Psychology › Therapy & treatment › Psychiatric medicine: drugs and brain treatments
The placebo effect
1955 (Beecher)· Finding or experiment
Mixed evidenceSome support; effects smaller or less consistent than once thought.
Improvement that comes from receiving a treatment and expecting it to help, rather than from the treatment's active ingredient.
See it on the Psychology map →What it says
In 1955 Harvard anesthesiologist Henry Beecher's paper 'The Powerful Placebo' claimed about 35 percent of patients improve on dummy treatments. Later analysis showed much of that was natural recovery and regression to the mean, the tendency of extreme symptoms to drift back toward average. Genuine expectation effects still exist, mainly for pain, nausea, fatigue and mood. Placebo pain relief can involve the body's own opioids. In depression trials, placebo groups often improve nearly as much as drug groups.
Why it matters
It is why new treatments must beat a fake in blinded trials, and a reminder that expectation and care are part of how any treatment works.
Think of it like
Like tenants rating a building higher after a lobby renovation, even though the plumbing and HVAC are unchanged.
The evidence
Status: mixed. A 2001 meta-analysis by Asbjørn Hróbjartsson and Peter Gøtzsche found little effect on objective outcomes but modest effects on self-reported pain. Opioid-blocker experiments since 1978 confirm a real pain mechanism.
Open questions
Why placebo responses in US drug trials have grown over the decades, and whether 'open-label' placebos, honestly labeled as dummies, truly help.
Connections (2)
▸The placebo effect sets the bar for Psychiatric medicine: drugs and brain treatmentsWell replicated
The connection
Because people with depression, anxiety and pain often improve on dummy pills, a psychiatric drug must beat placebo in blinded randomized trials to prove it works. High placebo responses make that bar hard to clear, and many promising drugs have failed it.
Who showed it and when
US law since the 1962 Kefauver-Harris Amendments requires proof of effectiveness, usually from placebo-controlled trials; Henry Beecher at Harvard had championed such controls.
The evidence
Analyses of FDA trial files by Arif Khan, Irving Kirsch and others show roughly 30 to 40 percent of placebo patients in antidepressant trials count as responders.
Open questions
Why placebo responses in psychiatric trials have risen over decades; theories include higher patient expectations, more trial sites and looser enrollment.
▸The placebo effect accounts for much of the benefit of SSRI antidepressants (e.g., Prozac)Mixed evidence
The connection
In antidepressant trials, people on placebo improve a lot: by Irving Kirsch's estimate, placebo groups achieve about 80 percent of the drug groups' improvement. So much of the benefit patients notice on SSRIs reflects natural recovery, expectation and care, with the drug adding a smaller extra effect.
Who showed it and when
Irving Kirsch and Guy Sapirstein raised the issue in 1998; Kirsch's 2008 analysis of FDA data on four newer antidepressants drew wide attention.
How solid is it?
FDA files, including unpublished trials, show an average drug-placebo gap of about 2 points on a 52-point depression scale; Cipriani's 2018 meta-analysis confirmed small but real effects.
Open questions
Averages can hide people who benefit greatly; researchers also dispute whether a 2-point gap matters clinically and whether side effects let patients guess they got the drug.