Supplements

Snake Oil Supplements

David Mendosa Health Guide January 06, 2008
  • Raise your right hand if you don't take any herbal supplements.

    Gee, I don't see any hands, and I probably wouldn't see more than a handful if you were here with me literally instead of virtually.

    People with diabetes probably take more supplements than other people. About 36 percent of American adults use complementary and alternative medicine (CAM), much of it in the form of supplements. We spend $20 billion a year for supplements.

    Aside from the drain on our bank accounts, we don't have any good proof that any supplements work. None of them.

    Supplements are, of course, supplemental to the foods that we eat, our diet. We certainly differ in what we think is a good diet, although we haven't subjected diets and individual foods to clinical trials. That could be coming, but we haven't yet mastered the testing of supplements.


    "There is no compelling, credible scientific evidence to suggest that any CAM therapy benefits any medical condition or reduces any medical symptom (pain or otherwise) better than a placebo." This is the conclusion that R. Barker Bausell, Ph.D., reaches on page 254 of his book, Snake Oil Science: The Truth About Complementary and Alternative Medicine (Oxford, 2007, $24.95). He is a bio-statistician and a professor at the University of Maryland. Dr. Bausell was formerly research director of the University of Maryland's Complementary Medicine Program.

     



    With his direct and entertaining style, Dr. Bausell's book is accessible to both a lay and professional audience. He addresses people with diabetes at several points, specifically on page 292, "If you have a potentially life-threatening condition, don't discontinue your insulin...or other treatment, because no CAM therapy is going to come close to matching these drugs' potency.

    Snake Oil Science came out in October, but I didn't hear about it until Christmas day, when I read Abigail Zuger's glowing review in The New York Times. I rushed right out and bought my own copy as soon as I could.

    Dr. Bausell's book is the perfect complement to The Natural Pharmacist and Complementary & Alternative Medicine (CAM) Supplement Use in People with Diabetes: A Clinician's Guide. I previously reviewed this website and book here.

    Why then do we subject ourselves to so many useless supplements? Dr. Bausell says that the placebo effect is the key to how well CAM works. "CAM therapies are nothing more than cleverly packaged placebos," he writes on page 275.

    The trouble is that most of these therapies do work -- although weakly, temporarily, and subjectively. While the placebo effect is the main culprit, other stumbling blocks include regression to the mean, the Hawthorne effect, and what he calls "natural history."

    Dr. Bausell uses the term natural history in the statistical rather than the usual sense. It's natural for a disease to get better or worse during the course of an illness whether or not it's treated. We also commonly start treatment when the disease is at its worst. When it naturally gets better, we credit the treatment instead of our body.


  • Because our minds play these tricks on us, we don't have any options besides randomized, double-blind, placebo-controlled trials when we evaluate any therapies. These trial are the gold standard. But while they are necessary to protect our bodies from our minds, they aren't sufficient.

    They need to be large studies too. Dr. Bausell says that credible trials must have at least 50 patients per group and preferably more than 100.

    Watch out too when a lot of the people in the trials drop out. For the trials to be credible they must have relatively minor attrition -- preferably less than 20 percent, but certainly less than 25 percent.


    Studies published in "high-quality journals" are better than the thousands of other journals, even if they are "peer-reviewed." The runaway leaders among general (non-disease-specific) American medical journals, Dr. Bausell writes on page 178, are the New England Journal of Medicine and the Journal of the American Medical Association (JAMA). My guess is that in diabetes specifically, they may be Diabetes and Diabetes Care, which the American Diabetes Association publishes.

    We have a relatively objective way to determine which journals are high quality. That is to look at what is called "journal impact," which is based upon the number of times a give journal's research articles are cited by other researchers.

    Also, those trials that independent investigators validate are the most credible. Multi-center trails are one of the best ways to get this independent replication.

    We also have to watch out for what Dr. Bausell calls bias from scientific acculturation, where certain countries publish essentially nothing but positive results. In practice that means that 98 percent of conventional Chinese trials produced positive results, as did 97 percent of Russian trials.

    If you want to keep taking the supplements in your pillbox or add new ones, these are the most important points to consider in your checklist. I take some supplements myself knowing full well that we have no proof of their effectiveness -- but I am too old to wait for proof.

    So I carefully consider the clinical trials of the supplements I'm interested it. Dr. Bausell brought to my attention studies of three supplements that I have been taking. As a result, as soon as my current supplies run out I will stop taking milk thistle (Silybum marianum) to protect against liver disease, saw palmetto to reduce the symptoms of BPH, and possibly ginkgo to improve my memory. The first two of these have highly credible studies showing that these supplements are no better than placebos. The ginkgo study meets Dr. Bausell's standards, but I wonder.

    The ginkgo study was a "randomized, double-blind, placebo-controlled, parallel-group trial." But it lasted only six weeks.

    I wondered if six weeks is long enough to prove or disprove whether herbs or drugs can establish the necessary neural connections. I wondered enough about this to ask Dr. Bausell directly.


  • "I agree with you to a certain extent about the length of, say, a six-week study," he replied. "But my guess is that if this particular trial had been extended it would have still been negative. You can always find reasons why a study doesn't come out the way you expect: the dose was wrong, the wrong outcome measure was employed, the length was wrong, the inclusion/exclusion criteria were inappropriate, and so forth. What I tried to do was look at the gestalt of evidence for the entire field."

    So I am surely wasting my money on some (if not all) of the supplements I take. So be it. Still, I want to warn you away from at least the wildest claims for snake oil supplements.