Prediction of functional dependence with digital clock drawing, word recall, and assessment of lifestyle and health factors

Abstract

AbstractBackgroundDistinguishing mild cognitive impairment (MCI) and Alzheimer’s disease‐related dementia (ADRD) requires assessing the clinical impact of cognitive deficits on a patient’s activities of daily living. To accurately monitor cognitive decline, it is thus critical to combine early screening of cognitive deficits with the assessment of functional impairment. The Digital Clock and Recall (DCR™), consisting of DCTclock™ and a 3‐word delayed recall, employs AI‐enabled performance analyses to evaluate memory, executive function, visuospatial abilities, and motor functions. The Life and Health Questionnaire (LHQ) is a brief digital survey that captures key lifestyle and health factors for cognitive decline/dementia. We evaluated whether the combination of the DCR and LHQ can predict functional status, measured by the Functional Abilities Questionnaire (FAQ) reported by an informant, in older individuals with statuses ranging from cognitively unimpaired to MCI and ADRD.Method849 participants from a multisite study (age mean±SD = 72.1±6.7; 56.8% female; years of education mean±SD = 15.4±2.7), classified as cognitively unimpaired (n = 369), MCI (n = 262), or ADRD (n = 215). Mild (mFD) and moderate (modFD) levels of functional dependence were defined as FAQ score ≥ 6 and ≥ 9, respectively. We conducted 10‐fold cross‐validated regularized logistic regressions using LHQ responses and DCR metrics to predict mFD or modFD, controlling for age, sex, and education.ResultLHQ significantly predicted mFD and modFD (p’s<0.001, R2 = 0.13, and R2 = 0.12). Adding DCR metrics to the model improved both models (p’s<0.001, R2’s = 0.18 and 0.13, respectively). Surviving LHQ questions for predicting mFD related to unintended weight loss, social engagement/volunteering, cognitively stimulating tasks, sleep, and life satisfaction/purpose. Surviving LHQ questions for predicting modFD related to unintended weight loss, social engagement/volunteering, cognitively stimulating tasks, and sleep.ConclusionA brief assessment that can be completed in primary care settings in <10 min consisting of an objective digital cognitive assessment and a survey of self‐reported lifestyle and health factors can not only identify early cognitive impairment but also reliably predict level of functional dependence in instrumental activities of daily living ‐ without need for an informant.