New real-world data platform launched to support Alzheimer’s research

A multi-institutional project, the M3AD Study and Real-World Data Metaplatform, brings together large-scale electronic health record data from three US cities to examine how multimorbidity and social factors shape dementia risk, with the aim of improving prediction, prevention, and care strategies.

A new large-scale research initiative, the Multimorbidity Three-City Alzheimer’s Disease EHR (M3AD) Study and Real-World Data Metaplatform, has been launched to support advances in Alzheimer’s disease and related dementias (AD/ADRD) research using real-world clinical data. The initiative, supported through a grant from the National Institute on Aging, is led by researchers at Columbia University Mailman School of Public Health, working with collaborators from the Vagelos College of Physicians and Surgeons, the School of Nursing, the University of Miami, and the University of Chicago.

Details of the platform, including its design and early feasibility, have been outlined in a recent publication in Alzheimer’s & Dementia. The project establishes a federated data infrastructure that harmonizes longitudinal electronic health record (EHR) data from three major US health systems: New York-Presbyterian’s Clinical Data Warehouse, the University of Chicago Clinical Research Data Warehouse, and the University of Miami Health System. The resulting dataset covers nearly 10 million patients, including approximately 60,000 individuals with AD/ADRD.

With more than 7.2 million older Americans living with AD, many also experience multiple chronic conditions. The metaplatform is designed to address this complexity by examining how dementia develops in the context of multimorbidity. By linking longitudinal clinical data with neighborhood-level census information, it will enable analysis of how clinical, behavioral, and social determinants of health interact over time to shape disease trajectories.

Highlighting the value of this approach, George Hripcsak, Vivian Beaumont Allen Professor of Biomedical Informatics at Columbia and a co-author of the study, explained the role of EHRs in enabling long-term analysis of disease trajectories:

“EHRs make it possible to analyze these interacting trajectories across decades of care. By examining longitudinal clinical histories, the platform may help identify previously unrecognized early warning signs of dementia, while capturing the broader context of patients’ lives.”

Advanced analytical methods, including machine learning, dynamic prediction models, and causal inference techniques, are embedded within the platform to generate deeper insights into AD/ADRD. These approaches aim to identify factors associated with increased or reduced risk, while supporting more individualized assessments of when interventions may be most effective.

The metaplatform also integrates predictive tools such as the Electronic Health Record Risk of Alzheimer’s and Dementia Assessment Rule (eRADAR), which uses routinely collected EHR data to identify individuals who may have undiagnosed dementia and could benefit from further clinical evaluation. Inclusion of the algorithm extends case detection beyond traditional coding-based approaches, supporting a more complete representation of dementia within research datasets.

In addition to risk prediction, the platform provides a resource for evaluating prevention strategies in real-world populations. The researchers will be able to examine how modifiable factors, including smoking, weight, and blood pressure in midlife, are associated with later cognitive outcomes, helping generate evidence to inform both clinical practice and public health strategies.

Tatjana Rundek, Professor of Neurology and Director of the Evelyn F McKnight Brain Institute at the University of Miami and co-author of the study, highlighted the importance of capturing the broader patient context in dementia research:

“To truly understand and prevent dementia, we need to look at the whole person, even before symptoms appear. This study provides an opportunity to identify early signals of dementia and generate insights that could inform real-world care, particularly for individuals managing multiple health conditions.”