ASSOCIATE EDITOR: MARTIN C. MICHEL Neuro-Bio-Behavioral...

1521-0081/67/3/697–730$25.00 http://dx.doi.org/10.1124/pr.114.009423PHARMACOLOGICAL REVIEWS Pharmacol Rev 67:697–730, July 2015Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics

ASSOCIATE EDITOR: MARTIN C. MICHEL

Neuro-Bio-Behavioral Mechanisms of Placebo andNocebo Responses: Implications for Clinical Trials and

Clinical PracticeManfred Schedlowski, Paul Enck, Winfried Rief, and Ulrike Bingel

Institute of Medical Psychology and Behavioral Immunobiology (M.S.) and Department of Neurology (U.B.), University Clinic Essen, Essen,Germany; Department of Internal Medicine VI, Psychosomatic Medicine, University Hospital Tübingen, Tübingen, Germany (P.E.); and

Department of Psychology, University of Marburg, Marburg, Germany (W.R.)

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698II. The Origin of the Placebo Concept: From the Ancient Healer to Modern Medicine . . . . . . . . . . . 699III. Placebo Effects and Effect Sizes in Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701

A. Effect Sizes of Symptom Improvement across Different Medical Conditions. . . . . . . . . . . . . . 701B. Influence of Patient Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702C. Influence of Randomized Clinical Trial Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703D. Head-to-Head Trials: No Placebo Arm but Even Stronger Placebo Effects . . . . . . . . . . . . . . . 703

IV. Neuro-Bio-Behavioral Mechanisms of Placebo Responses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704A. Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704

1. Placebo Analgesia Involves Changes in the Pain Processing Network. . . . . . . . . . . . . . . . . 7042. Placebo Analgesia Engages Descending Pain Modulatory Networks. . . . . . . . . . . . . . . . . . 7053. Neurotransmitter Systems Involved in Placebo Analgesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . 7054. Unresolved Issues and Remaining Questions Regarding the Mechanisms of

Placebo Analgesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706B. Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706C. Neuropsychiatric Diseases and Behavioral Disorders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707

1. Depression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7072. Schizophrenia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7073. Anxiety Disorders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

D. Immunologic Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7081. Studies in Experimental Animals.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7082. Human Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7093. Toward the Clinical Application of Learned Immune Responses. . . . . . . . . . . . . . . . . . . . . . 709

E. Neuroendocrine Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710F. Autonomic Organ Functioning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711

1. Cardiovascular Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7112. Pulmonary Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7123. Nausea. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712

G. Gastrointestinal System/Irritable Bowel Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713H. Sleep Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714

V. Neuro-Bio-Behavioral Mechanisms of Nocebo Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715A. Nocebo Effects in Clinical Trials and Clinical Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715B. Psychologic Mechanisms Contributing to Nocebo Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715C. Neurobiological Pathways of Nocebo Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715

This work was supported by grants from the German Research Foundation (DFG) for the Research Unit FOR 1328 (BI 789/2-1,2; EN 50/30-1; RI 574/21-1,2; RI 574-22-1; SCHE 341/17-1,2); and the German Federal Ministry of Education and Research (01GQ0808; to U.B.).

Address correspondence to: Dr. Manfred Schedlowski, Institute of Medical Psychology and Behavioral Immunobiology, UniversityHospital Essen, 45122 Essen, Germany. E-mail: [email protected]

dx.doi.org/10.1124/pr.114.009423.

697

by guest on October 23, 2020

Dow

nloaded from

http://dx.doi.org/10.1124/pr.114.009423

mailto:[email protected]

http://dx.doi.org/10.1124/pr.114.009423

VI. The Effect of Placebo Responses on Pharmacological Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . 716A. Neural Mechanisms Underlying the Effect of Expectations on Drug Efficacy. . . . . . . . . . . . . 717B. Additive versus Interactive Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718

VII. Predictors of Placebo and Nocebo Responses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718VIII. Relevance and Implications of Placebo Responses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720

A. Randomized Clinical Trials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720B. Training Health Care Professionals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721C. Health Care System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722

IX. Placebo Research: What Next? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723X. Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723

Abstract——The placebo effect has often been consid-ered a nuisance in basic andparticularly clinical research.This view has gradually changed in recent years due todeeper insight into the neuro-bio-behavioral mechanismssteering both the placebo and nocebo responses, the eviltwin of placebo. For the neuroscientist, placebo andnocebo responses have evolved as indispensable toolsto understand brain mechanisms that link cognitiveand emotional factors with symptom perception aswell as peripheral physiologic systems and end or-gan functioning. For the clinical investigator, betterunderstanding of the mechanisms driving placebo andnocebo responses allow the control of these responsesand thereby help to more precisely define the efficacy

of a specific pharmacological intervention. Finally, inthe clinical context, the systematic exploitation of thesemechanisms will help to maximize placebo responsesand minimize nocebo responses for the patient’s benefit.In this review, we summarize and critically examine theneuro-bio-behavioral mechanisms underlying placeboand nocebo responses that are currently known interms of different diseases and physiologic systems. Wesubsequently elaborate on the consequences of thisknowledge for pharmacological treatments of patientsand the implications for pharmacological research, thetraining of healthcare professionals, and for the healthcare system and future research strategies on placeboand nocebo responses.

I. Introduction

More than 55 years ago, Stewart Wolf (Wolf, 1959)published a paper in this journal entitled, “ThePharmacology of Placebo,” in which he stated thatthe chosen title itself may present a “picturesquecontradiction,” because by definition pharmacology isconcerned with the chemical properties of drugs andtheir effects on biologic mechanisms. It was commonknowledge at that time that the pharmacologicalaction of drugs as well as “other forces at play andthe circumstances surrounding their administration”contribute to the overall treatment effect (Wolf, 1950).Empirical data revealed that placebo effects thatendow inert agents with potency or modify the treat-ment effects of active pharmacological substances arenot only subjective in nature but that they can also beassociated with measurable and thus objective changesin end organ functions. However, the underlying mech-anisms remained largely unclear.The placebo effect itself—the symptom improvement

after inert treatments in clinical trials—is composed ofdifferent factors, such as the natural history of adisease or fluctuation of symptoms, response biases,effects of cointerventions, or statistical phenomena,

such as regression to the mean (Fig. 1). These factorscan be distinguished from the actual placebo responsethat is mediated via three interdependent factors:patients’ expectations about treatment benefits, thequality and quantity of doctor-patient communication,and associative (conditioning) learning processes (Fig.2). These psychologic factors trigger complex neurobi-ological phenomena distinctly involving the centralnervous system (CNS) as well as system-specific periph-eral physiologic and end-organ changes. Confusionpersists even within the scientific community abouthow to define the terms “placebo,” “placebo effect,” and“placebo response.” In this review, we will refer to theterm placebo responses, thereby focusing on the effectsof the neuropsychological mechanisms of expectation,communication, and conditioning that drive the pla-cebo response (Table 1).

The past two decades have witnessed groundbreak-ing advances in the understanding of neurobiologi-cal and neuropsychological mechanisms of placeboand nocebo responses in various medical conditions(Moerman, 2002; Brody, 2008; Benedetti, 2008, 2011).This knowledge is not only pivotal for more detailedanalyses of the neuropsychological mechanisms driving

ABBREVIATIONS: ANS, autonomic nervous system; BMJ, British Medical Journal; CCK, cholecystokinin; CER, comparative effectivenessresearch; CNS, central nervous system; CS, conditioned stimulus; CsA, cyclosporine A; EEG, electroencephalography; fMRI, functionalmagnetic resonance imaging, IBS, irritable bowel syndrome; IFN, interferon; IL, interleukin; PAG, periaqueductal gray; PET, positronemission tomography; PSG, polysomnographic assessments; rACC, rostral anterior cingulate cortex; RCT, randomized clinical trials; SOL,sleep onset latency; TST, total sleep time; US, unconditioned stimulus; WASO, wake after sleep onset.

698 Schedlowski et al.

the placebo response—it also paves the way for system-atic exploitation of this knowledge with the aim to min-imize placebo responses in clinical trials and in thecontext of drug development and to maximize theplacebo response in clinical practice to improve treatmentoutcomes and patient benefit (Jubb and Bensing, 2013).In this review we will particularly emphasize the

neuro-bio-behavioral mechanisms of placebo and noceboresponses, their effects on pharmacological treatments,and potential predictors of interindividual differences inthese effects. On this basis, we will outline the relevanceof placebo responses on the assay sensitivity in random-ized clinical trials (RCTs) and on therapeutic outcomesin clinical practice. Finally, we will discuss implicationsfor the training of health care professionals and thehealth care systems themselves and propose futureresearch directions.

II. The Origin of the Placebo Concept: From theAncient Healer to Modern Medicine

Since the dawn of experimental placebo research inthe 1990s, many papers (Fig. 3) and books (Spiro, 1986;Harrington, 1999; Thompson, 2005; Benedetti, 2008;Shapiro and Shapiro, 2010) have referred to the historyof the placebo concept and its origins long beforeevidence-based medicine became the gold standard inWestern medicine. In the following section, we brieflyrefer to three aspects of these historical roots of theplacebo concept: etymological considerations, method-ological aspects, and a “meaning” aspect (Moerman andJonas, 2002).

i. The origin of the word “placebo” from the Latinverb “placere” (pleasing), in the sense of “I mayplease” or “it may please,” has been greatly

Fig. 1. Unspecific or placebo effects from any medical treatment entail a number of subeffects such as the natural history of disease and spontaneoussymptom variation, statistical phenomenon such as regression to the mean, response biases and false positive responses, contextual factors, and effectsof cointerventions as well as the actual placebo response. The placebo response itself is mediated by associative learning or conditioning processes, aswell as cognitive factors such as a patient’s expectation of a benefit from a treatment and the quality of the doctor-patient communication. The equalparts are not meant to indicate a statistically balanced contribution.

Fig. 2. Placebo responses are primarily mediated by cognitive factors such as patients’ expectations of treatment benefits, by behavioral conditioning(associative learning) with pharmacological stimuli, and the quality of doctor-patient communication. These psychologic factors trigger complexneurobiological phenomena involving distinct CNS as well as system-specific peripheral physiologic and end-organ alterations.

Neuro-Bio-Behavioral Mechanisms of Placebo and Nocebo Responses 699

stressed as the background of a “placebo” appli-cation that helps without any underlying scien-tific background. There is evidence that an evenearlier use of the word referred to “singinga placebo” as someone mourning at a funeral asa paid service (Shapiro, 1964). The term placeboremains stuck with this negative connotation,because it is ethically regarded as deception ofthe patient and is thus only allowed underrestricted circumstances, e.g., those of the Dec-laration of Helsinki (World Medical Association,2013). The use of placebos is still predominantly

an issue of pharmacological research and isinevitably associated with clinical trials, althoughthe establishment of placebo-controlled trials wascertainly not the drug industry’s “invention.”

ii. Long before “placebo controls” became standardin pharmacological research in the 1940s, theidea of a placebo control was already in the mindof doctors and researchers, even those nowwith a questionable scientific reputation. WilliamCullen used the term for the first time in 1772,when he gave a patient a dose of mustard powderand wrote “… that I did not trust much to it, but I

TABLE 1Placebo/nocebo terminology

Terminology

Placebo The word placebo is the Latin term for “I shall please.”It is used to indicate sham treatments or inertsubstances such as sugar pills or saline infusions.

Placebo effect The placebo effect is defined as any improvement ina symptom or physiologic condition of subjects afterplacebo treatment. There are different mechanismsunderlying this phenomenon, including spontaneousremission, regression to the mean, natural course ofa disease, biases, and placebo responses.

Placebo response The placebo response refers to the outcome caused bya placebo manipulation. It reflects the neurobiologicaland psychophysiological response of an individual toan inert substance or sham treatment and ismediated by various factors within the treatmentcontext. Importantly, placebo responses are notrestricted to placebo treatments—they can alsomodulate the outcome of any active treatment.

Active placebo An active placebo is a substance or treatment thatmimics the side effects of the active compound beingtested and is thus by definition not an inertsubstance. In clinical trials, active placebos areadministered to avoid un-blinding due to the differentside-effect profiles of drugs and placebo treatments.

Nocebo The term nocebo (I shall harm) was introduced incontrast to “placebo” to distinguish the positive fromthe noxious effects of placebos, when an inertsubstance is given within a negative context, inducingnegative expectations about the outcome.

Fig. 3. Number of genuine placebo (solid line) and nocebo publications (broken line) in PubMed per year between 1950 and 2013 (from Weimer et al.,2015b, with permission).


gave it because it is necessary to give a medicine,and as what I call a placebo” (Cullen, 1772, citedin Jütte, 2013, pp 1). Later, Samuel Hahnemann,founder of homeopathy in the 18th century,frequently used treatments such as milk sugar(lactose), which he deemed ineffective: “In themeantime, until the second medicament is given,one can soothe the patient’s mind and desire formedicine with something inconspicuous such asa few teaspoons a day of raspberry juice or sugarof milk” (Hahnemann, 1814, cited in Jutte, 2014,pp 210). Although he never used the term placebo,Hahnemann was obviously aware of its clinical“potency.” Similarly, the famous “Nuremberg SaltTest” of 1835 (Stolberg, 2006) (designed to dis-prove Hahnemann’s homeopathic concept) used“pure snow water” as a control for the “salinetreatment” that was a popular homeopathic treat-ment around that time, also without mentioningthe term placebo. For more details/information onthis and other examples, we refer the interestedreader to Ted Kaptchuk’s article on the history ofblind assessment and placebo controls in medicine(Kaptchuk, 1998).

iii. A recent publication in the British MedicalJournal (BMJ) sought to investigate the “mean-ing” of the term placebo appearing in articlespublished in the BMJ in the early days of evidence-based medicine between 1840 and 1899, madepossible thanks to the fact that the BMJ’s com-plete archive has been digitized (Raicek et al.,2012). They identified 71 articles mentioningthe term “placebo,” of which 47 (66%) were inspecific sections in the BMJ such as “Correspon-dence” (10%), “Original communications” (10%),and “Reports of societies” (4%), with the remain-ing 42% distributed among 23 other categories.Twenty-four of the citations (34%) were in non-specified sections. They assigned the use of theterm placebo to nine categories: no effect orpejorative, (31%), natural history (25%), satisfypatient (20%), medical performance (10%), buytime (4%), financial gain (4%), placebo control(3%), has clinical effect (1%), and unclear (1%).Taken together, only two articles (of 1886 and1889) used placebos as controls to test the effectsof a medical treatment, and only one mentionedits being clinically effective. According to theauthors, all placebo applications were deceptiveand did not debrief the patients after completionof the study.

Our brief history of placebo use in medicine illustratesthe three aspects that are still evident in today’’s use ofplacebos: the negative connotation of the term placebo,the suspicion that the use of placebos implies the de-ception of patients, and the speculation that placebos

may be ineffective in helping patients. Experimentaland clinical data from the last two decades has clearlydemonstrated that it is about time these assumptionswere overcome.

III. Placebo Effects and Effect Sizes inClinical Trials

A. Effect Sizes of Symptom Improvement acrossDifferent Medical Conditions

When compared with the effect of drugs in random-ized placebo-controlled trials, placebo effects can varysubstantially, ranging from under 10% to over 60%,even within single clinical entities. Requiring a 50%symptom improvement to qualify as a treatment re-sponder resulted in 26% placebo responders in diabeticneuropathic pain (Arakawa et al., 2015) but this waslower in other pain conditions (e.g., dental pain: 16%;Averbuch and Katzper, 2001). Similar response rateswere reported in migraine (29%; Macedo et al., 2006),fibromyalgia (45%; Hauser et al., 2011), and pancreaticpain (20%; Capurso et al., 2012) investigations.

It is, however, almost impossible to compare theplacebo response data from different meta-analyses ofRCTs in various clinical conditions. RCTs with a binaryoutcome that allow for the differentiation of placeboresponders and nonresponders usually report thepercentage of responders in the placebo arm, and themeta-analyses of such trials report the “pooled placeboresponse” (Weimer and Enck, 2014). Most RCTs withgastroenterological patients take this approach, re-vealing pooled placebo response rates ranging between25% and 45% (Weimer et al., 2013a; Elsenbruch andEnck, 2015). Trials with continuous outcome measures(symptom scale improvements, e.g., in depression,schizophrenia, Parkinson’s disease, attention deficithyperactivity syndrome, etc.) usually report effect sizesin the placebo arm, but their interpretation depends ona clinically meaningful grading of the scale that canvary according to the condition. In sleep disorders, forinstance, the placebo response accounts for 60% of theresponse in the drug arm of respective trials whenassessed by polysomnographic measures (Winkler andRief, 2015).

Placebo effects are more pronounced when assessedvia patient-reported outcomes (symptoms, symptomratings, quality of life measures) compared withbiomarkers (Meissner, 2005) or disease markers. How-ever, even with biologic indicators of disease activitysuch as the Crohn’s Disease Activity Index or endoscopicassessment of disease activity, the placebo response canremain as high as 25% (Su et al., 2004). Biomarkerssuch as the “forced expiratory volume” in asthma,however, display very weak placebo responses (Wanget al., 2012), and they are known to correlate poorly withpatient-reported outcomes or clinical assessments ofdisease severity (Wechsler et al., 2011). The highest


(around 40%) and most homogeneous placebo responserates are reported in psychiatric trials [e.g., in de-pression (Papakostas and Fava, 2009)] but also infunctional gastrointestinal disorders, such as irritablebowel syndrome (IBS) or functional dyspepsia, disordersknown to coincide with depression and anxiety (Fordand Moayyedi, 2010). In contrast, placebo responses areknown to be lower in neurologic disorders (epilepsy,Parkinson’s disease) (Goetz et al., 2008; Rheims et al.,2008) and in treating addiction, i.e., smoking cessation(Moore and Aubin, 2012) (Tables 2 and 3).One of the pitfalls when assessing placebo responses

in RCTs is the fact that they can be confounded byspontaneous symptom improvements (Fig. 1) assumedto be similar across all the trial’s arms and thereforenot controlled for. Whether the placebo response inRCTs is still of a clinically relevant effect size afterignoring the contribution of spontaneous symptomvariation is debatable (Hrobjartsson and Gotzsche,2001, 2004). A meta-analysis of three-arm trials in 8different clinical conditions including a “no treatment”control group revealed that about 50% of the placeboresponse could be explained by spontaneous remission/variation (Krogsboll et al., 2009). Although only in-volving 10 trials, a similar meta-analysis in majordepression disorder (Rutherford et al., 2012) indicatedthat “waiting” had an effect size of approximately 0.5(Cohen’s d), the equivalence of a 4-point, clinicallyrelevant improvement on the Hamilton DepressionScale. However, a waiting list is a poor means of “notreatment control” (Weimer and Enck, 2014) because itinduces a dynamic (the expectation of being treated inthe near future) that differs from that in an observa-tion study arm only (Relton et al., 2010). It has evenbeen suggested that “waiting list” is a “nocebo” in-tervention, although evidence for such a worsening

effect is limited to analyses that failed to take samplesize into account (Furukawa et al., 2014).

B. Influence of Patient Characteristics

There tends to be very broad interindividual varia-tion in the placebo response of healthy individuals andpatients. Thus, to keep the placebo response low inRCTs, the search for predictors of placebo responses inRCTs applied post hoc reanalyses and sensitivity testsof individual RCTs to identify putative predictors of theplacebo response that could then be used to identifyand exclude so called “placebo responders” from trials.Among the patient characteristics frequently accused ofdriving the placebo response are sex (with larger placeboresponses in women) and (younger) age. Although thesefactors were shown to be relevant in some trials (Thijset al., 1990; Freeman and Rickels, 1999; Rheims et al.,2008; Cohen et al., 2010; Yildiz et al., 2011; Agid et al.,2013; Arakawa et al., 2015), a recent review (Weimeret al., 2015a) of 75 systematic reviews and meta-analyses including more than 1500 trials, 150,000patients, and 40 medical indications revealed that ageand sex were not significant predictors of placeboresponses in RCT, despite occasional evidence fromexperimental research (Aslaksen et al., 2007; Weimeret al., 2013b).

A more detailed meta-analysis of patient-relatedfactors driving the placebo effect in psychiatric disor-ders recently revealed consistent positive associationsbetween placebo effects with lower disease severity atbaseline and shorter disease duration in the treatment-naive patients (Weimer et al., 2015b). Similarly, lowsymptom severity at study entry was also found tocorrelate positively with a placebo response in otherdiseases such as fibromyalgia (Hauser et al., 2011), dia-betic neuropathic pain (Hauser et al., 2011), binge-eating

TABLE 2Systematic reviews and meta-analyses of the placebo response in RCTs in neurologic disorders, pain

syndromes, and in psychiatric disordersNote that this list is incomplete, as it reports only the meta-analysis with the largest number of studies included for the

respective disease.

Study N* Disease PR Is Higher with…**

Rheims et al., 2008 32 Epilepsy (children, adults) younger ageFulda and Wetter, 2008 36 Restless Leg Syndrome longer trial durationGoetz et al., 2008 11* Parkinson’s Disease higher baseline severityDiener et al., 1999 15 Migraine higher drug probabilityMacedo et al., 2006 98 Migraine European studiesHauser et al., 2011 72 Fibromyalgia syndrome lower baseline severityHauser et al., 2012 70 Diabetic neuropath pain lower baseline severityCapurso et al., 2012 7 Pancreatitis more study sitesPapakostas and Fava, 2009 182 Depression, adults higher drug probabilityBridge et al., 2009 12 Depression, children more study sitesAgid et al., 2013 50 Psychosis younger age, more recent trialsYildiz et al., 2011 38 Bipolar mania more study sites, female sexCohen et al., 2010 40 MDD, OCD, ANX (children) children than in adolescentsNewcorn et al., 2009 10 ADHD (children) comorbid MDD, non-whiteBuitelaar et al., 2012 2* ADHD (adults) higher baseline severityBlom et al., 2014 10* Binge Eating Disorder lower baseline severity

ADHD, attention deficit hyperactivity disorder; ANX, anxiety disorder; MDD, major depressive disorder; OCD,obsessive compulsive disorder; PR, placebo response.

*Indicates availability of individual patient data; number of RCTs included into analysis.**Only the most important influential variable listed.


disorder (Blom et al., 2014), or asthma (Wang et al.,2012), but not in attention deficit hyperactivitydisorder (Buitelaar et al., 2012) or Parkinson’s disease(Goetz et al., 2008). Meta-analyses of more trials indifferent medical conditions also failed to clearly identifypredictors of the placebo response, which can at least bepartially explained by limited access to individualizeddata, as Tables 2 and 3 illustrate. Individual patientdata were available in only 6 of the listed 30 meta-analyses, usually in small data sets. Although morerecent experimental approaches have begun focusing onidentifying behavioral or genetic traits of placebo ornocebo response differences, no consistent concept hasemerged so far (see section VI).

C. Influence of Randomized Clinical TrialCharacteristics

When it was first noted that the rate of placeboresponses had increased over the years in RCTs fordepression (Walsh et al., 2002; Bridge et al., 2009) andschizophrenia (Agid et al., 2013; Rutherford et al.,2014), this was interpreted as an indication of changesin the characteristics of RCTs. Several RCT character-istics that increase placebo response rates have beenidentified in individual re- and meta-analyses. Theseinclude numerous study sites participating in multi-center studies (Bridge et al., 2009; Yildiz et al., 2011;Capurso et al., 2012), longer trial duration (Su et al.,2004; Fulda et al., 2007), and RCTs performed inEurope rather than the United States (Macedo et al.,2006; Ford and Moayyedi, 2010). These factors havechanged during the last decades and might account forthese time trends. More importantly, in depression, ithas been shown that the trend is only apparent whenthe doctor rates treatment success but not when thepatient does (Rief et al., 2009b). The same study alsoreported evidence for the increasing homogeneity ofinvestigated patient samples over the years, which also

contributes to larger effect sizes but limits the externalvalidity of results. Furthermore, today’s RCTs alsoinvolve more frequent doctor-patient contacts that cancontribute to enhanced placebo responses (Ilnyckyjet al., 1997; Cho, 2005), as can the drug applicationfrequency (de Craen et al., 1999b) and its applicationroute (Narkus et al., 2013).

Unbalanced randomization refers to any deviationfrom a 50:50 drug:placebo randomization in RCTs. Thisapproach is often chosen to assign more patients to theactive drug in a two-arm trial for ethical reasons or totest more drug dosages against placebo by adding morestudy arms. As noted early on (Diener et al., 1999),unbalanced randomization may increase the overallplacebo response in RCTs, as shown in trials ondepression (Papakostas and Fava, 2009; Sinyor et al.,2010; Mancini et al., 2014), schizophrenia (Woodset al., 2005; Mallinckrodt et al., 2010), and psychosis(Agid et al., 2013). It is, however, unclear whether thisrepresents a specific effect in neuropsychiatric dis-orders, because it has not been observed in otherconditions, such as IBS (Ford and Moayyedi, 2010;Elsenbruch and Enck, 2015).

D. Head-to-Head Trials: No Placebo Arm but EvenStronger Placebo Effects

Because placebo responses are immanent in allmedical treatments, the omission of a placebo arm inRCTs does not prevent placebo responses—it onlyrenders their systematic assessment more difficult. Theevaluation of enrichment trials, or unbalanced random-ization in experiments and clinical trials (Weimer andEnck, 2014), has shown that providing a 100% prob-ability of receiving an active treatment increasesthe response in both drug and placebo groups, asopposed to the 50:50 probability in placebo-controlledtrials. In head-to-head trials, in which both treatmentarms receive an active component, it is impossible to

TABLE 3Systematic reviews and meta-analyses of the placebo response (pooled PR, percent) in RCTs in various

diseases in internal medicineNote that this list is incomplete, as it reports only the meta-analysis with the largest number of studies included for the

given disease.

Study N* Disease PR Is Higher with… **

Ilnyckyj et al., 1997 38 Ulcerative colitis higher number of visitsSu et al., 2004 21 Crohn’s Disease longer study durationFord et al., 2010 73 IBS European studiesTalley et al., 2006 4* FD inconsistent symptomsde Craen et al., 1999a 79 Duodenal ulcers# higher application frequencyCremonini et al., 2010 24 Reflux disease nonerosive reflux diseaseThijs et al., 1990 1* Hypertension older ageCho et al., 2005 29 CFS high intervention intensityFreeman et al., 1999 2* PMS younger ageLamel et al., 2012 31 Psoriasis higher drug probabilityNarkus et al., 2013 6 Allergy subcutaneous administrationWang et al., 2012 34 Asthma lower baseline severity

CFS, chronic fatigue syndrome; FD, functional dyspepsia; PMS, premenstrual syndrome; PR, placebo response.*Indicates availability of individual patient data; number of RCTs included into analysis.**Only the most influential variable listed.


specifically assess a placebo response. However, analy-ses of the treatment responses in head-to-head trialscompared with placebo-controlled trials enable us toindirectly assess the contribution of expectancy/placebomechanisms to the efficacy of active treatment (see sec-tion VI).A meta-analytic comparison of head-to-head trials

and placebo-controlled trials of the same drugs fortreating depression demonstrated that comparativetrials enhance the drug response compared withplacebo-controlled trials of the same compounds. Thiseffect is explained solely by the patient’s 100% assur-ance to receive a drug, and it resulted in an additional15% “placebo response” to the established average of40% from placebo-controlled drug trials for depression(Rutherford et al., 2012). Similar data have beenreported in schizophrenia (Woods et al., 2005). Thiscreates an ethical dilemma: head-to-head trials need upto four times more patients to statistically test “non-inferiority” than conventional placebo-controlled trials(Leon, 2012), a factor that contradicts the Declaration ofHelsinki request (World Medical Association, 2013) thatthe minimum number of patients be included in RCTs.These trials are also associated with substantiallyhigher trial costs, in particular when the appropriatecomparator drug selected is not the property of thesponsoring company and needs to be produced andthe provision of double-dummy technology is needed(Maru�si�c and Feren�ci�c, 2013). Finally, the comparator’sselection may force substantial methodological consid-erations and concerns if more than one potentialcomparator is available on the market (Dunn et al.,2013).In summary, the substantial effect sizes of placebo

treatments in RCTs are based on both patient-relatedfactors (e.g., age, sex, disease history, and severity),most of which vary greatly in their contribution to theplacebo response depending on the clinical condition,and by design-related factors that appear more homo-geneous between conditions (e.g., unbalanced random-ization, frequency of doctor-patient contacts, trialduration) but lack controllability across different con-ditions. Finally, public and scientific access to previousRCTs on the individual patient-data level are required(Tudur Smith et al., 2014; Lo, 2015) to further explorethe contribution of these factors to the effect size ofplacebos as long as placebo-controlled trials are the goldstandard in the evaluation of new drugs.

IV. Neuro-Bio-Behavioral Mechanisms ofPlacebo Responses

A. Pain

Among all the placebo responses, it is placeboanalgesia (referring to the phenomenon of reducedpain ratings after application of a sham treatment) thatis the most thoroughly examined and neurobiologically

best characterized placebo response (Tracey, 2010a). Theadvent of modern noninvasive neuroimaging techniqueshas provided the unique opportunity to study the neuralmechanisms of placebo analgesia and other placeboresponses (Benedetti, 2014).

1. Placebo Analgesia Involves Changes in the PainProcessing Network. One of the key questions thathas intrigued neuroscientists and clinicians alike iswhether the subjective reductions in pain ratingsduring placebo analgesia are associated with activitychanges in the "pain processing network"—the set ofbrain regions most closely associated with the expe-rience of pain—and, if so, in which of its components?

The majority of neuroimaging studies addressingplacebo analgesia report that the reduced painratings during placebo analgesia are accompaniedby decreased activity in the classic pain-processingareas, including the thalamus, insula, somatosensorycortex, and mid-cingulate regions (Wager et al., 2004;Bingel et al., 2006; Eippert et al., 2009a; Lui et al.,2010; Elsenbruch et al., 2012a; Geuter et al., 2013)(Fig. 4). Electroencephalography (EEG) studies havefurther confirmed that placebo analgesia is associ-ated with reduced amplitudes of event-related poten-tials to experimental pain stimuli (Wager et al., 2006;Watson et al., 2007; Aslaksen et al., 2011).

Fig. 4. The CNS mechanisms initiating and mediating placebo responsesare best characterized for placebo analgesia and involve the descendingpain modulatory network, which includes the dorsolateral prefrontalcortex (DLPFC), the anterior cingulate cortex (ACC), the amygdala (Am),and the PAG. Similar regions of the brain have been shown to contributeto emotional placebo responses. The shared and distinct contributions ofdifferent brain networks in other types of placebo responses are currentlyunknown.


Evidence from spinal cord forced magnetic resonanceimaging (fMRI), which only recently became techni-cally feasible in both humans and animals, has revealedthat pain-related activity in the ipsilateral dorsal horn,corresponding to painful stimulation, is substantiallyreduced under placebo (Eippert et al., 2009b) and risesunder expectations of increased pain (nocebo) (Geuterand Büchel, 2013). Together these studies support thenotion that altered pain experience during placeboanalgesia is not simply the result of report bias butcan, at least in part, result from the active inhibition ofnociceptive activity, even involving very early stages ofneural processing. However, this does not exclude otherbrain mechanisms’ relevance and contribution to pla-cebo analgesia.2. Placebo Analgesia Engages Descending Pain

Modulatory Networks. These observations have raisedthe question which brain processes mediate changes inpain processing and perception. The pioneering positronemission tomography (PET) study on placebo analgesiaby Petrovic et al. (2002) first revealed a shared neuralnetwork in the rostral anterior cingulate cortex (rACC)and the brain stem underlying both opioid and placeboanalgesia. The relevance of this network for placeboanalgesia has been substantiated by several subsequentstudies using various procedures to induce placeboanalgesia, including placebo analgesic creams, shamacupuncture, and others (Wager et al., 2004, 2007;Zubieta et al., 2005; Bingel et al., 2006; Kong et al.,2006; Eippert et al., 2009a). These studies showed thatplacebo analgesia involves the activation of cingulo-frontal brain regions together with subcortical struc-tures such as the midbrain periaqueductal gray (PAG),hypothalamus, and amygdala. Connectivity analysesfurther revealed that the behavioral placebo analgesiceffect is related to enhanced functional coupling of therACC with brain stem areas, such as the PAG (Wageret al., 2004, 2007; Eippert et al., 2009a), and thatindividual activity and connectivity changes within thisnetwork predict the behavioral analgesic effect (Wageret al., 2004; Eippert et al., 2009a).These studies support the notion that the top-down

activation of endogenous analgesic activity via thedescending modulatory system represents one mecha-nism of placebo analgesia. The prefrontal cortex seemsto play a crucial role in this mechanism. Activity in thedorsolateral prefrontal cortex was found in the periodpreceding noxious stimulation, which correlated withactivity in the PAG and the subsequent placeboanalgesic response (Wager et al., 2004; Eippert et al.,2009a; Lui et al., 2010). Intriguingly, both temporaryfunctional lesions in the prefrontal cortex by repetitivetranscranial magnetic stimulation (Krummenacheret al., 2010) as well as degenerated and disconnectedfrontal lobes in Alzheimer’s disease (Benedetti et al.,2006) are associated with a reduction in or completeloss of verbally induced placebo analgesic responses.

These findings implicate the prefrontal cortex asa brain region critical to the initiation of expectancy-related effects on pain perception.

3. Neurotransmitter Systems Involved in PlaceboAnalgesia. Levine et al. (1978) first demonstratedthat placebo analgesia can be antagonized by naloxone,suggesting the involvement of endogenously releasedopioids. Since then, the contribution of opioidergicneurotransmission in placebo analgesia has beencorroborated by indirect pharmacological approachesusing opioid antagonists and PET studies using in vivoreceptor binding approaches with opioidergic ligands(Amanzio and Benedetti, 1999; Zubieta et al., 2005;Wager et al., 2007; Eippert et al., 2009a). These PETstudies have substantiated that the placebo-relatedactivity and connectivity changes within cingulo-frontal and subcortical networks including the PAGobserved with fMRI involve m-opioidergic neurotrans-mission (for review, see Peciña and Zubieta, 2015). Thecontribution of opioidergic neurotransmission is fur-ther substantiated by pharmacological fMRI studies,i.e., by showing that naloxone blocks placebo-relatedactivity in the PAG and rostral ventromedial medullaand PAG-rACC connectivity along with impairedbehavioral analgesia (Eippert et al., 2009a), as wellas in seminal studies linking variability in genesmodulating opioidergic mechanisms with an individu-al’s placebo analgesic response (Hall et al., 2012;Peciña et al., 2013a,b, 2014, 2015).

The endogenous opioid system is not the only oneinvolved in placebo analgesia. Functional molecularimaging investigating changes in the binding potentialof 11C-labeled raclopride has shown increased dopami-nergic neurotransmission in the nucleus accumbens,putamen, and caudate nucleus that correlated with theindividual placebo analgesic response (Scott et al.,2007, 2008). Furthermore, Schweinhardt et al. (2009)reported a close positive relationship between graymatter density in the ventral striatum and the magni-tude of placebo analgesia, as well as dopamine-relatedpersonality traits (e.g., novelty, fun, and sensationseeking) using voxel-based morphometry, whereasothers have directly linked reward responsiveness andplacebo analgesia experimentally (Scott et al., 2007).These findings suggest a potentially relevant role of thedopaminergic system and the striatum in particularin placebo analgesia. It is, however, unclear whetherstriatal dopaminergic activity is causally involvedin generating analgesia or rather reflects reward pro-cesses associated with pain relief. The expectation ofdopamine release enhanced reward learning and mod-ulated learning-related signals in the striatum andventromedial prefrontal cortex (Schmidt et al., 2014).In a recent fMRI study on placebo analgesia, the ad-ministration of the D2/3 antagonist haloperidol blockedplacebo-related activity in the striatum but had no effecton analgesia or on activity in brain regions believed to


encode pain intensity (Wrobel et al., 2014). Taken to-gether, although these data suggest endogenous dopami-nergic pathways somehow contributing to the individualplacebo analgesic response, the distinct role of the do-paminergic system in placebo analgesia requires furtherinvestigation (Peciña et al., 2013a).Placebo analgesia was also recently linked to the

cannabinoid system (Benedetti et al., 2011). This systemseems to underlie placebo analgesia after pharmacolog-ical conditioning with the nonsteroidal anti-inflammatorydrug ketorolac. In this case, placebo analgesic responseswere reversed by the cannabinoid 1 receptor antagonistrimonabant, indicating that the effects elicited by non-steroidal anti-inflammatory drug conditioning are par-tially mediated by the endogenous release of cannabinoids(Benedetti et al., 2011).Overall, these findings support the notion that the

neurobiological effects of placebo analgesia are relatedto distinct brain mechanisms and neurochemical path-ways that are activated in various contexts contribut-ing to placebo analgesia.4. Unresolved Issues and Remaining Questions Re-

garding the Mechanisms of Placebo Analgesia. Al-though the aforementioned lines of evidence supportthe view that descending pain modulatory mechanisms,even those involving the spinal cord, are a criticalpathway underlying placebo analgesia, there is alsoevidence supporting the relevance of intracortical mech-anisms to placebo analgesia. Recent novel methodolog-ical approaches including multivariate pattern analysisand meta-analyses of brain-imaging studies suggestthat most behavioral variance in placebo analgesia isexplained by activity changes in intracortical, emotion-related circuitry, rather than changes in sensory brainareas (Wager et al., 2011), as one would expect wereplacebo analgesia determined solely by descending in-hibition. Defending this view, a recent laser-evokedpotential study reported that effective behavioral placeboanalgesia can occur without changes in N1 amplitude,which represents the earliest recordable in vivo corticalresponse to afferent spinothalamic input (Martini et al.,2015).However, in contrast to the latest perspectives that

either highlight descending or intracortical pathways,it is conceivable that several not mutually exclusivepathways can contribute to placebo analgesia, depend-ing on the individual disposition and the context inwhich placebo analgesia is induced. Further researchneeds to explore the interactions between the differentsystems involved in placebo analgesia under physio-logic and pathologic conditions (e.g., acute and chronicpain).

B. Parkinson’s Disease

Marked improvements in Parkinson symptoms havebeen observed in the placebo arm of clinical trialstesting pharmacological and surgical treatments for

Parkinson’s disease (Shetty et al., 1999; Goetz et al.,2002, 2008; McRae et al., 2004; Diederich and Goetz,2008). Although placebo rates in clinical trials do notallow for dissociating between “true” placebo responses(induced by positive expectation, doctor-patient com-munication, or prior experience) and natural fluctua-tions in the underlying disease unless a “no treatment”arm is included, these high placebo response rates inclinical trials of Parkinson’s disease have motivateddeeper examination of the neurobiological mechanismsunderlying clinical improvements following placebotreatments in Parkinson’s disease.

De la Fuente-Fernandez et al. (2001) were the first tostudy the involvement of the endogenous dopaminesystem for placebo responses in Parkinson patients byusing raclopride-PET. After the administration ofa placebo that patients believed to be apomorphine (apowerful anti-Parkinsonian treatment), the authorsobserved increasing dopaminergic neurotransmissionin the striatum. The release of dopamine in the dorsalstriatum was greater in those patients who reportedclinical improvement, suggesting a relationship be-tween placebo-induced changes in dopaminergic neu-rotransmission in the dorsal striatum and individualclinical benefit. Interestingly, no such association wasobserved between clinical benefit and changes indopaminergic neurotransmission in the ventral stria-tum, which was linked with patients’ expectations ofsymptom improvement that can also be consideredreward anticipation. The contribution of the dopami-nergic system to placebo responses in Parkinson’sdisease was later corroborated by two other dopamine-ligand PET studies on placebo responses in Parkinsonpatients (Strafella et al., 2006; Lidstone et al., 2010).The latter study by Lidstone and colleagues specificallyinvestigated how the strength of expectation of clinicalimprovement influences the changes in dopaminergicneurotransmission in the striatum in response toa placebo treatment. Parkinson patients were told theyhad a specific probability (25%, 50%, 75%, or 100%) ofreceiving an active dopaminergic medication, but theyin fact received a placebo. Intriguingly, significantdopamine release was only documented in thosepatients with the 75% probability expectation. Theresponse to prior medication appeared to be a majordeterminant of placebo-induced dopamine release in themotor (dorsal) striatum, which might be explained bya preconditioning phenomenon. In contrast, the expec-tation of clinical improvement was additionally requiredto trigger dopamine release in the ventral striatum(Lidstone et al., 2010), which likely reflects reward-related processes associated with the expectation ofclinical improvement (de la Fuente-Fernandez et al.,2004).

Single cell recordings in patients undergoing surgeryfor implantation of a deep brain stimulator (stimula-tion of the nucleus subthalamicus) provided the unique


opportunity to directly investigate any neuronal changesin the basal ganglia underlying placebo responses inParkinson’s disease. Benedetti et al. (2004) were the firstto record activity from single neurons in the subthalamicnucleus in awake Parkinson’s disease patients beforeand after placebo administration following a pharmaco-logical preconditioning procedure using apomorphine(Fig. 2). They reported a significant decrease in the firingrate and bursting activity after the placebo interventioncompared with baseline that was associated with thepatients’ subjective report of well-being and a reductionin muscle rigidity at the wrist, as rated by a blindedneurologist. In their later study, those observationswere expanded by recordings from thalamic nuclei andthe substantia nigra (Benedetti et al., 2009), showingcomplex changes in the entire subthalamic-nigral-thalamicmotor circuitry associated with the improvements follow-ing placebo administration (for a review, see Frisaldi et al.,2014).Taken together, several lines of evidence support the

notion that placebo responses in Parkinson’s diseaseare associated with changes in neuronal activity anddopaminergic neurotransmission in the brain circuitryinvolved in the pathophysiology of the disease itself (forreview, see Murray and Stoessl, 2013). These findingshave important implications for the interpretation ofclinical trials and the clinical care of Parkinsonpatients.How reward-related and potential disease-specific

processes associated with the placebo response inParkinson’s disease combine and interact and how longthese short-term experimental observations last andcontribute to clinical improvements over the longerterm (weeks to months) will have to be determined infuture studies.

C. Neuropsychiatric Diseases and BehavioralDisorders

Numerous clinical studies and meta-analyses pro-vide compelling evidence for pronounced placeboeffects in psychiatric conditions such as depression,schizophrenia, or anxiety disorders (Cavanna et al.,2007; Murray and Stoessl, 2013; Weimer et al., 2015b).Compared with pain or Parkinson’s disease, however,the neurobiological mechanisms underlying placeboresponses in these disorders are much less wellunderstood, which is partly because of the lack ofexperimental models suitable to investigate suchmechanisms in healthy volunteers and patients.1. Depression. Comprehensive reviews and meta-

analysis indicate that the drug-placebo difference inclinical studies of depression are relatively small,reporting placebo response rates up to 70% (Kirschand Sapirstein, 1998; Kirsch et al., 2008; Rief et al.,2009a; Gueorguieva et al., 2011; Mora et al., 2011) (seesection III). In addition, 79% of placebo respondersremained well within the continuation phase in the

studies lasting over 12 weeks, suggesting a long-lastingeffect of the antidepressive placebo response (Khanet al., 2008). The neurobiological mechanisms underly-ing these responses have been analyzed by employingquantitative electroencephalography (Leuchter et al.,2002). Placebo responders showed a significant increasein prefrontal brain activity that was not observed inplacebo nonresponders or in patients responding or notresponding to the medication.

Changes in glucose metabolism in depressive malepatients were analyzed after 6 weeks’ treatment withthe antidepressant fluoxetine or placebo via PET(Mayberg et al., 2002). The placebo response wasassociated with metabolic increases in prefrontal,anterior cingulate, premotor, parietal, and posteriorcingulate cortex and decreases in the subgenualcingulate, parahippocampus, and thalamus. Thesechanges in brain activity revealed at least some sharedneurobiological pathways for placebo responses in painand depression in conjunction with expectation andemotion- and reward-related circuitry involvement(Benedetti et al., 2005; Wernicke and Ossanna, 2010;Murray and Stoessl, 2013). These neurobiologicalfindings are accompanied by observations that in-creased subjective reward experience affects treatmentefficacy with antidepressive medication (Wichers et al.,2009). In addition, the expectancy of improvement isaffected by the probability of receiving active antide-pressant medication, which in turn seems to influencethe antidepressive response (Rutherford et al., 2013).These neuropsychological post hoc data justify admin-istering active placebos in clinical trials analyzing theeffectiveness of antidepressive treatment (Salamone,2000; Moncrieff et al., 2004; Edward et al., 2005; Encket al., 2013; Weimer et al., 2015b).

2. Schizophrenia. Placebo responses reported inclinical trials in schizophrenia are similar in magni-tude, quality, and impact to those observed in de-pression trials (Kinon et al., 2011). The difference inresponder rates between antipsychotic and placebotreatments in psychosis is reported to be 18% (Leuchtet al., 2009, 2013). Results suggest significant vari-ability in the magnitude of the placebo response due topatient characteristics and trial design factors, withthe magnitude of placebo effects increasing over time(Agid et al., 2013; Rutherford et al., 2014) (see sectionIII).

To date, the underlying neurobiological mechanismssteering placebo responses in schizophrenic patientshave not been studied; they are difficult to investigatemainly because there is no generally accepted exper-imental model for schizophrenia. Moreover, schizo-phrenia’s symptoms are classified into three groups:positive symptoms, such as hallucinations and delu-sions, negative symptoms, such as apathy and socialwithdrawal, and cognitive symptoms, such as poorexecutive functioning and working memory (Murray


and Stoessl, 2013; Kingwell, 2014), all of which couldreact differently to placebo mechanisms. Experimentalapproaches in this field might be further complicatedby the fact that patients’ expectations may vary con-siderably because of a false interpretation of the termplacebo, and positive psychotic symptoms may be clas-sified as rewarding (Dunn et al., 2006).There is thus an urgent need to characterize the

neuropsychological mechanisms underlying placeboresponses in schizophrenia to ensure continuedsupport for investment and progress in CNS drugdevelopment.3. Anxiety Disorders. For anxiety disorders such as

generalized anxiety, panic disorder, or social phobia,placebo effects in clinical trials have been reportedranging between 10% and 60% (Loebel et al., 1986;Mavissakalian, 1988; Mellergard and Rosenberg, 1990;Piercy et al., 1996; Huppert et al., 2004; Khan et al.,2005; Stein et al., 2006). Moreover, in RCTs investi-gating the anxiolytic effect of the benzodiazepinealprazolam, improvement in the placebo arm remainedstable after the pill intake was discontinued, whereaspatients in the drug arm suffered relapses to baselinelevels (Ballenger et al., 1988; Pecknold et al., 1988),indicating that placebo responses in anxiety disorderscan indeed be long lasting and clinically relevant. Thesubstantial and robust placebo responses observed inclinical studies of anxiety disorders motivated experi-mental investigations in their neurobiological under-pinnings and inspired the search for an individual’spredictors of placebo responses, although the results sofar have been vague (Dager et al., 1990; Woodmanet al., 1994; Feltner et al., 2009).PET analyses showed that placebo responses in

patients with social anxiety disorders were accompa-nied by attenuated amygdala activity, a brain regioncrucial for emotional processing. Intriguingly, this onlyapplied to those subjects homozygous for the long alleleof the serotonin transporter–linked polymorphic regionor the G variant of the G-703T polymorphism in thetryptophan-hydroxilase-2 gene promoter. In the samestudy, the tryptophan-hydroxilase-2 polymorphismsignificantly predicted placebo responses, wherebyhomozygosity for the G allele was associated withmore pronounced improvement in anxiety symptoms(Furmark et al., 2008). Likewise, anxiety-relievingplacebo responses in healthy subjects decreased activ-ity in emotionally responding brain areas such as theamygdala, insula dorsal ACC, and increased activity inthe subgenual anterior cingulate cortex and ventralstriatum, indicating a reward-related response inducedby cognitive factors such as volunteers’ expectations(Zhang et al., 2011). In patients with social anxiety,connectivity changes between the amygdala anddorsolateral prefrontal cortex and rACC areas havebeen shown to be associated with the individualanxiolytic response to SSRI treatment and placebo,

as demonstrated in PET. These observations supportthe notion that similar cognitive or expectancy-relatedmechanisms are involved in improving emotion regu-lation in patients taking anxiolytic drugs and placeboresponders (Faria et al., 2014).

Placebo responses in neuropsychiatric disordershave been documented in both RCTs and experimentalstudies (Weimer et al., 2015b). A few experimentalstudies have reported at least some shared neurobio-logical pathways for placebo responses in neuropsychi-atric disorders and pain with the involvement ofexpectation and emotion- and reward-related circuitry.A thorough understanding of the mechanisms steeringthese often profound placebo responses will be essen-tial to exploiting this knowledge and to maximizingthese unspecific treatment effects in daily clinical carefor the patient’s benefit and to minimize placeboresponses for the development of effective drugs (Encket al., 2013).

D. Immunologic Responses

Meta-analyses and reviews have demonstrated thesusceptibility to “inert” treatments of various immune-related pathologic conditions such as ulcerative colitis(Ilnyckyj et al., 1997), duodenal ulcer (de Craen et al.,1999b), Crohn’s disease (Su et al., 2004), or multi-ple sclerosis (La Mantia et al., 1996). However, fewattempts have been made to specifically investigateplacebo responses and the underlying mechanismsdirectly modulating peripheral immune functions inhuman subjects (Pacheco-Lopez et al., 2006). Theseexperimental data indicate that peripheral immunefunctions seem to be predominantly affected byassociative learning or conditioning paradigms ratherthan by cognitive factors alone, such as the expectationof subjects or patients of a given treatment’s benefit.

Learned placebo responses in immune functions canbe demonstrated by employing associative learningparadigms in experimental animals, healthy humans,and patients (Enck et al., 2008; Schedlowski andPacheco-Lopez, 2010). A prerequisite for the classicconditioning of immune functions is intense bidirec-tional communication between the CNS and the pe-ripheral immune system (Meisel et al., 2005; Tracey,2009, 2010b) that constantly exchange information onefferent and afferent pathways. The classic condition-ing of immune responses is one of the most impressiveexamples of communication between these majorphysiologic systems (Exton et al., 2001; Ader, 2003;Pacheco-Lopez et al., 2006; Schedlowski and Pacheco-Lopez, 2010).

1. Studies in Experimental Animals. Most insightsinto the neurobiological mechanisms steering learnedplacebo responses in immune functions came fromstudies in experimental animals. The majority ofexperiments in rodents usually employ the so-calledtaste aversion paradigm (Garcia et al., 1985) in which


a novel taste (conditioned stimulus [CS]) is paired withthe administration of an immunomodulating drug (un-conditioned stimulus [US]) during acquisition. When theCS is re-presented at a subsequent time point duringthe evocation or memory phase, animals demonstratea modification of immune parameters as a conditionedresponse that generally mimic the actual drug (US)effect. Experimental evidence over the last 3 decadeshas demonstrated behaviorally conditioned stimulatoryor inhibitory effects in rodents, both in humoral andcellular immunity, with behavioral conditioning able tore-enlist changes in lymphocyte circulation and pro-liferation, cytokine production, natural killer cell activ-ity, or endotoxin tolerance (reviewed in Exton et al.,2001; Hucklebridge, 2002; Ader, 2003; Pacheco-Lopezet al., 2006; Riether et al., 2008; Schedlowski andPacheco-Lopez, 2010).The underlying central, efferent, and afferent neu-

robiological mechanisms steering learned immunosup-pressive placebo responses are the best understood todate and are thus described here in greater detail.Employing the immunosuppressants cyclophoshamideor cyclosporine A (CsA) as a US demonstrated that theinsular cortex and amygdala are central key structuresin the behaviorally conditioned suppression of antibodyproduction (Ramirez-Amaya et al., 1996), lymphocyteactivity, as well as cytokine production [interleukin(IL)-2, interferon (IFN)-g] and cytokine mRNA ex-pression (Pacheco-Lopez et al., 2005; Schedlowski andPacheco-Lopez, 2010). Employing the CsA-taste condi-tioning paradigm, the learned immunosuppressiveresponses are mediated on the peripheral efferentarm via the splenic nerve via noradrenaline andadrenoceptor-dependent mechanisms (Exton et al.,1998, 1999, 2002; Pacheco-Lopez et al., 2009; Rietheret al., 2011). However, this neuroanatomical path-way with the splenic nerve mediating the learnedimmunosuppression appears to be just one of manyefferent neural routes mobilized during learned immu-nosuppression, because the learned inhibition of thecontact-hypersensitivity reaction (employing the identi-cal CsA-taste paradigm) was independent of sympa-thetic splenic innervation (Exton et al., 2000).Other approaches analyzed afferent mechanism(s)

enabling the CNS to receive information regardingspecific drug-induced immune changes or pharmaco-logical effects induced by the drug employed as a US.Acute peripheral CsA administration induced behav-ioral changes such as decreased ambulatory activity inthe open field 6 hours after CsA injection (von Horstenet al., 1998). These changes coincide with increasedneuronal activity detected in the insular cortex andamygdala 1 to 6 hours after intraperitoneal CsAinjection, as evident in intracortical EEG telemetry,as well as c-Fos expression and increased noradrena-line levels in the amygdala as determined by micro-dialysis (Doenlen et al., 2011; Pacheco-Lopez et al.,

2013). These immediate alterations appeared to bea direct effect of CsA and not indirectly mediated viathe vagus nerve, because a vagal deafferentation beforeCsA injection did not prevent the increased neuralactivity. However, CsA levels were detected in thecerebellum, insular cortex, and amygdala 2 to 4 hoursafter CsA administration (Pacheco-Lopez et al., 2013).Overall, at present it is incompletely understood howthe CNS detects the changes induced by differentsubstances and drugs employed in paradigms oflearned immune responses, and we still need toelucidate which molecules are the messengers thatactivate the brain during the acquisition or learningphase of a conditioning protocol (Hadamitzky et al.,2013).

2. Human Studies. The data from experimentalanimals together with deeper understanding of themechanisms responsible for learned immune functionsformed the basis for studying this interesting pla-cebo phenomenon in humans (Vits et al., 2011). TheCsA-taste-immune paradigm was expanded from justbeing applied in experimental animals to being ap-plied in healthy humans, revealing the behaviorallyconditioned significant suppression of T-cell functiondetected via impaired cytokine (IL-2 and IFN-g) pro-duction, reduced cytokine mRNA expression, andinhibited T-cell proliferation (Goebel et al., 2002). Thislearned immunosuppression can be repeatedly recalledby exposing the conditioned subjects to the CS againafter an 11-day break (Wirth et al., 2011). Moreover,plasma noradrenaline concentration, state anxiety,and baseline levels of IL-2 predicted nearly 60% ofthe conditioned immunosuppressive response (Oberet al., 2012), providing evidence for biologic and psycho-logic predictors of conditioned placebo responses ingeneral and learned immune responses in particular.Further studies employing the CsA-taste conditioningparadigm in humans showed that immunosuppressioncould not be induced by just manipulating the expec-tancy of test subjects (Albring et al., 2012). In addition,experimental evidence confirms previous observationsin rodents (Niemi et al., 2007), namely that inducinga pronounced learned response in immune functionsrequires multiple CS-US combinations during acquisi-tion and evocation (Goebel et al., 2005; Albring et al.,2012; Grigoleit et al., 2012).

3. Toward the Clinical Application of Learned Im-mune Responses. A number of studies in rodents havemeanwhile demonstrated the potential clinical rele-vance of learned responses in immune functions.Specifically, the morbidity and mortality of animals withautoimmune disease was abated via the learned im-munosuppressive response (Ader and Cohen, 1982;Klosterhalfen and Klosterhalfen, 1983, 1990; Joneset al., 2008). In addition, asthma-like symptoms, anaphy-lactic shock (Noelpp and Noelpp-Eschenhagen, 1951a,b;Djuric et al., 1988; Palermo-Neto and Guimaraes,


2000), histamine release (Peeke et al., 1987; Irie et al.,2001, 2002, 2004), or delayed-type hypersensitivityresponse (Bovbjerg et al., 1987; Roudebush andBryant, 1991; Exton et al., 2000) were affected bylearned immunosuppressive responses. Behavioralconditioning as supportive therapy has also beenstudied in the context of experimental cancer therapies(Hiramoto et al., 1991; Spector, 1996; Bovbjerg, 2003),in which the learned immune response reduced tumorgrowth and prolonged survival time in conditionedtumor-bearing mice (Ghanta et al., 1985, 1988, 1990).Another example of the potential clinical relevance ofthe behaviorally conditioned immune response isillustrated by grafting experiments (Gorczynski et al.,1982). Behaviorally conditioned CsA-immunosuppressiveeffects prolonged the survival time of heterotopicallytransplanted heart allografts in rats (Exton et al., 1998;Grochowicz et al., 1991). Moreover, combining theconditioning procedure with the administration of sub-therapeutic doses of the immunosuppressive drug CsAtogether with daily re-exposure to the CS led to long-term graft survival in 20–30% of the animals (Extonet al., 1999), indicating synergistic effects via a combina-tion of behaviorally conditioned immunosuppression andsubtherapeutic dosing of an immunosuppressive drug.Early observations in humans, namely, that allergic

symptoms can be induced in affected patients despitethe absence of allergens, support the notion thatlearning mechanisms contribute to asthma’s patho-physiology (Turnbull, 1962). This assumption is sup-ported by a study in patients with allergic rhinitis,whereby elevated measures of mast cell tryptase inmucosa were conditioned behaviorally (Gauci et al.,1994). Similarly, allergic subjects re-exposed to anolfactory cue (CS) formerly paired with a grass-allergenchallenge displayed increased histamine release (Barrettet al., 2000). Reversely, the antihistaminergic proper-ties of the H1-receptor antagonist desloratadine wasbehaviorally conditioned in patients suffering fromallergic house-dust-mite rhinitis, whereby the learnedplacebo response significantly reduced subjectivesymptoms, the allergic response to the skin prick test,and basophile activation (Goebel et al., 2008). Thatstudy was recently confirmed and expanded upon bydemonstrating reproducible placebo responses in theallergic response induced by both expectation andlearning (Vits et al., 2013). The effectiveness of theconditioning procedure on another type of allergicreaction (delayed-type hypersensitivity response) wastested in healthy volunteers undergoing monthlytuberculin skin tests. All subjects presented signifi-cantly blunted symptom severity as a result of theconditioning process (Smith and McDaniel, 1983).However, employing a similar approach, those resultscould not be replicated (Booth et al., 1995). Theefficiency of learned immune responses was also testedin patients with multiple sclerosis who receiving

cyclophosphamide infusions continuously paired witha novel taste during the learning phase. Re-exposureto the CS alone during evocation significantly reducedperipheral leukocyte numbers (Giang et al., 1996).Furthermore, by pairing subcutaneous IFN-g injec-tions with a strongly flavored drink (CS), elevatedlevels of neopterin and quinolinic acid serum wereinduced after re-exposing healthy volunteers to the CS(Longo et al., 1999).

Together, these experimental data form a “proof ofprinciple that associative learning protocols may betaken seriously as supportive treatment options duringimmune pharmacological regimens (Schedlowski andPacheco-Lopez, 2010; Doering and Rief, 2012; Encket al., 2013). However, as with other learning pro-cesses, behaviorally conditioned immunosuppression issubject to extinction, that is, the learned immunosup-pressive response gradually decreases over time. Thisconstitutes a considerable problem for the systematicapplication of conditioning paradigms as a treatmentoption supporting immunopharmacological regimens.Attempting to overcome this drawback, a recent studydemonstrated that extinction of the learned immuno-suppressive response (IL-2 production and IL-2 mRNA-expression) in healthy humans can be inhibited bycombining the CS re-exposure during evocation withadministration of subtherapeutic doses of CsA (Albringet al., 2014). Similarly, a recent partial reinforcementschedule in which patients received a full dose ofmedication 25 to 50% of the time and placebo medica-tion the other times significantly reduced the amount ofcorticosteroid needed to treat cutaneous lesions inpsoriasis patients (Ader et al., 2010).

Although the clinical exploitation of conditionedimmune responses is still in a very early stage, thesepreliminary results provide evidence that extinctionprocesses might be overcome by systematically modi-fying learning protocols such as partial reinforcementstrategies (Ader et al., 2010; Doering and Rief, 2012;Hadamitzky et al., 2013; Au Yeung et al., 2014).

E. Neuroendocrine Responses

The experimental designs employed in studies in-vestigating learned placebo responses in neuroendo-crine functions basically resemble the conditioningprotocols used to induce learned placebo responses inimmune functions. During acquisition, the condition-ing group receives the pairing of a CS (e.g., stimuluscompound, injection procedure, novel tasting drink, ornovel olfactory stimulus) and a US (e.g., administra-tion of adrenaline, insulin, dexamethasone, glucose,IFN-b-1a, sumatriptan), which induces alterations inneuroendocrine responses. The experimental group isthen re-exposed to the CS during evocation andalterations in neuroendocrine functions (e.g., concen-trations of adrenaline, glucose, cortisol, insulin, nor-epinephrine, glucagon, vasopressin, ACTH, somatropin)


are analyzed, reflecting the conditioned response.Although learned placebo responses in neuroendocrinefunctions have been demonstrated in experimentalanimals (Ader, 1976; Buske-Kirschbaum et al., 1996;Janz et al., 1996; Pacheco-Lopez et al., 2004), there arefew studies reporting these effects in humans, andthose that do mainly employed insulin as a USmeasuring blood glucose or insulin levels as a con-ditioned response (Fehm-Wolfsdorf et al., 1993;Stockhorst et al., 1999, 2004, 2011; Klosterhalfenet al., 2000; reviewed in Wendt et al., 2014b). Twohuman studies reported conditioned changes in plasmacortisol concentrations. One study observed an increasein plasma cortisol levels by re-exposing subjects toa novel tasting drink (CS) that had been paired withan injection of dexamethasone (US) (Sabbioni et al.,1997). A decrease in cortisol and an increase in growthhormone were observed when sumatriptan was used asa US during the conditioning procedure (Benedetti et al.,2003). These changes in cortisol and growth hormonelevels were induced via the associative learning protocolbut not via mere expectation, resembling the observa-tions in learned immune responses. However, there arecurrently no data on the neurobiological mechanismsmediating these conditioned neuroendocrine responses,which might explain why we are unaware of anysuch protocols that have been transferred to clinicalconditions.This experimental evidence demonstrates the poten-

tial applicability of such behavioral conditioning pro-tocols in clinical practice. However, future studies willhave to analyze the kinetics of the behaviorally condi-tioned endocrine response and to elucidate whether andto what extent these conditioned responses can bereconditioned on multiple occasions. Only with thisinformation and more detailed knowledge of the mech-anisms driving the CNS-endocrine system interactionwill it be possible to design conditioning protocols thatcan be employed in clinical situations to the patients’advantage.

F. Autonomic Organ Functioning

The innervation of most peripheral organs bysympathetic and the parasympathetic branches ofthe autonomic nervous system (ANS) forms the anat-omic and neurophysiological basis of placebo responsesin end-organ functioning (Jänig, 2006). The ANS isunder inhibitory and excitatory control of a number ofcortical and subcortical structures such as the anteriorand mid-cingulate cortices, the insula, dorsolateralprefrontal cortex, amygdala, hippocampal formation,and the hypothalamus (Beissner et al., 2013), brainareas known to mediate placebo responses in pain,nausea, and Parkinson’s disease. Among the ANS-regulated functions, cardiovascular, pulmonary, andintestinal functions have been subjected to placeboinvestigations.

1. Cardiovascular Functions. A drop in bloodpressure is regularly observed in the placebo groupsof clinical trials of hypertension (Preston et al., 2000;Weber, 2008). However, the inclusion of no-treatmentcontrol groups have revealed that these effects seem tobe largely due to confounding factors such as sponta-neous fluctuation and regression to the mean ofhabituation effects (Hrobjartsson and Gotzsche, 2010).

In studies of placebo analgesia, pain relief is ac-companied by a reduction in heart rate and heart ratevariability (Pollo et al., 2003). This placebo response inpain perception and heart rate was naloxone reversiblebut was unaffected by muscarinic blockade withatropine. In contrast, b-adrenoceptor blockade withpropranolol did not affect placebo analgesia but didantagonize a pain-induced heart rate increase, sug-gesting that placebo analgesia is independent fromheart rate changes (Pollo et al., 2003). This notion issupported by observations where placebo analgesiawas associated with a reduction in subjective stresslevels, heart rate, and heart rate variability. Thereduction in sympathetic activity appeared to be afactor involved in anticipatory placebo analgesia ratherthan simply a direct consequence of pain reduction(Aslaksen and Flaten, 2008; Aslaksen et al., 2011).These placebo responses reflected in the heart rateappear to be mediated via the prefrontal cortex, be-cause the impaired connectivity of the prefrontal lobesin Alzheimer patients has been associated with bothabsent placebo analgesia and a heart rate reduction(Benedetti et al., 2006). Whether placebo responsesevident in cardiovascular functions in placebo analge-sia studies are a direct cause of the placebo instructionor secondary to the altered pain perception is still notclear.

Placebo responses on cardiovascular functions out-side the context of pain-related changes in sympatheticactivity have been reported in experimental settings,but they are inconsistent and scarce. Placebo-induceddrops in systolic but not diastolic blood pressure innormo- and hypertensive patients have been reported(Agras et al., 1982; Suchman and Ader, 1992; Amigoet al., 1993). More recently, verbally induced expecta-tion reduced systolic blood pressure without affectingdiastolic blood pressure or other autonomic responses,such as skin conductance and sympathetic and para-sympathetic components of heart rate variability inhealthy volunteers (Meissner and Ziep, 2011). Anotherstudy reported stress responses induced by a placebo-spray administered with the suggestion that it wouldeither raise or lower blood pressure. Within the totalstudy sample, blood pressure was significantly higherafter the placebo spray independent of the associ-ated suggestions (Zimmermann-Viehoff et al., 2013).A systolic drop in blood pressure was evident afterplacebo intake plus verbal suggestion but vanishedwhen participants believed that the medication had


been switched from a trademark to a generic drug(Faasse et al., 2013). Finally, verbal suggestions havebeen shown to affect coronary diameter in chest painpatients. Surprisingly, the verbal suggestion of vaso-dilatation induced significant vasoconstriction but wasassociated with lower pain ratings in this patient groupwithout affecting stress ratings, heart rate, or bloodpressure (Ronel et al., 2011).Employing the open versus hidden infusion para-

digm, the b-blocker propranolol was more effective inreducing heart rate when given by the doctor (open)compared with the computer-infusion (hidden) condition(Colloca and Benedetti, 2005) (for details, see sectionVI). Similarly, the acetylcholine muscarinic antagonistatropine induced a more pronounced increase in heartrate when administered overtly compared with thehidden condition (Benedetti et al., 2003).Taken together, there is evidence of placebo responses

on systolic blood pressure, yet other effects on thecardiovascular system require further clarification.2. Pulmonary Functions. The autonomic nervous

system regulates smooth muscle constriction in therespiratory tract; sympathetic activation via cat-echolaminergic mechanisms induces bronchodilation,whereas bronchoconstriction is predominantly medi-ated via vagal efferents (Canning and Fischer, 2001).Several studies reported bronchoconstriction in asth-

matic patients induced by the use of a fake bronchocon-strictor containing pure saline solution (Isenberg et al.,1992a,b). Verbally induced expectations of broncho-constriction were antagonized by anticholinergic agents,suggesting these nocebo responses are mediated viaenhanced vagal activation of lung functions (Luparelloet al., 1968; Butler and Steptoe, 1986). More recently, inan experimental approach, a placebo bronchodilatorsignificantly reduced nonspecific airway hyperrespon-siveness in asthmatic patients (Kemeny et al., 2007). Incontrast, another study identified clear and pronouncedplacebo responses in subjective asthma symptoms butobserved no placebo responses on objective outcomemeasures as measured with spirometry (Wechsler et al.,2011). Taken together, whether and to what extentsubjective placebo and nocebo responses are associatedwith physiologic lung function parameters will have tobe analyzed in experimental approaches and clinicalinvestigations.Both heart rate and respiratory responses appear to be

affected during placebo analgesia. Patients treated withthe opioid buprenorphine for pain relief after surgerypresented a reduced respiratory response as one commonside effect of the opioid administration. When injectedwith NaCl instead of the opioid, patients reportedplacebo-induced pain relief and displayed a reducedrespiratory response that was antagonized with theopioid antagonist naloxone (Benedetti et al., 1998, 1999).3. Nausea. Nausea was known to demonstrate high

placebo response rates as early as 1955, when Beecher

(1955) cited a study with average placebo responserates of 40% with seasickness in individuals workingon seagoing vessels (Gay and Carliner, 1949). Theemetogetic properties of ipecac, a herbal drug to inducemild nausea in intoxication, was completely blockedby verbal suggestions of nausea relief in fistulatedindividuals (Wolf, 1950). A recent review on laboratoryexperiments in healthy volunteers addressing placeboresponses in nausea concludes that nausea symptomscan be alleviated with placebo interventions based onthe principles of Pavlovian conditioning, manipulatingexpectancies, or a combination of both (Quinn andColagiuri, 2015).

Attempts to relieve nausea by performing Pavlov-ian conditioning build on laboratory and clinical evi-dence that anticipatory nausea, the urge to vomit, tasteaversion, and rotation tolerance can be classicallyconditioned. Experimental techniques aiming to reduceconditioned responding and/or to enhance extinctionmay constitute promising tools leading to effectiveinterventions. Studies in healthy volunteers confirmthat anticipatory nausea is reduced by repetitive pre-exposure to the nausea-inducing environment (“latentinhibition”) (Klosterhalfen et al., 2005) and by pro-viding a different, salient beverage before nauseainduction (“overshadowing”) (Stockhorst et al., 2014).Only a single study to date has conducted conditioning-based interventions in a clinical setting to reducenausea in cancer patients undergoing chemotherapy(Stockhorst et al., 1998), although more work is un-derway in this promising field (Geiger and Wolfgram,2013).

Changing expectations about nausea has been ac-complished in laboratory experiments by the intake ofplacebo pills (Levine et al., 2006) or by deliveringpositive suggestions associated with distinct gustatorystimuli (Klosterhalfen et al., 2009; Weimer et al.,2012). Other studies attempting to modify expectationshave provided counterintuitive, i.e., reverse effects ofsuggestions (Levine et al., 2006) or negative results(Williamson et al., 2004). These inconsistencies may inpart be explained by sex differences (Klosterhalfenet al., 2009; Weimer et al., 2012) and/or complexinteractions between the participant’s sex and that ofthe experimenter (Aslaksen et al., 2007; Weimer et al.,2012). Ultimately, interventions combining the princi-ples of conditioning with optimized expectations maybe most promising for effective alleviation of nauseasymptoms. A combination of positive instructions andsurreptitiously reduced rotation speed in precedingtrials (i.e., conditioning) detected significantly allevi-ated symptoms, fewer nauseogenic head movements,and longer rotation tolerance in a subsequent trial(Horing et al., 2013).

Exploring the central mechanisms of nausea reliefwith placebo (and drugs) is difficult given the risk ofaspiration during brain-scanner investigations with


experimentally induced nausea in supine position(Kowalski et al., 2006). However, recent progress hasbeen made in a virtual-reality environment (Farmeret al., 2015). Also very recent has been the introductionof a novel experimental model that permits greaterfreedom during nausea experiments inside and outsidebrain scanners than the rotation paradigms currently inuse (Quinn et al., 2015). Galvanic vestibular stimulationinduced a sensory mismatch that proved highly effectivein conditioning nausea; it was also independent of otherexperimental context variables. This phenomenon willallow a more in-depth analysis of the central pathwaysmediating the placebo response in nausea in the future.Investigating the peripheral mechanisms underlying

placebo responses in nausea appears much easier.Studies have suggested that expectancies affect boththe behavioral effects of nausea and induce changes ingastric slow-wave rhythm as measured by electrogas-trography (Williamson et al., 2004). However, there isa paucity of findings and many are somewhat contra-dictory (Levine et al., 2006; Weimer et al., 2012; Horinget al., 2013).Efforts have been made to characterize peripheral

neuroendocrine and immune mediators associatedwith stress responses, such as cortisol and cytokines(Stockhorst et al., 1998, 2014; Klosterhalfen et al.,2005; Meissner, 2009). However, replication and ex-tension to healthy volunteers and patients is neededbefore the putative role of the hypothalamic-pituitary-adrenal axis and immune systems in placebo responsesfor nausea can be clarified.The evidence is accumulating that nausea in cancer

patients undergoing chemotherapy can be predicted bypretreatment expectancies (Colagiuri et al., 2011). Todate, the few interventional studies designed to improvenausea in the context of chemotherapy by enhancingpositive expectations have yielded conflicting results(Shelke et al., 2008; Roscoe et al., 2010). This is in linewith similar results originating in efforts to prevent (orreduce) seasickness symptoms by manipulating expect-ations (Eden and Zuk, 1995).

G. Gastrointestinal System/Irritable Bowel Syndrome

Of all the gastrointestinal disorders, functional boweldiseases of the IBS type represent the largest subgroupwith a population prevalence of around 15% in mostWestern countries (Choung and Locke, 2011). Its majorclinical characteristic is abdominal pain in the absenceof macroscopically identifiable organic causes. Thepathophysiology of IBS is largely unknown, but thesearch for biomarkers includes immunologic, endocri-nological, and genetic contributions (Elsenbruch, 2011).Experimental and clinical studies suggest that exper-imental visceral pain is affected by placebo interven-tions (Enck et al., 2012; Elsenbruch, 2014).A recent meta-analysis reported that in 73 eligible

RCTs including 8364 patients with IBS allocated to

placebo, the pooled placebo response rate across allRCTs was 37.5% (Ford and Moayyedi, 2010). Theseobservations are confirmed by experimental data inIBS patients in which augmented practitioner-patientcommunication increases the beneficial effect of pla-cebo acupuncture on symptom severity and quality oflife (Kaptchuk et al., 2008a) and in which an open-labelplacebo produced significantly higher global-improvementscores in these patients (Kaptchuk et al., 2010).

Our understanding of the neurobiological and neuro-psychological mechanisms steering placebo responsesin visceral pain is much more limited comparedwith somatic pain (Zhou and Verne, 2014). This is alsodue to differences in the processing of visceral andsomatosensory signals in the periphery and the brain(Aziz et al., 2000; Eickhoff et al., 2006). Most experi-mental placebo studies in the visceral-pain field em-ploy a rectal distension model in which pressure- orvolume-controlled distension of the gastrointestinalcompartments is carried out, resulting in reliablepain models. Verbal suggestions for pain relief induceda significant reduction in rectal distension-inducedpain intensity in IBS patients (Vase et al., 2003). Theseplacebo responses in visceral pain seemed to correlatenegatively with anxiety and negative emotions butwere not associated with endogenous opioidergic mech-anisms, because naloxone did not affect the placeboresponse (Vase et al., 2005). The brain responses of IBSpatients were analyzed employing glucose-PET imag-ing before and 3 weeks after a placebo intervention,revealing that prefrontal regions (the right ventrolateralprefrontal cortex) predicted self-reported symptomimprovement in these patients with this relationshipmeditated by changes in the dorsal anterior cingulatecortex (Lieberman et al., 2004). Similarly, an fMRIstudy reported that decreases in activity in pain-relatedbrain regions in the placebo condition were related toverbal suggestion and "habituation, attention, andconditioning" (Craggs et al., 2007, 2008; Price et al.,2007).

A series of studies analyzed the role of expectationin visceral placebo analgesia in healthy volunteers(Benson et al., 2012; Elsenbruch et al., 2012a,b; Kotsiset al., 2012; Schmid et al., 2013, 2015; Theysohn et al.,2014). When visceral pain stimuli were delivered witha cover story of receiving with varying likelihood (0%,50%, 100%) an intravenous pain killer, the volunteersreported a “dose-dependent” pain reduction (Elsenbruchet al., 2012a). This placebo analgesia was associatedwith activity changes in the thalamus, prefrontal,and somatosensory cortices, especially in the pain-anticipation phase, consistent with findings in condi-tioned esophagal placebo analgesia (Lu et al., 2010). Inaddition, in the “50% condition,” the placebo-inducedpain relief was more pronounced in those subjects whobelieved they were in the active-treatment group(Kotsis et al., 2012). A direct comparison in the placebo


response to visceral pain between IBS patients andhealthy volunteers demonstrated similar placeboresponses at the behavioral level (magnitude of placeboanalgesia) but a more pronounced neural response inaffective and cognitive brain regions (insula, cingulatecortex, prefrontal cortex), suggesting altered neuralprocessing of placebo-induced changes in pain percep-tion in IBS and in patients with ulcerative colitis inremission (Lee et al., 2012; Schmid et al., 2015).Together, these experimental data demonstrate that

placebo responses in visceral pain on the behavioraland a corresponding neural level are induced pre-dominantly by expectation in healthy volunteers andpatients with IBS, particularly those with gastrointes-tinal disorders (Zhou and Verne, 2014). Interest-ingly, only one study thus far has reported successfulconditioning of visceral (esophageal) analgesia inhealthy volunteers (Lu et al., 2010), whereas placeboanalgesia studies with lower gastrointestinal pain (asin IBS) are still lacking. Until cross-modality compar-isons of placebo effects involving different pain modelsand clinical conditions are available, it is impossibleto conclusively determine whether placebo effects invisceral and somatic pain are similar or whether theyinvolve different pathways. In the meantime, there isevidence to support “specificity” for the visceral domain(Eickhoff et al., 2006).

H. Sleep Disorders

Placebo responses in insomnia can be assessed onthe subjective and objective level. Objectively assessedsleep parameters derived from polysomnographicassessments (PSG) are total sleeping time (TST), sleeponset latency (SOL; time in bed before sleep onset),wake after sleep onset (WASO), and sleep quality(quotient between sleeping time and time in bed).Subjective and objective improvements in sleep

quality in placebo arms of insomnia trials were veryrecently subjected to meta-analyses (Winkler and Rief,2015). Results in the placebo groups accounted for 64%of the subjective benefits reported in the drug groupsand for 54% of physiologically assessed improvements.The effect sizes for objective versus subjective outcomeparameters in the placebo groups did not differ (e.g.,TST d = 0.43 derived from diaries; d = 0.42 derivedfrom PSG). No predictors for strong placebo responseswere identified, although other studies reported largebaseline variance in SOL and TST and longer SOL topredict placebo responses (Ogawa et al., 2011). Thesemeta-analytically described placebo responses are inaccordance with a re-analysis of placebo arms in fourdrug trials, also indicating the long-term persistenceand robustness of positive effects in the placebo arms(Perlis et al., 2005). Other meta-analyses reporteda modest difference between drug arms (Z-drugs) andplacebo arms in insomnia trials, and the differencesin some objective variables (TST, WASO) were even

nonsignificant (Huedo-Medina et al., 2012). To furtherdisentangle the effects of expectation, regression to themean, and social desirability on subjective outcomes inplacebo arms of insomnia studies, studies directlyexamining these influences have been conducted (McCallet al., 2011) that found evidence for the contribution ofall three effects (however, with varying degrees) ondifferent outcome variables. Significant improvementsin subjective TST were also reported for the compar-ison between placebo arms and waiting list groups/notreatment groups (McCall et al., 2005; Belanger et al.,2007).

Although clinical trials only enable limited conclu-sions about the underlying mechanisms, several ex-perimental approaches addressed placebo responsesin insomnia directly. In a pilot study, patients withinsomnia received a placebo pill for the second nightafter baseline assessments of the first night. Althoughthis study did not control the so-called “first-nighteffect” (reduced sleeping time during the first night inthe sleep laboratory), the improvements during thesecond night were larger than would be expected, witha TST increase of 1.3 hours and a WASO reduction to57% of the first-night assessments. Additionally, theworking group reported a trend toward longer rapideye movement sleep periods after placebo intake(Rogev and Pillar, 2013). Their result is in line withan experiment examining whether the first-night effectcan be reduced when placebo pills are administered.Using this laboratory model for insomnia in healthycontrols, the investigators showed that the first-nighteffect is associated with rapid eye movement sleepreductions and that this effect can be counteracted bytaking placebo pills (Suetsugi et al., 2007). Furtherstudies investigating healthy controls with experimen-tally balanced designs confirmed the positive effects ofplacebo pill intake versus no treatment on WASO onobjective and subjective parameters (Fratello et al.,2005). Placebo application can also stimulate alter-ations in circadian rhythms (Reinberg et al., 1994).

Learning has also been shown to influence sleepquality. Alpha/theta neurofeedback was used to learnthe transition into presleep states. This transition isaccompanied by reduced activity in the brain areasresponsible for external monitoring of sensory input(Kinreich et al., 2014), contributing to a reduction inphysiologic arousal.

The neurobiological underpinnings of placebo re-sponses in insomnia are poorly understood. Furtherinvestigation is needed to assess which functional andneurophysiological aspects of sleep regulation beyondPSG/EEG parameters are affected by placebo mecha-nisms. We need to discover which biologic system andphysiologic processes in sleep regulation are vulnera-ble to expectation and conditioning influences (e.g., theascending neuronal system including the reticularformation, thalamus, and forebrain neurons; the


descending neuronal pathways encompassing neuronalactivities in the hypothalamus, brain stem, and spinalcord; neurotransmitter systems such as the gluta-matergic system, GABA, noradrenaline, dopamine, andserotonin sensitive receptors; behavioral arousal andmuscle tone (Jones, 2011). Sleep is closely associatedwith the restoration of physiologic systems, such asnocturnal blood pressure dipping and changes in immunefunctions (Rief et al., 2010; Euteneuer et al., 2014), but itis also relevant for processes such as recreation andmemory consolidation (Weber et al., 2014), and placeboresponses in these sleep-associated factors also need to beaddressed.

V. Neuro-Bio-Behavioral Mechanisms ofNocebo Responses

Nocebo responses describe negative treatment ef-fects that are not directly attributable to the drug’spharmacokinetics. Two variants of these noceboresponses exist: one is characterized by new symptomsor a symptom aggravation associated with drug orplacebo intake, although the chemical agent itself isnot able to trigger these symptoms. Another variationof nocebo responses is the reduced efficacy of clinicalinterventions due to negative expectations or priorexperiences. The underlying mechanisms steeringthese two types of nocebo responses are not usuallydistinguished.

A. Nocebo Effects in Clinical Trials and ClinicalPractice

In double-blinded RCTs, the rate of unwanted sideeffects reported in placebo groups is notoriously high,although these symptoms obviously cannot be thedirect consequences of drug effects. Many patientsdiscontinue drug intake because of unwanted sideeffects, although being in the placebo arm, or becausethe reported adverse effects are in discordance with thepharmacokinetic of the drug (Rief et al., 2006, 2009a;de la Cruz et al., 2010; Stathis et al., 2013).Nocebo mechanisms are also considered responsible

for epidemic waves of reports of adverse events. In NewZealand, public blaming of a thyroxine product ontelevision led to a dramatic increase in adverse eventsreported to the authorities responsible for drugmonitoring (Faasse et al., 2012). The typical problemsobserved when switching from a brand to a genericdrug are related to nocebo mechanisms as documentedin antihypertensive treatments with placebo pills incounterbalanced experimental approaches (Faasse et al.,2013).Several studies have confirmed that the unwanted

side-effect profiles of the placebo arms “mimic” theexpected side-effect profiles of the drug arms. The sideeffects reported by placebo arms in multiple sclerosistreatments have been substantially higher when disease-

modifying drugs were investigated compared with trialsinvestigating symptomatic treatments (Papadopoulosand Mitsikostas, 2010). Similar effects were reportedfrom different groups of migraine drugs (Amanzio et al.,2009) and antidepressants (Rief et al., 2009a). Sucha mimicking effect can hamper adequate detection ofdrug-related adverse events.

B. Psychologic Mechanisms Contributing toNocebo Responses

Negative expectations are powerful determinantsand predictors of side effect development and drugtolerability (Nestoriuc et al., 2010). Research para-digms therefore use verbal suggestion to manipulateexpectations, frequently amplifying this effect throughnegative pre-experiences via Pavlovian conditioning.Although conditioning paradigms are more powerful intriggering placebo effects, both verbal suggestion andlearning induce similar effects on nocebo development(Colloca et al., 2008).

Observational learning such as watching and/orlistening to others who report serious problems aftertaking a specific drug is another powerful tool forsymptom development, sometimes even more power-ful than the mere verbal suggestion of side effects(Vögtle et al., 2013). The social dissemination ofsomatic symptoms was demonstrated elegantly ina recent study (Benedetti et al., 2014). The authorsinformed just one person in a group about hypoxia-induced headache before the group was to be exposedto hypobaric conditions in the Alps. This symptomreport “infected” other group members depending onthe intensity of social contacts with that particularindividual.

C. Neurobiological Pathways of Nocebo Responses

Neuroimaging studies have confirmed that the mereexpectation of symptoms can already activate the brainstructures responsible for symptom perception beforeany sensory stimulation has occurred, thus sensitizingthe person to the perception of discomfort (Koyamaet al., 2005; Keltner et al., 2006). Pain expectation isassociated with significant activations in nociceptiveregions in the thalamus, second somatosensory cortex,and insular cortex. When people expect highly intense,noxious thermal stimuli, significant differences appearin the caudal anterior cingulate cortex, head of thecaudate, cerebellum, and contralateral nucleus cunei-formis compared with the expectation of pain stimuli oflow intensity (Keltner et al., 2006).

Furthermore, although repeated pain stimulationleads to habituation, this effect can be blocked viaverbal instructions such as: “with every assessmentday, pain intensity will increase,” inducing negativeexpectations. This blockage of habituation is associatedwith diverse activities in the brain’s right parietaloperculum (Rodriguez-Raecke et al., 2010).


Different descriptions of a cream as being eithersensitizing for pain perception or neutral has revealeddifferent activities even on the spinal level (Geuter andBüchel, 2013). The authors demonstrated that activa-tion was higher under the nocebo instruction at thelevel of stimulated dermatomes C5/C6 in the spinalcord. Pain-related activity in the ipsilateral dorsal hornof the spinal cord was enhanced. The effect of expec-tancy, stimulus processing, and perception has beendemonstrated across different sensory modalities. Theseeffects are displayed on a behavioral and neuronal level(Summerfield and de Lange, 2014). Because expectationfacilitates the perception of a specific sensation and ofstimulus categories, this effect helps clarify why sideeffects often occur as a cluster of multiple symptoms.From a neurochemical point of view, nocebo

responses have been best characterized in the fieldof pain. Nocebo responses have been associated withvariations in the dopaminergic, opiodergic and chole-cystokinin (CCK) systems (Benedetti et al., 2007). CCKis involved in the activation of descending pronocicep-tive pathways from the midbrain PAG in mediatinganxiety-induced hyperalgesia as well as in the de-velopment and maintenance of hyperalgesia associatedwith peripheral neuropathy (Lovick, 2008). Accord-ingly, blockage of CCK pathways via the CCK antago-nist proglumide leads to the abolition of the noceboresponse (Benedetti et al., 1997).In the study investigating nocebo effects in high-

altitude conditions, Benedetti’s working group re-ported a higher frequency of headache and examinedbiologic trajectories of the nocebo via observationeffect. They identified significantly increased prosta-glandin levels (prostaglandin E2, prostaglandin F2) inthe verbally “infected” group that mediated vasodila-tation and subsequently the occurrence of headache(Benedetti et al., 2014).After examining pain relief caused by a placebo

mechanism versus pain amplification based on noceboeffects, Scott et al. (2008) postulated that these mayjust be different facets of the same pathways. However,the very same person can suffer simultaneously fromnocebo responses while experiencing benefits fromplacebo mechanisms. Even more, nocebo effects canbe (mis-)interpreted as signs of potent drugs, thusamplifying placebo effects. Consequently, moderatedrug-onset effects can amplify placebo responses inanalgesic therapies (Rief and Glombiewski, 2012). Thisshift in the meaning of somatic symptoms from dis-turbing side effect to indexing a powerful treatment isassociated with the activation of opioid and cannabi-noid systems, especially in pain symptoms (Benedettiet al., 2013). However, considering the broad range ofpotential physiologic systems and drugs’ side effects,various exchanges between placebo and nocebo mech-anisms may be occurring for each symptom andphysiologic system (Enck et al., 2008). A recent study

reported distinct patterns of neural activation inducedby positive or negative expectancies, however, thatresulted in a correlated placebo and nocebo behavioralresponse (Freeman et al., 2015).

Taken together, the underlying mechanisms ofnocebo responses are much less well understood thanthose of placebo responses. In particular, the contribu-tion of similar overlapping and distinct trajectoriesmediating nocebo versus placebo responses requirefurther investigation.

VI. The Effect of Placebo Responses onPharmacological Treatments

Placebo responses contribute significantly to clinicaloutcome in most if not all medical treatments includ-ing pharmacotherapy. Patients’ expectations are a keyfactor that codetermines treatment efficacy. Thecrucial role of expectation in the therapeutic outcomeis best illustrated in the so-called open/hidden drugparadigm. In this paradigm, identical concentrations ofthe same drug are administered under two conditions:an open condition in which the patient is aware of thetime point at which the medication is administered bya health care provider and of the intended treatmentoutcome (e.g., analgesia) and a hidden condition inwhich the patient is unaware of the medication beingadministered by a computer-controlled infusion. Thisparadigm enables the dissociation of the treatment’sgenuine pharmacodynamic effect (hidden treatment)from the additional benefit of the psychosocial contextin which the treatment is provided. Studies based onan open/hidden paradigm have revealed that psycho-social factors such as the awareness of a drug beinggiven can considerably enhance its analgesic effect(Levine and Gordon, 1984). Conversely, the hiddenadministration attenuates the analgesic effect of non-steroidal anti-inflammatory drugs to nonsignificance,and even the effects of opioids are substantially reducedby hidden application (Colloca et al., 2004). Because ofits hidden application, the drug dosage had to be doubledto achieve the same result as during open application, afact highlighting the economic relevance of placeboresponses.

This phenomenon is not limited to analgesics, be-cause similar pharmacotherapeutic effects have alsobeen reported in other domains, such as motor functionin Parkinson’s disease and anxiety-related disorders(Amanzio et al., 2001; Colloca et al., 2004). Findingsfrom these studies using the open/hidden drug para-digm are supported by investigations that explicitlymodulated the expectancy concerning a given drug byverbal instructions (Lyerly et al., 1964; Kirk et al.,1998; Metrik et al., 2009). The detrimental influence ofnegative expectations on the drug response became, forinstance, apparent in a behavioral experimental studyby Dworkin et al. (1983), who reported a reversal of


analgesia by nitrous oxide in dental pain whenparticipants expected the drug to increase awarenessof bodily sensations. Although this is not the focusof this review, it should be noted that substantialexpectation effects are also observed in nonpharmaco-logical interventions such as acupuncture (Linde et al.,2007).Similarly, learning effects modulate the response to

pharmacological treatments (see above). The effect ofprior experience on treatment outcome is evident inresults from crossover studies (i.e., drug trials inwhich subjects receive an active drug and a placeboin randomized order), showing that the order in whichthe placebo or active treatment is given substantiallymodulates the treatment response. Generally speak-ing, the response to the active treatment is strongerwhen active treatment is given first and weakerwhen it follows the placebo treatment. This can beexplained by the fact that the treatment experience,which can be assumed to be less substantial duringplacebo, carries over into the active treatment andvice versa. For instance, in a double-blind crossoverstudy on the effect of an antihypertensive drug, theplacebo did not affect the patients’ hypertension whenadministered first (Suchman and Ader, 1992). Yet,when placebos were administered after a week-longuse of atenolol, they produced a significantly greaterantihypertensive response than the "no treatment"condition. Sequence effects of drug efficacy have alsobeen proposed in conjunction with analgesic therapiesfor musculoskeletal pain (Batterman and Lower,1968), indicating that not only was the placebomedication more effective after active and effectivetreatment but also that the active drug was lesseffective when it followed an ineffective placebotreatment. Similar sequence effects were recentlyreported in conjunction with the analgesic effect ofactive compared with sham repetitive transcranialmagnetic stimulation for chronic pain (Andre-Obadiaet al., 2011). Observations from clinical trials wererecently corroborated experimentally in an analysisof whether prior experience with one treatmentcan influence the response to a second treatment byemploying experimental models of two analgesictreatments in healthy volunteers. In comparisonwith those subjects with a positive treatment history,participants who experienced no analgesia from thefirst treatment continued to show a substantiallyreduced response to a second, different analgesictreatment. These behavioral effects were substantiatedin neuroimaging data showing significant activationdifferences in brain regions coding pain and analge-sia between groups (Kessner et al., 2014). In sum,behavioral studies have proven that the effect of drugtreatment is determined by the drug’s pharmacologicalprofile and (at least partially) by expectancy and learn-ing mechanisms.

A. Neural Mechanisms Underlying the Effect ofExpectations on Drug Efficacy

Recent studies have begun exploring the effects ofexpectation and prior experience on opioid analgesia byusing functional brain imaging. Analgesia was studiedin response to the opioid remifentanil under threeconditions: without expecting analgesia (hidden appli-cation), expecting a positive analgesic effect, and notexpecting analgesia, i.e., the expectation of hyper-algesia (Bingel et al., 2011). Results show that thepositive treatment expectancy doubled the analgesicbenefit of remifentanil, whereas the negative treat-ment expectation interfered with the analgesic poten-tial of remifentanil so severely that its analgesic effectwas completely abolished. Importantly, these changesin pain perception were accompanied by significantalterations in the neural response to noxious thermalstimulation in core brain regions of the pain andopioid-sensitive brain networks such as the thalamus,mid-cingulate cortex, and primary somatosensorycortex, brain areas that have consistently displayedcorrelations with the intensity of nociceptive input andresultant pain perception (Apkarian et al., 2005;Tracey and Mantyh, 2007), and may therefore serveas an objective index of analgesic efficacy (Bingel et al.,2011). With respect to underlying mechanisms Bingelet al. observed that the individual benefit from positivetreatment expectancy during remifentanil analgesiawas associated with activity in the descending painmodulatory system, including cingulofrontal and sub-cortical brain areas, resembling mechanisms of placeboanalgesia. In contrast, the negative expectancy thatabolished the opioid’s analgesic effect was selectivelyassociated with increased activity in the hippocampusand medial prefrontal cortex. These brain areas havebeen implicated in the exacerbation of pain by moodand anxiety in patients and healthy controls (Ploghauset al., 2001; Schweinhardt et al., 2008).

These initial experimental data on the expectancymodulation of opioid analgesia substantiate the sig-nificant contribution of cognitive factors to the overallbenefit from pharmacological treatments. Similarinteractions between pharmacodynamic and psycho-logic effects on regulatory brain mechanisms have beenreported in conjunction with methylphenidate admin-istration in cocaine-addicted patients (Volkow et al.,2003). Together with evidence from experimentalplacebo and nocebo studies, these findings reveal thatthe effects of expectancy and prior experience convergewith pharmacological effects within the very samebiologic systems, involving distinct CNS and periph-eral physiologic mechanisms. However, despite theunequivocal effect of expectancy and learning on phar-macological treatments, research into its underlyingmechanisms remains at a very early stage, leaving manyunanswered questions.


B. Additive versus Interactive Effects

One of the crucial questions still unanswered iswhether cognitive effects and pharmacologically in-duced analgesia combine in an additive or interactivemanner (Fig. 5). This is the crucial basic assumptionbehind double-blinded RCTs, postulating that thedifference between drug and placebo arms reveals the“real” drug effect. However, depending on the drug,exogenous, pharmacologically induced mechanismsand endogenous cascades triggered by expectancy,learning, and their combination may combine in anadditive manner with one substance but combineinteractively with another. One approach to investi-gate additive versus interactive effects of drug and pla-cebo is to adopt a balanced placebo design (Rohsenowand Marlatt, 1981), in which the factors expectancyand drug alternate in a 2 � 2 factorial manner.Employing just such a design, Atlas et al. (2012) ob-served additive rather than interactive effects in con-junction with the influence of positive expectation onremifentanil analgesia. Their results were substan-tiated by a complementary fMRI study investigatingpain-related responses during the open and hiddenadministration of varying remifentanil dosages, alsoallowing inferences about potential interactions: nointeractive effects of drug and expectation were ob-served at the neural level. This finding does not ex-clude the possibility that other drugs, i.e., nonopioidanalgesics show interactive analgesic effects withexpectation-induced analgesia. Future neuroscientificinvestigations involving neuroimaging of the influenceof expectation and learning mechanisms on pharma-cological treatments constitute a new and promisingavenue of research. Instead of studying the effect of oneof them in isolation by controlling the other, it is timewe unraveled how both mechanisms combine on theneurobiological level. Deeper neurobiological under-standing of their potential interaction promises to

ultimately optimize treatment outcomes, encouragingthe development of personalized treatment strategies.

VII. Predictors of Placebo and Nocebo Responses

The presence of placebo responses varies tremen-dously in both healthy volunteers and patients inexperimental and clinical settings. The individualplacebo response can range from no effect ("nonres-ponders") to profound changes in symptom or diseaseseverity ("responders"). Given that placebo responsescontribute so substantially to the overall treatmentoutcome, knowledge about the individual magnitude oftheir occurrence would not just guide therapeuticdecisions to optimize treatment outcomes in clinicalpractice, it could also help to clarify poorly understoodinconsistencies in clinical trials (Enck et al., 2013).

How knowledge of an individual’s placebo re-sponsiveness may inform clinical decision making ishighlighted in an open/hidden study of local anesthesiain patients suffering from Alzheimer’s dementia (Benedettiet al., 2006). These patients, neurobiologically charac-terized by impaired connectivity of the prefrontal lobes(known to be placebo-relevant brain areas) presentedreduced pain relief from the open compared with thehidden application of lidocaine, a local anesthetic. Lossof their ability to form expectations reduced overalltreatment efficacy, and they needed dose increases toexperience adequate analgesia. This study illustratesthat the individual contribution of placebo mechanismsto therapeutic outcome is, at least in part, determinedby an individual’s neurobiological make-up; it highlightsthe necessity to adjust drug treatment approachesdepending on the individual predisposition for placeboresponses.

Seeking deeper knowledge of individual traits andstates susceptible to placebo responses, currently majoreffort is undertaken to identify any psychologic or

Fig. 5. The additive model assumes unspecific effects of equal size in both the placebo and drug arm or groups of studies or experiments. Theinteractive model, however, suggests that drug-specific effects interact with the placebo responses and will result in unequal placebo effects in the twostudy arms or experimental groups. (Modified from Enck et al., 2013, with permission.)


physiologic, neurobiological, or genetic variables thatmoderate individual placebo responsiveness over andabove the known established contribution of expectancyand learning mechanisms in different physiologic systemsand diseases.Along with the cumulative knowledge about the

underlying mechanisms, interindividual differencesare also best examined in conjunction with placeboanalgesia. Recent evidence supports the putative rele-vance of several psychosocial variables for an individu-al’s placebo responsiveness. These include anxiety andfear (Flaten et al., 2011; Lyby et al., 2011), hypnoticsuggestibility (De Pascalis et al., 2002), locus of controland self-efficacy (Horing et al., 2015), hostility (Peciñaet al., 2013a), coping abilities (Schneider et al., 2006),dispositional optimism (Morton et al., 2009; Geerset al., 2010), and empathy for socially induced placeboanalgesia (Colloca and Benedetti, 2009; Hunter et al.,2014). However, no consistent personality profile hasemerged so far across different clinical conditions(Horing et al., 2014).Genetic traits are also being increasingly explored

regarding their contribution to an individual’s pla-cebo analgesic responsiveness, a factor documentedin patients with IBS (Hall et al., 2012) assigned to oneof three treatment arms: no treatment (waiting list),placebo treatment with limited patient-health prac-titioner interaction, and placebo treatment withaugmented supportive patient-health practitionerinteraction. The primary outcome, namely the changefrom baseline in the IBS-Symptom Severity Scaleafter 3 weeks of treatment, correlated with thenumber of methionine alleles in the catechol-O-methytransferase gene Val158Met polymorphism(rs4633), which strongly influences endogenous dopa-minergic and opioidergic pathways. Patients who wereMet/Met homozygotes revealed the strongest placeboanalgesic effect under the augmented placebo arm,whereas those who were Val/Val homozygotes wereless responsive to warm and caring physicians andthus benefited minimally from placebo responses (Hallet al., 2012). An earlier search for endocrine andimmune biomarkers of the placebo response in thesame data set yielded inconsistent findings. Of the 10biomarkers tested, only one marker proved significantfor the placebo response (i.e., osteoprotegerin), butbecause the analysis did not correct for multipletesting, this result may reflect a Type I error (Kokkotouet al., 2010).Peciña et al. (2014) demonstrated an association in

healthy volunteers between the polymorphism in them-opioid receptor gene (OPRM1) A118G and placebo-induced changes in m-opioid binding potential (mea-sured with opioid-ligand PET) in the anterior insula,amygdala, nucleus accumbens, thalamus, and thebrain stem as well as placebo-induced changes in mood.Intriguingly, genetic variation in the endocannabinoid

system (in the gene coding fatty acid amide hydrolase,the major degrading enzyme of endocannabinoids) wasalso linked to the placebo analgesic response in thesame dataset (Peciña et al., 2014).

Furthermore, an individual’s brain anatomy (in-cluding structural and functional measures) predictsthe capacity for placebo analgesic responses in healthysubjects. Stein et al. (2012) indicated that white matterintegrity in the dorsolateral prefrontal and rostralanterior cingulate cortices and their pathways to theperiaqueductal gray are positively associated withindividual placebo analgesic responses in healthysubjects. This finding supports the importance ofstructural brain connectivity in determining the in-dividual ability to form placebo (analgesic) responses.Along these lines, resting-state functional connectivitybetween prefrontal and insular/parietal cortices isalso known to predict individual placebo analgesicresponses in patients suffering from chronic low backpain (Hashmi et al., 2012) and the expectancy-relatedmodulation of pain in healthy volunteers (Kong et al.,2013).

Psychologic, physiologic, and neurobiological predic-tors of placebo responses have also been identified forplacebo responses in other physiologic systems anddiseases. Placebo responses in anxiety disorders anddepression, for instance, have been correlated withpsychologic and genetic traits. An association withthe locus of control has been suggested for placeboresponses in depression (Reynaert et al., 1995), whereascultural variations seem to influence placebo responsesin anxiety disorders (Moerman, 2000). Serotonin-relatedgene polymorphisms have been found to influence theindividual placebo response in social anxiety at thebehavioral and neural level (Furmark et al., 2008), andgenetic polymorphisms modulating monoaminergic tonehave been related to degree of placebo responsiveness inmajor depressive disorder (Leuchter et al., 2009).Regarding learned placebo responses on immune func-tions, state anxiety and plasma noradrenaline levelswere identified as predictors for a learned placeboresponse in cytokine release (Ober et al., 2012) (see alsosection IV).

Beyond the known contribution of psychologic factorsto nocebo responses (e.g., general drug sensitivity,negative drug effects in the past, anxiety, etc.), little isknown about the neurobiological predictors of noceboresponses. Wendt et al. (2014a) reported more drug-specific and drug-unspecific side effects in healthymale subjects with Val/Val variations of the Val158Metpolymorphisms of the catechol-O-methytransferasegene. Their observations further confirm a link be-tween the dopaminergic system and the occurrence ofnocebo responses, as the valine genotype is associ-ated with a 3–4 times higher catabolic rate of braindopamine than the methionine variant (Lachmanet al., 1996).


The search of predictor variables in placebo andnocebo responses remains in a very preliminary stage.Studies have often enrolled small cohorts, which mightexplain the heterogeneous results (Kaptchuk et al.,2008b). Furthermore, it is statistically complicated tosufficiently dissociate the contribution of state fromtrait variables. Finally and most importantly, we donot know whether and, if so, how placebo responsive-ness in one system influences the placebo responsive-ness in another; large placebo responses in onecondition do not necessarily predict large placeboresponses in other conditions as well (Whalley et al.,2008). A deeper and more detailed understanding ofthe neurobiological and physiologic mechanisms ofplacebo responses will help unravel the shared anddistinct pathways and predictors for placebo responsesacross different system and conditions (Zhang et al.,2011).

VIII. Relevance and Implications ofPlacebo Responses

A. Randomized Clinical Trials

The use of placebos in RCTs is without a doubtnecessary to develop new drugs, despite the currentpreference of drug-approval authorities for head-to-head trials comparing novel compounds to drugs alreadyon the market (comparative effectiveness research[CER]). CER requires that more patients be includedfor noninferiority statistics (Gardiner et al., 2000).However, CER trials raise the placebo responsewithout being able to assess it (Weimer and Enck,2014). Some propose that future drug trials bedesigned as three-arm trials: the drug under investi-gation, a comparator (the best drug available or on themarket), and a placebo (Hida and Tango, 2011). Thiswould increase the likelihood of patients receivingeffective treatment, in case of 1:1:1 randomization, toa 66% drug chance compared with conventional 50% in1:1 drug:placebo randomized RCTs. Such “enrichmentdesign,” however, may increase the drug and placeboresponses in some conditions (depression, schizophre-nia, migraine) (Diener et al., 1999; Papakostas andFava, 2009; Rutherford et al., 2009) but not in otherssuch as IBS (Ford and Moayyedi, 2010; Elsenbruch andEnck, 2015). At the same time, it accommodates ethicalarguments that more patients require and receiveeffective medication (World Medical Association, 2013).Two major critical issues arise from the use of

placebos in RCTs that may also apply to CER: the riskof unblinding and the control for the spontaneousvariation of symptoms.The traditional concept to reduce between-subject

data variance is the use of crossover designs that carrythe intrinsic risk of unblinding because patients mayidentify the drug and the placebo phase by di-rectly comparing side-effect profiles rather than effects

(Boutron et al., 2006; Machado et al., 2008), whereas inparallel-group designs, this risk is much lower. At thesame time, crossover studies risk results being affectedby conditioning (learning) effects (Suchman and Ader,1992; Colloca, 2014).

In the early 21th Century, RCTs occasionallyemployed active placebos that mimic adverse events(e.g., in treatment trials of depression) to prevent orreduce such unblinding (Moncrieff et al., 2004), andsuch unblinding caused by adverse events substantiallydetermines treatment results (Rief and Glombiewski,2012). However, overall poorer discrimination betweendrug and active placebo will result in larger patientnumbers to prove a drug’s superiority over placebo,making them expensive. As an alternative, parallelgroup designs are currently the gold standard (Weimerand Enck, 2014).

Onset and offset effects may likewise enable theunblinding of study-arm allocation (Bello et al., 2014),an effect that is minimized by randomized run-in and/or withdrawal designs where patients are enrolled and/or withdrawn from active treatment with variable (butdouble-blinded) periods of placebo application beforeand thereafter, respectively (Ivanova and Tamura,2011; Enck et al., 2013).

Another more recently reported risk for RCTunblinding are social networks where patients enrolledin the same or other studies communicate their ex-perience. Via communicating effects and adverse eventprofiles of the drugs under investigation, patients mayeasily determine the treatment arm to which they havebeen allocated (Lipset, 2014). It has been suggested thatrather than allowing this in uncontrolled manner,investigators should try to incorporate network commu-nication into their patient management strategy in RCTs(Nicholas et al., 2013).

A meta-analysis of RCTs of IBS patients demon-strated that the higher overall adverse event profiles ofsome drugs (e.g., spasmolytics, tricyclic antidepres-sants) in comparison with others (e.g., local antibiotics,5-HT3 antagonists) resulted in higher overall drugefficacy (drug minus placebo). These observationsargue for studies to be unblinded—for the investiga-tors, but not for individual patients (Shah et al., 2014),something that clearly calls for improved blindingmethods, e.g., the re-implementation of “active placebos”(Edward et al., 2005; Rief and Glombiewski, 2012).

Finally, placebo responses may also be driven bypretrial beliefs in the drug being tested: In the mockinformed-consent forms of three putative trials testingeither an antidepressant (desipramine), a prokinetic(alosetron), or a local antibiotic (rifaximin), IBS patientsexpected the highest efficacy from treatments withwhich they were familiar (i.e., antibiotics), which theauthors called a “pre-cebo” effect, i.e., preconceivednotions from being familiar with the drug class (Kimet al., 2012).


To assess the placebo response in RCTs, it would beimportant to know how much of the overall placeboeffect is attributable to spontaneous symptom varia-tion and the disease’s natural course (Enck et al.,2013). Attempts have been made to estimate thiscontribution by comparing placebo controls to “notreatment controls” in trials where patients have beenallocated to a waiting list. Such a comparison is quitecommon under conditions where a placebo controlgroup is not readily available for methodologicalreasons, e.g., in psychotherapy trials. When waitingcontrol and placebo groups are compared, about 50% ofthe placebo effect is attributable to spontaneousvariation of symptoms, at least in the clinical con-ditions tested such as depression or nausea (Krogsbollet al., 2009). However, in most cases, especially incases of severe disease, a “no treatment” or waitingcontrol group would be unacceptable for ethical reasons(World Medical Association, 2013). Waiting controlsalso develop their own dynamics; on the one hand,patients may improve while waiting because of beingassured of receiving active treatment soon (Beck et al.,2015), similar to the effects seen in placebo run-inphases (Enck et al., 2009). On the other hand, it may beregarded as a punishment for patients not to beincluded in the active treatment condition (Furukawaet al., 2014). If waiting list controls are included,a “step-wedge” design may be superior, because itallows a “dose-response” assessment of waiting fortreatment (Fig. 6) (De Allegri et al., 2008; Weimer andEnck, 2014).Despite these complex considerations it remains

crucial to assess and dissociate the spontaneous coursesof disease from placebo and drug responses. The “cohortmultiple randomized controlled trial” (Relton et al.,2010) is a reinvention of the “Zelen design” (Zelen,1979) that recruits patients for a monitoring study (e.g.,via a patient registry) before recruiting and randomiz-ing a subgroup of the same patients for a placebo-controlled or CER study (Fig. 6). A number of approacheshave been published (Cockayne et al., 2014). Finally,initial attempts have been made to avoid involvingpatients that receive placebo and instead use historicplacebo controls from large trial databases (Desai et al.,2013).Growing knowledge about the mechanisms of pla-

cebo and nocebo responses is accompanied by a numberof critical consequences concerning the conduct ofRCTs, of which the two discussed above are major.First, if trials are not truly blinded for both patientsand doctors, the development of novel therapies isjeopardized. Second, if the natural courses of diseasesymptoms are not controlled, the validity of such noveltherapies is questionable. Third and most importantly,knowledge about the moderating and mediating factorsof placebo and nocebo responses enables us to asses,control, and homogenize these responses in a clinical

trial setting and thereby promises to improve assaysensitivity. For further details regarding these strate-gies, see Enck et al. (2013).

B. Training Health Care Professionals

Given the crucial contribution of placebo and noceboeffects to treatment outcomes, health care profession-als must be trained to comprehend the mechanisms ofaction of therapeutic interventions. Careful assessmentof patients’ expectations and pretreatment experiencesshould be incorporated when documenting medicalhistories. The physician should be able to judge whethertreatment expectations are helpful or dysfunctional andwhether pretreatment experiences might limit thetreatment response and/or make nocebo effects likely.Special emphasis must be given to patients who fulfillone of these risk factors for a suboptimal treatmentresponse.

Fig. 6. (A) The so-called step-wedge design is a modified waiting-listcontrol strategy. Patients are randomized to more than one waiting armwith variable waiting periods, enabling assessment of the dose-responsefunction of waiting for treatment. (Modified from Enck et al., 2013, withpermission.) (B) The so-called Zelen design separates the recruitment ofpatients for an observational study from recruitment for an interven-tional study, enabling the natural course of disease to be monitoredwithout randomizing patients to a no treatment control group. (Modifiedfrom Enck et al., 2013, with permission.)


At present, very few psychologic interventions havebeen evaluated that aim to optimize patient’s expect-ations and to minimize the risk of nocebo develop-ment. One successful approach assessed how to optimizepatients’ expectations after experiencing myocardialinfarction, replicated with an additional tool targetingpartners’ treatment expectation as well (Petrie et al.,2002; Broadbent et al., 2009). An optimal therapeuticrelationship seems to provide a solid basis from which tobenefit from placebo responses and to avoid noceboresponses (Kaptchuk et al., 2008a). Of note, even theserelationship effects are becoming better understoodthanks to neuroimaging: there is evidence that painstimuli are better tolerated in the presence of pictures ofsympathetic people and that this effect is associatedwith increased activity in the ventromedial prefrontalcortex (Eisenberger et al., 2011).The effects of treatment value (Waber et al., 2008;

Espay et al., 2015) and open applications reveal thenecessity to express the expected positive effects ofthe intervention and to draw patient’s attention to thevisible cues of open treatment applications. Finally, therole of observational learning can be used moresystematically. Role models (e.g., per video clips) couldbe presented to patients, expressing some pretreat-ment concerns but then confirming positive treatmenteffects. These models for observational learning couldbe selected according to mechanisms that facilitateobservational learning (e.g., similarities in age, gender,general attitudes).The systematic use of Pavlovian learning to improve

medical interventions needs further evaluation, de-spite initially promising results (Ader et al., 2010). Theuse of placebo-controlled drug reduction (Doering andRief, 2012) in particular promises to maintain drugefficacy despite a systematic reduction in drug dosage(Albring et al., 2014). Applying effective pretreatmentswith low side effects is another option to furtheroptimize patients’ treatment expectations.In addition to optimizing treatment effects for the

patient’s benefit, we now know that authentic andempathic doctor-patient communication protects fromunwanted side effects. For instance, medical jargon islikely to cause misunderstandings and trigger fear inpatients. A patient-centered communication style istherefore required when explaining diagnostic pro-cedures, their results, and the rationale and imple-mentation of any intervention (Barsky et al., 2002;Colloca and Finniss, 2012).Nocebo research has clinical implications for the

prevention of side effects. Health care professionalsshould be trained to offer balanced information onexpected positive treatment outcome and the risk ofside effects. Moreover, the patient’s own ability to copewith side effects (should they develop) should beemphasized. Although internet portals usually offerthreatening information, new Web-based information

systems should be developed and approved that offervaluable and accurate information (see below). Finally,if withdrawal symptoms are likely, “hidden with-drawal” procedures could be used to minimize negativeexpectation effects (for further recommendations, seeColloca and Finniss, 2012; Bingel, 2014).

Taken together, knowledge about placebo and noceboresponses and their underlying mechanisms should beintegral elements in the curricula for health careprofessionals. Furthermore, possessing the skills tomaximize placebo responses while minimizing noceboresponses is essential from a clinical applicationperspective.

C. Health Care System

The past decades have seen groundbreaking advan-ces in the development of new diagnostic tools andtreatments. However, this progress has also led toincreased specialization and shorter consultation timeand thus to frequently low therapeutic alliance dueto an inadequate doctor-patient communication, asreflected in the report that in the United States, 50% ofpatients leave their doctor’s office without havingadequately understood what their physician told them(Bodenheimer, 2008). In addition, patients explainingtheir problem to a physician were interrupted after anaverage of 23 seconds (Bodenheimer, 2008); similarnumbers have been reported for primary care practicesin European countries (Deveugele et al., 2002). Thesedisturbing findings are certainly not due to uncommu-nicative practitioners in general but because doctorsare often overstressed by having to see too manypatients in too little time. In addition, medical carecompensation systems such as the “diagnosis-relatedgroup” raise pressure on clinicians by often leavingthem with too little time for adequate doctor-patientcommunication. Serious communication takes time,which calls for a fundamental reorganization of re-imbursement structures in medical care to signifi-cantly upgrading the time spent communicating withthe patient.

In addition to improving the basic conditions foradequate communication, improved patient informationsystems such as drug information leaflets (packageinserts) should be designed to reduce negative expect-ations regarding unwanted treatment side effects(Bingel, 2014). At present, all potential adverse eventsmust be listed for legal reasons and in standardizedterminology, although the empirical evidence of acausal link between drug and unwanted side effectsis notoriously weak. Instead, lay and patient-orientedlanguage and the description of positive and unwantedeffects should be an integral part of any informationleaflet. This should include ways to convey abstractinformation (such as the probability of the occurrenceof side effects) in an intuitive fashion—a futurechallenge for the drug authorities.


Harnessing placebo mechanisms strategically withinhealth care systems promises to fundamentally im-prove the effectiveness of treatment strategies byoptimizing drug treatment regimens, drug efficacy,drug adherence, and the context characteristics withinthe overall medical setting. Ultimately, these approachesshould lower health care costs and improve patient care.

IX. Placebo Research: What Next?

The advances during the last two decades in ourunderstanding in the neurobiological and neuropsycho-logical mechanisms of placebo and nocebo responseshave been made possible by the employment of sophis-ticated experimental designs and tools, such as neuro-imaging, in vivo receptor binding, and single-neuronrecording in awake subjects and patients. However, ourknowledge about the mechanisms underlying theseresponses remains limited and several key issues needto be addressed in future research. We have to learnwhich medical conditions and physiologic systems areaffected by which placebo mechanisms and whetherthese effects are clinically relevant. For example,somatic placebo analgesia can be mediated by cognitivefactors such as patients’ expectation about a treatment’sbenefit, as well as by associative learning processes(Colloca et al., 2008; Tracey, 2010a; Colloca, 2014),whereas visceral placebo analgesia seems less respon-sive to conditioning. In contrast, neuroendocrine orimmune function appears to be affected primarily byconditioning and not by patient expectation (Wendtet al., 2014b). We also need to discover when placeboand nocebo responses occur and to analyze the specificsituational circumstances and patient characteristicsthat are particularly amenable to placebo or noceboresponses (Kaptchuk et al., 2008b). This knowledgeabout genetic and/or psychologic predictors will formthe basis from which to exploit placebo responses andavoiding nocebo responses in daily clinical routine.How placebo responses work remains a key issue,

because we need to better understand the brainmechanisms at both the macroscopic level (in particu-lar, the brain regions and their interactions withperipheral physiologic functions), as well as on themicroscopic (cellular and molecular) levels involved.The data demonstrating that placebo responses canaffect pharmacological treatments introduces fascinat-ing clinical perspectives. It will be important to betterunderstand the interactions between placebo mecha-nisms and pharmacological effects to exploit thisphenomenon in daily clinical care for the patient’sbenefit.Similarly, thorough knowledge of the basic mecha-

nisms steering the behavioral conditioning of pharma-cological responses will be essential, not just to betterunderstand the brain mechanisms involved in theselearning processes but especially to achieve the long-

term goal of learned pharmacological responses: toemploy these learning paradigms in clinical situationsas supportive therapy together with standard pharma-cological regimen, the aim being to maximize thetherapeutic outcome (Doering and Rief, 2012; Wendtet al., 2014b). Finally, another fascinating scientificquestion is why did placebo responses develop duringevolution (Kaptchuk, 2002, 2011)?

X. Conclusion

Placebo and nocebo responses are mediated byexpectations, associative learning processes, and thequality of the patient-physician interaction. Theymodulate fundamentally symptom perception, thecourse of diseases, and the efficacy and tolerability ofmedical treatment. Converging evidence from experi-mental and clinical studies has demonstrated that thesepositive and negative effects on health outcomes arebased on complex neurobiological phenomena involvingthe contribution of distinct CNS as well as peripheralphysiologic mechanisms.

Recent insights into the psychologic, neurobiolo-logical, and peripheral-physiologic processes under-lying placebo and nocebo responses provide theunique opportunity to systematically modulate theseresponses depending on the context in which theyoccur. In clinical care, the systematic use of placebomechanisms is a promising target to improve healthoutcomes. In the context of RCTs, placebo and noceboresponses represent a substantial risk of bias thathampers drug development and assay sensitivity.Strategies to homogenize and control placebo re-sponses and to prevent nocebo responses in a strategy-based manner promise to improve drug developmentby increasing the sensitivity to detect drug-specificeffects and improve the efficacy and tolerability ofdrug treatment regimens, drug adherence, and con-text characteristics within the general medicalsetting.

The full potential of these strategies criticallydepends on continued characterization of the neurobi-ological and peripheral-physiologic mechanisms un-derlying placebo and nocebo responses and, importantly,identification of the predictor variables that influence anindividual’s placebo and nocebo response in a context-and disease-specific manner. This knowledge will buildthe foundation for the exploitation of placebo and noceboresponses based on mechanism-based and personalizedclinical decisions.

Authorship Contributions

Wrote or contributed to the writing of the manuscript: Schedlowski,Enck, Rief, Bingel.

ReferencesAder R (1976) Conditioned adrenocortical steroid elevations in the rat. J CompPhysiol Psychol 90:1156–1163.


Ader R (2003) Conditioned immunomodulation: research needs and directions. BrainBehav Immun 17 (Suppl 1):S51–S57.

Ader R and Cohen N (1982) Behaviorally conditioned immunosuppression and mu-rine systemic lupus erythematosus. Science 215:1534–1536.

Ader R, Mercurio MG, Walton J, James D, Davis M, Ojha V, Kimball AB,and Fiorentino D (2010) Conditioned pharmacotherapeutic effects: a preliminarystudy. Psychosom Med 72:192–197.

Agid O, Siu CO, Potkin SG, Kapur S, Watsky E, Vanderburg D, Zipursky RB,and Remington G (2013) Meta-regression analysis of placebo response in anti-psychotic trials, 1970-2010. Am J Psychiatry 170:1335–1344.

Agras WS, Horne M, and Taylor CB (1982) Expectation and the blood-pressure-lowering effects of relaxation. Psychosom Med 44:389–395.

Albring A, Wendt L, Benson S, Nissen S, Yavuz Z, Engler H, Witzke O,and Schedlowski M (2014) Preserving learned immunosuppressive placebo re-sponse: perspectives for clinical application. Clin Pharmacol Ther 96:247–255.

Albring A, Wendt L, Benson S, Witzke O, Kribben A, Engler H, and Schedlowski M(2012) Placebo effects on the immune response in humans: the role of learning andexpectation. PLoS ONE 7:e49477.

Amanzio M and Benedetti F (1999) Neuropharmacological dissection of placebo an-algesia: expectation-activated opioid systems versus conditioning-activated specificsubsystems. J Neurosci 19:484–494.

Amanzio M, Corazzini LL, Vase L, and Benedetti F (2009) A systematic review ofadverse events in placebo groups of anti-migraine clinical trials. Pain 146:261–269.

Amanzio M, Pollo A, Maggi G, and Benedetti F (2001) Response variability toanalgesics: a role for non-specific activation of endogenous opioids. Pain 90:205–215.

Amigo I, Cuesta V, Fernández A, and González A (1993) The effect of verbalinstructions on blood pressure measurement. J Hypertens 11:293–296.

André-Obadia N, Magnin M, and Garcia-Larrea L (2011) On the importance of pla-cebo timing in rTMS studies for pain relief. Pain 152:1233–1237.

Apkarian AV, Bushnell MC, Treede RD, and Zubieta JK (2005) Human brainmechanisms of pain perception and regulation in health and disease. Eur J Pain 9:463–484.

Arakawa A, Kaneko M, and Narukawa M (2015) An investigation of factors con-tributing to higher levels of placebo response in clinical trials in neuropathic pain:a systematic review and meta-analysis. Clin Drug Investig 35:67–81.

Aslaksen PM, Bystad M, Vambheim SM, and Flaten MA (2011) Gender differences inplacebo analgesia: event-related potentials and emotional modulation. PsychosomMed 73:193–199.

Aslaksen PM and Flaten MA (2008) The roles of physiological and subjective stress inthe effectiveness of a placebo on experimentally induced pain. Psychosom Med 70:811–818.

Aslaksen PM, Myrbakk IN, Høifødt RS, and Flaten MA (2007) The effect of experi-menter gender on autonomic and subjective responses to pain stimuli. Pain 129:260–268.

Atlas LY, Whittington RA, Lindquist MA, Wielgosz J, Sonty N, and Wager TD (2012)Dissociable influences of opiates and expectations on pain. J Neurosci 32:8053–8064.

Au Yeung ST, Colagiuri B, Lovibond PF, and Colloca L (2014) Partial reinforcement,extinction, and placebo analgesia. Pain 155:1110–1117.

Averbuch M and Katzper M (2001) Gender and the placebo analgesic effect in acutepain. Clin Pharmacol Ther 70:287–291.

Aziz Q, Thompson DG, Ng VW, Hamdy S, Sarkar S, Brammer MJ, Bullmore ET,Hobson A, Tracey I, and Gregory L, et al. (2000) Cortical processing of humansomatic and visceral sensation. J Neurosci 20:2657–2663.

Ballenger JC, Burrows GD, DuPont RL Jr, Lesser IM, Noyes R Jr, Pecknold JC,Rifkin A, and Swinson RP (1988) Alprazolam in panic disorder and agoraphobia:results from a multicenter trial. I. Efficacy in short-term treatment. Arch GenPsychiatry 45:413–422.

Barrett JE, King MG, and Pang G (2000) Conditioning rhinitis in allergic humans.Ann N Y Acad Sci 917:853–859.

Barsky AJ, Saintfort R, Rogers MP, and Borus JF (2002) Nonspecific medication sideeffects and the nocebo phenomenon. JAMA 287:622–627.

Batterman RC and Lower WR (1968) Placebo responsiveness—influence of previoustherapy. Curr Ther Res Clin Exp 10:136–143.

Beck A, Burdett M, and Lewis H (2015) The association between waiting for psy-chological therapy and therapy outcomes as measured by the CORE-OM. Br J ClinPsychol 54:233–248.

Beecher HK (1955) The powerful placebo. J Am Med Assoc 159:1602–1606.Beissner F, Meissner K, Bär KJ, and Napadow V (2013) The autonomic brain: anactivation likelihood estimation meta-analysis for central processing of autonomicfunction. J Neurosci 33:10503–10511.

Bélanger L, Vallières A, Ivers H, Moreau V, Lavigne G, and Morin CM (2007) Meta-analysis of sleep changes in control groups of insomnia treatment trials. J SleepRes 16:77–84.

Bello S, Moustgaard H, and Hróbjartsson A (2014) The risk of unblinding was in-frequently and incompletely reported in 300 randomized clinical trial publications.J Clin Epidemiol 67:1059–1069.

Benedetti F (2014) Placebo effects: from the neurobiological paradigm to trans-lational implications. Neuron 84:623–637.

Benedetti F (2008) Placebo Effects: Understanding the Mechanisms in Health andDisease, Oxford University Press, Oxford.

Benedetti F (2011) The Patient’s Brain. The Neuroscience behind the Doctor-PatientRelationship, Oxford University Press, Oxford.

Benedetti F, Amanzio M, Baldi S, Casadio C, Cavallo A, Mancuso M, Ruffini E, OliaroA, and Maggi G (1998) The specific effects of prior opioid exposure on placeboanalgesia and placebo respiratory depression. Pain 75:313–319.

Benedetti F, Amanzio M, Baldi S, Casadio C, and Maggi G (1999) Inducing placeborespiratory depressant responses in humans via opioid receptors. Eur J Neurosci11:625–631.

Benedetti F, Amanzio M, Casadio C, Oliaro A, and Maggi G (1997) Blockade of nocebohyperalgesia by the cholecystokinin antagonist proglumide. Pain 71:135–140.

Benedetti F, Amanzio M, Rosato R, and Blanchard C (2011) Nonopioid placebo an-algesia is mediated by CB1 cannabinoid receptors. Nat Med 17:1228–1230.

Benedetti F, Arduino C, Costa S, Vighetti S, Tarenzi L, Rainero I, and Asteggiano G(2006) Loss of expectation-related mechanisms in Alzheimer’s disease makes an-algesic therapies less effective. Pain 121:133–144.

Benedetti F, Colloca L, Torre E, Lanotte M, Melcarne A, Pesare M, Bergamasco B,and Lopiano L (2004) Placebo-responsive Parkinson patients show decreased ac-tivity in single neurons of subthalamic nucleus. Nat Neurosci 7:587–588.

Benedetti F, Durando J, and Vighetti S (2014) Nocebo and placebo modulation ofhypobaric hypoxia headache involves the cyclooxygenase-prostaglandins pathway.Pain 155:921–928.

Benedetti F, Lanotte M, Colloca L, Ducati A, Zibetti M, and Lopiano L (2009) Elec-trophysiological properties of thalamic, subthalamic and nigral neurons during theanti-parkinsonian placebo response. J Physiol 587:3869–3883.

Benedetti F, Mayberg HS, Wager TD, Stohler CS, and Zubieta JK (2005) Neurobio-logical mechanisms of the placebo effect. J Neurosci 25:10390–10402.

Benedetti F, Pollo A, Lopiano L, Lanotte M, Vighetti S, and Rainero I (2003) Con-scious expectation and unconscious conditioning in analgesic, motor, and hormonalplacebo/nocebo responses. J Neurosci 23:4315–4323.

Benedetti F, Lanotte M, Lopiano L, and Colloca L (2007) When words are painful:unraveling the mechanisms of the nocebo effect. Neuroscience 147:260–271.

Benedetti F, Thoen W, Blanchard C, Vighetti S, and Arduino C (2013) Pain as a re-ward: changing the meaning of pain from negative to positive co-activates opioidand cannabinoid systems. Pain 154:361–367.

Benson S, Kotsis V, Rosenberger C, Bingel U, Forsting M, Schedlowski M, GizewskiER, and Elsenbruch S (2012) Behavioural and neural correlates of visceral painsensitivity in healthy men and women: does sex matter? Eur J Pain 16:349–358.

Bingel U; Placebo Competence Team (2014) Avoiding nocebo effects to optimizetreatment outcome. JAMA 312:693–694.

Bingel U, Lorenz J, Schoell E, Weiller C, and Büchel C (2006) Mechanisms of placeboanalgesia: rACC recruitment of a subcortical antinociceptive network. Pain 120:8–15.

Bingel U, Wanigasekera V, Wiech K, Ni Mhuircheartaigh R, Lee MC, Ploner M,and Tracey I (2011) The effect of treatment expectation on drug efficacy: imagingthe analgesic benefit of the opioid remifentanil. Sci Transl Med 3:70ra14.

Blom TJ, Mingione CJ, Guerdjikova AI, Keck PE Jr, Welge JA, and McElroy SL(2014) Placebo response in binge eating disorder: a pooled analysis of 10 clinicaltrials from one research group. Eur Eat Disord Rev 22:140–146.

Bodenheimer T (2008) Coordinating care—a perilous journey through the health caresystem. N Engl J Med 358:1064–1071.

Booth RJ, Petrie KJ, and Brook RJ (1995) Conditioning allergic skin responses inhumans: a controlled trial. Psychosom Med 57:492–495.

Boutron I, Estellat C, Guittet L, Dechartres A, Sackett DL, Hróbjartsson A,and Ravaud P (2006) Methods of blinding in reports of randomized controlled trialsassessing pharmacologic treatments: a systematic review. PLoS Med 3:e425.

Bovbjerg D, Kim YT, Siskind GW, and Weksler ME (1987) Conditioned suppressionof plaque-forming cell responses with cyclophosphamide. The role of taste aversion.Ann N Y Acad Sci 496:588–594.

Bovbjerg DH (2003) Conditioning, cancer, and immune regulation. Brain BehavImmun 17 (Suppl 1):S58–S61.

Bridge JA, Birmaher B, Iyengar S, Barbe RP, and Brent DA (2009) Placebo responsein randomized controlled trials of antidepressants for pediatric major depressivedisorder. Am J Psychiatry 166:42–49.

Broadbent E, Ellis CJ, Thomas J, Gamble G, and Petrie KJ (2009) Further de-velopment of an illness perception intervention for myocardial infarction patients:a randomized controlled trial. J Psychosom Res 67:17–23.

Brody H (2008) The Placebo Response, Harper Collins, New York.Buitelaar JK, Sobanski E, Stieglitz RD, Dejonckheere J, Waechter S, and Schäuble B(2012) Predictors of placebo response in adults with attention-deficit/hyperactivitydisorder: data from 2 randomized trials of osmotic-release oral system methyl-phenidate. J Clin Psychiatry 73:1097–1102.

Buske-Kirschbaum A, Grota L, Kirschbaum C, Bienen T, Moynihan J, Ader R, BlairML, Hellhammer DH, and Felten DL (1996) Conditioned increase in peripheralblood mononuclear cell (PBMC) number and corticosterone secretion in the rat.Pharmacol Biochem Behav 55:27–32.

Butler C and Steptoe A (1986) Placebo responses: an experimental study of psycho-physiological processes in asthmatic volunteers. Br J Clin Psychol 25:173–183.

Canning BJ and Fischer A (2001) Neural regulation of airway smooth muscle tone.Respir Physiol 125:113–127.

Capurso G, Cocomello L, Benedetto U, Cammà C, and Delle Fave G (2012) Meta-analysis: the placebo rate of abdominal pain remission in clinical trials of chronicpancreatitis. Pancreas 41:1125–1131.

Cavanna AE, Strigaro G, and Monaco F (2007) Brain mechanisms underlying theplacebo effect in neurological disorders. Funct Neurol 22:89–94.

Cho HJ (2005) Reviving the old sermon of medicine with the placebo effect. Rev BrasPsiquiatr 27:336–340.

Cho HJ, Hotopf M, and Wessely S (2005) The placebo response in the treatment ofchronic fatigue syndrome: a systematic review and meta-analysis. Psychosom Med67:301–313.

Choung RS and Locke GR 3rd (2011) Epidemiology of IBS. Gastroenterol Clin NorthAm 40:1–10.

Cockayne S, Adamson J, Corbacho Martin B, Fairhurst C, Hewitt C, Hicks K, Hull R,Keenan AM, Lamb SE, and Loughrey L, et al.; REFORM study (2014) The REFORMstudy protocol: a cohort randomised controlled trial of a multifaceted podiatry in-tervention for the prevention of falls in older people. BMJ Open 4:e006977.

Cohen D, Consoli A, Bodeau N, Purper-Ouakil D, Deniau E, Guile JM, and DonnellyC (2010) Predictors of placebo response in randomized controlled trials of psycho-tropic drugs for children and adolescents with internalizing disorders. J ChildAdolesc Psychopharmacol 20:39–47.


Colagiuri B, Livesey EJ, and Harris JA (2011) Can expectancies produce placeboeffects for implicit learning? Psychon Bull Rev 18:399–405.

Colloca L (2014) Placebo, nocebo, and learning mechanisms. Handbook Exp Phar-macol 225:17–35.

Colloca L and Benedetti F (2005) Placebos and painkillers: is mind as real as matter?Nat Rev Neurosci 6:545–552.

Colloca L and Benedetti F (2009) Placebo analgesia induced by social observationallearning. Pain 144:28–34.

Colloca L and Finniss D (2012) Nocebo effects, patient-clinician communication, andtherapeutic outcomes. JAMA 307:567–568.

Colloca L, Lopiano L, Lanotte M, and Benedetti F (2004) Overt versus covert treat-ment for pain, anxiety, and Parkinson’s disease. Lancet Neurol 3:679–684.

Colloca L, Sigaudo M, and Benedetti F (2008) The role of learning in nocebo andplacebo effects. Pain 136:211–218.

Craggs JG, Price DD, Perlstein WM, Verne GN, and Robinson ME (2008) The dy-namic mechanisms of placebo induced analgesia: Evidence of sustained and tran-sient regional involvement. Pain 139:660–669.

Craggs JG, Price DD, Verne GN, Perlstein WM, and Robinson MM (2007) Functionalbrain interactions that serve cognitive-affective processing during pain and placeboanalgesia. Neuroimage 38:720–729.

Cremonini F, Ziogas DC, Chang HY, Kokkotou E, Kelley JM, Conboy L, Kaptchuk TJ,and Lembo AJ (2010) Meta-analysis: the effects of placebo treatment on gastro-oesophageal reflux disease. Aliment Pharmacol Ther 32:29–42.

Dager SR, Khan A, Cowley D, Avery DH, Elder J, Roy-Byrne P, and Dunner DL(1990) Characteristics of placebo response during long-term treatment of panicdisorder. Psychopharmacol Bull 26:273–278.

De Allegri M, Pokhrel S, Becher H, Dong H, Mansmann U, Kouyaté B, Kynast-WolfG, Gbangou A, Sanon M, and Bridges J, et al. (2008) Step-wedge cluster-randomised community-based trials: an application to the study of the impact ofcommunity health insurance. Health Res Policy Syst 6:10.

de Craen AJ, Kaptchuk TJ, Tijssen JG, and Kleijnen J (1999a) Placebos and placeboeffects in medicine: historical overview. J R Soc Med 92:511–515.

de Craen AJ, Moerman DE, Heisterkamp SH, Tytgat GN, Tijssen JG, and Kleijnen J(1999b) Placebo effect in the treatment of duodenal ulcer. Br J Clin Pharmacol 48:853–860.

de la Cruz M, Hui D, Parsons HA, and Bruera E (2010) Placebo and nocebo effects inrandomized double-blind clinical trials of agents for the therapy for fatigue inpatients with advanced cancer. Cancer 116:766–774.

de la Fuente-Fernández R, Ruth TJ, Sossi V, Schulzer M, Calne DB, and Stoessl AJ(2001) Expectation and dopamine release: mechanism of the placebo effect inParkinson’s disease. Science 293:1164–1166.

de la Fuente-Fernández R, Schulzer M, and Stoessl AJ (2004) Placebo mechanismsand reward circuitry: clues from Parkinson’s disease. Biol Psychiatry 56:67–71.

De Pascalis V, Chiaradia C, and Carotenuto E (2002) The contribution of suggest-ibility and expectation to placebo analgesia phenomenon in an experimental set-ting. Pain 96:393–402.

Desai JR, Bowen EA, Danielson MM, Allam RR, and Cantor MN (2013) Creation andimplementation of a historical controls database from randomized clinical trials. JAm Med Inform Assoc 20 (e1):e162–e168.

Deveugele M, Derese A, van den Brink-Muinen A, Bensing J, and De Maeseneer J(2002) Consultation length in general practice: cross sectional study in six Euro-pean countries. BMJ 325:472.

Diederich NJ and Goetz CG (2008) The placebo treatments in neurosciences: Newinsights from clinical and neuroimaging studies. Neurology 71:677–684.

Diener HC, Dowson AJ, Ferrari M, Nappi G, and Tfelt-Hansen P (1999) Unbalancedrandomization influences placebo response: scientific versus ethical issues aroundthe use of placebo in migraine trials. Cephalalgia 19:699–700.

Djuri�c VJ, Markovi�c BM, Lazarevi�c M, and Jankovi�c BD (1988) Anaphylactic shock-induced conditioned taste aversion. II. Correlation between taste aversion andindicators of anaphylactic shock. Brain Behav Immun 2:24–31.

Doenlen R, Krügel U, Wirth T, Riether C, Engler A, Prager G, Engler H, SchedlowskiM, and Pacheco-López G (2011) Electrical activity in rat cortico-limbic structuresafter single or repeated administration of lipopolysaccharide or staphylococcalenterotoxin B. Proc Biol Sci 278:1864–1872.

Doering BK and Rief W (2012) Utilizing placebo mechanisms for dose reduction inpharmacotherapy. Trends Pharmacol Sci 33:165–172.

Dunn AG, Mandl KD, Coiera E, and Bourgeois FT (2013) The effects of industrysponsorship on comparator selection in trial registrations for neuropsychiatricconditions in children. PLoS ONE 8:e84951.

Dunn LB, Palmer BW, and Keehan M (2006) Understanding of placebo controlsamong older people with schizophrenia. Schizophr Bull 32:137–146.

Dworkin SF, Chen AC, LeResche L, and Clark DW (1983) Cognitive reversal ofexpected nitrous oxide analgesia for acute pain. Anesth Analg 62:1073–1077.

Eden D and Zuk Y (1995) Seasickness as a self-fulfilling prophecy: raising self-efficacy to boost performance at sea. J Appl Psychol 80:628–635.

Edward SJ, Stevens AJ, Braunholtz DA, Lilford RJ, and Swift T (2005) The ethics ofplacebo-controlled trials: a comparison of inert and active placebo controls.World JSurg 29:610–614.

Eickhoff SB, Lotze M, Wietek B, Amunts K, Enck P, and Zilles K (2006) Segregationof visceral and somatosensory afferents: an fMRI and cytoarchitectonic mappingstudy. Neuroimage 31:1004–1014.

Eippert F, Bingel U, Schoell ED, Yacubian J, Klinger R, Lorenz J, and Büchel C(2009a) Activation of the opioidergic descending pain control system underliesplacebo analgesia. Neuron 63:533–543.

Eippert F, Finsterbusch J, Bingel U, and Büchel C (2009b) Direct evidence for spinalcord involvement in placebo analgesia. Science 326:404.

Eisenberger NI, Master SL, Inagaki TK, Taylor SE, Shirinyan D, Lieberman MD,and Naliboff BD (2011) Attachment figures activate a safety signal-relatedneural region and reduce pain experience. Proc Natl Acad Sci USA 108:11721–11726.

Elsenbruch S (2011) Abdominal pain in Irritable Bowel Syndrome: a review of putativepsychological, neural and neuro-immune mechanisms. Brain Behav Immun 25:386–394.

Elsenbruch S (2014) How positive and negative expectations shape the experience ofvisceral pain. Handbook Exp Pharmacol 225:97–119.

Elsenbruch S, Kotsis V, Benson S, Rosenberger C, Reidick D, Schedlowski M, BingelU, Theysohn N, Forsting M, and Gizewski ER (2012a) Neural mechanisms medi-ating the effects of expectation in visceral placebo analgesia: an fMRI study inhealthy placebo responders and nonresponders. Pain 153:382–390.

Elsenbruch S, Schmid J, Bäsler M, Cesko E, Schedlowski M, and Benson S (2012b)How positive and negative expectations shape the experience of visceral pain: anexperimental pilot study in healthy women.Neurogastroenterol Motil 24:914–e460.

Elsenbruch S and Enck P (2015) Placebo effects and their determinants in gastro-intestinal disorders. Nat Rev Gastroenterol Hepatol, in press.

Enck P, Benedetti F, and Schedlowski M (2008) New insights into the placebo andnocebo responses. Neuron 59:195–206.

Enck P, Bingel U, Schedlowski M, and Rief W (2013) The placebo response in med-icine: minimize, maximize or personalize? Nat Rev Drug Discov 12:191–204.

Enck P, Horing B, Weimer K, and Klosterhalfen S (2012) Placebo responses andplacebo effects in functional bowel disorders. Eur J Gastroenterol Hepatol 24:1–8.

Enck P, Vinson B, Malfertheiner P, Zipfel S, and Klosterhalfen S (2009) The placeboresponse in functional dyspepsia—reanalysis of trial data. NeurogastroenterolMotil 21:370–377.

Espay AJ, Norris MM, Eliassen JC, Dwivedi A, Smith MS, Banks C, Allendorfer JB,Lang AE, Fleck DE, and Linke MJ, et al. (2015) Placebo effect of medication cost inParkinson disease: a randomized double-blind study. Neurology 84:794–802.

Euteneuer F, Mills PJ, Pung MA, Rief W, and Dimsdale JE (2014) Neighborhoodproblems and nocturnal blood pressure dipping. Health Psychol 33:1366–1372.

Exton MS, Elfers A, Jeong WY, Bull DF, Westermann J, and Schedlowski M (2000)Conditioned suppression of contact sensitivity is independent of sympatheticsplenic innervation. Am J Physiol Regul Integr Comp Physiol 279:R1310–R1315.

Exton MS, Gierse C, Meier B, Mosen M, Xie Y, Frede S, Goebel MU, Limmroth V,and Schedlowski M (2002) Behaviorally conditioned immunosuppression in the ratis regulated via noradrenaline and beta-adrenoceptors. J Neuroimmunol 131:21–30.

Exton MS, Herklotz J, Westermann J, and Schedlowski M (2001) Conditioning in therat: an in vivo model to investigate the molecular mechanisms and clinical impli-cations of brain-immune communication. Immunol Rev 184:226–235.

Exton MS, Schult M, Donath S, Strubel T, Bode U, del Rey A, Westermann J,and Schedlowski M (1999) Conditioned immunosuppression makes subtherapeuticcyclosporin effective via splenic innervation. Am J Physiol 276:R1710–R1717.

Exton MS, von Hörsten S, Schult M, Vöge J, Strubel T, Donath S, Steinmüller C,Seeliger H, Nagel E, and Westermann J, et al. (1998) Behaviorally conditionedimmunosuppression using cyclosporine A: central nervous system reduces IL-2production via splenic innervation. J Neuroimmunol 88:182–191.

Faasse K, Cundy T, Gamble G, and Petrie KJ (2013) The effect of an apparent changeto a branded or generic medication on drug effectiveness and side effects. Psy-chosom Med 75:90–96.

Faasse K, Gamble G, Cundy T, and Petrie KJ (2012) Impact of television coverage onthe number and type of symptoms reported during a health scare: a retrospectivepre-post observational study. BMJ Open 2:pil:e001607.

Faria V, Ahs F, Appel L, Linnman C, Bani M, Bettica P, Pich EM, Wahlstedt K,Fredrikson M, and Furmark T (2014) Amygdala-frontal couplings characterizingSSRI and placebo response in social anxiety disorder. Int J Neuropsychopharmacol17:1149–1157.

Farmer AD, Ban VF, Coen SJ, Sanger GJ, Barker GJ, Gresty MA, Giampietro VP,Williams SC, Webb DL, and Hellström PM, et al. (2015) Visually induced nauseacauses characteristic changes in cerebral, autonomic and endocrine function inhumans. J Physiol 593:1183–1196.

Fehm-Wolfsdorf G, Gnadler M, Kern W, Klosterhalfen W, and Kerner W (1993)Classically conditioned changes of blood glucose level in humans. Physiol Behav 54:155–160.

Feltner D, Hill C, Lenderking W, Williams V, and Morlock R (2009) Development ofa patient-reported assessment to identify placebo responders in a generalizedanxiety disorder trial. J Psychiatr Res 43:1224–1230.

Flaten MA, Aslaksen PM, Lyby PS, and Bjørkedal E (2011) The relation of emotionsto placebo responses. Philos Trans R Soc Lond B Biol Sci 366:1818–1827.

Ford AC, Marwaha A, Lim A, and Moayyedi P (2010) Systematic review and meta-analysis of the prevalence of irritable bowel syndrome in individuals with dys-pepsia. Clin Gastroenterol Hepatol 8:401–409.

Ford AC and Moayyedi P (2010) Meta-analysis: factors affecting placebo responserate in the irritable bowel syndrome. Aliment Pharmacol Ther 32:144–158.

Fratello F, Curcio G, Ferrara M, Marzano C, Couyoumdjian A, Petrillo G, Bertini M,and De Gennaro L (2005) Can an inert sleeping pill affect sleep? Effects on poly-somnographic, behavioral and subjective measures. Psychopharmacology (Berl)181:761–770.

Freeman EW and Rickels K (1999) Characteristics of placebo responses in medicaltreatment of premenstrual syndrome. Am J Psychiatry 156:1403–1408.

Freeman S, Yu R, Egorova N, Chen X, Kirsch I, Claggett B, Kaptchuk TJ, Gollub RL,and Kong J (2015) Distinct neural representations of placebo and nocebo effects.Neuroimage 112:197–207.

Freeman TB, Vawter DE, Leaverton PE, Godbold JH, Hauser RA, Goetz CG,and Olanow CW (1999) Use of placebo surgery in controlled trials of a cellular-based therapy for Parkinson’s disease. N Engl J Med 341:988–992.

Frisaldi E, Carlino E, Lanotte M, Lopiano L, and Benedetti F (2014) Characterizationof the thalamic-subthalamic circuit involved in the placebo response throughsingle-neuron recording in Parkinson patients. Cortex 60:3–9.

Fulda KG, Slicho T, and Stoll ST (2007) Patient expectations for placebo treatmentscommonly used in osteopathic manipulative treatment (OMT) clinical trials: a pilotstudy. Osteopath Med Prim Care 1:3.


Fulda S and Wetter TC (2008) Where dopamine meets opioids: a meta-analysis of theplacebo effect in restless legs syndrome treatment studies. Brain 131:902–917.

Furmark T, Appel L, Henningsson S, Ahs F, Faria V, Linnman C, Pissiota A, FransO, Bani M, and Bettica P, et al. (2008) A link between serotonin-related genepolymorphisms, amygdala activity, and placebo-induced relief from social anxiety.J Neurosci 28:13066–13074.

Furukawa TA, Noma H, Caldwell DM, Honyashiki M, Shinohara K, Imai H, Chen P,Hunot V, and Churchill R (2014) Waiting list may be a nocebo condition in psy-chotherapy trials: a contribution from network meta-analysis. Acta PsychiatrScand 130:181–192.

Garcia J, Lasiter PS, Bermudez-Rattoni F, and Deems DA (1985) A general theory ofaversion learning. Ann N Y Acad Sci 443:8–21.

Gardiner JC, Huebner M, Jetton J, and Bradley CJ (2000) Power and sampleassessments for tests of hypotheses on cost-effectiveness ratios. Health Econ 9:227–234.

Gauci M, Husband AJ, Saxarra H, and King MG (1994) Pavlovian conditioning ofnasal tryptase release in human subjects with allergic rhinitis. Physiol Behav 55:823–825.

Gay LN and Carliner PE (1949) The prevention and treatment of motion sickness;seasickness. Bull Johns Hopkins Hosp 84:470–490.

Geers AL, Wellman JA, Fowler SL, Helfer SG, and France CR (2010) Dispositionaloptimism predicts placebo analgesia. J Pain 11:1165–1171.

Geiger F andWolfgram L (2013) Overshadowing as prevention of anticipatory nauseaand vomiting in pediatric cancer patients: study protocol for a randomized con-trolled trial. Trials 14:103.

Geuter S and Büchel C (2013) Facilitation of pain in the human spinal cord by nocebotreatment. J Neurosci 33:13784–13790.

Geuter S, Eippert F, Hindi Attar C, and Büchel C (2013) Cortical and subcorticalresponses to high and low effective placebo treatments. Neuroimage 67:227–236.

Ghanta VK, Hiramoto NS, Solvason HB, Soong SJ, and Hiramoto RN (1990) Con-ditioning: a new approach to immunotherapy. Cancer Res 50:4295–4299.

Ghanta VK, Hiramoto RN, Solvason HB, and Spector NH (1985) Neural and envi-ronmental influences on neoplasia and conditioning of NK activity. J Immunol 135(2, Suppl)848s–852s.

Ghanta VK, Miura T, Hiramoto NS, and Hiramoto RN (1988) Augmentation ofnatural immunity and regulation of tumor growth by conditioning. Ann N Y AcadSci 521:29–42.

Giang DW, Goodman AD, Schiffer RB, Mattson DH, Petrie M, Cohen N, and Ader R(1996) Conditioning of cyclophosphamide-induced leukopenia in humans. J Neu-ropsychiatry Clin Neurosci 8:194–201.

Goebel MU, Hübell D, Kou W, Janssen OE, Katsarava Z, Limmroth V,and Schedlowski M (2005) Behavioral conditioning with interferon beta-1a inhumans. Physiol Behav 84:807–814.

Goebel MU, Meykadeh N, Kou W, Schedlowski M, and Hengge UR (2008) Behavioralconditioning of antihistamine effects in patients with allergic rhinitis. PsychotherPsychosom 77:227–234.

Goebel MU, Trebst AE, Steiner J, Xie YF, Exton MS, Frede S, Canbay AE, MichelMC, Heemann U, and Schedlowski M (2002) Behavioral conditioning of immuno-suppression is possible in humans. FASEB J 16:1869–1873.

Goetz CG, Leurgans S, and Raman R; Parkinson Study Group (2002) Placebo-associated improvements in motor function: comparison of subjective and objectivesections of the UPDRS in early Parkinson’s disease. Mov Disord 17:283–288.

Goetz CG, Wuu J, McDermott MP, Adler CH, Fahn S, Freed CR, Hauser RA, OlanowWC, Shoulson I, and Tandon PK, et al.; Parkinson Study Group (2008) Placeboresponse in Parkinson’s disease: comparisons among 11 trials covering medical andsurgical interventions. Mov Disord 23:690–699.

Gorczynski RM, Macrae S, and Kennedy M (1982) Conditioned immune responseassociated with allogeneic skin grafts in mice. J Immunol 129:704–709.

Grigoleit JS, Kullmann JS, Winkelhaus A, Engler H, Wegner A, Hammes F, OberbeckR, and Schedlowski M (2012) Single-trial conditioning in a human taste-endotoxinparadigm induces conditioned odor aversion but not cytokine responses. Brain BehavImmun 26:234–238.

Grochowicz PM, Schedlowski M, Husband AJ, King MG, Hibberd AD, and Bowen KM(1991) Behavioral conditioning prolongs heart allograft survival in rats. BrainBehav Immun 5:349–356.

Gueorguieva R, Mallinckrodt C, and Krystal JH (2011) Trajectories of depressionseverity in clinical trials of duloxetine: insights into antidepressant and placeboresponses. Arch Gen Psychiatry 68:1227–1237.

Hadamitzky M, Engler H, and Schedlowski M (2013) Learned immunosuppression:extinction, renewal, and the challenge of reconsolidation. J Neuroimmune Phar-macol 8:180–188.

Hall KT, Lembo AJ, Kirsch I, Ziogas DC, Douaiher J, Jensen KB, Conboy LA, KelleyJM, Kokkotou E, and Kaptchuk TJ (2012) Catechol-O-methyltransferase val158-met polymorphism predicts placebo effect in irritable bowel syndrome. PLoS ONE7:e48135.

Harrington A (1999) The Placebo Effect. An Interdisciplinary Exploration, HarvardUniversity Press, Boston, MA.

Hashmi JA, Baria AT, Baliki MN, Huang L, Schnitzer TJ, and Apkarian AV (2012)Brain networks predicting placebo analgesia in a clinical trial for chronic backpain. Pain 153:2393–2402.

Häuser W, Bartram C, Bartram-Wunn E, and Tölle T (2012) Adverse events at-tributable to nocebo in randomized controlled drug trials in fibromyalgia syndromeand painful diabetic peripheral neuropathy: systematic review. Clin J Pain 28:437–451.

Häuser W, Bartram-Wunn E, Bartram C, Reinecke H, and Tölle T (2011) Systematicreview: Placebo response in drug trials of fibromyalgia syndrome and painfulperipheral diabetic neuropathy-magnitude and patient-related predictors. Pain152:1709–1717.

Hida E and Tango T (2011) On the three-arm non-inferiority trial including a placebowith a prespecified margin. Stat Med 30:224–231.

Hiramoto RN, Hiramoto NS, Rish ME, Soong SJ, Miller DM, and Ghanta VK (1991)Role of immune cells in the Pavlovian conditioning of specific resistance to cancer.Int J Neurosci 59:101–117.

Horing B, Weimer K, Muth ER, and Enck P (2015) Prediction of symptom change inplacebo versus no-treatment group in experimentally induced motion sickness.Appl Psychophysiol Biofeedback, in press.

Horing B, Weimer K, Muth ER, and Enck P (2014) Prediction of placebo responses:a systematic review of the literature. Front Psychol 5:1079.

Horing B, Weimer K, Schrade D, Muth ER, Scisco JL, Enck P, and Klosterhalfen S(2013) Reduction of motion sickness with an enhanced placebo instruction: an ex-perimental study with healthy participants. Psychosom Med 75:497–504.

Hróbjartsson A and Gøtzsche PC (2001) Is the placebo powerless? An analysis ofclinical trials comparing placebo with no treatment. N Engl J Med 344:1594–1602.

Hróbjartsson A and Gøtzsche PC (2004) Is the placebo powerless? Update of a sys-tematic review with 52 new randomized trials comparing placebo with no treat-ment. J Intern Med 256:91–100.

Hróbjartsson A and Gøtzsche PC (2010) Placebo interventions for all clinical con-ditions. Cochrane Database Syst Rev (1):CD003974.

Hucklebridge F (2002) Behavioral conditioning of the immune system. Int RevNeurobiol 52:325–351.

Huedo-Medina TB, Kirsch I, Middlemass J, Klonizakis M, and Siriwardena AN(2012) Effectiveness of non-benzodiazepine hypnotics in treatment of adult in-somnia: meta-analysis of data submitted to the Food and Drug Administration.BMJ 345:e8343.

Hunter T, Siess F, and Colloca L (2014) Socially induced placebo analgesia: a com-parison of a pre-recorded versus live face-to-face observation. Eur J Pain 18:914–922.

Huppert JD, Schultz LT, Foa EB, Barlow DH, Davidson JR, Gorman JM, Shear MK,Simpson HB, and Woods SW (2004) Differential response to placebo amongpatients with social phobia, panic disorder, and obsessive-compulsive disorder. AmJ Psychiatry 161:1485–1487.

Ilnyckyj A, Shanahan F, Anton PA, Cheang M, and Bernstein CN (1997) Quantifi-cation of the placebo response in ulcerative colitis. Gastroenterology 112:1854–1858.

Irie M, Maeda M, and Nagata S (2001) Can conditioned histamine release occurunder urethane anesthesia in guinea pigs? Physiol Behav 72:567–573.

Irie M, Nagata S, and Endo Y (2002) Effect of isolation on classical conditionedhistamine release in guinea pigs. Neurosci Res 44:31–35.

Irie M, Nagata S, and Endo Y (2004) Diazepam attenuates conditioned histaminerelease in guinea pigs. Int J Psychophysiol 51:231–238.

Isenberg SA, Lehrer PM, and Hochron S (1992a) The effects of suggestion andemotional arousal on pulmonary function in asthma: a review and a hypothesisregarding vagal mediation. Psychosom Med 54:192–216.

Isenberg SA, Lehrer PM, and Hochron S (1992b) The effects of suggestion on airwaysof asthmatic subjects breathing room air as a suggested bronchoconstrictor andbronchodilator. J Psychosom Res 36:769–776.

Ivanova A and Tamura RN (2011) A two-way enriched clinical trial design: combiningadvantages of placebo lead-in and randomized withdrawal. Stat Methods Med ResDOI:10.1177/0962280211431023 [published ahead of print].

Jänig W (2006) The integrative action of the autonomic nervous system. Neurobiologyof homeostasis, Cambridge University Press, Cambridge.

Janz LJ, Green-Johnson J, Murray L, Vriend CY, Nance DM, Greenberg AH,and Dyck DG (1996) Pavlovian conditioning of LPS-induced responses: effects oncorticosterone, splenic NE, and IL-2 production. Physiol Behav 59:1103–1109.

Jones BE (2011) Neurobiology of waking and sleeping, in Handbook of ClinicalNeurology (Montagna P and Chokroverty S, ed) pp 131–149, Elsevier, Amsterdam.

Jones RE, Moes N, Zwickey H, Cunningham CL, Gregory WL, and Oken B (2008)Treatment of experimental autoimmune encephalomyelitis with alpha lipoic acidand associative conditioning. Brain Behav Immun 22:538–543.

Jubb J and Bensing JM (2013) The sweetest pill to swallow: how patient neurobiologycan be harnessed to maximise placebo effects. Neurosci Biobehav Rev 37:2709–2720.

Jütte R (2014) Hahnemann and placebo. Homeopathy 103:208–212.Jütte R (2013) The early history of the placebo. Complement Ther Med 21:94–97.Kaptchuk TJ (1998) Intentional ignorance: a history of blind assessment and placebocontrols in medicine. Bull Hist Med 72:389–433.

Kaptchuk TJ (2002) The placebo effect in alternative medicine: can the performanceof a healing ritual have clinical significance? Ann Intern Med 136:817–825.

Kaptchuk TJ (2011) Placebo studies and ritual theory: a comparative analysis ofNavajo, acupuncture and biomedical healing. Philos Trans R Soc Lond B Biol Sci366:1849–1858.

Kaptchuk TJ, Friedlander E, Kelley JM, Sanchez MN, Kokkotou E, Singer JP,Kowalczykowski M, Miller FG, Kirsch I, and Lembo AJ (2010) Placebos without de-ception: a randomized controlled trial in irritable bowel syndrome.PLoS ONE 5:e15591.

Kaptchuk TJ, Kelley JM, Conboy LA, Davis RB, Kerr CE, Jacobson EE, Kirsch I,Schyner RN, Nam BH, and Nguyen LT, et al. (2008a) Components of placebo effect:randomised controlled trial in patients with irritable bowel syndrome. BMJ 336:999–1003.

Kaptchuk TJ, Kelley JM, Deykin A, Wayne PM, Lasagna LC, Epstein IO, Kirsch I,and Wechsler ME (2008b) Do “placebo responders” exist? Contemp Clin Trials 29:587–595.

Keltner JR, Furst A, Fan C, Redfern R, Inglis B, and Fields HL (2006) Isolating themodulatory effect of expectation on pain transmission: a functional magnetic res-onance imaging study. J Neurosci 26:4437–4443.

Kemeny ME, Rosenwasser LJ, Panettieri RA, Rose RM, Berg-Smith SM, and KlineJN (2007) Placebo response in asthma: a robust and objective phenomenon. JAllergy Clin Immunol 119:1375–1381.

Kessner S, Forkmann K, Ritter C, Wiech K, Ploner M, and Bingel U (2014) The effectof treatment history on therapeutic outcome: psychological and neurobiologicalunderpinnings. PLoS ONE 9:e109014.


Khan A, Kolts RL, Rapaport MH, Krishnan KR, Brodhead AE, and Browns WA(2005) Magnitude of placebo response and drug-placebo differences across psychi-atric disorders. Psychol Med 35:743–749.

Khan A, Redding N, and Brown WA (2008) The persistence of the placebo response inantidepressant clinical trials. J Psychiatr Res 42:791–796.

Kim SE, Kubomoto S, Chua K, Amichai MM, and Pimentel M (2012) “Pre-cebo”: anunrecognized issue in the interpretation of adequate relief during irritable bowelsyndrome drug trials. J Clin Gastroenterol 46:686–690.

Kingwell K (2014) Schizophrenia drug gets negative results for negative symptoms.Nat Rev Drug Discov 13:244–245.

Kinon BJ, Potts AJ, and Watson SB (2011) Placebo response in clinical trials withschizophrenia patients. Curr Opin Psychiatry 24:107–113.

Kinreich S, Podlipsky I, Jamshy S, Intrator N, and Hendler T (2014) Neural dy-namics necessary and sufficient for transition into pre-sleep induced by EEGneurofeedback. Neuroimage 97:19–28.

Kirk JM, Doty P, and De Wit H (1998) Effects of expectancies on subjective responsesto oral delta9-tetrahydrocannabinol. Pharmacol Biochem Behav 59:287–293.

Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, and Johnson BT(2008) Initial severity and antidepressant benefits: a meta-analysis of data sub-mitted to the Food and Drug Administration. PLoS Med 5:e45.

Kirsch I and Sapirstein G (1998) Listening to prozac but hearing placebo: A meta-analysis of antidepressant medication. Prev Treat 1:article 2a.

Klosterhalfen S, Kellermann S, Braun S, Kowalski A, Schrauth M, Zipfel S, and EnckP (2009) Gender and the nocebo response following conditioning and expectancy. JPsychosom Res 66:323–328.

Klosterhalfen S, Kellermann S, Stockhorst U, Wolf J, Kirschbaum C, Hall G,and Enck P (2005) Latent inhibition of rotation chair-induced nausea in healthymale and female volunteers. Psychosom Med 67:335–340.

Klosterhalfen S and Klosterhalfen W (1990) Conditioned cyclosporine effects but notconditioned taste aversion in immunized rats. Behav Neurosci 104:716–724.

Klosterhalfen S, Rüttgers A, Krumrey E, Otto B, Stockhorst U, Riepl RL, Probst T,and Enck P (2000) Pavlovian conditioning of taste aversion using a motion sicknessparadigm. Psychosom Med 62:671–677.

Klosterhalfen W and Klosterhalfen S (1983) Pavlovian conditioning of immunosup-pression modifies adjuvant arthritis in rats. Behav Neurosci 97:663–666.

Kokkotou E, Conboy LA, Ziogas DC, Quilty MT, Kelley JM, Davis RB, Lembo AJ,and Kaptchuk TJ (2010) Serum correlates of the placebo effect in irritable bowelsyndrome. Neurogastroenterol Motil 22:285–e81.

Kong J, Gollub RL, Rosman IS, Webb JM, Vangel MG, Kirsch I, and Kaptchuk TJ(2006) Brain activity associated with expectancy-enhanced placebo analgesia asmeasured by functional magnetic resonance imaging. J Neurosci 26:381–388.

Kong J, Jensen K, Loiotile R, Cheetham A, Wey H-Y, Tan Y, Rosen B, Smoller JW,Kaptchuk TJ, and Gollub RL (2013) Functional connectivity of the frontoparietalnetwork predicts cognitive modulation of pain. Pain 154:459–467.

Kotsis V, Benson S, Bingel U, Forsting M, Schedlowski M, Gizewski ER,and Elsenbruch S (2012) Perceived treatment group affects behavioral and neuralresponses to visceral pain in a deceptive placebo study. Neurogastroenterol Motil24:935–e462.

Kowalski A, Rapps N, and Enck P (2006) Functional cortical imaging of nausea andvomiting: a possible approach. Auton Neurosci 129:28–35.

Koyama T, McHaffie JG, Laurienti PJ, and Coghill RC (2005) The subjective expe-rience of pain: where expectations become reality. Proc Natl Acad Sci USA 102:12950–12955.

Krogsbøll LT, Hróbjartsson A, and Gøtzsche PC (2009) Spontaneous improvement inrandomised clinical trials: meta-analysis of three-armed trials comparing notreatment, placebo and active intervention. BMC Med Res Methodol 9:1.

Krummenacher P, Candia V, Folkers G, Schedlowski M, and Schönbächler G (2010)Prefrontal cortex modulates placebo analgesia. Pain 148:368–374.

La Mantia L, Eoli M, Salmaggi A, and Milanese C (1996) Does a placebo-effect exist inclinical trials onmultiple sclerosis? Review of the literature. Ital J Neurol Sci 17:135–139.

Lachman HM, Papolos DF, Saito T, Yu YM, Szumlanski CL, and Weinshilboum RM(1996) Human catechol-O-methyltransferase pharmacogenetics: description ofa functional polymorphism and its potential application to neuropsychiatric dis-orders. Pharmacogenetics 6:243–250.

Lamel SA, Myer KA, Younes N, Zhou JA, Maibach H, and Maibach HI (2012) Placeboresponse in relation to clinical trial design: a systematic review and meta-analysisof randomized controlled trials for determining biologic efficacy in psoriasistreatment. Arch Dermatol Res 304:707–717.

Lee HF, Hsieh JC, Lu CL, Yeh TC, Tu CH, Cheng CM, Niddam DM, Lin HC, Lee FY,and Chang FY (2012) Enhanced affect/cognition-related brain responses duringvisceral placebo analgesia in irritable bowel syndrome patients. Pain 153:1301–1310.

Leon AC (2012) Challenges in designing comparative-effectiveness trials for anti-depressants. Clin Pharmacol Ther 91:165–167.

Leucht S, Arbter D, Engel RR, Kissling W, and Davis JM (2009) How effective aresecond-generation antipsychotic drugs? A meta-analysis of placebo-controlled tri-als. Mol Psychiatry 14:429–447.

Leucht S, Heres S, and Davis JM (2013) Increasing placebo response in antipsychoticdrug trials: let’s stop the vicious circle. Am J Psychiatry 170:1232–1234.

Leuchter AF, Cook IA, Witte EA, Morgan M, and Abrams M (2002) Changes in brainfunction of depressed subjects during treatment with placebo. Am J Psychiatry159:122–129.

Leuchter AF, McCracken JT, Hunter AM, Cook IA, and Alpert JE (2009) Monoamineoxidase a and catechol-o-methyltransferase functional polymorphisms and theplacebo response in major depressive disorder. J Clin Psychopharmacol 29:372–377.

Levine JD and Gordon NC (1984) Influence of the method of drug administration onanalgesic response. Nature 312:755–756.

Levine JD, Gordon NC, and Fields HL (1978) The mechanism of placebo analgesia.Lancet 2:654–657.

Levine ME, Stern RM, and Koch KL (2006) The effects of manipulating expectationsthrough placebo and nocebo administration on gastric tachyarrhythmia andmotion-induced nausea. Psychosom Med 68:478–486.

Lidstone SC, Schulzer M, Dinelle K, Mak E, Sossi V, Ruth TJ, de la Fuente-FernándezR, Phillips AG, and Stoessl AJ (2010) Effects of expectation on placebo-induced do-pamine release in Parkinson disease. Arch Gen Psychiatry 67:857–865.

Lieberman MD, Jarcho JM, Berman S, Naliboff BD, Suyenobu BY, Mandelkern M,and Mayer EA (2004) The neural correlates of placebo effects: a disruption account.Neuroimage 22:447–455.

Linde K, Witt CM, Streng A, Weidenhammer W, Wagenpfeil S, Brinkhaus B, WillichSN, and Melchart D (2007) The impact of patient expectations on outcomes in fourrandomized controlled trials of acupuncture in patients with chronic pain. Pain128:264–271.

Lipset CH (2014) Engage with research participants about social media.Nat Med 20:231.Lo B (2015) Sharing clinical trial data: maximizing benefits, minimizing risk. JAMA313:793–794.

Loebel AD, Hyde TS, and Dunner DL (1986) Early placebo response in anxious anddepressed patients. J Clin Psychiatry 47:230–233.

Longo DL, Duffey PL, Kopp WC, Heyes MP, Alvord WG, Sharfman WH, Schmidt PJ,Rubinow DR, and Rosenstein DL (1999) Conditioned immune response tointerferon-gamma in humans. Clin Immunol 90:173–181.

Lovick TA (2008) Pro-nociceptive action of cholecystokinin in the periaqueductalgrey: a role in neuropathic and anxiety-induced hyperalgesic states. NeurosciBiobehav Rev 32:852–862.

Lu HC, Hsieh JC, Lu CL, Niddam DM, Wu YT, Yeh TC, Cheng CM, Chang FY,and Lee SD (2010) Neuronal correlates in the modulation of placebo analgesia inexperimentally-induced esophageal pain: a 3T-fMRI study. Pain 148:75–83.

Lui F, Colloca L, Duzzi D, Anchisi D, Benedetti F, and Porro CA (2010) Neural basesof conditioned placebo analgesia. Pain 151:816–824.

Luparello T, Lyons HA, Bleecker ER, and McFadden ER Jr (1968) Influences ofsuggestion on airway reactivity in asthmatic subjects. Psychosom Med 30:819–825.

Lyby PS, Aslaksen PM, and Flaten MA (2011) Variability in placebo analgesia andthe role of fear of pain—an ERP study. Pain 152:2405–2412.

Lyerly SB, Ross S, Krugman AD, and Clyde DJ (1964) Drugs And Placebos: TheEffects Of Instructions Upon Performance And Mood Under Amphetamine SulphateAnd Chloral Hydrate. J Abnorm Psychol 68:321–327.

Macedo A, Farré M, and Baños JE (2006) A meta-analysis of the placebo response inacute migraine and how this response may be influenced by some of the character-istics of clinical trials. Eur J Clin Pharmacol 62:161–172.

Machado LA, Kamper SJ, Herbert RD, Maher CG, and McAuley JH (2008) Imperfectplacebos are common in low back pain trials: a systematic review of the literature.Eur Spine J 17:889–904.

Mallinckrodt CH, Zhang L, Prucka WR, and Millen BA (2010) Signal detection andplacebo response in schizophrenia: parallels with depression. PsychopharmacolBull 43:53–72.

Mancini M, Wade AG, Perugi G, Lenox-Smith A, and Schacht A (2014) Impact ofpatient selection and study characteristics on signal detection in placebo-controlledtrials with antidepressants. J Psychiatr Res 51:21–29.

Martini M, Lee MC, Valentini E, and Iannetti GD (2015) Intracortical modulation,and not spinal inhibition, mediates placebo analgesia. Eur J Neurosci 41:498–504.

Maru�si�c A and Feren�ci�c SF (2013) Adoption of the double dummy trial design toreduce observer bias in testing treatments. J R Soc Med 106:196–198.

Mavissakalian M (1988) The placebo effect in agoraphobia—II. J Nerv Ment Dis 176:446–448.

Mayberg HS, Silva JA, Brannan SK, Tekell JL, Mahurin RK, McGinnis S,and Jerabek PA (2002) The functional neuroanatomy of the placebo effect. Am JPsychiatry 159:728–737.

McCall WV, D’Agostino R Jr, Rosenquist PB, Kimball J, Boggs N, Lasater B,and Blocker J (2011) Dissection of the factors driving the placebo effect in hypnotictreatment of depressed insomniacs. Sleep Med 12:557–564.

Mccall WV, Perlis ML, Tu X, Groman AE, Krystal A, and Walsh JK (2005) A com-parison of placebo and no-treatment during a hypnotic clinical trial. Int J ClinPharmacol Ther 43:355–359.

McRae C, Cherin E, Yamazaki TG, Diem G, Vo AH, Russell D, Ellgring JH, Fahn S,Greene P, and Dillon S, et al. (2004) Effects of perceived treatment on quality of lifeand medical outcomes in a double-blind placebo surgery trial. Arch Gen Psychiatry61:412–420.

Meisel C, Schwab JM, Prass K, Meisel A, and Dirnagl U (2005) Central nervoussystem injury-induced immune deficiency syndrome. Nat Rev Neurosci 6:775–786.

Meissner K (2005) Preferring patient-reported to observer-reported outcomes sub-stantially influences the results of the updated systematic review on placebos byHróbjartsson and Gøtzsche. J Intern Med 257:394, author reply 395–396.

Meissner K (2009) Effects of placebo interventions on gastric motility and generalautonomic activity. J Psychosom Res 66:391–398.

Meissner K and Ziep D (2011) Organ-specificity of placebo effects on blood pressure.Auton Neurosci 164:62–66.

Mellergård M and Rosenberg NK (1990) Patterns of response during placebo treat-ment of panic disorder. Acta Psychiatr Scand 81:340–344.

Metrik J, Rohsenow DJ, Monti PM, McGeary J, Cook TA, de Wit H, Haney M,and Kahler CW (2009) Effectiveness of a marijuana expectancy manipulation:Piloting the balanced-placebo design for marijuana. Exp Clin Psychopharmacol 17:217–225.

Moerman D (2002) Meaning, Medicine and the ’Placebo Effect, Cambridge UniversityPress, Cambridge.

Moerman DE (2000) Cultural variations in the placebo effect: ulcers, anxiety, andblood pressure. Med Anthropol Q 14:51–72.

Moerman DE and Jonas WB (2002) Deconstructing the placebo effect and finding themeaning response. Ann Intern Med 136:471–476.

Moncrieff J, Wessely S, and Hardy R (2004) Active placebos versus antidepressantsfor depression. Cochrane Database Syst Rev (1):CD003012.


Moore RA and Aubin HJ (2012) Do placebo response rates from cessation trialsinform on strength of addictions? Int J Environ Res Public Health 9:192–211.

Mora MS, Nestoriuc Y, and Rief W (2011) Lessons learned from placebo groups inantidepressant trials. Philos Trans R Soc Lond B Biol Sci 366:1879–1888.

Morton DL, Watson A, El-Deredy W, and Jones AK (2009) Reproducibility of placeboanalgesia: Effect of dispositional optimism. Pain 146:194–198.

Murray D and Stoessl AJ (2013) Mechanisms and therapeutic implications of theplacebo effect in neurological and psychiatric conditions. Pharmacol Ther 140:306–318.

Narkus A, Lehnigk U, Haefner D, Klinger R, Pfaar O, and Worm M (2013) Theplacebo effect in allergen-specific immunotherapy trials. Clin Transl Allergy 3:42.

Nestoriuc Y, Orav EJ, Liang MH, Horne R, and Barsky AJ (2010) Prediction ofnonspecific side effects in rheumatoid arthritis patients by beliefs about medicines.Arthritis Care Res (Hoboken) 62:791–799.

Newcorn JH, Sutton VK, Zhang S, Wilens T, Kratochvil C, Emslie GJ, D’souza DN,Schuh LM, and Allen AJ (2009) Characteristics of placebo responders in pediatricclinical trials of attention-deficit/hyperactivity disorder. J Am Acad Child AdolescPsychiatry 48:1165–1172.

Nicholas A, Bailey JV, Stevenson F, and Murray E (2013) The Sexunzipped trial:young people’s views of participating in an online randomized controlled trial. JMed Internet Res 15:e276.

Niemi MB, Härting M, Kou W, Del Rey A, Besedovsky HO, Schedlowski M,and Pacheco-López G (2007) Taste-immunosuppression engram: reinforcement andextinction. J Neuroimmunol 188:74–79.

Noelpp B and Noelpp-Eschenhagen I (1951a) [Experimental bronchial asthma inguinea pigs, Part I. Methods for objective registration of asthma attacks]. Int ArchAllergy Appl Immunol 2:308–320.

Noelpp B and Noelpp-Eschenhagen I (1951b) [Role of conditioned reflex in bronchialasthma; experimental investigation on the pathogenesis of bronchial asthma].HelvMed Acta 18:142–158.

Ober K, Benson S, Vogelsang M, Bylica A, Günther D, Witzke O, Kribben A, EnglerH, and Schedlowski M (2012) Plasma noradrenaline and state anxiety levels pre-dict placebo response in learned immunosuppression. Clin Pharmacol Ther 91:220–226.

Ogawa A, Hinotsu S, Urushihara H, and Kawakami K (2011) Night-to-night vari-ability of sleep latency significantly predicts the magnitude of subsequent changein sleep latency during placebo administration. Sleep Med 12:565–571.

Pacheco-López G, Doenlen R, Krügel U, Arnold M, Wirth T, Riether C, Engler A,Niemi MB, Christians U, and Engler H, et al. (2013) Neurobehavioural activationduring peripheral immunosuppression. Int J Neuropsychopharmacol 16:137–149.

Pacheco-López G, Engler H, Niemi MB, and Schedlowski M (2006) Expectations andassociations that heal: Immunomodulatory placebo effects and its neurobiology.Brain Behav Immun 20:430–446.

Pacheco-López G, Niemi MB, Kou W, Härting M, Del Rey A, Besedovsky HO,and Schedlowski M (2004) Behavioural endocrine immune-conditioned response isinduced by taste and superantigen pairing. Neuroscience 129:555–562.

Pacheco-López G, Niemi MB, Kou W, Härting M, Fandrey J, and Schedlowski M(2005) Neural substrates for behaviorally conditioned immunosuppression in therat. J Neurosci 25:2330–2337.

Pacheco-López G, Riether C, Doenlen R, Engler H, Niemi MB, Engler A, Kavelaars A,Heijnen CJ, and Schedlowski M (2009) Calcineurin inhibition in splenocytes in-duced by pavlovian conditioning. FASEB J 23:1161–1167.

Palermo-Neto J and Guimarães RK (2000) Pavlovian conditioning of lung anaphy-lactic response in rats. Life Sci 68:611–623.

Papadopoulos D and Mitsikostas DD (2010) Nocebo effects in multiple sclerosis trials:a meta-analysis. Mult Scler 16:816–828.

Papakostas GI and Fava M (2009) Does the probability of receiving placebo influenceclinical trial outcome? A meta-regression of double-blind, randomized clinical trialsin MDD. Eur Neuropsychopharmacol 19:34–40.

Peciña M, Azhar H, Love TM, Lu T, Fredrickson BL, Stohler CS, and Zubieta JK(2013a) Personality trait predictors of placebo analgesia and neurobiological cor-relates. Neuropsychopharmacology 38:639–646.

Peciña M, Love T, Stohler CS, Goldman D, and Zubieta JK (2015) Effects of the Muopioid receptor polymorphism (OPRM1 A118G) on pain regulation, placebo effectsand associated personality trait measures. Neuropsychopharmacology 40:957–965.

Peciña M, Martínez-Jauand M, Hodgkinson C, Stohler CS, Goldman D, and ZubietaJK (2014) FAAH selectively influences placebo effects. Mol Psychiatry 19:385–391.

Peciña M, Stohler CS, and Zubieta JK (2013b) Role of m-opioid system in the for-mation of memory of placebo responses. Mol Psychiatry 18:135–137.

Peciña M and Zubieta JK (2015) Molecular mechanisms of placebo responses inhumans. Mol Psychiatry 20:416–423.

Pecknold JC, Swinson RP, Kuch K, and Lewis CP (1988) Alprazolam in panic dis-order and agoraphobia: results from a multicenter trial. III. Discontinuationeffects. Arch Gen Psychiatry 45:429–436.

Peeke HV, Dark K, Ellman G, McCurry C, and Salfi M (1987) Prior stress andbehaviorally conditioned histamine release. Physiol Behav 39:89–93.

Perlis ML, McCall WV, Jungquist CR, Pigeon WR, and Matteson SE (2005) Placeboeffects in primary insomnia. Sleep Med Rev 9:381–389.

Petrie KJ, Cameron LD, Ellis CJ, Buick D, and Weinman J (2002) Changing illnessperceptions after myocardial infarction: an early intervention randomized con-trolled trial. Psychosom Med 64:580–586.

Petrovic P, Kalso E, Petersson KM, and Ingvar M (2002) Placebo and opioid analgesia—imaging a shared neuronal network. Science 295:1737–1740.

Piercy MA, Sramek JJ, Kurtz NM, and Cutler NR (1996) Placebo response in anxietydisorders. Ann Pharmacother 30:1013–1019.

Ploghaus A, Narain C, Beckmann CF, Clare S, Bantick S, Wise R, Matthews PM,Rawlins JN, and Tracey I (2001) Exacerbation of pain by anxiety is associated withactivity in a hippocampal network. J Neurosci 21:9896–9903.

Pollo A, Vighetti S, Rainero I, and Benedetti F (2003) Placebo analgesia and theheart. Pain 102:125–133.

Preston RA, Materson BJ, Reda DJ, and Williams DW (2000) Placebo-associatedblood pressure response and adverse effects in the treatment of hypertension:observations from a Department of Veterans Affairs Cooperative Study. Arch In-tern Med 160:1449–1454.

Price DD, Craggs J, Verne GN, Perlstein WM, and Robinson ME (2007) Placeboanalgesia is accompanied by large reductions in pain-related brain activity in ir-ritable bowel syndrome patients. Pain 127:63–72.

Quinn VF and Colagiuri B (2015) Placebo interventions for nausea: a systematicreview. Ann Behav Med 49:449–462.

Quinn VF, MacDougall HG, and Colagiuri B (2015) Galvanic vestibular stimulation:a new model of placebo-induced nausea. J Psychosom Res 78:484–488.

Raicek JE, Stone BH, and Kaptchuk TJ (2012) Placebos in 19th century medicine:a quantitative analysis of the BMJ. BMJ 345:e8326.

Ramírez-Amaya V, Alvarez-Borda B, Ormsby CE, Martínez RD, Pérez-Montfort R,and Bermúdez-Rattoni F (1996) Insular cortex lesions impair the acquisition ofconditioned immunosuppression. Brain Behav Immun 10:103–114.

Reinberg AE, Bicakova-Rocher A, Gorceix A, Ashkenazi IE, and Smolensky MH(1994) Placebo effect on the circadian rhythm period tau of temperature and hand-grip strength rhythms: interindividual and gender-related difference. ChronobiolInt 11:45–53.

Relton C, Torgerson D, O’Cathain A, and Nicholl J (2010) Rethinking pragmaticrandomised controlled trials: introducing the “cohort multiple randomised con-trolled trial” design. BMJ 340:c1066.

Reynaert C, Janne P, Vause M, Zdanowicz N, and Lejeune D (1995) Clinical trials ofantidepressants: the hidden face: where locus of control appears to play a key rolein depression outcome. Psychopharmacology (Berl) 119:449–454.

Rheims S, Cucherat M, Arzimanoglou A, and Ryvlin P (2008) Greater response toplacebo in children than in adults: a systematic review and meta-analysis in drug-resistant partial epilepsy. PLoS Med 5:e166.

Rief W, Avorn J, and Barsky AJ (2006) Medication-attributed adverse effects inplacebo groups: implications for assessment of adverse effects. Arch Intern Med166:155–160.

Rief W and Glombiewski JA (2012) The hidden effects of blinded, placebo-controlledrandomized trials: an experimental investigation. Pain 153:2473–2477.

Rief W, Mills PJ, Ancoli-Israel S, Ziegler MG, Pung MA, and Dimsdale JE (2010)Overnight changes of immune parameters and catecholamines are associated withmood and stress. Psychosom Med 72:755–762.

Rief W, Nestoriuc Y, von Lilienfeld-Toal A, Dogan I, Schreiber F, Hofmann SG,Barsky AJ, and Avorn J (2009a) Differences in adverse effect reporting in placebogroups in SSRI and tricyclic antidepressant trials: a systematic review and meta-analysis. Drug Saf 32:1041–1056.

Rief W, Nestoriuc Y, Weiss S, Welzel E, Barsky AJ, and Hofmann SG (2009b) Meta-analysis of the placebo response in antidepressant trials. J Affect Disord 118:1–8.

Riether C, Doenlen R, Pacheco-López G, Niemi MB, Engler A, Engler H,and Schedlowski M (2008) Behavioural conditioning of immune functions: how thecentral nervous system controls peripheral immune responses by evoking asso-ciative learning processes. Rev Neurosci 19:1–17.

Riether C, Kavelaars A, Wirth T, Pacheco-López G, Doenlen R, Willemen H, HeijnenCJ, Schedlowski M, and Engler H (2011) Stimulation of b₂-adrenergic receptorsinhibits calcineurin activity in CD4(+) T cells via PKA-AKAP interaction. BrainBehav Immun 25:59–66.

Rodriguez-Raecke R, Doganci B, Breimhorst M, Stankewitz A, Büchel C, Birklein F,and May A (2010) Insular cortex activity is associated with effects of negativeexpectation on nociceptive long-term habituation. J Neurosci 30:11363–11368.

Rogev E and Pillar G (2013) Placebo for a single night improves sleep in patients withobjective insomnia. Isr Med Assoc J 15:434–438.

Rohsenow DJ and Marlatt GA (1981) The balanced placebo design: methodologicalconsiderations. Addict Behav 6:107–122.

Ronel J, Mehilli J, Ladwig KH, Blättler H, Oversohl N, Byrne RA, Bauer A,Schneider S, Linde K, and Henningsen P, et al. (2011) Effects of verbal suggestionon coronary arteries: results of a randomized controlled experimental investigationduring coronary angiography. Am Heart J 162:507–511.

Roscoe JA, O’Neill M, Jean-Pierre P, Heckler CE, Kaptchuk TJ, Bushunow P, ShayneM, Huston A, Qazi R, and Smith B (2010) An exploratory study on the effects of anexpectancy manipulation on chemotherapy-related nausea. J Pain SymptomManage 40:379–390.

Roudebush RE and Bryant HU (1991) Conditioned immunosuppression of a murinedelayed type hypersensitivity response: dissociation from corticosterone elevation.Brain Behav Immun 5:308–317.

Rutherford BR, Marcus SM, Wang P, Sneed JR, Pelton G, Devanand D, Duan N,and Roose SP (2013) A randomized, prospective pilot study of patient expectancyand antidepressant outcome. Psychol Med 43:975–982.

Rutherford BR, Mori S, Sneed JR, Pimontel MA, and Roose SP (2012) Contribution ofspontaneous improvement to placebo response in depression: a meta-analytic re-view. J Psychiatr Res 46:697–702.

Rutherford BR, Pott E, Tandler JM, Wall MM, Roose SP, and Lieberman JA (2014)Placebo response in antipsychotic clinical trials: a meta-analysis. JAMA Psychiatry71:1409–1421.

Rutherford BR, Sneed JR, and Roose SP (2009) Does study design influence outcome?The effects of placebo control and treatment duration in antidepressant trials.Psychother Psychosom 78:172–181.

Sabbioni ME, Bovbjerg DH, Mathew S, Sikes C, Lasley B, and Stokes PE (1997)Classically conditioned changes in plasma cortisol levels induced by dexamethasonein healthy men. FASEB J 11:1291–1296.

Salamone JD (2000) A critique of recent studies on placebo effects of antidepressants:importance of research on active placebos. Psychopharmacology (Berl) 152:1–6.

Schedlowski M and Pacheco-López G (2010) The learned immune response: Pavlovand beyond. Brain Behav Immun 24:176–185.

Schmid J, Langhorst J, Gaß F, Theysohn N, Benson S, Engler H, Gizewski ER,Forsting M, and Elsenbruch S (2015) Placebo analgesia in patients with functional


and organic abdominal pain: a fMRI study in IBS, UC and healthy volunteers. Gut64:418–427.

Schmid J, Theysohn N, Gaß F, Benson S, Gramsch C, Forsting M, Gizewski ER,and Elsenbruch S (2013) Neural mechanisms mediating positive and negativetreatment expectations in visceral pain: a functional magnetic resonance imag-ing study on placebo and nocebo effects in healthy volunteers. Pain 154:2372–2380.

Schmidt L, Braun EK, Wager TD, and Shohamy D (2014) Mind matters: placeboenhances reward learning in Parkinson’s disease. Nat Neurosci 17:1793–1797.

Schneider A, Enck P, Streitberger K, Weiland C, Bagheri S, Witte S, Friederich HC,Herzog W, and Zipfel S (2006) Acupuncture treatment in irritable bowel syndrome.Gut 55:649–654.

Schweinhardt P, Kalk N, Wartolowska K, Chessell I, Wordsworth P, and Tracey I(2008) Investigation into the neural correlates of emotional augmentation of clin-ical pain. Neuroimage 40:759–766.

Schweinhardt P, Seminowicz DA, Jaeger E, Duncan GH, and Bushnell MC (2009)The anatomy of the mesolimbic reward system: a link between personality and theplacebo analgesic response. J Neurosci 29:4882–4887.

Scott DJ, Stohler CS, Egnatuk CM, Wang H, Koeppe RA, and Zubieta JK (2007)Individual differences in reward responding explain placebo-induced expectationsand effects. Neuron 55:325–336.

Scott DJ, Stohler CS, Egnatuk CM, Wang H, Koeppe RA, and Zubieta JK (2008)Placebo and nocebo effects are defined by opposite opioid and dopaminergicresponses. Arch Gen Psychiatry 65:220–231.

Shah E, Triantafyllou K, Hana AA, and Pimentel M (2014) Adverse events appear tounblind clinical trials in irritable bowel syndrome. Neurogastroenterol Motil 26:482–488.

Shapiro AK (1964) A Historic and Heuristic Definition of the Placebo. Psychiatry 27:52–58.

Shapiro AK and Shapiro E (2010) The Powerful Placebo: From Ancient Priest toModern Physician, Johns Hopkins University Press, Baltimore.

Shelke AR, Roscoe JA, Morrow GR, Colman LK, Banerjee TK, and Kirshner JJ (2008)Effect of a nausea expectancy manipulation on chemotherapy-induced nausea:a university of Rochester cancer center community clinical oncology programstudy. J Pain Symptom Manage 35:381–387.

Shetty N, Friedman JH, Kieburtz K, Marshall FJ, and Oakes D; Parkinson StudyGroup (1999) The placebo response in Parkinson’s disease. Clin Neuropharmacol22:207–212.

Sinyor M, Levitt AJ, Cheung AH, Schaffer A, Kiss A, Dowlati Y, and Lanctôt KL(2010) Does inclusion of a placebo arm influence response to active antidepressanttreatment in randomized controlled trials? Results from pooled and meta-analyses.J Clin Psychiatry 71:270–279.

Smith GR Jr and McDaniel SM (1983) Psychologically mediated effect on the delayedhypersensitivity reaction to tuberculin in humans. Psychosom Med 45:65–70.

Spector NH (1996) Neuroimmunomodulation: a brief review. Can conditioning ofnatural killer cell activity reverse cancer and/or aging? Regul Toxicol Pharmacol24:S32–S38.

Spiro HM (1986) Doctors, Patients and Placebo, Yale University Press, New Haven,CT.

Stathis P, Smpiliris M, Konitsiotis S, and Mitsikostas DD (2013) Nocebo as a potentialconfounding factor in clinical trials for Parkinson’s disease treatment: a meta-analysis.Eur J Neurol 20:527–533.

Stein DJ, Baldwin DS, Dolberg OT, Despiegel N, and Bandelow B (2006) Whichfactors predict placebo response in anxiety disorders and major depression? Ananalysis of placebo-controlled studies of escitalopram. J Clin Psychiatry 67:1741–1746.

Stein N, Sprenger C, Scholz J, Wiech K, and Bingel U (2012) White matter integrityof the descending pain modulatory system is associated with interindividual dif-ferences in placebo analgesia. Pain 153:2210–2217.

Stockhorst U, de Fries D, Steingrueber HJ, and Scherbaum WA (2011) Un-conditioned and conditioned effects of intranasally administered insulin vs placeboin healthy men: a randomised controlled trial. Diabetologia 54:1502–1506.

Stockhorst U, Gritzmann E, Klopp K, Schottenfeld-Naor Y, Hübinger A, BerresheimHW, Steingrüber HJ, and Gries FA (1999) Classical conditioning of insulin effectsin healthy humans. Psychosom Med 61:424–435.

Stockhorst U, Hall G, Enck P, and Klosterhalfen S (2014) Effects of overshadowing onconditioned and unconditioned nausea in a rotation paradigm with humans. ExpBrain Res 232:2651–2664.

Stockhorst U, Mahl N, Krueger M, Huenig A, Schottenfeld-Naor Y, Huebinger A,Berresheim HW, Steingrueber HJ, and Scherbaum WA (2004) Classical condi-tioning and conditionability of insulin and glucose effects in healthy humans.Physiol Behav 81:375–388.

Stockhorst U, Wiener JA, Klosterhalfen S, Klosterhalfen W, Aul C, and SteingrüberHJ (1998) Effects of overshadowing on conditioned nausea in cancer patients: anexperimental study. Physiol Behav 64:743–753.

Stolberg M (2006) Inventing the randomized double-blind trial: the Nuremberg salttest of 1835. J R Soc Med 99:642–643.

Strafella AP, Ko JH, and Monchi O (2006) Therapeutic application of transcranialmagnetic stimulation in Parkinson’s disease: the contribution of expectation.Neuroimage 31:1666–1672.

Su C, Lichtenstein GR, Krok K, Brensinger CM, and Lewis JD (2004) A meta-analysis of the placebo rates of remission and response in clinical trials of activeCrohn’s disease. Gastroenterology 126:1257–1269.

Suchman AL and Ader R (1992) Classic conditioning and placebo effects in crossoverstudies. Clin Pharmacol Ther 52:372–377.

Suetsugi M, Mizuki Y, Yamamoto K, Uchida S, and Watanabe Y (2007) The effect ofplacebo administration on the first-night effect in healthy young volunteers. ProgNeuropsychopharmacol Biol Psychiatry 31:839–847.

Summerfield C and de Lange FP (2014) Expectation in perceptual decision making:neural and computational mechanisms. Nat Rev Neurosci 15:745–756.

Talley NJ, Locke GR, Lahr BD, Zinsmeister AR, Cohard-Radice M, D’Elia TV, Tack J,and Earnest DL (2006) Predictors of the placebo response in functional dyspepsia.Aliment Pharmacol Ther 23:923–936.

Theysohn N, Schmid J, Icenhour A, Mewes C, Forsting M, Gizewski ER, SchedlowskiM, Elsenbruch S, and Benson S (2014) Are there sex differences in placeboanalgesia during visceral pain processing? A fMRI study in healthy subjects.Neurogastroenterol Motil 26:1743–1753.

Thijs L, Amery A, Birkenhäger W, Bulpitt CJ, Clement D, de Leeuw P, De SchaepdryverA, Dollery C, Forette F, and Henry JF, et al. (1990) Age-related effects of placebo andactive treatment in patients beyond the age of 60 years: the need for a proper controlgroup. J Hypertens 8:997–1002.

Thompson WG (2005) The Placebo Effect and Health: Combining Science and Com-passionate Care, Prometheus Books, Amherst, NY.

Tracey I (2010a) Getting the pain you expect: mechanisms of placebo, nocebo andreappraisal effects in humans. Nat Med 16:1277–1283.

Tracey I and Mantyh PW (2007) The cerebral signature for pain perception and itsmodulation. Neuron 55:377–391.

Tracey KJ (2009) Reflex control of immunity. Nat Rev Immunol 9:418–428.Tracey KJ (2010b) Understanding immunity requires more than immunology. NatImmunol 11:561–564.

Tudur Smith C, Dwan K, Altman DG, Clarke M, Riley R, and Williamson PR (2014)Sharing individual participant data from clinical trials: an opinion survey re-garding the establishment of a central repository. PLoS ONE 9:e97886.

Turnbull JW (1962) Asthma conceived as a learned response. J Psychosom Res 6:59–70.

Vase L, Robinson ME, Verne GN, and Price DD (2003) The contributions of sug-gestion, desire, and expectation to placebo effects in irritable bowel syndromepatients. An empirical investigation. Pain 105:17–25.

Vase L, Robinson ME, Verne GN, and Price DD (2005) Increased placebo analgesiaover time in irritable bowel syndrome (IBS) patients is associated with desire andexpectation but not endogenous opioid mechanisms. Pain 115:338–347.

Vits S, Cesko E, Benson S, Rueckert A, Hillen U, Schadendorf D, and Schedlowski M(2013) Cognitive factors mediate placebo responses in patients with house dustmite allergy. PLoS ONE 8:e79576.

Vits S, Cesko E, Enck P, Hillen U, Schadendorf D, and Schedlowski M (2011)Behavioural conditioning as the mediator of placebo responses in the immunesystem. Philos Trans R Soc Lond B Biol Sci 366:1799–1807.

Vögtle E, Barke A, and Kröner-Herwig B (2013) Nocebo hyperalgesia induced bysocial observational learning. Pain 154:1427–1433.

Volkow ND, Wang GJ, Ma Y, Fowler JS, Zhu W, Maynard L, Telang F, Vaska P, DingYS, and Wong C, et al. (2003) Expectation enhances the regional brain metabolicand the reinforcing effects of stimulants in cocaine abusers. J Neurosci 23:11461–11468.

von Hörsten S, Exton MS, Vöge J, Schult M, Nagel E, Schmidt RE, Westermann J,and Schedlowski M (1998) Cyclosporine A affects open field behavior in DA rats.Pharmacol Biochem Behav 60:71–76.

Waber RL, Shiv B, Carmon Z, and Ariely D (2008) Commercial features of placeboand therapeutic efficacy. JAMA 299:1016–1017.

Wager TD, Atlas LY, Leotti LA, and Rilling JK (2011) Predicting individual differ-ences in placebo analgesia: contributions of brain activity during anticipation andpain experience. J Neurosci 31:439–452.

Wager TD, Matre D, and Casey KL (2006) Placebo effects in laser-evoked painpotentials. Brain Behav Immun 20:219–230.

Wager TD, Rilling JK, Smith EE, Sokolik A, Casey KL, Davidson RJ, Kosslyn SM,Rose RM, and Cohen JD (2004) Placebo-induced changes in FMRI in the anti-cipation and experience of pain. Science 303:1162–1167.

Wager TD, Scott DJ, and Zubieta JK (2007) Placebo effects on human mu-opioidactivity during pain. Proc Natl Acad Sci USA 104:11056–11061.

Walsh BT, Seidman SN, Sysko R, and Gould M (2002) Placebo response in studies ofmajor depression: variable, substantial, and growing. JAMA 287:1840–1847.

Wang X, Shang D, Ribbing J, Ren Y, Deng C, Zhou T, Guo F, and LuW (2012) Placeboeffect model in asthma clinical studies: longitudinal meta-analysis of forced expi-ratory volume in 1 second. Eur J Clin Pharmacol 68:1157–1166.

Watson A, El-Deredy W, Vogt BA, and Jones AK (2007) Placebo analgesia is not dueto compliance or habituation: EEG and behavioural evidence. Neuroreport 18:771–775.

Weber FD, Wang J-Y, Born J, and Inostroza M (2014) Sleep benefits in parallelimplicit and explicit measures of episodic memory. Learn Mem 21:190–198.

Weber MA (2008) Using placebo in hypertension clinical trials: new information andnew opportunities. J Hum Hypertens 22:653–655.

Wechsler ME, Kelley JM, Boyd IO, Dutile S, Marigowda G, Kirsch I, Israel E,and Kaptchuk TJ (2011) Active albuterol or placebo, sham acupuncture, or nointervention in asthma. N Engl J Med 365:119–126.

Weimer K, Colloca L, and Enck P (2015a) Age and sex as moderators of the placeboresponse – an evaluation of systematic reviews and meta-analyses across medicine.Gerontology 61:97–108.

Weimer K, Colloca L, and Enck P (2015b) Placebo effects in psychiatry: mediatorsand moderators. Lancet Psychiatry 2:246–257.

Weimer K and Enck P (2014) Traditional and innovative experimental and clinicaltrial designs and their advantages and pitfalls. Handbook Exp Pharmacol 225:237–272.

Weimer K, Gulewitsch MD, Schlarb AA, Schwille-Kiuntke J, Klosterhalfen S,and Enck P (2013a) Placebo effects in children: a review. Pediatr Res 74:96–102.

Weimer K, Horing B, Stürmer J, Klosterhalfen S, Zipfel S, and Enck P (2013b)Nicotine stimulus expectancy differentially affects reaction time in healthy non-smokers and smokers depending on sex: a pilot study. Exp Clin Psychopharmacol21:181–187.

Weimer K, Schulte J, Maichle A, Muth ER, Scisco JL, Horing B, Enck P,and Klosterhalfen S (2012) Effects of ginger and expectations on symptoms ofnausea in a balanced placebo design. PLoS ONE 7:e49031.


Wendt L, Albring A, Benson S, Engler H, Engler A, Hinney A, Rief W, Witzke O,and Schedlowski M (2014a) Catechol-O-methyltransferase Val158Met poly-morphism is associated with somatosensory amplification and nocebo responses.PLoS ONE 9:e107665.

Wendt L, Albring A, and Schedlowski M (2014b) Learned placebo responses inneuroendocrine and immune functions. Handbook Exp Pharmacol 225:159–181.

Wernicke JF and Ossanna MJ (2010) The placebo response in pain and depression: insearch of a common pathway. Front Biosci (Schol Ed) 2:106–111 Schol Ed.

Whalley B, Hyland ME, and Kirsch I (2008) Consistency of the placebo effect. JPsychosom Res 64:537–541.

Wichers MC, Barge-Schaapveld DQ, Nicolson NA, Peeters F, de Vries M, MengelersR, and van Os J (2009) Reduced stress-sensitivity or increased reward experience:the psychological mechanism of response to antidepressant medication. Neuro-psychopharmacology 34:923–931.

Williamson MJ, Thomas MJ, and Stern RM (2004) The contribution of expectationsto motion sickness symptoms and gastric activity. J Psychosom Res 56:721–726.

Winkler A and Rief W (2015) Effect of placebo conditions on polysomnographicparameters in primary insomnia: a meta-analysis. Sleep 38:925–931.

Wirth T, Ober K, Prager G, Vogelsang M, Benson S, Witzke O, Kribben A, Engler H,and Schedlowski M (2011) Repeated recall of learned immunosuppression: evi-dence from rats and men. Brain Behav Immun 25:1444–1451.

Wolf S (1950) Effects of suggestion and conditioning on the action of chemical agentsin human subjects; the pharmacology of placebos. J Clin Invest 29:100–109.

Wolf S (1959) The pharmacology of placebos. Pharmacol Rev 11:689–704.

Woodman CL, Noyes R Jr, Ballenger JC, Lydiard RB, Sievers G, and Mihalko D(1994) Predictors of response to alprazolam and placebo in patients with panicdisorder. J Affect Disord 30:5–13.

Woods SW, Gueorguieva RV, Baker CB, and Makuch RW (2005) Control group biasin randomized atypical antipsychotic medication trials for schizophrenia. Arch GenPsychiatry 62:961–970.

World Medical Association (2013) World Medical Association Declaration of Helsinki:ethical principles for medical research involving human subjects. JAMA 310:2191–2194.

Wrobel N, Wiech K, Forkmann K, Ritter C, and Bingel U (2014) Haloperidol blocksdorsal striatum activity but not analgesia in a placebo paradigm. Cortex 57:60–73.

Yildiz A, Vieta E, Tohen M, and Baldessarini RJ (2011) Factors modifying drug andplacebo responses in randomized trials for bipolar mania. Int J Neuro-psychopharmacol 14:863–875.

Zelen M (1979) A new design for randomized clinical trials.N Engl J Med 300:1242–1245.Zhang W, Qin S, Guo J, and Luo J (2011) A follow-up fMRI study of a transferableplacebo anxiolytic effect. Psychophysiology 48:1119–1128.

Zhou Q and Verne GN (2014) Neurogastroenterology: Neuronal correlates of placeboin chronic FGIDs. Nat Rev Gastroenterol Hepatol 11:460–461.

Zimmermann-Viehoff F, Meissner K, Koch J, Weber CS, Richter S, and Deter HC (2013)Autonomic effects of suggestive placebo interventions to increase or decrease bloodpressure: a randomized controlled trial in healthy subjects. J Psychosom Res 75:32–35.

Zubieta JK, Bueller JA, Jackson LR, Scott DJ, Xu Y, Koeppe RA, Nichols TE,and Stohler CS (2005) Placebo effects mediated by endogenous opioid activity onmu-opioid receptors. J Neurosci 25:7754–7762.


ASSOCIATE EDITOR: MARTIN C. MICHEL Neuro-Bio-Behavioral...

Documents

Transcript of ASSOCIATE EDITOR: MARTIN C. MICHEL Neuro-Bio-Behavioral...