Unobtrusive remote monitoring to identify and track meaningful change in daily function in community dwelling older adults at risk for Alzheimer's disease

Resubmission 1 R01 AG058687-01 PI: Adriana Seelye, Ph.D. Project Summary. Alzheimer's disease (AD) is a growing public health concern that has a profound negative impact on individuals living with the disease and their families, the healthcare system, and the economy. Early identification and accurate assessment of mild cognitive and functional deterioration in older adults at risk for AD will be critical in order to intervene at the earliest stages of the disease and to reduce the cost and consequences of functional declines. Project Objectives. The overall goal of this research is to use unobtrusive in-home sensor technologies in the home environment to objectively assess high-level IADL functioning in a new way that will more effectively identify the earliest subtle declines in functioning that slowly emerge and ultimately threaten independence. The specific aims for this research are: Aim 1 will establish the most robust remotely monitored IADL variables and combinations of variables to discriminate between MCI and intact cognition based on sensitivity, specificity, accuracy and overall ROC AUC using cross-sectional analyses. Aim 2 will determine the discriminatory ability of the new remotely monitored IADL variables compared to available IADL questionnaires to differentiate those with MCI relative to those with intact cognition based on sensitivity, specificity, accuracy and overall ROC AUC in cross-sectional analyses. Aim 3 will identify and characterize the longitudinal trajectories (slopes) of remotely monitored IADL functioning over time between those with intact cognition and those with MCI using generalized linear mixed effects models, with person-specific IADL distributions and their changes as outcomes. Project Methods. The proposed project will apply innovative computing and ambient sensing technologies directly in participants' home and community environments to objectively assess IADL performance and variability among older adults with and without MCI in multiple clinically relevant functional domains for up to 4 years. Study involvement will also include annual neuropsychological and clinical testing. Machine learning computational approaches will be used to examine a large number of sensor-based IADL candidate variables generated through this study to determine the relative importance of these variables for discriminating between MCI and intact cognition groups, cross-sectionally and longitudinally. Project Impact. The approaches used in this study will allow researchers, physicians, and caregivers to proactively identify and monitor increasing risks for deteriorating cognitive function (progressing from normal aging to MCI and from MCI to AD) in a way that is not currently possible, transforming AD prevention trials and significantly reducing the cost and consequences of functional decline in our aging population.