10/07/2026
The fear-avoidance model has given us a useful heuristic for understanding how threatening interpretations of pain can become self-reinforcing. But I find it is often misunderstood through an excessively cognitive focus, leading us to see people in pain as somehow irrationally fearful, especially in chronic pain.
The model proposes how pain is interpreted as danger, then movement becomes something to excessively control or avoid, and short-term protection can gradually narrow activity, confidence and function. But a common dissonance arises where a person can come to cognitively understand that a movement is unlikely to be harmful, while their body still responds with apprehension, bracing, heightened arousal or a strong urge to stop.
Varangot-Reille, Pezzulo, and Thacker (2024) offer a useful reframing of this problem through predictive processing and active inference. In their account, threat is inferred through an embodied, multisensory process. The nervous system is not only drawing on tissue information or conscious belief. It is integrating movement sensation, prior learning, context, pain, and interoceptive cues such as elevated heart rate, sweating, sympathetic arousal or unease. When these internal signals are given high precision, threat may become the most coherent explanation for the whole sensory situation, even when external evidence suggests the movement is safe.
This has important implications for graded exposure and rehabilitation. Accurate reassurance is of course necessary, but it is often insufficient on its own. The clinical challenge is to help create conditions where safety can be experienced in a tolerable, repeatable and meaningful way.
Meaning we need to more closely attend to the person’s physiological state, the context of the task, their expectations, and the degree of uncertainty or threat present during movement.
Reference:
Varangot-Reille, C., Pezzulo, G., & Thacker, M. (2024). The fear-avoidance model as an embodied prediction of threat. Cognitive, Affective, & Behavioral Neuroscience, 24, 781–792. https://doi.org/10.3758/s13415-024-01199-4
