Interoception concerns the afferent vagal and spinothalamic lamina I systems, and their projection to regions of the brain comprising the central autonomic network (CAN). At the level of the cortex, the CAN includes regions such as the insula, medial prefrontal and cingulate
cortices, which interact with subcortical and brainstem networks to regulate autonomic, neuroendocrine, immune, and other visceral functions of the body. Interoception is an important
concept linking ‘primitive’ homeostatic functions of the brain to its ‘high-order’ cognitive functions. This view is supported by an increasing body of experimental evidence indicating the
relevance of interoceptive neural systems to motivational drives, mood, emotion, self-awareness, body-ownership, somatic disorders and psychopathology. However, constructs, paradigms and
other methodology for investigating neural interoception in humans require additional development and validation. Additionally, neural interoceptive processing in psychopathology has not been thoroughly characterized, hence limiting the translational relevance of findings
from this field. Given the emerging role of interoception in many psychological functions, a key question would be whether we could access and modulate neural interoceptive systems in
humans. Hence, the first aim of this thesis was to investigate whether interoceptive neural processing can be modulated through non-invasive stimulation of the cortex or through peripheral nerve stimulation. To accomplish these aims, transcranial magnetic stimulation (TMS) and transcutaneous vagus nerve stimulation (tVNS) were used to modulate heart-brain interactions. A second aim of this thesis was to determine whether traumatic stress exposure in
female psychiatric patients alters the degree to which neural interoceptive systems are engaged when asked to attend to somatic and visceral feelings during mindful breathing.
**STUDY 1** is a randomized, sham-controlled investigation to determine whether tVNS affects cardiovagal responses and neurocardiac integration in interoceptive cortices. The ability of tVNS to evoke cardiovagal responses was mixed. tVNS was found to increase baroreceptor sensitivity, but not heart rate variability, whereas both sham and tVNS elicited reductions in heart rate. At the level of the brain, tVNS increased electroencephalographic (EEG) functional
connectivity between regions of the CAN. In particular, stronger functional connectivity was obtained for the right somatosensory and anterior insula in the beta frequency band. The effect of
tVNS on an evoked potential reflecting neural cardiac interoceptive processing (the heart-evoked potential or ‘HEP’) was also assessed. At the sensor-levels, tVNS was associated with greater HEP negativity in left-lateralized frontal, temporal, parietal and central electrodes. Source localized functional connectivity between regions where HEPs have been observed intracranially revealed patterns of greater and lesser connectivity in several frequency bands. Insula-prefrontal
connectivity features correlated with heart rate during tVNS. Altogether, the results indicate that tVNS modulates neural systems relevant to cardiac interoceptive processing, which may be
relevant to the mechanisms of action by which tVNS improves cardiovascular autonomic function in somatic and psychiatric conditions.
**STUDY 2** applied transcranial magnetic stimulation to the right frontotemporal cortex to test whether modulating cortical excitability within regions putatively accessing the CAN alters cardiovascular autonomic responses. Intermittent theta-burst stimulation increased vagally mediated heart rate variability, but this effect appears to have been confounded by stimulation induced state anxiety. However, continuous theta-burst stimulation increased pulse-transit time
latency, an effect that was not explained by stimulation-induced anxiety. This study supports the use of TMS for modulating ‘top-down’ neurocardiac integration, and discusses approaches for optimizing TMS for investigating neural interoceptive and visceromotor processing, and its translational relevance.
**STUDY 3** investigated the functional MRI correlates of respiratory interoception in women in residential treatment for stimulant dependence (SUD) who have varying histories of physical,
psychological and/or sexual trauma. A subset of patients had a concurrent diagnosis of posttraumatic stress disorder (PTSD). Reduced functional connectivity of an interoception-linked
network was found in women with SUD-PTSD comorbidity. Specifically, an orbitofrontal network showed diminished strength of correlation with the insular, somatosensory and cognitive control regions during a mindfulness-based breathing task. Additionally, orbitofrontal network
strength was negatively associated with sexual violence exposure beyond the contribution of PTSD diagnosis alone. This study contributes to scientific understanding concerning interoceptive dysfunction in psychopathology and potential mechanisms through which psychobehavioral techniques such as mindfulness may improve mental health.
OVERALL, these results of this dissertation support the utility of non-invasive cortical or peripheral nerve stimulation in accessing and modulating neural interoceptive systems related to cardiovascular autonomic regulation. The results also support the utility of using certain psychobehavioral techniques, such as mindfulness, to engage interoceptive brain systems, and they highlight how different psychopathological conditions may respond differently to treatment
modalities involving interoceptive manipulations. Altogether, this work enhances basic understanding of brain-body interactions, and advances the translational value that can be derived from interoceptive theoretical frameworks.