Anxious individuals shift emotion control from lateral frontal pole to dorsolateral prefrontal cortex

Anxiety disorders are prevalent and challenging to treat due to the avoidance of feared situations, preventing exposure-based learning. Emotional action selection, crucial for overcoming avoidance, involves a distributed network centered around the lateral frontopolar cortex (FPI), interacting with the amygdala, parietal, and sensorimotor cortices. FPI's role is both mechanistic and clinically significant; its interference disrupts emotional action selection, and its recruitment predicts resilience against emotional disorders. This study explores functional, structural, and neurochemical FPI properties to explain variations in emotional behavior control between highly anxious and non-anxious individuals. Contemporary models of anxiety, based on rodent studies, highlight hippocampal-amygdala pathways in avoidance, while medial frontal recurrent signals facilitate approach behaviors. However, translating these rodent findings to human anxiety disorders has been difficult, partly due to the expansion of the human granular prefrontal cortex, particularly the FPI, which lacks a direct homologue in rodents. The human FPI's unique connectivity allows access to both medial and lateral cortical circuits and direct input from the amygdala. This study hypothesizes that aberrant FPI recruitment underlies difficulties in controlling emotional action tendencies in anxious individuals.

The existing literature strongly supports the crucial role of the lateral frontopolar cortex (FPI) in regulating emotional responses and actions. Studies have shown that interfering with FPI function impairs the ability to control emotional behaviors (Volman et al., 2011), and individual differences in FPI recruitment predict resilience to developing emotional disorders (Kaldewaij et al., 2021). Rodent models have shed light on the involvement of the hippocampal-amygdala circuit in avoidance behaviors (Adhikari et al., 2010, 2011; Likhtik et al., 2014). However, translating these findings to human anxiety disorders is challenging due to significant differences in prefrontal cortex anatomy between humans and rodents (Preuss & Wise, 2022). Research suggests that the human FPI's unique connectivity allows it to integrate information from diverse brain regions and play a pivotal role in selecting actions that are incongruent with emotional impulses (Bramson et al., 2020). The current research builds upon this foundation to investigate the FPI's role in individuals with anxiety.

This study combined Magnetic Resonance Spectroscopy (MRS), Diffusion Weighted Imaging (DWI), and functional MRI (fMRI) to investigate neurochemical, structural, and functional properties of the FPI in individuals with high anxiety. 52 participants with high anxiety (Liebowitz Social Anxiety Scale (LSAS) score > 30) and 41 non-anxious male participants (matched for age) were recruited. Participants performed a social approach-avoidance task (AA task) involving approaching or avoiding happy and angry faces using a joystick. This task requires overriding automatic emotional tendencies. fMRI data was acquired during task performance. MRS was used to measure GABA/glutamate ratio (as an index of neuronal excitability) in the right FPI, left sensorimotor cortex (SMC), and left occipital cortex (control). DWI was employed to measure the strength of amygdala projections to the FPI via the amygdalo-fugal bundle. Bayesian mixed-effects models were used to analyze behavioral data, relating error rates to neural excitability and connectivity. fMRI data analysis used GLM approach to identify brain activation patterns during emotional action control. Whole-brain group comparisons and correlations between neural measures and behavioral performance were also conducted.

The study yielded three main findings: 1. During emotional action control, high-anxiety participants relied on the dorsolateral prefrontal cortex (dIPFC) rather than the FPI, unlike their non-anxious counterparts. Non-anxious participants showed increased FPI activation during incongruent (emotionally challenging) trials, while no such effect was observed in the high-anxiety group. 2. High-anxiety participants exhibited a more excitable FPI (lower GABA/glutamate ratio) and stronger amygdalofugal projections to the FPI compared to the non-anxious group. This difference was specific to FPI, not observed in SMC or occipital cortex. In the non-anxious group, a more excitable FPI correlated with better emotional control, but this relationship was reversed in the high-anxiety group. 3. The strength of amygdalofugal projections to the FPI predicted the degree of the FPI-to-dIPFC shift during emotional control in anxious individuals. Non-anxious participants with stronger amygdalofugal projections exhibited better emotional control; there was no such relationship in the high-anxiety group. Furthermore, stronger amygdala connections to the FPI correlated with stronger dIPFC activation during incongruent trials only in high-anxiety participants, potentially indicating compensatory recruitment of dIPFC due to FPI dysfunction.

These findings indicate that anxiety is associated with inefficient FPI engagement during emotional control, shifting the control mechanism from FPI to dIPFC. The increased FPI excitability in high-anxiety individuals, coupled with stronger amygdala input, may saturate FPI circuitry, limiting its ability to fine-tune responses. This aligns with the clinical features of anxiety: hypersensitivity to threat and overgeneralization. The shift to dIPFC, while maintaining behavioral performance in this mild emotional challenge, might create vulnerabilities when facing stronger challenges. FPI's ability to integrate affective and contextual information for flexible emotion control contrasts with dIPFC's simpler role in maintaining task rules. These findings suggest potential therapeutic interventions targeting FPI activity, such as modulating FPI inhibition or dampening dIPFC overactivation.

This study reveals that anxiety is linked to inefficient FPI use during emotional control, with a functional shift to dIPFC. This shift is associated with increased FPI excitability and stronger amygdala-FPI connections. This suggests interventions to normalize FPI activity for improved emotional regulation in anxiety disorders. Future research could explore the subjective experience of anxiety in relation to the observed structure-function relationships and examine the effects of varying emotional challenge intensity on this neural shift.

The study included only male participants in the non-anxious group, limiting the generalizability of the findings to females. While the study maximized signal-to-noise ratio, future large-scale studies are needed to confirm the observed effects. High-field MRS could enhance the precision of FPI excitability measures, which is crucial for tailoring individualized treatment interventions.