Background

In daily life, people need to execute multiple tasks simultaneously during routine activities. This is called ‘dual-task’ ability, and it allows individuals to do activities such as walking and talking or walking while carrying a tray with glasses. It is also essential to preserve balance when performing simultaneous activities [1, 2]. Usually, motor and cognitive activities are executed synchronously, and in the case of any cognitive or motor impairment, the ability to perform these activities is impaired [3].

Gait disorders are the core features of Parkinson’s disease (PD), and simultaneous activities impair the gait of patients with PD. In the late stage of the disease, cognitive deficits and motor impairments including gait and balance problems become more significant. Thus, the dual-task capacity decreases and the functionality of patients PD is impacted negatively [4, 5]. Anatomically, basal ganglia structures are involved in the execution and control of automatic motor actions such as standing and walking. It has been shown that brain regions associated with attention were more active when walking and simultaneously executing an activity compared with the ‘walking only’ condition [6, 7]. In recent studies, it was observed that during dual-task walking, patients with PD were more likely to use the attention sources to perform two tasks simultaneously than healthy people [8]. This may lead to further deterioration of gait under dual-task walking in patients with PD who already have damaged movement control and insufficient attentional resources [9]. Furthermore, the failure of ability to perform a secondary task while walking could be a strong predictor of future falls in patients with PD [10].

Several studies showed associations between dual-task ability and age [11,12,13], motor status, treatment, fatigue [14], freezing of gait [15], cognitive impairment [16], and depression [17] in PD. However, few studies have addressed the types of cognitive tasks that affect dual-task ability [18, 19].

The aims of this study were to compare the cognitive dual-task performance of patients with PD and control subjects and to analyze the effect of different types of cognitive activities on gait performance in patients with PD.

We hypothesized that the gait of patients with PD would be more impaired in dual-task conditions compared with healthy controls, and different cognitive tasks would create different levels of interference on the dual-task walking of patients with PD.

Methods

The study was approved by the Interdisciplinary Clinical Research Ethics Committee, (Protocol Number: 646). All participants were informed about the procedure and gave informed consent. The research also followed the STROBE statement.

Participants

This study is a single-center, descriptive, controlled study. Sixty-five consecutive individuals with Parkinson’s disease (PD) who attended the Faculty of Medicine, Movement Disorders Outpatient Clinic were recruited to the study between December 2015 and May 2017. Also, 57 healthy adults who live in a nursing home were assessed for eligibility to participate in control group. The inclusion criteria for all participants were as follows: (a) having the ability of independent walking without a walking device over flat ground; (b) having a score of 21 or above in the Montreal Cognitive Assessment Test (MoCa) and reaching the maximum target of subtests (delayed memory, digit span-forwards, digit span-backwards, serial sevens) of MoCa [20]; and (c) the ability to reach more than 31.75 cm in the Functional Reach Test (FRT) [21]. The inclusion criteria for patients with PD were (a) having a clinical diagnosis of PD according to the UK Brain Bank criteria [22]; (b) having mild-to-moderate PD (“Hoehn and Yahr” score < 4) [23, 24]; and (c) having a score of 3 or less in item 3 of the Freezing of Gait Questionnaire (FOGQ) [25]. The exclusion criteria were: (a) severe hearing or vision problems; (b) having any cardiac or orthopedic disorder that could interfere with gait; (c) any other neurologic disorders (e.g., dementia, cerebrovascular disease); and (d) having an education level of less than 5 years.

Outcome measures

The Timed Up and Go test

TUG is a simple, widely used, and quick test for the assessment of mobility, balance, and risk of falls. This test is the modified version of the Get up and Go test [26]. To perform the test, participants need to stand up from a standard chair, walk comfortably towards an object on the floor 3 m away from the starting point, turn and walk back to the chair, and to sit down on the same chair. All steps of the test were measured using a chronometer. In the current study, four different dual-task conditions including delayed memory task, digit span-forwards, digit span-backward, and the serial subtraction item of the MoCa were also added to the standard TUG test. The task conditions used in the study were as follows:

  1. 1.

    TUG (Single-Task): The participants were asked to stand up from the chair, walk to the line painted on the floor (3 m away) with average speed, walk back to the chair, and sit down again.

  2. 2.

    TUG + Delayed Memory Task (Dual-Task): The participants were asked to remember 5 words (nose, velvet, mosque, chamomile, purple) before the walk and they needed to repeat these words while performing the TUG test. The cognitive task was performed once simultaneously with TUG.

  3. 3.

    TUG + Digit Span-forwards (Dual-Task): The participants were asked to repeat the 5 digits (2-1-8-5-4) forwards while performing the TUG test. The cognitive task was performed once simultaneously with TUG.

  4. 4.

    TUG + Digit Span-backwards (Dual-Task): The participants were asked to repeat the 3 digits (7-4-2) backwards while performing the TUG test. The cognitive task was performed once simultaneously with TUG.

  5. 5.

    TUG + Serial Sevens (Dual-Task): The participants were asked to subtract 7 serially from 100 while performing the TUG test. The subtractions were performed during TUG.

The patients with PD were evaluated in their ‘on’ period. No instruction was given to focus on either motor or cognitive task before or during the task. Each participant’s entire assessment was completed in the same session. The order of the assessments in each patient was as follows:

  • First: TUG (single task), second: TUG + Delayed Memory Task, third: TUG + Digit Span-forwards, fourth: TUG + Digit Span-backwards, and fifth: TUG + Serial Sevens.

Calculation of the deviations of cognitive performance

The deviations of cognitive performance during the dual-task walking compared with the performance before the TUG test were calculated as a percentage. It was calculated as 20% for each word that could not be recalled in delayed memory (total of 5 words); 20% for each digit that could not be counted in digit span-forwards (total of 5 digits); 33.3% for each digit that could not be counted in digit span-backwards (total of 3 digits); 20% for each incorrect subtraction in serial sevens (subtracting seven from 100, five steps). (Deviation (%) = 100 × [number of deviations during TUG test / number of total items]).

Dual-task interference measurement

To quantify the ability of the individuals to execute two tasks at the same time, we used the traditional formula for evaluating the DTI [5].

$$ \mathrm{DTI}\ \left(\%\right)=\left(\mathrm{single}\ \mathrm{task}-\mathrm{dual}-\mathrm{task}\right)/\mathrm{single}\ \mathrm{task}\times 100. $$

Statistical analysis

Power and Sample Size Calculation Version 3.1.2 was used to determine the sample size. The number of participants that should be included in the study was calculated as 84 (42 tests / 42 controls) in total with the difference of 3.5 s (minimal clinical significance value for the Timed Get up and Go Test for Parkinson’s disease, standard deviation: 3.4) [27], and 90% power (type 1 error = 0.05).

Data were analyzed using the Statistical Package for Social Sciences (SPSS) Version 21.0 software package (SPSS Inc. Chicago, IL, USA). The Shapiro-Wilk test was used to confirm the distribution of the data. To compare quantitative variables between the two groups, the Chi-square test was used for categorical data, and Student’s t test and the Mann-Whitney U test were used to compare continuous variables. The Friedman test was used to compare more than two independent sample groups. A value of p < 0.05 was considered statistically significant for all tests. Post hoc analysis using the Wilcoxon signed-rank tests was conducted with Bonferroni correction with a significance level of p < 0.008.

Results

A total of 122 participants (65 patients with PD, 57 healthy subjects) were assessed for eligibility. Nineteen (14 patients with PD, 5 healthy subjects) were excluded according to the criteria used in the study. Four patients (2 patients with PD, 2 healthy subjects) did not accept to participate in the study, and 3 patients (1 patient with PD, 2 healthy subjects) did not complete the study protocol. In total, 48 patients with PD and 48 age-matched healthy subjects were included in the final analysis (Fig. 1). The demographic and clinical features of the groups are presented in Table 1; there was no statistically significant difference between the two groups in terms of these parameters.

Fig. 1
figure 1

Study diagram

Full size image
Table 1 Demographic and clinical features of groups
Full size table

The median (minimum-maximum) scores of the TUG completion times and the comparison of the two groups are shown in Table 2. There were statistically significant differences between the groups in the completion times of TUG with single tasks and dual-task activities. The PD group completed both the single-task and dual-task activities over a longer time. On the other hand, there were no statistically significant differences in DTI measurements of both groups (Table 3).

Table 2 Comparison of TUG scores under single and dual-task conditions
Full size table
Table 3 Comparison of the dual-task interferences
Full size table

The deviation of cognitive performance in the delayed memory task was statistically higher in the PD group during TUG compared with the control group (p = 0.01). The most increased cognitive performance deviation was found in the ‘serial sevens’ test in both the PD group and control group (PD group: 50.83%, control group: 53.33%) (Table 4).

Table 4 Comparison of percentages of deviations
Full size table

There was a statistically significant difference between the TUG single tasks and each TUG + Dual-Task activity in the PD group (p < 0.05). When the completion time of TUG with different cognitive tasks (delayed memory, digit span-forwards, digit span-backwards, digits serial sevens) were compared, there were significant differences between TUG + serial sevens and all other cognitive tasks (p < 0.001). The completion time of TUG + serial sevens was significantly longer than the other dual-task walking conditions (Table 5).

Table 5 The comparison of different cognitive tasks in PD group
Full size table

Discussion

In the current study, patients with PD were found to have increased completion time of TUG during single- and dual-task conditions. On the other hand, dual-task interference rates were similar between the groups for all cognitive dual tasks. Among the other tasks, the serial seven test was found to have the most significant impact on the TUG test. Although the patients with PD exhibited more errors in the dual-task condition with the memory task compared with healthy subjects, dual-task conditions with more complex attention tests (e.g., serial sevens) impaired the gait speed of the patients with PD more significantly compared with the other conditions.

Studies addressing the effect of dual-task activity on gait and balance in patients with PD yielded a variety of findings. One of the main factors responsible for this variability is the different methodologies used in these studies to assess dual-task performance. Furthermore, dual-task performance can be influenced by many factors such as cognitive status, balance and gait ability, cognitive load, and environmental factors [16]. We designed our study to minimalize the effect of these confounders as recommended in various systematic reviews [1, 28]. As has been shown in previous studies, if a cognitive activity becomes more complex during walking, the dual-task interference increases [29]. It is unclear which cognitive domain causes the most interference under dual-task walking.

One of the targets in this study was to compare the completion time of TUG between patients with PD and healthy age-matched individuals under different amounts of cognitive load. In the literature, although there are studies that use the TUG test with dual-task activities, few studies included healthy subjects for comparison. In one of these studies, Campbell et al. evaluated the participants under different conditions [30]. They observed that the completion time of the gait test was increased in the PD group compared with the control group under both single- and dual-task activities, as we observed in our study. Wild et al. demonstrated decreased gait speed under single- and dual-task activities with the TUG test in patients with PD without dementia compared with healthy subjects. When the results were interpreted according to DTI, similar to our findings, they also found no statistically significant DTI between the PD and healthy control group in the majority of secondary cognitive tasks [31]. In another study, dual-task conditions caused an inreased DTI in patients with PD with freezing of gait [32]. In our study, DTI was not statistically significantly different between the PD and control group. Also, all DTIs had negative values, which indicates that all cognitive tasks were able to create dual-task cost on the completion time of gait in both groups. The cognitive tasks might have caused similar interference on gait in the current study because all participants had similar cognitive level and only PD patients without freezing of gait were included in the study.

Another aim of our study was to analyze the effect of different types of cognitive activities on the completion time of TUG. Serial sevens which was related to complex attention was found to have the most negative impact on gait in the TUG test in our study. Maclean et al. compared high (serial sevens) and low cognitive load (serial threes) on gait, and they showed the speed of gait became slower when the cognitive load was increased in a geriatric population [33]. Li et al. also compared two dual-task paradigms: walking while reciting alternate letters of the alphabet and walking while counting backwards in sevens, and they showed that the impact of dual tasks on gait was associated with the complexity of the cognitive task [34]. Strouwen et al. investigated determinants of dual-task performance in patients with PD by comparing different dual-task activities [35]. The tasks were (1) digit span-backwards while walking, (2) auditory response while walking, and (3) typing the date of the test using the keyboard of a mobile phone while walking. They observed that the speed of gait was decreased in all dual-task activities, but the digit span-backwards task decreased the speed of gait more than the other tasks. In our study, the even backward digits dual-task activity was performed, but another complex attention test (serial sevens) was found to have the most significant effect on gait. A recent review suggested choosing cognitive tasks that require cognitive flexibility, which is an early cognitive impairment in PD for dual-task gait analysis [36]. It has been shown that more challenging walking paradigms require more cognitive flexibility than simple paradigms in older people [37]. From our results, we could make the same comment for the PD population. The serial sevens test, which requires complex attention to accomplish the test, significantly impaired the gait of patients with PD. However, the digit span-backwards test, another test that demands complex attention, did not decrease the speed of gait as much as serial sevens. This may be due to the different nature of the two tests, and the serial sevens test may more impair the supraspinal control of gait.

The other result of this study was that a statistically greater deviation of cognitive performance under TUG + Delayed Memory was found in the PD group compared with the control group. This may indicate that cognitive disturbances related to delayed memory, which are not apparent under single-task conditions, may be uncovered while under dual-task walking in patients with the initial stages of PD. On the other hand, the completion time of TUG + Delayed Memory did not increase as much as in TUG + serial sevens, which had the greatest percentage of deviation in cognitive performance during walking. This result can be interpreted according to capacity theory, which reflects that when two tasks are performed concurrently, competition for limited resources results in deterioration in performance of “one or both tasks” [1]. In the current study, TUG + serial sevens seemed to cause the most distribution in both cognitive and gait performances among other dual-task conditions in patients with PD.

This study has strengths and limitations. First, we controlled factors that may have an effect on dual-task abilities such as balance skill, cognitive status, and freezing of gait through our inclusion criteria. We also used subtests of a standardized cognitive tests and compared the performance of cognitive tasks while walking with prior baseline conditions in our study. One of the limitations in our study was recording only the completion time of the TUG test and not including other spatio-temporal parameters of walking such as cadence, step length, step symmetry, and stride time. Another limitation was the inclusion of cognitive tasks but not motor tasks in the dual-task conditions. Future studies should include both motor and cognitive tasks in dual-task conditions to compare the effects of cognitive and motor activity on the performance of gait. Longitudinal studies will provide more detailed information about the specificity and sensitivity of the serial sevens test with TUG to predict the possibility of falls in patients with PD.

Conclusions

We determined that individuals with PD have more dual-task interference when compared with healthy controls. TUG with the complex attention function was the most effective dual-task assessment to deteriorate the walking function and balance in PD. It should be taken into consideration that complex cognitive activities will slow down walking in PD from the early phase of the disease.