press coverage was plentiful. But few articles noted that the conference was co-sponsored by Hoffmann-LaRoche, the largest vitamin manufacturer in the U.S. None of the Academy's materials mentioned the company.

In a way, you can't blame the industry. For years, it has labored to pry open the minds of scientists who refuse to consider advising people to take supplements. And health authorities like the Surgeon General don't exactly scramble to recognize new scientific evidence.

And it's no coincidence that the industry is on the offensive now. In 1990, Congress told the FDA to decide whether to allow supplement and food labels to make claims linking antioxidants and cancer. Last November, the FDA said that "significant scientific agreement" on a link doesn't exist. We agree.

That doesn't mean that claims should never appear, or that people shouldn't take antioxidants now. At such little cost and risk to health, it's not a bad bet. But you should know about the missing or disappointing research as well as the promising leads.

THE THEORY.

The theory that antioxidants can reduce the risk of cancer is a plausible one: damage caused by oxidation plays a role in causing cancer. Antioxidants--including beta-carotene, vitamin C, and vitamin E--can block oxidation. Therefore, antioxidants should help prevent cancer, right?

Not necessarily.

"It's possible that the damage caused by oxidation that we're exposed to is small compared to exposures to other carcinogenic hazards," says Regina Ziegler, an epidemiologist at the National Cancer Institute. "It's also possible that the body has protective systems in place to repair that damage."

That's why we need persuasive evidence from human studies that antioxidants are protective, says Ziegler. And so far, we don't have it.

THE OBSERVATIONS SO FAR.

Some of the human evidence linking antioxidants to cancer comes from what scientists call "observational" studies, which compare the diets or blood levels of people who have (or later get) cancer to those who don't.

(Observational studies are different--and less reliable--than "intervention" trials, in which researchers give people either a supplement or a placebo and wait to see who gets cancer.)

Observational studies show a strong link between some antioxidants and some cancers. For example, people with higher levels of beta-carotene in their diets or blood consistently have a lower risk of lung cancer. And people with more vitamin C in their diets have a lower risk of stomach cancer.

For other cancers, though, evidence of a link is less consistent.

Figure 1 (see next page) is a summary of six studies on beta-carotene compiled by George Comstock of Johns Hopkins University.

It shows only the results of studies in which blood levels of beta-carotene were collected before anyone got sick (although studies that asked people about their diets have yielded fairly similar results). One advantage of blood analyses is that they don't depend on people remembering what they ate years before they got cancer.

Comstock looked at whether the blood beta-carotene levels of people who later got cancer (cases) were at least 10 or 20 percent lower than those of people who didn't get cancer (controls). The strongest evidence is represented by circles to the left of the dashed purple line.

While the evidence looks promising for stomach cancer (which is no longer common in the U.S.), only for lung cancer are all the circles in the right place. In other words: people with lower beta-carotene levels consistently have a higher risk of lung cancer.

"It is becoming clearer with every report that beta-carotene is not universally protective against free-radical carcinogenesis as was originally hypothesized," says Comstock.

Figure 2 shows the results for vitamin E. There is no impressive evidence for any cancer. One study shows that Finnish men who had lower blood vitamin E levels had a higher risk of pancreatic cancer. But that wasn't true for Finnish women, and the results need confirmation.

Comstock has no graph for vitamin C because, so far, only one study was designed to analyze the vitamin C in the blood before it degraded. In that study, people who had lower levels of vitamin C had a higher risk of cancer of the stomach, but not the lung, colon, or rectum.(1)

BETA OR BROCCOLI?

If the observational studies are consistent for lung cancer, why not put that information on labels of foods and supplements that contain beta-carotene?

That would certainly help smokers, who have 20 times the usual risk of lung cancer (even though studies suggest that beta-carotene may only cut that risk in half).

Many scientists, however, aren't convinced that it's the beta-carotene that's responsible for the lower risk of lung cancer. The same goes for other antioxidants and other cancers.

"We have persuasive evidence that diets high in fruits and vegetables are protective against lung and probably other cancers," says the NCI's Regina Ziegler. "But whether it's because of antioxidants or other constituents, or the overall dietary patterns of people who eat fruits and vegetables, we can't say."

In other words: the protection might come from the extra fiber, the absence of fat, or something else in diets that are rich in fruits and vegetables. For example, broccoli is loaded with beta-carotene. But it also contains sulforaphane, which reportedly has potent anti-cancer properties (see "Quick Studies," June 1992).

It's also possible that people who eat few fruits and vegetables do other things that increase their cancer risk. For example, women who eat few fruits and vegetables often smoke or have other risk factors for cervical cancer, like human papilloma virus, low socioeconomic status, a large number of sexual partners, and a young age at first intercourse. Those factors--not antioxidant intake--may account for their increased risk.

There is only one way to distinguish between the diet, the food, and the nutrient: an intervention trial.

LOVE AT FIRST SITE.

An intervention trial is the gold standard of research. The idea is to divide your participants randomly into two or more groups. One gets the nutrient in question, the other gets a look-alike, but inactive, placebo. You wait several years. Then you count the cancer cases in each group. That's it.

You don't have to worry that people can't remember what they ate ten years before they got cancer, or that people who eat more fruits and vegetables may also do something else that reduces their risk. That's why researchers love intervention trials.

True, trials have limitations. One is that they can't readily be done for the less common cancers--like pancreatic, bladder, and ovarian--because it would require either too many people or too long a wait to get enough cancers to see a difference between groups.

Nevertheless, many researchers want intervention trials to confirm that it's the antioxidants--and not something else in fruits and vegetables--that reduce the risk of at least some cancers.

"If we had trial data for some cancers, then we'd be more inclined to trust the observational data for other cancer sites," says Meir Stampfer of the Harvard School of Public Health.

A few intervention trials have already been done. None has turned up a strong link between antioxidants and cancer.

* Two trials found that vitamins C and E had no significant effect on the incidence of colon polyps, which are precursors of colon cancer. One, a four-year study in New York, gave people a daily dose of 4,000 mg of vitamin C (67 times the USRDA) and 400 mg of vitamin E (40 times the RDA for men).(2) A two-year Canadian study used a daily dose of 400 mg each of vitamins C and E.(3) But it's possible that those trials were too small to see an effect.

* A five-year study run by Dartmouth Medical School showed that 50 mg of beta-carotene a day (equal to 83,000 IU, or roughly 17 times the USRDA for vitamin A) had no effect on the incidence of non-melanoma skin cancers.(4)

That's it as far as intervention trials go, which explains why the National Cancer Institute is "not ready to say that taking a supplement will reduce your cancer risk," says Peter Greenwald, NCI's Director of Cancer Prevention and Control. But that could change as ongoing trials are completed.

Two trials are testing beta-carotene's ability to reduce the risk of lung cancer in smokers. "We should have that answer in a few years," says Greenwald.

Other trials are testing the effect of beta-carotene--sometimes in combination with vitamin C, E, or other nutrients--on cancers or precancerous lesions of the colon, cervix, skin, stomach, and esophagus. The Physicians Health Trial, which is run by the Harvard School of Public Health, should be able to detect any major effect of beta-carotene on the most common cancers in men: prostate, colon, and lung.

"The bottom line," says Howerde Sauberlich of the University of Alabama, "is that the evidence is sufficiently strong that people should consume fruits and vegetables rich in antioxidants."

(1) Am. J. Clin. Nutr. 45: 1368, 1987.

(2) J. Nat. Cancer Inst. 81: 1290, 1989.

(3) Cancer Res. 48: 4701, 1988.

(4) N. Eng. J. Med. 323: 789, 1990.