RSS feed of this blog

Doctor takes a close look at the power of the placebo effect

Editor’s note:  This column by Dr. Patrick Neustatter will appear in the Healthy Living section on Sunday, March 25, 2012.


It is an unsettling realization that I am one of those people who might as well go to a witch doctor—who get better just because they believe the treatment they are taking is doing them good.

A couple weeks’ ago, columnist Chris Lillis, a local physician, confessed to being obese. And now I’d like to make my own confession: I take Prozac.

I mentioned this in a column a few years ago, but I’m still not comfortable admitting it to people. I rationalize it to myself that I only take a small dose and only irregularly.

But I am willing to share my guilty secret, partly because we need to break down this attitude that our mental foibles are so taboo that we can’t mention them.  I am not at all embarrassed to tell you I take ibuprofen for my aching back, by contrast.

But primarily, I’m sharing that I take Prozac because while it seems to work—on things like impatience, irritability and negativity, which I fear are ingrained in my nature—there is evidence that the chemical in it isn’t really working.


Zillions of other people take antidepressants. But if you look at statistics from drug trials, there is reason to believe we’re responding to the hocus–pocus factor. The placebo effect is what I’m talking about. (Placebos are those supposedly inactive sugar pills.)

Since Prozac was released by Eli Lilly in 1987, sales of antidepressants have increased 400 percent, and these “tremendously effective” medicines have become an $11.3 billion business, according to the drug industry’s parent company PhRMA.

But the claim from a more in-depth analysis of the trials on antidepressants is that it is not the chemical in the medicine that is getting people better; it is the placebo effect.

“People get better when they take the drug, but it’s not the chemical ingredients of the drug that are making them better,” claimed Dr. Irvin Kirsch, in an interview on the news program “60 Minutes.”

Kirsch is associate director of the Program in Placebo Studies and the Therapeutic Encounter, which Harvard Medical School was prompted to set up to investigate this issue.

As may be expected, this is a bit controversial, as many patients taking antidepressants and the doctors prescribing them find them effective. There are detractors from the idea that the effectiveness all stems from the placebo effect, including Dr. Michael Face, professor of psychiatry at the University of Pennsylvania. The claim of people like Face is that antidepressants are effective—for severe depression, at least.

Regardless of the controversy over antidepressants, the placebo effect is the issue I’m interested in, and it seems to be a powerful business.


The Harvard program and other researchers have discovered much about the power of the placebo, including:

  • Placebos used for pain relief cause the brain to produce its own version of pain-relieving endogenous morphine (endorphins). And these, just like opiate pain medicines, can be blocked by the narcotic antidote Naloxone.
  • Different size, color and type of pills or capsules influence how effective the placebo is.
  • Valium doesn’t work unless patients know they are taking it.
  • There is also a “nacebo effect,” where if you believe something is going to have an adverse effect, it is much more likely to. This is the downside of being a fully informed consumer and getting that massive list of possible adverse effects the pharmacist gives you with your prescription.
  • If you rub poison ivy on one arm and tell the person it’s some benign leaf, and then rub the benign leaf on the other arm and say it’s poison ivy, the person will react on the side believed to have been touched by the poison ivy (even though it’s not).
  • Even if the patient knows he’s taking a placebo, it still works.

This last point is a crucial new finding from a recent trial of 80 Irritable Bowel Syndrome patients done at Harvard where one group was given no treatment and the other group placebos—with full disclosure.

“We actually had ‘Placebo’ printed on the bottle,” explained alternative therapist Ted Kaptchuk,  director of the Harvard program.

By the end of the three-week trial, nearly twice as many subjects in the placebo group reported adequate symptom relief (59 percent versus 35 percent)—equivalent to the effects of the most powerful IBS medicine, Kaptchuk notes.

The placebo effect has been recognized by the medical profession for years, but it has been treated as an embarrassing and funky quirk that is best not talked about.

And the idea of giving patients placebos intentionally was considered unethical—though giving someone an expensive pill with the potential for dangerous side effects that may be no better than a placebo seems to me to be the ethically questionable option.


“What exactly is the placebo effect?” That’s the question everyone is asking. The answer seems to lie in what a person believes in.  In the Harvard trial, researchers told their subjects that placebos are shown to have an effect even though they are only a “sugar pill.”

Just having someone, like a doctor or other therapist, doing something that you think is going to get you better seems to be the key. The improvement is reinforced by that person being supportive—which is where alternative therapists so often win out over regular doctors.

It has always been my belief that the boost in well-being that people so often enthuse about when they are on some elimination, cleansing or detoxification diet—or other kind of radical life style change—is as much because they are “doing something” and taking charge of their lives as because of the specifics of the diet it self.

This is what I’m referring to when I say you might as well go to a “witch doctor.” Witch doctors invoke intensely powerful beliefs—to the point where people can die from a negative belief.

This placebo effect can be a serious confounding factor when you are trying to test drugs or other treatments in a clinical trial. As soon as you enroll people, you are “doing something.” Hypertensives subjects (people with high blood pressure) have been noted to drop their blood pressure just by being recruited, before getting any treatment.

Another study of asthmatic patients, who were treated with a new drug, illustrates the blinkered attitudes of the medical profession and elegantly epitomizes this whole issue.

The trial was considered a failure because it had no impact on the chemical marker being used to decide if the new drug was effective. But the patients felt better!

There is a saying in medicine that the more empathetic, less scientific doctors have been quoting for years. “Treat the patient, not the test.”

ON THE NET: Harvard Program in Placebo Studies & The Therapeutic Encounter:

Dr. Patrick Neustatter can be reached at healthyliving@freelance


To understand how this claim that antidepressants are no better than placebo can come about, you have to understand the whole process of drug evaluations.

To bring a drug to market, the manufacturer must present at least two trials to the FDA showing the drug to be statistically more effective than a placebo. What the manufacturers do not have to do is show any trials that didn’t demonstrate the drug was significantly superior to placebo. There may be lots of those, but they get buried and never seen—normally.

This is euphemistically called “publication bias.” So, the FDA only gets to see the trials that look good.

Even the published trials on antidepressants didn’t show them to be all that much better than placebos—something I used to torment the drug reps with.

“Effexor was sixty percent effective in relieving depression,” the enthusiastic rep would come and tell us in his bold and assertive way, showing us a slick and glossy brochure. But if you looked carefully, the brochure showed the placebo was  50 percent effective. (These may not be the exact numbers, but they’re something like that.) “The placebo’s pretty damn effective,” I would chide.

The research showing that antidepressants had no statistically significant advantage over placebos came from a Freedom of Information Act request. Researchers got the unpublished data on antidepressants and then performed a meta analysis—where they did a statistical overview of the results of all the studies.

The researchers concluded that the antidepressant effect of these expensive medicines, with their potential side effects, is all just placebo effect.

In 2010, an article in the Journal of the American Medical Association seemed to validate that view, saying, “The magnitude of benefit of antidepressant medication compared with placebo  may be minimal or nonexistent, on average, in patients with mild or moderate symptoms.”

In other words, for people with mild or moderate depression, you might as well take M&M’s—providing you believe in them.

And this is the point. Whether you go along with the proponents or detractors on antidepressants’ effectiveness, what seems indisputable is that placebos have a significant effect.

It doesn’t mean antidepressants don’t work or that doctors don’t see improvement in their patients—it’s a matter of how the drugs work.

The placebo effect is powerful; it seems to come from the patient’s belief that he or she is receiving some beneficial treatment—which also has implications for diets and alternative treatments.

—Dr. Patrick Neustatter