The probability of self-related stimuli modulates heartbeat-evoked potentials.

Interoception, the cortical processing of visceral signals, has been proposed as a foundation of self-processing. Heartbeat-evoked potentials (HEPs) provide a measure of interoceptive processing and have been linked to self-processing, but whether HEP modulations reflect self-specificity rather than familiarity and depend on probabilistic context remains unclear. We recorded EEG and ECG during passive listening to subject's own names (SON), familiar names (FN), and unfamiliar name (UN) under equal- and biased-probability conditions. In the equal-probability condition, HEP amplitudes were more positive for SON than for FN and UN over centro-parietal-occipital sites at 164-242 and 258-296 ms after the R peak, whereas FN and UN did not differ. This self-specific pattern was not detected under biased probabilities. Within-name comparisons revealed probability-related HEP effects for SON and UN with distinct spatiotemporal characteristics, indicating that stimulus probability did not produce a uniform effect across name types. The late positive potential showed a comparable probability-dependent self-specific pattern, and SON-evoked LPP amplitudes positively covaried with HEP amplitudes during SON perception across conditions. These findings indicate that self-specific HEP modulations depend on probabilistic context and is associated with relatively late external self-related processing, highlighting a context-sensitive relationship between interoceptive and external self-related processing.