Our main finding was that pleasurable music modulated MOR system activity in several cortical and subcortical regions and that pleasurable chills during music listening were associated with BPND in the nucleus accumbens, linking striatal MOR release to subjective experience. The fMRI results revealed that acutely music-induced pleasure was associated with activity in regions implicated in interoception, emotion, and reward. Finally, the PET-fMRI fusion analysis demonstrated that higher baseline MOR availability was associated with stronger pleasure-dependent BOLD responses particularly in the reward circuit as well as in motor regions. Altogether, our results represent the first in vivo neuroimaging evidence supporting the mediating role of the MOR system in the experience of music-induced pleasure and underline how interindividual variability in the MOR system is associated with individual differences in the tendency for reward responses to music.

The PET data revealed that pleasurable music modulated opioidergic activity in central nodes of the reward circuits such as the ventral striatum and OFC containing”hedonic hotspots” regulated by opioids [9]. While prior [11C]carfentanil PET studies on reward processing have mainly focused on biologically salient primary rewards such as feeding [32], sex [14], and sociability [15], our results indicate that aesthetic rewards, such as music, that are strongly influenced by cultural learning, also modulate MOR system activity. This suggests that MORs mediate pleasure across domains, driving humans to pursue primary and non-primary rewards and make choices in the face of conflicting options [33]. Due to the well-established role of endogenous opioids in affiliative behavior and social rewards, the MOR system has been hypothesized to contribute particularly to the prosocial effects of musical activities involving synchronized movements in groups, such as joint music-making and dancing [2, 20]. Our results, however, indicate that even solitary listening to pleasurable music without overt movement can modulate MOR system activity.

We observed higher BPND in the music condition compared to the control condition. A decrease in BPND is typically interpreted as evidence for heightened endogenous neurotransmitter release, in line with competition between the radiotracer and synaptic neurotransmitters [34, 35], while an elevation in [11C]carfentanil binding may signify MOR "deactivation", a reduction in synaptic endogenous opioids [25]. However, BPND encompasses both receptor density and affinity, and augmented radioligand binding may indicate an increase in the available receptors or enhanced binding affinity. While prior studies indicate that pleasurable stimuli often decreases [11C]carfentanil BPND, pleasurable touch, laughter, and social acceptance [15, 25, 36] have also been reported to heighten regional MOR availability. The current design does not allow us to disentangle whether the observed effects are due to increased or decreased opioid tone.

However, we found that the number of music-induced chills was negatively correlated with [11C]carfentanil binding in the nucleus accumbens. As chills are widely recognized as a proxy for intense music-induced pleasure, this finding indicates that such pleasurable experiences are associated with increased endogenous opioid release in the NAcc, which competes with carfentanil for receptor binding [34]. This result aligns with animal studies demonstrating the central role of opioid release in the NAcc in mediating pleasure and parallels evidence of striatal dopamine release during music-induced pleasure [6]. The increased [11C]carfentanil binding in the music scan relative to the baseline scan, may reflect a more general effect, such as heightened arousal, rather than pleasure specifically. In contrast, the observed opioid release in the NAcc appears to be more directly tied to the pleasurable aspects of the music experience itself.

While endogenous opioids have often been hypothesized to mediate musical hedonia [5, 17, 37], an alternative proposal is that the dopamine system constitutes the main neurochemical pathway for music-induced pleasure [28]. This assertion is supported by a PET study demonstrating heightened dopamine release during music-induced chills [6] and pharmacological data showing that dopamine receptor blockade diminishes music-induced pleasure and dopamine agonist administration had the opposite effect [38]. Animal studies have, however, demonstrated that injecting opioid agonists into the striatum increases liking reactions, whereas dopamine antagonist injections in these sites or chemical (6-OHDA) lesions of the striatal dopaminergic neurons do not reduce liking responses [9, 39]. Moreover, one human study has reported opioid antagonist-dependent weakening of pleasure induced by music [26], yet this outcome has not been replicated in subsequent research [27, 28]. The dopamine and opioid systems interact at the molecular level [40]. For instance, opioid release in the ventral tegmental area (VTA) has been shown to modulate dopamine release in the nucleus accumbens (NAc), providing a potential pathway through which both systems could contribute to music-induced pleasure [41]. It is thus likely that the interplay the opioid and dopamine system, and possible interactions with other systems such as oxytocin [42], plays a role in shaping the experience of music-induced pleasure and arousal [37].

Our fMRI results revealed that activity in the orbitofrontal cortex (OFC) and dorsal striatum correlated with the subjective experience of music-induced pleasure. Prior research shows that OFC activity tracks subjective pleasure from rewards like food and sexual cues [9]. The dorsal striatum, the caudate and putamen are consistently activated by rewards, including liked music [30, 43, 44].

Pleasure-dependent activation was also observed in the ACC and insula [44, 45], linked to visceral signal processing and the physiological arousal accompanying music-induced pleasure [46, 47]. Pleasurable music elicited increased heart rate and stronger pupillary responses, indicating heightened autonomic activity, consistent with prior studies showing that music-induced pleasure coincides with emotional arousal [48]. Insula activation, common to music and food rewards [4], may reflect its role in emotional and bodily responses central to music. Activation in the right pre/postcentral gyri and supramarginal gyrus indicates somatomotor system involvement, mirroring bodily sensations or movement simulations triggered by liked music [49, 50]. The pleasure-dependent BOLD responses co-localized with music-induced BPND changes in the ACC, and frontal pole, among other regions, suggesting that increased blood flow during music-induced pleasure may partly reflect the MOR system's metabolic demands.

Baseline MOR availability was associated with hemodynamic pleasure-dependent responses: Participants with a higher concentration of MORs exhibited stronger haemodynamic pleasure responses, particularly in the ACC, insula, and auditory cortex as well as in the NAc. Thus, our results indicate that individual variation in MOR tone influences pleasure responses in regions associated with bodily, auditory, and reward processing during pleasurable music listening, which may explain individual differences in subjective music-induced emotional experience. It is noteworthy, however, that prior studies have shown that baseline MOR availability is associated with functional BOLD responses to both positive, reward-related stimuli, such as food pictures and laughter sounds, as well as negative stimuli, like violent videos [16, 31, 51]. MOR availability has also been associated with haemodynamics responses reflecting emotional arousal irrespective of valence. Together, these findings suggest the endogenous opioid system has a highly general role in modulating the processing of a wide range of emotionally evocative events irrespective of valence and biological saliency.

Due to the complexity of PET neuroreceptor imaging, our sample size was relatively small, which may have compromised the ability to detect small effects. However, the PET results were robust, consistent across participants, and localized in all the expected regions, suggesting that the statistical power allowed us to detect biologically meaningful effect sizes. For reasons outlined in the supplementary methods, we only included female participants, which may limit the generalizability of our results to males.