Health Atlas
Menu
Evidence-led reference
Health Atlas/Prevention & care

Prevention & care

Placebo and nocebo: context effects, real symptoms and careful comparisons

10 sources · Latest source check 2026-10-20 UTC
On this page

A placebo is a comparison intervention without the specific active treatment being tested. Expectations and interactions around care can influence outcomes. Calling something a placebo does not mean every change after it was caused by expectations.[1]

Separate the intervention, response and effect

A placebo response is the overall change after an inactive intervention. Symptoms may naturally fluctuate, people may enrol when symptoms are unusually bad, and other care can continue. A placebo effect is the part attributable to contextual mechanisms. An untreated or otherwise appropriate comparison helps separate these possibilities.[3]

Nocebo does not mean imaginary illness

Nocebo describes adverse effects linked to negative expectations or the treatment context. These mechanisms can operate with an inactive comparator or alongside an active treatment. The umbrella review describes psychological and biological processes, with varied evidence across settings. The label should not become a diagnosis that dismisses a new symptom.[2]

Discuss symptoms and expectations without withholding care

Tell the treating professional about symptoms and concerns, and discuss treatment changes with them. Placebo research does not replace assessment or establish that a particular person's symptoms are harmless. NCCIH's explanation is educational and does not substitute for clinical care.[1]

The expert consensus favours honest explanation, trust and empathy, and advises against taking unnecessary treatment risks to increase placebo effects. It treats communication as part of care, not permission to deceive or conceal necessary information. These are expert recommendations rather than proof that one script works for everyone.[10]

Evidence & uncertainty

Placebo and nocebo effects concern changes attributable to treatment context; a placebo-group response includes all observed change, including natural recovery and measurement effects. Different outcomes and conditions can respond differently, so one impressive percentage is not a universal rule.[2]

Selected-source overview · Independent review pending

Published questions

Is improvement in a placebo group the same as a placebo effect?

Selected-source overview · Independent review pending. Read the evidence context.

No. Overall change can include natural recovery, symptom fluctuation, other care and reporting effects. Comparing placebo with an appropriate untreated group helps investigate the additional contextual effect, although awareness of group assignment can itself affect that comparison.[3]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

If context affects a symptom, does that make it imaginary?

Selected-source overview · Independent review pending. Read the evidence context.

No. Placebo and nocebo research concerns psychological and biological processes. Context can influence experience without establishing the original cause of illness. A contextual contribution does not invalidate the symptom or remove the need for assessment.[2]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Can feeling better mean an underlying problem is unchanged?

Selected-source overview · Independent review pending. Read the evidence context.

Yes. In the asthma experiment, reported relief and measured airflow did not move identically. Both outcomes matter, but relief alone did not establish the same physiological benefit as the active inhaler. The finding is not a reason to substitute placebo for asthma care.[4]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Do nocebo findings prove medicines never cause the symptoms people report?

Selected-source overview · Independent review pending. Read the evidence context.

No. SAMSON excluded dangerous reactions and studied selected people with prior intolerance using one regimen. Average similarity to placebo cannot rule out pharmacological harms in other people or explain an individual's new symptom. Discuss concerns with the prescriber.[5]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

What should I do if symptoms start after a medicine?

Selected-source overview · Independent review pending. Read the evidence context.

Tell the treating professional what changed, when and how severely. Discuss any treatment changes with them. A possible expectation effect is one consideration; this guide cannot identify the cause or replace an appropriate clinical assessment.[1]

Evidence behind this answer
  • Placebo Effect

    agency explanation · 2020
    Full relevant public explanation read; updated January 2020 · Inactive comparators, anticipation and clinical-care boundaries

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Can a placebo have effects when people know it is a placebo?

Selected-source overview · Independent review pending. Read the evidence context.

Some open-label trials report benefit. The back-pain trial used openly described pills alongside a positive explanation and usual care. Its self-reported short-term results do not establish that an inert pill alone, an unrelated product or every condition will respond.[6]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Are open-label placebo benefits the same for symptoms and objective tests?

Selected-source overview · Independent review pending. Read the evidence context.

The 2025 review found larger effects for self-reports and no statistically clear pooled objective-outcome benefit. These categories contained different studies and outcomes. The finding cannot predict every test, but it cautions against translating symptom improvement into disease modification.[7]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Is open-label placebo a proven long-term replacement for treatment?

Selected-source overview · Independent review pending. Read the evidence context.

No such broad conclusion follows from the review. Many studies were brief and small, and results varied across conditions and measures. Discuss evidence for the specific condition and continuing care rather than replacing an established plan on the strength of the label.[7]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Can hearing about a side effect make it more likely to be noticed?

Selected-source overview · Independent review pending. Read the evidence context.

In the sham-capsule experiment, information influenced reported symptoms compared with no treatment. That does not make every reported adverse event expectation-driven or inevitable. The experiment was short and cannot determine what caused a particular person's symptoms.[8]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Can the same risk be explained in a less alarming way?

Selected-source overview · Independent review pending. Read the evidence context.

A risk can be described by how many people are affected and how many are not, without changing the numbers. The leaflet experiment found less symptom attribution with positive framing, but not clearly fewer or milder symptoms. Clear communication should preserve relevant information.[9]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Should clinicians hide side effects to avoid nocebo effects?

Selected-source overview · Independent review pending. Read the evidence context.

No. The expert consensus emphasizes balancing honest, transparent information with communication that avoids unnecessary harm. Tailoring explanations follows provision of necessary information; it is not a reason to remove meaningful choice or conceal important risks.[10]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Does failing to improve mean I did not believe strongly enough?

Selected-source overview · Independent review pending. Read the evidence context.

No reliable individual blame rule follows from these studies. Open-label effects differ across people, settings and outcomes; pooled benefits cannot promise a personal response. Reassess symptoms and care needs rather than judging someone's effort or character.[6]

Evidence behind this answer

This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.

Broad reviews show why the outcome matters

The 2010 Cochrane review compared placebo with no treatment across 234 trials, with usable outcomes in 202. It did not find important clinical effects generally, but found possible benefits for some patient-reported outcomes, especially pain. Effects varied, most trials had bias concerns and unblinded no-treatment comparisons can affect reporting or other care. Its search ended in 2008; it is a historical methods illustration, not a current verdict on every condition.[3]

Feeling better and improved airflow were different outcomes

In a small asthma crossover experiment, 46 people were randomized and 39 completed repeated active-inhaler, placebo-inhaler, sham-acupuncture and no-intervention visits. Active albuterol improved measured airflow more than the other conditions. Reported relief was similar with the active inhaler and both placebos, and lower with no intervention. This short experiment shows why symptom experience and physiological change should both be considered; it does not support replacing asthma treatment.[4]

A tablet-related pattern need not identify its ingredient

SAMSON enrolled 60 people who had stopped statins because of symptoms; 49 completed twelve months of randomized statin, placebo and no-tablet periods. Average symptoms were similar with statin and placebo and higher than during no-tablet months. Starting and stopping symptom patterns did not distinguish the two tablets. Dangerous previous reactions were excluded. These findings concern selected participants and one statin regimen, not proof that medicines cannot cause harm.[5]

Some placebo studies tell participants exactly what they receive

A three-week back-pain trial randomized 97 adults; 83 completed it. Adding openly described placebo pills to usual care improved self-reported pain and disability more than usual care alone. Participants also received a positive explanation. The study was small, unblinded and short; it could not isolate the pill from its explanation or establish durable benefit. Its authors called for further research before clinical recommendations.[6]

An updated synthesis finds larger self-reported than objective effects

A 2025 review included 60 open-label placebo trials with varied populations and controls. It found a small average benefit, larger for self-reported outcomes; the pooled objective-outcome estimate was not statistically clear. Most trials were small and many were brief. The authors adapted bias rules for the unblinded interventions and did not perform GRADE. The result supports further condition-specific evaluation rather than a universal prescription.[7]

Information wording can affect some short-term responses

In a 112-person sham-capsule experiment, positively framed side-effect information reduced reported side effects during the session compared with negative framing, but the difference was absent at 24 hours. Both informed groups reported more of the listed symptoms than a no-treatment group. This does not establish durable protection from nocebo effects in routine care.[8]

A separate 203-person placebo-tablet experiment found fewer healthy volunteers attributed symptoms to the tablet after a positively framed leaflet. The total number and severity of attributed symptoms did not clearly differ. Framing can affect interpretation without improving every symptom outcome; this one-hour experiment needs careful application to patients.[9]

There is no single placebo strength or responder label

The umbrella review combines diverse outcomes, reviews and experiments. Overlapping evidence, publication bias and methodological differences remain relevant. It cannot establish one effect size, identify every person's future response or turn all symptoms reported during placebo treatment into proven nocebo effects.[2]

References

  1. Placebo Effect

    Agency explanation

    National Center for Complementary and Integrative Health · Content/publication date: 2020
    Accessed: 2026-10-20 UTC · Full relevant public explanation read; updated January 2020
    Cited sections: Inactive comparators, anticipation and clinical-care boundaries

    Agency overview, not a universal placebo treatment recommendation.

  2. Placebo and nocebo effects and mechanisms associated with pharmacological interventions: an umbrella review

    Umbrella review

    Frisaldi, Shaibani and Benedetti · Content/publication date: 2023
    Accessed: 2026-10-20 UTC · Original public abstract, introduction and methodological limits read
    Cited sections: Response versus effect, context mechanisms and heterogeneous evidence

    Overlapping reviews and varied outcomes cannot supply one universal effect size.

  3. Placebo interventions for all clinical conditions

    Systematic review

    Hróbjartsson and Gøtzsche; Cochrane · Content/publication date: 2010
    Accessed: 2026-10-20 UTC · Original complete abstract and relevant review passages read; search ended March 2008
    Cited sections: Placebo versus no treatment, outcome differences and response bias

    Historical broad review; not a current exhaustive condition-by-condition verdict.

  4. Active albuterol or placebo, sham acupuncture, or no intervention in asthma

    Randomized crossover trial

    Wechsler and colleagues · Content/publication date: 2011
    Accessed: 2026-10-20 UTC · Complete original abstract and displayed outcome figures read; full report not independently appraised
    Cited sections: Subjective improvement versus measured airflow in a small asthma experiment

    Short-term experiment; not an asthma treatment plan.

  5. Side Effect Patterns in a Crossover Trial of Statin, Placebo, and No Treatment

    Randomized crossover trial

    Howard and colleagues · Content/publication date: 2021
    Accessed: 2026-10-20 UTC · Original full publisher text retrieved; abstract, methods and relevant results read
    Cited sections: Statin, placebo and empty-bottle months; selection and timing limits

    Selected prior intolerance; dangerous reactions were excluded; not a universal causality test.

  6. Open-label placebo treatment in chronic low back pain: a randomized controlled trial

    Randomized trial

    Carvalho and colleagues · Content/publication date: 2016
    Accessed: 2026-10-20 UTC · Original full text retrieved; abstract, design and limitations read; linked erratum read
    Cited sections: Transparent placebo added to usual care for three weeks; self-report limits

    A 2017 erratum corrected a bothersomeness effect size; that metric is not reproduced.

  7. Effects of open-label placebos across populations and outcomes: an updated systematic review and meta-analysis of randomized controlled trials

    Systematic review meta analysis

    Fendel and colleagues · Content/publication date: 2025
    Accessed: 2026-10-20 UTC · Original full text retrieved; abstract, methods, discussion and limitations read; search through November 2023
    Cited sections: Sixty trials, self-report versus objective outcomes and methodological adaptations

    Broad pooled outcomes, small trials, adapted bias rules and no formal GRADE assessment.

  8. The Influence of Side Effect Information Framing on Nocebo Effects

    Randomized experiment

    Faasse and colleagues · Content/publication date: 2019
    Accessed: 2026-10-20 UTC · Complete original PubMed abstract read; full report not read
    Cited sections: Sham capsules, positive framing and loss of difference by follow-up

    Laboratory-style study; immediate framing benefit did not persist at 24 hours.

  9. Positively Framed Risk Information in Patient Information Leaflets Reduces Side Effect Reporting: A Double-Blind Randomized Controlled Trial

    Randomized experiment

    Webster, Weinman and Rubin · Content/publication date: 2018
    Accessed: 2026-10-20 UTC · Complete original abstract and metadata read on authors' institutional repository
    Cited sections: Healthy volunteers, symptom attribution versus number and severity

    One-hour placebo study; attribution changed, not all symptom outcomes.

  10. Implications of Placebo and Nocebo Effects for Clinical Practice: Expert Consensus

    Expert consensus

    Evers and colleagues · Content/publication date: 2018
    Accessed: 2026-10-20 UTC · Original public abstract, methods, recommendations and discussion read
    Cited sections: Honest communication, consent, non-deception and no unnecessary risks

    Survey and meeting of invited experts; not an implementation-effectiveness trial.

Scope & review

Population
Readers interpreting placebo-controlled research and people discussing symptoms or treatment expectations with clinicians
Exposure or intervention
Inactive comparators, contextual effects, open-label placebo and side-effect communication
Comparison
No intervention, usual care, active treatment and alternative information wording
Outcomes
Symptoms, objective measures, attribution, understanding and ethical boundaries
Timeframe
Short experiments, selected three-week and twelve-month trials, and explicitly dated syntheses
Jurisdiction
International educational explanation; selected trial populations and expert-consensus scope retained

Targeted agency explanation, original experiments, updated syntheses and expert-consensus recommendations read with access, overlap and methodological limits recorded. Qualified clinical and methods review not completed.

Sources checked on their individual access dates. Selected sources were checked; ongoing surveillance and qualified review have separate statuses on the coverage page.

Revision history
  • 2026-10-20 UTC: Initial educational authorship; foundations precede source examples; adjacent bodies and publication identities checked; qualified subject-matter review pending.

What this article still leaves open

  • Which context-based interventions produce durable benefits for specific conditions without impairing informed choice?
  • How can trials distinguish experienced symptom relief, reporting changes and objective physiological outcomes more clearly?
Suggest a correction

Submit a content correction to the editorial queue, or download a local note. Leave out personal medical details. Submitted text is stored privately with a browser receipt session and is visible to this browser and the editor. Clearing cookies removes access to your submission history. This is not an account or a medical service.

Explore the topic map

Browse the health domains connected to this article. The map also identifies topics whose articles remain unwritten.