Frailty may not be a predictor of progression from MCI to dementia

Abstract

AbstractBackgroundSome studies have reported that an individual’s degree of frailty influences the risk of progression from MCI to dementia and correlates with the extent to which those with AD manifest the clinical symptoms of dementia. However, the construct of frailty generally includes factors that can directly compromise cerebro‐vascular function (e.g., cardiac disease or hypertension) or be manifestations of mental and cognitive impairment due to brain pathology. This broad definition of frailty may be problematic in that it measures what it seeks to predict.MethodTo explore the utility of frailty to predict progression from MCI to dementia, we tracked clinical, health, demographic, executive function and behavioral features from the National Alzheimer’s Coordinating Center (NACC) dataset. In 5923 samples over 65 years, with a diagnosis of MCI and at least one visit after MCI diagnosis, 3359 remained MCI at the end of the study while 2564 progressed to dementia. We analyzed common frailty contributors, as well as all other possible variables with sufficient data coverage.ResultAmong these, only 7 of the 40 most commonly considered contributors to frailty were statistically significant differences between MCI patients progressing to dementia and those remaining MCI. The 7 significant features were difficulty in remembering dates, in paying bills, in shopping, in traveling around the neighborhood, in preparing meals, feelings of apathy and change in appetite. Using statistical tests to determine significant categorical features differentiating MCI patients progressing to dementia from those remaining MCI, we found a total of 34 features, of which 21 might be considered executive functioning measures and 5 behavioral symptoms.ConclusionOur findings show that the metrics of frailty that are predictive of progression from MCI to dementia are essentially cognitive measures and thus the construct of frailty may simply be identifying patients with more advanced brain disease, who predictably would be at a greater risk of becoming diagnosed with dementia. The frailty construct may thus mainly offer a metric of dementia severity, and not really a predictor of developing it.