UCLA Health researchers have uncovered four distinct diagnostic pathways leading to Alzheimer’s disease by analysing data from 25,000 patients
Alzheimer’s disease, traditionally linked to a few common risk factors, is now understood to develop in a more complex manner. In a groundbreaking study, UCLA Health researchers examined thousands of patient health records and identified four distinct diagnostic patterns that contribute to the condition. This significant discovery provides a more comprehensive understanding of how Alzheimer’s progresses, shifting the focus from isolated risk factors to the broader, step-by-step journey that many patients experience.
The results are detailed in the journal eBioMedicine.
New insights into how Alzheimer’s develops
Researchers analysed longitudinal health data from over 25,000 patients in the University of California Health Data Warehouse and confirmed their findings in the nationally diverse All of Us Research Program. After filtering the data, the team focused on 5,762 patients who provided 6,794 unique trajectories of Alzheimer’s progression. Using advanced computational methods, including dynamic time warping (a technique for measuring the similarity between two sequences), machine learning clustering (a method for grouping data points into clusters), and network analysis (a method for studying complex systems)—the researchers mapped the temporal relationships between diagnoses leading to Alzheimer’s disease.
“We found that multi-step trajectories can indicate greater risk factors for Alzheimer’s disease than single conditions,” said first author Mingzhou Fu, a medical informatics pre-doctoral student at UCLA. “Understanding these pathways could fundamentally change how we approach early detection and prevention.”
The research identified four major trajectory clusters:
- Mental health pathway: Psychiatric conditions leading to cognitive decline
- Encephalopathy pathway: Brain dysfunction conditions that escalate over time
- Mild cognitive impairment pathway: Gradual cognitive decline progression
- Vascular disease pathway: Cardiovascular conditions that contribute to dementia risk
Each pathway exhibited unique demographic and clinical characteristics, indicating that different populations may be prone to varying progression routes. The study found that 26% of diagnostic progressions displayed a consistent directional order. For instance, hypertension often occurred before depressive episodes, which subsequently increased the risk of developing Alzheimer’s disease.
By recognising these sequential patterns instead of concentrating on diagnoses in isolation, clinicians could potentially transform the diagnosis of Alzheimer’s disease. This perspective comes from lead author Dr. Timothy Chang, an assistant professor of Neurology at UCLA Health, who believes that this approach could greatly enhance patient outcomes.
Predicting disease risk more accurately than single diagnoses
The researchers confirmed their findings with an independent population and discovered that multi-step trajectories predicted the risk of Alzheimer’s disease more accurately than single diagnoses alone. This indicates that healthcare providers could utilise these trajectory patterns for:
- Enhanced risk stratification: Identifying high-risk patients earlier in the progression of the disease.
- Targeted interventions: Interrupting harmful sequences before they advance.
- Personalised prevention: Tailoring strategies based on individual pathways.
The validation process in the All of Us Research Program, which includes a diverse and nationally representative cohort, confirmed that these trajectory patterns are consistent across various populations and demographics. This strong validation establishes a solid foundation for applying these findings in clinical practice, reinforcing confidence in their reliability and relevance.