Multimorbidity in neurodegenerative diseases: a network analysis.

Area of Science:

• Medical Informatics
• Network Science
• Public Health

Background:

• Neurodegenerative diseases (NDs) impose substantial socioeconomic burdens.
• Comorbidities significantly exacerbate these costs and impact patient outcomes.
• Understanding disease associations is crucial for effective healthcare management.

Purpose of the Study:

• To analyze prevalent comorbidities associated with neurodegenerative diseases (NDs) using network analysis.
• To identify key diseases that frequently co-occur with NDs.
• To provide a framework for improved resource allocation and patient care.

Main Methods:

• Construction of a multimorbidity network (MN) from a large longitudinal electronic health record (EHR) dataset (93,647,498 diagnoses, 824,847 patients).
• Application of Phi-correlation and Cosine Index (CI) to measure disease associations.
• Utilized network centrality measures to rank prevalent comorbidities.
• Generated class-level networks for disease prevalence and strength.

Main Results:

• The general MN comprised 928 diseases and 337,253 associations.
• Networks at 99% confidence revealed 575 relationships (73 diseases via Phi-correlation, 102 via CI).
• Thirteen diseases were identified as the most notable multimorbidities based on centrality measures.
• Specific analysis of NDs identified key associated conditions and relationships.

Conclusions:

• Network analysis provides a robust framework for understanding complex disease comorbidities in neurodegenerative diseases (NDs).
• Identification of key multimorbidities can guide clinical practice, preventive strategies, and resource allocation.
• This approach can lead to improved quality of life for patients and caregivers.