Deficient prefrontal-amygdalar connectivity underlies inefficient face processing in adolescent major depressive disorder

Major Depressive Disorder (MDD) is a debilitating mental illness with a significant increase in prevalence during adolescence, a critical period of brain development. A core feature of adolescent MDD is aberrant emotion processing, often associated with functional alterations in the prefrontal-amygdala circuitry. Existing research has demonstrated amygdala reactivity and aberrant activity in the sgACC and lateral prefrontal cortex (LPFC) in adolescents with MDD during emotional face processing. Furthermore, studies suggest disrupted functional coupling within the prefrontal-amygdala network in this population. This study aimed to investigate the functional architecture of this circuitry using a dynamic emotional face-matching task and dynamic causal modeling (DCM) to analyze effective connectivity. The researchers hypothesized that adolescents with MDD would exhibit aberrant emotion processing, reflected in differences in evidence accumulation, and that this would be related to altered connectivity within the prefrontal-amygdala network. Specifically, they expected disrupted connectivity between the sgACC, LPFC, and fusiform face area (FFA), and reduced top-down influence of prefrontal regions on the amygdala.

A substantial body of literature links the cognitive bias observed in MDD to dysregulation of the prefrontal-amygdala network. Adolescence is a particularly vulnerable period due to the stress-sensitive developmental trajectory of this network. Previous studies have consistently shown amygdala hyperactivity in adolescents with or at risk for MDD during emotional face processing. Furthermore, aberrant activity in the sgACC and LPFC has also been implicated. Studies have also indicated disrupted functional coupling within the prefrontal-amygdala network, with specific alterations noted between the sgACC and amygdala, and between the sgACC and FFA. Longitudinal studies further suggest that prefrontal-amygdala interactions not only characterize adolescent depression but also predict treatment response.

Thirty adolescents with MDD and 33 age-, IQ-, sex-, and handedness-matched healthy controls participated. Participants completed a dynamic face- and shape-matching task, considered more ecologically valid than static tasks. The task involved matching probes to dynamic stimuli that gradually transitioned from a neutral expression to a target emotion (positive, negative, or neutral). A linear ballistic accumulator (LBA) model was fitted to the behavioral data to quantify evidence accumulation. Functional magnetic resonance imaging (fMRI) was used to assess brain activity during the task. Dynamic causal modeling (DCM) was then employed to analyze effective connectivity within the prefrontal-amygdala network, focusing on the FFA, LPFC, sgACC, and amygdala. Behavioral data were analyzed using linear mixed-effects models, and fMRI data were preprocessed and analyzed using SPM12. Regions of interest for DCM were identified based on previous studies and the LBA model results.

Behavioral results showed no significant group differences in response time or accuracy, but MDD adolescents exhibited slower evidence accumulation for neutral faces in the LBA model, indicating reduced processing efficiency. Arousal ratings were significantly higher for negative faces in the MDD group, and valence ratings were lower for positive faces. fMRI analysis revealed that slower evidence accumulation was associated with stronger deactivation in the sgACC during neutral face processing. DCM analysis indicated altered functional coupling in the FFA-LPFC-sgACC pathway and reduced connectivity between the sgACC and amygdala in the MDD group. These alterations were consistent across all facial emotions. Further analysis showed that SSRI use was associated with decreased sgACC self-connectivity, suggesting increased input sensitivity. Finally, valence-dependent modulations of prefrontal-amygdala connectivity were observed in both groups, with differences in amygdala-FFA, amygdala-LPFC, and amygdala-sgACC connectivity.

The findings suggest that inefficient face processing in adolescent MDD is associated with impaired connectivity within a distributed network involving the FFA, LPFC, sgACC, and amygdala. The sgACC appears to play a crucial role as a gatekeeper, modulating the interaction between prefrontal and limbic systems. Disruption of sgACC engagement during ambiguous social situations may lead to maladaptive emotional responses. The observed valence-dependent connectivity modulations highlight the complex interplay between emotion and cognition. The association of SSRI use with increased sgACC input sensitivity suggests a potential neural mechanism underlying the therapeutic effects of these medications. The study's methodology, combining computational modeling and neuroimaging, provides a detailed understanding of the cognitive and neural mechanisms involved.

This study demonstrated diminished cognitive efficiency and altered functional brain circuits supporting emotion processing in adolescents with MDD. The impaired connectivity within the prefrontal-amygdalar network, particularly involving the sgACC, may contribute to difficulties in processing nuanced facial expressions and lead to maladaptive emotional responses. Future research should investigate the longitudinal aspects of these findings and explore the specific roles of various emotional cues in shaping these network interactions. These findings may inform the development of targeted interventions, including psychotherapy, pharmacotherapy, and potentially neurofeedback approaches.

The relatively small sample size and the cross-sectional design limit the generalizability of the findings. The high accuracy rates for positive and negative faces in the task may have reduced the sensitivity of the LBA model. Further research incorporating a larger sample, longitudinal design, and more nuanced emotional stimuli is needed. The majority of patients were taking SSRIs; therefore, the findings may not be generalizable to unmedicated individuals.