This study examined the impact of cognition on physical function, specifically on walking ability among non-demented community-dwelling older adults, by evaluating whether executive function affects walking ability, with a higher impact on tasks that present particular challenges. The results of this study showed that the association between executive function and gait speed is task dependent and varies according to the degree of locomotor and sensorial adaptation required for the performance of complex walking tasks.
Executive function ability plays an important role in older adults’ ability both to effectively adapt to complex environments and to adequately allocate the attentional resources that are necessary to successfully complete a given task [3
]. Previous research using dual-task paradigms has shown an association between cognition and sensorimotor adaptation. Walking and balance control are increasingly dependent on cognitive control as a compensatory mechanism for sensory impairments, decreased attentional resources and other age-related deficits [23
Executive function had an effect on gait speed for some (pick up an object, walking over obstacles, carrying a weighted vest) but not for others (talking while walking, carrying a large package) of the selected complex walking tasks included in this study. Previous studies have shown that during level walking, carrying a light package that obscures vision has no effect on gait velocity and lower extremity kinematics; in this case, few gait adaptations may be required for the performance of this task [28
]. Consequently, subsequent demands for appropriate executive control may also be low. In contrast, picking up an object or stepping over obstacles requires both the modification of gait kinematics and changes in the frequency and duration of visual sampling of the environment [29
]. Demands on executive control would be correspondingly high in response to the increased demands for sensorimotor adaptation inherent in these tasks.
In this study, participants with poor executive function exhibited significantly slower gait speed compared with those with good executive function, for both walking and picking up an object and walking over obstacles tasks. These findings are consistent with previous studies that have reported the existence of an association between executive control and walking speed, as well as increased difficulty in performing complex instrumental activities of daily living among subjects with executive dysfunction [3
Executive function differences in per cent change were also dependent on the nature of the dual tasks. In the walking while carrying a package task, the greatest per cent decrement in gait speed was found among the good executive function performance group. In contrast, in the walk over obstacles task, poor executive function performers showed significantly higher per cent decline in gait speed compared with good performers. In the walk and talk tasks, there were no significant executive function group differences. These results are consistent with recent experimental work that has reported an association between the nature/characteristics of dual tasks and cost on gait speed [30
]. Baddeley and Hitch’s [31
] model of a central executive that controls cognitive and attentional resources offers a way of understanding these findings. In dual-task paradigms, the cost of performance on the primary task is greater when the two tasks are in the same perceptual domain. The complex walking tasks selected for this analysis represent dual tasks within similar (e.g. walking with obstacles) and different (e.g. walking while talking) perceptual domains.
Poor executive function may also be associated with decreased awareness of a risk inherent in a task, and subsequently the need to modify behaviour in response to that risk. Indeed, declines in gait speed can be an appropriate adaptive behaviour under certain conditions such as walking on icy surfaces, walking under low-light conditions or walking on uneven surfaces while carrying a load [11
]. In these particular situations, failure to adapt gait speed to a change in environmental conditions may represent a judgment error consistent with lower executive function control. The fact that good executive function performers exhibited per cent declines in gait speed for the walk while carrying the package task which were significantly greater than those of poor performers (5 versus 0%; P
<0.0031) suggests that elements such as perceived task-associated risk, task-accuracy estimation and self-monitoring all of which are considered to be under executive function control [1
] may play a relevant role in the performance of this task and could possibly explain the group differences observed in this study.
Results from our study support the concept that examining gait speed on complex walking tasks in relationship to a reference baseline task can be used to infer an individual’s ability to adapt gait to changing task demands. However, our findings suggest that a reduction in gait speed is not always indicative of impaired locomotor adaptation. We suggest that a reduced ability to adapt gait speed to task demands can manifest either as an inappropriate (or disproportionate) reduction in gait speed or alternatively as a failure to modify gait speed appropriately to increased task demands. Finally, our results suggest that the clinical interpretation of complex tests of physical function, particularly among older people, must include a careful consideration of both the nature of the selected complex task and the intra-individual characteristics of the target population.
One limitation of this study is that we used cross-sectional data that do not allow us to assess causal relationships. Longitudinal research aimed at assessing the predictive value of executive function on walking ability is needed to better understand the mechanisms whereby the central nervous system affects physical function in later years and to identify possible target areas for preventive interventions. Another important limitation is that executive function in the InChianti study was assessed solely with the TMT. Although this is a widely used test, there are other measures of executive control that in conjunction with the TMT would have provided a more comprehensive evaluation of executive control functions.
Complex walking tasks measure the ability to walk in diverse and challenging environments and may provide a better indication of the ability to function in daily life than simple standardised gait speed measurement. Findings of this study suggest that executive function significantly affects older adults’ ability to appropriately adapt to environmental challenges and that this effect varies by the complexity of the task. These results further highlight the importance of cognitive flexibility and psychomotor speed for the performance of certain complex tasks. Further research is needed to determine whether improvement in executive function abilities translates to better physical function and performance on selected complex walking tasks.