Individual Submission Summary
Share...

Direct link:

Poster #63 - Multimodal Examination of Amygdala-prefrontal White Matter Microstructure and Amygdala Reactivity in a Large Adolescent Sample.

Fri, March 22, 7:45 to 9:15am, Baltimore Convention Center, Floor: Level 1, Exhibit Hall B

Integrative Statement

Introduction: The amygdala is fundamentally involved in processing emotion and altered amygdala function is implicated in a range of emotion-based traits and conditions, including inhibited temperament, depression, and anxiety. Through bidirectional connections, the prefrontal cortex (PFC) is hypothesized to influence amygdala reactivity. However, research that elucidates the nature of amygdala-PFC interactions – through mapping amygdala-PFC tracts, quantifying variability among tracts, and linking this variability to amygdala activation – is lacking. The present study used probabilistic tractography to quantify the maximum-likelihood of amygdala white matter (WM) connectivity with seven PFC regions that share structural connections with the amygdala based on tract tracer studies in non-human primates: Brodmann’s Area (BA)9, BA10, BA11, BA24, BA25, BA32, BA47. We examined the relation between amygdala WM connectivity with the PFC and amygdala activation from an emotional faces functional MRI task.

Hypotheses: When compared across the PFC regions, areas where neural connections with the amygdala have been stronger in non-human primates (BA25, BA47, BA10, BA11) would have a higher maximum-likelihood of amygdala connectivity. Increased amygdala-PFC WM connectivity would correspond to attenuated amygdala activation and a subset of PFC regions would drive that association.

Study Population: 142 adolescents (age 15-17) from the Detroit, Toledo, and Chicago subsamples of the Fragile Families and Child Wellbeing Study (FFCWS).

Methods: Diffusion MRI data was analyzed using probabilistic tractography to estimate the probability of WM connectivity between the PFC and the amygdala. Group peaks in connectivity were identified for each amygdala-PFC pair and the mean value was then extracted at the individual level. T-tests were done to determine which regions had the highest likelihood of amygdala connectivity. Functional activation was extracted for the left and right amygdala. Multiple linear regression was done using a machine learning ridge regression model in Python’s sci-kit-learn package. In this model, the maximum-likelihood of WM connectivity between the ipsilateral amygdala and PFC regions predicted ipsilateral amygdala activation to threat faces. Significance testing of this model was done using random permutation sampling.

Results: BA25 (subgenual cingulate) and BA47 (orbitofrontal cortex (OFC)) had the highest likelihood of amygdala WM connectivity followed by BA11 (OFC) and BA10 (dorsal medial PFC) (Fig.1). Maximum-likelihood of amygdala WM connectivity with the PFC regions significantly predicted amygdala activation in both hemispheres (Fig.2). In the right hemisphere, mean squared error (MSE) for the model was 0.33 (p<0.01) compared to the mean MSE from the randomized permutations of 0.39. Amygdala connectivity with BA47 and BA10 were the strongest predictors. In the left hemisphere, MSE for the model was 0.29 (p<0.01) compared to the mean MSE from the randomized permutations of 0.33. Amygdala connectivity with BA10 and BA11 were the strongest predictors.

Conclusion: Results demonstrate widespread, but variable, amygdala WM connectivity in the human PFC. In addition, this research provides novel evidence in humans that emotion processing may be facilitated by amygdala-PFC WM connectivity with regions traditionally responsible for sensory integration – OFC, dmPFC. By examining the association between specific amygdala-PFC tracts and amygdala activation, we elucidate the nature of this emotion-based circuit.

Authors