As people were getting ready for the holiday season and its accompanying waist expansion late last year, Dr. Mehmet Oz let viewers of his TV show in on a timely little secret. "Everybody wants to know what's the newest, fastest fat buster," said the board-certified cardiothoracic surgeon and one of People magazine's sexiest men alive. "How can I burn fat without spending every waking moment exercising and dieting?"

He then told his audience about a "breakthrough," "magic," "holy grail," even "revolutionary" new fat buster. "I want you to write it down," America's doctor urged his audience with a serious and trustworthy stare. After carefully wrapping his lips around the exotic words "Garcinia cambogia," he added, sternly: "It may be the simple solution you've been looking for to bust your body fat for good."

In Dr. Oz's New York City studio, garcinia extract - or hydroxycitric acid found in fruits like purple mangosteen - sounded fantastic, a promising new tool for the battle against flab. Outside the Oprah-ordained doctor's sensational world of amazing new diets, there's no real debate about whether garcinia works: The best evidence is unequivocally against it.

The miracle cure isn't really a miracle at all. It's not even new. Garcinia cambogia has been studied as a weight-loss aid for more than 15 years. A 1998 randomized controlled trial looked at the effects of garcinia as a potential "antiobesity agent" in 135 people. The conclusion: The pills were no better than placebo for weight and fat loss.

More recently, a group of researchers summarized the evidence for this "breakthrough" extract in a systematic review of 12 randomized trials involving 706 participants. Some trials reported short-term slimming, but the overall effects were so small and most studies were so methodologically flawed that the authors were unable to conclude that garcinia extract has an impact on body weight.

One of those authors, Edzard Ernst, has dedicated his career to analyzing research on alternative and complementary medicine; he found that the supplement may be linked to adverse gastrointestinal effects. He told us, "Dr. Oz's promotion of this and other unproven or disproven alternative treatments is irresponsible and borders on quackery."

Still, people march into pharmacies or their physicians' offices every day asking for Dr. Oz-endorsed treatments - even when these treatments are backed by the barest of evidence or none at all. Oz's satellite patients spend tremendous amounts of money on products he recommends, a phenomenon that has been dubbed the "Oz Effect." After he promoted neti pots, for example, Forbes magazine reported sales and online searches for the nasal irrigation system rose by 12,000 percent and 42,000 percent, respectively.

Who can blame his viewers? Oz may be the most credentialed of celebrity health promoters. He's a professor and vice-chair of surgery at Columbia University College of Physicians and Surgeons. He earned his degrees at Ivy League universities, namely Harvard and the University of Pennsylvania. He's won a slew of medical awards (in addition to his Emmys) and co-authored hundreds of academic articles. He's clearly a smart guy with qualifications, status and experience. It's reasonable to assume he is well-versed in the scientific method and the principles of evidence-based medicine. "Because he's a physician, that lends a certain authority and credibility to his opinions," said Steven Novella, a clinical neurologist and assistant professor at Yale University School of Medicine who has taken Oz to task for his science. "But it lends credibility to anything he says."

And lately, a lot of what Oz has been saying is downright wrong.

To support the awesome assertions about the flab-fighter Garcinia cambogia, the doctor created on his TV show an atmosphere of accessible scientific certainty. He brought out researchers and physicians in white coats who discussed what they said was compelling evidence for the weight-loss panacea. There was an inspiring testimonial from a member of the audience. Plastic models even demonstrated how garcinia could suppress appetite and stop fat from being made. The show had the same easy manner as Oprah discussing Leo Tolstoy with her book club.

Throughout the episode, Oz maintained his trademark boyish wonder and excitement as he delivered a message many of us long to hear: A pill could help us "burn fat without spending every waking moment exercising and dieting" and even combat "emotional eating." Oz peppered his excitement with some caution: "Please, listen carefully," he said with a shrug of his shoulders and his hands raised defensively in the air, "I don't sell the stuff. I don't make any money on this. I'm not going to mention any brands to you, either. I don't want you conned."

Oz has acknowledged on air that as soon as he mentions a product, manufacturers clamor to get up websites claiming their brand was endorsed by him. They put his face on pill bottles and placards in health-food stores. They link to his show's website and columns. His PR man, Tim Sullivan, told us that with every product Oz talks about, "the next morning I wake up to 50 Google alerts from companies saying 'Oz recommends raspberry ketones.' " He says Oz's legal team prosecutes these unauthorized endorsements "aggressively."

Still, Oz seems to have a penchant for peddling products. Millions follow his advice through the TV and radio, as well as his books, newspaper columns and magazine articles. And examples of his pseudoscience abound.

Take a breaking-news segment about green coffee-bean supplements that "can burn fat fast for anyone who wants to lose weight." Oz cited a new study that showed people lost 17 pounds in 22 weeks by doing absolutely nothing but taking this "miracle pill."

A closer look at the coffee-bean research revealed that it was a tiny trial of only 16 people, with overwhelming methodological limitations. It was supported by the Texas-based company Applied Food Sciences Inc., a manufacturer of green coffee-bean products. Oz didn't mention the potential conflict of interest, but he did say he was skeptical. To ease his mind, he conducted his own experiment: It involved giving the pills to two audience members for five days and seeing what would happen. Unsurprisingly, both women reported being less hungry, more energetic and each losing two and six pounds.

There are many reasons why this made-for-TV "study" would not be published in any reputable medical journal or meet the approval of Oz's peers: The sample size was minuscule. The women were not followed for long enough to know whether the effects of the supplement were real. They were neither randomly selected nor unaware of what they were taking. They also knew they were going to have to announce their weight in public to millions of viewers. That pressure, combined with a strong placebo effect, was the most likely cause of their shape change, if one can call it that at all.

A legion of doctor-bloggers has dedicated thousands of hours to dissecting and debunking Oz's claims. One of them is Steven Charlap, a preventive medicine physician in Delray Beach, Fla. "Patients were bringing in shopping carts full of different pills," Charlap recalls. "When I would ask them, 'Why do you take a certain pill?' I found very often, the response was, 'I heard about it on the Oz show.' "

To understand where his patients were getting their health advice, Charlap began watching the program. "I was shocked that someone with his credentials - someone who apparently still operates on patients and therefore must still be fully cognizant of a physician's first priority, which is to do no harm - would be recommending all types of different pills, many that had never undergone rigorous scientific scrutiny, as miracle cures or magic pills to a very susceptible audience."

One of the first Oz-approved products Charlap looked into was milk thistle. Oz suggested the supplement as a "quick fix" for nights when you have one too many gin and tonics. The herbal remedy, according to Oz, "boosts your liver's enzyme function, which helps to detoxify the body from excess alcohol."

But Charlap noted that while some studies suggest milk thistle may be helpful for people with liver disease, the evidence is unclear on its other health benefits. So when a woman on milk thistle came into Charlap's clinic for her annual wellness visit, he asked if she had any problems with her liver, an abnormal liver-function test, or any other medically justifiable reason for using the product. The answer was no: She used milk thistle because her "other doctor" told her to do so. She was taking Oz's medicine.

Oz, Charlap noted, has also encouraged people to take two baby aspirin every night before bed to prevent heart attacks. For people at high risk for coronary heart disease, the authoritative U.S. Preventive Services Task Force would agree with him. But for healthy and older folks, aspirin can have damaging side effects - including bleeding ulcers - which are well-documented and may outweigh any potential benefits. Aspirin can also hurt patients who are on anti-coagulants or who have a history of gastric or stomach ulcers, a warning Oz does not mention on his show. An exasperated Charlap asked: "Where is the 'first, do no harm' when he does something like that?

Beyond potential damage to people's health and purses, this kind of peddling can also foster doubt and mistrust of science. As Edzard Ernst put it: "Prominent people like Oz do have considerable influence. If this influence is used to promote quackery, bogus treatments will seem credible. Using bogus treatments for serious conditions may cost lives."

So how are we supposed to tell medicine from miracles? As a general rule, said Victor Montori, an evidence-based medicine guru at the Mayo Clinic, "If studies are cited, then this cannot be, at the same time, a secret revealed just to you now. If the studies are any good, the effects are usually very small." Referring to Oz's holiday weight-loss advice, he added: "It is very unlikely that an important compound hidden in the garcinia could have a big effect."

We can also arm ourselves with the knowledge that not all evidence is created equally, and celebrities - even famous doctors - are not credible sources of health information.

Some other rules for sifting nuggets of good evidence from gobbledygook include: Research involving humans is typically more relevant than animal models; prospective, randomized, controlled trials are usually better than retrospective, observational analyses; large studies are better than small studies; multisite studies are better than single-site; and systematic syntheses of all the available evidence are more informative than individual studies presented out of context.

Our own doctors will probably know more about what's good for us than our favorite celebrities, their doctors, or even America's doctor, the wonderful wizard that is Oz.

------

Julia Belluz is the Science-ish blogger, a senior editor with the Medical Post, and contributing writer for Maclean's magazine.

Steven J. Hoffman is an assistant professor at McMaster University and the McMaster Health Forum, a Trudeau Scholar, and an instructor at the Harvard School of Public Health.