Département de gériatrie, CHU d’Angers, France, FORMADEP, Korian, Paris, France, Département de neurologie, CHU de Genève, Suisse, Département de réhabilitation et gériatrie, CHU de Genève, Suisse, Département de gériatrie, CHU de Nantes, France
Simple and efficient detection of the risk of falling in older adults is an objective yet to be reached. Ten years ago, a study published by Lundin-Olsson et al., entitled «Stops walking when talking as a predictor of falls in elderly people», showed that an interruption of walking while talking was a predictor of fall. Although a majority of studies have reported that fallers under dual-task conditions perform more poorly than non-fallers, predicting falls with dual-task related gait changes remains an issue as no reliable clinical test with acceptable sensitivity and specificity is available. Few studies on the relation between dual-task related gait changes and the occurrence of falls are available, and their results are contradictory. The divergence may probably be related to three main reasons. First of all, there is a lack of understanding of the theoretical models of dual-task interferences. Secondly, confusion factors such as a cognitive impairment or depressive symptoms were not taken into account. Thirdly, the analysis of dual-task related gait change depends on visual observation that does not give sufficient measures of gait and has poor inter-rater reliability related to its subjectivity. A better understanding of the nature of dual-task interferences, the standardization of test conditions as well as the use of user-friendly portable gait analysis systems that provide simple and objective gait measurements with less technical equipment and lower costs will most likely allow the development of efficient dual-task-based fall risk assessment tests. Additionally, the analysis of changes in the performance of attention-demanding tasks seems to constitute a relevant new orientation of research in the detection of risk of falling.