The placebo effect gets used as a dismissal — it is just placebo — and as a justification, on the grounds that if it works it works. Both misunderstand what it actually is.
What it is not
It is not imagination, and it is not the same as feeling better because you believe you should.
It is a measurable set of physiological and psychological responses to the context of treatment, which can be studied and partially blocked.
The mechanism evidence
The strongest demonstration comes from pain research.
Placebo analgesia can be blocked by a drug that blocks opioid receptors, which establishes that endogenous opioid release is involved.
That is a specific biochemical mechanism, demonstrated experimentally, and it is why the effect cannot be dismissed as merely psychological.
Dopamine release has been demonstrated in placebo response in Parkinson's disease, which is a second identified pathway.
What it affects and what it does not
The critical distinction.
Placebo responses are substantial for subjective outcomes — pain, nausea, fatigue, mood, and reported symptom severity.
They are minimal or absent for objective disease processes — tumour size, infection clearance, bone healing, laboratory markers of organ function.
Which means placebo can genuinely make someone feel better without affecting the underlying condition at all.
That gap is where the danger sits, since feeling better while a treatable condition progresses is a bad outcome.
The components
Research has separated several contributors.
Expectation, generated by information about what the treatment will do.
Conditioning, where previous experience of effective treatment produces a learned response to the associated cues.
The therapeutic encounter itself — attention, explanation, empathy — which has been shown to produce measurable effects independent of any treatment given.
Natural history and regression to the mean, which are not placebo effects at all but are frequently attributed to them.
Open-label placebo
The most surprising finding in the field.
Studies giving people placebos while explicitly telling them they are placebos have found symptom improvement in several conditions.
Which challenges the assumption that deception is necessary, and the findings have been replicated enough to be taken seriously.
The proposed explanation involves conditioning and the ritual of treatment rather than belief in efficacy, and it is not fully understood.
Nocebo
The inverse and considerably less discussed.
Negative expectations produce negative effects, including measurable symptoms.
Studies have found that informing patients about potential side effects increases their occurrence, which creates a genuine ethical tension with informed consent.
Media coverage of side effects has been shown to increase reported side effects in populations, which is a documented and awkward finding.
What this means for evaluating anything
If a treatment addresses a subjective symptom, and the evidence for it comes from uncontrolled observation, placebo response is a sufficient explanation for what is observed.
Which is why controlled trials exist, and why any therapy claiming effects on subjective outcomes without them has not established anything.
It also means that a practitioner who is warm, attentive and provides a coherent explanation will produce better outcomes than one who does not, regardless of what they are doing.
That is a real effect, worth having, and it should not be confused with the specific treatment working.
The line
Using this to feel better about ordinary discomfort is fine.
Using it in place of treatment for a condition that has one is not, and the distinction is whether there is an underlying process that placebo does not touch.
Which is a question for a doctor rather than for a practitioner selling the treatment.
The size varies enormously
Worth knowing since placebo response is frequently described as a fixed quantity.
Response rates in trials differ by condition, by outcome measure, by the form of the placebo and by the setting.
Injections produce larger responses than tablets. More tablets produce larger responses than fewer. Branded packaging produces larger responses than plain.
Procedures produce the largest responses of all, which is why sham surgery trials have been so revealing when they have been conducted.
Several established surgical procedures have been found no better than sham in controlled trials, which is among the more uncomfortable findings in modern medicine.
Placebo response is rising
An observed trend in trial data, particularly for pain and psychiatric conditions, and particularly in some regions.
Which makes demonstrating drug efficacy harder over time, since the comparison arm is improving.
Proposed explanations include changes in trial design, participant expectations, recruitment methods and the intensity of contact participants receive.
None is established, and the trend is real enough to affect drug development decisions.
The ethics question
Deliberately prescribing placebos raises an obvious problem with informed consent.
Surveys of clinicians have found that impure placebos — treatments given without expectation of specific effect — are used more often than most people assume.
Which sits uncomfortably with consent requirements, and the open-label findings offer a possible route around it that is not yet established enough to be standard practice.