Impaired executive function can predict recurrent falls in Parkinson's disease

Abstract

Objective: To examine whether impairment in executive function independently predicts recurrent falls in people with Parkinson's disease (PD). Design: Prospective cohort study. Setting: University motor control research laboratory. Participants: A convenience sample of community-dwelling people with PD (N=144) was recruited from a patient self-help group and movement disorders clinics. Interventions: Not applicable. Main Outcome Measures: Executive function was assessed with the Mattis Dementia Rating Scale Initiation/Perseveration (MDRS-IP) subtest, and fear of falling (FoF) with the Activities-specific Balance Confidence (ABC) Scale. All participants were followed up for 12 months to record the number of monthly fall events. Results: Forty-two people with PD had at least 2 falls during the follow-up period and were classified as recurrent fallers. After accounting for demographic variables and fall history (P=.001), multiple logistic regression analysis showed that the ABC scores (P=.014) and MDRS-IP scores (P=.006) were significantly associated with future recurrent falls among people with PD. The overall accuracy of the prediction was 85.9%. With the use of the significant predictors identified in multiple logistic regression analysis, a prediction model determined by the logistic function was generated: Z = 1.544 + .378 (fall history) − .045 (ABC) − .145 (MDRS-IP). Results: Forty-two people with PD had at least 2 falls during the follow-up period and were classified as recurrent fallers. After accounting for demographic variables and fall history (P=.001), multiple logistic regression analysis showed that the ABC scores (P=.014) and MDRS-IP scores (P=.006) were significantly associated with future recurrent falls among people with PD. The overall accuracy of the prediction was 85.9%. With the use of the significant predictors identified in multiple logistic regression analysis, a prediction model determined by the logistic function was generated: Z = 1.544 + .378 (fall history) − .045 (ABC) − .145 (MDRS-IP).