What can movies tell us about the anxious brain? In our recent papers, we explore how anxiety shapes with brain and cardiac dynamics during movie-watching.

A well characterized amygdala-prefrontal circuit is thought to be crucial for threat vigilance during anxiety. However, studies to date have primarily investigated this using static stimuli (i.e. faces) presented without context. Consequently, the relationship between this circuit and anxiety in relatively more dynamic, naturalistic contexts remains poorly understood. Extending study of this circuitry to more naturalistic stimuli offers the opportunity to validate these findings in more ecologically-rich settings and observe how this circuit may be modulated as a function of dynamic, contextual features. We have now investigated this using naturalistic ‘movie fMRI’ data from the Naturalistic Neuroimaging Database (N=86), Human Connectome Project (N=178), and CamCAN (N=630) datasets. Firstly, we report a relationship between trait anxiety and stimulus-dependent activations in the Naturalistic Neuroimaging Database. Specifically, self-reported anxiety was associated with decreased activation to words in the primary auditory cortex and increased activation to faces in the superior parietal lobe. 

Static connectivity analyses (which treat brain connectivity as time-invariant) of the Naturalistic Neuroimaging Database and Human Connectome Project failed to demonstrate any relationship between amygdala-prefrontal connectivity and anxiety during movie-watching. However, dynamic analyses (which treat brain connectivity as relatively time-varying) of the CamCAN database revealed trait anxiety to have a variable association with amygdala-prefrontal connectivity dependent on the degree of suspense during the movie. Specifically, we observed effects in the inverse direction to what we predicted: anxiety was associated with increases in right amygdala-prefrontal connectivity primarily during low suspense scenes (r = -.14, p = .04; figure 2). Additionally, a planned exploratory analysis suggested a measure of threat-relevant attentional bias was also sensitive to these effects (r = -.24, p = .001). We offer three potential interpretations of these findings: 1) high anxiety individuals may chronically engage threat circuitry irrespective of suspenseful scenes in a movie; 2) high anxiety individuals anticipate suspenseful scenes quicker, engaging threat circuitry at earlier time points; and/or 3) high anxiety individuals are slower to disengage threat circuitry following suspenseful scenes.

There is also evidence to suggest anxiety impacts the degree of communication between the brain and the autonomic nervous system. Using a suspenseful movie-watching paradigm (n=29-), we demonstrated that anxiety was associated with reduced coherence between heart rate and: amygdala-dorsomedial prefrontal dynamic connectivity; amygdala-subgenual anterior cingulate dynamic connectivity; precuneus activity; vmPFC activity; and bilateral putamen activity. Taken together, we provide preliminary evidence for altered neural-autonomic coherence as a function of anxiogenic movie-watching.