Cardiovascular synchronisation to music : blood pressure entrainment to expressive musical structures

Abstract Background Music is increasingly recognised as a potential non-pharmacological intervention for cardiovascular regulation 1. Prior research suggests that music features such as loudness and tempo influence autonomic responses, including blood pressure (BP) 2,3; however, how music properties entrain BP remains unclear. Entrainment, the synchronisation of physiological rhythms with external stimuli, can be quantified using Earth Mover’s Distance (EMD) to produce a similarity measure between music and physiology. Understanding BP-music entrainment could inform personalised music-based cardiovascular therapies. Purpose This study aims to evaluate the effect of anticipatory mechanisms and music feature modulations on BP entrainment to different music pieces and their performed variations to advance the use of music in cardiovascular therapy. Methods Continuous BP was recorded from 92 participants (60 women, mean age 42 (95%CI:39.2-44.6)) while each individual listened to 9 out of a pool of 30 piano tracks 4. A Bayesian algorithm 5 computationally detected tempo and loudness phrase arc boundaries, which denote arc-like changes in expressive music features and drive listeners' responses. The BP waveform was interpolated to the music’s beat time. Its envelope, computed using the Hilbert transform, was compared with the phrase arc outputs for tempo and loudness. Results for each track were compared to surrogate data (generated from participants' responses to other tracks) to compute statistical significance. This helps determine whether BP response was influenced by the music rather than expected variation in the individual’s physiology. Paired t-tests compared BP entrainment to loudness vs. tempo. Pearson’s r coefficient assessed relationships between phrase arc statistics and entrainment strength. Results In 25 out of 30 tracks, BP entrains more to loudness than tempo, with significantly smaller curve differences (P5.5e-07). Predictable tracks in the dataset are characterised by having (1) short mean phrase arc lengths with small standard deviations and (2) longer track durations with an increased number of phrase arcs. Increased predictability allows the listener to anticipate phrase changes leading to higher synchronisation. BP entrainment has average r=0.57 and r=0.65 for the mean arc length and standard deviation, and r=-0.66 and r=-0.57 for the number of arcs and track duration. Coupling between BP and tempo is greatest when tempo arcs closely resemble loudness arcs (Fig.1c and 1d) with r=0.87. See Fig.2 for correlations and visualisations. Conclusion Blood pressure response entrains more strongly to loudness than tempo. Tracks with predictable music structures led to stronger entrainment, supporting the hypothesis that anticipatory mechanisms play a role in cardiovascular synchronisation to music. Future research will investigate personalised music-based BP modulation strategies for patients with hypertension or autonomic dysfunction.