A longitudinal model for disease progression was developed and applied to multiple sclerosis

Area of Science:

• Neurology
• Biostatistics
• Epidemiology

Background:

• Multiple sclerosis (MS) is a chronic neurological disease characterized by progressive disability.
• Modeling disease progression is crucial for understanding patient trajectories and treatment effects.
• Existing models may not fully capture the complexities of chronic disease progression.

Purpose of the Study:

• To develop and validate a robust statistical model for chronic disease progression using multiple sclerosis (MS) as an exemplar.
• To identify key factors influencing the trajectory of disability in MS patients.
• To establish a framework for modeling other chronic diseases.

Main Methods:

• Utilized two large, longitudinal observational cohorts: the University of Wales MS (UoWMS) and British Columbia MS (BCMS) databases.
• Employed multilevel modeling to estimate the Expanded Disability Status Scale (EDSS) trajectory over time.
• Addressed methodological challenges including time axis definition, observation-level variation, relapse adjustments, and autocorrelation.

Main Results:

• A nonlinear function of time since onset best described the EDSS trajectory in the UoWMS cohort.
• Measurement error decreased over time, and ad hoc methods effectively reduced autocorrelation and relapse effects.
• The developed model was successfully validated in the independent BCMS cohort, showing similar time coefficients.

Conclusions:

• Robust models for chronic disease progression can be developed.
• Explicit validation of such models is essential due to inherent methodological complexities.
• This approach provides a foundation for modeling disability progression in other chronic conditions.