Chiropractic identity rethink: pain as prediction, not just mechanics

Chiropractic has long been anchored to a body-first worldview, one that treats the spine and its tissues as the principal driver of symptoms and improvement. A new commentary published in Chiropractic & Manual Therapies argues that this inherited, Cartesian-style separation of body from mind and context is increasingly out of step with contemporary pain science. The authors, led by Dave Newell and colleagues, propose instead that pain is not a direct readout of tissue state but an embodied, context-sensitive process of inference, and they argue that the profession’s future identity should be rebuilt around this understanding.

The paper, published as a comment article on 10 August 2026, draws on Active Inference, a framework in which perception and action are treated as inseparable processes of prediction and error correction. Rather than locating pain in tissues or in the brain alone, the authors describe pain as an emergent aspect of embodied sense-making, shaped by prior experience, current context, and inferences concerning threat and safety. They argue that this perspective helps explain why pain may persist in the absence of clear pathology and why touch, movement, communication, and the wider therapeutic context may be central to clinical change in non-specific musculoskeletal pain.

The commentary opens by noting that musculoskeletal pain is a leading cause of disability worldwide, with low back and neck pain among the largest contributors to years lived with disability. Persistent pain makes up a disproportionate share of this burden, and a substantial proportion of patients presenting for chiropractic care have had pain for greater than three months, a duration consistent with the ICD-11 classification of chronic pain as a unique pain condition. Yet common chiropractic explanations still tend to privilege local tissue findings and structural disturbance, a mismatch the authors argue can inadvertently reinforce threat-based beliefs and protective behaviours.

To illustrate the gap between structure and experience, the authors describe two patients with ostensibly similar radiographic findings of lumbar disc degeneration, normal inflammatory markers, and no spinal canal compromise. If pain were simply a reflection of structural damage, these two patients should report similar symptoms, yet in practice there are often marked differences in their pain, disability, and lived experience. The authors argue such differences are increasingly understood within a framework that sees pain as shaped by how bodily signals are interpreted in relation to prior experience, expectations, meaning, and context.

Active Inference, as the authors present it, rests on three central variables. Priors are learned generative models about the likely causes of afferent signals, ranging from low-level sensorimotor regularities to higher-level abstract narratives such as “my spine is fragile” or “movement will injure me.” Sensory evidence consists of real-time signals from tissues and nerves, which are inherently noisy, ambiguous, and incomplete. Precision-weighting is a context-sensitive probabilistic process that determines what is trusted more in the moment, prior expectations or sensory evidence, and it is shaped by threat, safety, attention, prior learning, and social context. Perception emerges from the dynamic interplay between these three elements.

In the context of pain, the authors explain that if high precision is assigned to higher-level threat-based priors, ambiguous sensory signals are interpreted through the lens of those priors and experienced as dangerous and painful. Conversely, if credible cues of safety or successful action experiences reduce the precision of threat priors and increase the weighting of benign interpretations, pain can reduce even when sensory input remains unchanged. They stress this is not because pain is imaginary, but because perception is always the result of inference under uncertainty. The authors use Adelson’s checker-shadow illusion as an analogy: just as the visual system can perceive different shades from identical luminance when context implies different illumination, the system generating pain can feel different levels of pain from similar bodily signals when context and prior experience imply danger.

The commentary then revisits Contextually Aided Recovery, or CARe, a concept the authors introduced in earlier work to describe the clinical encounter as replete with cues and affordances that can influence outcomes in addition to any perceived specific intervention delivered. They argue these contextual dynamics are not incidental but fundamental, and worthy of explicit recognition and deliberate curation before, during, and after hands-on care. Every clinical encounter is saturated with signals that inform high-level generative models, including how the patient is greeted, the appearance of the setting, the confidence of the clinician’s voice, and the metaphors used to describe the body and pain.

In manual therapy settings, these elements are commonly referred to as contextual factors, and the authors cite a growing body of work suggesting they can influence pain, function, satisfaction, and recovery trajectories. Within an Active Inference framing, contextual factors can be understood as modulators of priors and precision, shaping what the patient expects will happen and what their system is prepared to treat as credible evidence. The authors argue that curating context is not a soft skill or an optional add-on to the real work of hands-on care, but one of the principal levers through which chiropractic encounters shape outcomes, and in many cases may be the more decisive one.

The paper offers concrete examples of how this framework might operate in practice. Preparing a patient with persistent low back pain for spinal manipulation, the authors suggest a formulation that reattributes the audible cavitation from a potential threat signal to a benign bodily cue, reducing the precision afforded to priors such as “my spine is fragile” or “something is out of place.” They also describe graded exposure as the deliberate shaping of informative prediction error, where movements are carefully adjusted in dose or intensity so that what is experienced differs sufficiently from what is predicted to support updating, but not so markedly that the experience confirms threat and reinforces protective responses. Pain science education, they note, has been shown to reduce catastrophising and to shift unhelpful pain- and body-related conceptualisations, with self-efficacy, distress, and fear found to mediate the relationship between pain and disability.

The authors are careful to acknowledge limitations. They state that Active Inference is valued not because it offers a complete or conclusively verified theory of pain, but because it provides a unifying explanatory framework. They note that modern formulations are relatively recent, that direct empirical tests of Active Inference in pain remain limited but are beginning to emerge, and that critics are right to note the framework may be challenging to operationalise clinically. They point to placebo and nocebo literature as particularly strong experimental support for inferential accounts of pain, noting that expectations induced by suggestion, learning, and context reliably modulate pain even when sensory input is held constant. They also highlight that a 2011 study suggested less than 1% of medical curricula addressed explicit education on pain mechanisms, with more recent publications continuing to identify substantial deficits below recommended standards for pain education, and they argue chiropractic curricula must explicitly prioritise the science of perception, action, pain, and context alongside manual skills.

Subject of Research: A commentary proposing Active Inference and Contextually Aided Recovery as a framework for a contemporary chiropractic professional identity.

Article Title: Prediction, pain, and the power of context: toward a contemporary identity for the chiropractic profession

Article References: Newell, D., Segal, S., Rossettini, G., O’Neill, S., Lothe-Raven, T., Lothe, L., Weiniger, S., DuRose, A., Beira, B., Walker, K., Esteves, J. E., & Thacker, M. (2026). Prediction, pain, and the power of context: toward a contemporary identity for the chiropractic profession. Chiropractic & Manual Therapies, 34(1), Article 35. https://doi.org/10.1186/s12998-026-00669-5

Image Credits: AI Generated

DOI: 10.1186/s12998-026-00669-5

Keywords: chiropractic, pain, Active Inference, contextual factors, manual therapy, low back pain, professional identity, pain education, biopsychosocial, placebo, prediction, contextually aided recovery