The Predictive Brain in Pain: Control, Expectation, and Updating Maladaptive Priors

Session Title: The Predictive Brain in Pain: Control, Expectation, and Updating Maladaptive Priors

Topic: EXPERIMENTAL (NON-CLINICAL) HUMAN STUDIES

Description of Workshop: Chronic pain remains one of the most complex phenomena in neuroscience, not only for its sensory dimension but for its deep entanglement with cognition, emotion, and learning. This symposium brings together leading researchers including Christian Büchel, and Pavel Goldstein and the early career investigator Nandini Raghuraman, to explore how predictive coding and Bayesian models illuminate the dynamic interplay between expectation, perceived control, and pain modulation.
Christian Büchel opens the session by framing chronic pain through the Bayesian pain model, where pain persistence is conceptualized as the result of over-precise, negatively biased priors, expectations of threat and pain that dominate over sensory evidence. Drawing on a series of neuroimaging experiments, Büchel demonstrates how perceived control can recalibrate these priors by altering both their mean (the expected magnitude of pain) and their precision (the confidence in these predictions). His data show that control over pain treatment or stimulus intensity not only reduces perceived pain but also enhances top-down regulation within the descending pain modulatory system. These findings position control as a crucial mechanism for modifying maladaptive pain expectations, an insight with direct implications for chronic pain interventions.
Nandini Raghuraman extends this framework by dissecting the distinct contributions of expectations and endogenous opioid mechanisms in placebo hypoalgesia. Using a rigorous 2×2 double-blind naloxone-controlled design, Raghuraman from Colloca lab demonstrates that placebo-induced pain relief is mediated by endogenous opioids but remains independent from expectations themselves. Building on this, she presents novel evidence from an open-label placebo (OLP) trial showing that six weeks of OLP use significantly reduces chronic pain intensity. Mediation analyses reveal that lower pain expectations partially explain the analgesic effects, suggesting that expectation updating rather than expectations per se drives pain relief. These findings reinforce the notion that expectation modulation is a fundamental mechanism in pain regulation and a viable therapeutic target.
Finally, Pavel Goldstein introduces Personalized Danger Signal Reprocessing (PDSR), a novel group-based intervention that operationalizes predictive coding principles into clinical practice. Rooted in the concept of threat priors and maladaptive learning, PDSR combines psychoeducation, somatosensory and emotional reprocessing, and daily-life integration coaching to promote model updating and perceived control. Across two studies, PDSR yielded robust improvements—over 50% of participants achieved sustained pain recovery, accompanied by significant reductions in anxiety, depression, and PTSD symptoms. Mediation analyses identified perceived control and curiosity as key drivers of recovery, suggesting that fostering flexible, curiosity-driven reappraisal of pain predictions may underlie therapeutic change.
Together, these talks provide a compelling framework for computational, neurobiological, and behavioral approaches to better understanding and treating chronic pain. By linking expectation precision, perceived control, and model updating, this session highlights a paradigm shift from managing symptoms to modifying the predictive processes that generate them.

Engagement and Q&A will be facilitated by Luana Colloca who will serve only as a chair (no speaker) helping with moderating the session.
This workshop includes diversity of gender, culture (US, Germany, and Israel), and career stages.

Speakers