Children and adolescents’ neural response to emotional faces and voices: Age-related changes in common regions of activation

ABSTRACT

The perception of facial and vocal emotional expressions engages overlapping regions of the brain. However, at a behavioral level, the ability to recognize the intended emotion in both types of nonverbal cues follows a divergent developmental trajectory throughout childhood and adolescence. The current study a) identified regions of common neural activation to facial and vocal stimuli in 8- to 19-year-old typically-developing adolescents, and b) examined age-related changes in blood-oxygen-level dependent (BOLD) response within these areas. Both modalities elicited activation in an overlapping network of subcortical regions (insula, thalamus, dorsal striatum), visual-motor association areas, prefrontal regions (inferior frontal cortex, dorsomedial prefrontal cortex), and the right superior temporal gyrus. Within these regions, increased age was associated with greater frontal activation to voices, but not faces. Results suggest that processing facial and vocal stimuli elicits activation in common areas of the brain in adolescents, but that age-related changes in response within these regions may vary by modality.

KEYWORDS:

• Social cognition
• emotion
• neural response
• development
• adolescence

Acknowledgments

We are thankful to Brooke Fuller, Meika Travis, Roberto French, Connor Grannis, and Andy Hung for their help with this project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ To verify whether different patterns of age-related change in accuracy were evident before and after the putative peak around age 14, we conducted an exploratory piece-wise regression with free-knot splines (with one knot set at age 14). For facial ER accuracy, the spline model represented an improvement in fit above the linear regression described above, F(1, 18) = 6.11, p =.02. The model indicated a rise in P_r, B = 0.03, t(1, 17) = 3.70, p =.002, before the age of 14, followed by a decrease in P_r, B = −0.04, t(1, 17) = −2.47, p =.02, after age 14. This pattern suggests that improvements in facial ER across childhood and adolescence may be most notable before 14 years of age. For vocal ER accuracy, the spline regression did not represent a significant improvement over the linear model, F(1, 19) = 3.75, p =.07 – supporting the finding that vocal ER may follow a linear increase across childhood and adolescence. See Supplementary Materials for graphical depictions of the spline models.

² To investigate whether there were quadratic effects of age on neural activation to faces (as in the ER performance data), an exploratory model was conducted with an additional quadratic term for age. No clusters of significance were noted at p <.001, nor at a less conservative threshold of p =.05.

Additional information

Funding

This work was supported by internal funds at Nationwide Children’s Hospital and the Fonds de recherche du Québec – Nature et technologies [grant number 207776].