In our everyday lives, we frequently walk while performing one or more cognitive tasks, such as conversing with a companion or observing the traffic to anticipate a safe time to cross the street. Consequently, individuals who experience deteriorations in gait during cognitive-motor dual-task conditions may be quite limited in their daily lives and be at risk for falling and serious injury. Improving gait performance in dual-task situations is becoming an important focus of rehabilitation for people with neurological disorders. There has been an increase in the number of studies examining the effects of various intervention protocols on dual-task performance.

In this issue, the systematic review by Fritz et al1 synthesizes the existing literature on dual-task training in people with Parkinson disease, brain injury, and Alzheimer disease or dementia. Fritz et al1 show that despite the growing interest in training dual-task performance for people with neurological disorders, there is currently a limited amount of available literature. Of the 14 studies identified for inclusion in this review, only 3 were randomized controlled trials, and all had small sample sizes. The evidence available for this systematic review indicates that there is some support for the value of dual-task training for improving single-task and dual-task performance in various aspects of gait, balance, and cognition; however, the results were inconsistent across studies. The discrepancies are not surprising given the differences in study populations, measures of dual-task performance, and treatment duration and intensity. Importantly, of the 14 included studies, there were only 2 studies that the authors identified as having a clinically meaningful effect (as shown in Table 1 in the Fritz et al1 article). As clinicians, we typically look for sources of synthesized evidence to guide our decision making in clinical practice. Unfortunately, as Fritz et al1 acknowledge, this systematic review is unable to provide any practical recommendations; however, it does highlight how little we currently know about how to improve dual-task performance.

When planning future studies of interventions to improve dual-task performance, there are a few important issues that we as clinicians and researchers need to address. First, what constitutes "dual-task training"? Fritz et al1 included 4 studies that used virtual reality or gaming. By definition, a dual-task is the simultaneous performance of 2 discrete tasks. This is different from a cognitively demanding motor task-a single task that places high demands on attentional resources. Indeed, this operational definition has been proposed in a new taxonomy for dual-task performance.2 This is not to say that cognitively demanding motor tasks, such as following dance steps on a monitor, cannot improve dual-task performance, but this type of intervention cannot be considered dual-task training because the cognitive component is not a discrete task, independent of the motor task. The second important consideration is whether we even need to use a dual-task training intervention to improve dual-task performance. Fritz et al1 included only dual-task training studies. A recent meta-analysis of exercise interventions in older adults found that physical exercise interventions regardless of whether they involved dual-task training could improve dual-task gait speed more than control interventions.3

To determine the optimal intervention to train dual-task performance, it is necessary to understand the mechanisms by which dual-task interference occurs. Is the important issue one of attentional capacity issue, whereby any intervention that improves the automaticity of gait (ie, decreases the attentional demands of gait) will result in reduced dual-task interference? Or rather is the issue one of attention shifting/task coordination, whereby task-specific practice of dual-task walking is needed to improve dual-task skills? Until we have a better understanding of the mechanisms underlying dual-task interference and the factors that influence dual-task interference in different clinical populations, designing effective interventions will be challenging, and will need to be informed more by clinical experience than evidence.

Ultimately, as clinicians, we aim to maximize functional independence while maintaining patient safety. To date, research on training dual-task performance has emphasized the importance of reducing the dual-task declines in gait (especially gait speed). Although walking slower in dual-task situations may have important functional implications, if this is a self-selected adaptive strategy to maximize safety, then it is not necessarily a negative phenomenon that we should try to change. Rather, we may need to shift our focus to whether patients spontaneously select the safest strategy in a given situation. This may be particularly relevant for individuals who have limited capacity to improve their dual-task gait speed. Thus, when considering treatment outcomes related to dual-task performance, it is important to consider how the attention strategy changes with treatment, and which treatment approaches optimize flexibility of attention allocation during dual-task performance. To this end, we cannot ignore the role that cognitive function plays in dual-tasking.

Treatment approaches for training dual-task performance may need to differ on the basis of the individual's cognitive abilities. For example, if a person is experiencing a degenerative condition such as Alzheimer disease, introducing dual-task training as an intervention may not realistically lead to long-term improvements that decrease falls risk and improve functional mobility. Conversely, if an individual has experienced a brain injury and is demonstrating consistent clinical improvements in cognition and physical skills, dual-task training may be instrumental for achieving safe community ambulation. A final important point to consider when evaluating dual-task training is that outcomes of dual-task performance can only be accurately determined when the dual-task effects on both the cognitive and motor tasks are measured.4 Tradeoffs between the 2 tasks and changes in the attention (prioritization) strategy must be examined to determine whether dual-task performance really has "improved" in a way that is meaningful for real-world function.

REFERENCES