Digital speech-based markers to advance prognosis in Alzheimer's disease

Validated prognostic tools are essential to advance drug development and clinical care in Alzheimer's disease, particularly as the field shifts toward the prevention of cognitive decline. While progress has been made in developing blood-based biomarkers for the early detection of amyloid pathology, amyloid positivity alone does not reliably predict progression to symptomatic disease. Digital markers can serve as complementary prognostic tools to inform early intervention strategies. Among digital markers, speech-based markers offer a scalable, non-invasive, and cost-effective approach to predicting and monitoring cognitive decline. However, the development of validated speech-based tools has been constrained by the lack of large, multilingual datasets with longitudinal sampling, deep phenotyping, harmonized clinical and biomarker data, and adequate representation of preclinical populations. SpeechDx is a 3-year, multinational, multilingual observational study (n = 2006) designed to address these gaps and accelerate the development of speech-based tools to inform early risk assessment, enable timely intervention, and guide personalized care.