To our knowledge, this is the first study to report multi-area brain activation and walking performance during different types (cognitive and motor) of dual task walking in individuals with PD. Both types of dual tasks exerted negative impact on gait performance including speed, stride length, and swing and stride time. The interference of the cognitive task was greater than the motor task on gait performance, with higher and sustained activation in the PMC and SMA during late phase of cognitive dual task walking than motor dual task walking in participants with PD.
The negative effects of motor and cognitive dual tasks on walking noted in this study were in line with previous results in people with PD (1, 2, 12). In addition, the cognitive dual task exerted more negative effects on cadence and temporal variability than the motor dual task. Galletly and Brauer found the gait speed and stride length was significantly decreased more during walking with the serial subtraction task than walking with the button press task(40). Rochester and colleagues also found that the cognitive dual task had a greater effect on gait performance in PD subjects than motor dual task(41). Taking this together, the dual task walking was a challenge to people with PD and cognitive dual task walking was even more challenging, as also indicated by the higher dual task cost noted in present study.
Previously, we found that the secondary tasks, both the cognitive and the motor, exerted a negative influence on walking performance in healthy adults33, healthy elderly(42), and people with stroke34. However, the cognitive task did not seem to exert greater impact than the motor task on walking, as indicated by gait variability and dual task cost in healthy adults, elderly, and in people with stroke33, 34, 42. It has been suggested that performing dual tasks requires specific cognitive functions, such as set shifting, divided or alternating attention, and response inhibition(43). The global cognitive function of the participants in above mentioned studies33, 34, (42) was similar to that of the participants in present study (MMSE ≥ 24). The particularly negative impact of the cognitive dual task on gait performance may thus be specific to people with PD, which may be associated with impairments in executive function(41, 44, 45). Executive function, one of the cognitive domains, is needed for planning, monitoring, and executing a sequence of motor acts(23). Impairments in executive function have been reported in people with PD(23) and such deficits may limit those with PD to compensate their walking abilities through cognitive strategies, especially during cognitive dual task walking(2). In addition to dual task cost, the particular challenge of the cognitive dual task on walking was also demonstrated by increasing the gait variability. It is noted that the higher gait variability indicates less stability during gait and greater fall risks(46). This important message is suggested to advise people with PD, since cognitive dual task walking is frequently performed during daily activities.
The PFC has been recognized as the key area for executive function(26, 47, 48). Executive function is involved in many daily activities, especially walking. Another task in addition to walking increases attentional demand, prefrontal load, and interferes with gait performance(26, 28). However, brain activity may react differently in neurologically involved people, such as in people with PD. It is interesting to note that the bilateral PFC did not activate more during dual task walking than during SW in the present study. However, our previous study and other studies showed that the PFC was activated significantly during dual task walking in healthy adults and elderly(20-22, 26). In line with present study, Maidan et al., reported no activation difference in the PFC while executing cognitive dual task walking as compared with SW in people with PD(23). In addition, they also found hyperactivity in the PFC during SW in PD population as compared with healthy age-matched adults(23). It is possible that the PFC is highly activated in people with PD to compensate for the deficit in automaticity, even in relatively simple tasks(49, 50). Wu et al., found that PD patients had greater activity in the prefrontal cortex, cerebellum, premotor cortex, parietal cortex, and precuneus while performing automatic movement when compared with healthy controls, as demonstrated by fMRI(51). This compensatory phenomenon of higher activation during a single task may reflect the lower efficiency of the neural network(50). Therefore, people with PD may recruit other brain areas for complex activities. As shown in the present study, recruiting more PMC and SMA activity may be the strategy used to cope with the challenging cognitive dual task walking. We further noted that almost all the brain areas measured in this study were activated more while performing WCT than when performing WMT. Taking together, even by activating more brain areas during WCT, the interference or the negative impact of the cognitive task on walking was still more significant than WMT in people with PD.
The PMC is known to involve motor planning and selection and locomotion(24). The SMA is responsible for posture control and for coordinating movements(25). The PMC and SMA were significantly activated during both cognitive and motor dual task walking according to our previous study in healthy adults(26). Therefore, additional neural processing occurred at least in the PMC and SMA for dual task walking. However, in people with PD, the activations of PMC and SMA during motor dual task walking were not significantly higher than during SW, but were significantly higher during cognitive dual task walking. The activity in different brain areas during dual task walking are less studied and need further elucidation in people with PD. However, according to our results on brain activation and walking performance, it is reasonable to suggest that the cognitive dual task gait training may be a potential intervention to promote plasticity of the motor cortex and dual task walking ability in individuals with PD.
We also examined the relationships between brain activation and walking performance in this study, and found a negative correlation between brain activation and gait performance in general. That is, the higher the brain activation during single and dual task walking, the worse the gait demonstrated. These relationships were also noted in patients with stroke34. Therefore, we speculate that in people with impaired neuromotor control, the increased brain activation in the motor- and cognitive-related areas may not be able to sufficiently fulfill the walking demands, even during single walking.
The present study looked into the different dual task gait performance and multi-area brain activities in people with PD, which has not yet been reported by previous studies. However, there are several limitations that should be noted. First, the sample size of the present study was relatively small, and a study with a larger sample size is needed to validate our findings. Despite the small sample size in our present study, the statistical power was greater than 0.99, indicating that the possibility of a Type 2 error was very low. Second, the clinical disability of most included patients was mild (Hoehn and Yahr stage I-II), therefore, the generalizations of our findings need to consider the severity of the disease. Third, all the participants were tested during the “on-medication” status, thus, it is unclear whether the results could be applied to the off-medication status.