Cognitive and Behavioral Predictors of Fall Risk in Parkinson Disease

Date

2015-07-24

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Falls and their related injuries are a significant health issue for individuals with Parkinson disease (PD). Several factors have been identified that increase fall risk, including cognitive impairment, impulsiveness, and lower balance confidence, as well as PD-related characteristics. However, to date, no definitive predictor profile has been identified. As a result, there is a need to develop a comprehensive model incorporating elements from each of the areas known to have a relationship with fall behaviors in PD. Such information could result in improved identification and treatment of PD patients at higher risk of falls. This study used stepwise logistic regression analyses to identify predictors of retrospectively reported falls from four domains, which included separate cognitive, impulsiveness/impulsive-compulsive disorder (ICD) related behaviors, disease characteristics, and balance confidence models. Each stepwise logistic regression yielded significant results (p < .20), and all of the significant predictor variables were included in a fifth combined model. The combined stepwise logistic regression was significant for postural instability (odds ratio = 8.66), verbal learning (California Verbal Learning Test-2 Total Learning T Score [CVLT-II]) (odds ratio = 0.95), and self-reported behavioral impulsiveness (Barratt Impulsiveness Scale-11 [BIS-11]) (odds ratio = 1.10). Model comparisons using net reclassification improvement (NRI) and the Hanley and McNeil (1983) method were conducted to determine if the combined model was significantly better at predicting fall risk than the domain-specific models. The combined model had the highest rate of accurately predicting fall risk (83%); however, the combined model was not significantly better at predicting fall risk than the impulsiveness/ICD or balance confidence models. These results showed that postural instability was the best predictor of fall risk; however, incorporating cognitive and impulsiveness measures improved prediction of fall risk. In light of these findings, screening for impulsiveness and, when possible, verbal learning, could be incorporated into routine clinical PD evaluations for better identification of patients at higher risk of falls.

General Notes

Table of Contents

Subjects

Accidental Falls, Parkinson Disease, Postural Balance, Psychomotor Performance, Risk Assessment

Citation

Related URI