Abstract
Emotions play a pivotal role in guiding our behaviour within society and our environment. In particular, emotions enable interpersonal social interactions through non-verbal communication that may be below conscious awareness. However, when there is some disruption to normal emotional processing, such as in anxiety disorders, quality of life of the individual can be severely disrupted. Anxiety disorders account for nearly a quarter of all mental health diagnoses, however the aetiology and underpinning neural correlates of anxiety are still not fully understood. This thesis sought to investigate the neurobiological mechanisms of emotion processing, specifically in the amygdala, in a healthy sub-clinical cohort. Six different studies are presented using quantitative methodology to explore amygdala activation and connectivity during emotion processing, and structural differences, as modulated by gender and sub-clinical anxiety. Overall results reveal a modulating effect of sub-clinical anxiety on amygdala habituation, fronto-amygdala connectivity (at rest and during emotion processing) and neural structure. In addition, results presented in this thesis suggest that there may be an attentional component to characteristic hyper-responsivity of the amygdala during emotion processing seen in clinical anxiety patients that should be incorporated into future models of maladaptive emotion. Furthermore, various different chapters in this thesis present evidence that the left amygdala appears to be more specialised for responses to more socially salient stimuli and the right amygdala appears to be more responsive to threat related stimuli indicating that key theoretical models of emotion (the dual processing model, and the salience detection model) should be integrated into one cohesive model of emotion processing. In addition to these theoretical implications, results demonstrating the modulating effect of anxiety and gender presented in this thesis suggest that research on emotion should account for individual differences as a matter of standard practice. This thesis also supports the use of resting state -functional magnetic resonance imaging (fMRI) as a low cost, valid alternative, to task based fMRI within the study of anxiety. Finally, results suggest that investigation of structural differences in sub-clinical populations, and the use of analytical methods such machine learning classification techniques, could aid the development of diagnostic tools that can track disease progression and identify individuals at risk of developing anxiety disorders. The possibility of identifying such neural biomarkers will allow research to look for therapeutic treatments and interventions, which could prevent individuals from transitioning from sub-clinical anxiety to chronic anxiety disorders.