Plain-language takeaway
In both Multiple Sclerosis and Parkinson's Disease, anxiety sits at the centre of the symptom network — but the pathways leading out from it differ between the two diseases.
This study is cross-sectional and does not show causation. The arrows represent directional dependency structures learned from data — statistical hypotheses about how symptoms relate, not confirmed cause-and-effect. The authors frame these results as testable hypotheses for future longitudinal and interventional research.
Study at a glance
- DesignCross-sectional
- MS samplen = 104
- PD samplen = 54
- Structure learningHill-ClimbingGaussian BIC scoring
- Bootstrap resamples500edge stability
- Stability threshold≥ 50%
Missing data (< 20%) were handled by multiple imputation (predictive mean matching). Relationship strengths (β) were estimated by linear regression on z-scored variables.
Symptom dependency networks
Each network shows the key reported directed edges from the paper. Arrow direction is from cause-candidate to effect-candidate. Edge thickness is proportional to |β|; colour encodes the sign of the relationship (blue = positive, orange = negative). The hub node (Anxiety) is enlarged. Hover or tap any node or edge for exact β and p-values.
MS: 18 nodes, 27 directed edges total — key reported edges shown. 21 / 27 edges (77.8%) were bootstrap-stable at ≥ 50%.
PD: 14 nodes, 15 directed edges total — key reported edges shown. 14 / 15 edges (93.3%) were bootstrap-stable at ≥ 50%.
MS — 18 nodes / 27 edges · 77.8% stable
PD — 14 nodes / 15 edges · 93.3% stable
Text description of all reported edges (accessible list)
Key reported pathways
Every edge shown below is reported verbatim in the paper. Click a column header to sort; type to filter.
What's shared vs different
Shared
Anxiety is the central hub in both diseases. In MS it radiates outward to physical activity, sleep quality, physical fatigue and pain severity; in PD it drives pain interference and (inversely) balance. This convergence points to anxiety as a candidate intervention target in both conditions.
Distinct to MS
The MS network is denser (27 edges) and features pain interference as a key mediator, relaying influence from sleep and depression onward to cognitive and physical fatigue. Fatigue is reached through multiple converging routes.
Distinct to PD
The PD network is more parsimonious (15 edges) and dominated by age- and anxiety-related pathways: depression strongly drives anxiety (β = 0.78), while age erodes physical activity and balance.
Bootstrap edge stability
Share of directed edges that recurred in ≥ 50% of 500 bootstrap resamples. Higher is more robust.
Conclusion
MS and Parkinson's Disease show distinct, disease-specific symptom dependency structures. Anxiety is a shared central hub in both, but downstream pathways diverge: in MS, pain interference is a key mediator, whereas the PD network is more parsimonious and dominated by age- and anxiety-related pathways. These findings generate hypotheses for longitudinal and interventional work and highlight anxiety as a candidate intervention target in both diseases.