Multimodal Neuroimaging Approaches to Understanding Suicide-Related Thoughts, Behaviours and Clinical Risk Factors in Treatment Resistant Depression

Abstract

Despite extensive research contributions, suicide risk assessment still relies heavily on subjective and clinician judgement. In recent years, the search for suicide biomarkers has gained significant momentum, and an expansive body of literature has provided diverse and convincing evidence of biological involvement in suicide. This thesis is presented as a collection of articles with three primary aims. First, to use structural and resting-state functional magnetic resonance imaging (MRI) techniques to study gray matter volume, cortical thickness, and functional connectivity correlates of suicidal ideation (SI) severity in treatment-resistant depression (TRD). Second, to use resting-state functional MRI to examine functional connectivity differences among healthy controls and participants with TRD and a lifetime history of SI only (SI group), and those with a lifetime history of both SI and suicide attempt (SA group). Additionally, associations between functional connectivity alterations and clinical features that may differentiate SI and SA groups were examined, including hopelessness, anxiety, depression, aggression, impulsivity, and childhood trauma. Third, to use proton magnetic resonance spectroscopy (1H-MRS) to compare levels of glutamate (Glu), Glx (combined glutamine/glutamate), myo-inositol (mI), N-aceylaspartate (NAA), and creatine (Cr) in the dorsal anterior cingulate cortex (ACC) in participants with TRD and a lifetime history of SI to a healthy control group. Associations between neurometabolite concentrations and suicide-related symptoms and traits, including depression, suicidal ideation severity, impulsivity, and aggression were also investigated. Participants underwent a single MRI session, and structural, resting-state functional, and 1H-MRS data were collected. Overall, findings included: 1) increased SI severity correlated with reduced right lateral occipital cortical thickness, and widespread functional connectivity correlates centered on the right posterior parietal cortex (Manuscript 1), 2) functional connectivity differences in fronto-temporo-limbic regions differentiating participants with SA history from those with SI only, with connectivity markers meaningfully related to clinical features in the SA group (Manuscript 2), and 3) glutamatergic alterations in the dorsal ACC in participants with TRD and SI associated with higher trait impulsivity (Manuscript 3). Collectively, this thesis provides evidence of neural abnormalities associated with suicide risk in TRD, thereby advancing the search for clinically meaningful suicide biomarkers which may inform identification of treatment targets.