Table 2 Conceptual and practical comparison of linear, exponential and sigmoidal progression models in PD
From: Balancing practicality and complexity in neuroimaging models of Parkinson’s disease progression
Feature | Linear Model | Exponential Model | Sigmoidal Model |
|---|---|---|---|
Assumed rate of change | Constant over time | Rapid early loss, then deceleration | Slow onset, acceleration, then plateau |
Pathophysiological plausibility | Moderate: Oversimplifies complex processes | Good: Reflects probabilistic neuron loss | High: Consistent with prion-like alpha-synuclein spread |
Simplicity and interpretability | High: Easy to communicate | Moderate | Moderate to low |
Data requirements | Low: Few timepoints sufficient | Moderate: Multiple timepoints including early and late | High: Requires data across all disease stages |
Applicability to clinical imaging trials | Widely used | Occasionally used; limited examples | Emerging use |
Sensitivity to floor/ceiling effects | Low | Moderate | High |
Captures early compensation mechanisms | No: Assumes immediate decline without buffering | No: Rapid initial decline leaves little room for compensation | Yes: Initial plateau/slow decline reflects preserved function despite pathology |
