Impaired Processing of Bodily Signals Is Associated With Borderline Personality Traits: Insights From the Predictive Coding Model

Borderline Personality Disorder (BPD) is marked by emotional instability and intense negative affect. The Predictive Coding Model posits that impaired prediction of bodily signals may underlie such deficits, though empirical validation remains lacking. In this study, we compared a high BPD-trait group (score ≥ 6 on the Chinese Version of the McLean Screening Instrument for BPD, without a BPD diagnosis) with a low BPD-trait group (score ≤ 2). Participants were examined using the Multidimensional Assessment of Interoceptive Awareness-2 (MAIA-2) for interoceptive sensitivity (IS), the Heartbeat Counting Task (HBCT) for interoceptive accuracy (IA) and heartbeat-evoked potentials (HEP) as the neural correlates of cardiac interoceptive processing. Results showed that the high BPD-trait group scored lower in the two subscales of the MAIA-2 (Not-Worrying and Trusting); however, no differences were observed for the HBCT. Furthermore, HEP amplitudes were significantly modulated by attention in the low BPD-trait group, with higher amplitudes during the heartbeat-focused condition, whereas the high BPD-trait group exhibited consistently high amplitudes across both conditions. Within the predictive coding framework, the observed pattern (intact IA, low interoceptive trust and altered HEP responses) may indicate a rigidity in precision weighting that impairs adaptive interoceptive modulation in individuals with BPD traits.