Autonomic entrainment to expressive musical phrase structures

Aims: Music reliably influences autonomic physiology, yet the mechanisms linking expressive music structures to cardiorespiratory responses remain poorly understood. This study investigated how phrase-level organization in loudness and tempo relates to autonomic entrainment during music listening. Methods and Results: We analysed multimodal physiologic recordings [electrocardiogram(ECG)-derived R-to-R intervals, respiration, continuous blood pressure (BP)] from 126 participants in the HeartFM study, each of whom listened to nine expressive solo-piano tracks drawn from a set of 30 (8 original + 22 feature-modulated versions) rendered on a reproducing piano. Music phrase structures were extracted using a Bayesian arc-detection algorithm, generating beat-level boundary profiles for loudness and tempo. Physiologic signals were converted into smooth envelopes and resampled onto the music-beat grid. Entrainment strength was quantified using Earth Mover's Distance (EMD) between boundary profiles and physiologic envelopes, and evaluated against surrogate distributions. Participant-specific surrogate testing showed that original-vs.-surrogate separation was heterogeneous across physiological signals, with the strongest trend observed for tempo-BP waveform pairings. In contrast, loudness phrases showed consistently lower EMD than tempo phrases across most tracks. Phrase-arc summary statistics (mean and variability of arc length) and music-feature similarity provided measures of predictability. These metrics correlated with entrainment strength; more regular phrase structures were associated with greater synchronization. Conclusion: Autonomic activity can be mapped to expressive phrase structures, with loudness showing stronger morphology-based alignment than tempo. Predictable phrasing and cross-feature congruence led to greater synchronisation. These findings characterize music-physiology coupling as a structure-dependent, quantifiable phenomenon, providing a mechanistic basis for feature-specific selection of music for eliciting autonomic responses.