| Abstract (english) | Introduction: The analgesic placebo effect refers to the phenomenon that a certain treatment that should not lead to a reduction of pain (e.g. consumption of a pill that does not contain medicinal substances with a pharmacological effect) leads to a reduction in the sensation of pain thanks primarily to specific expectations about this effect. Explanations of how and why the placebo reaction occurs in the experience of pain can be found within the framework of various theoretical models. From the very beginnings of research on the analgesic placebo effect, there has been and is still a debate whether the placebo effect is based largely on conscious expectations (Price and Barell, 2000) or different learning mechanisms, primarily through classical conditioning (Meissner et al., 2011). However, the theoretical framework that is most accepted today is the integrative model proposed by Colloca and Miller (2011). They proposed a unique model in which the placebo effect is the result of the creation of expectations that are triggered by a psychosocial context (e.g. verbal, contextual or social), including among others conditioned stimuli (learning through classical conditioning). Conscious or unconscious expectations are conditioned on the one hand by instructions about the effectiveness of a substance or procedure, by one's own experience of exposure to a certain effective treatment for pain, or by social learning (observation of other people whose pain is reduced due to the influence of a certain substance or treatment), and on the other hand are determined by contextual stimuli, prior beliefs and previous experiences. It is believed that expectations play a key role in achieving the placebo effect. It was shown that participants who were presented with painful stimuli of equal intensity in two situations, with the only difference in the induced expectations by a verbal instruction about the intensity of the stimulus, gave lower pain estimates when they expected the stimulus to be of lower intensity and less painful, compared to the situation when they had expectations that the stimulus will be very painful (Koyama et al., 2005; Colloca et al., 2008). That the expectation of positive outcomes of the therapy is the core of the placebo effect is shown by experimental studies that have shown that only the manipulation of expectations, without applying any treatment, is effective in reducing the experience of pain (Price et al., 1980; Wang et al., 2011; Ružić et al., 2017; Ritter et al., 2019). In experimental studies of the analgesic placebo effect, expectations are most often manipulated by verbal instruction given to the participants, which states what effect the applied treatment will have and what kind of sensation it should lead to. Research clearly shows that, among other things, conditioning is important for the occurrence of the placebo effect, i.e. the temporal and spatial pairing of a neutral stimulus (placebo) and an active drug or treatment. This implicitly (but also explicitly) teaches the connection between the presence of a substance (treatment) and the outcome in behavior or experience. In this linking process, a neutral stimulus (placebo substance or procedure) becomes conditioned and can cause the same reaction as an unconditioned stimulus (one that would normally lead to a reaction by itself), which is most often a certain drug or therapeutic procedure. A typical paradigm used in experimental studies of the analgesic placebo effect using conditioning consists of causing pain to the participants and giving them an analgesic (e.g. via infusion). Its use changes the experience of pain, the pain threshold or tolerance becomes higher due to the effect of the drug. In the next step, usually the next day - with the same prior stimulation, the subjects are given a placebo substrate. If after the application of a placebo, which the participants think is the medicine they received in an earlier session, there is a reduction in pain, it can be said that a placebo effect has occurred (Ivanec, 2015). Current knowledge shows that the placebo effect is not a universal phenomenon, i.e. that there are large inter-individual and intra-individual differences. Estimates are that in some clinical or experimental research, an average of 30-40 % of patients or participants show a placebo reaction (Noon, 1999). In order for the placebo effect to be considered a real phenomenon, it was first necessary to investigate its biological and neurophysiological basis. Advances in technology, particularly the development of brain imaging techniques, have provided insight into the biological mechanisms involved in the placebo effect. According to the assumed model of the formation of the analgesic placebo effect, cognitive psychological processes such as expectations interact with structures that normally participate in the processing of nerve impulses from pain receptors (ACC, S1, insula, S2, thalamus, PAG area). The assumption is that their activation results in the release of various chemical substances, from natural endogenous opiates, but also other non-opiate endogenous substances (e.g. cannabinoid substances). The consequence of this is the modulation of the painful experience, which occurs due to the inhibition of nociceptive activation, either at the level of the spinal cord or in the brain. Such a hypothesized model is called the top-down pain modulation system (Colloca et al., 2013). Although the findings of neurophysiological basis indicate that the placebo effect is a real phenomenon, there are meta-analyses that question the strength and universality of this phenomenon. Hrobjartsson and Gotzsche (2001) showed that the placebo effect appeared only in studies that used continuous measures versus dichotomous data and only in cases where subjective measures were used. These shortcomings in the research of the analgesic placebo effect were observed relatively quickly, and there is a certain body of research that aimed to examine the objective parameters of the placebo effect. There are studies that have shown that objective parameters do not follow changes in subjective parameters of pain experience. Research by Roelofs et al. (2000) in which, in addition to the subjective evaluation of pain on the VAS scale, they also directly measured objective physiological parameters, showed that there is no significant correlation between subjective and objective measures of the analgesic effect. Findings from research that used measures of evoked potentials are particularly significant. Such measures have the advantage of registering changes in nociceptive processes that occur before any evaluation or decision-making process (Wager et al., 2006), that is, they reflect changes at the level of processing nerve impulses caused only by the strength of stimulation (Wager et al., 2006; Colloca et al., 2008). One of the objective parameters that has been shown to be directly related to the intensity of sensations from different sensory modalities are the characteristics of the motor reaction, such as reaction speed and force. None of the studies published so far have dealt with the relationship between reaction speed and force in the context of the placebo effect. The correlation between the strength of the stimulus and the characteristics of the motor reaction and the latency and amplitude of the evoked potentials, speaks in favor of the fact that variations in the speed and force of the reaction can be considered an indicator of relatively early processing of signals from nociceptors. Knowing that the placebo effect manifests itself on the measures of evoked potentials, the question is whether the placebo will also manifest itself on other objective measures, such as the speed and force of a simple motor reaction. The main basis for using reaction speed in placebo effect research lies in the fact that there is a negative relationship between stimulus intensity and reaction speed (Jaskowski and Sobieralska, 2004; Bell et al., 2006; Carreiro et al., 2011; Janssen, 2015). Also, numerous studies have shown that as the intensity of the stimulus increases, so does the force of the reaction (Jaskowski et al., 1995; Mattes and Ulrich, 1997). Although the most common objective correlates of the analgesic placebo effect are measures obtained using brain imaging techniques (fMRI, PET) which indicate which neural structures are involved in the realization of the placebo effect, it is not entirely clear whether they are an indicator of early sensory processing or they are an indicator of later higher cognitive processes, that is, decisions about giving answers are based primarily on the creation of expectations. Therefore, the basic goal in this research is to use behavioral measures that are an almost direct indicator of the strength of sensation in general, and thus the sensation of pain. Research aims and problem: It is reasonable to expect that the characteristics of a simple motor reaction can be an objective indicator of the change in the intensity of sensation that painful stimulation can cause in situations of the presence of placebo treatment compared to the control situation without such treatment. Therefore, the aim of this research, by collecting objective behavioral measures, is to verify whether the placebo effect is mostly symbolic, since it is primarily expressed through subjective assessments that may be burdened by the relationship between the participant/experimenter and the patient/therapist, as well as the context of the research itself, or whether it is a matter of changes in the nociceptive process, which is also reflected in objective behavioral measures. Such measures may be in line with subjective assessments in the evaluation of placebo treatment, but may additionally indicate a different size of effect in subjective and objective measures. On the other hand, there is a possibility that such measures are inconsistent with subjective evaluations, which would indicate that subjective evaluations are influenced by some other processes that are not closely related to the early processing of sensation intensity. An additional aim of the research was to verify whether the strength of the placebo effect would change depending on how the manipulation was carried out, by inducing expectations by giving specific instructions to the participants, or by implementing the process of classical conditioning by covertly reducing the stimulus intensity while giving instructions about the effectiveness of the placebo cream. If differences between these two situations were to be shown on the subjective measures of the assessment of the experienced pain, the question is whether the same differences would be shown on the objective measures of reaction speed and force. The research problem was to test the differences in subjective assessments of the intensity and unpleasantness of the experienced pain, the force and speed of the reaction to painful stimuli, before and after inducing the placebo effect in groups in which the placebo was induced only by expectations and by a combination of classical conditioning and inducing expectations in relation to a control group where no manipulation was performed. Methods: Participants. A total of 68 participants took part in the research, of which 50 (73.5 %) were female, with an average age of 22.13 years (SD = 3.059, min = 19, max = 34). There were 22 participants in the control group, the same number in the group whose expectations were induced only by the instruction, and 24 participants in the group whose expectations were induced both by the instruction and the conditioning process. Measures. To induce pain, electrical stimuli were used, which were given using two electrodes, connected to a voltage-controlled constant current generator (Digitimeter DS5, manufactured by Digitimer Limited, Welwyn Garden City, UK), and using a personal computer with an installed program for giving electrical stimuli. A device for measuring reaction speed and force was connected to the apparatus for giving stimuli and the computer. As a placebo, ultrasound gel was used, with a transparent blue texture, pharmacologically inactive, to which almond essential oil was added in order to obtain a specific smell similar to the smell of analgesic gels or creams. The participants rated the intensity of the experienced pain on a scale from 1 to 30, and the unpleasantness of the experienced pain on a scale from 0 - no discomfort to 10 - extremely unpleasant. In order to equalize the participants by group according to the relevant variables, the following questionnaires were applied: 1) Croatian version of The State - Trait Anxiety Inventory (STAI) used to examine anxiety as a trait and state (Spielberg et al., 1983; Spielberger, Lushene and McAdoo, 1977), 2) The Pain Catastrophizing Scale, Sullivan et al., 1995, which measures the use of catastrophizing as a coping strategy, 3) The Emotional Stability Scale taken from the International Personality Item Pool – IPIP, Goldberg (1992) and 4) Pain Sensitivity Questionnaire (PSQ; Ruscheweyh et al., 2009). Procedure. All participants came to the examination twice, where the time period between the two visits was 2 weeks. When the participants first came to the test, they were introduced to the experimental situation, the application of electrical stimuli via two electrodes to the index and ring finger of the non-dominant hand, and they were trained in the task of measuring the speed and force of the reaction, in three series of 30 reactions each, a total of 90 stimuli. All subjects were then given a total of 48 stimuli, in two series of 24 stimuli each, and the task for the participants was to press the button that measured the speed and force of the reaction as quickly as possible after the presentation of each individual stimulus. Within each series, the stimuli were presented in sequences of three stimuli of equal strength, after which the participants had to give their assessment of the perceived intensity and unpleasantness of the stimuli, on the basis of which the participants were divided into three groups. At the end of the stimulation, the participants filled out questionnaires in paper-pencil form. In the second visit, manipulation was first carried out in the experimental groups. In the expectation-only group, along with the use of the placebo cream, instructions were given that it is an extremely effective analgesic. In the second experimental group, along with instructions and the use of a placebo cream, a covert reduction in the intensity of the stimulus was carried out. Then, in all three groups, a measurement was carried out with stimuli equal to those during the first visit. The difference in the measures for assessing the intensity and discomfort of the experienced pain between the measurements during the first visit and the measurements after the manipulation in the experimental groups is an indicator of the existence of the placebo effect. Dana analysis. The normality of the distributions of the examined variables was tested with the Shapiro Wilk test. In order to test whether all three groups of participants were initially uniform in terms of all relevant variables, testing was performed using one-way analysis of variance. The existence of a placebo effect can be concluded on the basis of a comparison between the control group and the two experimental ones from the last measurement, that is, the measurement after manipulation in the experimental groups. These differences in subjective measures of perceived pain intensity and unpleasantness between groups were verified by one-way analysis of variance. The same statistical procedure was applied to test differences in reaction speed and force between groups. Results and discussion: Although earlier research has established a fairly clear biological basis of the placebo effect, it is not entirely clear whether the changes occur in the placebo effect already in the early stages of nociceptive processing of information, or whether it is a question of the involvement of higher processes, primarily the role of expectations. If objective measures were found to be inconsistent with subjective assessments, this would indicate that subjective assessments are influenced by some other processes that are not closely related to the early processing of sensation intensity. In order to be able to test the research hypotheses, it was important that the groups were initially well balanced in terms of all relevant variables, which the results showed. Contrary to the basic hypothesis of the research, no statistically significant differences were found in the evaluations of the intensity and unpleasantness of the experienced pain between the groups in the last measurement, that is, after the manipulation in the experimental groups. In other words, despite the manipulation of inducing expectations by verbal instructions in one group and the combination of the conditioning process and application of verbal instructions in the other group, no placebo effect was shown in either group. It is not uncommon for studies in which the placebo effect, despite the manipulation, was not shown. Beecher, who conducted one of the first systematic reviews of research on the placebo effect, showed that out of the 14 experiments presented, only one used a control group and no significant differences between the control and treatment groups were shown (Beecher, 1955). In the remaining 13 experiments in which the placebo effect was demonstrated, no control group was used. Hrobjatsson and Gortzsche, with their meta-analysis of 114 studies (2001) and an additional 44 studies (2004), showed that the placebo effect appeared only in studies that used continuous subjective assessments and in studies that primarily included the measurement of pain experience. For measures that were based on observer assessments and dichotomous variables, the placebo effect was not significant. Their meta-analysis triggered an avalanche of research on the placebo effect, which was conducted with the aim of verifying whether the placebo effect is real or the result of a series of methodological errors such as statistical regression, spontaneous recovery, or something else. Some of the earlier studies showed that using the conditioning process does not produce a placebo effect (Flaten et al., 2018; Rhudy et al., 2018). While others have shown that even the use of specific instructions does not lead to a placebo effect (Roelofs et al., 2000; Vambheim et al., 2021). When looking at the average estimates of pain intensity in individual groups, it is evident that the stimuli were experienced as medium strong, that is, the stimulus intensities were high enough to reduce the pain intensity experience in the participants. It is recommended that, in order to demonstrate the placebo effect, participants with pain assessments of at least 3 or 4 on a scale of 0 to 10 should be included in the research (Dworkin et al., 2010). Intensity ratings were highest in the control group, which is in line with hypothesis, but were slightly lower in the group in which expectations were induced only by the instruction, although it was expected that the manipulation of both the instruction and the conditioning process would produce a greater placebo effect than the manipulation of the instruction only. Previous research has shown that participants who have positive previous experiences with analgesics show a greater placebo effect than those who have had less positive experiences (Benedetti, 2014; Finniss et al., 2010; Ropper et al., 2020; Benedetti et al., 2022). Almost half of the participants stated that they do not use analgesics at all or use them a couple of times a year, so this may be a potential reason why the placebo effect was not shown. When examining the placebo effect, it is also necessary to ensure that all environmental cues are convincing so that the participants do not suspect placebo manipulation. Most often, for this reason, a double-blind methodology is used in research, where neither the participants nor the technician directly conducting experiment are aware of which participants are receiving the real drug and which are placebo. In this case, the research was conducted by only one female experimenter, so the influence of the experimenter's expectations, if it existed, was constant during the entire research. An attempt was made to make the research environment suitable for real pharmacological research as much as possible, but it is possible that additional work could have been done on it. For example, it would be more appropriate to conduct this type of research in a medical institution where the initial expectations of the participants - those with which the participants come to the research, would probably be more in the direction of belief in the effectiveness of the applied 'medicine'. Other potential methodological limitations of the research were discussed in the paper, and recommendations were given for further research on objective parameters in the placebo effect. Conclusion: Using one way analysis of variance, no statistically significant differences were found in the evaluations of the intensity and unpleasantness of the experienced pain between the groups after conducting the manipulation in the experimental groups. In other words, despite the manipulation of inducing expectations by verbal instructions in one group and the combination of the conditioning process and application of verbal instructions in the other group, no placebo effect was shown in either group. There were no significant differences in the speed and force of the reaction between control and experimental groups. |
