Neural Correlates of Aberrant Heart–Brain Coupling in Schizophrenia

Schizophrenia disrupts integration of internal (interoceptive) and external (exteroceptive) signals, splitting subjective from objective reality. Early psychosis features positive symptoms linked to exteroceptive deficits; stable-phase patients show blunted autonomic responses and negative symptoms, indicating interoceptive dysfunction. Conventional fMRI or HEP measures capture only static or early neural markers, missing the dynamic, multi-scale heart–brain interplay vital for interoceptive processing. We propose aberrant heart–brain coupling—the bidirectional interaction of cardiac rhythms and brain activity—as a core link between interoceptive failure and symptom diversity. In health, coupling balances bottom-up interoceptive input and top-down predictive control; in schizophrenia, altered dynamics may reveal dominance of top-down priors (misattributions, hallucinations) or bottom-up signals (fragmented interoceptive states, affective flattening). Measuring directional coupling directly tests these predictive coding accounts. In 80 participants (40 clinically stable patients, 40 healthy controls), we combine a heartbeat detection task with simultaneous ECG and EEG. Cardiac activity is indexed via cleaned interbeat intervals (mean and log-transformed variation per epoch). Instantaneous HRV is extracted from ECG wavelets. HEPs are derived from EEG locked to R-peaks, without baseline correction. Directional coupling is assessed using Gaussian Copula Mutual Information: EEG is filtered into low-frequency bands, Hilbert-transformed, copula-normalized, and paired with ECG phase; systematic phase offsets isolate bottom-up (MIHEART–BRAIN) and top-down (MIBRAIN–HEART) components. This design identifies neural markers of interoceptive deficits, distinguishes tonic from task-specific coupling abnormalities, and links multimodal signatures to dimensional symptoms and traits measured by PANSS, MCCB, MAIA, SPQ, and THISQ. We predict patients will show reduced HEP amplitudes, weaker directional MI (especially MIBRAIN–HEART), and altered prediction error signaling (smaller P300). Task-related HEP–accuracy and HRV–reaction time couplings will be diminished, correlating with motivational and perceptual deficits; heightened HEP–P300 coupling during ambiguity may reflect maladaptive uncertainty monitoring. By mapping direction-sensitive, frequency-resolved heart–brain dynamics, we transform “interoceptive dysfunction” into precise biomarkers, clarify whether deficits arise from chronic autonomic imbalance or context-specific inference failure, and advance a trans-scale model of schizophrenia grounded in disrupted interoceptive–allostatic processing.