The Placebo Effect: How Expectations Shape Symptoms and Treatment

Expectations, learning and the setting in which care happens can influence some symptoms and treatment experiences. Understanding placebo and nocebo effects means taking those experiences seriously while distinguishing symptom relief from changes in disease.

A person takes a tablet and their pain eases. The improvement matters. Understanding why it happened requires a different question: what would have happened without that intervention?

The tablet might contain an effective medicine. The pain might also have been settling anyway. Expectations, previous treatment experiences and the clinical encounter may contribute. Several influences can operate together, making a convincing story about recovery less informative than a carefully designed comparison.

Placebo research examines some of these contextual influences. Nocebo research examines how negative expectations and related processes can contribute to worse experiences. Both concern interactions between psychological and physiological processes, rather than a division between real symptoms and imaginary ones (Colloca & Barsky, 2020).

A placebo intervention, response and effect are different things

Researchers do not always use these terms consistently. The following distinctions help separate what is observed from what causes it (Hafliðadóttir et al., 2021).

Term	What it describes
Placebo intervention	An intervention lacking the particular active component being investigated. Its status depends on the research question; it need not be inactive in every possible respect.
Placebo response	The observed change following that intervention, including changes arising for reasons unrelated to placebo mechanisms.
Placebo effect	The component attributable to placebo-related mechanisms, such as learned associations and expectations, rather than unrelated influences.

A placebo response therefore does not automatically demonstrate a placebo effect. Nor does a response to placebo establish that someone's symptoms had a psychological cause.

Why improvement alone cannot identify the mechanism

Many conditions improve naturally or fluctuate. People often seek help when symptoms are unusually severe. A subsequent measurement may be closer to their usual level even if treatment has done nothing: a statistical pattern called regression to the mean. This is different from biological recovery.

Measurement error, changes in other treatment and ordinary variation also contribute. Participants may interpret a questionnaire differently after entering a study, monitor sensations more closely or report what they think a researcher expects. A lower score can reflect changed experience, changed reporting, or both (Finniss et al., 2010).

Consider an illustrative example: someone enters a headache study during an exceptionally difficult week and reports fewer headaches a fortnight later after receiving a placebo. Without an appropriate comparison, we cannot determine how much reflects the intervention, the usual course of their headaches or other changes in their life.

Randomising participants to placebo and a suitable no-intervention comparison helps estimate differences beyond changes occurring in both groups. Even then, the groups may differ in attention, treatment rituals, expectations and awareness of allocation. The comparison tests a package of influences; isolating a particular mechanism requires further experimental work.

How expectations can shape symptom experience

Expectations are predictions about what is likely to happen. They can influence attention, anticipation and the interpretation of bodily sensations. Someone expecting discomfort may monitor their body differently from someone expecting relief. Treatment credibility and previous encounters can shape those predictions (Wager & Atlas, 2015).

Anticipatory anxiety and symptom monitoring are among the processes examined in nocebo research. A systematic review identified negative suggestions and expectations as relevant contributors, while finding that evidence for many proposed individual predictors remained inconsistent (Webster et al., 2016). These processes are not simply deliberate decisions to feel better or worse.

Pain is particularly well studied. Its intensity and unpleasantness depend on processing that extends beyond incoming signals about potential tissue damage. A change in the experience can be meaningful even when the original injury remains. Conversely, feeling less discomfort does not by itself demonstrate that an underlying disease has improved.

Nausea provides another example of an outcome that can be influenced in some placebo comparisons. The size and reliability of effects differ across symptoms and study settings. Evidence about pain cannot simply be transferred to every sensation, diagnosis or biological system.

Treatment experiences can become learned associations

Classical conditioning describes learning through association. When a cue repeatedly accompanies an effective intervention, it may later help elicit a related response. In treatment, cues can include a particular-looking tablet or aspects of the setting in which relief has previously occurred.

Laboratory studies sometimes pair a purported pain-relieving cream with secretly reduced stimulus intensity during a learning phase. Researchers then compare responses when the physical stimulation is matched. This distinguishes learning about relief from simply experiencing a weaker stimulus. Colloca and colleagues found that conditioning strengthened behavioural and neurophysiological placebo analgesic responses relative to verbal suggestion alone (2008).

Expectation and conditioning can overlap: an experienced association may change what someone predicts next. They are not necessarily competing explanations. In experiments involving pain, motor performance and hormone secretion, Benedetti and colleagues found that verbal suggestions and conditioning had different influences across the responses being measured (2003). Learning also features in how avoidance can maintain anxiety, although that literature addresses a different set of questions.

Those findings do not establish that people can consciously regulate any hormone or cure illness through learned cues. Experimental procedures are specific, often brief, and may combine several forms of learning. How reliably their effects transfer to routine care remains a separate question.

What the neuroscience of placebo analgesia shows

Some placebo analgesia involves the body's endogenous opioid systems. In a classic postoperative dental-pain experiment, naloxone, which blocks opioid receptors, increased pain in participants who had responded to placebo. The results supported an opioid contribution in that setting (Levine et al., 1978).

Later positron emission tomography research linked placebo administration during sustained experimental pain with changes in endogenous activity at μ-opioid receptors and lower pain ratings (Zubieta et al., 2005). A spinal imaging experiment also found reduced pain-related spinal-cord activity under placebo (Eippert et al., 2009). Such findings provide evidence that contextual manipulations can have physiological consequences.

They do not identify a universal placebo pathway. A participant-level meta-analysis of 20 imaging studies, involving 603 healthy participants, found small changes across several pain-related brain areas and considerable variation in some patterns between studies (Zunhammer et al., 2021).

A larger individual study of 392 participants found reduced reported pain and changes in brain measures associated with higher-level contributions to pain, but no reduction in selected measures linked more closely to nociceptive processing (Botvinik-Nezer et al., 2024). Nociception concerns the processing of potentially damaging stimulation; it is related to, but distinguishable from, pain experience.

A July 2026 meta-analysis compared suggestions alone with suggestions combined with conditioning across 16 studies and 409 participants. It found shared and partly distinct neural patterns (Spisak et al., 2026). Because these induction methods also differed across studies, such comparisons do not perfectly isolate conditioning from every other procedural difference.

Neurochemical explanations also need testing. A preregistered experiment manipulating dopamine signalling did not find the predicted changes in treatment expectations or subsequent placebo analgesia (Kunkel et al., 2024). Brain scans and neurotransmitter findings answer specific questions under specific conditions. A striking image cannot establish one mechanism for every patient or symptom.

Clinical context also accompanies active treatment

Placebo-related mechanisms can operate while someone receives an effective medicine. They are not confined to sugar pills. In a heat-pain experiment with healthy volunteers receiving a fixed concentration of remifentanil, positive and negative treatment expectations changed the observed analgesic benefit (Bingel et al., 2011).

This demonstrates an interaction between treatment information and analgesic experience in that experiment. It does not show that expectations can reliably cancel any drug's biological action, or that a clinician should promise an outcome to obtain a better response.

An irritable bowel syndrome trial compared a waiting list, sham acupuncture with limited interaction, and sham acupuncture accompanied by a supportive practitioner interaction. Reported outcomes improved most in the augmented-interaction group (Kaptchuk et al., 2008). The study investigated the clinical encounter without establishing which individual ingredient caused each improvement.

This connects with the therapeutic relationship: understandable explanations and a workable partnership can matter alongside the intervention itself. Recognising that context has effects does not establish that different treatments are equally effective.

Symptom relief and disease outcomes can diverge

A small asthma crossover study offers an instructive example. Albuterol improved measured lung function more than a placebo inhaler, sham acupuncture or no intervention. Yet participants reported similar subjective improvement after albuterol and the two placebo procedures (Wechsler et al., 2011).

The placebo procedures did not improve the lung-function measure beyond the no-intervention comparison. Feeling better and breathing out more air were different outcomes. This short-term pilot cannot settle every question about asthma, but it shows why symptom reports cannot stand in for all physiological measurements.

A broad Cochrane review found no evidence of important clinical effects of placebo interventions in general, while identifying effects on some patient-reported outcomes, especially pain and nausea. Effects varied, and reporting bias was difficult to separate from changed experience (Hróbjartsson & Gøtzsche, 2010). Its search ended in 2008, so it is a methodological reference point rather than a complete account of today's literature.

Subjective outcomes remain valuable: pain and nausea are experiences people seek help for. Their importance does not make them interchangeable with infection clearance, tumour growth, lung function or survival. Each claim requires evidence for the particular outcome being claimed.

What placebo-controlled trials can establish

In a well-conducted randomised trial, comparing an active intervention with a credible placebo estimates the effect of assignment to those alternatives under the trial's conditions. Matching the treatment experience and maintaining blinding help reduce differences in expectations and assessment.

A two-group trial does not independently measure the placebo effect. Both groups may improve through natural recovery, background care and study participation. Nor can researchers always assume that a drug effect and a contextual effect simply add together in fixed amounts (Finniss et al., 2010).

Side effects can reveal allocation; sham procedures may have effects of their own; withdrawals can complicate comparisons. A result showing no clear advantage over placebo also needs interpretation in light of uncertainty, sample size, adherence and the chosen outcome. It does not automatically establish equivalence or explain why either group changed.

Can a placebo help when people know it is a placebo?

Open-label placebos are explicitly described as placebos. In an early trial, 80 adults with irritable bowel syndrome were randomised to open-label placebo or no treatment with matched provider interaction. After three weeks, the placebo group reported greater improvement on several symptom measures (Kaptchuk et al., 2010).

This challenges the assumption that concealment is always necessary. It does not demonstrate that information and expectations no longer matter: participants received an explanation about placebo effects, and knew which intervention they had been assigned.

A 2025 meta-analysis included 60 randomised trials across clinical and non-clinical samples. The pooled effect was small overall: a standardised mean difference of 0.35. Effects were clearer for self-reported outcomes; the pooled estimate for objective outcomes was 0.09, with a 95% confidence interval from −0.12 to 0.29, providing no clear evidence of benefit on that combined category (Fendel et al., 2025).

That overall average does not establish a clinically important benefit for every condition. Trials were often small, outcomes differed and participants could not be blinded to receiving an openly described placebo. The review searched through November 2023. Longer follow-up, independent replication and stronger attention controls remain important for judging practical value.

Nocebo effects: when expectations contribute to worse symptoms

A nocebo effect is an adverse change attributable to negative expectations or related contextual mechanisms. Warnings, previous difficult experiences and learning may contribute. The possibility does not imply that a symptom is invented or voluntarily produced (Colloca & Barsky, 2020).

As with placebo, a response and an effect need separating. In a meta-analysis of COVID-19 vaccine trials, around 35% of placebo recipients reported systemic adverse events after the first injection (Haas et al., 2022). Without an untreated comparison, that proportion cannot identify how much arose from expectations rather than background headaches, fatigue, ordinary illness or other influences. An adverse-event rate is not a direct measure of a psychological mechanism.

The SAMSON trial used statin, placebo and no-tablet months in 60 participants who had previously stopped statins because of symptoms. Average symptom scores were similar in statin and placebo months and lower in no-tablet months (Howard et al., 2021). This supports a contribution from taking tablets and their associated context in that selected sample.

Participants knew when they were taking no tablets, and individual experiences varied. The findings cannot establish that every statin-related symptom, or every adverse drug reaction, is a nocebo effect. Suspected side effects still deserve assessment; an average trial result cannot identify the cause of an individual's symptoms.

Communication should support understanding and informed consent

Clinicians face a real responsibility: explain possible harms accurately without presenting them as inevitable. A review of framing studies found promising evidence that the presentation of side-effect information can influence experiences, alongside gaps concerning durability, mechanisms and risk understanding (Barnes et al., 2019).

For an illustrative risk of 10 in 100, explaining both that 10 people may experience an effect and that 90 may not conveys the same frequency from two perspectives. This is an example of balanced framing, not a risk estimate for a particular treatment. Relevant severity, uncertainty and what to do if symptoms occur still need explanation.

Withholding material information to avoid a nocebo effect undermines informed consent. So does describing uncertain benefits as guaranteed. Asking about previous experiences, checking understanding and inviting reports of symptoms can make communication more useful without dismissing concerns.

Ethical use depends on honesty and appropriate care

Deceptively substituting a placebo for indicated treatment can damage trust and delay effective care. Open-label use avoids one form of deception, but still requires accurate explanations, meaningful choice and evidence appropriate to the intended use. Research findings do not make every commercial placebo product a supported treatment.

Research ethics also constrain the use of placebo controls. The 2024 Declaration of Helsinki permits them where no proven intervention exists, or where compelling methodological reasons apply and withholding the best proven intervention does not add risks of serious or irreversible harm (World Medical Association, 2024). Scientific usefulness alone is insufficient.

Uncertainty remains about which effects persist, who benefits and how laboratory findings translate to care. Differences in induction procedures, comparators and outcome measures complicate synthesis. Small studies, selective publication and participants' awareness of what researchers hope to observe can distort estimates. Neither an enthusiastic success story nor a single null result resolves the whole field.

The practical implication is to treat expectations and learning as part of the conditions in which healthcare occurs. They can influence some treatment experiences, and deserve careful study. They do not establish that disease is imaginary, that belief alone cures it, or that someone who remains unwell has failed to think positively.

About this article: This educational overview does not provide an individual assessment or treatment plan. Examples are illustrative. Do not change prescribed treatment on the basis of placebo or nocebo research; discuss symptoms and treatment concerns with an appropriate healthcare professional.

References

Barnes, K., Faasse, K., Geers, A. L., Helfer, S. G., Sharpe, L., Colloca, L., & Colagiuri, B. (2019). Can positive framing reduce nocebo side effects? Current evidence and recommendation for future research. Frontiers in Pharmacology, 10, Article 167. https://doi.org/10.3389/fphar.2019.00167

Benedetti, F., Pollo, A., Lopiano, L., Lanotte, M., Vighetti, S., & Rainero, I. (2003). Conscious expectation and unconscious conditioning in analgesic, motor, and hormonal placebo/nocebo responses. Journal of Neuroscience, 23(10), 4315–4323. https://doi.org/10.1523/JNEUROSCI.23-10-04315.2003

Bingel, U., Wanigasekera, V., Wiech, K., Ni Mhuircheartaigh, R., Lee, M. C., Ploner, M., & Tracey, I. (2011). The effect of treatment expectation on drug efficacy: Imaging the analgesic benefit of the opioid remifentanil. Science Translational Medicine, 3(70), Article 70ra14. https://doi.org/10.1126/scitranslmed.3001244

Botvinik-Nezer, R., Petre, B., Ceko, M., Lindquist, M. A., Friedman, N. P., & Wager, T. D. (2024). Placebo treatment affects brain systems related to affective and cognitive processes, but not nociceptive pain. Nature Communications, 15, Article 6017. https://doi.org/10.1038/s41467-024-50103-8

Colloca, L., & Barsky, A. J. (2020). Placebo and nocebo effects. New England Journal of Medicine, 382(6), 554–561. https://doi.org/10.1056/NEJMra1907805

Colloca, L., Tinazzi, M., Recchia, S., Le Pera, D., Fiaschi, A., Benedetti, F., & Valeriani, M. (2008). Learning potentiates neurophysiological and behavioral placebo analgesic responses. Pain, 139(2), 306–314. https://doi.org/10.1016/j.pain.2008.04.021

Eippert, F., Finsterbusch, J., Bingel, U., & Büchel, C. (2009). Direct evidence for spinal cord involvement in placebo analgesia. Science, 326(5951), 404. https://doi.org/10.1126/science.1180142

Fendel, J. C., Tiersch, C., Sölder, P., Gaab, J., & Schmidt, S. (2025). Effects of open-label placebos across populations and outcomes: An updated systematic review and meta-analysis of randomized controlled trials. Scientific Reports, 15, Article 29940. https://doi.org/10.1038/s41598-025-14895-z

Finniss, D. G., Kaptchuk, T. J., Miller, F., & Benedetti, F. (2010). Biological, clinical, and ethical advances of placebo effects. The Lancet, 375(9715), 686–695. https://doi.org/10.1016/S0140-6736(09)61706-2

Haas, J. W., Bender, F. L., Ballou, S., Kelley, J. M., Wilhelm, M., Miller, F. G., Rief, W., & Kaptchuk, T. J. (2022). Frequency of adverse events in the placebo arms of COVID-19 vaccine trials: A systematic review and meta-analysis. JAMA Network Open, 5(1), Article e2143955. https://doi.org/10.1001/jamanetworkopen.2021.43955

Hafliðadóttir, S. H., Juhl, C. B., Nielsen, S. M., Henriksen, M., Harris, I. A., Bliddal, H., & Christensen, R. (2021). Placebo response and effect in randomized clinical trials: Meta-research with focus on contextual effects. Trials, 22, Article 493. https://doi.org/10.1186/s13063-021-05454-8

Howard, J. P., Wood, F. A., Finegold, J. A., Nowbar, A. N., Thompson, D. M., Arnold, A. D., Rajkumar, C. A., Connolly, S., Cegla, J., Stride, C., Sever, P., Norton, C., Thom, S. A. M., Shun-Shin, M. J., & Francis, D. P. (2021). Side effect patterns in a crossover trial of statin, placebo, and no treatment. Journal of the American College of Cardiology, 78(12), 1210–1222. https://doi.org/10.1016/j.jacc.2021.07.022

Hróbjartsson, A., & Gøtzsche, P. C. (2010). Placebo interventions for all clinical conditions. Cochrane Database of Systematic Reviews(1), Article CD003974. https://doi.org/10.1002/14651858.CD003974.pub3

Kaptchuk, T. J., Friedlander, E., Kelley, J. M., Sanchez, M. N., Kokkotou, E., Singer, J. P., Kowalczykowski, M., Miller, F. G., Kirsch, I., & Lembo, A. J. (2010). Placebos without deception: A randomized controlled trial in irritable bowel syndrome. PLOS ONE, 5(12), Article e15591. https://doi.org/10.1371/journal.pone.0015591

Kaptchuk, T. J., Kelley, J. M., Conboy, L. A., Davis, R. B., Kerr, C. E., Jacobson, E. E., Kirsch, I., Schyner, R. N., Nam, B. H., Nguyen, L. T., Park, M., Rivers, A. L., McManus, C., Kokkotou, E., Drossman, D. A., Goldman, P., & Lembo, A. J. (2008). Components of placebo effect: Randomised controlled trial in patients with irritable bowel syndrome. BMJ, 336(7651), 999–1003. https://doi.org/10.1136/bmj.39524.439618.25

Kunkel, A., Asan, L., Krüger, I., Erfurt, C., Ruhnau, L., Caliskan, E. B., Hackert, J., Wiech, K., Schmidt, K., & Bingel, U. (2024). Dopamine has no direct causal role in the formation of treatment expectations and placebo analgesia in humans. PLOS Biology, 22(9), Article e3002772. https://doi.org/10.1371/journal.pbio.3002772

Levine, J. D., Gordon, N. C., & Fields, H. L. (1978). The mechanism of placebo analgesia. The Lancet, 2(8091), 654–657. https://doi.org/10.1016/S0140-6736(78)92762-9

Spisak, T., Hartmann, H., Zunhammer, M., Kincses, B., Wiech, K., Wager, T. D., Bingel, U., & Placebo Imaging Consortium. (2026). Meta-analytic evidence for distinct neural correlates of conditioned versus verbally induced placebo analgesia. Nature Communications, 17, Article 6538. https://doi.org/10.1038/s41467-026-74743-0

Wager, T. D., & Atlas, L. Y. (2015). The neuroscience of placebo effects: Connecting context, learning and health. Nature Reviews Neuroscience, 16(7), 403–418. https://doi.org/10.1038/nrn3976

Wechsler, M. E., Kelley, J. M., Boyd, I. O. E., Dutile, S., Marigowda, G., Kirsch, I., Israel, E., & Kaptchuk, T. J. (2011). Active albuterol or placebo, sham acupuncture, or no intervention in asthma. New England Journal of Medicine, 365(2), 119–126. https://doi.org/10.1056/NEJMoa1103319

Webster, R. K., Weinman, J., & Rubin, G. J. (2016). A systematic review of factors that contribute to nocebo effects. Health Psychology, 35(12), 1334–1355. https://doi.org/10.1037/hea0000416

World Medical Association. (2024). WMA Declaration of Helsinki: Ethical principles for medical research involving human participants. https://www.wma.net/policies-post/wma-declaration-of-helsinki/

Zubieta, J.-K., Bueller, J. A., Jackson, L. R., Scott, D. J., Xu, Y., Koeppe, R. A., Nichols, T. E., & Stohler, C. S. (2005). Placebo effects mediated by endogenous opioid activity on μ-opioid receptors. Journal of Neuroscience, 25(34), 7754–7762. https://doi.org/10.1523/JNEUROSCI.0439-05.2005

Zunhammer, M., Spisák, T., Wager, T. D., Bingel, U., & The Placebo Imaging Consortium. (2021). Meta-analysis of neural systems underlying placebo analgesia from individual participant fMRI data. Nature Communications, 12, Article 1391. https://doi.org/10.1038/s41467-021-21179-3