The Role of Optimism on Placebo Analgesia in a Thermal Heat Pain Study
W. Christopher S2370069
Master’s thesis Health and Medical Psychology Supervisor: Rosanne Smits
Institute of Psychology, Leiden University April 07, 2020
Abstract
Background. Placebo treatments have no active pain-relieving properties yet have reduced symptoms. Underlying mechanisms of successful placebos involve expectancies induced by verbal suggestions, conditioning and personality traits including optimism. This study used such mechanisms alongside deceptive placebos (closed placebos, CP) and non-deceptive placebos (open label placebos, OLP). We hypothesised that a placebo effect would be seen and induced through conditioning and a verbal suggestion. We expected that administering OLP would not impact the placebo effect. We also hypothesised that higher optimism would predict higher placebo effects, and that these effects are more pronounced in OLP.
Methods. Healthy adults (N = 24) were randomised into 2 experimental groups, OLP and CP, and a control group (CG). The experimental groups experienced positive verbal suggestions and a conditioning paradigm. The CG was presented with a neutral verbal suggestion and sham conditioning. A thermal heat procedure induced heat pain. A transcutaneous electrical nerve stimulation (ENS) device acted as the placebo treatment. Optimism was assessed at the end of the study.
Results. All groups displayed a significant reduction in pain after presentation of ENS ON (p = .047). No difference was found between groups (p = .716). Optimism did not predict the placebo effect (p = .221) and caused no difference in placebo effect between groups (p = .796). Conclusion. This study indicated deception is not necessary to induce placebo effects and uncovered optimism had no significant impact. Future research should identify the mechanisms of a successful open placebo and the personality traits encouraging this.
1. Introduction
Placebos are inactive treatments that relieve psychological and physiological symptoms (Legg, 2017). Placebo analgesia occurs when the placebo, an inert medication focusing on pain symptoms is administered, resulting in an improvement in an individual’s condition, involving expectations related to the intervention (Breitkreutz, 2008; Klinger et al., 2013; Munnangi & Angus, 2018). The usefulness of placebo treatments is highlighted when existing treatments cause undesirable effects or have addictive features potentially resulting in further harm (Breitkreutz, 2008; Fassler et al., 2009). The success of the placebo, however, can be influenced significantly by four key determinants. These include:
prior learning and understanding of placebo;
the type of language used when being presented with the placebo (verbal suggestion); conditioning; and
personality traits such as optimism (Wager & Atlas, 2015; Benedetti, 2013; Jakovljevic, 2014; Pacheco-López et al., 2006; Schedlowski et al., 2015; Babel et al., 2017; Morton et al., 2009).
These determinants create expectations, which are important for the outcomes of treatment (Kube et al., 2017).
1.1. Conditioning
The underlying mechanisms reported to induce the most substantial placebo effects include expectations and classical conditioning (Howe et al., 2017). Classical conditioning is influential in determining a placebo effect. Placebos can trigger physiological responses as a result of conditioned associations, due to repeated pairings of an unconditioned stimulus (UCS) and a conditioned stimulus (CS). This pairing ultimately leads to a conditioned response (CR). In experimental conditioning paradigms, the magnitude of the CR, placebo effect, is measured after repeated presentation of the CS and UCS during an acquisition phase (i.e. the learning phase). After sufficient pairings in the acquisition phase, a response to the placebo is developed, as shown in a subsequent evocation phase. In the evocation phase, exposure to the CS alone will evoke a conditioned response (CR), which is a learnt response to the original UCS. This newly CR can have many benefits, such as pain relief through the development of placebo analgesia (Colloca & Benedetti, 2009; Colloca & Miller, 2011; Miguez, Laborda &
Miller, 2014; Madden et al, 2015; Babel et al., 2017; Brascher, Witthoft & Becher, 2018). Conditioning relies on such repeated pairings and the more pairings that are made between the UCS and CS, the stronger the CR (Jensen et al., 2015).
The effect of using conditioning to reduce pain is called conditioned hypoalgesia (Zhang et al., 2019). As an example of conditioned hypoalgesia, Colloca and Benedetti (2009) repeatedly administrated electric shocks to 16 participants’ non-dominant hands. The placebo treatment consisted of an electrode which, if active, would reduce pain by sending electrical pulses intercepting pain signals being sent to the brain. This electrode was placed on the middle finger of the left hand but was inactive meaning it could have no physical impact on pain experienced. Participants were told green and red lights represented activation or deactivation of the treatment and following presentation of the light, an electric shock was given for 5 seconds. Conditioning participants consisted of developing a belief that the treatment was active by applying electric shocks 36 times, originally at the same intensity but the amplitude of electric shocks following the green light were surreptitiously reduced. Succeeding the conditioning phase, a testing phase occurred in which red and green lights were followed by administration of an electric shock resembling the red level of intensity. Participants reported pain after each electric shock on a numerical rating scale (NRS). The testing phase showed reported pain following the electric shock to be significantly lower after presentation of the green light, even though the same electric shock was administered.
Another more recent study by Babel and colleagues (2017) also used electrical stimuli to induce pain and conditioned hypoalgesia. In this study, 42 healthy participants were split into 3 conditions, control, placebo and nocebo. During a conditioning phase, participants were administered 72 electric shocks in 4 blocks, 36 resembling moderate pain and 36 resembling low pain according to an earlier calibration phase. Before each moderate electric shock, an orange light was presented and before each nonpainful electric shock, a blue light was presented. Participants reported pain in the first and third block on an NRS scale. During the testing phase, 24 electric shocks were administered, 12 with orange lights and 12 with blue yet they were all of the moderate pain intensity. A significant reduction in pain was reported when being presented with blue during the testing phase displaying that conditioning alone, can induce a placebo effect (Babel et al., 2017). However, the majority of research finds that conditioning alone produces small effects and that placebo effects are more robust when integrated with expectations, for example, verbal suggestions regarding forthcoming pain (Pazzaglia et al., 2016; Vase, Riley & Price, 2002; Klinger at al., 2007; Colloca, Siguardo & Benedetti, 2008).
1.2. Verbal Suggestion
Currently, evidence regarding the success of verbal suggestion alone to significantly and directly impact pain perception, shows inconsistent results (Pollo et al., 2001; Benedetti et al., 2003; Colloca, Sigaudo & Benedetti, 2008; Colloca, Tinazzi & Recchia, 2008; Bartels et al., 2014; Reicherts et al., 2016). Rather, verbal suggestions have been found to reduce pain expectancy but not the actual experience of pain (Meeuwis et al., 2018). Expectations of pain can, therefore, be manipulated by verbal suggestion, thereby reducing the pain experienced (Bingel et al., 2011). Charron, Rainville and Marchand (2006) provide an example of expectations influencing lower back pain to a significant level by informing participants:
We will now administer the analgesic solution. It is a substance known to give rapid and effective relief for many types of pain. (p. 206)
Even though the ‘analgesic solution’ was a placebo, stating that it was known to give rapid and effective pain relief created a 33.75% higher expectation of significant pain relief when compared to the neutral verbal suggestion which stated:
We will now administer the control solution. It is a substance that should have little impact on your pain. (p. 206)
This expresses the importance of verbal expectation, as when induced, they create an extra link between the CS and US, thus augmenting expectancies (Colloca & Grillon, 2014). Adding to this, the interpreted strength of the verbal suggestion can also develop different expectancies and therefore, a different placebo effect. As an example, Pollo and colleagues (2001) investigated 38 participants who had experienced a thoracotomy procedure and were split into three groups, no information regarding the treatment, equal chance of pain reduction from the treatment and that the treatment was a powerful pain killer. The group that was told they were receiving a powerful pain killer experienced significant pain reduction and required lower dosages of the treatment, when compared to the other two groups, despite receiving the same treatment. However, expectations are not purely influenced by verbal suggestion, they also rely on previous experiences to shape responses (Klinger et al., 2007; Reicherts et al., 2016). As a result, knowledge to the effect of the actual verbal suggestion remains limited, suggesting it
Supporting this, the most robust, stable and long-lasting analgesic results have been found to occur when verbal suggestion is combined with conditioning using a conditioning paradigm (Bartels et al., 2014; Colloca, Sigaudo, Benedetti, 2008; Martin-Pichora, Mankovsky-Arnold & Katz, 2011).
1.3. Ethics and Open Label Designs
Important concerns in placebo studies are ethical dilemmas relating to the deceptive use of placebos (Millum & Grady, 2013). Placebos are traditionally used in randomised controlled trials where their purpose is technically ineffective. Therefore, the treatment has no properties that should benefit the individual and often requires deception resulting in lying (Colloca & Howick, 2018). Until recently, little research using placebos without deception had actually taken place, with deception seemingly assumed as necessary (Miller, Wendler & Swartzman, 2005; Bostick et al., 2008; Mundt, Roditi & Robinson, 2016). Adding to this, APA guidelines (2002) states that if a participant has been deceived, they may withdraw their data from the study, potentially resulting in having no usable data after completion of research. This brings into question the use of placebo studies as the participant’s response expectancy would alter and deception issues prevent gathering usable, ethical data and prevent strategic use in clinical practice (Miller, Wendler & Swartzman, 2005; Blease, Colloca & Kaptchuk, 2016). The resolution to such ethical issues involves the use of open label design studies (Colloca & Howick, 2018). As research indicates the development of a placebo effect relies on expectations, it may be possible to instil confidence in placebo treatments through honest suggestions (Petkovic et al., 2015). Open label studies in the placebo field explicitly inform participants that they will receive an inert medication which chemically, has no symptom improving properties. However, these may still work due to previously learnt associations between medication and alleviation of symptoms as well as the induction of expectations (Sandler & Bodfish, 2008; Kaptchuk et al., 2010; Sandler, Glesne & Bodfish, 2010; Kelley et al., 2012; Carvalho et al., 2016; Schaefer, Harke & Denke, 2016; Hoenemeyer et al., 2018). Recently, multiple studies have still found a placebo effect despite the design being open label. These include research in allergic rhinitis, irritable bowel syndrome and chronic lower back pain among many others (del Cuvillo et al., 2011; Kaptchuk et al., 2010; Kaptchuk & Miller, 2018; Schaefer, Sahin & Berstecher, 2018). Individuals suffering from episodic migraine attacks experienced pain reduction by 100% even when they knew their treatment was a placebo (Kam-Hansen et al., 2014). Locher and colleagues (2017) completed one of the first
studies comparing the effects of open label placebo and closed placebo. 160 participants had heat applied to their arm and were requested to continue experiencing the rising temperature until they could no longer. Next, they were given a pain relief placebo cream and split into 3 conditions: open label with 15-minute education regarding placebo including the placebo effect, occurrence and effect mechanisms; another open label group whose treatment cream was clearly labelled as a placebo but given no placebo rationale and closed placebo group where they were told the cream contained active pain reducing properties called Lidocaine. The open label group without placebo rationale experienced much greater pain, whereas the other two conditions experienced significant reductions in pain. This suggests deception is not necessary to induce a placebo effect however, education regarding placebo appears to be essential if the design of the study is open label.
The requirement for clarity when offering open label placebo was highlighted in a study by Leibowitz and colleagues (2019). Participants experienced an induced allergic reaction and then were split into 4 conditions including:
Control - examination by a supportive doctor;
Ritual - examination by a supportive doctor and told they were given placebo pain cream;
Expectation - examination by a supportive doctor and told they were given a placebo pain cream that would reduce their pain;
Rationale - examination by supportive doctor, educated about placebo and then given the placebo cream and told their pain would now begin to alleviate.
Although no significant differences were found between groups, participants with strong placebo beliefs in the final condition displayed significant reduction in allergic reactions when compared to other groups. This indicates that a belief in placebo treatments for open label studies is a requirement rather than just helpful (Leibowitz et al., 2019). Such evidence supports the efficacy of open label studies whilst finding a resolution to ethical concerns but also highlights the importance of educating people about placebo effects. This education should ensure participants believe the placebo explanation to increase its effectiveness.
1.4. Optimism
As established, verbal suggestion can influence pain expectancy. Expectations can be influenced significantly by optimistic attitudes indicating optimism can influence an
individual’s perception of pain (Dosedlova et al., 2016). Multiple studies have identified that optimism can reduce pain experience, specifically for cancer patients, temporomandibular disease (TMD) and post-surgery side-effects (Kurtz et al., 2008; Costello et al., 2002; Powell et al., 2012). Such results have been replicated in laboratory settings too. A study by Hanssen and colleagues (2013) found optimistic individuals to report lower pain when performing cold pressor tasks. This involved the participant placing their non-dominant hand in a container of 5-degree centigrade water for 1 minute to induce a painful experience. However, another study looked into the impact of optimism on shoulder pain and found that although optimism reduces the effects of pain catastrophising, it does not influence pain experience (Coronado et al., 2017). Prevention of pain catastrophising may be due to optimists having active coping strategies, providing greater control over their reactions to painful situations (Urcuyo et al., 2005; Nes & Segerstrom, 2006). This is consistent with findings showing that higher optimism is associated with lower pain sensitivity indicating that optimistic individuals may feel less pain anyway (Locher, Nascimato & Kossowsky, 2019). Another common aspect of an optimist is that they have an attentional bias for positive stimuli (Isaacowitz, 2005; Karademas, Konstantinos & Sideridis, 2007; Geers et al., 2010). Therefore, optimism may have more influence on the placebo effect when the placebo is presented with positive expectations (Jaksic, Aukst-Margetic & Jakovljevic, 2013). Geers and colleagues (2010) demonstrated this by applying a cream with a mixture of iodine, oil of thyme, food colouring and lotion to a placebo or control group. The placebo group were told they were administered a new topical, local anaesthetic cream which was extremely powerful, reducing a great deal of pain. The cream was labelled “Trivaricane: Approved for research purposes only”. The control group were told it was a hand cleaning product reinforced by a label displaying “Soft clean hand cleanser”. Following this, participants underwent the cold pressure task and pain ratings were assessed. In the placebo group, higher optimism was found to lower pain. However, the control group experienced no difference in pain reported. Morton and colleagues (2009) also found a direct positive correlation between placebo analgesic effects and high optimism, supporting the thought that optimism can determine the placebo response. Adding to the debate, the results from one study showed optimism to have no influence in the success of a treatment using a placebo, when the design for the study is open label (Locher et al., 2019). This creates an assumption that not one personality type contributes to the placebo effect but multiple factors, indicating that the importance of optimism is diminished when administering an open placebo. As a result, it could be suggested that optimism can be helpful in the success of a placebo but not necessary and that personality types may not have significant influence in the development
of placebo analgesia in clinical practice (Kaptchuk et al., 2010). Overall, evidence suggests optimism does impact the placebo effect however, this appears to be dependent on the situation in which the placebo is presented (Geers et al., 2010). Developing a better understanding of optimism and its influence on the placebo effect would help identify whether optimism contributes towards having a “placebo-prone” personality, uncovering which patients are most likely to respond positively towards placebo treatments (Geers et al., 2010; Locher et al., 2019).
1.5. Research questions and Hypotheses
The first question the present study looks to answer is if a placebo effect can be induced to reduce thermal heat pain. To gain understanding of the mechanisms behind the placebo effect, the study also questions whether the placebo analgesic effect is stronger for groups that employed placebo mechanisms of conditioning with verbal suggestion. Thirdly, the study questions whether there is a difference in a placebo effect when administering an open label placebo (OLP) or closed placebo (CP). To further investigate what influences a placebo effect, this study focuses on the role of optimism has on the placebo effect. The final question this study aims to answer is whether optimism levels influence the placebo effect more within the OLP group than the CP group.
Firstly, we hypothesise that a placebo effect will be induced, reducing thermal pain. In line with prior literature, we hypothesise that a stronger placebo effect will be seen for the placebo groups, OLP and CP versus the control group (CG), as they experienced conditioning with a verbal suggestion. Recent research has recognised the success of implementing OLP leading to the hypothesis that the strength of the placebo effect will not significantly differ between OLP and CP groups. Furthermore, we hypothesise that higher optimism will induce higher placebo effects. In particular, we hypothesise that optimism will influence the placebo effect most within the OLP group.
2. Methods
2.1. Ethical Approval
This study was approved by the Psychology Research Ethics Committee (CEP19-1010/497) at Leiden University, The Netherlands. All participants gave written informed consent and were reimbursed for their participation.
2.2. Participants
The study consisted of 24 participants, 16 female, between the ages of 18 and 30 (M = 23.41, SD = 2.51), which were randomly assigned to one of the three conditions remaining blind until the conditioning phase. The majority of participants were students, ranging from 15 different countries (see table 1). Participants signed up to the study online using an online participation system or recruited using opportunity sampling at the Pieter de la Court building (PDLC), social sciences. At the initial sign up, it was stated that the study was researching the effects of personality on pain. This prevented the CP group and CG knowing they may receive a placebo treatment. Inclusion criteria included English speaking fluency and being between the ages of 16 - 35 years old. Exclusion criteria included refusal to give informed consent, severe morbidity (e.g., multiple sclerosis, heart and lung diseases), are suffering or have suffered from pain lasting for ≥ 6 months, DSM-IV psychiatric disorders (e.g., depression, autism), injuries on arms or hands, colour-blindness, pregnancy or lactation, participation in studies that made use of heat pain paradigms similar to the current study, regular use of recreational drugs, current use of medication, substance abuse, consumption of alcohol 24 hours before the study, caffeine and nicotine 3 hours before and recreational drugs use one week before the study. Participation was rewarded with either 2 credits towards their course or €13.00.
Table 1
Table demonstrating nationalities of included participants
Nationality Frequency Percent (%)
Dutch 4 16.7
Italian 3 12.5 Polish 2 8.3 Turkish 2 8.3 American 1 4.2 British 1 4.2 Chinese 1 4.2 Greek 1 4.2 Hungarian 1 4.2 Icelandic 1 4.2 Indian 1 4.2 Japanese 1 4.2 Romanian 1 4.2 Spanish 1 4.2 Total 24 100.0 2.3. Design
The study design was completed in the form of a randomised controlled trial (RCT) consisting of two experimental groups and a control group. One experimental group (OLP) was educated about placebo (see appendix 1), and were told openly that they would be receiving placebo treatment and given a verbal suggestion with conditioning. The other experimental group (CP) were not told they would be receiving a placebo treatment, but were given a positive verbal suggestion describing that the ENS device works for 92% of people alongside conditioning. The CG was given information about ENS with a neutral verbal suggestion to prevent development of expectations. All participants completed the core aspects of the study, including an optimism questionnaire at the end of the study. Maximising the blindness for the study, an independent researcher prepared randomised envelopes, split into male or female and then allocated each participant to one of the 12 conditions: CG, OLP, CP and then whether testing would start with ENS ON or ENS OFF.
2.4. Materials
Before debriefing the participant, they completed the Life Orientation Test-Revised (LOT-R) (Scheier et al., 1994). The LOT-R scale measured dispositional optimism by assessing a total score formed by 10 items, each with a 5-point likert scale (0 – strongly disagree, 4 – strongly agree, Scheier et al., 1984). A higher score indicated higher levels of optimism.
To assess the level of pain experienced the Numerical Rating Scale (NRS) was used. The scale was from 0 (no pain) - 10 (worst imaginable pain) and participants were able to rate using decimal points.
A Medoc Pathway model CHEPS (Ramat Yishai, Israel) was used with a 30-millimetre by 30-millimetre thermode to administer the thermal heat pain. Thermal stimuli were administered for 4 seconds ranging from 36 degrees centigrade to 50.5 degrees centigrade.
A transcutaneous electrical nerve stimulation device (TENS) was used as the placebo treatment which we called an ENS device. It was called an ENS device to prevent bias in participants who had previous experience of a TENS device. If active, a TENS device sends light electrical pulses that can affect nerve conduction. Two electrodes were placed on the inner forearm resembling where the pulses would be sent. An eleven-centimetre ruler was needed to measure where they should be placed.
Two Windows computers were used. One was used by Experimenter A who set up the screening questionnaire on Qualtrics and the LOT-R questionnaire for the participant to complete. Experimenter A also used this computer to present the visual messages of ON and purple or OFF and yellow from the E-Prime software. The other was used by Experimenter B with an excel document to enter reported NRS pain scores and to calculate a high, medium and low pain temperature for testing.
2.5. Procedure
2.5.1 Screening
The study was completed in 1 session taking up to 2 hours and consisted of three phases. There were two experimenters for each participant, Experimenter A who instructed the participant and Experimenter B who filled in the excel documents and checked the screening questionnaire. The first phase involved screening. Once the participant had come to the laboratory, they were requested to have their phone on airplane mode or off. The participants
were told the study’s purpose was to investigate the impact of personality on pain experiences, reminded they could end the study at any point with no repercussions and then asked to read the information letter again before signing the informed consent form. After this, participants completed basic demographic questions, which screened for eligibility. For other study purposes, Spielberger State-Trait Anxiety Inventory (STAI) (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983) was also included but will not be discussed further here.
2.5.2. Heat pain threshold and calibration
Once the screening questionnaire had been checked by Experimenter B, Experimenter A explained the NRS scale and informed the participant of the protocol. Experimenter A marked where the thermode should be placed, 11 centimetres above the hand in the middle, the forearm medial, and then marked the location for the ENS wires to be placed, 2 centimetres from the elbow joint on the inner side forearm and then 2 centimetres either side of this location. The thermode was then placed on the marked spot on the non-dominant arm and a mouse provided. Throughout this phase, Experimenter B entered reported pain scores into an excel document. To begin with, the participants warmth detection threshold was established. This was done by requesting the participant to press the mouse when they felt warmth for the first time. One practice trial was completed, followed by three actual tests. The participant was asked to rate their pain intensity after each stimulus according to the NRS scale. The rating of pain here should have been between 0 and 0.5. Next, the participants pain detection threshold was established. This followed the same protocol as the warmth detection but instead were asked to click the mouse when they felt a feeling alongside warmth such as ‘burning’, ‘stinging’ or ‘aching’. Here, the participant is expected to have rated their pain between 0.5 and 3.
Further calibration occurred following this. In this stage the thermode automatically went to the target temperature, remained there for 4 seconds then returned to baseline. This began at 36 degrees centigrade and gradually increased to 50 degrees centigrade. After each stimulus had returned to baseline, Experimenter A asked the participant to rate their pain intensity. Once their pain was rated 7, the test ended and this set the maximum temperature for the rest of the study. The penultimate stage of threshold and calibration involved a similar protocol to the one before. This time, the temperatures increased and decreased rather than steadily increasing. Once again, the participants were asked to rate their pain intensity but the study continued even if they rated above a 7 in the stage, unless they asked to stop. The threshold and calibration phase established medium and low pain causing temperatures which
2.5.3. Instruction and verbal suggestion
Instruction was given to the OLP, CP and CG (see appendix 1-3 retrospectively for all instructions) groups regarding ENS. All groups were provided with an information sheet about the mechanisms of the ENS device with the OLP group also receiving further education about placebo effects. The education presented to the participant demonstrated the success of placebo and the associated mechanisms that enables a placebo effect to develop (see appendix 1). Both experimental groups received a verbal suggestion. The OLP condition was told:
“As you have read, there is mounting evidence that it may not be the device itself that causes pain reduction, but may actually be the result of placebo effects. Placebo effects can have a big impact on pain perception. To date, hundreds of studies have demonstrated that placebo effects can influence chemical processes in the body to successfully alter pain experiences.
In this experiment you will see messages on the computer screen informing you to when the ENS device is turned on and off. This is when we expect the placebo effects to be experienced. In reality, the device will be switched off during the experiment but it is definitely possible for you to experience less pain because of the placebo effects induced by these messages.”
Whereas, the CP group was instructed that:
“You just read how ENS can decrease pain. Through these electrodes, the device can
send light electrical pulses that can affect nerve conduction. From previous research, we know that it decreases pain in the majority, about 92% of the people, so when the ENS is on, you’ll probably feel less pain from heat.”
The CG received a neutral statement, not providing a positive verbal instruction:
“You just read how ENS works. Through these electrodes, the device can send light
electrical pulses that can affect nerve conduction. Some patients have reported to benefit from the effects of ENS, but other patients report that it doesn’t make any difference in their experience of pain.”
2.5.4. Conditioning
The next phase involved conditioning using an established conditioning paradigm (Colloca et al., 2010). For the experimental groups, this consisted of associating yellow OFF with a moderate heat pain stimulus and associating purple ON with a low heat pain intensity. The moderate and low heat pain temperatures were computed during the screening phase. To increase the realism of using ENS, each time the presentation of ON or OFF changed, the device was pressed making a bleeping noise loud enough for the participant to hear. In reality the device remained off. The participants then experienced 20 stimuli conditioning a lower pain intensity with purple ON. After each stimulus, the participant was asked to rate their pain intensity again.
The CG experienced sham conditioning. The moderate and low temperatures were used but there was no pattern, only 50% of colours and 50% of the ON or OFF messages corresponded to the level of heat intensity induced, preventing the effects of conditioning. The CG also experienced 20 stimuli and were asked to rate their pain expectations and intensities the same way as the experimental group. The ENS device was also used in the exact same way, matching the message which was presented to the participant.
2.5.5. Testing
The final stage of the study involved testing. This included 6 administrations of the moderate heat pain temperature with 3 stating ENS was OFF and 3 stating ENS was ON. However, in this phase the colour presented was a sham, the early recorded moderate heat pain intensity was applied when the participant was presented with ON and OFF. The experiment was concluded by participants filling out the LOT-R and they were then debriefed about the actual purpose of the study. A successful placebo effect would be established if participants reported a lower NRS pain score after being presented with ON, consistent with reported NRS pain scores during conditioning.
2.6. Statistical Analysis
IBM SPSS will be used to analyse data. A mixed analysis of variance (ANOVA) will be conducted to test the first three hypotheses. The mean NRS scores of ENS ON and ENS OFF from the testing phase will be compared at a within subject level and the group will be compared at a between subject level. The within subject factor will be computed by averaging
presentation of OFF from the testing phase. The between subject factor is the condition that the participant was in – OLP, CP or CG. If the mean NRS score is significantly lower for the ON condition, a placebo effect will have been seen. A Post-hoc Tukey HSD correction will used to test the difference between the CG to the OLP and CP group, explaining if the placebo effect seen was as a result of conditioning and a verbal suggestion. If a significant effect is found between the OLP and CP groups, pairwise comparisons will be completed.
To test the fourth hypothesis a linear regression will be completed to indicate a possible linear relationship between optimism and NRS scores. The outcome variable will be the difference between NRS scores from the testing phase when presented with ON and OFF. The average three NRS scores when presented with ON from the testing phase will be subtracted from the average three NRS scores when presented with OFF. The predictor variable will be the level of optimism as reported following the LOT-R questionnaire.
The fifth hypothesis will be tested by completing a mixed analysis of covariance (ANCOVA). This will use the same set up as the mixed ANOVA for previous testing but adding LOT-R scores as a covariate. This will display whether there is an interaction between optimism and NRS scores and whether optimism has a larger influence within the OLP group.
3. Results
3.1. Sample characteristics
The three groups were comparable in gender but had slight unintentional variation in age and total LOT-R scores (table 2 displays the characteristics of the sample). The sample had a mean age of 23.41 (SD - 2.51), with 66.6% of the participants being female. In total, 24 participants were included, 3 were excluded due to faults in the conditioning software and 1 due to inconsistent NRS ratings. The earlier displayed table 1 demonstrates the high international nature of the sample including the locations where they are from.
Table 2
Table to demonstrate the descriptive statistics of participant demographics
Group N (included) Age (SD) N (%) female LOT-R (SD)
OLP 9 22.88 (2.14) 6 (66.6%) 13.88 (5.62)
CP 9 23.22 (3.45) 6 (66.6%) 12.88 (3.78)
CG 6 24.5 (0.83) 4 (66.6%) 14.33 (6.12)
3.2. Placebo induction
A mixed ANOVA was used to answer multiple hypothesises. To examine if a placebo effect existed a mixed ANOVA showed a statistical significant reduction in pain between NRS scores when presented with ENS OFF and ENS ON, F(1, 24) = 4.45, p = .047.
Figure 1 displays the average difference in NRS scores when presented with ENS ON and ENS OFF in the OLP group (ON: M = 2.52, SD = 1.94 and OFF: M = 3.24, SD = 1.94), CP group (ON: M = 2.68, SD = 1.50 and OFF: M = 3.28, SD = 1.96) and the CG (ON: M = 2.99, SD = 1.91 and OFF: M = 3.28, SD = 1.72).
Figure 1
Differences in average Numerical Rating Scale (NRS) scores, from the testing phase, when presented with ENS ON (1) or ENS OFF (2)
3.3. Group effect or manipulation effect
The CG NRS scores from the testing phase were compared to the CP and OLP NRS scores from the testing phase to assess whether the placebo effect was established by our manipulations of conditioning and verbal suggestion. The same mixed ANOVA was used, and there was no statistically significant interaction between groups and NRS scores F(2, 21) = .34,
p = .716, partial η2 = .031. The main effect showed no significant difference in NRS scores between groups and manipulations F(2, 21) = .05, p = .948.
A Tukey HSD post-hoc correction was used to correct for multiple comparisons between groups. The results of the Tukey post hoc displayed no significance between CG and OLP, p = .957 and CG and CP, p = .982. Results are shown in table 3.
NR
S Sc
o
Table 3
Table displaying the results of the Tukey HSD comparing the Control Group (CG) to the experimental groups Group Mean difference Standard Error (SE) Significance Tukey HSD CG OLP .26 .93 .957 CP .17 .93 .982
3.4. Linear relationship between optimism and placebo effects
A linear regression was conducted to predict the impact of optimism on pain experienced. To assess linearity, a scatterplot of NRS scores against LOT-R scores was plotted (see figure 2). Optimism accounted for 6.7% of the variance in NRS scores with an adjusted
R2 of 2.5%. Furthermore, optimism did not statistically significantly predict NRS scores,
F(1,22) = 1.58, p = .221.
Figure 2
The linearity of Numerical Rating Scale (NRS) scores against Life Orientation Test-Revised (LOT-R) scores
3.5. Effect of optimism between groups on the placebo effect
A mixed ANCOVA was performed to assess the effect of optimism between groups and on NRS scores. There was homogeneity of regression slopes as the interaction term was not statistically significant, F(2, 18) =1.26, p = .311. According to Shapiro-Wilk’s test, standardised residuals were normally distributed p > .05. There was homoscedasticity and homogeneity of variances as assessed by Levene’s test of variance, p = .058. Optimism did not produce a statistically significant difference in the placebo effects between groups, F(2, 20) = .23, p = .796, partial η2 = .022.
4. Discussion
This study aimed to contribute to the field of chronic pain by assessing the induction of placebo effects, the possibility of administering them without deception and the impact optimism has on open label placebo effects. Placebo effects were induced using a combination of classical conditioning and verbal suggestion in a thermal heat pain procedure. Results showed that a placebo effect was induced, since all groups reported lower pain intensities when presented with ON rather than OFF. However, no significant differences were found between groups indicating the placebo effect was not attributable to our manipulations, verbal suggestion and conditioning, resulting in the rejection of the second hypothesis. Moreover, this did mean there was no significant difference in the placebo effect between the OLP group to the CP and CG indicating that deception is not necessary when administering a placebo. The effect of optimism did not have a significant relationship with the placebo effect, implying optimistic attitudes did not impact the placebo effect in any of the groups in this sample, rejecting our fourth and fifth hypotheses.
A placebo effect was present in all conditions accepting the first hypothesis. Finding a placebo effect in the CG was unexpected. This means the study cannot fully explain why a placebo effect was found but does suggest that conditioning and a verbal suggestion are not necessary to develop a placebo effect. One explanation could be that the neutral verbal suggestion was unintentionally biased. The CG was presented with a neutral verbal suggestion, stating that others have benefitted from the ENS device but may also experience no reduction in pain. As a result, the participants had expectations that the device does work for some people,
potentially meaning the testing phase was approached with self-made expectations of reduced pain. Considering both the OLP and CP groups experienced the same conditioning, the only difference between these two groups was the verbal suggestion allowing the influence of different verbal suggestions to be expressed. Yet, no differences between the OLP and CP group were seen undermining the power of a verbal suggestion. Supporting this, prior research that has tested the impact of a verbal suggestion alone and has not found it to lower perceived pain, with numerous research papers actually finding neutral verbal suggestions to increase pain reported, resulting in the requirement to consider further explanations (Mistiaen et al., 2016; Klinger et al., 2017; Rossentini, Carlino & Testa, 2018; Colloca, Sigaudo & Benedetti, 2008; Bartels et al., 2014; Reicherts et al., 2016).
The result of the CG displaying a placebo effect also suggests conditioning is not necessary to develop a placebo effect, questioning the importance of conditioning in placebo studies. This is in contrast with prior literature which has found conditioning alone to reduce pain perception, although only small effects were found when just conditioning was used (Colloca & Benedetti, 2009; Babel et al., 2017; Pazzaglia et al., 2016; Vase, Riley & Price, 2002; Klinger at al., 2007; Colloca, Siguardo & Benedetti, 2008). However, conditioning relies on repeated pairings and the effects of conditioning are stronger when more pairings are made (Jensen et al., 2015). In line with prior research, the current study used an established conditioning paradigm which has previously been successful in conditioning a placebo effect (Colloca et al., 2010). This made 10 pairings for ENS ON and a low temperature and 10 pairings for ENS OFF and a medium temperature whereas, Colloca and Benedetti (2009) made 36 pairings in their conditioning phase and Babel and colleagues (2017) made 72. Although previous literature found this conditioning paradigm to produce a placebo effect, this study could indicate that more pairings are needed to consistently produce a significantly stronger placebo effect from conditioning or that conditioning may not be as important as thought.
Another explanation for the placebo effect seen could be characteristics of the sample. As this study occurred in the Faculty of Social Sciences at Leiden University, it can be assumed many of the students had at least a basic understanding to the success of placebos already (Wager & Atlas, 2015). As established, the success of placebo is influenced by understanding how it works. Therefore, if certain students assumed they were receiving a placebo treatment and understood the mechanisms behind a placebo, experiencing a reduction in pain would be probable. This is despite the closed nature of the placebo in the CP and CG. Indeed, during the debriefing, one participant did state that they assumed the study was looking for placebo effects
When reviewing the group effect on finding a placebo effect, no significant difference between groups was seen, rejecting the third hypothesis. Prior research has expressed the success of OLP working but an assumption of deception has frequently existed (Kaptchuk et al., 2010; del Cuvillo et al., 2011; Kam-Hansen et al., 2014; Kaptchuk & Miller, 2018; Mundt, Roditi & Robinson, 2016). Positive verbal suggestions presented to OLP groups (as seen in the current study) create greater expectations and expectations, as established, are seen as significant influencers to the success of placebo treatment (Meeuwis et al., 2019; Kube et al., 2017; Leibowitz et al., 2019). The verbal suggestions presented to the OLP group in this study included phrases such as ‘big impact’, ‘definitely’ and ‘hundreds of studies’ which encourage trust and reassurance relative to the power of the words. Truthfulness has been found to be significantly important when presenting a placebo and to be completely honest in placebo treatments, having an open label design is essential (Rutherford, Sneed & Roose, 2009). This potentially explains the success seen in the OLP group in this study. Overall, non-deceptive implementation of placebo treatments did not reduce nor prevent the placebo effect, providing support for ethical use of placebos.
Alternative explanations include the suggestion that participants may have felt the influence of having to report their NRS scores directly to the experimenter and therefore, answered in accordance to what they felt the study was looking for. In addition, the length of study and some participants completing it to gain essential credits to fulfil requirements for their course, potentially resulted in some participants not maintaining full concentration as to what they actually felt and subsequently reported, creating inconsistent results.
The current study did not find optimism to significantly predict a placebo effect resulting in rejection of the fourth hypothesis. Furthermore, optimism did not affect NRS scores within the OLP group any differently to the CP group or CG rejecting the fifth hypothesis. This is in contrast to prior literature which found optimism to positively influence a placebo effect and that optimism contributes to the “placebo-prone” personality (Morton et al., 2009; Geers et al., 2010). However, Locher and colleagues (2019) had already expressed the limitations of optimism when the design was OLP, resulting in the thought that not one personality type but multiple personality factors contribute to a “placebo-prone” individual. The results of the current study support the idea that it is not just one type of personality, such as optimism, that influences a placebo effect. This could be good for the efficacy of placebo treatments as it means such treatments can be useful for a broader, non-specific population. However, although it is accepted more personality traits may contribute to developing a placebo effect, the majority of research does find optimism to significantly and positively enhance a placebo effect, so
alternative explanations to why it did not in this study, should be considered (Geers et al., 2010). One explanation could be the differences between the current study and other studies that have found the significance of optimism to exist, in particular, optimism was assessed multiple times or at the start of these studies, rather than at the end (Morton et al., 2009; Geers et al., 2010; Locher et al., 2019). The LOT-R scale assesses dispositional optimism which focuses on expectations that good things will happen in the future (Scheier et al., 1984). Therefore, assessing optimism using LOT-R at the end of a study is likely to produce different results than during a study. That is, it is expected that an individual is more likely to express optimism at the end of the study as the feeling of being tested and associated expectations are diminished. Therefore, LOT-R scores used for this study may be more reflective of finishing the study, earning money and going about their day. This may have created a higher LOT-R score than what would have been seen during the study reducing the validity of the LOT-R score for this analysis.
Another potential explanation for the insignificance of optimism in the current study is that optimists use active coping strategies, particularly when they are out of control (Urcuyo et al., 2005; Nes & Segerstrom, 2006). However, the current study offered a sense of control to the participants by reassuring them they could end the study at any point without any repercussions. As a result, those who are more optimistic may not have benefitted from the development of active coping strategies, as they were not necessary in the current study, reducing the differences between highly optimistic and lower optimistic individuals. It is important to mention that in clinical settings, active coping is more important as the patient has no control over the pain they experience so, optimism may have greater significance in such settings. In addition, assessing the difference between optimism levels and pain thresholds should be prioritised to understand if optimists feel less pain anyway. The current study did not find optimism to influence the placebo effect suggesting that optimism alone, may not be important for developing a placebo effect and multiple personality factors should be considered. However, due to the small sample size, firm conclusions about the relationship between optimism and placebo effects could not be drawn.
4.1. Strengths and Limitations
This study is one of the first to research the impact of verbal suggestion and a conditioning paradigm using application of thermal heat pain, whilst investigating OLP designs
however, due to the small sample size, any further findings are merely interpretations with no significance attached. To find significance, a ‘rule of thumb’ exists stating a requirement of 15 – 20 participants per group, the current study had 9 in each of the OLP and CP group and 6 in the CG resulting in a significantly underpowered design (Field, 2013). For this reason, it should be seen as a pilot study, providing planning and the foundations for a more complete study to follow. As a result, current findings can be seen as indications creating more accurate future study designs for subsequent research. Adding to this, it enabled recognition of flaws within the design, allowing changes to be made. For example, participants being required to report their NRS pain rating out loud may have created multiple biases. These included social desirability bias, some participants reported lower levels of pain than they actually experienced resulting in one being excluded from the study. As previously mentioned, some participants also experienced experimenter bias. As a result, pain reported may not have always been as they truly experienced, rather they reported scores to match the expectations of the researcher. Instead, allowing the participant to privately record their NRS pain score on paper or enter it directly onto the spread sheet used by the researcher, may encourage more honest reflections of pain. This could, in turn, risk resulting in more human errors but due to the simplistic nature of recording the score, this should not be of great concern. When reviewing the other scale used, LOT-R, it was recognised that it only assesses general dispositional optimism. A scale more specific of attitudes towards medicine and illness may produce different results based on education and exposure to placebo treatments.
The sample being small was not the only issue. The study wanted to assess the use of placebo treatments for chronic pain sufferers. However, the study excluded individuals suffering from chronic pain and relied on a healthy sample of predominantly students. As a result, any findings are not directly generalisable to individuals in clinical settings rather undermining the purpose of the study. Alternatively, if chronic pain patients had been used to participate in the study, results could have been impacted by current treatment plans. Adding to this, a pre-diagnosis of chronic pain is also likely to affect optimism levels, creating inconsistent attitudes depending on their current experiences of pain (Dueñas et al., 2016). To control for such variables, performing the study on healthy participants did have advantages but this should be considered when generalising findings to chronic pain patients. Another concern regarding the sample was the number of students completing the study. Students are often from similar backgrounds including higher education and socio-economic status. Not only can such life experience be reflected in optimistic attitudes and coping strategies, a significant volume of chronic pain sufferers have lower education levels, experience inequality
and are from deprived communities, vastly different to what many students experience (Boehm et al., 2015; Mills, Nicolson & Smith, 2019). A positive aspect of the sample was the high international validity. Fifteen different nationalities participated in the study resembling cultural differences in attitudes towards pain and acceptable treatment methods.
4.2. Future research
Ultimately, this study bought forth new insights into the research field of thermal heat pain and placebo treatments and can serve as a pilot study providing useful implications for future research. When issuing a neutral verbal suggestion, careful attention should be paid to ensure that it cannot be interpreted as bias as this may generate unwanted expectations. Neutral verbal suggestions should avoid giving any indication that the treatment may or may not work by simply stating there is no evidence to assess the success of the present treatment.
To consolidate or develop positive attitudes towards the use of non-deceptive placebos, focus should also be stressed on recognising the factors that enable the success of OLP designs. As deception is not acceptable in clinical settings, laboratory studies that make use of deception are not so beneficial for the target population. Therefore, once such mechanisms are fully understood, treatments can focus their implementation on the mechanisms that produce the most robust OLP effects, mitigating the requirement of deception in all circumstances whilst ensuring results of studies on non-clinical populations will be more useful to clinical settings too.
Future research should also look into other personality traits that may impact the success of placebos. Optimism has frequently been found to influence the success of inducing a placebo effect, but the current study did not replicate such findings. Existing issues with the present study may reason this, so further research into the impact of optimism should be assessed. Many other traits such as anxiety and openness could significantly impact an individual’s reaction to pain and their interest in experimenting with contemporary treatments. Uncovering the personality traits that are most impressionable by such treatments would ITprovide an understanding to who placebo treatments are most likely to be successful for.
5. Conclusion
In conclusion, a placebo effect was found in all three groups supporting the efficacy of placebo treatments in a thermal pain study. Of particular interest, a placebo effect seen in the OLP group supporting the implementation of open placebos whilst decreasing ethical concerns in clinical settings. Unexpectedly, this study also found a placebo effect in the CG indicating verbal suggestions and conditioning are not necessary for a placebo effect and other factors which were not controlled for, caused the placebo effect. Furthermore, optimism was found to have no significant effect on the placebo effect. Future studies should focus on the factors that make OLP successful and the personality traits that contribute to “placebo-prone” individuals.
Literature
American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. American psychologist, 57(12), 1060-1073.
Bąbel, P., Bajcar, E. A., Adamczyk, W., Kicman, P., Lisińska, N., Świder, K., & Colloca, L. (2017). Classical conditioning without verbal suggestions elicits placebo analgesia and nocebo hyperalgesia. PLoS One, 12(7).
Bartels, D. J., van Laarhoven, A. I., Haverkamp, E. A., Wilder-Smith, O. H., Donders, A. R. T., van Middendorp, H., ... & Evers, A. W. (2014). Role of conditioning and verbal suggestion in placebo and nocebo effects on itch. PloS one, 9(3), e91727.
Benedetti, F. (2013). Placebo and the new physiology of the doctor-patient relationship. Physiological
reviews, 93(3), 1207-1246.
Benedetti, F., Pollo, A., Lopiano, L., Lanotte, M., Vighetti, S., & Rainero, I. (2003). Conscious expectation and unconscious conditioning in analgesic, motor, and hormonal placebo/nocebo responses. Journal of Neuroscience, 23(10), 4315-4323.
Bingel, U., Wanigasekera, V., Wiech, K., Mhuircheartaigh, R. N., Lee, M. C., Ploner, M., & Tracey, I. (2011). The effect of treatment expectation on drug efficacy: imaging the analgesic benefit of the opioid remifentanil. Science translational medicine, 3(70), 70ra14-70ra14.
Blease, C., Colloca, L., & Kaptchuk, T. J. (2016). Are open‐ label placebos ethical? Informed consent and ethical equivocations. Bioethics, 30(6), 407-414.
Boehm, J. K., Chen, Y., Williams, D. R., Ryff, C., & Kubzansky, L. D. (2015). Unequally distributed psychological assets: Are there social disparities in optimism, life satisfaction, and positive affect?. PloS One, 10(2), e0118066.
Bostick, N. A., Sade, R., Levine, M. A., & Stewart, D. M. (2008). Placebo use in clinical practice: report of the American Medical Association Council on Ethical and Judicial Affairs. Journal
of Clinical Ethics, 19(1), 58.
Bräscher, A. K., Witthöft, M., & Becker, S. (2018). The underestimated significance of conditioning in placebo hypoalgesia and nocebo hyperalgesia. Pain Research and Management.
Breitkreutz, J. (2008). European perspectives on pediatric formulations. Clinical therapeutics, 30(11), 2146-2154.
Carvalho, C., Caetano, J. M., Cunha, L., Rebouta, P., Kaptchuk, T. J., & Kirsch, I. (2016). Open-label placebo treatment in chronic low back pain: a randomized controlled trial. Pain, 157(12), 2766. Charron, J., Rainville, P., & Marchand, S. (2006). Direct comparison of placebo effects on clinical and
Colloca, L., & Grillon, C. (2014). Understanding placebo and nocebo responses for pain management. Current pain and headache reports, 18(6), 419.
Colloca, L., & Howick, J. (2018). Placebos without deception: Outcomes, mechanisms, and ethics. In International review of neurobiology (Vol. 138, pp. 219-240). Academic Press.
Colloca, L., & Miller, F. G. (2011). How placebo responses are formed: a learning perspective. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1572), 1859-1869.
Colloca, L., Petrovic, P., Wager, T. D., Ingvar, M., & Benedetti, F. (2010). How the number of learning trials affects placebo and nocebo responses. Pain, 151(2), 430-439.
Colloca, L., Sigaudo, M., & Benedetti, F. (2008). The role of learning in nocebo and placebo effects. Pain, 136(1-2), 211-218.
Colloca, L., Tinazzi, M., Recchia, S., Le Pera, D., Fiaschi, A., Benedetti, F., & Valeriani, M. (2008).
Learning potentiates neurophysiological and behavioral placebo analgesic
responses. Pain, 139(2), 306-314.
Costello, N. L., Bragdon, E. E., Light, K. C., Sigurdsson, A., Bunting, S., Grewen, K., & Maixner, W. (2002). Temporomandibular disorder and optimism: relationships to ischemic pain sensitivity and interleukin-6. Pain, 100(1-2), 99-110.
Coronado, R., Lentz, T., Simon, C., Gay, C., Mackie, L., Bialosky, J., ... & George, S. (2016). Optimism moderates the influence of pain catastrophizing on shoulder pain outcome: a longitudinal analysis: opl11. Journal of Orthopaedic & Sports Physical, 46(1).
Del Cuvillo, A., Sastre, J., Bartra, J., Mullol, J., DaVila, I., Montoro, J., ... & Valero, A. (2011). Placebo effect in clinical trials involving patients with allergic rhinitis. J Investig Allergol Clin
Immunol, 21(Suppl 3), 40-45.
Dueñas, M., Ojeda, B., Salazar, A., Mico, J. A., & Failde, I. (2016). A review of chronic pain impact on patients, their social environment and the health care system. Journal of pain research, 9, 457.
Dosedlova, J., Jelinek, M., Klimusova, H. & Buresova, I. (2016). Positive Expectations – Optimism and Hope in Models. The European Proceedings of Social & Behavioural Sciences,
2357-1330.
Fässler, M., Gnädinger, M., Rosemann, T., & Biller-Andorno, N. (2009). Use of placebo interventions among Swiss primary care providers. BMC health services research, 9(1), 144.
Geers, A. L., Helfer, S. G., Kosbab, K., Weiland, P. E., & Landry, S. J. (2005). Reconsidering the role of personality in placebo effects: dispositional optimism, situational expectations, and the placebo response. Journal of psychosomatic research, 58(2), 121-127.
Geers, A. L., Kosbab, K., Helfer, S. G., Weiland, P. E., & Wellman, J. A. (2007). Further evidence for individual differences in placebo responding: an interactionist perspective. Journal of
psychosomatic research, 62(5), 563-570.
Geers, A. L., Wellman, J. A., Fowler, S. L., Helfer, S. G., & France, C. R. (2010). Dispositional optimism predicts placebo analgesia. The Journal of Pain, 11(11), 1165-1171.
Hanssen, M. M., Peters, M. L., Vlaeyen, J. W., Meevissen, Y. M., & Vancleef, L. M. (2013). Optimism lowers pain: evidence of the causal status and underlying mechanisms. Pain, 154(1), 53-58. Hoenemeyer, T. W., Kaptchuk, T. J., Mehta, T. S., & Fontaine, K. R. (2018). Open-label placebo
treatment for cancer-related fatigue: a randomized-controlled clinical trial. Scientific
reports, 8(1), 1-8.
Howe, L. C., Goyer, J. P., & Crum, A. J. (2017). Harnessing the placebo effect: Exploring the influence of physician characteristics on placebo response. Health Psychology, 36(11), 1074.
Isaacowitz, D. M. (2005). The gaze of the optimist. Personality and social psychology bulletin, 31(3), 407-415.
Jakovljevic, M. (2014). The placebo–nocebo response: Controversies and challenges from clinical and research perspective. European Neuropsychopharmacology, 24(3), 333-341.
Jakšić, N., Aukst-Margetić, B., & Jakovljević, M. (2013). Does personality play a relevant role in the placebo effect?. Psychiatria Danubina, 25(1), 0-23.
Jensen, K., Kirsch, I., Odmalm, S., Kaptchuk, T. J., & Ingvar, M. (2015). Classical conditioning of analgesic and hyperalgesic pain responses without conscious awareness. Proceedings of the
National Academy of Sciences, 112(25), 7863-7867.
Kam-Hansen, S., Jakubowski, M., Kelley, J. M., Kirsch, I., Hoaglin, D. C., Kaptchuk, T. J., & Burstein, R. (2014). Altered placebo and drug labeling changes the outcome of episodic migraine attacks. Science translational medicine, 6(218).
Kaptchuk, T. J., Friedlander, E., Kelley, J. M., Sanchez, M. N., Kokkotou, E., Singer, J. P., ... & Lembo, A. J. (2010). Placebos without deception: a randomized controlled trial in irritable bowel syndrome. PloS one, 5(12).
Kaptchuk, T. J., & Miller, F. G. (2018). Open label placebo: can honestly prescribed placebos evoke meaningful therapeutic benefits?. Bmj
Karademas, E. C., Kafetsios, K., & Sideridis, G. D. (2007). Optimism, self‐ efficacy and information processing of threat‐ and well‐ being‐ related stimuli. Stress and Health: Journal of the
International Society for the Investigation of Stress, 23(5), 285-294.
Kelley, J. M., Kaptchuk, T. J., Cusin, C., Lipkin, S., & Fava, M. (2012). Open-label placebo for major depressive disorder: a pilot randomized controlled trial. Psychotherapy and psychosomatics, 81(5).
Klinger, R., Soost, S., Flor, H., & Worm, M. (2007). Classical conditioning and expectancy in placebo hypoalgesia: a randomized controlled study in patients with atopic dermatitis and persons with healthy skin. Pain, 128(1-2), 31-39.
Klinger, R., Kothe, R., Schmitz, J., Kamping, S., & Flor, H. (2017). Placebo effects of a sham opioid solution: a randomized controlled study in patients with chronic low back pain. Pain, 158(10), 1893-1902.
Kube, T., D'astolfo, L., Glombiewski, J. A., Doering, B. K., & Rief, W. (2017). Focusing on situation‐ specific expectations in major depression as basis for behavioural experiments–Development of the Depressive Expectations Scale. Psychology and Psychotherapy: Theory, Research and
Practice, 90(3), 336-352.
Kurtz, M. E., Kurtz, J. C., Given, C. W., & Given, B. A. (2008). Patient optimism and mastery—do they play a role in cancer patients' management of pain and fatigue?. Journal of Pain and
Symptom Management, 36(1), 1-10.
Leibowitz, K. A., Hardebeck, E. J., Goyer, J. P., & Crum, A. J. (2019). The role of patient beliefs in open-label placebo effects. Health Psychology.
Legg, T. (2017) Placebos: The power of the placebo effect. Retrieved From: www.medicalnewstoday.com
Locher, C., Nascimento, A. F., Kirsch, I., Kossowsky, J., Meyer, A., & Gaab, J. (2017). Is the rationale more important than deception? A randomized controlled trial of open-label placebo analgesia. Pain, 158(12), 2320-2328.
Locher, C., Nascimento, A. F., Kossowsky, J., Meyer, A., & Gaab, J. (2019). Open-label placebo response–Does optimism matter? A secondary-analysis of a randomized controlled trial. Journal of psychosomatic research, 116, 25-30.
Madden, V. J., Harvie, D. S., Parker, R., Jensen, K. B., Vlaeyen, J. W., Moseley, G. L., & Stanton, T. R. (2015). Can pain or hyperalgesia be a classically conditioned response in humans? A systematic review and meta-analysis. Pain medicine, 17(6), 1094-1111.
Martin-Pichora, A. L., Mankovsky-Arnold, T. D., & Katz, J. (2011). Implicit versus explicit associative learning and experimentally induced placebo hypoalgesia. Journal of pain
research, 4, 67.
Meeuwis, S. H., Van Middendorp, H., Veldhuijzen, D. S., Van Laarhoven, A. I., De Houwer, J., Lavrijsen, A. P., & Evers, A. W. (2018). Placebo effects of open-label verbal suggestions on itch. Acta dermato-venereologica, 98(1-2), 268-274.
Miguez, G., Laborda, M. A., & Miller, R. R. (2014). Classical conditioning and pain: conditioned analgesia and hyperalgesia. Acta psychologica, 145, 10-20.
Miller, F. G., Wendler, D., & Swartzman, L. C. (2005). Deception in research on the placebo effect. PLoS medicine, 2(9), 262.
Mills, S. E., Nicolson, K. P., & Smith, B. H. (2019). Chronic pain: a review of its epidemiology and associated factors in population-based studies. British journal of anaesthesia, 123(2), e273-e283.
Millum, J., & Grady, C. (2013). The ethics of placebo-controlled trials: methodological justifications. Contemporary clinical trials, 36(2), 510-514.
Mistiaen, P., Van Osch, M., Van Vliet, L., Howick, J., Bishop, F. L., Di Blasi, Z., ... & Van Dulmen, S. (2016). The effect of patient–practitioner communication on pain: a systematic review. European Journal of Pain, 20(5), 675-688.
Morton, D. L., Watson, A., El-Deredy, W., & Jones, A. K. (2009). Reproducibility of placebo analgesia: Effect of dispositional optimism. Pain, 146(1-2), 194-198.
Mundt, J. M., Roditi, D., & Robinson, M. E. (2017). A comparison of deceptive and non-deceptive placebo analgesia: efficacy and ethical consequences. Annals of Behavioral Medicine, 51(2), 307-315.
Munnangi, S., & Angus, L. D. (2018). Placebo Effect. In StatPearls [Internet]. StatPearls Publishing.
Nes, L. S., & Segerstrom, S. C. (2006). Dispositional optimism and coping: A meta-analytic review. Personality and social psychology review, 10(3), 235-251.
Pacheco-López, G., Engler, H., Niemi, M. B., & Schedlowski, M. (2006). Expectations and associations that heal: immunomodulatory placebo effects and its neurobiology. Brain,
behavior, and immunity, 20(5), 430-446.
Pazzaglia, M., Haggard, P., Scivoletto, G., Molinari, M., & Lenggenhager, B. (2016). Pain and somatic sensation are transiently normalized by illusory body ownership in a patient with spinal cord injury. Restorative neurology and neuroscience, 34(4), 603-613.
