Harnessing brain imaging data to personalise management of fatigue in inflammatory arthritis

Stefanov, Kristian Ivanov (2024) Harnessing brain imaging data to personalise management of fatigue in inflammatory arthritis. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2023stefanovphd.pdf] PDF
Download (8MB)


Rheumatoid arthritis and psoriatic arthritis are chronic inflammatory conditions in which chronic fatigue persists in the majority of patients despite successful management of disease activity. This multidimensional, disabling fatigue correlates with various brain characteristics. Current treatments inadequately address fatigue, emphasising the importance of exploring its neural underpinnings and what potential imaging the brain has to inform the management of fatigue in these inflammatory arthritis conditions. To do so, I applied brain measures to stratify inflammatory arthritis patients into fatigue-related subgroups with potentially amendable biological differences, identify correlates of different subdimensions of fatigue, and predict fatigue follow-up after fatigue-specific or pharmacological treatments in different inflammatory arthritis cohorts of rheumatoid and psoriatic arthritis. I hypothesised that there are (1) subtypes of fatigue in patients with rheumatoid arthritis, illustrated by distinct subgroups stratified by a relationship between neuroimaging brain characteristics and fatigue; (2) statistically significant correlates of subcomponents of fatigue; (3) statistically significant predictors of fatigue scores after non-pharmacological treatments in rheumatoid arthritis; (4) statistically significant predictors of fatigue scores after pharmacological treatments in rheumatoid and psoriatic arthritis; (5) models that can predict individual fatigue outcomes above chance in a trial of non-pharmacological treatments in rheumatoid arthritis using machine learning to combine multiple neuroimaging and clinical variables.

I found a link between neuroimaging brain connectivity and distinct subgroups in rheumatoid arthritis related to fatigue subdimensions, albeit only within a specific cohort. Associations emerged between brain imaging metrics and baseline fatigue subcomponents, showing varied correlations with different metrics. In rheumatoid arthritis patients undergoing exercise or cognitive-behavioural interventions, baseline brain imaging predictors of fatigue centred on structural connectivity from the precuneus to the anterior cingulate cortex. In contrast, I did not find significant neuroimaging predictors of fatigue in rheumatoid arthritis patients who started a new disease-modifying antirheumatic drug. However, I did find such predictors in psoriatic arthritis patients, encompassing cortical thickness of the visual pericalcarine cortex and functional connectivity within the default mode and salience networks, involving the inferior parietal lobule and anterior cingulate cortex. Finally, models using diverse neuroimaging and clinical modalities along with different machine learning algorithms outperformed models using solely the baseline median fatigue. Significantly, these models did not surpass chance level or replicate their utility in usual care patients in an independent rheumatoid arthritis cohort. Overall, despite not finding a model that can predict individual fatigue outcomes, this research advanced our understanding by pinpointing different fatigue-related brain circuits, delineating associations with subcomponents, and identifying group-level predictors of fatigue. If such findings are utilised by future studies using molecular and brain stimulation techniques, neuroimaging can offer innovative solutions to patients to significantly improve their quality of life.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: R Medicine > RC Internal medicine > RC0321 Neuroscience. Biological psychiatry. Neuropsychiatry
Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Infection & Immunity
Supervisor's Name: Basu, Professor Neil and Cavanagh, Professor Jonathan
Date of Award: 2024
Depositing User: Theses Team
Unique ID: glathesis:2024-84203
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 03 Apr 2024 10:09
Last Modified: 03 Apr 2024 10:25
Thesis DOI: 10.5525/gla.thesis.84203
URI: https://theses.gla.ac.uk/id/eprint/84203
Related URLs:

Actions (login required)

View Item View Item


Downloads per month over past year