The placebo effect: more than a control condition

For most of the twentieth century, the placebo was treated as something to subtract from a clinical trial — the noise to be removed to reveal the signal. The patient who improved on a sugar pill was a methodological inconvenience. The drug worked when its effect exceeded the placebo's.

Beginning in the 1990s, researchers led by Ted Kaptchuk at Harvard and Fabrizio Benedetti in Turin began studying the placebo as a phenomenon in its own right. The findings have rearranged how the field thinks about the mind-body relationship.

1. The placebo effect is biological

When a patient takes a placebo for pain and reports relief, fMRI studies show changes in the same brain regions that respond to actual analgesics — particularly the descending pain-modulation system. The placebo isn't producing imagined relief; it's producing actual neurochemical changes (Wager et al., 2004; Benedetti, 2014).

The clearest evidence: placebo-induced pain relief can be blocked by administering naloxone, an opioid antagonist. This means the placebo had triggered endogenous opioid release. The pain reduction was as real as any pharmacological reduction; the trigger was different.

2. The conditioning component

Placebos work better when patients have been previously conditioned. Studies in which patients receive an active analgesic for several days, then are switched to a placebo without their knowledge, show larger placebo responses than placebos given to opioid-naive patients. The body has learned to release endogenous opioids in response to the ritual of taking a pill.

This explains why placebos are stronger in conditions with long treatment histories (chronic pain, depression) and weaker in acute conditions where the patient has no learned response.

3. Open-label placebos

The most striking modern finding: placebos work even when patients know they're taking a placebo. Kaptchuk's 2010 trial gave IBS patients placebos with full disclosure that the pills were inert sugar. The patients improved more than the no-treatment control group, on objective and subjective measures (Kaptchuk et al., 2010).

This is counterintuitive but well-replicated. Subsequent open-label placebo trials in chronic pain, depression, allergic rhinitis, and ADHD have found similar effects. The mechanism isn't deception — it's the ritual of treatment combined with the patient's expectation of improvement, even when they consciously know the pill is inert.

4. The clinical implication

For pain management, depression, fatigue, and several other subjectively-experienced conditions, the placebo response accounts for a substantial fraction of standard treatment response. A common figure: the placebo response in depression trials is roughly 30-40%, and the additional benefit of active medication over placebo is often only 5-15% (Kirsch, 2014).

This doesn't make antidepressants useless. The placebo + drug combined effect can be substantial. But it does mean that treatments that seem to work in open trials may be working primarily through placebo mechanisms — a fact that affects how to evaluate alternative therapies.

5. The honest summary

Placebo effects are real biological events, not imaginary. They are larger for subjective conditions, smaller for objective ones (a placebo doesn't shrink a tumor). They are partly conditionable, work even with disclosure, and depend on the ritual of care — being attended to by a clinician — as much as on the pill itself.

The practical implication for medicine: optimizing the placebo component of treatment — clinician warmth, ritual, expectation-setting — is not just ethical; it produces additional therapeutic effect on top of any pharmacological action. The placebo isn't the noise. Sometimes it's a substantial part of the signal.