Neurobiologically, panic attacks in humans hold similarities to what in animals can
be elicited by fear conditioning: A fear network, centred in the amygdala and involving
interactions with the medial prefrontal cortex (mPFC) is assumed to mediate emotional,
endocrine, motor and autonomous reactions during acute fear. This study focusses on
the hypothesis that an altered mPFC/amygdalar interaction represents a neural substrate
of panic disorder (PD). We probe this by applying an emotional conflict paradigm during
fMRI to 20 patients diagnosed with PD. Congruent and incongruent image/word pairs
are presented as mixed even-related/block design. Analysis involves dissociating a
conflict monitoring phase from a resolution phase by accounting for the order of the
individual trials. The design allows for analysing general deactivation of default
mode activity during cognitive demanding task, conflict resolution versus monitoring,
and mPFC/amygdala interactions. Drug-naïve patients are followed up after initiation
of citalopram medication. 20 control subjects are scanned repeatedly to control for
habituation. So far, controls (N=10) but not PD patients (N=10) showed faster reaction
times in high versus low conflict resolution trials. Second level random effect fMRI
analysis revealed more activation in the rostral ACC/mPFC during high conflict resolution
trials in controls than PD patients (p<0.05). Dynamic causal modelling is applied
to investigate the mPFC/amygdala interpla
