Symptoms and the body: Taking the inferential leap

This review addresses a fundamental challenge in medicine: why the subjective experience of physical symptoms often fails to correspond with objective physiological dysfunction. The paper focuses on medically unexplained symptoms (MUS)—prevalent in primary care affecting up to 81.6% of the population—and critiques the traditional somatosensory amplification model, which attributes MUS to stress-related arousal, hypervigilance, and misinterpretation of bodily sensations. The authors demonstrate that existing evidence does not convincingly support the amplification model, particularly regarding peripheral physiological abnormalities, heightened interoceptive accuracy, or cognitive misinterpretations as primary causes.

The authors propose a comprehensive new framework grounded in predictive coding theory, which views symptom perception as an inferential process rather than a direct recording of bodily signals. According to this model, the brain continuously generates and tests hypotheses about internal bodily states based on prior expectations and past experience, weighted by confidence in those predictions. Symptoms emerge when the brain interprets sensory information in light of these predictions, a process heavily influenced by precision optimization—how the brain weights incoming signals relative to existing beliefs. The framework integrates neurobiology, interoception research, and existing theoretical models into a unified account explaining both "explained" and "unexplained" symptoms dimensionally rather than categorically.

The model identifies key factors influencing symptom perception including varieties of bodily input, attentional processes, gender differences, threat perception, and negative affect. Critically, the authors show that the relationship between physiological dysfunction and symptom reports varies substantially even in well-defined diseases like asthma, COPD, and diabetes, suggesting that many symptoms in medically explained conditions could technically be classified as "unexplained" if measured systematically. This finding dissolves the artificial distinction between explained and unexplained symptoms.

The predictive coding framework has significant clinical and theoretical implications. It suggests that MUS should be understood as somatovisceral illusions reflecting normal perceptual processes rather than pathological mechanisms, potentially reducing stigma and enabling more effective interventions. The authors propose that future treatments should target the precision weighting of bodily signals and illness-related predictions rather than assuming peripheral dysfunction. The model also explains diagnostic implications—reconceptualizing somatic symptom disorder beyond the mind-body dichotomy—and suggests specific testable hypotheses for empirical validation, offering a foundation for integrating biomedical and psychological perspectives on symptom perception.