The Florey Dementia Index for Alzheimer’s onset prediction

A predictive tool for determining the age at which individuals may develop mild cognitive impairment (MCI) or Alzheimer’s dementia (AD) has demonstrated the ability to predict MCI onset within 2.78 years and AD onset within 1.48 years.

Developed by Florey Institute of Neuroscience and Mental Health researchers, collaborating with the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and the Australian Imaging, Biomarker, and Lifestyle (AIBL) study, the team has released validation results of the Florey Dementia Index (FDI).

FDI predicts the age at onset of MCI and AD with high accuracy, potentially enabling patients and clinicians to better plan care and prioritize early treatment before the presentation of classic symptoms.

Alzheimer’s disease is a progressive neurodegenerative disorder that develops gradually from MCI to dementia, leading to significant loss of independence. While previous research has seen early indication signals related to the gut microbiome, no clinical tools currently exist for reliably predicting the age at onset.

Allowing for interventions with emerging disease-modifying therapies

Transitioning from MCI to AD escalates the need for comprehensive care. Reliable methods to forecast the onset age of these conditions, especially in preclinical phases, would allow for interventions with emerging disease-modifying therapies.

In the study ¹, the researchers introduce the FDI statistical modeling approach for predicting disease onset and test the index in an Alzheimer’s-specific cohort.

The FDI model incorporates age and Clinical Dementia Rating Sum of Boxes (CDR-SB) scores, relying solely on noninvasive data collection methods to balance predictive accuracy with accessibility.

The prognostic study utilized data from 3,694 participants across the AIBL and ADNI cohorts, from October 2004 to March 2023. Data from 93 participants in the Anti-Amyloid Treatment in Asymptomatic Alzheimer (A4) study were also employed in a simulated trial to assess the FDI’s applicability. The study excluded individuals with MCI or dementia not associated with Alzheimer’s disease.

Results indicate that among AIBL and ADNI participants, the FDI accurately predicted the onset of MCI and AD. The mean absolute error (MAE) for predicting the age of MCI onset was 2.78 years, with an MAE of 1.48 years for AD onset. Lower prediction errors were observed in the simulated A4 trial, a group with preclinical Alzheimer’s disease, with MAEs of 1.57 years for MCI and 0.70 years for AD.

No adjustments for comorbidities or demographics

No adjustments were made initially for medical comorbidities or demographic factors, including sex, and results still demonstrated consistency across datasets, supporting the model’s robustness and generalizability.

When well-characterized comorbidities (hypertension, stroke, neurologic disorders, psychiatric disorders) were included in the assessment, FDI showed a slightly enhanced performance over the generalized version. Adjustments for sex slightly improved performance for predicting AD onset but not MCI onset.

Survival analysis revealed a sharp decline in dementia-free probability at specific FDI thresholds. MCI onset was linked to an FDI threshold of 79, while AD onset corresponded to a threshold of 85.

FDI offers a novel approach to predicting the onset of MCI and AD, providing clinicians with a low cost, accessible tool that relies on noninvasive, widely available metrics. In clinical terms, it could allow early planning of treatment options, or prioritize patients for emerging treatments such as disease-modifying monoclonal antibody drugs.

It would also be useful for patients planning for future challenges with dementia and home care, allowing them to make autonomous decisions while they are best able to do so.

Though validation in broader and more diverse cohorts is necessary, this tool holds promise for advancing personalized dementia care and improving patient outcomes.