In randomized clinical trials, the placebo group routinely improves. Sometimes substantially. People in placebo arms of depression studies, pain studies, and even some asthma studies report feeling better at rates that are striking — and that improvement isn’t fake. It’s the underlying behavior of the body and the mind doing what they often do when given time and attention. Understanding why patients improve without active treatment is one of the most useful frames in medicine, and it changes how to think about whether a given intervention is actually doing anything.
The placebo effect is real, but smaller than you’ve heard
The popular version of the placebo effect overstates it. Sugar pills don’t cure cancer or set broken bones. What they do reliably affect is subjective symptoms — pain, nausea, fatigue, mood, anxiety — and conditions that fluctuate naturally. The mechanism appears to involve real changes in the brain’s pain and reward pathways, with some studies showing measurable endorphin release and shifts in inflammatory markers. Open-label placebos, where patients are told they’re taking a sugar pill, also produce measurable effects, suggesting the ritual of treatment matters as much as the deception. None of this means placebos are a substitute for medicine. It means part of what we attribute to medicine is the placebo effect riding along with it.
Regression to the mean does most of the work
The bigger and less discussed factor is regression to the mean. People typically seek treatment when symptoms are worst — at the peak of a flare, after a particularly bad week, when something has gotten frightening. Whatever caused that peak usually wasn’t going to stay at peak intensity. Symptoms drift back toward their average naturally. If a treatment is started at the worst point, the subsequent improvement gets credited to the treatment when much of it was going to happen anyway. This isn’t a critique of treatment; it’s a critique of how cause and effect get inferred without controls. It’s why properly designed trials with placebo arms are essential.
Many conditions are self-limiting
A substantial share of common medical visits — viral upper respiratory infections, mechanical low back pain, mild gastroenteritis, tension headaches, certain skin rashes — resolve on their own within days to weeks, regardless of treatment. Antibiotics for viral infections, muscle relaxants for back pain, and many cough suppressants offer minimal benefit beyond placebo and time. The pressure to “do something” at a visit can lead to prescriptions that don’t help and occasionally cause harm. Watchful waiting is often the most evidence-based course, even if it doesn’t feel like care.
What this means for the patient
None of this argues against seeking medical care. Serious conditions do require diagnosis and active treatment, and undertreatment is a real problem in plenty of contexts. The takeaway is more nuanced: when a treatment “worked,” ask whether it was the treatment, the natural course, the placebo effect, or some combination. The answer changes what you do next time. For chronic or recurring conditions, that distinction matters a lot.
The bottom line
The body is more resilient than the medical system gives it credit for. A meaningful fraction of the recoveries we credit to treatment would have happened anyway. Knowing that doesn’t make medicine useless. It makes it more honest.
Leave a Reply