McGill University, Canada
Title: Fall prediction in older adults: From dual-task to artificial neural network
Biography: Olivier Beauchet
Falls are highly prevalent among adults aged 65 years and over with prevalence estimated around 35%. Falls lead to injuries, hospitalization, loss of independence and social disability, which impose high costs to public health and social services. Identification of fallers, which is the first step of falls prevention strategies in older community dwellers, is therefore crucial for their efficiency and cost-effectiveness. There are complex interactions between host-related motor behaviors and environmental characteristics for the mechanism of falls. Due to the interplay between risk factors of falls, the prediction of falls remains difficult and depends on combinations of risk factors of falls. Over the past years, dual task-related gait changes have frequently been reported among older adults. However, published data are heterogeneous and show that impaired dual tasking is and is not associated with falls, or is even an irrelevant fall risk indicator compared to impaired single task performance. More recently it has been reported that Artificual Nerula Networks (ANN), using a set of clinical characteristics corresponding to the most commonly reported risk factors for falling, was an efficient way for the identification of recurrent fallers in older community-dwellers. Thus, these previous results suggest that ANNs could improve the predictive performance of the tools designed to predict falls. The aim of this presenttaion is systematic review all published data which examined the relationship between fall and changes in gait and/or attention-demanding task performance while dual tasking and to open a new perspective with ANN.