A powerful new real-world data platform could change how scientists understand and predict Alzheimer’s disease and related dementias (AD/ADRD), according to a study led by researchers at Columbia University Mailman School of Public Health and their collaborators. The initiative, known as the M3AD Study and Real-World Data Metaplatform, brings together expertise from multiple institutions to advance research on ageing, dementia prevention, and care. The findings are published in Alzheimer’s & Dementia.
The platform uses electronic health records collected from three major U.S. cities and includes data from around 60,000 individuals living with Alzheimer’s disease or related dementias. By combining these records, researchers can follow how health conditions, behaviours, and social factors interact over time to influence dementia risk. This creates one of the most comprehensive datasets ever assembled for studying dementia in real-world settings.
Unlike traditional research that often focuses on a single disease, this platform takes a broader approach. It examines how multiple chronic conditions, lifestyle factors, and social environments work together and change over time. This allows scientists to capture the complexity of ageing better and to improve predictions about who may develop dementia and how the disease may progress.
The need for this kind of approach is increasing as populations age. In the United States, more than 7.2 million older adults are living with Alzheimer’s disease, including a significant share of those aged 85 and older. At the same time, most older adults are managing more than one chronic illness. This pattern, known as multimorbidity, makes both diagnosis and care more challenging and highlights the limits of studying diseases in isolation.
Researchers stress that dementia does not occur on its own. Instead, it develops through complex interactions between health conditions, behaviours, and life circumstances over many years. By analysing long-term clinical histories, the platform may help identify early warning signs of dementia that have previously been overlooked, creating opportunities for earlier detection and intervention.
The data come from three major health systems: NewYork-Presbyterian Hospital, the University of Chicago, and the University of Miami. Together, they provide decades of patient information covering millions of individuals. The dataset also reflects a diverse population, including people from different racial and ethnic backgrounds, allowing researchers to study dementia risk across a wide range of communities.
In addition to its scale, the platform uses advanced analytical tools, including machine learning, to identify patterns in complex data. It also relies on a federated system that allows institutions to work together while keeping patient data secure. Over time, the platform can be expanded to include additional sources of information, such as imaging, genetic data, and new biomarkers.
Beyond improving risk prediction, the platform will help researchers test prevention strategies in real-world populations. It can be used to study how factors like smoking, healthy weight, and blood pressure in midlife influence later cognitive decline. By linking clinical data with social and environmental information, the initiative also supports a more holistic understanding of dementia, helping to guide better care and future research.
More information: Moise Desvarieux et al, Accelerating real-world prediction and research in Alzheimer’s: The M3AD study, Alzheimer’s & Dementia. DOI: 10.1002/alz.71174
Journal information: Alzheimer’s & Dementia Provided by Columbia University’s Mailman School of Public Health
