CP13CH04-Wager ARI 13 April 2017 17:10

ANNUAL
REVIEWS Further
Click here to view this article's
online features:
• Download figures as PPT slides
• Navigate linked references
• Download citations
• Explore related articles
Brain Mechanisms of the
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

• Search keywords
Placebo Effect: An Affective
Appraisal Account
Yoni K. Ashar,1 Luke J. Chang,2 and Tor D. Wager1,3
1
Department of Psychology and Neuroscience, University of Colorado, Boulder,
Colorado 80309
2
Department of Psychological and Brain Sciences, Dartmouth College, Hanover,
New Hampshire 03755
3
Institute of Cognitive Science, University of Colorado, Boulder, Colorado 80309;
email: tor.wager@colorado.edu

Annu. Rev. Clin. Psychol. 2017. 13:73–98 Keywords

First published online as a Review in Advance on expectation, emotion, default mode, pain, mood disorders, Parkinson’s
March 27, 2017
disease
The Annual Review of Clinical Psychology is online at
clinpsy.annualreviews.org Abstract
https://doi.org/10.1146/annurev-clinpsy-021815- Placebos are sham medical treatments. Nonetheless, they can have substan-
093015
tial effects on clinical outcomes. Placebos depend on a person’s psychological
Copyright  c 2017 by Annual Reviews. and brain responses to the treatment context, which influence appraisals of
All rights reserved
future well-being. Appraisals are flexible cognitive evaluations of the per-
sonal meaning of events and situations that can directly impact symptoms
and physiology. They also shape associative learning processes by guiding
what is learned from experience. Appraisals are supported by a core net-
work of brain regions associated with the default mode network involved
in self-generated emotion, self-evaluation, thinking about the future, social
cognition, and valuation of rewards and punishment. Placebo treatments for
acute pain and a range of clinical conditions engage this same network of
regions, suggesting that placebos affect behavior and physiology by changing
how a person evaluates their future well-being and the personal significance
of their symptoms.

73
 CP13CH04-Wager ARI 13 April 2017 17:10

Contents
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
PLACEBO RESPONSES AND PLACEBO EFFECTS
IN CLINICAL CONTEXTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Do Placebos Have Meaningful Effects on Clinical Outcomes? . . . . . . . . . . . . . . . . . . . . 75
How Reliable Is the Placebo Effect? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Which Disorders Respond Most to Placebo Treatments? . . . . . . . . . . . . . . . . . . . . . . . . . 78
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

MECHANISMS OF PLACEBO EFFECTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Appraisal and Precognitive Association: Separable but Interacting Mechanisms . . . . 80
Brain Mechanisms Underlying Placebo Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
A Brain System for Affective Appraisal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
CONCLUSIONS AND FUTURE DIRECTIONS: HOW DO
APPRAISALS HEAL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

INTRODUCTION
“One of the most successful physicians I have ever known, has assured me that he used more bread pills, drops of
colored water, & powders of hickory ashes, than of all other medicines put together.” —Thomas Jefferson, Letter
to Caspar Wistar, 1807

In a recent clinical trial for Parkinson’s disease, physicians surgically injected a viral vector designed
to enhance dopaminergic function directly into patients’ brains (Olanow et al. 2015). Similar
treatments had worked in nonhuman primates, and results from an open-label trial in patients
were promising. Buoyed by these findings, the researchers conducted a large, multisite, double-
blind randomized trial comparing the novel treatment to a sham surgery.
The patients injected with the treatment showed marked, sustained improvement over 2 years,
as the researchers hoped. Yet, surprisingly, patients who underwent the sham surgery improved
at the same rate over the same time period—and these gains were maintained for at least 2 years
(Figure 1). Thus, the trial failed, signaling a potential finale to a decade-long program of ground-
breaking research and triggering the sale of the company that funded it. However, it provided
a remarkable demonstration of placebo-related improvements in a neurodegenerative disorder
typically characterized by progressive decline.
This trial is not an isolated phenomenon. Clinically meaningful placebo effects have been
observed in depression (Cuijpers et al. 2012, Khan et al. 2012), chronic pain (Hróbjartsson &
Gøtzsche 2010, Madsen et al. 2009), irritable bowel syndrome (IBS) (Kaptchuk et al. 2008, 2010),
and other conditions (Hróbjartsson & Gøtzsche 2010). In each case, patients given placebos fare
substantially better than those in no-treatment conditions (natural history), demonstrating causal
effects of placebos (Figure 2). Placebo effects also contribute to the effectiveness of many active
treatments. For instance, the analgesic effects of several commonly used painkillers are markedly
reduced when patients do not know they are receiving them (Atlas et al. 2012, Benedetti et al. 2003a,
Colloca et al. 2004), and patients who adhere to medication for heart disease live longer—even
if the medication is a placebo and even when controlling for a number of potential confounding
variables (Pressman et al. 2012).
Placebos are by definition inert. They are sham medical treatments—drugs, devices, or other
treatments with no inherent potency. How, then, can they heal? Their therapeutic potential lies

74 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Real versus sham surgery in Parkinson’s disease

0

Improvement in symptoms (UPDRS)

–2

–4 Sham surgery
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

–6

Real surgery

–8
Baseline 15 24
Month
Figure 1
The placebo response. Sham brain surgery led to improvements in Parkinson’s symptoms that were as large
as active treatment (verum surgery) up to 2 years later (Olanow et al. 2015). Symptoms were measured with
Unified Parkinson’s Disease Rating Scale (UPDRS) motor off scores.

in the patient’s brain and is driven by the patient’s responses to the psychosocial context in which
the placebo treatment is delivered (Benedetti 2014, Büchel et al. 2014, Wager & Atlas 2015).
Key elements of the treatment context include the patient’s relationships with care providers and
other cues and rituals, such as visiting a doctor’s office or taking a pill; these influence patients’
appraisals about how a treatment will affect them, including expectations for recovery (Colloca &
Miller 2011, Finniss et al. 2010, Frank & Frank 1993, Price et al. 2008). Placebo effects thus offer
a window into how psychosocial processes impact health and disease.
Here, we review placebo effects on clinical outcomes and explore their behavioral and brain
mechanisms. We argue that placebo effects are created largely by psychological appraisals. Ap-
praisals depend on cognitive beliefs and also influence precognitive learning processes to create
placebo effects that do not depend on cognitive beliefs or expectations. We present meta-analytic
findings showing that placebo treatments engage a brain system that mediates multiple varieties of
appraisals, including self-generated emotions, expectations, valuation, self-evaluations, and beliefs
about others. This system overlaps with the default network, a brain system that is involved in
spontaneous thought and feeling (Gusnard & Raichle 2001, Raichle et al. 2001) and is implicated
in multiple mood disorders (Etkin & Wager 2007, Kaiser et al. 2015). This colocalization pro-
vides a neurobiological connection among placebo effects, emotional appraisal, and mood and
pain disorders.

PLACEBO RESPONSES AND PLACEBO EFFECTS

IN CLINICAL CONTEXTS

Do Placebos Have Meaningful Effects on Clinical Outcomes?

The response to placebo treatment is often as large as the response to pharmacological treatment,
especially for chronic pain (Tuttle et al. 2015) and depression (Fournier et al. 2010, Kirsch et al.
2008). However, only some of the observed response is caused by the placebo treatment. Patients
often enroll in trials when fluctuating symptoms are at their worst and then improve due to

www.annualreviews.org • Mechanisms of Clinical Placebo 75

 CP13CH04-Wager ARI 13 April 2017 17:10

Placebo effects: placebo versus no treatment

Standard mean difference (SMD)

0.6
a N = 11
(1,610)

N = 44
(4,045)
0.4 N = 109 N = 61
N = 158 trials (8,000) (3,922)
(10,525 patients) N = 49
(2,513) N = 53
(2,546)
0.2
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

0
Overall Studies Patient- Observer- Pill Psychological Physical
with low reported reported placebo placebo placebo
risk of bias outcomes outcomes

Placebo effects
in acupuncture for pain Placebo effects in depression
Worse N = 13 trials c
b (3,025 patients)
2.0 Specific
Pain at end of trial

Wait list Drug Wait list

28% 33% factors
18%
1.5 17%

1.0 Placebo Common factors

54% 50%
0.5

0 Pharmacotherapy Psychotherapy
Standard Sham Acupuncture
care acupuncture

Figure 2
The placebo effect. (a) Meta-analysis of clinical trials comparing placebo treatments to treatment as usual or
no treatment (Hróbjartsson & Gøtzsche 2010). Numbers of trials and patients are shown. Psychological
placebos include, for example, nondirective, supportive conversations. Physical placebos include, for
example, sham acupuncture and sham surgery. (b) The placebo effect accounts for most of the response to
acupuncture in the treatment of clinical pain conditions; standardized pain scores at the end of treatment are
shown from the Madsen et al. (2009) meta-analysis. (c) Estimated effect sizes for placebo (Khan et al. 2012)
and common factors (Cuijpers et al. 2012) in the treatment of depression. Common factors are closely
conceptually related to placebo effects. They include patient expectations, the patient-provider relationship,
and other factors common to almost all psychotherapies.

regression to the mean or spontaneous recovery (see, e.g., Wager & Fields 2013). Similarly,
patients with the worst outcomes on placebo may drop out of studies, creating sampling biases
that increase the apparent placebo response. Thus, it is crucial to distinguish the placebo effect—
the mind-brain response to the placebo specifically—from the placebo response, or the overall
response to placebo treatment.
Placebo effects in clinical disorders can be estimated in at least four ways. One common ap-
proach is to compare placebo treatment to no-treatment (natural history) or treatment-as-usual
control groups in randomized clinical trials. Hróbjartsson and Gøtzsche have conducted several
influential meta-analyses of such trials (most recently, Hróbjartsson & Gøtzsche 2010). Collaps-
ing across all disorders and placebo treatment modalities, they found small but significant placebo
effects [Standardized Mean Difference (SMD) = −0.23, 95% confidence interval (CI) (−0.28

76 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

−0.17)], with stronger effect sizes in trials of high methodological quality [SMD = −0.38, 95%
CI (−0.55 −0.22)] (Figure 2a). Other studies have provided more focused examinations of placebo
effects in specific clinical contexts.
Focusing on pain, a meta-analysis including over 3,000 patients found comparable ther-
apeutic effects of acupuncture and sham acupuncture, both of which provided substantially
more relief than standard care based on physical therapy or medication regimes (Madsen
et al. 2009) (Figure 2b). Focusing on depression, Khan et al. (2012) conducted a meta-
analysis of trials with placebo, drug, and wait-list conditions (Figure 2c, left). They decom-
posed the overall drug response into active drug effects (18% of the total effect), placebo effects
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

(54%), and improvement in wait-list conditions (28%). Highly similar results were reported by
Cuijpers et al. (2012) in a meta-analysis that decomposed the effects of psychotherapy treatments
for depression into nontreatment factors (i.e., wait list, 33%); treatment-specific factors, such as
mindfulness or cognitive restructuring (17%); and common factors, such as patient expectations
and the patient-therapist relationship, which are conceptually closely related to placebo effects
(50%) (Figure 2c, right) (Wampold et al. 2015). Focusing on anxiety disorders, Bandelow et al.
(2015) conducted a meta-analysis of clinical trials, including 234 studies with over 37,000 pa-
tients. They found treatment effect sizes of d = 1.29 for placebo pills, d = 0.83 for psychological
placebos, and d = 0.20 for wait-list controls, demonstrating substantial placebo effects.
A second approach to estimating placebo effects is to compare open drug treatment with
hidden drug treatment—when patients are aware versus unaware that they are receiving a drug.
These studies have demonstrated substantial placebo effects on experimental pain (Atlas et al.
2012), postoperative pain, and Parkinson’s disease (Colloca et al. 2004), with awareness of drug
administration treatment accounting for half or more of drug effects.
A third approach to estimating placebo effects is to compare outcomes across clinical trials in
which patients had different probabilities of receiving active treatment or placebo. Increased prob-
ability of receiving active treatment enhances patient expectations of improvement (Rutherford
2016). Schizophrenia patients had twice as large a response to the same medication when ad-
ministered in a comparator trial (comparing two or more active drugs without a placebo arm)
relative to when administered in a placebo-controlled trial (Rutherford et al. 2014, Woods et al.
2005). Similarly, among depressed patients, both drugs and placebos achieved an approximately
10% higher response rate in comparator trials relative to placebo-controlled trials (Papakostas &
Fava 2009, Sinyor et al. 2010, Sneed et al. 2008). Similar results are observed for several anxiety
disorders (Rutherford et al. 2015), highlighting the importance of treatment context.
A fourth approach to estimating placebo effects is to compare different types of placebo treat-
ments (Figure 2a). In a systematic review of migraine prophylaxis, Meissner et al. (2013) found a
modest (26%) response rates to sham pills, injections, and herbs; a larger response rate to sham
acupuncture (38%); and an even larger response rate to sham surgery (58%). A similar pattern
of increased response to more invasive placebo treatments was found in a meta-analysis of 149
trials of osteoarthritis pain (Bannuru et al. 2015) and in a meta-analysis of 11 trials of Parkinson’s
disease (Goetz et al. 2008), although one meta-analysis did not find differences among different
placebo modalities (Fässler et al. 2015). Thus, the modality of the placebo treatment contributes
to the placebo effect, likely due to patients’ appraisals of a modality’s potency.
Placebo effects can last for months to years. For example, in clinical trials for neuropathic pain,
Parkinson’s disease, and depression, the placebo response appears to grow over the course of the
trial and is reliably observed for months or even years after initiating placebo treatment (Khan
et al. 2008, Marks et al. 2010, Olanow et al. 2015, Quessy & Rowbotham 2008, Tuttle et al. 2015).
A better understanding of when placebo effects are sustained vs. when they naturally extinguish
may help shed light on placebo mechanisms.

www.annualreviews.org • Mechanisms of Clinical Placebo 77

 CP13CH04-Wager ARI 13 April 2017 17:10

How Reliable Is the Placebo Effect?

How reliably will a person respond to identical administrations of a placebo treatment? And will
a person’s response to one placebo treatment correlate with his or her response to a different
placebo treatment?
In one of the most rigorous investigations of placebo reliability, Whalley et al. (2008) found
that the analgesic responses to a placebo cream had moderately high test-retest reliability at one
week, r = 0.60–0.77, agreeing with an estimate of r = 0.55 from a similarly designed experiment
(Morton et al. 2009). This suggests that responses to the same placebo treatment are reliable over
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

time. Yet, Whalley et al. (2008) also found that responses to one placebo cream were uncorre-
lated with responses to another placebo cream that was differently labeled but otherwise identical,
r(69) = 0.10, p < 0.41. This pattern of reliable responding to identical placebo treatments but un-
reliable responding to different placebo treatments is supported by other previous studies (Kessner
et al. 2013, 2014). For example, Kong et al. (2013) reported that participants had uncorrelated
analgesic responses to sham acupuncture and pill placebo treatments.
Little evidence indicates that placebos elicit reliable responses in clinical contexts. One study did
find that responses to an oral and intravenous antidepressant placebo treatment were correlated,
r = 0.35; p = 0.04, and that response to the oral placebo also predicted later response to an active
medication (Peciña et al. 2015). However, findings from other studies point to low reliability
across contexts. Liberman (1967) found that placebo responses were uncorrelated across three
types of pain, including experimental pain and the pain of childbirth. Müller et al. (2016) found
that placebo analgesia in experimental pain was uncorrelated with responses to a placebo treatment
for chronic pain. Similarly, meta-analyses of clinical trials for depression have found that patients’
gains during the placebo lead-in phase are not related to their placebo responses during the active
phase (Posternak et al. 2002).
One interpretation of these findings is that placebo effects depend strongly on individuals’
appraisals of the treatment context. Thus, they can be reliable when the treatment context (and
corresponding appraisals) are held constant, but they can change dramatically with even relatively
minor changes in the treatment context (Koban et al. 2013) such as changing the name of a cream
(Whalley et al. 2008).

Which Disorders Respond Most to Placebo Treatments?

Clinical studies suggest substantial placebo responses in many disorders. These include multiple
pain conditions (Tuttle et al. 2015, Vase et al. 2002), such as osteoarthritis (Bannuru et al. 2015,
Moseley et al. 2002), migraine (Kam-Hansen et al. 2014, Meissner et al. 2013), IBS (Kaptchuk
et al. 2010, Vase et al. 2005), and labor pain (Liberman 1967). Substantial placebo responses are
also observed in depression (Cuijpers et al. 2012, Fournier et al. 2010, Kirsch et al. 2008), anxiety
(Bandelow et al. 2015), Parkinson’s disease (Goetz et al. 2008), schizophrenia (Rutherford et al.
2014), asthma (Wechsler et al. 2011), urological conditions (Sorokin et al. 2015), menopausal hot
flashes (Freeman et al. 2015), and other conditions.
Estimating relative effect sizes across disorders is challenging. One meta-analytic comparison of
responses to placebo pills across six patient groups found the largest placebo effects in generalized
anxiety disorder and panic disorder and the smallest effects in psychosis and obsessive-compulsive
disorder (Khan et al. 2005) (Figure 3a). Effects on depression and posttraumatic stress disorder
were in between.
In a comparison of sham surgical interventions across disorders, Jonas et al. (2015) found
moderately large placebo responses in pain, gastroesophageal reflux disease, and other conditions

78 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Placebo pill responses Placebo surgery responses

N = 10
70 a Placebo 2.5 b (287) Sham surgery
N=2
Drug Real surgery

Treatment response (SMD)

60
Improvement CGI-S (%)

2
50
N=5 N=8
N=8 (655)
40 1.5 (342)
N=4 N = 15 trials
N = 42
30 (1,584
N = 11 1 patients)
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

20
N = 6 trials
0.5
10

0 0
Psychosis OCD PTSD Panic MDD GAD Pain Obesity GERD Other

Figure 3
Placebo responses across disorders. (a) Pill placebo responses in different psychiatric disorders (Khan et al. 2005). (b) Sham surgery
responses across different conditions ( Jonas et al. 2015). The number of trials and patients (if available) in each condition is shown.
Abbreviations: CGI-S, Clinical Global Impressions of Severity Scale; GAD, generalized anxiety disorder; GERD, gastroesophageal
reflux disease; MDD, major depressive disorder; OCD, obsessive-compulsive disorder; PTSD, posttraumatic stress disorder; SMD,
standardized mean difference.

(SMD ≈ 0.5) (Figure 3b). For pain, the effects of sham surgery were nearly as large as real surgery,
and were statistically indistinguishable. Because sham surgery studies typically do not include no-
treatment control groups, the observed response includes factors not due to the sham surgery
specifically, such as regression to the mean.
Although placebo treatments can affect physiological outcomes, such as hormone production
or urinary flow rate (Meissner 2011, Meissner et al. 2007, Sorokin et al. 2015, Wager & Atlas
2015), they are largest for psychological outcomes (Figure 2a) (Hróbjartsson & Gøtzsche 2010).
For example, one study found that a placebo treatment for asthma improved subjective symptom
severity but did not improve forced expiratory volume, an objective measure of lung function
(Wechsler et al. 2011). This suggests that there is a larger potential for placebo effects on outcomes
that are more closely linked to patients’ emotional and motivational states.

MECHANISMS OF PLACEBO EFFECTS

Theories of placebo mechanisms during the past several decades have focused primarily on expec-
tations and learning (Kirsch 1985, Stewart-Williams & Podd 2004). Recent work has elaborated
our understanding of how each class of processes works at both behavioral and brain levels of
analysis (Benedetti 2014, Büchel et al. 2014, Enck et al. 2008, Price et al. 2008, Wager & Atlas
2015). Though the terms expectations and learning are used across many disciplines, they often
have domain-specific meanings, which can lead to confusion. Here, we recast theories of expecta-
tion and learning in a somewhat broader context, focusing on interactions between precognitive
associations, which refer to learned associations that can operate without cognitive awareness or
intervention, and appraisals, which refer to interpretations of the meaning of events in a given
context. These two broad classes of mechanisms interact with one another to create and maintain
placebo effects. In this paper, we differentiate appraisals from precognitive associations and outline
their critical role in the mechanisms underlying placebo effects. A key aspect of our argument is
that the appraisal system is supported by the coordinated functioning of multiple brain systems

www.annualreviews.org • Mechanisms of Clinical Placebo 79

 CP13CH04-Wager ARI 13 April 2017 17:10

involved in emotion and cognition, converging in the ventromedial prefrontal cortex (vmPFC)
and other regions in the so-called default mode network.

Appraisal and Precognitive Association: Separable but Interacting Mechanisms

Placebo effects are supported by two distinct but interacting processes: precognitive associations
and appraisals.

Precognitive associations. Precognitive associations are responses to stimuli that are automatic
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

in the sense that they can occur without cognitive effort and are largely invariant to cognitive
context and goals. These associations are learned based on experience and are mediated by plasticity
in stimulus- and response-specific neural pathways that are distributed throughout the nervous
system. For example, in classical conditioning paradigms, Aplysia and Drosophila exhibit single-trial
learning in response to aversive events (shocks), which manifests in the strengthening of specific
neural pathways associated with defensive responses (Carew et al. 1983). Anencephalic animals and
humans deprived of a forebrain and cortex can also learn to generate complex affective behavior
(Berntson & Micco 1976) and autonomic responses via classical conditioning to shocks (Berntson
et al. 1983). Likewise, the isolated spinal cord can learn complex motor responses, including
learning to anticipate and avoid shocks (Grau 2014).
Precognitive associations with drug and context cues are likely to underlie some forms of
placebo effects. Placebo effects are readily obtainable in rodents (Guo et al. 2011, Herrnstein
1962, Woods & Ramsay 2000) and humans by pharmacological conditioning, which involves
repeated pairing of drugs with drug cues, usually over several days, and then testing by delivering
the cues alone. Such procedures can produce effects on hormonal and immune responses (Goebel
et al. 2002, Schedlowski & Pacheco-Lopez 2010), which after conditioning appear to be insensitive
to verbal instructions about the treatment and, thus, presumably to patients’ beliefs (Benedetti
et al. 2003b, Wendt et al. 2013). Similarly, placebo analgesic responses become stronger and more
durable with longer conditioning (Carlino et al. 2014, Colloca et al. 2010, Colloca et al. 2008a).
After several days of training, responses can persist even after subjects are explicitly told that the
treatment is inert (Schafer et al. 2015), suggesting a shift from being driven by beliefs to being
driven by more stable precognitive associations.

Appraisals. Appraisals are cognitive evaluations of events and situations (Smith & Ellsworth
1985). This simple definition belies complexity: Situations are integrated mental representations
of multiple kinds of information, including precognitive associations, long-term memories, expec-
tations, goals, representations of others’ mental states, and interoception of internal bodily states
(Roy et al. 2012). Appraisals are not simple perceptions but rather constructed interpretations of
events (Wilson-Mendenhall et al. 2011). Whereas precognitive associations are reactive responses
to events, appraisals are conceptual acts (Barrett 2014). The appraisals that generate emotions are
those with personal meaning—they matter to the self and one’s future well-being. This sense of
personal meaning is thought to be central in generating both emotions (Barrett 2012, Ellsworth &
Scherer 2003, Lazarus & Folkman 1984, Ortony et al. 1988, Scherer 2001) and placebo responses
(Moerman & Jonas 2002).
Appraisals play a critical role in many kinds of active treatments. They stand at the center
of many psychotherapies, which aim to explicitly alter patients’ appraisals of clinically relevant
events and stimuli through reframing, cognitive restructuring, or other techniques. For example,
depressed patients’ use of cognitive restructuring skills during psychotherapy sessions predict
relapse rates one year posttreatment, controlling for a number of confounding variables (Strunk

80 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

et al. 2007). Pretreatment expectations of psychotherapy efficacy can also account for substantial
variance in treatment outcomes (Gaston et al. 1989, Joyce & Piper 1998, Sotsky et al. 1991) and
are often related to perceived treatment credibility (Hardy et al. 1995, Kazdin & Krouse 1983)
and to the competence of care providers (Frank & Frank 1993). Relatedly, patient appraisals
of perceived doctor empathy predict reduced severity and duration of cold symptoms and an
increased immune response (Rakel et al. 2011).
Appraisals play a central role in placebo effects, especially those induced by verbal suggestion.
Verbal suggestions alone have been shown to modulate adrenocorticotropic hormone and cortisol
responses to ischemic pain (Benedetti et al. 2006), autonomic responses to painful events ( Jepma
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

& Wager 2015, Nakamura et al. 2012), and skin conductance during threat of shock (Meyer et al.
2015). Similarly, information about how other people had experienced painful stimuli robustly
modulated pain-related autonomic responses (Koban & Wager 2016), although this information
was never systematically reinforced. Finally, suggestions (in the form of package labeling) that
a milkshake was “indulgent” resulted in reduced blood levels of the hunger-related hormone
ghrelin, relative to a milkshake labeled as “sensible” (Crum et al. 2011). Together, these findings
emphasize the role of the appraisal system in mediating placebo effects on multiple outcomes,
including physiological responses.

Interactions between precognitive associations and appraisal. Appraisals and precognitive

associations interact to create placebo effects. Placebo effects are typically small when they are
induced solely by conditioning (Carlino et al. 2014, Montgomery & Kirsch 1997, Vase et al. 2002)
or solely by verbal suggestions (Colloca et al. 2008b, de Jong et al. 1996). The largest placebo
effects are induced when suggestions are reinforced via conditioned experience (Carlino et al. 2014,
Colloca et al. 2008b, Schafer et al. 2015, Vase et al. 2002). This suggests that obtaining robust
placebo effects requires repeated success experiences coupled with a cognitive attribution of benefit
to the treatment. By governing attributions, appraisals guide precognitive associative learning
processes, shaping what is learned from experience (for further discussion see Wager & Atlas 2015).

Brain Mechanisms Underlying Placebo Effects

Because there is “not one placebo effect but many” (Finniss et al. 2010, p. 687), placebo effects are
not restricted to one brain system. Rather, multiple systems and mechanisms are involved, and
understanding the principles by which appraisals and precognitive associations map onto brain
systems is a challenge. The study of placebo effects can teach us something about these mappings,
and conversely, understanding these mappings contributes to understanding the neurophysiology
of placebo effects.
Placebo effects have been studied with functional magnetic resonance imaging (fMRI); electro-
and magnetoencephalography (EEG and MEG); and positron emission tomography (PET)-based
imaging of glucose, dopamine, and opioid activity. Most studies—approximately 50 to date—have
focused on placebo analgesia, which has allowed quantitative meta-analyses of consistent placebo-
induced analgesia findings across laboratories and paradigms (e.g., Amanzio et al. 2013, Atlas &
Wager 2014). The placebo analgesia literature provides a foundation for insights into placebo
effects in general and for comparing placebo-induced brain-activation patterns with those related
to appraisal and other processes.

Pain-related processes reduced by placebo treatment. One question relates to the depth
of placebo effects: Can placebo treatments influence symptoms in fundamental ways? As shown
in Figure 4, placebo analgesics have been found to reduce pain-related activity in the cortex

www.annualreviews.org • Mechanisms of Clinical Placebo 81

 CP13CH04-Wager ARI 13 April 2017 17:10

Placebo-induced changes in brain function

fMRI activity decreases during pain

1
S1 ANTERIOR
dlPFC
S2
dpIns
aIns
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

vlPFC
aMCC POSTERIOR
OFC
Opioid release ANTERIOR
Thalamus 2

Hy NAc
PAG vmPFC

POSTERIOR Amy
PBN

RVM (NRM) Spinal fMRI Brain stem activation

3 4
Noxious
stimuli
Hypothalamus
L T
LEF

PAG

Spinal cord
DOR
DO
ORRSAL
SSAAL RVM

x = –4

Figure 4
Brain mechanisms and pathways involved in placebo analgesia. Pathways involved in pain representation are shown in blue. Regions
that modulate activity in pain-encoding circuits are shown in orange. Clockwise, from upper right:  fMRI results showing brain
regions in red that decrease during pain (Wager et al. 2004);  regions with placebo-induced increases in µ-opioid activity (red and
yellow; Wager et al. 2007);  pain-related spinal cord activity ( yellow square) reduced by placebo treatment (Eippert et al. 2009b); and
 brain stem regions activated by placebo treatment (Eippert et al. 2009a). Abbreviations: aIns, anterior insula; Amy, amygdala;
aMCC, anterior midcingulate; dlPFC, dorsolateral prefrontal cortex; dpIns, dorsal posterior insula; fMRI, functional magnetic
resonance imaging; Hy, hypothalamus; NAc, nucleus accumbens; OFC, orbitofrontal cortex; PAG, periaqueductal gray; PBN,
parabrachial nucleus; RVM (NRM), rostral ventral medulla (nucleus raphe magnus); S1 and S2, primary and secondary somatosensory
cortex; vlPFC, ventral lateral prefrontal cortex; vmPFC, ventromedial prefrontal cortex.

(Wager et al. 2004) and spinal cord (Eippert et al. 2009b) and to activate the endogenous opioid
system (Wager et al. 2007) and specific brain stem nuclei associated with pain control (Eippert
et al. 2009a).
These examples are supported by a quantitative meta-analysis of published studies (Wager
& Atlas 2015). Placebos can, under some circumstances, reduce pain-related brain responses in
most or all of the cortical and subcortical targets of pain-related somatosensory input (Figure 5,
blue). The most consistent reductions are in the anterior midcingulate, thalamus, and mid- and
anterior insula. In a number of studies, these brain reductions correlate with the magnitude of
reductions in pain (for a detailed review, see Wager & Atlas 2015). Reductions in sensorimotor

82 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Individual studies Most consistent reductions

a b
S2/dplns
aMCC
Amygdala
alns S2/dplns
aMCC dmPFC

Thalamus
LATERAL Thalamus
alns
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

Amygdala

Thalamus
MEDIAL

Individual studies Most consistent increases

c d

dIPFC
pTPJ
vmPFC

LATERAL vIPFC

dIPFC

ORBITAL
PAG
NAc/VS
MEDIAL

Figure 5
Consistent findings in neuroimaging studies of placebo analgesia. (a,c) Peak activation locations in studies of placebo analgesia. Each
sphere is a finding from an activation map, with (a) blue spheres indicating decreases in pain-related activity (21 studies) and (c) yellow
spheres indicating increases in pain- and anticipation-related activity (19 studies) (for details, see Wager & Atlas 2015). Locations from
the same map within 12 mm were averaged into one sphere. (b,d ) Consistent activations with at least three studies reporting effects
within 10 mm. (b, blue) Consistent reductions during pain occur in the S1 and S2, thalamus, dACC, and aIns, which are associated with
pain encoding. (d, yellow and pink) Consistent increases with placebo occur in the vmPFC, NAc/VS, PAG, dlPFC and vlPFC, and
pTPJ. Abbreviations: aIns, anterior insula; aMCC, anterior midcingulate; dlPFC, dorsolateral prefrontal cortex; dmPFC, dorsomedial
prefrontal cortex; dpIns, dorsal posterior insula; fMRI, functional magnetic resonance imaging; NAc, nucleus accumbens; OFC,
orbitofrontal cortex; PAG, periaqueductal gray; PBN, parabrachial nucleus; pTPJ, posterior temporal-parietal junction; RVM (NRM),
rostral ventral medulla (nucleus raphe magnus); S1 and S2, somatosensory regions 1 and 2; vlPFC, ventral lateral prefrontal cortex;
vmPFC, ventromedial prefrontal cortex.

cortex and amygdala activity are less common, but they are consistent across a subset of studies. In
parallel, EEG and MEG studies show placebo-induced reductions in cortical responses to painful
laser stimuli at ∼150–300 ms post-stimulus (Colloca et al. 2008b). These studies demonstrate that
placebo treatments can affect multiple components of pain-related responses, sometimes at a deep
(i.e., early sensory) level.

Brain generators of placebo analgesia. In addition to reductions in pain-related processes,

placebo analgesia studies have also identified consistent increases in brain activity. The most

www.annualreviews.org • Mechanisms of Clinical Placebo 83

 CP13CH04-Wager ARI 13 April 2017 17:10

consistent placebo-related increases are shown in Figure 5; they include engagement of the
dorsolateral and ventrolateral prefrontal cortex (dlPFC/vlPFC), vmPFC, medial orbitofrontal
cortex (OFC), and mid-lateral OFC. Increases in activity in these areas are also correlated with the
magnitude of reported analgesia (Wager & Atlas 2015). A number of studies have also reported
increases in activity in the nucleus accumbens (NAc)/ventral striatum and periaqueductal gray
(PAG)—two areas most closely associated with the opioid system—converging with molecular
imaging studies that identified placebo-induced increases in opioid system activity (Scott et al.
2008, Wager et al. 2007). Many of these regions show anticipatory increases prior to pain, and
these anticipatory increases are some of the strongest predictors of the strength of an individual’s
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

placebo analgesic response (Wager et al. 2011). The engagement of these regions in anticipation
of pain suggests that their role in placebo analgesia may not be pain-specific but rather may be
tied to broader appraisal and expectation processes.

Beyond pain: clinical placebo effects across disorders. Most studies of placebo mechanisms
have studied placebo analgesia in experimentally induced pain in healthy subjects. A smaller liter-
ature has investigated the brain mechanisms of placebo effects on clinical disorders, most notably
Parkinson’s disease and depression. Results from these investigations converge with findings from
the experimental placebo analgesia literature and also suggest the involvement of disorder-specific
brain mechanisms.
A line of research on Parkinson’s disease has demonstrated placebo effects on three systems:
(a) the mesolimbic dopaminergic pathway, which projects from the ventral tegmental area to the
ventral striatum and vmPFC; (b) the nigrostriatal dopaminergic pathway, which projects from the
substantia nigra to the dorsal striatum; and (c) the subthalamic nucleus-thalamocortical pathway
associated with Parkinson’s motor dysfunction. In a landmark early study using radiolabeled raclo-
pride PET imaging, de la Fuente-Fernandez et al. (2001) found enhanced dopamine activity in
the dorsal striatum after patients took a sham medication as compared to a control condition. A
larger follow-up study replicated these effects, but only for patients randomized to receive (false)
instructions that the placebo treatment was 75% likely to be an active drug (Lidstone 2010),
which suggests a key role for appraisals in this response. Another metabolic PET study identified
a pattern of increased brain metabolism in the vmPFC and striatum, among other regions, that
both predicted and correlated with clinical improvement following double-blind sham surgery in
Parkinson’s patients (Ko et al. 2014). In addition, an fMRI study of learning-related brain function
in the mesolimbic dopamine pathway found that placebo medication enhanced performance in a
reward (but not punishment) learning task and altered learning-related activity in the NAc/ventral
striatum and the vmPFC (Schmidt et al. 2014). Taken together, these studies demonstrate ro-
bust placebo responses in the mesolimbic and nigrostriatal dopaminergic pathways in Parkinson’s
patients.
Another paradigm for studying the placebo effect in Parkinson’s disease has used sham stim-
ulation of the subthalamic nucleus (STN), a brain structure implicated in the pathophysiology
of Parkinson’s motor dysfunction. Sham STN stimulation compared with no treatment resulted
in improved motor function and reduced neural firing in the STN (Benedetti et al. 2004). In a
follow-up study, this effect was shown to depend on prior learning: Both thalamic neuronal and
clinical responses to placebo treatment increased as patients were administered a greater number
of active drug conditioning trials prior to the placebo treatment (Benedetti et al. 2016).
In major depression, the placebo response has been well-documented (Fournier et al. 2010,
Kirsch et al. 2008), and studies investigating its brain bases have implicated prefrontal and striatal
regions as well as the opioid system. In an innovative study, Peciña et al. (2015) imaged µ-opioid
activity in depressed patients during administration of an intravenous placebo antidepressant.

84 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Placebo treatment increased µ-opioid neurotransmission in the vmPFC and NAc, among other
regions. These increases predicted improvement in depressive symptoms following both a 1-week
treatment with a pill placebo and a later 10-week trial of an antidepressant medication. These
findings connect the acute placebo brain response (to the intravenous placebo), sustained clinical
improvement to sham antidepressant pills, and the opioidergic system. They also converge with
earlier findings linking pill placebo responses with increased activity in prefrontal and posterior
cingulate cortices, as measured with fMRI (Mayberg 2002) and EEG (Leuchter et al. 2002).
A growing literature has investigated placebo responses in several chronic pain disorders. In IBS
patients undergoing fMRI scanning during painful rectal distension, placebo treatments have been
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

shown to reduce activity in multiple nociceptive areas and to increase activity in the midfrontal
gyrus (MFG), the superior temporal lobe, and posteromedial cortex (Craggs et al. 2008, 2014;
Lieberman et al. 2004). In patients with chronic knee osteoarthritis, altered functional connectivity
between the right MFG, the perigenual ACC, and the posterior cingulate (PCC) and the rest
of the brain predicted placebo treatment response in two independent cohorts (Tétreault et al.
2016). Similarly, mPFC–insula functional connectivity predicted placebo treatment response in a
randomized controlled trial of patients with chronic back pain (Hashmi et al. 2012).
The brain bases of placebo effects have also been investigated in patients with substance use
disorders. In active cocaine abusers undergoing raclopride PET imaging, methylphenidate and
placebo administration lead to statistically indistinguishable levels of dopamine release in stri-
atal regions, demonstrating a drug-mimicking effect of placebo on striatal dopaminergic systems
(Volkow et al. 2011). Similarly, methylphenidate administered to cocaine abusers was also found
to elicit a ∼50% larger increase in thalamic and cerebellar brain metabolism and a ∼50% increase
in mood when subjects believed they were receiving the drug versus when they believed they were
receiving a placebo (Volkow et al. 2003). Parallel findings have been observed in cigarette smokers,
who reported significantly reduced craving and a significant correlation between insular activity
and craving after smoking a cigarette that they believed contained nicotine, but no such changes
when they were told that the cigarette did not contain nicotine (Gu et al. 2016). Belief that a
cigarette did not contain nicotine, as compared with believing that it did, also strongly diminished
brain responses in the striatum related to value and reward prediction errors during a learning
task (Gu et al. 2015).
In summary, findings in clinical disorders converge with those from placebo analgesia in acute
pain and point to disorder-specific placebo mechanisms. They highlight the involvement of medial
prefrontal, posteromedial, and temporal cortex in the genesis of placebo responses, including a role
for dopaminergic—and perhaps opioidergic—pathways. At the same time, these findings suggest
that elements of the placebo response are localized in disorder-specific brain systems, such as the
subthalamic nucleus for Parkinson’s disease.

A Brain System for Affective Appraisal

The brain regions engaged during placebo analgesia (Figure 5) are encompassed by the de-
fault mode network, which includes the dorsal medial prefrontal cortex (dmPFC) and vmPFC,
PCC, temporal-parietal junction (TPJ), and superior temporal sulcus (STS), among other regions
(Figure 6). The default mode network supports a broad category of processes related to affective
appraisal. It is associated with spontaneous thought (Mason et al. 2007), autobiographical memory
retrieval (Vincent et al. 2006), prospection (Buckner & Carroll 2007), generating negative and
positive emotion (Lindquist et al. 2012, 2016; Wager et al. 2015), assessing others’ mental states
(Frith & Frith 2006), self-related processing (Gusnard et al. 2001), representing expected value
(Hare et al. 2008), and more. A central mechanism of placebo treatments is their engagement of

www.annualreviews.org • Mechanisms of Clinical Placebo 85

 CP13CH04-Wager ARI 13 April 2017 17:10

Default mode network Self-generated positive emotion

Placebo increases Self-generated negative emotion

Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

Common areas: core affective appraisal

Self-focused cognition
Value and reward

Social cognition

Figure 6
Appraisal-related processes converge in the default network. Meta-analyses converge on a core appraisal system. (Clockwise from top)
The default mode system (red ) is based on a parcellation of 1,000 resting-state connectivity scans (Yeo et al. 2011) (subcortical regions
are not included in this map). Meta-analyses depicted along the perimeter show activity during recall and imagery techniques for
self-generating positive ( yellow; 21 study maps) or negative (blue; 56 study maps) emotional states (data from Linquist et al. 2012);
( purple) value-related activity from 375 studies of “reward” and “value” from http://Neurosynth.org (Yarkoni et al. 2011); ( green)
social, other-focused cognition across 48 studies (Denny et al. 2012); (orange) self-focused cognition engaged by self-referential
judgments across 48 studies (Denny et al. 2012).

appraisal processes in default mode network regions. Developing an integrative understanding of

the role and function of the default network will be critical for advancing understanding of placebo
brain mechanisms. In this section, we describe key appraisal-related functions of the default mode
network and their relation to placebo effects.

Self-generated emotion. Key regions of the default mode network are prominently activated
when participants are asked to self-generate both negative and positive emotional responses by
recalling or simulating (imagining) events and situations (Figure 6; yellow and blue maps, data

86 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

from Lindquist et al. 2012). Unlike the majority of emotion studies that examine brain responses
elicited by affective stimuli, these activations occur in the absence of any external stimulation and
are generated purely by participants’ thoughts. Self-generated negative emotions are associated
with activation in the vmPFC, dmPFC, and PCC along with a wide swath of limbic regions,
including the amygdala, insula, striatum, and PAG. The medial prefrontal activity overlaps with
activity related to instructed fear, in which anxiety is generated by conceptual knowledge about
associations between cues and shocks, without reinforcement (Mechias et al. 2010). Self-generated
positive emotion consistently activates an overlapping, but more restricted, set of regions, includ-
ing all the major elements of the mesolimbic dopamine system—the vmPFC, striatum, and ventral
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

tegmental area. Placebo treatments likely elicit patients’ memories of previous treatment expe-
riences to influence clinical outcomes and placebo-induced brain responses (Kessner et al. 2013,
2014).

Inferences about others. Another key function of the default network is social cognition—in
particular, the ability to infer the intentions, beliefs, and mental and affective states of others. This
inferential process is known as mentalizing or theory-of-mind and reliably recruits a network of
regions described as the social brain (Blakemore 2008) that include the dmPFC, PCC, STS, and
TPJ (Amodio & Frith 2006, Frith & Frith 2006, Van Overwalle 2009)—all of which are included
in the default mode network and overlap with systems implicated in emotional appraisal (Etkin
et al. 2011). These regions mature in late adolescence (Blakemore 2008) and are important for
inferring the preferences of another individual (Mitchell et al. 2006) or the intensity of another’s
affective experience (Krishnan et al. 2016, Morelli et al. 2015). They are preferentially engaged
by observing social interactions (Wagner et al. 2016) and by processing social information. For
example, when participants made judgments about how their friends ranked on a particular trait,
activity in this network increased as participants were asked to rank a greater number of friends,
suggesting a role in integrating increasing amounts of social information (Meyer et al. 2012).
One recent social cognition study manipulated whether participants felt understood or not
understood after sharing personal experiences. Feeling understood versus not understood activated
different components of the default mode network, including the dmPFC and the precuneus
(Morelli et al. 2014). Feeling understood and cared for by a provider is thought to be a central
component of the placebo effect (Colloca & Miller 2011, Frank & Frank 1993, Wager & Atlas
2015), and three clinical trials have found that patients randomized to more supportive versus less
supportive physician interactions experience superior health outcomes in the cases of the common
cold (Rakel et al. 2011), IBS (Kaptchuk et al. 2008), and chronic back pain (Fuentes et al. 2014).

Self-focused cognition and self-concept. Other-focused and self-focused cognition engages

broadly overlapping brain systems (Buckner & Carroll 2007). However, whereas thinking about
others preferentially engages the dmPFC, self-referential processing preferentially engages the
vmPFC, particularly the pregenual cingulate and anterior medial prefrontal cortex (Denny et al.
2012). This includes tasks as simple as judging the degree to which words describe oneself (Kelley
et al. 2002), and it extends to self-evaluations across a variety of domains, including preferences,
personality, mental states, and physical attributes ( Jenkins & Mitchell 2011, Kelley et al. 2002,
Mitchell et al. 2006). Self-referential processing is also involved in judgments of others, but par-
ticularly to the degree that others are seen as similar to or close to oneself ( Jenkins et al. 2008,
Mitchell et al. 2006).
Self-focused cognition revolves around a concept of the self. Notions of what it means to have
a self-concept are evolving, but one idea is that an implicit or explicit representation of the self
serves as a reference point for valuing other concepts and relating them to one’s goals and values.

www.annualreviews.org • Mechanisms of Clinical Placebo 87

 CP13CH04-Wager ARI 13 April 2017 17:10

Thus, default mode regions are also involved in autobiographical thought (Andrews-Hanna et al.
2014), imagining potential future situations the self may encounter (Buckner & Carroll 2007), and
self-relevance biases in memory encoding (Kelley et al. 2002, Rogers et al. 1977)—all of which
involve positioning the self in a context.

Value. Value is an abstract concept that describes the worth of an item or outcome. Value is
typically operationalized as the amount of resources or effort that an agent would spend to obtain
the outcome. At its heart, however, value is an appraisal of the gain or cost (economic, social,
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

or physical) to current and future well-being, made in reference to the self and in consideration
of one’s goals. Studies have consistently observed that the vmPFC and NAc/ventral striatum are
associated with subjective value (Bartra et al. 2013, Hare et al. 2008, Padoa-Schioppa 2011). These
regions are among the cortical areas most richly innervated by both dopamine and opioids, which
are key players in emotion, motivation, and hedonic pleasure (Berridge & Kringelbach 2008).
Though often considered in terms of reward, responses in these regions and in the dopamine
system more generally show many hallmarks of encoding conceptual appraisals. Neurons in the
vmPFC–lateral OFC group code for anticipated reward (Tremblay & Schultz 1999) and antici-
pated punishment (Morrison et al. 2011), with separate populations of dopamine neurons related
to each (Matsumoto & Hikosaka 2009). They also code for relative value among rewarding op-
tions, rather than physical reward properties of a given stimulus (Tremblay & Schultz 1999), and
they change rapidly with learning as reward contingencies change (Kim & Hikosaka 2013). Im-
portantly, these value representations in vmPFC and ventral striatum are computed with respect
to higher-order goal states and can instantly shift from repulsion to pleasure based on shifts in
internal homeostatic states. In one study, rats initially repulsed by an intensely salty liquid were
highly motivated to obtain the liquid when in a sodium-deprived state, without any additional
learning. Activity markers in the mesolimbic dopamine system, including ventral tegmental area,
NAc, and OFC, were associated with this change (Robinson & Berridge 2013).
Converging evidence comes from lesion studies. Lesions to the vmPFC-OFC do not ap-
pear to impair basic value preferences (Izquierdo et al. 2004), emotional responses (Rudebeck
et al. 2013), or simple forms of value learning (Milad & Quirk 2002) (for a review, see Stalnaker
et al. 2015), but instead disrupt the ability to make value-guided choices in the context of an
animal’s current goals and homeostatic states (Roy et al. 2012, Rudebeck & Murray 2014).
Correspondingly, human fMRI activity in this system appears to reflect a form of expected
affective value related both to pursuit of reward (Chib et al. 2009) and to avoidance of punishment
(Roy et al. 2012). Value-related vmPFC activity is sensitive to diverse forms of conceptual infor-
mation, including personal goals (Hare et al. 2008), homeostatic motivational states (Gottfried
et al. 2003), and verbal suggestions about how others value items (e.g., “this is expensive wine”)
(Plassmann & Wager 2014), which are closely related to placebo effects. Perceptions of value are
known to modulate the effectiveness of placebo treatments, with more expensive placebos exerting
greater analgesic effects and recruiting increased medial prefrontal cortex activity relative to less
expensive placebos (Geuter et al. 2013).

An integrated view of appraisal. These findings suggest that a range of appraisal processes—
which are crucial for most placebo effects—engage a common core brain system (Figure 6). This
appraisal system is adapted for representing schemas or situations, including representations of
one’s goals and well-being in the context of events and stimuli. This system is highly integrative,
involving brain systems supporting memory, prospection, social cognition, emotion, interocep-
tion, and autonomic and neuroendocrine control. Brain networks important for each of these

88 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

domains are partly differentiable but involve points of convergence, particularly in the vmPFC,
PCC, and inferior TPJ, all of which are part of the default mode network (Figure 6).
The vmPFC, in particular, may be a critical hub in the appraisal process (Roy et al. 2012). It is
anatomically and functionally connected to (a) portions of the ventral striatum and lateral OFC
that encode the value of rewarding and aversive events (Pauli et al. 2016, Price 1999, Wallis 2007),
(b) portions of the hippocampus and parahippocampal cortex (Kahn et al. 2008) that participate in
episodic and semantic memory (Binder et al. 2009), and (c) specific portions of the hypothalamus
and PAG (Keay & Bandler 2001, Price 1999) that are central to emotion and the governance
of physiological responses. Converging evidence also suggests that the vmPFC is critical for
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

representing structured, conceptual relationships (Binder et al. 2009, Constantinescu et al. 2016,
Doeller et al. 2010). For example, the vmPFC is centrally involved in semantics (Binder et al. 2009).
Semantic meaning is flexible and context dependent, such that the meaning of “boxer” can shift
depending on the context (i.e., “fighting” versus “poodle”). The vmPFC is also centrally involved
in self-referential cognition (Denny et al. 2012, Jenkins et al. 2008, Kelley et al. 2002, Mitchell
et al. 2006) and may encode “abstract value signals” (Wallis 2007, p. 46) involving “predictions
about specific outcomes associated with stimuli, choices, and actions. . .based on current internal
states” (Rudebeck & Murray 2014, p. 1143). Thus, one view of the appraisal system’s underlying
function is that it allows the mental construction of a conceptual space (Constantinescu et al.
2016), positioning one’s concept of self in relation to valued situations and events. This enables
projections about future events and alternative courses of action by imagining their impact on our
overall well-being.

CONCLUSIONS AND FUTURE DIRECTIONS: HOW DO

APPRAISALS HEAL?
Placebo treatments can have large and sustained effects on clinical outcomes in multiple disor-
ders, particularly those in which emotion and motivation play a central role. Understanding how
these effects arise and are maintained is an open challenge. We believe that interactions between
associative learning systems and appraisals play a central role. Learning can occur in many neural
circuits, but appraisal appears to be supported by a specialized system—a collection of midline
cortical and temporoparietal regions associated with the so-called default mode network. This
network is involved in emotion generation, social and self-referential cognition, and value-based
learning and decision making, pointing to a common core function of flexible, conceptual, and af-
fective thought. This system allows us to simulate potential outcomes and to develop expectations
about future events. It also allows us to relate those events to a representation of the self, including
our broader goals and overall well-being. This system is engaged by placebo treatments for pain,
mood disorders, Parkinson’s disease, and other conditions. Taken together, these findings suggest
that placebos work by engaging systems in the default mode network that govern how a person
conceptualizes their future well-being and the personal significance of their symptoms.
Yet, it remains unclear how appraisals create sustained, long-term improvements in health.
We offer here four ideas that address this question. First, conceptual representations in the
vmPFC and associated appraisal system strongly influence goal-directed decision making—what
to eat, when to sleep, whether to exercise—which can have profound influences on health over
time. Second, the appraisal system is important for assigning affective value and significance
to thoughts. Dysregulation of this system is prominently related to depressive rumination,
catastrophizing, and posttraumatic stress disorder (Etkin & Wager 2007, Kaiser et al. 2015).
Third, when imagined or perceived events are conceived of as close to the self, they create
strong, organism-wide emotional responses. These responses directly influence physiological

www.annualreviews.org • Mechanisms of Clinical Placebo 89

 CP13CH04-Wager ARI 13 April 2017 17:10

processes, including changes in autonomic output and hormone release, which are relevant for
both behavior and physical health and may directly influence disease pathophysiology.
Fourth, appraisals and associative learning systems may form positive feedback loops. The
pain literature reviewed here and elsewhere (Büchel et al. 2014) shows that positive appraisals
influence how symptoms are perceived. The more positive initial expectations are, the less
pain is perceived—which then reinforces the initial expectation of low pain. Put simply, ap-
praisals can become self-fulfilling prophecies, by virtue of their direct influences on mood, phys-
iology, and behavior. In the context of positive feedback loops, over time positive appraisals
may become more automatic and ingrained, transitioning from conceptual appraisal systems
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

to circuits encoding learned, precognitive associations (Schafer et al. 2015). Positive feedback
loops may be one reason that cognitive- and emotion-focused therapies work to change partic-
ipants’ conceptions of themselves and their situations, and a reason that placebo treatments—
which are injections of ideas into the course of a treatment—can have long-lasting therapeutic
effects.

SUMMARY POINTS
1. Placebo treatments can have large, clinically relevant therapeutic effects on pain, mood
disorders, and Parkinson’s disease.
2. Placebo effects are mediated by multiple mechanisms. Two main mechanisms are:
a. Precognitive associations—relatively stable, stimulus-response associations that can
be learned by diverse brain circuits, and
b. Appraisals—cognitive evaluations of situations integrating multiple kinds of infor-
mation in a cohesive, constructed conceptualization with personal meaning.
3. Appraisals are supported by a core brain system associated with the default mode network
that includes the ventromedial prefrontal cortex. This system is instrumental in forming
conceptual expectations and beliefs, self-generating emotion, representing knowledge
about oneself and others, and integrating information into calculations of anticipated
value. This appraisal system is reliably engaged by placebo treatments.

FUTURE ISSUES
1. Many studies have investigated the biological mechanisms underlying immediate, short-
term placebo effects on health, physiology, and performance. What mechanisms underlie
sustained, long-term placebo effects? Why don’t placebo effects naturally extinguish?
2. How do placebo mechanisms relate across diverse disorders and outcomes?
3. How do the mechanisms of placebo treatments relate to the mechanisms of other psy-
chosocial treatments (e.g., psychotherapy) that explicitly aim to initiate changes in pa-
tients’ appraisals?
4. How do patient characteristics and treatment contexts interact to create appraisals sup-
porting or obstructing treatment?

90 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS
We thank Mark Olanow and Wayne Jonas for providing data for Figure 1 and Figure 3b,
respectively, and Jennifer Winer for help with a literature review. T.D.W. is funded by
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

NIH grants R01MH076136 and R01DA035484. Matlab code for analyses is available at
https://github.com/canlab.

LITERATURE CITED
Amanzio M, Benedetti F, Porro CA, Palermo S, Cauda F. 2013. Activation likelihood estimation meta-analysis
of brain correlates of placebo analgesia in human experimental pain. Hum. Brain Mapp. 34:738–52
Amodio DM, Frith CD. 2006. Meeting of minds: the medial frontal cortex and social cognition. Nat. Rev.
Neurosci. 7:268–77
Andrews-Hanna JR, Smallwood J, Spreng RN. 2014. The default network and self-generated thought: com-
ponent processes, dynamic control, and clinical relevance. Ann. N. Y. Acad. Sci. 1316:29–52
Atlas LY, Wager TD. 2014. A meta-analysis of brain mechanisms of placebo analgesia: consistent findings
and unanswered questions. In Placebo, ed. by F Benedetti, pp. 37–69. Berlin: Springer
Atlas LY, Whittington RA, Lindquist MA, Wielgosz J, Sonty N, Wager TD. 2012. Dissociable influences of
opiates and expectations on pain. J. Neurosci. 32:8053–64
Bandelow B, Reitt M, Röver C, Michaelis S, Görlich Y, Wedekind D. 2015. Efficacy of treatments for anxiety
disorders: a meta-analysis. Int. Clin. Psychopharmacol. 30:183–89
Bannuru RR, McAlindon TE, Sullivan MC, Wong JB, Kent DM, Schmid CH. 2015. Effectiveness and impli-
cations of alternative placebo treatments: a systematic review and network meta-analysis of osteoarthritis
trials. Ann. Int. Med. 163:365–72
Barrett LF. 2012. Emotions are real. Emotion 12:413–29
Barrett LF. 2014. The conceptual act theory: a précis. Emot. Rev. 6:292–97
Bartra O, McGuire JT, Kable JW. 2013. The valuation system: a coordinate-based meta-analysis of BOLD
fMRI experiments examining neural correlates of subjective value. NeuroImage 76:412–27
Benedetti F. 2014. Placebo effects: from the neurobiological paradigm to translational implications. Neuron
84:623–37
Benedetti F, Amanzio M, Vighetti S, Asteggiano G. 2006. The biochemical and neuroendocrine bases of the
hyperalgesic nocebo effect. J. Neurosci. 26:12014–22
Benedetti F, Colloca L, Torre E, Lanotte M, Melcarne A, et al. 2004. Placebo-responsive Parkinson patients
show decreased activity in single neurons of subthalamic nucleus. Nat. Neurosci. 7:587–88
Benedetti F, Frisaldi E, Carlino E, Giudetti L, Pampallona A, et al. 2016. Teaching neurons to respond to
placebos. J. Physiol. 594:5647–60
Benedetti F, Maggi G, Lopiano L, Lanotte M. 2003a. Open versus hidden medical treatments: The patient’s
knowledge about a therapy affects the therapy outcome. Prev. Treat. 6. https://dx.doi.org/10.1037/1522-
3736.6.1.61a
Benedetti F, Pollo A, Lopiano L, Lanotte M, Vighetti S, Rainero I. 2003b. Conscious expectation and uncon-
scious conditioning in analgesic, motor, and hormonal placebo/nocebo responses. J. Neurosci. 23:4315–23
Berntson GG, Micco DJ. 1976. Organization of brainstem behavioral systems. Brain Res. Bull. 1:471–83
Berntson GG, Tuber DS, Ronca AE, Bachman DS. 1983. The decerebrate human: associative learning. Exp.
Neurol. 81:77–88
Berridge KC, Kringelbach ML. 2008. Affective neuroscience of pleasure: reward in humans and animals.
Psychopharmacology 199:457–80

www.annualreviews.org • Mechanisms of Clinical Placebo 91

 CP13CH04-Wager ARI 13 April 2017 17:10

Binder JR, Desai RH, Graves WW, Conant LL. 2009. Where is the semantic system? A critical review and
meta-analysis of 120 functional neuroimaging studies. Cereb. Cortex 19:2767–96
Blakemore S-J. 2008. The social brain in adolescence. Nat. Rev. Neurosci. 9:267–77
Büchel C, Geuter S, Sprenger C, Eippert F. 2014. Placebo analgesia: a predictive coding perspective. Neuron
81:1223–39
Buckner RL, Carroll DC. 2007. Self-projection and the brain. Trends Cogn. Sci. 11:49–57
Carew TJ, Hawkins RD, Kandel ER. 1983. Differential classical conditioning of a defensive withdrawal reflex
in Aplysia californica. Science 219:397–400
Carlino E, Torta D, Piedimonte A, Frisaldi E, Vighetti S, Benedetti F. 2014. Role of explicit verbal information
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

in conditioned analgesia. Eur. J. Pain 19:546–53

Chib VS, Rangel A, Shimojo S, O’Doherty JP. 2009. Evidence for a common representation of decision values
for dissimilar goods in human ventromedial prefrontal cortex. J. Neurosci. 29:12315–20
Colloca L, Lopiano L, Lanotte M, Benedetti F. 2004. Overt versus covert treatment for pain, anxiety, and
Parkinson’s disease. Lancet Neurol. 3:679–84
Colloca L, Miller FG. 2011. How placebo responses are formed: a learning perspective. Philos. Trans. R. Soc.
Lond. B Biol. Sci. 366:1859–69
Colloca L, Petrovic P, Wager TD, Ingvar M, Benedetti F. 2010. How the number of learning trials affects
placebo and nocebo responses. Pain 151:430–39
Colloca L, Sigaudo M, Benedetti F. 2008a. The role of learning in nocebo and placebo effects. Pain 136:211–18
Colloca L, Tinazzi M, Recchia S, Le Pera D, Fiaschi A, et al. 2008b. Learning potentiates neurophysiological
and behavioral placebo analgesic responses. Pain 139:306–14
Constantinescu AO, O’Reilly JX, Behrens TE. 2016. Organizing conceptual knowledge in humans with a
gridlike code. Science 352:1464–68
Craggs JG, Price DD, Perlstein WM, Verne NG, Robinson ME. 2008. The dynamic mechanisms of placebo
induced analgesia: evidence of sustained and transient regional involvement. Pain 139(3):660–69
Craggs JG, Price DD, Robinson ME. 2014. Enhancing the placebo response: functional magnetic resonance
imaging evidence of memory and semantic processing in placebo analgesia. J. Pain 15(4):435–46
Crum AJ, Corbin WR, Brownell KD, Salovey P. 2011. Mind over milkshakes: mindsets, not just nutrients,
determine ghrelin response. Health Psychol. 30:424–29; discussion 430–31
Cuijpers P, Driessen E, Hollon SD, van Oppen P, Barth J, Andersson G. 2012. The efficacy of non-directive
supportive therapy for adult depression: a meta-analysis. Clin. Psychol. Rev. 32:280–91
de Jong PJ, van Baast R, Arntz A, Merckelbach H. 1996. The placebo effect in pain reduction: the influence
of conditioning experiences and response expectancies. Int. J. Behav. Med. 3:14–29
de la Fuente-Fernández R, Ruth TJ, Sossi V, Schulzer M, Calne DB, Stoessl J. 2001. Expectation and dopamine
release: mechanism of the placebo effect in Parkinson’s disease. Science 293:1164–66
Denny BT, Kober H, Wager TD, Ochsner KN. 2012. A meta-analysis of functional neuroimaging studies of
self- and other judgments reveals a spatial gradient for mentalizing in medial prefrontal cortex. J. Cogn.
Neurosci. 24:1742–52
Doeller CF, Barry C, Burgess N. 2010. Evidence for grid cells in a human memory network. Nature 463:657–61
Eippert F, Bingel U, Schoell ED, Yacubian J, Klinger R, et al. 2009a. Activation of the opioidergic descending
pain control system underlies placebo analgesia. Neuron 63:533–43
Eippert F, Finsterbusch J, Bingel U, Buchel C. 2009b. Direct evidence for spinal cord involvement in placebo
analgesia. Science 326:404
Ellsworth PC, Scherer KR. 2003. Appraisal processes in emotion. In Handbook of Affective Sciences, ed. RJ
Davidson, KR Scherer, HH Goldsmith, pp. 572–95. Oxford, UK: Oxford Univ. Press
Enck P, Benedetti F, Schedlowski M. 2008. New insights into the placebo and nocebo responses. Neuron
59:195–206
Etkin A, Egner T, Kalisch R. 2011. Emotional processing in anterior cingulate and medial prefrontal cortex.
Trends Cogn. Sci. 15:85–93
Etkin A, Wager TD. 2007. Functional neuroimaging of anxiety: a meta-analysis of emotional processing in
PTSD, social anxiety disorder, and specific phobia. Am. J. Psychiatry 164:1476–88
Fässler M, Meissner K, Kleijnen J, Hróbjartsson A, Linde K. 2015. A systematic review found no consistent
difference in effect between more and less intensive placebo interventions. J. Clin. Epidemiol. 68:442–51

92 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Finniss DG, Kaptchuk TJ, Miller F, Benedetti F. 2010. Biological, clinical, and ethical advances of placebo
effects. Lancet 375:686–95
Fournier JC, DeRubeis RJ, Hollon SD, Dimidjian S, Amsterdam JD, et al. 2010. Antidepressant drug effects
and depression severity: a patient-level meta-analysis. JAMA 303:47–53
Frank JD, Frank JB. 1993. Persuasion and Healing: A Comparative Study of Psychotherapy. Baltimore, MD: Johns
Hopkins Univ. Press
Freeman EW, Ensrud KE, Larson JC, Guthrie KA, Carpenter JS, et al. 2015. Placebo improvement in
pharmacologic treatment of menopausal hot flashes. Psychosom. Med. 77(2):167–75
Frith CD, Frith U. 2006. The neural basis of mentalizing. Neuron 50:531–34
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

Fuentes J, Armijo-Olivo S, Funabashi M, Miciak M, Dick B, et al. 2014. Enhanced therapeutic alliance
modulates pain intensity and muscle pain sensitivity in patients with chronic low back pain: an experimental
controlled study. Phys. Ther. 94(4):477–89
Gaston L, Marmar CR, Gallagher D, Thompson LW. 1989. Impact of confirming patient expectations of
change processes in behavioral, cognitive, and brief dynamic psychotherapy. Psychotherapy 26:296–302
Geuter S, Eippert F, Hindi Attar C, Büchel C. 2013. Cortical and subcortical responses to high and low
effective placebo treatments. NeuroImage 67:227–36
Goebel MU, Trebst AE, Steiner J, Xie YF, Exton MS, et al. 2002. Behavioral conditioning of immunosup-
pression is possible in humans. FASEB J. 16:1869–73
Goetz CG, Wuu J, McDermott MP, Adler CH, Fahn S, et al. 2008. Placebo response in Parkinson’s disease:
comparisons among 11 trials covering medical and surgical interventions. Mov. Disord. 23:690–99
Gottfried JA, O’Doherty J, Dolan RJ. 2003. Encoding predictive reward value in human amygdala and or-
bitofrontal cortex. Science 301:1104–7
Grau JW. 2014. Learning from the spinal cord: how the study of spinal cord plasticity informs our view of
learning. Neurobiol. Learn. Mem. 108:155–71
Gu X, Lohrenz T, Salas R, Baldwin PR, Soltani A, et al. 2015. Belief about nicotine selectively modulates
value and reward prediction error signals in smokers. PNAS 112(8):2539–44
Gu X, Lohrenz T, Salas R, Baldwin PR, Soltani A, et al. 2016. Belief about nicotine modulates subjec-
tive craving and insula activity in deprived smokers. Front. Psychiatry 7:126. https://doi.org/10.3389/
fpsyt.2016.00126
Guo JY, Yuan XY, Sui F, Zhang WC, Wang JY, et al. 2011. Placebo analgesia affects the behavioral despair
tests and hormonal secretions in mice. Psychopharmacology 217:83–90
Gusnard D, Akbudak E, Shulman GL, Raichle ME. 2001. Medial prefrontal cortex and self-referential mental
activity: relation to a default mode of brain function. PNAS 98:4259–64
Gusnard DA, Raichle ME. 2001. Searching for a baseline: functional imaging and the resting human brain.
Nat. Rev. Neurosci. 2:685–94
Hardy GE, Barkham M, Shapiro DA, Reynolds S, Rees A, Stiles WB. 1995. Credibility and outcome of
cognitive-behavioural and psychodynamic-interpersonal psychotherapy. Br. J. Clin. Psychol. 34:555–69
Hare T, O’Doherty J, Camerer C, Schultz W, Rangel A. 2008. Dissociating the role of the orbitofrontal cortex
and the striatum in the computation of goal values and prediction errors. J. Neurosci. 28:5623–30
Hashmi JA, Baria AT, Baliki MN, Huang L, Schnitzer TJ, Apkarian AV. 2012. Brain networks predicting
placebo analgesia in a clinical trial for chronic back pain. Pain 153:2393–402
Herrnstein RJ. 1962. Placebo effect in the rat. Science 138:677–78
Hróbjartsson A, Gøtzsche PC. 2010. Placebo interventions for all clinical conditions. Cochrane Database Syst.
Rev. 20:CD003974
Izquierdo A, Suda RK, Murray EA. 2004. Bilateral orbital prefrontal cortex lesions in rhesus monkeys disrupt
choices guided by both reward value and reward contingency. J. Neurosci. 24:7540–48
Jenkins AC, Macrae CN, Mitchell JP. 2008. Repetition suppression of ventromedial prefrontal activity during
judgments of self and others. PNAS 105:4507–12
Jenkins AC, Mitchell JP. 2011. Medial prefrontal cortex subserves diverse forms of self-reflection. Soc. Neurosci.
6:211–18
Jepma M, Wager TD. 2015. Conceptual conditioning: mechanisms mediating conditioning effects on pain.
Psychol. Sci. 26:1728–39

www.annualreviews.org • Mechanisms of Clinical Placebo 93

 CP13CH04-Wager ARI 13 April 2017 17:10

Jonas WB, Crawford C, Colloca L, Kaptchuk TJ, Moseley B, et al. 2015. To what extent are surgery and invasive
procedures effective beyond a placebo response? A systematic review with meta-analysis of randomised,
sham controlled trials. BMJ Open 5:e009655
Joyce AS, Piper WE. 1998. Expectancy, the therapeutic alliance, and treatment outcome in short-term indi-
vidual psychotherapy. J. Psychother. Pract. Res. 7:236–48
Kahn I, Andrews-Hanna JR, Vincent JL, Snyder AZ, Buckner RL. 2008. Distinct cortical anatomy linked
to subregions of the medial temporal lobe revealed by intrinsic functional connectivity. J. Neurophysiol.
100:129–39
Kaiser RH, Andrews-Hanna JR, Wager TD, Pizzagalli DA. 2015. Large-scale network dysfunction in major
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

depressive disorder: a meta-analysis of resting-state functional connectivity. JAMA Psychiatry 72:603–11

Kam-Hansen S, Jakubowski M, Kelley JM, Kirsch I, Hoaglin DC, et al. 2014. Altered placebo and drug
labeling changes the outcome of episodic migraine attacks. Sci. Transl. Med. 6:218ra5
Kaptchuk TJ, Friedlander E, Kelley JM, Sanchez MN, Kokkotou E, et al. 2010. Placebos without deception:
a randomized controlled trial in irritable bowel syndrome. PLOS ONE 5:e15591
Kaptchuk TJ, Kelley JM, Conboy LA, Davis RB, Kerr CE, et al. 2008. Components of placebo effect: ran-
domised controlled trial in patients with irritable bowel syndrome. BMJ 336:999–1003
Kazdin AE, Krouse R. 1983. The impact of variations in treatment rationales on expectancies for therapeutic
change. Behav. Ther. 14:657–71
Keay K, Bandler R. 2001. Parallel circuits mediating distinct emotional coping reactions to different types of
stress. Neurosci. Biobehav. Rev. 25:669–78
Kelley WM, Macrae CN, Wyland CL, Caglar S, Inati S, Heatherton TF. 2002. Finding the self? An event-
related fMRI study. J. Cogn. Neurosci. 14:785–94
Kessner S, Forkmann K, Ritter C, Wiech K, Ploner M, Bingel U. 2014. The effect of treatment history on
therapeutic outcome: psychological and neurobiological underpinnings. PLOS ONE 9:e109014
Kessner S, Wiech K, Forkmann K, Ploner M, Bingel U, et al. 2013. The effect of treatment history on
therapeutic outcome: an experimental approach. JAMA Intern. Med. 173:1468–69
Khan A, Faucett J, Lichtenberg P, Kirsch I, Brown WA. 2012. A systematic review of comparative efficacy of
treatments and controls for depression. PLOS ONE 7:e41778
Khan A, Kolts RL, Rapaport MH, Krishnan KR, Brodhead AE, Brown WA. 2005. Magnitude of placebo
response and drug-placebo differences across psychiatric disorders. Psychol. Med. 35:743–49
Khan A, Redding N, Brown WA. 2008. The persistence of the placebo response in antidepressant clinical
trials. J. Psychiatr. Res. 42:791–96
Kim HF, Hikosaka O. 2013. Distinct basal ganglia circuits controlling behaviors guided by flexible and stable
values. Neuron 79:1001–10
Kirsch I. 1985. Response expectancy as a determinant of experience and behavior. Am. Psychol. 40:1189–202
Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT. 2008. Initial severity and
antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLOS
Med. 5:e45
Ko JH, Feigin A, Mattis PJ, Tang CC, Ma Y, et al. 2014. Network modulation following sham surgery in
Parkinson’s disease. J. Clin. Investig. 124:3656–66
Koban L, Ruzic L, Wager TD. 2013. Brain predictors of individual differences in placebo responding. In
Placebo and Pain, ed. L Colloca, MA Flaten, K Meissner, pp. 89–102. Boston: Acad. Press
Koban L, Wager TD. 2016. Beyond conformity: social influences on pain reports and physiology. Emotion
16:24–32
Kong J, Spaeth R, Cook A, Kirsch I, Claggett B, et al. 2013. Are all placebo effects equal? Placebo pills, sham
acupuncture, cue conditioning and their association. PLOS ONE 8:e6748
Krishnan A, Woo C-W, Chang LJ, Ruzic L, Gu X, et al. 2016. Somatic and vicarious pain are represented by
dissociable multivariate brain patterns. eLife 5:e15166
Lazarus RS, Folkman S. 1984. Stress, Appraisal, and Coping. New York: Springer
Leuchter AF, Cook IA, Witte EA, Morgan M, Abrams M. 2002. Changes in brain function of depressed
subjects during treatment with placebo. Am. J. Psychiatry 159:122–29
Liberman RP. 1967. The elusive placebo reactor. Neuro-Psycho-Pharmacology: Proc. Int. Congr. Coll. Int. Neuro-
Psycho-Pharmacol., 5th, Washington, DC, pp. 557–66. Amsterdam/New York: Excerpta Medica

94 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Lidstone SC. 2010. Effects of expectation on placebo-induced dopamine release in Parkinson disease. Arch.
Gen. Psychiatry 67:857–65
Lieberman MD, Jarcho JM, Berman S, Naliboff BD, Suyenobu BY, et al. 2004. The neural correlates of
placebo effects: a disruption account. NeuroImage 22(1):447–55
Lindquist KA, Satpute AB, Wager TD, Weber J, Barrett LF. 2016. The brain basis of positive and negative
affect: evidence from a meta-analysis of the human neuroimaging literature. Cereb. Cortex 26:1910–22
Lindquist KA, Wager TD, Kober H, Bliss-Moreau E, Feldman Barrett L. 2012. The brain basis of emotion:
a meta-analytic review. Behav. Brain Sci. 35:121–43
Madsen MV, Gotzsche PC, Hrobjartsson A. 2009. Acupuncture treatment for pain: systematic review of
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

randomised clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups. BMJ
338:a3115
Marks WJ, Bartus RT, Siffert J, Davis CS, Lozano A, et al. 2010. Gene delivery of AAV2-neurturin for
Parkinson’s disease: a double-blind, randomised, controlled trial. Lancet Neurol. 9:1164–72
Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, Macrae CN. 2007. Wandering minds: the
default network and stimulus-independent thought. Science 315:393–95
Matsumoto M, Hikosaka O. 2009. Two types of dopamine neuron distinctly convey positive and negative
motivational signals. Nature 459:837–41
Mayberg HS. 2002. The functional neuroanatomy of the placebo effect. Am. J. Psychiatry 159(5):728–37.
http://doi.org/10.1176/appi.ajp.159.5.728
Mechias M-L, Etkin A, Kalisch R. 2010. A meta-analysis of instructed fear studies: implications for conscious
appraisal of threat. NeuroImage 49:1760–68
Meissner K. 2011. The placebo effect and the autonomic nervous system: evidence for an intimate relationship.
Philos. Trans. R. Soc. Lond. B Biol. Sci. 366:1808–17
Meissner K, Distel H, Mitzdorf U. 2007. Evidence for placebo effects on physical but not on biochemical
outcome parameters: a review of clinical trials. BMC Med. 5:3
Meissner K, Fässler M, Rücker G, Kleijnen J, Hróbjartsson A, et al. 2013. Differential effectiveness of placebo
treatments. JAMA Intern. Med. 173:1910–41
Meyer B, Yuen KSL, Ertl M, Polomac N, Mulert C, et al. 2015. Neural mechanisms of placebo anxiolysis. J.
Neurosci. 35:7365–73
Meyer ML, Spunt RP, Berkman ET, Taylor SE, Lieberman MD. 2012. Evidence for social working memory
from a parametric functional MRI study. PNAS 109:1883–88
Milad M, Quirk G. 2002. Neurons in medial prefrontal cortex signal memory for fear extinction. Nature
420:70–74
Mitchell JP, Macrae CN, Banaji MR. 2006. Dissociable medial prefrontal contributions to judgments of similar
and dissimilar others. Neuron 50:655–63
Moerman DE, Jonas WB. 2002. Deconstructing the placebo effect and finding the meaning response. Ann.
Intern. Med. 136:471–76
Montgomery GH, Kirsch I. 1997. Classical conditioning and the placebo effect. Pain 72:107–13
Morelli SA, Sacchet MD, Zaki J. 2015. Common and distinct neural correlates of personal and vicarious
reward: a quantitative meta-analysis. NeuroImage 112:244–53
Morelli SA, Torre JB, Eisenberger NI. 2014. The neural bases of feeling understood and not understood. Soc.
Cogn. Affect. Neurosci. 9:1890–96
Morrison SE, Saez A, Lau B, Salzman CD. 2011. Different time courses for learning-related changes in
amygdala and orbitofrontal cortex. Neuron 71:1127–40
Morton DL, Watson A, El-Deredy W, Jones AKP. 2009. Reproducibility of placebo analgesia: effect of
dispositional optimism. Pain 146:194–98. https://doi.org/10.1016/j.pain.2009.07.026
Moseley JB, O’Malley K, Petersen NJ, Menke TJ, Brody BA, et al. 2002. A controlled trial of arthroscopic
surgery for osteoarthritis of the knee. N. Engl. J. Med. 347:81–88
Müller M, Kamping S, Benrath J, Skowronek H, Schmitz J, Klinger R, Flor H. 2016. Treatment history
and placebo responses to experimental and clinical pain in chronic pain patients. Eur. J. Pain 1–12.
http://doi.org/10.1002/ejp.877
Nakamura Y, Donaldson GW, Kuhn R, Bradshaw DH, Jacobson RC, Chapman CR. 2012. Investigating
dose-dependent effects of placebo analgesia: a psychophysiological approach. Pain 153:227–37

www.annualreviews.org • Mechanisms of Clinical Placebo 95

 CP13CH04-Wager ARI 13 April 2017 17:10

Olanow WC, Bartus RT, Baumann TL, Factor S, Boulis N, et al. 2015. Gene delivery of neurturin to
putamen and substantia nigra in Parkinson disease: a double-blind, randomized, controlled trial. Ann.
Neurol. 78:248–57
Ortony A, Clore GL, Collins A. 1988. The Cognitive Structure of Emotions. New York: Cambridge Univ. Press
Padoa-Schioppa C. 2011. Neurobiology of economic choice: a good-based model. Annu. Rev. Neurosci. 34:333–
59
Papakostas GI, Fava M. 2009. Does the probability of receiving placebo influence clinical trial outcome? A
meta-regression of double-blind, randomized clinical trials in MDD. Eur. Neuropsychopharmacol. 19:34–40
Pauli WM, O’Reilly RC, Yarkoni T, Wager TD. 2016. Regional specialization within the human striatum
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

for diverse psychological functions. PNAS 113:1907–12

Peciña M, Bohnert ASB, Sikora M, Avery ET, Langenecker SA, et al. 2015. Association between placebo-
activated neural systems and antidepressant responses. JAMA Psychiatry 72:1087–94
Plassmann H, Wager TD. 2014. How expectancies shape consumption experiences. In The Interdisciplinary
Science of Consumption, ed. SD Preston, ML Kringelbach, B Knutson, pp. 219–40. Cambridge, MA: MIT
Posternak MA, Zimmerman M, Keitner GI, Miller IW. 2002. A reevaluation of the exclusion criteria
used in antidepressant efficacy trials. Am. J. Psychiatry 159:191–200. https://doi.org/10.1176/appi.ajp.
159.2.191
Pressman A, Avins AL, Neuhaus J, Ackerson L, Rudd P. 2012. Adherence to placebo and mortality in the Beta
Blocker Evaluation of Survival Trial (BEST). Contemp. Clin. Trials 33:492–98
Price DD, Finniss DG, Benedetti F. 2008. A comprehensive review of the placebo effect: recent advances and
current thought. Annu. Rev. Psychol. 59:565–90
Price JL. 1999. Prefrontal cortical networks related to visceral function and mood. Ann. N. Y. Acad. Sci.
877:383–96
Quessy SN, Rowbotham MC. 2008. Placebo response in neuropathic pain trials. Pain 138:479–83
Raichle ME, MacLeod AM, Snyder AZ, Powers WJ, Gusnard DA, Shulman GL. 2001. A default mode of
brain function. PNAS 98:676–82
Rakel D, Barrett B, Zhang Z, Hoeft T, Chewning B, et al. 2011. Perception of empathy in the therapeutic
encounter: effects on the common cold. Patient Educ. Couns. 85(3):390–97
Robinson MJF, Berridge KC. 2013. Instant transformation of learned repulsion into motivational “wanting.”
Curr. Biol. 23:282–89
Rogers TB, Kuiper NA, Kirker WS. 1977. Self-reference and the encoding of personal information. J. Pers.
Soc. Psychol. 35:677–88
Roy M, Shohamy D, Wager TD. 2012. Ventromedial prefrontal-subcortical systems and the generation of
affective meaning. Trends Cogn. Sci. 16:147–56
Rudebeck PH, Murray EA. 2014. The orbitofrontal oracle: cortical mechanisms for the prediction and eval-
uation of specific behavioral outcomes. Neuron 84(6):1143–45
Rudebeck PH, Saunders RC, Prescott AT, Chau LS, Murray EA. 2013. Prefrontal mechanisms of behavioral
flexibility, emotion regulation and value updating. Nat. Neurosci. 16(8):1140–45
Rutherford BR. 2016. Patient expectancy as a mediator of placebo effects in antidepressant clinical trials. Am.
J. Psychiatry 174:135–42
Rutherford BR, Bailey VS, Schneier FR, Pott E, Brown PJ, Roose SP. 2015. Influence of study design on
treatment response in anxiety disorder clinical trials. Depress. Anxiety 32:944–57
Rutherford BR, Pott E, Tandler JM, Wall MM, Roose SP, Lieberman JA. 2014. Placebo response in antipsy-
chotic clinical trials: a meta-analysis. JAMA Psychiatry 10032:1409–21
Schafer SM, Colloca L, Wager TD. 2015. Conditioned placebo analgesia persists when subjects know they
are receiving a placebo. J. Pain 16:412–20
Schedlowski M, Pacheco-Lopez G. 2010. The learned immune response: Pavlov and beyond. Brain Behav.
Immun. 24:176–85
Scherer KR. 2001. Appraisal considered as a process of multilevel sequential checking. In Appraisal Processes
in Emotion: Theory, Methods, Research, ed. KR Scherer, A Schorr, T Johnstone, pp. 92–120. Oxford, UK/
New York: Oxford Univ. Press
Schmidt L, Braun EK, Wager TD, Shohamy D. 2014. Mind matters: placebo enhances reward learning in
Parkinson’s disease. Nat. Neurosci. 17:1793–97

96 Ashar · Chang · Wager

 CP13CH04-Wager ARI 13 April 2017 17:10

Scott DJ, Stohler CS, Egnatuk CM, Wang H, Koeppe RA, Zubieta J-K. 2008. Placebo and nocebo effects are
defined by opposite opioid and dopaminergic responses. Arch. Gen. Psychiatry 65:220–31
Sinyor M, Levitt AJ, Cheung AH, Schaffer A, Kiss A, et al. 2010. Does inclusion of a placebo arm influence
response to active antidepressant treatment in randomized controlled trials? J. Clin. Psychiatry 71:270–79
Smith CA, Ellsworth PC. 1985. Patterns of cognitive appraisal in emotion. J. Pers. Soc. Psychol. 48:813–38
Sneed JR, Rutherford BR, Rindskopf D, Lane DT, Sackeim HA, Roose SP. 2008. Design makes a difference:
a meta-analysis of antidepressant response rates in placebo-controlled versus comparator trials in late-life
depression. Am. J. Geriatr. Psychiatry 16:65–73
Sorokin I, Schatz A, Welliver C. 2015. Placebo medication and sham surgery responses in benign prostatic
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

hyperplasia treatments: implications for clinical trials. Curr. Urol. Rep. 16(10). 16:73. https://doi.org/
10.1007/s11934-015-0544-4
Sotsky SM, Glass DR, Shea MT, Pilkonis PA, Collins JF, et al. 1991. Patient predictors of response to
psychotherapy and pharmacotherapy: findings in the NIMH Treatment of Depression Collaborative
Research Program. Am. J. Psychiatry 148:997–1008
Stalnaker TA, Cooch NK, Schoenbaum G. 2015. What the orbitofrontal cortex does not do. Nat. Neurosci.
18:620–27
Stewart-Williams S, Podd J. 2004. The placebo effect: dissolving the expectancy versus conditioning debate.
Psychol. Bull. 130:324–40
Strunk DR, DeRubeis RJ, Chiu AW, Alvarez J. 2007. Patients’ competence in and performance of cognitive
therapy skills: relation to the reduction of relapse risk following treatment for depression. J. Consult. Clin.
Psychol. 75(4):523–30
Tétreault P, Mansour A, Vachon-Presseau E, Schnitzer TJ, Apkarian AV, Baliki MN. 2016. Brain connectivity
predicts placebo response across chronic pain clinical trials. PLOS Biology 14(10):e1002570
Tremblay L, Schultz W. 1999. Relative reward preference in primate orbitofrontal cortex. Nature 398:704–8
Tuttle AH, Tohyama S, Ramsay T, Kimmelman J, Schweinhardt P, et al. 2015. Increasing placebo responses
over time in U.S. clinical trials of neuropathic pain. Pain 156:2616–26
Van Overwalle F. 2009. Social cognition and the brain: a meta-analysis. Hum. Brain Mapp. 30:829–58
Vase L, Riley JL, Price DD. 2002. A comparison of placebo effects in clinical analgesic trials versus studies of
placebo analgesia. Pain 99:443–52
Vase L, Robinson ME, Verne GN, Price DD. 2005. Increased placebo analgesia over time in irritable bowel
syndrome (IBS) patients is associated with desire and expectation but not endogenous opioid mechanisms.
Pain 115:338–47
Vincent JL, Snyder AZ, Fox MD, Shannon BJ, Andrews JR, et al. 2006. Coherent spontaneous activity identifies
a hippocampal-parietal memory network. J. Neurophysiol. 96:3517–31
Volkow ND, Wang G-J, Ma Y, Fowler JS, Zhu W, et al. 2003. Expectation enhances the regional brain
metabolic and the reinforcing effects of stimulants in cocaine abusers. J. Neurosci. 23:11461–68
Volkow ND, Wang G, Fowler JS, Tomasi D, Telang F, et al. 2011. Addiction: beyond dopamine reward
circuitry. PNAS 108(37):15037–42
Wager TD, Atlas LY. 2015. The neuroscience of placebo effects: connecting context, learning and health.
Nat. Rev. Neurosci. 16:403–18
Wager TD, Atlas LY, Leotti LA, Rilling JK. 2011. Predicting individual differences in placebo analgesia:
contributions of brain activity during anticipation and pain experience. J. Neurosci. 31:439–52
Wager TD, Fields HL. 2013. Placebo analgesia. In Textbook of Pain, ed. S McMahon, M Koltzenburg, I
Tracey, DC Turk, pp. 362–73. Philadelphia: Elsevier Health Sci.
Wager TD, Kang J, Johnson TD, Nichols TE, Satpute AB, Barrett LF. 2015. A Bayesian model of category-
specific emotional brain responses. PLOS Comput. Biol. 11:e1004066
Wager TD, Rilling JK, Smith EE, Sokolik A, Casey KL, et al. 2004. Placebo-induced changes in fMRI in the
anticipation and experience of pain. Science 303:1162–67
Wager TD, Scott DJ, Zubieta J-K. 2007. Placebo effects on human µ-opioid activity during pain. PNAS
104:11056–61
Wagner DD, Kelley WM, Haxby JV, Heatherton TF. 2016. The dorsal medial prefrontal cortex responds
preferentially to social interactions during natural viewing. J. Neurosci. 36:6917–25

www.annualreviews.org • Mechanisms of Clinical Placebo 97

 CP13CH04-Wager ARI 13 April 2017 17:10

Wallis JD. 2007. Orbitofrontal cortex and its contribution to decision-making. Annu. Rev. Neurosci. 30:31–56
Wampold BE, Imel ZE. 2015. The Great Psychotherapy Debate: The Evidence for What Makes Psychotherapy Work.
New York: Routledge. 2nd ed.
Wechsler ME, Kelley JM, Boyd IOE, Dutile S, Marigowda G, et al. 2011. Active albuterol or placebo, sham
acupuncture, or no intervention in asthma. N. Engl. J. Med. 365:119–26
Wendt L, Albring A, Ober K, Engler H, Freundlieb C, et al. 2013. Placebo-induced immunosuppression in
humans: role of learning and expectation. Brain Behav. Immun. 29(Suppl.):S17
Whalley B, Hyland ME, Kirsch I. 2008. Consistency of the placebo effect. J. Psychosom. Res. 64:537–41
Wilson-Mendenhall CD, Barrett LF, Simmons WK, Barsalou LW. 2011. Grounding emotion in situated
Downloaded from www.annualreviews.org. Guest (guest) IP: 31.205.199.39 On: Tue, 25 Feb 2025 20:39:15

conceptualization. Neuropsychologia 49:1105–27

Woods SC, Ramsay DS. 2000. Pavlovian influences over food and drug intake. Behav. Brain Res. 110:175–82
Woods SW, Gueorguieva RV, Baker CB, Makuch RW. 2005. Control group bias in randomized atypical
antipsychotic medication trials for schizophrenia. Arch. Gen. Psychiatry 62:961–70
Yarkoni T, Poldrack RA, Nichols TE, Van Essen DC, Wager TD. 2011. Large-scale automated synthesis of
human functional neuroimaging data. Nat. Methods 8:665–70
Yeo BTT, Krienen FM, Sepulcre J, Sabuncu MR, Lashkari D, et al. 2011. The organization of the human
cerebral cortex estimated by intrinsic functional connectivity. J. Neurophysiol. 106:1125–65

98 Ashar · Chang · Wager