Nocebo effects are stronger and more persistent than placebo effects in healthy individuals

Reviewer #1 (Public review):

Summary:

This manuscript presents a study on expectation manipulation to induce placebo and nocebo effects in healthy participants. The study follows standard placebo experiment conventions with the use of TENS stimulation as the placebo manipulation. The authors were able to achieve their aims. A key finding is that placebo and nocebo effects were predicted by recent experience, which is a novel contribution to the literature. The findings provide insights into the differences between placebo and nocebo effects and the potential moderators of these effects.

Specifically, the study aimed to:

(1) assess the magnitude of placebo and nocebo effects immediately after induction through verbal instructions and conditioning
(2) examine the persistence of these effects one week later, and
(3) identify predictors of sustained placebo and nocebo responses over time.

Strengths:

An innovation was to use sham TENS stimulation as the expectation manipulation. This expectation manipulation was reinforced not only by the change in pain stimulus intensity, but also by delivery of non-painful electrical stimulation, labelled as TENS stimulation.

Questionnaire-based treatment expectation ratings were collected before conditioning and after conditioning, and after the test session, which provided an explicit measure of participants' expectations about the manipulation.

The finding that placebo and nocebo effects are influenced by recent experience provides a novel insight into a potential moderator of individual placebo effects.

Weaknesses:

There are a limited number of trials per test condition (10), which means that the trajectory of responses to the manipulation may not be adequately explored.

On day 8, one stimulus per stimulation intensity (i.e., VAS 40, 60, and 80) was applied before the start of the test session to re-familiarise participants with the thermal stimulation. There is a potential risk of revealing the manipulation to participants during the re-familiarization process, as they were not previously briefed to expect the painful stimulus intensity to vary without the application of sham TENS stimulation.

The differences between the nocebo and control conditions in pain ratings during conditioning could be explained by the differing physiological effects of the different stimulus intensities, so it is difficult to make any claims about expectation effects here.

A randomisation error meant that 25 participants received an unbalanced number of 448 trials per condition (i.e., 10 x VAS 40, 14 x VAS 60, 12 x VAS 80).

Reviewer #2 (Public review):

Summary:

Kunkel et al aim to answer a fundamental question: Do placebo and nocebo effects differ in magnitude or longevity? To address this question, they used a powerful within-participants design, with a very large sample size (n=104), in which they compared placebo and nocebo effects - within the same individuals - across verbal expectations, conditioning, testing phase, and a 1-week follow-up. With elegant analyses, they establish that different mechanisms underlie the learning of placebo vs nocebo effects, with the latter being acquired faster and extinguished slower. This is an important finding for both the basic understanding of learning mechanisms in humans and for potential clinical applications to improve human health.

Strengths:

Beyond the above - the paper is well-written and very clear. It lays out nicely the need for the current investigation and what implications it holds. The design is elegant, and the analyses are rich, thoughtful, and interesting. The sample size is large which is highly appreciated, considering the longitudinal, in-lab study design. The question is super important and well-investigated, and the entire manuscript is very thoughtful with analyses closely examining the underlying mechanisms of placebo versus nocebo effects.

Weaknesses:

There were two highly addressable weaknesses in my opinion:

(1) I could not find the preregistration - this is crucial to verify what analyses the authors have committed to prior to writing the manuscript. Please provide a link leading directly to the preregistration - searching for the specified number in the suggested website yielded no results.

(2) There is a recurring issue which is easy to address: because the Methods are located after the Results, many of the constructs used, analyses conducted, and even the main placebo and nocebo inductions are unclear, making it hard to appreciate the results in full. I recommend finding a way to detail at the beginning of the results section how placebo and nocebo effects have been induced. While my background means I am familiar with these methods, other readers will lack that knowledge. Even a short paragraph or a figure (like Figure 4) could help clarify the results substantially. For example, a significant portion of the results is devoted to the conditioning part of the experiment, while it is unknown which part was involved (e.g., were temperatures lowered/increased in all trials or only in the beginning).

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

This manuscript presents a study on expectation manipulation to induce placebo and nocebo effects in healthy participants. The study follows standard placebo experiment conventions with the use of TENS stimulation as the placebo manipulation. The authors were able to achieve their aims. A key finding is that placebo and nocebo effects were predicted by recent experience, which is a novel contribution to the literature. The findings provide insights into the differences between placebo and nocebo effects and the potential moderators of these effects.

Specifically, the study aimed to:

(1) assess the magnitude of placebo and nocebo effects immediately after induction through verbal instructions and conditioning

(2) examine the persistence of these effects one week later, and

(3) identify predictors of sustained placebo and nocebo responses over time.

Strengths:

An innovation was to use sham TENS stimulation as the expectation manipulation. This expectation manipulation was reinforced not only by the change in pain stimulus intensity, but also by delivery of non-painful electrical stimulation, labelled as TENS stimulation.

Questionnaire-based treatment expectation ratings were collected before conditioning and after conditioning, and after the test session, which provided an explicit measure of participants' expectations about the manipulation.

The finding that placebo and nocebo effects are influenced by recent experience provides a novel insight into a potential moderator of individual placebo effects.

We thank the reviewer for their thorough evaluation of our manuscript and for highlighting the novelty and originality of our study.

Weaknesses:

There are a limited number of trials per test condition (10), which means that the trajectory of responses to the manipulation may not be adequately explored.

We appreciate the reviewer’s comment regarding the number of trials in the test phase (i.e., 10 trials per condition). This trial number was chosen to ensure comparability with previous studies employing similar designs and research questions (e.g. Colloca et al., 2010). Our primary objective was to directly compare placebo and nocebo effects within a within-subject design and to examine their persistence one week after the first test session. While we did not specifically aim to investigate the trajectory of responses within a single testing session, we fully agree that a comprehensive analysis of the trajectories of expectation effects on pain would be a valuable extension of our work. We will acknowledge this limitation and future direction in the revised manuscript.

On day 8, one stimulus per stimulation intensity (i.e., VAS 40, 60, and 80) was applied before the start of the test session to re-familiarise participants with the thermal stimulation. There is a potential risk of revealing the manipulation to participants during the re-familiarization process, as they were not previously briefed to expect the painful stimulus intensity to vary without the application of sham TENS stimulation.

We thank the reviewer for the opportunity to clarify that participants were informed at the beginning of the experiment that we would use different stimulation intensities to re-familiarize them with the stimuli before the second test session. We are therefore confident that participants perceived this step as part of a recalibration rather than associating it with the experimental manipulation. We will add this information to the revised version of the manuscript.

The differences between the nocebo and control conditions in pain ratings during conditioning could be explained by the differing physiological effects of the different stimulus intensities, so it is difficult to make any claims about expectation effects here.

We appreciate the reviewer’s comment and agree that, despite the careful calibration of the three pain stimuli, we cannot entirely rule out the possibility that temporal dynamics during the conditioning session were influenced by differential physiological effects of the varying stimulus intensities (e.g., intensity-dependent habituation or sensitization). We will address this in the revision of the manuscript, but we would like to emphasize that the stronger nocebo effects during the test phase are statistically controlled for any differences in the conditioning session.

A randomisation error meant that 25 participants received an unbalanced number of 448 trials per condition (i.e., 10 x VAS 40, 14 x VAS 60, 12 x VAS 80).

We agree that it is unfortunate that 25 participants were conditioned with an unbalanced number of trials per condition during the conditioning session. In the revised version of the manuscript, we will include additional analyses to demonstrate that this imbalance did not systematically bias the results and that the findings observed during the test phase remain robust despite this error.

Reviewer #2 (Public review):

Summary:

Kunkel et al aim to answer a fundamental question: Do placebo and nocebo effects differ in magnitude or longevity? To address this question, they used a powerful within-participants design, with a very large sample size (n=104), in which they compared placebo and nocebo effects - within the same individuals - across verbal expectations, conditioning, testing phase, and a 1-week follow-up. With elegant analyses, they establish that different mechanisms underlie the learning of placebo vs nocebo effects, with the latter being acquired faster and extinguished slower. This is an important finding for both the basic understanding of learning mechanisms in humans and for potential clinical applications to improve human health.

Strengths:

Beyond the above - the paper is well-written and very clear. It lays out nicely the need for the current investigation and what implications it holds. The design is elegant, and the analyses are rich, thoughtful, and interesting. The sample size is large which is highly appreciated, considering the longitudinal, in-lab study design. The question is super important and well-investigated, and the entire manuscript is very thoughtful with analyses closely examining the underlying mechanisms of placebo versus nocebo effects.

We thank the reviewer for their positive evaluation of our manuscript and for acknowledging the large sample size, methodological rigor, and the significant implications for clinical applications and the broader research field.

Weaknesses:

There were two highly addressable weaknesses in my opinion:

(1) I could not find the preregistration - this is crucial to verify what analyses the authors have committed to prior to writing the manuscript. Please provide a link leading directly to the preregistration - searching for the specified number in the suggested website yielded no results.

We apologize that the registration number alone does not directly lead to the preregistration of this study. We thank the reviewer for pointing this out and will include a link to the preregistration in the revised manuscript. This study was pre-registered with the German Clinical Trial Register (registration number: DRKS00029228; https://drks.de/search/de/trial/DRKS00029228).

(2) There is a recurring issue which is easy to address: because the Methods are located after the Results, many of the constructs used, analyses conducted, and even the main placebo and nocebo inductions are unclear, making it hard to appreciate the results in full. I recommend finding a way to detail at the beginning of the results section how placebo and nocebo effects have been induced. While my background means I am familiar with these methods, other readers will lack that knowledge. Even a short paragraph or a figure (like Figure 4) could help clarify the results substantially. For example, a significant portion of the results is devoted to the conditioning part of the experiment, while it is unknown which part was involved (e.g., were temperatures lowered/increased in all trials or only in the beginning).

We thank the reviewer for this comment and suggestion. In the revised version, we will restructure the manuscript and include more detailed information about the key experimental procedures and design at the beginning of the Results section to enhance clarity and improve the interpretability of the reported findings.