Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Placebo, Nocebo, and Therapeutic Context

Abstract. This fundamental research monograph examines placebo effects, nocebo effects, and therapeutic context as experimentally tractable interactions among expectation, associative learning, social information, patient–practitioner communication, symptom processing, and selected neurobiological pathways. Through critical narrative synthesis of foundational experiments, mechanistic pharmacology, controlled clinical studies, methodological analyses, individual-participant neuroimaging syntheses, and systematic reviews through 2026, it establishes precise distinctions between an observed placebo response and a causal placebo effect, and between adverse-event reporting and experimentally identified nocebo effects. The review analyzes opioid-mediated placebo analgesia, learning-dependent neural variation, open-label placebo evidence, clinical interaction, trial blinding, risk communication, evidence heterogeneity, and ethical boundaries. The main conclusion is that therapeutic context can affect specific subjective and physiological outcomes under identifiable conditions, while evidence for general disease modification and universal clinical efficacy remains limited. The synthesis advances a causal, plural, and ethically grounded framework for interpreting these findings without claiming a new original experiment or a registered systematic review.

Keywords: placebo effect; placebo response; nocebo effect; therapeutic context; treatment expectancy; associative learning; classical conditioning; observational learning; analgesia; endogenous opioids; cholecystokinin; patient–practitioner interaction; clinical communication; informed consent; open-label placebo; randomized controlled trials; clinical outcomes; neuroimaging; placebo controls; symptom perception; Artificial Era

Scientific Scope and Research Questions

Placebo, nocebo, and therapeutic context constitute a field of inquiry into how information, prior experience, interpersonal interaction, and treatment environments shape health-related experience and physiological regulation. A placebo effect is a change causally attributable to the contextual and psychological features of treatment rather than to the specific mechanism attributed to the intervention. A nocebo effect is a harmful or unpleasant change caused by analogous processes, particularly negative expectations and learned threat associations. Therapeutic context encompasses verbal information, learned cues, clinician behavior, ritual, perceived credibility, and social and institutional conditions. These processes can operate alongside pharmacological and procedural interventions and can alter symptom reporting, behavior, and measurable biological responses (Colloca & Barsky (2020); Finniss & Benedetti (2010)).

The central questions are causal and mechanistic. Under what experimental contrasts can contextual effects be separated from the course of illness and from specific treatment effects? Which kinds of expectation, associative learning, and observation produce symptom change? Through which neural and peripheral pathways can those processes operate? How do communication, consent, clinical relationships, and sociocultural expectations influence therapeutic outcomes? What distinguishes a replicated laboratory response from meaningful long-term clinical benefit? The research program must also account for harmful contextual responses without minimizing genuine adverse drug reactions or underlying disease.

The scope of this dossier is the organization and effects of therapeutic context, across symptom experience and selected physiological systems. Research on a medication's molecular efficacy, a particular diagnosis, or the full methodological architecture of psychopharmacology trials belongs to adjacent scientific objects. Accordingly, this monograph treats treatment-specific efficacy as an empirical comparator and studies the contextual mechanisms that may accompany it. It is an evidence-based critical narrative synthesis of primary studies, clinical trials, mechanistic experiments, methodological papers, and systematic reviews; it does not claim to be a registered systematic review, an original clinical experiment, or a quantitative meta-analysis.

Method of Evidence Appraisal

The evidence is organized by the causal question each design can answer. Randomized placebo-versus-no-treatment comparisons are especially informative about effects attributable to a placebo procedure under specified conditions. Placebo-versus-active-treatment contrasts estimate the effect of a treatment strategy against a comparator that may carry its own contextual meaning. Open-label placebo experiments test a different intervention: a disclosed inert treatment delivered with a particular rationale and encounter. Conditioning studies manipulate prior pairings of cues and outcomes. Neuropharmacological and neuroimaging experiments probe plausible mediators, while observational studies of clinical communication describe associations that require appropriate controls before causal interpretation.

A review can preserve scientific rigor without borrowing a label of systematic review. The selected literature emphasizes identifiable original studies, independently synthesized findings, experimental mechanisms, methodological controversy, and more recent evidence through 2026. For each landmark claim, the manuscript identifies the comparison, outcome class, and limits of inference. Evidence strength depends on randomization and comparator adequacy, measured expectations, blinding, outcome objectivity, sample characteristics, follow-up, and independence of replication. A positive standardized effect across heterogeneous outcomes does not establish a common biological mechanism or equal clinical value.

The distinction between self-reported improvement and pathophysiological modification is fundamental. Pain, fatigue, nausea, and symptom-related distress are real outcomes even when expectation-sensitive; however, a patient's reported reduction in pain does not by itself demonstrate that a lesion healed, infection resolved, or a disease process reversed. Objective markers, functional measures, and long-term endpoints are necessary when the scientific claim concerns disease modification. This distinction is central to interpreting both established placebo research and contemporary clinical applications (Hróbjartsson & Gøtzsche (2010); Price & Benedetti (2008)).

Definitions and the Causal Placebo Effect

A placebo intervention is a procedure or substance whose proposed condition-specific active mechanism is absent, while its administration may still create information, expectation, conditioned associations, and a recognizable treatment encounter. An inert pill is inert with respect to the specified pharmacological target, not necessarily inert as an event in the person's life. A sham procedure may alter sensation, attention, and expectancy; a verbal reassurance may become a component of care. Precisely defining the intervention matters because different placebo procedures can contain different contextual components.

The observed placebo response is improvement recorded in a group receiving placebo. It may include a causal contextual effect, regression to the mean, spontaneous remission, cyclic symptom variation, concurrent treatments, reporting behavior, and measurement error. By contrast, the causal placebo effect is the outcome difference that would arise from assigning a defined placebo context rather than an appropriate comparison condition, all else being equal. A placebo-group before-and-after change cannot isolate that causal effect. The historical habit of treating all improvement after placebo administration as a placebo effect was challenged by the placebo-versus-no-treatment literature and remains a recurring interpretive error (Hróbjartsson & Gøtzsche (2010)).

Several components may be estimated separately. A clinician's interpersonal style can be varied while a sham procedure is held constant; treatment labeling can be varied while pill content remains the same; a conditioning history can be manipulated before a common test. Each contrast estimates a context-specific effect, not an immutable quantity called the placebo effect. This framework also explains why an effective treatment and a favorable context can coexist: the clinical outcome may depend on molecular action, behavior, and expectation, with interactions among them rather than simple arithmetic addition (Enck & Rief (2013)).

Nocebo Definitions and Attribution

Nocebo phenomena arise when treatment-related cues, anticipatory beliefs, observation, or learning contribute causally to unpleasant symptoms or reduced benefit. They include increased experimental pain after warning, heightened symptom vigilance, and expectation-related adverse experiences. The term does not identify every side effect reported after a placebo or active medication. Adverse-event reports can reflect ordinary background symptoms, underlying illness, real pharmacological harms, concurrent exposures, and spontaneous temporal variation. Identifying a nocebo effect therefore requires a comparison capable of evaluating expectation or contextual causation (Colloca & Barsky (2020); Rooney & Colagiuri (2024)).

Nocebo mechanisms cannot be inferred merely from a person's disappointment or from an adverse event's subjective nature. Pain and nausea may be experientially real despite a contextual trigger; physiological arousal and attention can also contribute. In a medication trial, both genuine drug toxicity and expectation-mediated symptom amplification may occur. Their coexistence is a scientific reason to improve measurement and communication, while preserving accurate risk disclosure, pharmacovigilance, and individual clinical evaluation.

This causal approach avoids converting the patient's complaint into a judgment about character or suggestibility. It locates explanation in experimentally tractable processes and in the interaction between individual learning histories and institutional practices. The relevant question is which portion of observed change is attributable to a specific contextual manipulation, in which population, over what interval, and for which outcome.

Historical Development of Placebo Research

Modern placebo research developed from therapeutic observations into an experimental science of treatment context. Henry K. Beecher's 1955 article The Powerful Placebo became historically influential because it assembled clinical observations of improvement in placebo-treated patients and foregrounded the role of nonspecific influences on treatment. Its celebrated estimates arose from placebo-group outcomes rather than the causal comparisons now required to separate contextual effects from natural history and regression to the mean. The article therefore marks an important historical hypothesis and a methodological lesson, not a timeless numerical law (Beecher (1955)).

In the following decades, mechanistic experiments transformed the field. The naloxone study by Levine, Gordon, and Fields in 1978 connected a form of placebo analgesia with endogenous opioid processes. Conditioning research showed that learned associations can influence responses even when explicit verbal beliefs are experimentally constrained. The late twentieth century saw increasingly refined distinctions between expectation-induced and conditioning-induced analgesia and early experiments on nocebo hyperalgesia. Neurobiological syntheses later connected several partly distinct placebo pathways to research on pain and brain regulation (Benedetti & Zubieta (2005); Levine & Fields (1978); Amanzio & Benedetti (1999)).

The early twenty-first century expanded the empirical agenda to functional neuroimaging, patient–practitioner interaction, ethically disclosed placebo, negative expectations in drug trials, and systematic reviews with control-group comparisons. These developments shifted the focus from the inert object toward identifiable learning, communication, and neurobiological processes. Their central methodological achievement was to make the treatment encounter itself an experimental variable. Recent individual-participant neuroimaging syntheses and clinical meta-analyses continue that program while showing the heterogeneity of contextual interventions (Zunhammer & Placebo Imaging Consortium. (2021). Meta-analysis of neural systems underlying placebo analgesia from individual participant fMRI data. Nature Communications (2021); Fendel & Schmidt (2025)).

Expectancy as a Psychological Mechanism

Expectation is a prediction about future sensation, symptom evolution, or treatment success. Clinical instructions, prior relief, a familiar prescription, and statements from trusted clinicians can create prospective beliefs about what the body will experience. Expectations may alter selective attention, appraisals of ambiguous sensations, anticipatory anxiety, symptom reporting, behavior, and descending modulation of nociception. Experimental evidence is strongest when investigators manipulate information while holding drug dose, procedure, and relevant exposure constant.

In an elegant analgesia experiment, Bingel and colleagues maintained pharmacological administration of the opioid remifentanil while varying treatment expectations. Positive information enhanced analgesia, while negative expectations markedly reduced reported benefit under the studied conditions. Neuroimaging provided convergent mechanistic evidence, but the design does not imply that expectation can generally replace an effective drug or reverse its every physiological action (Bingel & Tracey (2011)). Expectancy is therefore an empirically manipulable moderator of treatment experience and, for some outcomes, of treatment-related neural activity.

Expectations are multidimensional. A person may expect benefit from a drug while fearing its side effects; may trust the clinician while doubting personal improvement; or may anticipate pain relief but not restored mobility. Their effects depend on outcome domain, credibility, personal history, and congruence between instructions and experience. Manipulation checks and prospective measures of expectancy are preferable to explanations reconstructed after the outcome is known. Expectation is an important mechanism, yet neither a single necessary mediator nor a universal explanation for every placebo effect (Price & Benedetti (2008)).

Associative Learning and Classical Conditioning

Classical conditioning provides a second fundamental mechanism: treatment-related cues acquire significance through repeated pairing with biologically active outcomes. A pill's appearance, a clinical room, the temporal sequence of administration, or a clinician's gesture may later evoke learned components of a response. Such associations can influence pain-related reactions, anticipatory arousal, and selected physiological processes. The relevant causal structure requires an acquisition phase, a subsequent test phase, and a design that distinguishes learned responding from uncontrolled carryover of the active treatment.

Ader and Cohen's 1975 work on behaviorally conditioned immunosuppression demonstrated in animals that immunological responses could be influenced by learned associations. The demonstration is historically important because it broadened the concept beyond conscious report, yet animal conditioning findings are not a warrant for assuming human clinical disease control through symbols alone (Ader & Cohen (1975)). In human pain experiments, Montgomery and Kirsch showed that certain conditioned analgesic responses were mediated by explicit expectations under their test conditions. Amanzio and Benedetti later presented evidence for partially distinct systems recruited by expectation and conditioning paradigms (Montgomery & Kirsch (1997); Amanzio & Benedetti (1999)).

The most fruitful theoretical formulation treats explicit prediction and learned association as interacting processes. Conditioning can change what is expected; expectation can influence the salience of a cue and the interpretation of subsequent experience. Repetition, timing, reinforcement contingencies, extinction, and the match between learning and testing environments influence generalization. These parameters are rarely identical across laboratories and clinics, which limits efforts to translate one laboratory effect directly into a standardized therapeutic procedure.

Observation and Social Learning

People learn about treatments through other people's experiences. Observation of apparent relief, descriptions of adverse reactions, patient communities, media narratives, and clinician examples can shape expectations before an intervention begins. This social transmission can be experimentally manipulated. Colloca and Benedetti demonstrated placebo analgesia through observation of another person's experience in a controlled pain paradigm, showing that direct conditioning is not the only route by which contextual cues acquire analgesic meaning (Colloca & Benedetti (2009)).

Social learning also matters for nocebo outcomes. A systematic review and meta-analysis by Saunders and colleagues assembled experimental evidence that observing or receiving socially transmitted negative information can produce measurable nocebo-related outcomes. Its laboratory designs and heterogeneous paradigms require care when estimating the size of effects in ordinary health care or across distinct conditions (Saunders & Barnes (2024)).

Digital networks create additional social exposures. Repeated accounts of severe side effects can change perceived likelihood and attention to bodily sensations, while credible accounts of recovery can change treatment expectations. Such pathways are plausible extensions of established social learning, but the effect of a particular social media exposure requires direct study. A scientific explanation must distinguish verified drug-safety signals, inaccurate information, the frequency of symptom mention, and the individual's response to that information.

Predictive Processing and Symptom Construction

Predictive-processing theories interpret perception as the integration of prior expectations with sensory evidence. In this family of models, incoming bodily signals are assessed in relation to anticipated states, uncertainty, and the apparent reliability of different sources of information. Placebo analgesia can be described as a context-driven alteration in the interpretation or modulation of nociceptive input; nocebo hyperalgesia can be described as an increase in anticipated threat or the salience assigned to bodily signals. Such descriptions offer mechanistic hypotheses rather than direct proof that one specific computational architecture explains all responses.

Experimental studies of pain anticipation support components of this framework. Wager and colleagues found changes in functional MRI activity associated with placebo-induced expectations and pain experience. Subsequent work suggests that learning history and verbal information can engage overlapping and distinguishable neural patterns (Wager & Cohen (2004); Spisak & Bingel (2026)). Imaging correlations themselves do not establish the direction of every causal influence, and computational parameters cannot be assumed merely from a regional activation map.

A rigorous predictive account specifies what counts as a prior, how prediction error is measured, and which observations distinguish it from conditioning or attentional accounts. One testable prediction is that expectation manipulations have different effects when sensory evidence is ambiguous than when it is strong and reliable. Another is that learning should update expectations according to prior reliability and reinforcement history. These predictions can guide pre-registered experiments without treating predictive processing as an all-purpose explanation.

Attention, Interpretation, and Symptom Reporting

A clinical encounter changes the allocation of attention. Instructions may direct a patient to notice relief, monitor a symptom, anticipate risk, or interpret an unfamiliar sensation as clinically meaningful. These processes can change the experienced intensity of pain and discomfort, but they can also change how identical experiences are remembered, categorized, and reported on a scale. Symptom reporting is therefore both an outcome and a measurement process sensitive to context.

This is especially important when a study's principal endpoint is a patient-reported rating. A randomized improvement may reflect a genuine change in the experienced symptom, a shift in reporting threshold, or a combination. Blinded outcome assessment, standardized measures, pre-specified endpoints, and triangulation with functioning or physiological indicators can improve inference. These precautions are not a dismissal of subjective symptoms: pain intensity and distress are often inherently experiential. They clarify what the study can claim and support transparent comparisons with outcomes that have different levels of contextual susceptibility (Hróbjartsson & Gøtzsche (2010)).

The words chosen for a rating scale, the clinician's response to a complaint, and the time elapsed after reassurance may influence responses. Repeated measurements can themselves become part of the treatment context. Researchers should therefore document assessor roles, question wording, feedback, and monitoring procedures, particularly when interpreting small changes in highly variable symptoms.

Endogenous Opioids and Placebo Analgesia

The endogenous opioid system supplies one of the best-studied biological pathways for placebo analgesia. Levine, Gordon, and Fields reported that naloxone, an opioid receptor antagonist, interfered with placebo-induced pain relief in an experimental context. This pharmacological challenge implicated opioid-mediated processes in at least some forms of placebo analgesia. It did not establish that every placebo response in every organ system uses opioids, and the inference is specific to the study's design and endpoints (Levine & Fields (1978)).

Later neuropharmacological research separated paradigms with different learning histories. Amanzio and Benedetti found that expectation-activated analgesia and conditioning-related responses could show different sensitivity to pharmacological antagonism. Positron-emission tomography studies by Zubieta and colleagues connected placebo-induced analgesic experience with endogenous μ-opioid receptor activity in pain-related networks (Amanzio & Benedetti (1999); Zubieta & Stohler (2005)).

These studies establish that a psychosocial intervention can be associated with regulated neurochemical activity. They also impose scientific boundaries: a neural mediator is demonstrated under defined experimental conditions, often with modest samples and acute endpoints. The finding supports mechanistic reality without guaranteeing generalization to chronic disease, long-term functional outcomes, or every therapeutic ritual.

Descending Pain Modulation and Brain Networks

Placebo analgesia recruits interacting neural systems involved in anticipation, valuation, affective appraisal, and pain modulation. Functional imaging studies have implicated frontal, cingulate, insular, and brainstem regions, although specific topographies vary with task, analysis, and expectation manipulation. Petrovic and colleagues described shared neuronal network features in placebo and opioid analgesia; Eippert and colleagues provided evidence for engagement of the opioidergic descending pain control system in placebo analgesia (Petrovic & Ingvar (2002); Eippert & Büchel (2009)).

A large individual-participant neuroimaging meta-analysis by Zunhammer and colleagues pooled evidence across studies of placebo analgesia. The value of this design lies in cross-study assessment of convergence and variability rather than in identifying a single anatomical placebo center. Placebo effects involve distributed processes, and different experimental components may activate partially different networks (Zunhammer & Placebo Imaging Consortium. (2021). Meta-analysis of neural systems underlying placebo analgesia from individual participant fMRI data. Nature Communications (2021)).

Interpretation must distinguish a change in BOLD signal from direct measurement of neurotransmitter release, behavioral response, or a clinical cure. Imaging analyses are sensitive to stimulus design, subtraction contrasts, preprocessing, sample size, and the independence of regions selected for analysis. Convergent pharmacological blockade, imaging, behavioral manipulation, and replication yield stronger causal evidence than imaging alone.

Conditioned and Verbally Induced Analgesia in Recent Neuroimaging

Recent work has tested whether analgesia induced by language and analgesia reinforced through experience have identical neural signatures. In 2026, Spisak and colleagues reported a meta-analysis of individual-participant functional MRI data across experimental comparisons of verbally induced and conditioned placebo analgesia. The analysis assembled sixteen studies with 409 participants and found evidence for partly distinct neural correlates across induction methods (Spisak & Bingel (2026)).

This finding has direct theoretical significance. A placebo procedure is not a uniform physiological stimulus. An instruction that changes anticipated pain and a learning history that pairs treatment cues with actual analgesia may recruit different computational, affective, and neural processes while producing superficially similar subjective relief. Consequently, pooling placebo paradigms without coding their induction histories can conceal mechanistic heterogeneity.

The result also illustrates the strengths and limitations of neuroimaging synthesis. Individual-participant pooling permits harmonized testing beyond separate published maps, but it remains constrained by the included tasks, regions, outcomes, and available datasets. Differences in neural correlates do not by themselves show that one pathway causes all behavioral effects in its condition. Future designs can jointly randomize instructions and conditioning, collect prospective expectations, test mediation, and evaluate whether neural differences predict clinically relevant outcomes.

Neurochemical Mechanisms of Nocebo Hyperalgesia

Nocebo hyperalgesia provides an important test of biological mediation. Experimental warning can increase anticipated pain and produce an increase in experienced pain under controlled stimulation. Benedetti and colleagues reported that the cholecystokinin antagonist proglumide blocked experimentally induced nocebo hyperalgesia in their paradigm, supporting a role for this neurochemical system in a particular expectation-related pain response (Benedetti & Maggi (1997)).

The implication is mechanistic specificity, not a general formula in which negative thoughts cause every symptom. Pain enhancement can involve anxiety, attention, learning, and peripheral responses, with neurochemical mediation depending on the experimental protocol. Positive and negative expectancy manipulations need not be perfect mirror images: learning history, threat, and the consequences of missing a danger signal can create different behavioral dynamics.

Nocebo neurobiology should therefore be connected to experimentally defined outcomes, antagonist studies, validated symptom scales, and adequate safety controls. Broad statements about stress hormones, immune responses, or disease worsening require domain-specific evidence. The clinical relevance of a mechanistic finding depends on whether the outcome is acute laboratory pain, a sustained clinical symptom, medication adherence, or a verified physiological abnormality.

Therapeutic Ritual and Contextual Cues

A therapeutic ritual is a structured sequence of actions associated with care: examination, explanation, prescription, administration, follow-up, and reassessment. The ritual can communicate that a problem has been understood and that a credible intervention is occurring. Material cues such as pill appearance, equipment, and environmental organization may become meaningful through previous experience and social learning. These effects can be experimentally studied by holding a procedure constant while changing selected components of its presentation.

The context of intervention also shapes practical behavior. A person who understands a regimen may adhere more consistently, while a person frightened by an encounter may avoid follow-up or interrupt treatment. Such changes are behaviorally mediated effects of the treatment environment and should not automatically be relabeled as placebo effects. A broad account of therapeutic context must track these distinct pathways, including informational, behavioral, physiological, and measurement mechanisms (Finniss & Benedetti (2010)).

The strength of a treatment ritual can be assessed only through controlled variation and meaningful outcomes. More elaborate rituals may enhance credibility in some settings but also impose cost, deception risks, unequal access, or procedural harms. The appropriate scientific goal is to identify which components support informed care and measurable benefit without relying on an exaggerated claim that treatment theatre is itself a medical intervention for all conditions.

Patient Practitioner Interaction and Clinical Warmth

The patient–practitioner relationship is a central component of therapeutic context because clinicians provide explanations, model certainty, respond to symptoms, and support behavior across time. Warmth, attentive listening, opportunity for questions, and credible explanations may influence perceived support, expectations, treatment engagement, and symptom experience. The causal contribution of these features can be tested by randomizing aspects of the encounter while keeping other treatment conditions as constant as practicable.

In a three-arm randomized trial involving patients with irritable bowel syndrome, Kaptchuk and colleagues investigated components of a placebo intervention, including an augmented supportive relationship. Outcomes differed across the experimental packages, indicating that encounter-related factors can influence patient-reported symptoms. The design demonstrates a contextual contribution under the trial conditions; it does not identify an effect that should be generalized unchanged to every clinician, disease, and clinical system (Kaptchuk & Lembo (2008)).

A scientifically responsible reading also distinguishes interpersonal quality from persuasive certainty. A confident but inaccurate statement may create expectations while violating informed consent. Effective clinical communication aims at clarity, empathy, and realistic explanation. The most useful treatment context sustains the patient's ability to understand what is known, participate in care, report harms, and follow an appropriate evidence-based plan (Evers et al. (2018)).

Open Label Placebo and Disclosure

Open-label placebo research asks whether a person can benefit from an inert intervention while knowing that it is inert. The intervention in this paradigm includes disclosure, an explanation of possible context-related responses, a ritual of administration, and an encounter that often differs from usual care. It therefore tests a disclosed placebo package, not an effect of hidden deception.

Kaptchuk and colleagues randomized eighty adults with irritable bowel syndrome to an openly described placebo or a no-treatment comparison in a three-week trial. Patient-reported global improvement favored the open-label placebo condition. The result is a landmark demonstration that deception is not a necessary condition for a beneficial contextual response under all circumstances, but the sample, follow-up duration, and subjective endpoint limit broader inferences (Kaptchuk & Lembo (2010)).

Later randomized studies examined open-label placebo in chronic low back pain and cancer-related fatigue. These studies extend the research beyond a single symptom domain while still requiring attention to intervention wording, differences in contact, co-treatments, participant expectations, and long-term follow-up (Carvalho & Kirsch (2016); Hoenemeyer & Fontaine (2018)). Open-label placebo is scientifically distinct from covert use of a placebo and from an active treatment presented under a neutral or positive label.

Clinical Evidence Across Open Label Placebo Trials

A 2025 systematic review and meta-analysis by Fendel and colleagues synthesized randomized open-label placebo studies across populations and outcomes. It included sixty randomized controlled trials and reported a pooled benefit on the studied outcomes, with meaningful between-study heterogeneity. The reported pooled standardized mean difference was approximately 0.35, with a 95% confidence interval from 0.26 to 0.44; the review also found larger pooled effects for self-reported than for objective outcomes (Fendel & Schmidt (2025)).

A meta-analytic average is not a universal expected clinical benefit. The included interventions differed in explanation, duration, population, comparison group, and outcome measurement. Between-study heterogeneity and a prediction interval crossing the null indicate that a future trial can plausibly find little or no benefit even if the overall mean estimate favors open-label placebo. Small-study effects, expectancy-sensitive outcomes, variation in provider contact, and differences between pragmatic and explanatory designs all warrant scrutiny.

The strongest clinical implication is that contextual components can sometimes be studied and potentially used transparently, especially for particular patient-reported outcomes. It does not follow that an inert intervention can substitute for antimicrobial therapy, disease-modifying treatment, urgent diagnosis, or proven preventive care. The ethically relevant question is whether an additional context-based procedure has sufficient demonstrated benefit for a specified condition and comparison, while preserving access to effective treatment.

Irritable Bowel Syndrome as an Experimental Model

Irritable bowel syndrome has been influential in placebo research because symptoms are recurrent, clinically important, and often measured through patient report. Its variability creates both a plausible opportunity for contextual effects and a serious risk of mistaking spontaneous changes for a causal response. The 2008 components trial used multiple conditions to investigate the contribution of the encounter and placebo procedure; the 2010 open-label trial separately addressed the necessity of concealment (Kaptchuk & Lembo (2008); Kaptchuk & Lembo (2010)).

The studies answer different questions. The components trial concerns variation in ritual and interpersonal context; the later trial concerns a disclosed placebo against a comparator without that placebo package. Their results should not be combined into the claim that the same causal mediator has been isolated. Differential expectations, attention, explanation, reassurance, and contact are all plausible contributors requiring targeted experiments.

More broadly, a fluctuating disorder demands control for baseline symptom severity, time since flare, spontaneous remission, and the distinction between symptom perception and gastrointestinal physiology. Durable benefit should be assessed with longer follow-up, reproducible functional endpoints, appropriate adverse-event monitoring, and transparent disclosure of concomitant therapy.

Migraine and Treatment Labeling

Treatment labels can influence outcomes even when drug content is held constant. Kam-Hansen and colleagues examined episodic migraine treatment under different labeling conditions involving rizatriptan and placebo. Their randomized within-person design varied the information accompanying the intervention across migraine attacks and demonstrated that treatment context could modify recorded therapeutic outcomes (Kam-Hansen & Burstein (2014)).

The study has particular value because it examined an active medication and placebo within a common experimental framework. It supports the proposition that expectations may moderate experienced drug benefit, while preserving evidence that pharmacological content also matters. The findings should be interpreted in relation to the selected population, attack-level design, symptom measure, and acute treatment window; they do not imply that a label can duplicate all therapeutic effects of an active migraine medication.

Clinically, accurate discussion of what a treatment can do may influence the experience of treatment. Scientifically, labeling manipulations can help separate expectations from drug identity and identify interactions between them. Future studies should examine durability, functional recovery, and variation across patients while maintaining ethically appropriate disclosure.

Chronic Pain and Functional Outcomes

Persistent pain involves sensory, cognitive, affective, behavioral, and social processes, making it an important setting for context research. At the same time, chronic pain is heterogeneous and frequently accompanied by changes in sleep, mood, function, and treatment use. A reduction in symptom intensity over a brief interval may matter to a patient while leaving disability or the underlying cause unaffected. Trials should explicitly distinguish these endpoints.

Carvalho and colleagues tested an open-label placebo procedure for chronic low back pain in a randomized clinical trial. The outcome contributes evidence about a transparent contextual intervention and a defined symptom population, rather than establishing a general treatment for structural spinal pathology (Carvalho & Kirsch (2016)). The comparison should be interpreted with the study's contact pattern, co-interventions, and patient-reported endpoints in mind.

For chronic conditions, the crucial next step is not simply to maximize immediate analgesic expectancy. Researchers should evaluate functioning, rehabilitation participation, adverse experiences, treatment avoidance, maintenance of benefit, and consistency across care settings. When a contextual adjunct is studied alongside evidence-based care, designs should preserve access to effective treatment and report whether observed differences reflect the combined strategy or isolated contextual components.

Nocebo Effects in Medication Safety

Medication information can change the likelihood that ordinary bodily sensations are noticed and attributed to treatment. Warnings are necessary for informed consent, yet alarming or imprecise risk language may unnecessarily amplify threat expectations. Nocebo science investigates how probabilities, framing, clinician communication, and prior experiences affect adverse-symptom reporting while preserving recognition of genuine toxicities. The question is how to communicate accurately without creating avoidable additional distress.

The SAMSON N-of-1 study randomized periods of statin, placebo, and no tablets among people who had previously stopped statins because of perceived side effects. Many recorded symptoms during placebo periods resembled those during statin periods, highlighting substantial background and context-linked symptom burden in this selected cohort. The trial did not demonstrate that statins have no adverse effects; its comparison concerns symptoms in a particular previously intolerant population, with an individual-level crossover structure (Wood & Francis (2020)).

Adverse-event interpretation requires attention to the treatment's known mechanisms, serious warning signs, baseline symptom frequency, disease course, and comparator reporting. An event's occurrence after treatment is evidence of temporal succession, not automatic proof of treatment causation; equally, its occurrence during placebo exposure does not exclude a pharmacological adverse reaction in other circumstances. Accurate pharmacovigilance and contextual research are complementary.

Vaccine Trials and the Meaning of Nocebo Responses

Placebo arms of vaccine trials provide a high-profile example of adverse symptoms occurring in the absence of vaccine antigen exposure. Haas and colleagues conducted a systematic review and meta-analysis of adverse-event frequencies in placebo recipients in randomized COVID-19 vaccine trials. The reports showed substantial frequencies of systemic symptoms in placebo groups, especially after the first dose, emphasizing the need to account for background symptoms and treatment expectations (Haas & Kaptchuk (2022)).

The presence of a symptom in a placebo recipient is often described as a nocebo response, but the causal definition demands greater precision. Some adverse reports would have occurred in the absence of trial participation; others may be influenced by monitoring, expectation, stress, or the encounter. Only contrasts designed to isolate these pathways can determine the contribution of nocebo mechanisms. Estimates derived by comparing adverse-event frequencies across trial arms therefore must not be interpreted as the proportion of symptoms directly caused by negative expectations.

For risk communication, these findings justify providing balanced information about expected reactions and reliable measures of serious adverse outcomes. They do not justify dismissing reported symptoms or altering evidence-based vaccination advice. The scientific contribution lies in improved causal interpretation of symptom reports and in ethically designed communication experiments.

Nocebo Meta Analysis and Outcome Heterogeneity

Experimental nocebo studies differ in induction method, population, and dependent variable. A verbal warning that a stimulus will hurt more, observation of another person's distress, a previously unpleasant experience, and an adverse-event list on a consent form do not necessarily produce the same psychological process. Their effects may vary across experimentally delivered pain, nausea, itch, fatigue, anxiety, and medication-related symptoms. Direct learning experiments illustrate how verbal suggestions and conditioned cues interact in experimental nocebo pain (Colloca & Benedetti (2008)).

Rooney and colleagues synthesized nocebo evidence across health outcomes in a systematic review and meta-analysis, supporting the existence of negative contextual effects while drawing attention to heterogeneity in the literature (Rooney & Colagiuri (2024)). Saunders and colleagues separately examined the role of social learning, a narrower induction route requiring careful attention to study design and generalizability (Saunders & Barnes (2024)).

A methodological synthesis should avoid transferring one pooled standardized estimate to all conditions or prescribing contexts. Researchers need negative and neutral control groups, preregistered outcomes, reliable symptom measurement, and appropriate handling of multiple tests. Particularly important are experiments comparing truthful risk descriptions that preserve the same numerical facts while varying emphasis, order, framing, and opportunity for questions. Such designs can test preventable communication-related harms without withholding material information.

Blinding and Placebo Controls in Clinical Trials

Randomized placebo-controlled trials estimate treatment effects under particular comparison and masking conditions. Randomization can balance measured and unmeasured baseline causes in expectation, but it does not eliminate bias from informative dropout, differential co-interventions, selective outcome reporting, or compromised blinding after allocation. Contextual responses complicate these issues because recognizable sensations may reveal treatment identity and modify expectations.

Placebo and active-treatment groups may share some ritual features without sharing identical beliefs. Active treatments with noticeable effects can reveal allocation, creating differential symptom interpretation and rater expectations. Conversely, placebo interventions can elicit genuine symptom changes. The relevant estimand should therefore be stated explicitly: a drug versus placebo strategy under trial conditions, a drug versus active comparator, or a particular component of expectation. These are scientifically different questions (Hróbjartsson & Gøtzsche (2010)).

A trial should prespecify primary outcomes, identify who was masked, monitor treatment guesses when feasible, document informative side effects, and maintain access to safety procedures. After-treatment guesses can reflect actual outcomes, so an observed association between guessing and symptom improvement does not by itself quantify the bias produced by unblinding. Mechanistic experiments and robust trial design must be interpreted together rather than using the mere existence of placebo effects to discount efficacy evidence.

Regression to the Mean and Natural History

Many people enter studies when symptoms are unusually severe, after deterioration, or when they are actively seeking help. Subsequent measurements may be closer to their personal average even when the intervention has no causal effect. This statistical tendency, regression to the mean, is especially strong when entry is conditional on an extreme score and the measure is imperfectly reliable. Spontaneous remission and cyclic illness can produce further improvement that resembles a therapeutic response.

These processes explain why a placebo group can improve substantially without establishing a large causal placebo effect. The Cochrane review by Hróbjartsson and Gøtzsche compared placebo interventions with no-treatment conditions across clinical contexts and found that effects differed by outcome class, with the clearest signals for some patient-reported continuous outcomes and considerable variability (Hróbjartsson & Gøtzsche (2010)). This evidence disciplines claims based on pre-post changes.

Design solutions include adequate randomization, repeated baseline measurement when appropriate, appropriate waitlist or no-treatment comparisons, objective endpoints, and longitudinal modeling of symptom trajectories. Ethical constraints are crucial: no-treatment comparison is not automatically justified when effective treatment is needed. Add-on designs, delayed interventions with safeguards, and carefully chosen conditions can support causal inference while protecting participants.

Expectancy Measurement and Mediation

Expectancy research requires explicit definitions. Belief in treatment credibility, predicted symptom improvement, anticipated pain intensity, fear of side effects, and general optimism are related but distinct constructs. Their instruments should specify time frame, response scale, target outcome, and whether the item measures confidence or desire. Treating all positive attitudes as a common expectancy variable obscures mechanisms and may inflate apparent convergence across studies.

A randomized experiment can identify the effect of an expectation-inducing instruction, but mediation through a self-report expectancy score is a more demanding claim. The mediator may be measured with error, may change after early symptom response, or may be confounded by traits and care experiences. Post-randomization adjustment can remove meaningful parts of an intervention effect or introduce bias. Stronger evidence comes from repeated prospective assessment, independently varied induction techniques, careful temporal ordering, and sensitivity analysis.

Mechanistic triangulation can combine pharmacological blockade, expectancy manipulation, quantitative symptom reports, and neurophysiological signals. A successful mediation model should clarify what changes first, what remains after controlling plausible alternative pathways, and what the model predicts in new participants or settings. The goal is explanatory precision, not assigning a numerical percentage of all health change to expectation in the absence of an identified causal model.

Individual Differences and Development

Contextual responses differ across people because experience, learning, symptom history, interpersonal trust, anxiety, attentional habits, and social information differ. These variables can influence baseline expectations and responsiveness, yet evidence for a universal stable placebo-responder personality remains limited. Apparent response types may be unstable across symptom modalities, trials, and time. Individual differences should therefore be tested as pre-specified moderators with adequate power and replication rather than inferred from dramatic cases (Price & Benedetti (2008)).

Development introduces further questions. A child's beliefs about medical procedures may rely heavily on caregiver communication and observed reactions. Older adults may have extensive prior treatment histories but also complex medication regimens and competing explanations for symptoms. Studies across age groups should distinguish developmental mechanisms from differential disease burden, previous exposure, communication accessibility, and selection into research.

Cultural and linguistic variation also shapes what a treatment ritual communicates and how symptoms are described. Credibility, explanations of illness, respect, and expectations about clinicians vary across settings. Direct cross-cultural comparisons require validated measures and contextual descriptions; simply exporting a standardized placebo script may test a different intervention once its meaning changes.

Clinical Setting and Health System Organization

Therapeutic context extends beyond the consultation room. Appointment duration, continuity of care, access to treatment, medication packaging, institutional reputation, and opportunities for follow-up may alter expectations and actual treatment behavior. Some of these factors operate through psychological learning; others operate through adherence, information quality, material resources, and clinical safety. A comprehensive model must distinguish these pathways while examining how they interact.

The health system determines whether an intervention can be delivered consistently and ethically. A supportive interaction cannot compensate for lack of an indicated treatment, a missed diagnosis, or an inaccessible follow-up pathway. Conversely, an effective therapy may be undermined by confused instructions and fear-driven discontinuation. Context is therefore an integral feature of treatment implementation, though its causal effects must be estimated for each endpoint.

Future pragmatic trials can embed randomized communication or contextual enhancements within routine care, preserve standard treatment across groups, and measure adherence, symptom burden, functioning, satisfaction, and serious adverse events. Such studies can estimate whether an intervention improves real-world treatment strategies, while mechanistic laboratory studies continue to identify the components responsible for measured changes.

Ethics of Deception and Open Label Interventions

Deceptive placebo administration raises questions about patient autonomy, trust, informed consent, and the legitimacy of treatment claims. An intervention may be ethically unacceptable even if it changes a symptom, especially when it misrepresents a therapy's established efficacy or delays effective medical care. The existence of open-label placebo research changes the conceptual debate by showing that concealment is not always necessary for a contextual response, but it does not make every open-label protocol clinically warranted.

In clinical research, consent procedures may explain that participants could receive an inactive comparator and that the scientific comparison is intended to evaluate benefits and harms. Ethical approval must consider withholding or delaying established care, rescue options, adverse-event monitoring, participant vulnerability, and the scientific value of the design. The most defensible studies align valid inference with participant protection (Finniss & Benedetti (2010)).

In ordinary clinical care, transparent support, careful communication, and evidence-based treatment can incorporate beneficial contextual features without misrepresentation. Open-label placebo interventions remain condition-specific research questions: a supportive explanation alone does not prove that giving inert tablets adds value, and an observed trial benefit must be weighed against opportunity cost, adherence effects, and patient preferences.

Clinical Relevance and Safety Boundaries

The strongest empirical findings concern selected outcomes such as pain and other patient-reported symptoms, alongside specific neurobiological pathways demonstrated in controlled experiments. The presence of a contextual effect can be clinically meaningful because symptoms affect daily functioning, sleep, distress, and participation in care. Its relevance depends on magnitude, duration, reliability, availability of effective therapy, and whether the outcome valued by the patient actually improves.

Safety boundaries are equally fundamental. The literature provides no general basis for replacing proven treatments for infections, malignancies, endocrine deficiencies, or emergencies with placebo rituals. A symptom improvement may coexist with ongoing pathology. When studies involve serious disease, investigators must separately examine biomarkers, clinically meaningful events, function, and harms. Placebo research is most useful as a science of adjunctive context, trial interpretation, and ethically grounded communication, with occasional transparent interventions evaluated through condition-specific evidence.

Nocebo research likewise supports reducing misleading or unnecessarily alarming communication while protecting informed consent. A warning-related symptom may require full clinical assessment; attributing it prematurely to expectations risks missing genuine adverse drug effects or a new illness. Accurate science supports nuanced care rather than categorical dismissal.

Quantitative Synthesis and Evidence Quality

Evidence must be weighed according to the question under study. Mechanistic analgesia experiments provide relatively direct tests of expectation and neuropharmacological mediation but commonly measure brief responses in highly controlled conditions. Clinical randomized trials supply patient-centered outcomes under more realistic conditions yet often use small samples, limited follow-up, and expectation-sensitive measures. Systematic reviews summarize a wider distribution of studies but may inherit their confounding, selective reporting, and measurement heterogeneity.

The open-label placebo meta-analysis by Fendel and colleagues found an average favorable effect across its included trials while showing variation between domains and greater effects for subjective than objective outcomes. The distinction matters: a pooled effect of roughly one third of a standard deviation cannot establish disease-specific clinical importance without identifying the instrument, its minimally important difference, and the comparator (Fendel & Schmidt (2025)). Evidence about nocebo induction similarly varies by cue and symptom class (Rooney & Colagiuri (2024)).

Publication bias and selective outcome reporting remain concerns because interesting positive placebo demonstrations can attract disproportionate attention. Independent replication, preregistered primary outcomes, complete reporting of null findings, appropriate power, and access to anonymized individual-level data can strengthen the field. Scientific synthesis should maintain separate categories for experimental causality, mediator evidence, short-term patient-reported improvement, and durable treatment effectiveness.

Competing Explanations and Scientific Controversies

Several interpretive controversies recur. One concerns whether placebo effects chiefly reflect conscious expectations or associative learning. The evidence supports both routes in appropriately designed experiments and increasingly suggests different pathways depending on induction procedure and outcome. A second concerns whether reported improvement reflects changed perception, changed response thresholds, or both. Careful measurement and physiological triangulation are needed to distinguish these possibilities (Amanzio & Benedetti (1999); Spisak & Bingel (2026)).

Another controversy concerns clinical significance. An inert treatment may outperform no treatment on a symptom scale while failing to alter a more objective disease marker. These findings can both be true. Their reconciliation requires a precise statement of the outcome, comparison, and duration. Equating subjective effects with imaginary experiences is incorrect; equating symptom relief with general disease modification is equally unsupported.

Finally, some debate arises from classification. Changes in adherence caused by a reassuring explanation are context-related, but they are not necessarily the same mechanism as opioid-mediated placebo analgesia. A credible taxonomy should distinguish expectation, conditioning, social learning, behavior, neural mediation, and reporting without imposing an artificial single cause on a multicomponent clinical event.

Research Design Priorities

The next generation of research should separate induction mechanisms through factorial designs that independently vary verbal information, conditioning history, and interpersonal context. Studies should register the primary endpoint, include appropriate no-treatment or attention-controlled comparisons where ethical, assess expectations before outcomes, and identify whether differences in contact, adherence, or co-interventions could explain results. When the scientific question concerns disease modification, outcomes must include validated physiological or clinical indicators.

Clinical replication requires diverse populations, longer follow-up, transparent intervention scripts, accurate accounting for harms, and tests of treatment-context interactions. Multisite trials can determine how a laboratory effect travels across providers and cultures. Data from negative or null findings are necessary to estimate heterogeneity and reduce selective-publication bias. Individual-participant synthesis can test moderators more rigorously than ad hoc subgroup claims.

Mechanistic research should integrate conditioning designs, pharmacological antagonists where ethically appropriate, standardized sensory testing, and independently validated neural analyses. The 2026 evidence differentiating neural responses to conditioning and verbal induction offers a concrete testbed for experiments that identify mediating pathways and examine whether short-term neural markers predict durable, patient-important benefits (Spisak & Bingel (2026)).

Therapeutic Context in the Artificial Era

Digital interfaces, conversational artificial systems, symptom trackers, and algorithmically generated health messages increasingly participate in the informational context of care. Such systems can affect when a person seeks information, what risks are emphasized, how symptoms are recorded, and which explanations are considered credible. Established research on expectation, learning, and social information makes these processes legitimate hypotheses for studying human interaction with artificial communicators.

A rigorous experiment could randomize scientifically accurate versions of a health explanation across human and artificial delivery channels, hold substantive risk information constant, and compare comprehension, expectation, symptom experience, adherence, and safety. The aim would be to test whether interface characteristics and the perceived source of information have identifiable effects. An artificial interlocutor's fluency is not evidence of clinical efficacy, and persuasive communication alone does not justify autonomous medical decision-making.

Within the broader historical horizon of Artificial Era · From Homo to Artificial, therapeutic context becomes a useful example of how communication systems can alter human interpretation and behavior. The empirical scope remains explicit: no psychological states or medical authority are inferred for artificial systems without independently testable definitions and evidence. The research object in this dossier is the contextual organization of human treatment experience, including new digital mediators, not a speculative assertion that an artificial agent possesses a clinical healing capacity.

Critical Scientific Discussion

The accumulated literature establishes that contextual information can produce causally demonstrable changes in selected health-related experiences and in some measurable neural and physiological processes. The most persuasive convergence comes from randomized expectation and conditioning manipulations, pharmacological antagonist experiments, imaging results, and trials with carefully chosen comparators. Endogenous opioid involvement in some forms of placebo analgesia and cholecystokinin-related findings in certain nocebo paradigms demonstrate that contextual interventions can engage biological systems (Levine & Fields (1978); Benedetti & Maggi (1997)).

The clinical literature supports a more differentiated conclusion. Transparent open-label placebo interventions produce average benefits in some studied contexts, especially on patient-reported outcomes; these effects vary across trials and are substantially less established for objective disease outcomes. The observed placebo response in a controlled trial remains a composite of causal context and other sources of change. Reports of adverse events in a placebo group similarly require careful distinction between background events and demonstrable negative-expectation effects (Hróbjartsson & Gøtzsche (2010); Fendel & Schmidt (2025); Haas & Kaptchuk (2022)).

The theoretical synthesis is therefore plural and causal. Expectations, associative learning, social observation, clinical communication, behavioral response, and neurobiological regulation form a family of experimentally separable and sometimes interacting mechanisms. Their effects are conditional on the treatment context, learning history, outcome definition, and study design. Ethical practice follows from these scientific distinctions: accurate and supportive communication can be cultivated alongside established care while claims about specific treatment efficacy, serious risks, and disease modification remain grounded in independent evidence.

Conclusion

Placebo, nocebo, and therapeutic context describe a substantive domain of psychological and psychobiological causation within health care. Their scientific investigation requires distinctions between placebo-group improvement and a causal placebo effect, between adverse-event reporting and a demonstrated nocebo mechanism, and between symptom experience and the underlying course of disease. These distinctions make the field experimentally tractable and protect it from misleading claims.

The strongest mechanistic evidence implicates experimentally induced expectations, learned associations, social observation, endogenous opioid pathways in selected analgesia paradigms, and additional neurochemical and distributed neural processes. Controlled clinical research shows that the manner of delivering treatment can influence some outcomes, including outcomes under openly disclosed placebo conditions. Recent quantitative syntheses show average benefits alongside clinically important heterogeneity and weaker evidence for objective than for subjective endpoints.

The resulting scientific program integrates causal comparisons, explicit outcome measurement, mechanistic triangulation, ethical communication, and independent replication. Therapeutic context can contribute to how treatments are experienced and implemented. Its responsible study increases the precision of clinical research and the quality of care while preserving the indispensable distinction between contextual benefit and evidence that a specific medical intervention changes disease.

References

  • Ader, R., & Cohen, N. (1975). Behaviorally conditioned immunosuppression. Psychosomatic Medicine, 37(4), 333–340. https://doi.org/10.1097/00006842-197507000-00007
  • Amanzio, M., & Benedetti, F. (1999). Neuropharmacological dissection of placebo analgesia: Expectation-activated opioid systems versus conditioning-activated specific subsystems. Journal of Neuroscience, 19(1), 484–494. https://doi.org/10.1523/JNEUROSCI.19-01-00484.1999
  • Beecher, H. K. (1955). The powerful placebo. JAMA, 159(17), 1602–1606. https://doi.org/10.1001/jama.1955.02960340022006
  • Benedetti, F., Amanzio, M., Casadio, C., Oliaro, A., & Maggi, G. (1997). Blockade of nocebo hyperalgesia by the cholecystokinin antagonist proglumide. Pain, 71(2), 135–140. https://doi.org/10.1016/S0304-3959(97)03346-0
  • Benedetti, F., Mayberg, H. S., Wager, T. D., Stohler, C. S., & Zubieta, J.-K. (2005). Neurobiological mechanisms of the placebo effect. Journal of Neuroscience, 25(45), 10390–10402. https://doi.org/10.1523/JNEUROSCI.3458-05.2005
  • 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), 70ra14. https://doi.org/10.1126/scitranslmed.3001244
  • Carvalho, C., Caetano, J. M., Cunha, L., Rebouta, P., Kaptchuk, T. J., & Kirsch, I. (2016). Open-label placebo treatment in chronic low back pain: A randomized controlled trial. Pain, 157(12), 2766–2772. https://doi.org/10.1097/j.pain.0000000000000700
  • 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., & Benedetti, F. (2009). Placebo analgesia induced by social observational learning. Pain, 144(1–2), 28–34. https://doi.org/10.1016/j.pain.2009.01.033
  • Colloca, L., Sigaudo, M., & Benedetti, F. (2008). The role of learning in nocebo and placebo effects. Pain, 136(1–2), 211–218. https://doi.org/10.1016/j.pain.2008.02.006
  • Eippert, F., Bingel, U., Schoell, E. D., Yacubian, J., Klinger, R., Lorenz, J., & Büchel, C. (2009). Activation of the opioidergic descending pain control system underlies placebo analgesia. Neuron, 63(4), 533–543. https://doi.org/10.1016/j.neuron.2009.07.014
  • Enck, P., Bingel, U., Schedlowski, M., & Rief, W. (2013). The placebo response in medicine: Minimize, maximize or personalize? Nature Reviews Drug Discovery, 12(3), 191–204. https://doi.org/10.1038/nrd3923
  • Evers, A. W. M., Colloca, L., Blease, C., Annoni, M., Atlas, L. Y., Benedetti, F., Bingel, U., Büchel, C., Carvalho, C., Colagiuri, B., Crum, A. J., Enck, P., Gaab, J., Geers, A. L., Howick, J., Jensen, K. B., Kirsch, I., Meissner, K., Napadow, V., … Weimer, K. (2018). Implications of placebo and nocebo effects for clinical practice: Expert consensus. Psychotherapy and Psychosomatics, 87(4), 204–210. https://doi.org/10.1159/000490354
  • 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, 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), e2143955. https://doi.org/10.1001/jamanetworkopen.2021.43955
  • Hoenemeyer, T. W., Kaptchuk, T. J., Mehta, T. S., & Fontaine, K. R. (2018). Open-label placebo treatment for cancer-related fatigue: A randomized-controlled clinical trial. Scientific Reports, 8, 2784. https://doi.org/10.1038/s41598-018-20993-y
  • Hróbjartsson, A., & Gøtzsche, P. C. (2010). Placebo interventions for all clinical conditions. Cochrane Database of Systematic Reviews, 2010(1), CD003974. https://doi.org/10.1002/14651858.CD003974.pub3
  • Kam-Hansen, S., Jakubowski, M., Kelley, J. M., Kirsch, I., Hoaglin, D. C., Kaptchuk, T. J., & Burstein, R. (2014). Altered placebo and drug labeling changes the outcome of episodic migraine attacks. Science Translational Medicine, 6(218), 218ra5. https://doi.org/10.1126/scitranslmed.3006175
  • 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
  • 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), e15591. https://doi.org/10.1371/journal.pone.0015591
  • 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
  • Montgomery, G. H., & Kirsch, I. (1997). Classical conditioning and the placebo effect. Pain, 72(1–2), 107–113. https://doi.org/10.1016/S0304-3959(97)00016-X
  • Petrovic, P., Kalso, E., Petersson, K. M., & Ingvar, M. (2002). Placebo and opioid analgesia—Imaging a shared neuronal network. Science, 295(5560), 1737–1740. https://doi.org/10.1126/science.1067176
  • Price, D. D., Finniss, D. G., & Benedetti, F. (2008). A comprehensive review of the placebo effect: Recent advances and current thought. Annual Review of Psychology, 59, 565–590. https://doi.org/10.1146/annurev.psych.59.113006.095941
  • Rooney, T., Sharpe, L., Todd, J., Tang, B., & Colagiuri, B. (2024). The nocebo effect across health outcomes: A systematic review and meta-analysis. Health Psychology, 43(1), 41–57. https://doi.org/10.1037/hea0001326
  • Saunders, C., Tan, W., Faasse, K., Colagiuri, B., Sharpe, L., & Barnes, K. (2024). The effect of social learning on the nocebo effect: A systematic review and meta-analysis with recommendations for the future. Health Psychology Review, 18(4), 934–953. https://doi.org/10.1080/17437199.2024.2394682
  • Spisak, T., Hartmann, H., Zunhammer, M., Kincses, B., Wiech, K., Wager, T. D., & Bingel, U. (2026). Meta-analytic evidence for distinct neural correlates of conditioned versus verbally induced placebo analgesia. Nature Communications, 17, 6538. https://doi.org/10.1038/s41467-026-74743-0
  • Wager, T. D., Rilling, J. K., Smith, E. E., Sokolik, A., Casey, K. L., Davidson, R. J., Kosslyn, S. M., Rose, R. M., & Cohen, J. D. (2004). Placebo-induced changes in fMRI in the anticipation and experience of pain. Science, 303(5661), 1162–1167. https://doi.org/10.1126/science.1093065
  • Wood, F. A., Howard, J. P., 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. (2020). N-of-1 trial of a statin, placebo, or no treatment to assess side effects. New England Journal of Medicine, 383(22), 2182–2184. https://doi.org/10.1056/NEJMc2031173
  • 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., & Placebo Imaging Consortium. (2021). Meta-analysis of neural systems underlying placebo analgesia from individual participant fMRI data. Nature Communications, 12, 1391. https://doi.org/10.1038/s41467-021-21179-3