Expectations in Pain Research: Measured or Assumed?

Abstract

This journal recently published a paper by Lunde et al. (2026) entitled ‘To ask or not to ask: Does inquiring about participants' expectations influence the placebo effect? A placebo mechanism study in healthy volunteers’. The study addresses whether measuring expectations influences placebo analgesia.

In 64 healthy volunteers, placebo effects were numerically larger when participants rated expected pain before stimulation, but between-group differences were not statistically significant. As the authors acknowledge, the study was not powered to detect these differences or establish equivalence. The results therefore leave the question of measurement reactivity open and call for further evaluation of expectation assessment. The study is important because it also prompts two broader questions: what evidence establishes expectations as a mechanism, and what exactly do expectation measures capture?

Expectations are widely proposed as a principal mechanism of placebo and nocebo effects, and these effects are sometimes defined in terms of expectations. Yet their involvement is sometimes inferred without measurement. When verbal suggestion, classical conditioning, observational learning or their combination produces placebo hypoalgesia/analgesia or nocebo hyperalgesia, the outcome alone does not establish that expectations changed or caused the effect. Induction procedures and their mechanisms must be distinguished. Inferring expectation mediation from the occurrence of a placebo or nocebo effect treats a theoretical explanation as an empirical finding (Bąbel 2019).

This problem also concerns the definition of expectations. In Kirsch's response expectancy theory, expectations concern anticipated nonvolitional reactions, such as pain (Trusz and Bąbel 2016). These expectations are understood as consciously accessible and, in principle, reportable (Bąbel 2019). Lunde et al.'s interviews are informative here: participants predominantly described expectations as thoughts or assumptions about future outcomes. Measuring expectations is therefore crucial for understanding their role.

Proposals that expectations can be unconscious or implicit require an explicit definition and evidence independent of the effect being explained. Inferring unreported expectations simply because pain changed makes the explanation circular (Bąbel 2019). Neither missing expectation measures nor unsuccessful mediation analyses establish that implicit expectations produced the effect. Unconscious learning remains a possible explanation, but labelling it expectancy without independent evidence obscures distinctions between mechanisms. Conversely, a null association with reported expectations does not rule out their involvement; measurement sensitivity and statistical precision must be considered.

Expectations and pain should also be examined as separate outcomes. Learning procedures may change expected pain even when no corresponding change in experienced pain is detected. Without expectation measures, this could be mistaken for a wholly unsuccessful manipulation, although expectations have changed. The converse pattern is equally informative: placebo hypoalgesia/analgesia or nocebo hyperalgesia can occur without evidence that measured expectations predict or mediate it (Bąbel 2019). Measuring both outcomes helps distinguish whether a procedure modifies expectations and whether those expectations contribute to pain.

In Lunde et al.'s study, between-condition differences in expected pain intensity and unpleasantness predicted the corresponding placebo effects. These results support measuring expectations, although prediction alone does not demonstrate mediation. The authors also note in their introduction that asking about expected pain intensity versus expected pain relief can yield different conclusions about expectations' contribution. The instructions and rating scale therefore need to be taken into account when interpreting expectation ratings. Retrospective reports can add context but may be influenced by subsequent experience.

The implications extend beyond placebo and nocebo effects. Lunde et al. also found associations between expected and experienced pain in the no-treatment condition. These findings are relevant to understanding the relationship between expectations and pain regardless of whether a placebo intervention is used, although they do not establish causality. Distinguishing expectations from pain outcomes remains essential in both contexts.

Future studies should combine clear definitions and prospective assessment with adequately powered tests of measurement reactivity. Randomising the timing or wording of questions could help distinguish expectations' contribution from the contribution of asking about them. Testing mediation also requires designs and analyses suited to the proposed causal pathway. These efforts would help clarify when expectations mediate pain changes and when other mechanisms require consideration.

Przemysław Bąbel conceived, wrote and revised the commentary, approved the final manuscript, and accepts responsibility for all aspects of the work.

This commentary was supported by the National Science Centre, Poland, under grant no. 2020/38/A/HS6/00470.

The author declares no conflicts of interest.

Lunde et al., To Ask or Not to Ask: Does Inquiring About Participants' Expectations Influence the Placebo Effect? A Placebo Mechanism Study in Healthy Volunteers, https://doi.org/10.1002/ejp.70384.