Parasympathetic Nervous System Functioning Moderates the Associations between Callous-Unemotional Traits and Emotion Understanding Difficulties in Late Childhood

Abstract

Background: Callous-unemotional (CU) traits are characterized by low empathy, guilt, and prosociality, putting children at risk for lifespan antisocial behavior. Elevated CU traits have been linked separately to difficulties with emotion understanding (i.e., identifying emotional states of others) and disrupted parasympathetic nervous system (PNS) functioning. However, no study has investigated how PNS functioning and emotion understanding are jointly related to CU traits.

Method: We explored associations between CU traits, emotion understanding, and PNS functioning (indexed via respiratory sinus arrhythmia [RSA]) among children aged 7-10 years old (n = 55). We also tested whether deficits in emotion understanding differ across specific emotions (i.e., fear, pain, happiness, anger). Each child's RSA was continuously recorded while they watched a film that included emotionally evocative social interactions. To assess emotion understanding, children identified emotions replayed in 1s animations of scenes from the film. Parents reported on child CU traits, conduct problems, and demographic information.

Results: Higher CU traits were related to lower emotion understanding (β = -0.43, p = 0.03). PNS activity during the film moderated this association (β = -0.47, p < 0.001), such that CU traits were associated with lower emotion understanding among children with mean (B = -0.01, t = -2.46, p = 0.02) or high (i.e., 1 SD > M; B = -0.02, t = -3.00, p < 0.001) RSA levels during the film, but not among children with low RSA levels (i.e., 1 SD < M; B = 0.00, t = -0.53, p = 0.60). Moreover, we found that the observed moderated associations are driven by deficits in fear, specifically.

Conclusions: The link between poorer emotion understanding, fear understanding in particular, and CU traits was attenuated for children who demonstrated patterns of PNS functioning consistent with attentional engagement while viewing the emotion stimuli.

Keywords: callous unemotional traits; emotion understanding; parasympathetic nervous system; physiological regulation.

Figure 1

An overview of the Pixar task to assess emotion understanding, including pictorial response scale and stills from GIFs. (A) Example presentation of the task screen depicting a still from the GIF of the bird displaying “fear”. Cartoon images of prototypical facial expressions are presented below the film GIFs during the emotion understanding component of the task. GIFs presented in 1 s loops at 38 frames per second. (B) Examples of stills depicting angry, happy, and pained emotions during the task. Two different examples of each emotion were presented in separate trials throughout the task (n = 10 trials). Images ©2009 Pixar.

Figure 2

Higher levels of CU traits are related to lower emotion understanding accuracy at mean and high levels of RSA during the animated film, but not at low levels of RSA. Note: * p < 0.05, ** p < 0.01, n.s. = Not Significant. (A) Simple slopes plot showing that higher CU traits were associated with lower emotion understanding accuracy for children showing mean (B = −0.01, t = −2.46, p = 0.02) or high levels (i.e., 1 SD > M; B = −0.02, t = −3.00, p < 0.001) of RSA during the animated film, but not for children with low levels of RSA (i.e., 1 SD < M; B = 0.00, t = −0.53, p = 0.60). (B) Johnson-Neyman plot showing the range of significant slope values for the prediction of emotion understanding accuracy using CU traits for children just below mean task RSA and above (upper bound threshold of −0.10, range of observed, centered values were −1.68 to 2.05). The lower bound RoS for RSA was outside the data range and was not interpreted.

Figure 3

Higher levels of CU traits are related to lower fear accuracy at mean and high levels of RSA during the animated film, but not at low levels of RSA. Note: * p < 0.05, ** p < 0.01, n.s. = Not Significant. (A) Simple slopes plot showing that higher CU traits were associated with lower emotion understanding accuracy for children showing high levels (i.e., 1 SD > M; B = −0.03, t = −2.23, p = 0.03) of RSA during the animated film, but not for children with mean (B = 0.01, t = −0.99, p = 0.33) or low levels (i.e., 1 SD < M; B = −0.01, t = −1.24, p = 0.22) of RSA. (B) Johnson-Neyman plot showing the range of significant slope values for the prediction of emotion understanding accuracy using CU traits for children just below mean task RSA and above (upper bound threshold of −0.10, range of observed, centered values were −1.68 to 2.05). The lower bound RoS for RSA was outside the data range and was not interpreted.