Heart rate and insula activity increase in response to music in individuals with high interoceptive sensitivity

Abstract

Interoception plays an important role in emotion processing. However, the neurobiological substrates of the relationship between visceral responses and emotional experiences remain unclear. In the present study, we measured interoceptive sensitivity using the heartbeat discrimination task and investigated the effects of individual differences in interoceptive sensitivity on changes in pulse rate and insula activity in response to subjective emotional intensity. We found a positive correlation between heart rate and valence level when listening to music only in the high interoceptive sensitivity group. The valence level was also positively correlated with music-elicited anterior insula activity. Furthermore, a region of interest analysis of insula subregions revealed significant activity in the left dorsal dysgranular insula for individuals with high interoceptive sensitivity relative to individuals with low interoceptive sensitivity while listening to the high-valence music pieces. Our results suggest that individuals with high interoceptive sensitivity use their physiological responses to assess their emotional level when listening to music. In addition, insula activity may reflect the use of interoceptive signals to estimate emotions.

Fig 1. Schematic diagram of the analysis of the heartbeat discrimination task.

The horizontal axis shows the delay time of the beep from R-peak, while the vertical axis shows the percentage of trials in which the participants responded that the timing was matched. The triangular symbol indicates the extrapolated condition considering the periodicity of the heartbeat (see details in the text). The dotted line shows the Gaussian approximation.

Fig 2. Changes in pulse rate grouped by interoceptive sensitivity defined by heartbeat discrimination task performance.

The participants were divided into two groups based on the heartbeat discrimination task amplitude parameter. The graph on the left shows the average result of the high-IS group, while the graph on the right shows the average result of the low-IS group. The vertical axis shows the change in pulse rate based on the fixation time before the trial. The zero on the horizontal axis indicates the time when the music started. IS, interoceptive sensitivity.

Fig 3. Spearman’s correlation coefficient between valence level and pulse rate.

The solid dot indicates each participant and the open dot indicates the median for each group. IS, interoceptive sensitivity.

Fig 4. Relationship between the average valence score for all music pieces and interoceptive sensitivity.

The relationship between the average valence score for all music pieces and interoceptive sensitivity was defined by the amplitude parameter of the heartbeat discrimination task. Each point indicates each participant’s data. The solid line shows a result of linear regression.

Fig 5. Activations associated with valence level observed by the parametric modulation analysis.

Uncorrected p < 0.001 at voxel-level, FWE-corrected p < 0.05 at cluster level. SMA, supplementary motor area; STG, Superior temporal gyrus; aIns, anterior Insula; FWE, family-wise error.

Fig 6. Mean beta values of parametric modulation with valence level in the insula subregion.

Each panel represents the insula subregions showing significant differences between the high- and low-IS groups in the ROI (region of interest) analysis. The error bars represent ±1 standard error.