Antioxidants: still hazy after all these years.

"Coffee is America's No. 1 source of antioxidants," reported ABC News in August. "In chocolate, cocoa means higher antioxidant activity," announced the U.S. Department of Agriculture in April. "Healthy antioxidants!" beckons the label of Bigelow Green Tea.

In recent years, major clinical trials have found that antioxidants like beta-carotene and vitamin E failed to reduce the risk of heart disease and cancer. Yet the buzzword is showing up on more foods than ever.

"People don't like to let go of easy, simple answers." suggests Alice Lichtenstein of the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University in Boston.

And now there's a new wrinkle: antioxidants may help people with certain variations of some genes but not others. Here's the latest on what was supposed to be the cure-all, prevent-all answer to cancer, heart disease, and more.

"The evidence that antioxidants play a role in preventing cancer has diminished in the last 10 to 15 years, and we're not facing up to the rather dismaying facts," says Regina Ziegler of the National Cancer Institute in Bethesda, Maryland.

However, Ziegler and other experts are not ready to completely rule out the possibility that antioxidants are protective. Indeed, scientists are racing to uncover new evidence that some antioxidants may reduce the risk of cancer, but only in people with some variations of certain genes.

Yet the simple image of antioxidants as valiant warriors that protect the body from rampaging free radicals is, well, toe simplistic. Unfortunately, that message hasn't reached most people ... and the companies that sell antioxidant-rich foods and supplements would clearly like to keep it that way.

Cancer & Carotene 

The antioxidant theory has always sounded convincing. Free radicals cause cancer. Antioxidants can quench or capture free radicals.

"If oxidation damages DNA, it could cause cancer by preventing the DNA from replicating accurately," explains the National Cancer Institute's Regina Ziegler. "Beta-carotene, vitamins C and E, and glutathione peroxidase, a selenium-containing enzyme, were supposed to limit oxidative damage."

It wasn't just a hypothesis. Evidence from test-tube experiments and animal studies backed it up. The most intriguing observation in humans: smokers who ate more fruits and vegetables rich in beta-carotene had a lower risk of lung cancer.

In the early 1980s, the NCI sprang into action. It launched three large randomized trials with nearly 70,000 people to see if high-dose beta-carotene pills (33,000 to 50,000 IU a day) could cut the risk of lung cancer.

The results: in one trial (mostly of nonsmoking doctors), beta-carotene did nothing. (1) That was the good news.

The two other trials (mostly of smokers) had to be cut short because beta-carotene takers had a slightly higher risk of lung cancer and heart disease than placebo takers. (2,3)

"We later found that at high levels, beta-carotene gets metabolized into compounds that are carcinogenic in smokers," says Tufts University's Alice Lichtenstein. In animals exposed to smoke, high doses of beta-carotene caused precancerous lesions, while low doses (like those in foods) seemed to curb lung damage. (4)

The beta-carotene flop doesn't mean that all antioxidants are useless or dangerous. The NCI is still testing whether selenium and vitamin E can reduce the risk of prostate cancer in 35,000 men in the SELECT trial. But some of the enthusiasm has dissipated.

"As we've learned more about cancer, we haven't accumulated persuasive evidence that oxidative stress or damage plays a critical role in cancer progression," explains Ziegler.

E for the Heart?

As with cancer, researchers had good reason to think that antioxidants could thwart heart disease.

People who reported taking vitamin E supplements had a lower risk of heart disease. And experiments on LDL ("bad") cholesterol seemed to back up the theory.

"If you isolate LDL from blood and subject it to oxidizing conditions in a test tube, the more vitamin E the person consumes, the less oxidation there's going to be," explains Lichtenstein.

"But you can get burned by extrapolating from test tube studies to people," she adds, "because in the body there are so many things that buffer that reaction."

How did researchers get burned? In seven trials on 81,000 people who already had or were at high risk of heart disease, those who were given vitamin E did no better than those who took a placebo. (5)

In July, researchers struck out in the first big trial testing vitamin E on healthy women. Nearly 40,000 women were randomly assigned to take vitamin E (600 IU of natural alpha-tocopherol), aspirin (100 mg), both, or a placebo every other day for ten years. The researchers found no difference in heart attacks, strokes, or cancers of the lung, colon, or breast. (6)

"We found no overall benefits of vitamin E," says JoAnn E. Manson of Brigham and Women's Hospital and the Harvard School of Public Health in Boston.

Vitamin E takers who were at least 65 years old did have a 34 percent lower risk of heart attacks. But that was probably just due to chance, says Manson.

It's not biologically plausible that vitamin E would help only older women without heart disease, she contends. "Other trials found no benefit in people with prior heart disease, and we found no benefit for younger women."

If it doesn't help women early or late in the artery-clogging process, "what sense would it make for vitamin E to benefit women in the middle?" asks Manson.

What's more, some studies suggested that people who were given vitamin E had a slightly higher risk of dying or of congestive heart failure. (7,8)

Add it all up, says Manson, and it's clear that "vitamin E is not the elixir of youth or the answer to cardiovascular disease prevention."

"The theory that antioxidants may have benefits was promising," she adds, "but there's no convincing evidence that taking vitamin E or other antioxidants will confer cardiovascular benefits."

Good Genes

Have antioxidants faded off the research radar screen? Not quite. Dozens of recent reports suggest that differences in genes may explain whether antioxidants reduce the risk of disease.

"The whole field is in a state of rapid change as we try to get a grip on the genetic data," says Michael Leitzmann of the National Cancer Institute.

For example, he and colleagues at Harvard University found that higher blood levels of selenium, vitamin E, or lycopene were linked to a lower risk of prostate cancer, but only in men with one version of a gene that directs the body to make an enzyme called manganese superoxide dismutase (MnSOD).

MnSOD acts as an antioxidant in mitochondria--the power-producing centers in every cell of the body. If men with the more-active AA version of the MnSOD gene had low blood levels of antioxidants, their risk of aggressive prostate cancer was ten times higher than if they had high levels of antioxidants. (9) But antioxidants had less impact on men with the less-active VV or VA version of the MnSOD gene.

"We need more data to make sure that it isn't just a fluke," says Harvard University's Meir Stampfer, who collaborated with Leitzmann on the study. "But if these findings hold up, it means that higher levels of antioxidants would be most beneficial for people with this common version of the gene."

(Researchers found it in one out of four white men, but have no data on most other ethnic groups.)

The study also found no lower risk of fatal prostate cancer in men who were given beta-carotene supplements. But when the researchers looked only at men with the AA version of the MnSOD gene, beta-carotene cut the risk by more than 60 percent.

"These studies have the potential to clean up the many inconsistencies reported in diet and cancer," explains Leitzmann. "If we find that subsets of the population are genetically more prone to react to an exposure, it would help diminish some of the noise in the system."

Dozens of other studies have reported a lower risk of some cancer (like breast, lung, colon, or stomach) in people who eat more of something (like fruits, vegetables, fish, or green tea), but only if they have one version of a gene (with an alphabet soup of names like GST1, COMT, and XRCC1). And those studies are not just looking at antioxidants.

For example, people have different versions of a gene (called methylenetetrahydrofolate reductase, or MTHFR) that metabolizes folate, a B-vitamin. (10) One version "slows down the metabolism of one form of folate to another," says Stampfer. "So these people have a buildup of the upstream form, which seems to protect against colon cancer."

Those people are more sensitive to folate intake, he says. "When they eat enough folate, they get more protection than people who eat the same amount of folate but don't have that version of the gene."

Don't Jump

Tantalizing as those results are, it's too early to jump to conclusions.

"Some of these analyses have problems being reproduced," says Leitzmann. "It would be premature to base any recommendations on these first studies."

Part of the problem is that many genes may alter the risk of disease. "In the gene world, there are tens of thousands of alterations in the genetic code that could make a difference," says Leitzmann. "We have to look at complex combinations to see if patterns emerge."

What's more, when scientists conclude that their findings are statistically significant, they mean that the odds are only 1 in 20 that the results are due to chance. But given enough results, 1 in 20 is a sizeable number.

"If you look at 10 cancers, 10 genes, and 10 antioxidants, you've got 1,000 possible interactions," explains the National Cancer Institute's Regina Ziegler. "If 1 in 20 is significant by chance alone, that's 50 significant interactions that may not mean anything."

"And the odds are actually higher that promising results will make news because studies that find nothing are less likely to get published."

The answer is to look not just at single genes, but at a bunch that work together, say researchers. And they're pooling their data so they can look at these genes in thousands of people.

Researchers also need large numbers of people because they're dividing the population into smaller groups.

"If you start with 1,000 people and divide them into four levels of, say, vitamin E intake and then you look at one or two genes that control how the body handles vitamin E for each level of vitamin E intake, you'll have only a handful of people in each category," explains the NCI's Demetrius Albanes. "So the big wave is to combine studies."

In any case, scientists agree that people can't use the findings yet. "I wouldn't go out and genotype myself," says Stampfer. "I would eat enough healthful foods, and if I'm lucky to have the genotype that's sensitive to them, that's good."

The Bottom Line

What to do while researchers sort it all out?

First, it may be useful to stop thinking of antioxidants as a category. "Vitamin E is very different from selenium, which is very different from lycopene, for example," says Stampfer. "They share antioxidant properties but they also have unique properties."

Second, it's useful to remember that the evidence for antioxidants started with food. "We lump all antioxidants together and they have this aura of being wonderful," says Lichtenstein. "But in reality, it's dietary patterns high in fruits and vegetables that are associated with a lower risk of disease."

It may not be antioxidants, but something else in fruits and vegetables, that protects health. Or it could be that people who eat more fruits and vegetables eat fewer damaging foods.

"We say we're going to eat fruits and vegetables because of their antioxidants or their phytochemicals," says Lichtenstein. "But one of the biggest reasons to eat fruits and vegetables is that it means you're not eating lots of brownies, candy, or fat-free ice cream or drinking a lot of soda."

This timeline shows only a selection of clinical trials, but researchers also consider evidence from other human (and animal) studies before they reach conclusions. Most of the clinical trials looked not just at heart attacks and strokes or cancer, but at angioplasty, coronary bypass, angina, congestive heart failure, and other cardiovascular events. The number of participants in each study has been rounded and may include people on statin drugs, aspirin, and other treatments.

29,000 Finnish male smokers take beta-carotene (33,000 IU), vitamin E (110 IU), both, or a placebo daily for 5 to 8 years. Lung cancer rates are slightly (18%) higher in beta-carotene takers. In a later study of the same men, beta-carotene takers who had previous heart attacks were more likely to die of heart disease.

New Eng. J. Med. 330: 1029, 1994.

In the Physicians' Health Study, 22,000 healthy men take beta-carotene (83,000 IU) or a placebo on alternate days for 12 years. No difference in the risk of cancer, heart disease, or diabetes.

New Eng. J. Med. 334: 1145, 1996 and J. Amer. Med. Assoc. 282: 1073, 1999.

In the Heart Protection Study, 20,000 people at high risk of heart disease take vitamin E (2,300 IU), vitamin C (250 mg), and beta-carotene (33,000 IU) or a placebo daily for 5 years. No difference in heart attack rates.

Lancet 360: 23, 2002.

In the CARET study, 18,000 smokers, former smokers, and workers exposed to asbestos take beta-carotene (50,000 IU) and vitamin A (25,000 IU) or a placebo daily for 4 years. Supplement takers have a 28% higher risk of cancer.

New Eng. J. Med. 334: 1150, 1996.

In the HOPE-TOO trial, 7,000 people with either cardiovascular disease or diabetes take vitamin E (400 IU) or a placebo daily for 7 years. Vitamin E takers have slight (13%) increased risk of heart failure and no lower rate of cancer.

J. Amer. Med. Assoc. 293: 1338, 2005.

In the Women's Health Study, 40,000 healthy female health professionals take either vitamin E (600 IU), aspirin (100 mg), both, or a placebo on alternate days for 10 years. No difference in heart disease, stroke, or cancer rates.

J. Amer. Med. Assoc. 294: 56, 2005.

In the Women's Antioxidant and Cardiovascular Study, 8,000 women with cardiovascular disease ease are taking vitamin E (600 IU every other day), vitamin C (500 mg every day), and/or beta-carotene (83,000 IU every other day), or a placebo.

2006

In the SELECT study, 35,000 men with elevated PSA levels are taking vitamin E (400 IU), selenium (200 mcg), both, or a placebo daily.

2013

In the Physicians' Health Study II, 15,000 male physicians are taking vitamin E (400 IU every other day), vitamin C (500 mg every day), and/or beta-carotene (83,000 IU every other day), or a placebo.

2007

(1) New England Journal of Medicine 334: 1145, 1996.

(2) New England Journal of Medicine 330: 1029, 1994.

(3) New England Journal of Medicine 334: 1150, 1996.

(4) Carcinogenesis 21: 2245, 2000.

(5) Lancet 361: 2017, 2003.

(6) Journal of the American Medical Association 294: 56, 2005.

(7) Journal of the American Medical Association 293: 1338, 2005.

(8) Annals of Internal Medicine 142: 37, 2005.

(9) Cancer Research 65: 2498, 2005

(10) Journal of Nutrition 133: 3745S, 2003.

Beyond Vitamins

Researchers may be losing interest in antioxidant vitamins, but the food industry is just starting to stir it up. You'll find vitamins C and E added to foods like Tropicana Healthy Heart orange juice, Kellogg's Smart Start Antioxidants cereal, and Quaker Take Heart oatmeal.

What's more, the industry is moving beyond vitamins to other candidates. "Now we have a new crop of potential antioxidants," says National Cancer Institute investigator Regina Ziegler.

Furthest along the way to celebrity status are polyphenols. They include anthocyanins (in grapes), catechins (in chocolate, tea, and red wine), flavonols (in onions and apples), hydroxycinnamic acids (in coffee), isoflavones (in soy), and more. (1)

"But people have to recognize that we haven't found consistent evidence for these new antioxidants, either," notes Ziegler. For example, recent studies found no link between flavonoids and breast cancer or heart disease. (2,3)

Among the foods that are gaining antioxidant star appeal are chocolate, coffee, and green tea.

(1) American Journal of Clinical Nutrition 81: 230S, 2005.

(2) International Journal of Cancer 114: 628, 2005.

(3) Annals of Internal Medicine 125: 384,1996.

CHOCOLATE

The chocolate industry is working hard to turn candy into a health food.

Last year, Nestle sponsored the 1st International Conference on Polyphenols and Health in Vichy, France. In April, the U.S. Department of Agriculture reported that it had analyzed antioxidants in chocolates, thanks to funding from the industry's American Cocoa Research Institute.

And in July, Mars (makers of M&Ms and other candies) and the Harvard Medical School put on a two-day seminar in Switzerland to discuss the latest research on cocoa's potential health benefits.

"Mars says it is in talks with large pharmaceutical companies for a licensing or joint venture agreement to reproduce the compounds in cocoa shown to improve blood flow," noted USA Today in its report on the conference.

That's fine with researcher Alice Lichtenstein of the USDA Human Nutrition Research Center on Aging at Tufts University in Boston.

"If polyphenols in chocolate are effective at lowering blood pressure or improving blood flow, they should take the compounds out of chocolate and test them as rigorously as any drug," she says. "That way, people can eat them without all the fat and sugar in chocolate."

A recent Italian study reported that 3V2 ounces of dark chocolate lowered blood pressure and boosted insulin sensitivity. (1) "With only 15 people, it was a very small study," Lichtenstein points out. More importantly, "these people were eating 480 calories' worth of chocolate a day."

(It may not be polyphenots, but theobromine--a cousin of caffeine--or something else in chocolate that's responsible, note the authors of the Italian study. (2) That would be bad news for antioxidant advocates, but good news for chocolate eaters.)

Do Americans really need encouragement to eat more candy? "If you put hot fudge on premium ice cream, you're not neutralizing what's in the ice cream, you're just getting fatter," says Lichtenstein. "If people start adding chocolate to their diets, it would be a travesty because we're all too fat anyway."

(1) American Journal of Clinical Nutrition 81: 611, 2005.

(2) American Journal of Clinical Nutrition 82: 487, 2005.

COFFEE

Coffee is a newcomer to the antioxidant scene. It leaped to fame with a study that crowned coffee "America's No. 1 source of antioxidants."

"Antioxidants are your army to protect you from the toxic free radicals which come from breathing oxygen and eating sugar, that start chronic diseases," study author Joe Vinson of the University of Scranton in Pennsylvania told ABC News in August. "Anti-oxidants help stave off cancer, heart disease, diabetes, and stroke."

But coffee leads the pack in Vinson's study in part because we drink so much of it. "If we were eating more fruits and vegetables, they would be our biggest source of antioxidants," explains Lichtenstein.

What's more, it's not clear that Vinson or others know how to measure antioxidants in food, she adds. "They measure something called antioxidant capacity in test tubes, but we don't know what that means."

Researchers have no data to show that antioxidant capacity in the lab translates into how well the body staves off oxidation. For example, Lichtenstein explains, "at high levels, some antioxidants become pro-oxidants"--that is, they promote oxidation.

More importantly, so many people drink so much coffee that studies can pick up any links with disease (whether or not anti-oxidants are responsible).

Though coffee doesn't seem to protect against heart disease or most cancers, some studies suggest that coffee drinkers have a lower risk of Parkinson's disease' and gallstones (2) and that people who drink four or more cups a day have a lower risk of diabetes. (3) And liver cancer is less common in coffee drinkers in Japan (where the disease is more common than in the U.S.). (4)

"It's not as though someone is trying to withhold some incredible findings about coffee," says Lichtenstein.

(1) Annals of Neurology 50: 56, 2001.

(2) Gastroenterology 123: 1823, 2002.

(3) Journal of the American Medical Association 294: 97, 2005.

(4) International Journal of Cancer 116: I50, 2005.

GREEN TEA

Lipton Green tea "naturally contains protective antioxidants," says the box. "This is important because anti-oxidants oxidants can help the body protect itself against radicals--molecules that can damage cells."

To look at Lipton's labels, ads, and Web site (www.teaandhealth.com), you'd think that drinking green tea is a slam dunk for anyone who wants to avoid cancer.

Yet in June, the Food and Drug Administration turned down a petition from a small company, Dr. Lee's TeaForHealth, to claim that green tea reduces the risk of cancer. Well, the FDA didn't exactly deny the request. Instead, the agency agreed to allow this not-exactly-glowing claim:

"Two studies do not show that drinking green tea reduces the risk of breast cancer in women, but one weaker, more limited study suggests that drinking green tea may reduce this risk. Based on these studies, FDA concludes that it is highly unlikely that green tea reduces the risk of breast cancer."

Of course, you'll never see that claim on a label (though you'll find a misleading distortion of it on TeaForHealth's Web site). Nor will you see a similar claim that the FDA approved for green tea and prostate cancer. And the agency decided that "there is no credible evidence supporting a relationship between green tea consumption" and colon lung, stomach and six other cancers. (1)

Even if studies find a lower risk of, say, breast cancer green tea drinkers, something else about those women may De protecting them. "Green tea may De an indicator of an Asian lifestyle that includes lower bogy weight, more exercise, lower consumption of meat, and a greater maKe of fruits, vegetables, and grains," explains the National Cancer Institute's Regina Ziegler.

And a recent study suggests that green tea may lower the risk of breast cancer only in women with a less-active version of a certain enzyme. (2)

"Green tea may not help people with the more-active enzyme because the tea doesn't hang around long enough for you to derive the benefit," speculates researcher Anna Wu of the University of Southern California.

Interesting, but way too early to Know.

(1) www.ctsan.fda.gov/~dms/qhc-gtea.html.

(2) Cancer Research 63: 7526, 2003