PAINCODE – PAINCODE

Chronic primary pain and comorbid anxiety are major health burdens, disproportionately affecting women and poorly treated by current therapies. Both conditions may share a root in disrupted predictive processing - rigid threat-related priors and impaired belief updating - often shaped by early adversity. These symptoms are thought to arise from dysfunction in a conserved cortico-thalamo-limbic circuit including the medial prefrontal cortex (mPFC), paraventricular thalamus (PVT), central amygdala (CeA), and nucleus accumbens (NAc). Yet, how circuit-level dynamics cause symptom persistence remains unclear.
PAINCODE addresses this gap using a cross-species, forward-translational approach. In rodents, we combine miniscopic calcium imaging, in vivo electrophysiology, and optogenetics to decode and causally test predictive signals within the mPFC-PVT-CeA-NAc circuit. In humans with fibromyalgia, parallel predictive learning tasks assess belief updating using EEG and fMRI. A shared hierarchical Bayesian model links neural dynamics to behavior across species and tracks symptom trajectories over six months.
PAINCODE delivers: (1) mechanistic, cross-species biomarkers of predictive dysfunction; (2) causal evidence for predictive mPFC-PVT dysfunction in pain and anxiety; and (3) validated prognostic markers for personalized care. This work reframes chronic pain as a predictive coding disorder, paving the way for biologically grounded diagnosis and intervention.