Abstract
Background: Although anxiety is a natural adaptive response part of our evolutionary defense
system, it can become pathological if it becomes disproportionately exaggerated, functionally
disruptive, or chronic and severe. Anxiety disorders are the most prevalent group of psychiatric
disorders in New Zealand - affecting more than 1 in 10 New Zealanders (almost 1 in 5 Māori)
annually, and accounting for about 1 in 5 patients seen in primary care. Anxiety disorders are
also very costly, both as a burden on the healthcare system, with anxiety patients frequently
accessing primary care resources in Western countries, and by generating additional costs to
society as work days lost due to illness. Currently, we lack a proper understanding of the
mechanisms underlying anxiety, which poses a major challenge in its diagnosis and
management. The diagnosis of anxiety disorders relies on clinical criteria using a superficial
list of signs and symptoms, and lacks a more objective, cause-based means of assessment.
Aims: This study held two objectives: 1) to investigate the effects of acute citalopram
administration on the resting-state functional connectivity of the amygdala (a deep brain grey
matter structure heavily implicated in the neural circuitry of anxiety and fear); 2) to explore the
resting-state functional connectivity of the amygdala across all study participants.
Methods: This study was conducted using resting-state fMRI (RS-fMRI) data collected from
a recent clinical trial set up to help develop a novel anxiety-specific biomarker. The phase of
this wider clinical trial from which this study sourced its data studied the acute effects of
commonly prescribed medications (e.g., citalopram) on deep brain structures implicated in the
neural circuitry of anxiety and fear (e.g., the amygdala) in anxious, phobic, and healthy
participants. The citalopram group recruited 15 patients with DSM-V diagnosed generalized
anxiety disorder (GAD) or social anxiety disorder (SAD) alongside 15 matched healthy
volunteers. RS-fMRI data was collected about 45 minutes after intervention administration
(20 mg citalopram or placebo pill). Data analysis was performed via seed-based correlation
analysis (SCA) using the left and right amygdala as the seed regions. First-level (single
subject) analysis extracted blood oxygen dependent (BOLD) signal data from the left and
right amygdala, middle-level analysis combined the contents of first-level analysis for both
the left and right amygdala, and finally higher-level (group) analysis compared the two study
groups and explored the amygdala’s functional connectivity at rest.
Results: Two participants from the citalopram group were excluded from data analysis as one
received the incorrect drug (fluoxetine instead of citalopram), and another contained
excessive motion artefact in their RS-fMRI scan. Higher-level analysis tested three contrasts
for each of the four lower-level contrasts (group mean, left > right, left amygdala, and right
amygdala) across all study participants. The first two higher-level contrasts (C1 = C - P, C2 =
P - C) aimed to investigate any BOLD signal differences between the citalopram (C) and
placebo (P) groups. The third contrast (C3 = total mean) combined the timeseries data of both
study groups to explore RS-FC of the amygdala after acute citalopram administration. C1 and
C2 returned with completely unremarkable seed-based connectivity maps for all four lower
level contrasts (i.e., not a single voxel cleared the Z-score threshold). This implies there were
no voxels with a statistically significant (Z 3.1) difference in BOLD signal timeseries data
between the citalopram and placebo groups. Fortunately, higher-level C3 (total mean
activation) returned with widespread voxel activation for most lower-level contrasts (except
for C2, which assessed BOLD signal differences between the left and right amygdala). The
most significant cluster identified consisted of the bilateral amygdala, hippocampus, and
parahippocampal gyrus. The remaining activated voxels were scattered throughout the pons,
basal ganglia, and cerebral cortex (mainly within the frontal and temporal lobes, with
significant involvement of the anterior cingulate and paracingulate gyrus). Furthermore, there
was a statistically significant difference in neuronal activity (as indirectly measured via
BOLD signal data) between the left and right amygdala across all subjects.
Conclusions: This thesis examined the effects of acute citalopram administration on the
resting-state functional connectivity of the amygdala. However, it demonstrated no
statistically significant difference across all four lower-level contrasts tested between
participants from the citalopram group and those from the placebo group. The diagnosis and
management of anxiety disorders continues to represent a major challenge for clinicians
across the globe, with the all-encompassing solution being further research to gain a complete
understanding of the mechanisms underlying anxiety.