Inflammation and oxiditative stress in early phase psychosis – a multimodal neuroimaging and machine learning approach

Murray, Alex (2024). Inflammation and oxiditative stress in early phase psychosis – a multimodal neuroimaging and machine learning approach. University of Birmingham. Ph.D.

Murray2024PhD.pdf
Text - Accepted Version
Available under License Creative Commons Attribution.
Download (4MB)

Abstract

Inflammation and oxidative stress are two central interlinked processes that are thought to play pivotal roles in the pathophysiology of schizophrenia spectrum disorders. This thesis presents novel findings highlighting the roles of these processes in disorder progression and presenting novel methodological advancements in glutathione quantification and patient stratification. After general introduction in Chapters 1 and 2, Chapter 3 follows presenting a systematic review and meta-analysis of 1H-MRS studies across all psychosis phases identifying glutathione reductions specific to the later stages of the disorder. While MRS acquisition methodologies were not found to significantly affect results, year of study was a confounding factor, perhaps indicating the effect of uncontrolled methodological factors and improved reporting standards. In Chapter 4, work classifying distinct immune-related clusters based on four peripheral inflammatory biomarkers (CRP, IL-6, TNF-a and INF-y) is presented. Employing machine learning clustering models on cytokine profiles, four distinct schizophrenia-specific clusters were identified, each with unique patterns of grey matter volume alterations and clinical symptoms. Proinflammatory cytokine clusters IL-6 and CRP were associated with more severe symptoms and extensive grey matter reductions, providing novel validation of the clustering approach and suggesting potential targets for immune-focused treatments. In Chapter 5, brain-age prediction models are described demonstrating higher brain age gaps in established schizophrenia patients compared to controls, correlating with symptom severity. This accelerated ageing may be linked to poor defence against oxidative stress, as evidenced by correlations with MRS derived glutathione levels, however no such associations were found with peripheral inflammatory biomarkers. Summarised in Chapter 6, this work suggests that inflammation and oxidative stress make excellent candidates for monitoring of disorder progression and future pharmacological interventions in a stratified cohort; however further research is required to elucidate how peripheral markers are related to central mechanisms and downstream symptomatology and potential future directions for this work are suggested.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):