How expectations shape pain—and what cannabis research can learn

Placebo and nocebo effects on pain provide a model for investigating how cognitive and contextual factors modulate the physiological processes underlying nociception and its regulation. However, existing accounts remain fragmented and lack a unifying mechanistic framework linking brain, behaviour and autonomic responses. Here, we synthesize evidence from neuroimaging, electrophysiology, pharmacological and experimental paradigms within a predictive processing framework, conceptualizing pain as an inferential process arising from the integration of top-down predictions and bottom-up nociceptive input weighted by their precision. Across studies, placebo analgesia is consistently associated with increased engagement of prefrontal-cingulate circuits and descending modulatory pathways, including periaqueductal grey and rostral ventromedial medulla, alongside activation of opioidergic, dopaminergic and cannabinoid systems. In contrast, nocebo hyperalgesia involves increased activity in limbic-interoceptive networks, recruitment of stress-related systems and cholecystokinin-mediated facilitation of nociceptive transmission. Within this framework, placebo and nocebo effects emerge as precision-dependent modulations of predictive signals that shape both central processing and peripheral physiological outputs, including autonomic and neuroendocrine responses. We further identify methodological approaches to operationalize priors, precision and prediction errors, and discuss how inter-individual variability may reflect differences in computational and physiological phenotypes. This integrative review offers a mechanistic bridge between cognitive context and physiological regulation, offering a novel perspective on how endogenous predictions can be harnessed to enhance treatment efficacy and mitigate adverse outcomes in clinical practice.