The placebo question, taken seriously

The charge that a traditional treatment works only through placebo is common and frequently made loosely. It is worth being precise about what is being claimed.

What placebo effects actually are

Placebo effects are genuine changes in symptoms resulting from the context of treatment rather than from any specific active component.

They are not imaginary. Neuroimaging and pharmacological studies have documented real physiological mechanisms — endogenous opioid release in placebo analgesia, which can be blocked by opioid antagonists, and dopaminergic changes in placebo response in Parkinson's disease.

They are also not universal. Placebo effects are substantial for subjective outcomes — pain, nausea, fatigue, mood, itching — and much smaller or absent for objective ones. A placebo does not shrink a tumour, lower blood glucose in the way medication does, or clear an infection.

This distinction is the crux of it. A condition where the outcome is how someone feels is one where context effects operate powerfully. A condition where the outcome is measurable pathology is not.

What contributes to the effect

Several components, and disentangling them is difficult.

Expectation. Believing a treatment will help changes reported symptoms measurably.

Conditioning. Prior experience of treatment producing relief conditions a response.

The therapeutic encounter. Time, attention, being listened to, a coherent explanation and a plan all affect symptom reporting. Studies varying the quality of the practitioner interaction while holding the intervention constant find measurable differences in outcome.

Natural history and regression to the mean, which are not placebo effects at all but are frequently attributed to them. Illnesses fluctuate and improve, and improvement after any intervention gets credited to it.

This last point matters. A substantial part of what is loosely called placebo in uncontrolled settings is simply the illness getting better.

The uncomfortable question for traditional medicine

Traditional systems generally deliver the components associated with strong context effects: long consultations, individualised attention, a coherent explanatory framework, elaborate preparation, and a practitioner who takes the patient's account seriously.

Which means one should expect substantial improvement in subjective outcomes from these encounters, independent of any specific effect of the preparations used.

That is not an accusation of fraud. It is a prediction about what a well-designed trial should find, and it is why blinded comparison against placebo is the relevant test.

Where such trials have been done, results have generally been modest, which is consistent with this prediction.

The uncomfortable question for conventional medicine

Worth stating for balance.

Placebo effects operate in conventional practice too, and conventional medicine has systematically stripped out the components that produce them — consultation times have shortened, continuity of care has weakened, and explanation is frequently rushed.

There is a reasonable argument that conventional practice under-uses context effects that are free, safe and demonstrably real, and that some of the appeal of alternative practitioners is that they supply what has been removed.

Research on the therapeutic encounter supports this. Studies of clinician communication style and consultation quality find measurable effects on outcomes.

What follows practically

Placebo effects are worth having. Reducing someone's pain or nausea through context is a real benefit to them. If a practice is safe and cheap and makes people feel better, that is not nothing.

They are not sufficient for serious disease. Feeling better is not the same as being better, and for conditions where pathology progresses regardless of symptoms, relying on context effects is dangerous.

This is the central practical point. Someone with cancer, diabetes or heart failure may genuinely feel better on a treatment with no specific effect, and the disease continues.

They do not require deception. Studies of open-label placebo — where participants are told they are receiving an inert treatment — have found symptom improvement in some conditions, which is a genuinely surprising finding with implications still being worked out.

The size varies enormously by condition, and is largest in conditions with fluctuating subjective symptoms.

How to hold this

If a practice is safe, affordable, and makes you feel better, the mechanism matters less than whether you are also getting appropriate treatment for anything that requires it.

If a practice is expensive, carries risk, or is replacing effective treatment, the mechanism matters a great deal, and the question of whether it does anything specific becomes decisive.

The dismissive use of "just placebo" is unhelpful because it ignores that the effect is real and valuable. The defensive rejection of the question is unhelpful because it avoids the only test that would settle whether a preparation does anything.

Both moves prevent the useful conversation, which is: does this help, how much, at what cost, and is it substituting for something that would help more?

Feeling better is not evidence that a disease is improving. Continue prescribed treatment and monitoring for any diagnosed condition.