A new method for predicting dementia, developed by researchers at Queen Mary University of London, has demonstrated over 80% accuracy and can identify the onset of the condition up to nine years before a formal diagnosis. This innovative approach surpasses traditional methods such as memory tests and measurements of brain shrinkage, which are currently used to diagnose dementia.

The significance of this advancement is underscored by the global impact of dementia, which affects over 55 million people worldwide. Early and accurate diagnosis has been a longstanding challenge in the medical community.

“Predicting who is going to get dementia in the future will be vital for developing treatments that can prevent the irreversible loss of brain cells that causes the symptoms of dementia,” said Charles Marshall, Professor and Honorary Consultant Neurologist, who led the research team within the Centre for Preventive Neurology at Queen Mary’s Wolfson Institute of Population Health. Their groundbreaking work was published today in Nature Mental Health, with Sam Ereira, of Queen Mary’s Center for Preventive Neurology, as the first author.

The researchers developed their predictive method by analyzing functional MRI (fMRI) scans for changes in the brain’s default mode network (DMN). The DMN connects various regions of the brain to perform specific cognitive functions and is the first neural network affected by Alzheimer’s disease.

Using fMRI scans from over 1,100 volunteers from the UK Biobank—a comprehensive biomedical database containing genetic and health information from half a million UK participants—the team estimated the effective connectivity between ten regions of the brain that constitute the DMN. Each patient was assigned a “probability of dementia value” based on how closely their brain connectivity patterns matched those indicative of dementia. These predictions were then compared to the patients’ medical data from the UK Biobank.

The model successfully predicted the onset of dementia with greater than 80% accuracy up to nine years before an official diagnosis. Furthermore, for participants who developed dementia, the model could estimate, within a two-year margin of error, the time until diagnosis.

The research also explored whether changes in the DMN might be influenced by known risk factors for dementia. They discovered a strong association between genetic risk for Alzheimer’s disease and connectivity changes in the DMN, indicating these changes are specific to Alzheimer’s. Additionally, social isolation was found to increase dementia risk through its impact on DMN connectivity.

“Although we are getting better at detecting the proteins in the brain that can cause Alzheimer’s disease, many people live for decades with these proteins without developing symptoms of dementia. We hope that the measure of brain function we have developed will allow us to be much more precise about whether someone is actually going to develop dementia, and how soon, so that we can identify whether they might benefit from future treatments,” said Marshall.

Sam Ereira added, “Enormous potential exists to apply these methods to different brain networks and populations, to help us better understand the interplays between environment, neurobiology, and illness, both in dementia and possibly other neurodegenerative diseases. fMRI is a non-invasive medical imaging tool, and it takes about six minutes to collect the necessary data on an MRI scanner, so it could be integrated into existing diagnostic pathways, particularly where MRI is already used.”

This breakthrough in dementia prediction represents a significant step forward in the early detection and treatment of this debilitating condition, offering hope for improved patient outcomes in the future.

By Impact Lab