The Placebo Effect: How It Works

This thought experiment demonstrates how the placebo effect works.

Posted Jan 10, 2012

Let's do a thought experiment. Round up 300 harried commuters with headaches—not hard to do on the New York subway any workday at rush hour. Of course, they are probably shouting and whining, which only worsens their headaches, which is precisely what you want. You reassure them that you'll get their names listed on the society pages of the New York Times in recognition of their public service (you can't afford to pay them), and that settles them down long enough for you to herd them into three soundproof rooms, 100 headaches per room.

Now the fun begins. You do nothing with the first 100. They get to glare at one another Big-Apple-style and ruminate on their throbbing temples. You make an eloquent speech to the second group, informing them that they are the lucky recipients of a newly developed and powerful painkilling miracle drug. (It's actually aspirin with codeine, a proven pain reliever.) Then you leave them alone with each other and their pain, contemplating their lawsuits against you. You make the same speech to the third group, but you are lying to them. They think you are giving them a pain-relieving drug. In truth, they get a sugar pill.

After a half-hour, you ask your 300 captives to report on their headaches. In the "do nothing" group, 20 say their headaches are gone. Eighty are still suffering. In the second group, 90 report the complete disappearance of pain; that drug is certainly a miracle potion, the people say, and they wonder where they can purchase it. In the third group, the ones you deceived, 45 still have headaches, but 55 do not. That pill did the trick, they say, happily reboarding the subway pain-free. Your experiment was a success and you are off the hook, unless one of your subjects is a liability lawyer.

But forget the legal ramifications for now. Look at what the experiment revealed. A sugar pill has no physiological action that will cure a headache, but 35 of your headache-free subjects in the third group provide evidence to the contrary. (Why 35 and not 55? Because the results from the "do nothing" group show headache pain will cease in 20 percent of your subjects after one half-hour regardless.) Thus, for 35 percent of the subjects in our thought experiment, the sugar pill was just as much of a miracle drug as the painkiller that the members of the "real drug" group received. This "cure" in the absence of any truly therapeutic agent is the placebo effect, and it's more than a curiosity. It's a direct result of brain activity. But how?

Before we answer that question, we need to define precisely what the placebo effect is. It is not spontaneous remission. That's what the 20 people in the first group (and presumably 20 more in each of the other two groups) experienced. Some of us, no matter what the disease, get better for unknown reasons. The disease process simply reverses itself without any intervention. Whether remission is mere chance or the result of some self-healing process remains anybody's guess.


Neither is the placebo effect deception or self-delusion. The people whose headaches disappear after ingestion of the sugar pill are not lying, cheating, simple-minded, or insane. Their pain disappears—and not because they consciously wish it to. In study after study, where both subjects and experimenters are "blind" to the experimental conditions—that is, no one, including the researchers, knows who is getting the placebo—measurable, clinically replicable improvements in disease conditions occur in a sizeable fraction of all cases.


Furthermore, the placebo effect is no small or insignificant statistical aberration. Estimates of the placebo cure rate range from a low of 15 percent to a high of 72 percent. The longer the period of treatment and the larger the number of physician visits, the greater the placebo effect.


Finally, the placebo effect is not restricted to subjective self-reports of pain, mood, or attitude. Physical changes are real. For example, studies on asthma patients show less constriction of the bronchial tubes in patients for whom a placebo drug works.


The placebo effect is not deception, fluke, experimenter bias, or statistical anomaly. It is, instead, a product of expectation. The human brain anticipates outcomes, and anticipation produces those outcomes. The placebo effect is a self-fulfilling prophecy, and it follows the patterns you'd predict if the brain were, indeed, producing its own desired outcomes. Researchers have found, for example:

  • Placebos follow the same dose-response curve as real medicines. Two pills give more relief than one, and a larger capsule is better than a smaller one.
  • Placebo injections do more than placebo pills.
  • Substances that actually treat one condition but are used as a placebo for another have a greater placebo effect than sugar pills.
  • The greater the pain, the greater the placebo effect. It's as if the more relief we desire, the more we attain.
  • You don't have to be sick for a placebo to work. Placebo stimulants, placebo tranquilizers, even placebo alcohol produce predictable effects in healthy subjects.

As in all brain actions, the placebo effect is the product of chemical changes. Numerous studies have supported the conclusion that endorphins in the brain produce the placebo effect. In patients with chronic pain, for example, placebo responders were found to have higher concentrations of endorphins in their spinal fluid than placebo non-responders.

At one time, researchers viewed the placebo effect as an impediment—a statistical annoyance that got in the way of objectively evaluating the efficacy of potentially legitimate therapies. That view has changed. The placebo effect is today seen as an important part of the healing process. It's been studied as a treatment for Parkinson's disease, depression, chronic pain, and more. For large numbers of patients—the placebo responders—belief in the therapy will create or enhance its effectiveness.

In some respects, the placebo effect offers the best of all possible alternatives: therapeutic effects without the risk of negative side effects. That's why dozens of brain researchers are working to sort through the complexity of the numerous brain regions and neurotransmitters that produce placebo results. Theirs is no easy task. The placebo effect is not a single phenomenon, but the result of the complex interplay of anatomical, biochemical, and psychological factors. The same can be said for all our perceptions, I suspect. We see, hear, taste, touch, and smell pretty much what we expect to.

This post has been excerpted from Chapter 4, "Pain and the Placebo Effect," from the book Brain Sense.