A standardized video-based asynchronous neurological examination (VANE) deployed across 25 clinical sites in the Diabetes Prevention Program Outcomes Study (DPPOS) successfully completed 1,237 of 1,284 reviewable exams (96%) without errors, with only 6 (0.4%) showing major deviations. Designed to detect diabetic cranial neuropathies, stroke signs, and parkinsonism in a 25-year cohort of adults with pre-diabetes and type 2 diabetes, the 10–15 minute iPad-recorded exam required no direct examiner-participant contact and was centrally reviewed by three physicians across 55 trained research staff.
This preprint — not yet peer-reviewed — presents primarily a methodological feasibility demonstration rather than a clinical discovery, but its implications for dementia research infrastructure are meaningful. The intersection of metabolic disease and neurodegeneration is a high-priority research frontier: type 2 diabetes is an established risk factor for Alzheimer's disease, and large longitudinal cohorts like DPPOS are rare and valuable. Standardizing neurological phenotyping remotely addresses a chronic bottleneck in multi-site aging studies — specialist availability and inter-examiner variability. By using commodity hardware (iPad) and scripted protocols in English and Spanish, the VANE model lowers barriers for underrepresented populations. Limitations include the absence of inter-rater reliability data and validation against in-person gold-standard exams, which the authors do not yet report. If validated, this approach could accelerate dementia screening in metabolic disease cohorts globally — an incremental but practically significant contribution to epidemiological methodology.