Directional Symptom Dependencies in Multiple Sclerosis and Parkinson's Disease: A Comparative Bayesian Network...

Area of Science:

• Neuroscience
• Computational Biology
• Clinical Psychology

Background:

• Multiple sclerosis (MS) and Parkinson's disease (PD) are complex neurological disorders.
• These conditions involve interconnected motor, psychological, sleep, fatigue, and pain symptoms.
• Conventional methods struggle to determine the direction of influence among these symptoms.

Purpose of the Study:

• To apply Bayesian Network (BN) analysis to identify and compare directional pathways among disease characteristics, physical function, psychological measures, sleep, fatigue, pain, and physical activity in MS and PD patients.
• To uncover distinct dependency structures within each disease cohort.
• To identify potential therapeutic targets based on network analysis.

Main Methods:

• Cross-sectional data from 104 MS patients and 54 PD patients were analyzed.
• Bayesian networks were estimated using the Hill-Climbing algorithm with Gaussian BIC scoring.
• Bootstrap analysis assessed edge reliability, and linear regression quantified relationship strengths.

Main Results:

• The MS network showed anxiety as a central hub, directly predicting physical activity, sleep quality, physical fatigue, and pain. Pain interference mediated links between sleep/depression and fatigue.
• The PD network also featured anxiety as a central hub, influenced by depression and predicting pain interference and balance. Age directly impacted physical activity and balance.
• Both networks exhibited distinct structures, with anxiety as a common central node.

Conclusions:

• Bayesian Network analysis revealed unique dependency structures for MS and PD.
• Anxiety emerged as a shared central hub, but its downstream effects differed between the conditions.
• Findings suggest anxiety is a plausible intervention target for both MS and PD, warranting further longitudinal and interventional studies.