Development of an Innovative, Non-invasive, Wearable Multi-parameter System for the Early Prediction of Cognitive Decline and Dementia in Older Adults

BACKGROUND: Early detection of preclinical symptoms and prediction of potential development of mild cognitive impairment (MCI) and Alzheimer's disease (AD) could improve non-pharmacologic, life-style and exercise related preventative interventions' efficacy and slow-down disease progression. To achieve this goal, discriminating the earliest preclinical stage of MCI/AD from healthy state would be necessary. However, this is still challenging and current clinical methods are not feasible for preventative screening in larger populations of older adults, as they involve invasive sampling of molecular blood or cerebrospinal fluid biomarkers, as well as expensive brain imaging and extensive neuropsychological testing. Recently, several non-invasive alternative measures, including electroencephalography (EEG), gait analysis, heart rate variability (HRV), and core body temperature (Tc), were shown to be associated with preclinical symptoms of MCI/AD and to predict disease progression.

AIM: The investigators aim to combine these measures in a novel non-invasive multi-parameter prediction model, which better reflects multimodal symptomatology compared to currently used methods and, therefore, allows discriminating healthy persons from MCI state with adequate sensitivity (i.e. >80%).

METHODS: A cohort of 85 older adults, ≥65 years of age, including healthy persons and patients with MCI, will be recruited. Assessments will be performed at baseline, after 2 months (within these two 2 months one group will follow a cognitive-motor training intervention, while the other serves as passive control), and at 12-month follow-up. Assessments include EEG, gait analysis, HRV, and Tc at rest and during walking, and will be compared to reference measures of MCI status, including neuropsychological tests, to develop the prediction model and evaluate its sensitivity.