The Brain Health Platform: Combining Resilience, Vulnerability and Performance to assess brain health and risk of Alzheimer’s disease and related disorders

Abstract Background One of the great challenges in conducting ADRD research in community samples is the standardized measurement of medical, social, and environmental factors that may increase the risk of ADRD (vulnerability) or may provide some protective benefits from ADRD (resilience). While all physicians can diagnose mild cognitive impairment (MCI) and ADRD, it is much more difficult to assess the status of brain health and identify those at risk for future impairment. We developed and validated a novel clinical tool, the Brain Health Platform, to quantify brain health and identify risk factors for MCI and ADRD. Method We combined the Resilience Index (RI), a brief, quantifiable measure of 6 modifiable factors of brain health and risk of MCI and ADRD, the Vulnerability Index (VI), a weighted measure of cognitive impairment risk that assesses modifiable and non‐modifiable risk factors, and the Number Symbol Coding Task (NSCT), a brief, 90‐second executive task that incorporates attention, planning, and set‐shifting, to measure objective cognitive performance, into the Brain Health Platform to assess overall brain health and the presence of cognitive impairment. We tested this in a combined clinical and research sample (n = 400; 53 controls, 347 impaired) compared with a Gold Standard clinical‐cognitive‐behavioral evaluation. Result The Brain Health Platform was able to correctly segregate healthy controls from MCI and ADRD in time‐ and cost‐efficient fashion. We applied a linear discriminant analysis using the NSCT, RI and VI to predict control, MCI and ADRD groups performing 5‐fold cross validation with 1000 replications. The area under the curve for Controls vs MCI was 0.823±0.09, for Control vs ADRD was 0.974±0.03 and for MCI vs ADRD was 0.884±0.06. In an exploratory fashion, we also demonstrated the ability of the Brain Health Platform to assist in differential diagnosis. Conclusion The Brain Health Platform can intake the inputs of three brief assessments (RI, VI, NSCT) and output a patient risk score to determine the cross‐sectional status of an individual’s brain health, from which effective, personalized, precision‐medicine‐like prevention interventions, including lifestyle changes, medications and clinical trial enrollment, can be implemented in at‐risk individuals to decrease the development or progression of cognitive decline.