Music and ‘chills’ – the brain’s reward system in action

Article: Valorie N Salimpoor, Mitchel Benovoy, Kevin Larcher, Alain Dagher & Robert J Zatorre.  Anatomically distinct dopamine release during anticipation and experience of peak emotion to music. Nature Neuroscience.  Free Access Here

See BBC Coverage here

A few weeks ago I wrote a blog about the experience of musical chills or ‘shivers down the spine’ – that unique, intense, peak emotional as well as physical experience that we can get when listening to our favourite piece of music. Pavarotti singing ‘Panis Angelicus’ does it for me – what a voice!

Well, this topic must be in fashion as today sees the pre-release online launch of a new Nature Neuroscience article looking at the dynamic time course of the chills response in the brain. The article reveals that the brain reacts differently to both the anticipation of chills and to the chills themselves, in a kind of two stage evolution of emotional reaction which is mirrored in the way our brain reacts to other biologically significant stimuli, such as food and sex, as well as synthetics that mimic their neurochemical effects, such as drugs.

This paper adds a unique finding to the growing force of literature which suggests that music is linked to the deepest and, evolutionarily speaking, oldest emotion and reward mechanisms in the brain. 

Valorie Salimpoor has written about the rewarding aspects of musical listening and musical chills before, and her methods maintain a uniquely high level of validity by using real music, while still employing the rigors of scientific control and neuroscientific techniques.

Her method involves recruiting a large number of people who experience musical chills and asking them to report the tunes that reliably trigger the experience. Then, in order to create a controlled condition, she uses those same chill tunes on participants for whom that music does not trigger chills. So in both the chill and control conditions, overall, the same music is played. Simple but beautifully effective!

In her new paper she and her team then examine the time course of the chills response. Typically chills are a very fleeting experience – they don’t come along very often and they don’t last when they do. So you can play a whole piece of music and see what occurs in the period leading up to the passage where the chill is likely to occur (according to the participant) and also what happens afterwards. Using the latest techniques the authors are also able to track the temporal dynamics of the activity within the brains reward centre at the same time.

We know from previous studies that chills are associated with dopamine release in the mesolimbic system, our reward brain centre; namely in the nucleus accumbens (Blood & Zatorre, 1999). But 10 years on we can now use machines that are able to look at a higher level of neurochemical resolution and with better hemodynamic response characteristics at the second by second changes that occur in the brain as a musical chill approaches.

The authors of the present report do this by combining the temporal specificity of fMRI with the neurochemical specificity of PET.

The authors predicted that the regions of the brain revealing dopamine activity in the PET data (striatal reward systems) would show the largest increases in fMRI signal during chills. Chills were measured and defined by asking participants about their experience of pleasure during a piece of music, but also by measuring known physiological correlates of a chill experience such as heart rate, respiration and electrodermal response, as well as temperature and blood volume pulse amplitude.

The authors did find as they expected – during chills there was increased BOLD response in the right nucleus accumbens, reflecting an increase in dopamine in the region. An additional interesting finding, that I believe to be completely new, is that in the run up to chill epochs they found reliably increased BOLD response in the right caudate, a subregion of the striatum.

Effectively the authors had found brain evidence for a dissociation for dopamine neurotransmitter activity during the ‘anticipatory’ and the ‘consummatory’ phases of intensely pleasurable music listening.

The authors suggest these two stages might correspond to the ‘wanting’ and ‘liking’ phases we go through when presented with a typically biologically rewarding stimulus. The ‘anticipatory phase’ triggers expectation of the chill experience and creates a sense of wanting and reward prediction. Then the reward comes along in the ‘consummatory’ phase, which in music is thought to be largely abstract; we get reward from the fulfilling of our musical expectations and a sense of resolution to the musical structure.

This specific two stage pattern of dopaminergic activity has so far been linked only to biologically adaptive behaviours (food, sex) and the ingestion of substances that synthetically mimic their brain effects (drugs). The fact that music can also trigger this complex reward and motivation system indicates why music is so very valued, and probably has been for a large part of human history.

How it became so highly valued is another, interesting question. See my blogs on the evolution of music psychology for one or two ideas on this point – and of course, feel free to comment if you have any other ideas about why humans apparently came to value the rewards of music alongside other basic biological needs.