Ah, the magic of.... placebo

Placebos are incredibly neat. This is why I am disappointed that they are typically described in undergraduate classrooms simply as a type of control. While they are certainly critical research controls, they are also a powerful example of how the brain, as much as the external world, creates our reality.

Work on placebos started after World War II, primarily from the work of Henry Beecher. Beecher, who had tended to wounded soldiers during the war, had witnessed saline solutions providing powerful pain relief to wounded soldiers during times when morphine supplies had run out.

Generally, we know that a placebo can provide relief from a variety of symptoms, including pain, Parkinson’s symptoms, ADHD, depression, allergies, epilepsy and more.  The relief from placebo isn’t just from patients telling doctors what they want to hear –when naloxone, a drug that blocks opiate receptors, is given to people who are taking a placebo pain reliever, the pain relief vanishes, suggesting that the placebo was activating some of the same neural systems activated when “real” drugs are taken. Similarly, dopamine is released during a placebo treatment for Parkinson’s, just as with the active treatment. An excellent summary of the neurobiology of placebo research can be found here.

Of course, not all placebos are created equal. For example, placebos that are injected are more effective than pills, and among pills, gelatin capsules are more effective than tablets. The price of the placebo also seems to alter its effectiveness. For example, in a clever 2008 study from Dan Ariely’s research group, participants were given mild electric shocks to their wrists before and after taking a drug they were told was an experimental pain killer that is similar to codeine. The “drug” was, of course, a placebo. Critically, half of the participants were told that the pills cost $2.50 each, while the other half of participants were told that each pill had been discounted to $0.10 each. It turns out that over 85% of participants in the high-priced drug condition experienced pain relief while only 60% of participants in the low-priced drug did.

To further complicate matters, not all people respond to a placebo to the same degree. In general, the more willing you are to expect a good outcome, the more of a placebo effect you will have. This is why children have more of an effect than adults (I mean, they believe in the Tooth Fairy, too!), and severely depressed patients have very little of a placebo effect.

It is crucial to consider studies such as these when considering clinical trials for new drugs. The gold standard for such trials is to compare the new drug to a placebo. If the drug has a greater effect than placebo and has an acceptable level of negative side effects, then it is typically approved by the FDA. While it is unlikely that clinical researchers are making their placebos intentionally less effective by making them less believable (these studies are done “double blind” meaning that the researchers don’t know which participants are on the drug, and which are on the placebo until the end of the study, so doing this would make the drug less effective too), we do need to consider that the effect size of the same drug is going to vary widely as a function of the placebo used, and the type of participants in the study.

This speaks directly to the recent controversy over the effectiveness of anti-depressant medication. The media creates headlines such as “The depressing news about anti-depressants”, which hide a rather nuanced issue: if the severity of depression affects the strength of the placebo effect, then not all studies of depressed patients will show a statistically significant effect of anti-depressants. This is shown graphically in this figure from a 2008 meta-analysis of antidepressant effects: while the effect size of the anti-depressants (red triangles) were more or less constant over a wide-range of depression levels, the effect size of the placebo (gray circles) decreased as the severity of depression increased. This means that in only the most depressed patients is there a significant effect of the antidepressant, but not because the drug is less effective in less depressed people.

…And placebos are not just limited to drugs. It turns out that most building thermostats and city street crosswalk buttons don’t actually do anything other than give you a sense of control! And I’m just about the biggest jerk on the planet as the placebo effect goes away once you know about it, so you can blame me when you are sitting in your office freezing and knowing that you can’t do a thing about it! Sorry.

EDIT: 1-1-11
And the movement of your progress bar might also be another placebo!