Can contagious itch be affected by positive and negative suggestions?

Abstract Contagious itch can be evoked by observing people scratching. Verbal suggestions about to‐be‐received itch can influence itch intensity, as shown by placebo research, but it is unknown whether this extends to contagious itch. The current study aimed to replicate prior findings that listening to scratching and rubbing sounds elicits contagious itch, and to investigate whether suggestions can modulate this process. Healthy participants (n = 140) received positive or negative suggestions about itch in response to the sounds (aimed to decrease or increase expected itch, respectively), or no specific suggestions as a control. Participants listened to a number of audio fragments with scratching and rubbing sounds. The amount of expected itch as well as itch sensation after each audio fragment were measured by self‐report. Suggestions had no effect on the expected itch. Both rubbing and scratching sounds significantly elicited itch in all groups. Scratching sounds induced more itch than rubbing sounds exclusively in the control group. These findings indicate that short suggestions might be not effective enough to modify the expectations of people regarding contagious itch. Furthermore, suggestions modulate contagious itch to some degree, but not in the hypothesized direction. Potential similarities and differences in the neurobiological mechanisms of contagious itch and nocebo effects are discussed.

gious itch, and to investigate whether suggestions can modulate this process. Healthy participants (n = 140) received positive or negative suggestions about itch in response to the sounds (aimed to decrease or increase expected itch, respectively), or no specific suggestions as a control. Participants listened to a number of audio fragments with scratching and rubbing sounds. The amount of expected itch as well as itch sensation after each audio fragment were measured by self-report. Suggestions had no effect on the expected itch. Both rubbing and scratching sounds significantly elicited itch in all groups. Scratching sounds induced more itch than rubbing sounds exclusively in the control group. These findings indicate that short suggestions might be not effective enough to modify the expectations of people regarding contagious itch.
Furthermore, suggestions modulate contagious itch to some degree, but not in the hypothesized direction. Potential similarities and differences in the neurobiological mechanisms of contagious itch and nocebo effects are discussed.

K E Y W O R D S
contagious itch, expectancy, itch, nocebo effects, placebo effects Psychological processes play an important role in how itch is experienced. A biopsychosocial model of itch has been proposed that emphasizes the role of psychological factors in itch experience. 8,9 For instance, stress and anxiety exacerbate chronic itch in patients with dermatological conditions. 10,11 The influence of individuals' expectancies on itch is also underlined by research into placebo and nocebo effects. [12][13][14][15][16][17] Placebo effects are positive effects (e.g., reduced itch) that do not emerge due to active treatment components, but are rather elicited by non-active components of the treatment or its psychosocial context, for instance when these factors evoke positive expectancies about treatment outcomes. 18 Nocebo effects are negative effects such as increased itch, side effects or reduced treatment efficacy elicited by the context. 19 Placebo and nocebo effects have often been induced in cutaneous itch through expectancy modulation, that is, by giving positive or negative suggestions about some form of a sham treatment. 12,13,15 For example, Darragh and colleagues 15 found that participants experienced less itch after a cream was applied, when they were told that it was an antihistamine.
Little is known about whether itch evoked by non-cutaneous stimuli may also be sensitive to expectancy effects. So far, a single study investigated whether suggestions can influence contagious itch in a 3 (catastrophizing information, simple information, no information) by 2 (patients, healthy controls) design. 20 The authors found that catastrophizing information (i.e., suggestions that the induced itch would be extremely unpleasant) led to increases in scratching during and after itch-related audiovisual stimuli, compared with a group that was simply informed that they may experience itch in response to the stimuli. The group that was not warned about itch did not differ from the group that received catastrophizing information, which could be related to heightened arousal. 20 Moreover, the effects of suggestions were only seen in patients with dermatological conditions, and not in healthy controls, suggesting that patients may be more sensitive to itch-related catastrophizing information. 20 Itch intensity was not altered in response to the catastrophizing information though, in either patients or healthy controls. While these findings provide us with information about the potential impact of negative information and expectations, it is not clear how contagious itch may change when positive information evokes expectancies of low itch. For instance, it is possible to prevent nocebo effects by explaining them as typical human responses to health warnings. [21][22][23] Similarly, such positive explanations may help to alleviate contagious itch. Negative suggestions, on the contrary, could draw attention to and as such increase itch elicited by audiovisual cues. 24 The present study had two main goals: first, we aimed to replicate the results of previous research and investigate whether sounds of scratching induce sensations of itch to a higher extent than rubbing sounds. Secondly, we studied whether it is possible to change the expectancies of participants regarding contagious itch and modulate contagious itch intensity using negative and positive suggestions.
Additionally, we explored whether interindividual differences in, for instance, anxiety, worrying and optimism, may moderate the effects of suggestions on contagious itch.

| Study design
The study was approved by the Psychology Research Ethics Committee of Leiden University (number 2020-05-08-A.W.M. Evers-V3-2410). The study protocol was preregistered on Open Science Framework (osf.io/muqcj). A randomized between-within subject study design was applied. Participants were randomly allocated to one of three groups: (1) negative suggestions group, (2) positive suggestions group or (3) control group (no specific suggestions).
The distribution of males and females (i.e., based on self-reported sex) was equal across groups.

| Participants
Healthy participants between 18 and 60 years old were recruited for this study. Participants were recruited online using social media sites such as Facebook, and Sona; a participant database of Leiden University (ul.sona-syste ms.com). The exclusion criteria were as follows: chronic itch experienced within the past three months, use of painkillers, sleep-inducing medication, alcohol or other drugs 24 h before participation, caffeine consumption 1 h before participation.
A power analysis was conducted in GPower version 3, 25 with the aim to determine the optimal sample size to detect differences in the intensity of contagious itch among the three study groups. Input for the power analysis was derived from two studies: (1) a study that investigated the influence of positive suggestions on itch intensity following topical histamine application 14 was used for calculating a sample size needed to detect the between-group differences (the effect of suggestions); (2) a study that investigated the effect of scratching and rubbing sounds on the subjective itch intensity 5 was used for calculating a sample size needed to detect any within-group differences (i.e., any difference in contagious itch elicited by scratching relative to rubbing sounds), in order to ensure that we could replicate prior research findings. The power analysis indicated that to detect between-group differences in subjective itch experience using analysis of variance with 3 groups, with an estimated effect size of f = 0.42, 14 a critical alpha level of α = 0.05 and a power of 1−β = 0.95, a number of 31 participants per group would be needed.
To detect within-group differences in contagious itch using an analysis of variance with an estimated effect size of f = 0.17, 5 a critical alpha level of α = 0.05 and a power of 1−β = 0.95, a number of 46 participants per group would be needed.
Taking into consideration that the manner of providing suggestions differs from prior work (i.e., a combination of verbal and written suggestions in Darragh and colleagues, 14 versus completely online in the current study), and that the auditory itch induction in the previous study 5 was also induced in a laboratory setting, the number of participants per group was adjusted to a more conservative 50 participants per group.

| Experimental interventions
The suggestions and the quality of the sounds were checked in a pilot that included 17 participants. The length of the suggestions was matched in all groups except the control group, which received slightly shorter instructions. The full text of the suggestions and results of the pilot are presented in Appendix 1.

| Negative suggestions group
In this condition, participants were given information about contagious itch and were informed that most people experience itch after hearing the scratching and rubbing sounds, even when being aware of the phenomenon.

| Positive suggestions group
In this condition, participants received information on contagious itch, and were informed that being aware of the contagiousness of itch will minimize its impact.

| Control group
In this condition, participants did not receive any explicit suggestions and were merely told that in the following part they would be asked to listen to various sounds, and to rate the amount of itch they experienced afterwards.

| Itch induction
The sound task used in this study for itch induction is described in detail in previous research. 5 The sounds comprised recordings of scratching and rubbing different targets, including 3 body parts (beard, hand, leg) and 3 non-body materials (polyester, denim, leather). Each sound was administered three times, with different compositions: high frequency (HF) tones above 1000 Hz were either increased or decreased in amplitude (i.e., volume) by 10 dB relative to other tones in the audio recordings. This resulted in 3 different versions of each sound: HF tones' amplitude −10 dB, HF tones' amplitude unchanged and HF tones' amplitude +10 dB. In total, participants listened to 36 sounds, of which 18 were scratching and 18 were rubbing sounds, and each sound lasted for 20 s. The sounds were presented to participants in three blocks, using a pseudorandom order (i.e., every block consisted of 6 scratching and 6 rubbing sounds, and the blocks were randomized).

| Measurements and questionnaires
Expected itch of participants was measured after the suggestions were given, but before the sounds were presented. Participants were asked to predict how much itch they think they will experience during listening to the sounds on a 0-10 numeric rating sale (NRS; 0-"no itch" to 10-"worst itch imaginable").
Experienced itch was measured after the presentation of each sound. Participants were asked to indicate on a 0-10 NRS how much itch they experienced while listening to the sound (0-"no itch" to 10-"worst itch imaginable").

| Questionnaires
Itch experience questionnaire was developed by one of the authors (HH) for familiarizing participants with an 11-point NRS for itch.
Participants were asked whether they ever experienced: (1) a mosquito bite; (2) an itchy scalp (for example, after using wrong kind of shampoo); (3) itchy feet/toes (for example, athletes' foot). In case participants positively replied to these questions, they were asked to use the NRS and rate how much itch these experiences elicited (0-"no itch" to 10-"worst itch imaginable"). As this scale was primarily used for familiarizing participants with rating itch, no total score was computed.
Interindividual differences in the following constructs were as-

| Procedure
The study was advertised online as a study that investigated individual differences in itch sensitivity. Interested volunteers were sent a link to the online survey platform Qualtrics (Provo, UT & Seattle, WA, USA), containing the study's information letter. In case they wanted to participate, they were asked to sign an online informed consent form.
Participants were then asked to fill out a short online screening questionnaire, that checked the inclusion and exclusion criteria. In case of ineligibility for participation, they were automatically referred to the end of the survey and participation discontinued. Eligible participants were asked to listen to a sample sound (i.e., a beeping tone) and to report how well they could hear it. They were asked to wear headphones (noise-cancelling if available) during the task to improve the quality of the sounds. Then, they were given the itch experience questionnaire. Subsequently, participants were randomly allocated to one of the three groups: negative suggestions, positive suggestions or a control group and received various instructions depending on their group allocation (see paragraph "Experimental interventions"). Participants then rated how much itch they expected to experience while listening to the audio recordings. Next, participants listened to the sounds of skin and material scratching and rubbing.
After each of the 20 s sounds, participants were asked to report the level of itch that they felt on the NRS. Subsequently, the participants filled out the rest of the questionnaires. Following this, participants were debriefed. The study lasted around half an hour and was conducted in English. Participants were compensated with either 1 course credit (for Psychology students of Leiden University) or the option to take part in a lottery and win 15 euros.

| Statistical analysis
All analyses were conducted using SPSS 26.0 for windows (IBM SPSS Inc., Chicago, Illinois, USA). Prior to analyses, all variables were checked for normal distribution and the assumptions of the preplanned statistical methods were checked. All variables were normally distributed except for age and mean experienced itch scores. χ 2 tests, analysis of variance (ANOVA) and a Kruskal-Wallis nonparametric test (i.e., for age) were used to examine between-group differences in demographic variables and individual characteristics.
Because mean itch scores were not normally distributed, and no non-parametric equivalent of the mixed model repeated-measures ANOVA exists, square root transformation was performed on itch scores. First, a one-sample t-test was used to compare the wholesample mean for expected itch to zero, to check whether participants expected the sounds to elicit itch. This was followed by an ANOVA to assess whether the positive suggestions, negative suggestions and control group differed in expected itch.
Second, we assessed whether contagious itch could be elicited with a one-sample t-test, in which we compared whole-sample mean itch for scratching and rubbing sounds across all groups to zero. Next, a 2 (sound type: scratching vs. rubbing) × 3 (HF amplitude: −10 dB, original, +10 dB) within-subjects repeated-measures ANOVA was conducted to assess whether prior study findings 5 could be replicated. Because the suggestions in the experimental groups were aimed at manipulating itch levels, this analysis was conducted within the control group exclusively (a secondary analysis of the replicability of prior research findings including the positive and negative suggestion groups was conducted. The results of this analysis can be found in Appendix 2). Bonferroni post hoc tests were performed for each factor within this analysis.
Third, the effects of verbal suggestions on itch were assessed using a 3 × 2 mixed ANOVA, with group allocation (positive suggestions, negative suggestions, control) as between-subjects factor and sound type (scratching, rubbing) as within-subjects factor. Itch ratings were averaged across HF amplitude for this analysis. Planned Bonferroni-corrected comparisons for itch were performed: (1) between positive suggestions and the control group, averaged across sound type; (2) between negative suggestions and the control group, averaged across sound type; (3) between positive suggestions and the control group for each sound type (scratching, rubbing) separately; (4) between negative suggestions and the control group for each sound type (scratching, rubbing) separately; and (5) between scratching and rubbing sounds within each separate group. To estimate whether the care with which participants read the instructions may influence the results, a sensitivity analysis was performed based on the total duration of the survey. Participants who were overly quick or slow (>1.5 standard deviation (SD) slower or faster than the mean duration time) were excluded in this sensitivity analysis.
Fourth, explorative analyses were conducted to investigate whether individual characteristics modulated the effects of suggestions on itch elicited by the scratching and rubbing sounds. For these analyses, itch ratings were also averaged across HF amplitude.
As a first step, Pearson's correlation coefficients were calculated to explore whether interindividual differences were associated with average itch elicited by scratching and rubbing sounds in general.
To explore whether interindividual differences influenced the effects of the suggestions on contagious itch elicited by the scratching and rubbing sounds, multiplicative moderation analysis was then conducted with the MEMORE macro. 31 The moderation analyses were conducted twice: once to compare the positive suggestions group and once to compare the negative suggestions group with the control group. Prior to the analysis, assumptions for ordinary least square (OLS) regression were checked. Moderation effects were probed using the "pick-a-point" approach and set at −1 SD, mean and +1 SD for the continuous variables. Confidence intervals (CI's) were generated by percentile bootstrapping set at 5000 samples.
All values are reported as arithmetic means ± SD unless stated otherwise. An alpha <0.05 is considered statistically significant.
Because the itch scores were negatively skewed and violated the assumption of normal distribution of the residuals, square root transformations were performed on the data prior to all analyses.

| Participants
In total, 157 participants started the study and were randomized into the three groups. Four participants did not complete the study, and data of 17 participants were excluded post hoc because they did not meet the inclusion criteria. The final sample consisted of 140 participants (84.3% female) with an age between 18 and 57 years old (20.64 ± 5.02). Most of the participants had university education (80.9%) and were of Dutch nationality (59.6%). χ 2 tests, analysis of variance and Kruskal-Wallis non-parametric tests demonstrated no differences between the groups in demographics or on the assessed questionnaires (all p ≥ 0.12; Table 1).

| Expected itch
A one-sample t-test showed that, prior to listening to the scratching and rubbing sounds, participants in all groups expected to experience itch when listening to the sounds; t(135) = 20.03, p < 0.001.
The suggestions did not modulate the amount of itch participants expected to experience prior to listening to the sounds F(2,133) = 2.09, p = 0.13, η 2 partial = 0.03. For an overview of the means per group for each outcome, see Table 2.

| Experienced itch
An overview of the itch scores can be found in

| Effects of verbal suggestions on contagious itch
The 3 × 2 mixed ANOVA demonstrated no significant main effect of group on itch sqrt ; F(2,130) = 0.07, p = 0.94, η 2 partial < 0.01. This indicates that overall, the suggestions did not modulate the experience of contagious itch. The planned comparisons confirm that itch in general did not differ between the positive suggestions and control group (both p ≥ 0.11), nor between the negative suggestions and control group (both p ≥ 0.11).
A main effect of sound type was found within the analysis; F(1,130) = 20.43, p < 0.001, η 2 partial = 0.14. Across all groups, participants reported more itch following the scratching sounds  The largest categories have been described -all other nationalities (n = 20 in total, with n < 5 for each respective nationality) have been listed as "other." b When participants answered "yes" on the question whether they experienced itch before, for example, from a mosquito bite or itchy scalp (the proportion that answered "yes" is presented in the table), they were then asked to rate itch intensity on a 0-10 numeric rating scale (Mean ± SD in the table).
exclusively: the planned comparisons showed, that within this group, more itch was elicited by the scratching than the rubbing sounds (p < 0.001; Figure 1). In the positive and negative suggestion groups, on the contrary, itch ratings following scratching and rubbing sounds did not differ significantly (both p ≥ 0. 19). No differences were detected between the positive suggestions and control group, or the negative suggestions and control group, in itch elicited by the scratching and rubbing sounds separately (all p ≥ 0.11; Figure 1).
A sensitivity analysis revealed that these findings did not change, when participants who filled out the survey very quickly or slowly (>1.5 SD of the mean duration; 33.2 ± 14.65 in min) were excluded.

| Moderation of group effects by interindividual differences
In general, interindividual differences (i.e., in anxiety, worrying, optimism and personality) were not correlated with itch (all p ≥ 0.067; Table S1). Moreover, interindividual differences did not moderate the effects of positive or negative suggestions on contagious itch (Tables S2 and S3)

| DISCUSS ION
The current study had a twofold aim: (1)  That contagious itch can be induced by scratching and rubbing sounds is in line with previous research findings. 5,32 Similar to the study conducted by Swithenbank and colleagues, 5 scratching sounds elicited itch to a higher extent than rubbing sounds and increases in HF amplitude were associated with higher auditory itch contagion, which confirmed our first hypothesis. Contrary to our second hypothesis, the positive and negative suggestions about contagious itch failed to modulate participants' itch expectancies as well as their general itch experience. These findings are somewhat unexpected in light of previous work that shows that verbal suggestions can modulate expectancies about, and the experience of, cutaneous itch e.g., 12,33,34 On the contrary, in previous research catastrophizing information only influenced scratching responses and did not affect contagious itch. 20 These effects were moreover only found in patients with dermatological conditions, but not in healthy controls. Speculatively, patients may be more sensitive to catastrophizing information, 20 and this may also be the case for information that is positively framed. However, positive expectancies about itch may be more easily induced in itch-free populations; given that patients likely have a more extensive history with this symptom and may therefore hold more persistent negative expectancies towards itch-evoking stimuli.
An interaction between suggestions and sound type (scratching versus rubbing) was found, which indicated that the pattern in itch ratings found in prior research 5 and the control group-that is, that scratching sounds elicit significantly more itch compared with rubbing sounds-was absent following suggestions. Instead, itch ratings following rubbing sounds were somewhat higher, and itch ratings following scratching sounds somewhat lower after positive and negative suggestions relative to the control group (although between-group comparisons were non-significant). Thus, the suggestions equalized itch responses to scratching and rubbing sounds.
Participants may have assumed following suggestions that all sounds would elicit itch, instead of monitoring carefully how each sound felt, which would account for this finding. However, this interpreta- Finally, it has been proposed that attention and expectancy may be separate constructs, that are both able to influence contagious itch. 24 While the suggestions aimed to elicit positive expectancies and lower itch, they may have also redirected more attention towards physical sensations, which has been shown to be associated with increased itch sensitivity. 36  It is not yet known whether the underlying neurobiological mechanisms of contagious itch and nocebo effects in cutaneously evoked itch are similar. Conceptually, the experience of contagious itch has been linked to nocebo effects before. 16,37,38 In a sense, this comparison seems reasonable: contagious itch can be elicited by stimuli from the surrounding context (i.e., itch-related cues). Experimental studies also show that initially neutral contextual cues can become associated with itch and may exacerbate physical itch sensations (i.e., conditioned nocebo effects in itch, for instance, using abstract cues such as colours 39,40 ). However, nocebo effects have always been studied explicitly with physical or chemical induction of itch (i.e., cutaneous itch), which is different from contagious itch elicited by non-physical cues such as sounds or pictures. The neurobiological underpinnings of nocebo effects in itch, as well as those of contagious itch, are not yet well understood. Current evidence points towards a role of brain areas involved in the somatosensory processing of contagious itch. 1,[41][42][43] It has also been hypothesized that activation of the mirror neuron system may be relevant. 4,44 There is some data supporting this: among other, brain regions involved in the simulation of actions also appear involved in contagious itch. 1,45 These areas differ from the brain areas that have been associated with nocebo effects fol- In the current study, higher self-assessed sensitive skin was associated with larger differences in contagious itch intensity between scratching and rubbing sounds in healthy, itch-free individuals.
These findings complement prior findings that itch and scratching contagion are elevated in patients with skin disorders e.g., 3,5,43,47 For instance, prior research demonstrates that itch in response to scratching sounds is amplified in psoriatic patients compared with healthy individuals. 5 This is the first time that an association between more sensitive skin and higher auditory itch contagion following scratching sounds has been found outside of a clinical population. Sensitive skin is associated with more frequent and intense subjective complaints, such as itch, in response to external stimuli. 48 Individuals with more sensitive skin may be more likely to experience itch and may, as a consequence, be exposed more frequently to scratching sounds. This could in theory have sensitized them to contagious itch. More research is needed to replicate these findings.
Some limitations of the current work need to be addressed.

AUTH O R CO NTR I B UTI O N S
All authors contributed to the design of the study. Stefanie H.

ACK N OWLED G EM ENTS
None.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available  The difference in itch elicited by scratching compared to rubbing sounds changed significantly across sensitive skin (SS10) ratings for the control group, but not for the negative suggestions group (see Table S3 for the statistical data). Moderation analysis indicates that the difference in itch between scratching and rubbing sounds was significant for medium (M) and high (+1 SD), but not for low (−1 SD) levels of sensitive skin in the control group. The difference was non-significant for the negative suggestions group regardless of sensitive skin ratings

FIGURE S4
Itch levels evoked by the scratching sounds (A) and rubbing sounds (B) respectively, within the negative suggestions group and control group and plotted across low (−1 SD), medium (M) and high (+1 SD) levels of sensitive skin (SS10). Even though no significant group × SS10 interaction effect was found for itch elicited by either scratching or rubbing sounds, differences in how itch changes across levels of sensitive skin for each group may have contributed to the significant group × SS10 × movement type interaction (see also Figure S3 and Table S3)

FIGURE S5
The difference in auditory itch elicited by scratching compared to rubbing sounds changed significantly across sensitive skin (SS10) ratings for the control group, but not for the positive suggestions group (see Table S2 for the statistical data). Moderation analysis indicates that the difference in auditory itch between scratching and rubbing sounds was significant for medium (M) and high (+1 SD), but not for low (−1 SD) levels of sensitive skin in the control group. The difference was non-significant for the positive suggestions group regardless of sensitive skin ratings interaction (see also Figure S5 and Table S2)