As youth mature, they become better able to identify other people’s emotional states based on their facial and vocal expressions (“emotion recognition”, or ER). How does this happen? This is still an open question! Amongst other changes to cognition and social interactions, changes to brain functioning during childhood and adolescence could play a role in the growth of emotion recognition skills. Although previous research has investigated how brain activation to facial expressions of emotion changes with age and across time, this study examined change in neural response to vocal emotional expressions.
We recruited children and teenagers aged 8 to 19 to participate in the study. While they were in an MRI scanner, we asked them to complete a vocal emotion recognition task: they were presented with recordings of other people saying things in different tones of voice, and had to select what emotion the speaker was expressing. Participants then came back one year later to do the same task again. We looked at how participants’ brains activated when they were hearing the emotional voices, compared to the neutral voices that we used as a baseline.
We asked three main questions:
Did participants’ brain response to the emotional voices depend on their age?
Did brain response change across 1 year’s time (i.e., from visit 1 to visit 2)?
And, did the rate of change in brain response across visits depend on participants’ age?
We found that certain areas of the brain did respond differently in younger vs. older adolescents (Question 1, above): notably, there was an area of the prefrontal cortex that responded more strongly to emotional voices (than to neutral voices) in the older participants than the younger ones. This was a bit of a surprising finding, because previous research that asked similar questions about how the brain responds to emotional faces found that activation in this area of the brain decreased with age! We’re not sure why this result is different, but this may be because voices elicit different patterns of neural response than do faces.
We also found that regions like the dorsal striatum and inferior frontal gyrus—which have previously been involved in coding the value of certain outcomes, and in ‘mentalizing’ tasks where people are asked to infer others’ emotions or mental states—showed less activation at the second visit than at the first (Question 2). Although we have to speculate about what this pattern means, it could be reflecting increased efficiency (i.e., needing less response in these areas) with repeated practice. Lastly, and perhaps most interestingly, we found that activation patterns in the right temporo-parietal junction (TPJ)—an area thought to be crucial to understanding others’ emotions and intentions—varied as a function of both age and time (Question 3). For younger teens, we saw increases in TPJ response between visits. But, for older teens, we found the opposite pattern. This inverted U-shaped pattern is consistent with the predictions of a prominent theory of neurodevelopment called the “interactive specialization” model (Johnson et al., 2000). According to this theory, early development is characterized by broad and diffuse patterns of activation (meaning that a given task would require lots of activation in many parts of the brain); with maturation, however, less activation is required, and less parts of the brain need to participate.
What do these patterns of activation mean in the real-world? Some of these patterns were linked to performance on the task. Specifically, people who showed decreases in dorsal striatum and TPJ activation across timepoints did best on the vocal emotion recognition task. Therefore, our findings suggest that some of these changes in neural activation to emotional voices may be part of the neural mechanisms that are supporting increases in vocal ER skills across childhood and adolescence. Our findings help shed some light on how changes to our brain’s function may be helping us better navigate our social world during adolescence.