Pain-Reporting Variability and Reliability as Predictors of Placebo Effects: A Cross-Sectional Experimental Pain Study

Placebo analgesia varies substantially across individuals, yet the sources of this heterogeneity remain incompletely understood. Pain-reporting variability may represent an underrecognized predictor of placebo responsiveness. This study examined whether variability and reliability of pain reporting predict placebo analgesia in an experimental pain setting.

Eight hundred and three participants (401 individuals with temporomandibular disorder and 400 healthy controls) completed a standardized thermal pain paradigm involving calibration, placebo conditioning, and testing phases. We obtained repeated heat temperatures (four) during thermal calibration and pain ratings during conditioning test (24 trials). We used these measurements to quantify within-person pain-reporting variability using standard deviation (SD) and coefficient of variation (CoV), and reliability using intraclass correlation coefficients (ICC). Placebo analgesia was calculated as the difference in pain ratings between control and placebo cue trials during testing. Regression models examined associations between reporter characteristics and placebo analgesia controlling for age, sex, race and experimenters.

Variability of thermal pain responses during calibration did not predict placebo analgesia. Greater pain-reporting variability during conditioning was associated with reduced placebo analgesia (higher SD and CoV), and this effect was mediated by slower acquisition of the cue– pain contingency. Conditioning-phase reliability was not associated with placebo analgesia, whereas three clusters of learning profiles predicted placebo effects. Thus, individual differences in pain-reporting variability during conditioning, rather than baseline sensory variability, contribute to heterogeneity in placebo analgesia. These findings suggest that variability during acquisition processing matters more than general sensory variability influencing the magnitude of placebo effects.