Impact of Chronic Obstructive Pulmonary Disease on Cognitive and Motor Performances in Dual-Task Walking

ABSTRACT

When two tasks are performed simultaneously, they compete for attentional resources, resulting in a performance decrement in one or both tasks. Patients with attention disorders have a reduced ability to perform several tasks simultaneously (e.g., talking while walking), which increases the fall risk and frailty. This study assessed the cognitive and motor performances of patients with COPD and healthy controls within a dual-task walking paradigm. A subobjective was to assess the impact of a pulmonary rehabilitation program on the dual-task performances in COPD. Twenty-five patients with COPD and 20 controls performed a cognitive task (subtraction) and a 15-m walking test separately (single-task; ST) and jointly (dual-task; DT). In addition, a subsample of 10 patients performed the same evaluations 5 weeks later after a pulmonary rehabilitation program following current recommendations. Cognitive and gait performances in ST showed no differences between patients with COPD and controls (all p > 0.05). However, COPD patients exhibited a greater increase in gait variability than controls in DT (4.07 ± 1.46% vs. 2.17 ± 0.7%, p < 0.001). The pulmonary rehabilitation program had no effect on the dual-task impairment for the subsample of patients (p = 0.87).

This study provides evidence of insufficient attentional resources to successfully deal with DT in patients with COPD, and this was expressed through an exaggerated increase in gait variability in DT walking. Given the high risk of falls and disability associated with altered gait variability, dual-task training interventions should be considered in pulmonary rehabilitation programs.

KEYWORDS:

• Brain impairment
• cognitive interference
• gait variability
• motor control

Introduction

Many daily activities involve the performance of several tasks simultaneously, such as walking while talking or avoiding obstacles. However, the human capacity for processing information is limited. When cognitive and motor tasks are performed simultaneously, they compete for attentional resources (1) and information-processing neural pathways (2). This conflict may cause a decrease in the performance of one or both tasks, which is called cognitive-motor interference. In elderly people or patients with brain disorders like Alzheimer's disease, Parkinson's disease or stroke, dual-task interference may be inordinately increased when the person is both walking and performing a cognitive task, due to the frontal lobe dysfunction and the associated decrease in attention capacity (3–6).

In patients with COPD, frontal lobe dysfunction with altered attention capacity has been described (7,8), and the ability to perform complex multiple-task activities such as driving or visuo-guided muscle contractions is impaired (9–11). In addition, a recent study reported that patients with COPD take longer to perform a functional test with a cognitive task (12). Although it thus seems very likely that these patients would display a gait deficit during dual-task walking, no study to date has investigated this point. Yet such a deficit would increase the risk of falls and disability (13–15). This being a major clinical issue, the reversibility of this phenomenon should also be investigated. Further, pulmonary rehabilitation might be an appropriate therapy for the dual-task impairment in COPD because of its positive effects on cognitive functions like planning, selective attention and verbal memory (16,17).

The aim of this study was to determine the impact of COPD on cognitive and motor function in dual-task walking. We hypothesized that the alteration in cognitive and gait performances during dual-task walking would be worse in patients with COPD than in healthy controls. In addition, we assessed the effect of pulmonary rehabilitation on dual-task walking performances in a subset of the patients.

Material and methods

Subjects

Twenty-five clinically stable patients with COPD and admitted to a pulmonary rehabilitation program and 20 age- and sex-matched healthy controls were included in this prospective controlled study. The patients had been diagnosed with COPD according to the Global Initiative for Chronic Obstructive Lung Disease criteria (Stages II to IV) and some were receiving long-term oxygen therapy. The healthy subjects were nonsmoking volunteers with no significant medical history. The non-inclusion criteria were the inability to give written consent, the inability to perform the experimental maneuvers, impaired visual function, a neurological, cognitive or psychiatric medical history, foot pain and/or injury, a knee and/or hip prosthesis, and the need for a walking aid (i.e., a cane or walker). All subjects gave written consent. The procedures were approved by the Internal Ethics Committee (Research Committee, 5 Santé Group, France) and complied with the principles of the Declaration of Helsinki for human experimentation. The study was registered at www.clinicaltrials.gov as NCT02493790.

Study design

Subject characteristics

Before taking part in the study, all subjects underwent a medical examination, including the evaluation of body composition, anthropometric and clinical parameters, and resting pulmonary function tests. All underwent plethysmography (V6200 Autobox, Sensormedics Corp., Yorba Linda, CA, USA), with the measurement of forced vital capacity (FVC) and forced expiration volume in 1 second (FEV₁). The presence of persistent airflow obstruction and thus COPD was defined by a post-bronchodilatator FEV₁/FVC ratio <70%. Body composition was evaluated with a TANITA DC-430 S MA and provided information on weight and the fat-free mass index for each subject. We also tested frontal lobe function with the Frontal Assessment Battery (FAB). The FAB assesses the six main cognitive domains of the frontal lobe involving executive functions and attention: conceptualization, mental flexibility, motor programming, sensitivity to interference, inhibitory control and environmental autonomy (18). A score ≤16 was used to determine frontal lobe dysfunction (19). Two healthy controls exhibiting a score ≤16 were excluded from the study sample.

Cognitive and gait tasks

The protocol comprised a series of evaluations of cognitive and motor performances in single and dual tasks. In the single-task (ST) condition, the subjects performed the cognitive and gait tests separately. In the dual-task (DT) condition, they were instructed to perform the two tasks simultaneously (i.e., cognitive-motor dual-task) without giving priority to a task. The ST and DT conditions were randomly performed. To minimize the methodological bias related to the experimenter, the same evaluator tested and gave the same instructions to each subject. Each subject performed each task twice: once for familiarization and once for the evaluation.

Cognitive task

The subjects were seated in a quiet space and asked to count backwards by 3 s from 100 out loud. The evaluator monitored the answers without giving feedback. During the ST condition, the cognitive task was performed over 25 seconds to be closer to the duration of the gait task. During the DT condition, the cognitive task lasted until the gait test was finished. To take into account both correct and incorrect answers, the cognitive performance was assessed by calculating the rate of correct responses per second (RCR) as follows (20): $$\begin{array}{ccc}\hfill RCR& =& \left(\frac{\mathit{Number}\mathit{of}\mathit{correct}\mathit{answers}}{\mathit{Total}\mathit{time}\mathit{of}\mathit{exercise}}\right)\hfill \\ & & \times \left(\frac{\mathit{Number}\mathit{of}\mathit{correct}\mathit{answers}}{\mathit{Total}\mathit{number}\mathit{of}\mathit{given}\mathit{answers}}\right)\hfill \end{array}$$

Gait task

The subjects were asked to walk at their own pace for 15 m (20). The task was performed in a quiet space along a flat corridor. No encouragement was given during the test. Numerical and graphical gait parameters were collected using a Locomètre (Locomètre® – Satel, Blagnac, France). The locomètre is a device developed for human gait analysis to characterize the spatiotemporal parameters of gait in the longitudinal dimension. Each foot movement along a predefined walking perimeter is transmitted via an inextensible thread to an optic probe. The thread is kept under tension by an electrical motor sending out a mechanical recall force that is maintained constant by a micro-computer system. Analysis of the recordings provides quantitative data to determine average velocity, cadence and stride length. Gait performance in the ST and DT conditions was assessed through mean gait velocity and stride time variability (20), with stride time variability corresponding to the coefficient of variation in the gait cycle duration.

Rehabilitation program

To evaluate the impact of pulmonary rehabilitation on DT walking performance in COPD, a subsample of patients (n = 10) performed the same series of evaluations at the end of a pulmonary rehabilitation program. The rehabilitation program followed the latest recommendations from the American Thoracic Society/European Respiratory Society (21). It lasted 5 weeks and was composed of exercise training at the first ventilatory threshold (aerobic exercise), muscle strengthening, and education sessions (including talks on nutrition, stress and general disease management).

Data analysis and statistics

All statistical analyses were performed using Statistica software (StatSoft, Inc., version 13.0, Tulsa, OK, USA). Normal Gaussian distributions of the data were verified using the Shapiro–Wilk test. Differences in subject characteristics were assessed using an unpaired Student's t-test or a Mann–Whitney test (if non-normality was observed). To assess the differences between ST and DT and patients and controls, a two-way analysis of variance (ANOVA) was performed for each dependent variable with group as the between-subject factor (COPD and controls) and condition (ST and DT) as the within-subject factor. To determine the implication of frontal lobe dysfunction in the dual-task interference, supplementary analyses of covariance (ANCOVA) with adjustment for the FAB score were conducted when the ANOVA F interaction was significant. To assess the effect of rehabilitation on DT performance, a two-way analysis of variance (ANOVA) was used for each dependent variable with condition (ST and DT) and time of rehabilitation (pre and post) as the within-subject factors. The underlying assumptions of ANOVA were checked using a Levene test (homogeneity of variance) and a Mauchly test (sphericity of variance). When the ANOVA F ratio was significant (p < 0.05), the means were compared with an LSD post hoc test. Data are reported as means and standard deviation (SD).

Results

Subject characteristics

The subject characteristics are given in Table 1. No differences between patients and controls were observed for age, anthropometric measures or FAB scores (p > 0.05). According to the FAB cut-off score, 9 patients (36%) had frontal lobe dysfunction. Patients with COPD presented a significantly lower FEV₁/FVC ratio and FEV₁ (expressed in liters and percentage of predicted value, all p < 0.001) than healthy controls. Four patients were on long-term oxygen therapy (LTOT) due to chronic respiratory failure. None of them was in the subsample of patients with post-pulmonary rehabilitation data.

Table 1. Characteristics of the subjects included in the study.

	COPD (n = 25)	Healthy (n = 20)	p Value
Age (years)	64.6 ± 7.4	69.3 ± 6.9	ns (0.06)
Height (cm)	164.6 ± 8.6	165.2 ± 10	ns (0.87)
Weight (kg)	70.9 ± 22.8	71.4 ± 16.3	ns (0.51)
BMI (kg.m^−2)	25.8 ± 6.3	25.7 ± 4	ns (0.74)
FFMI (kg.m^−2)	17.8 ± 2.4	17.9 ± 2.6	ns (0.66)
FEV1 (L)	1.21 ± 0.5	2.5 ± 0.7	< 0.001
FEV1 (%pred)	47.9 ± 15.5	111.6 ± 24.1	< 0.001
FEV1/FVC (%pred)	51.3 ± 10.4	98.9 ± 12.4	< 0.001
FAB score	16.1 ± 2	17.4 ± 0.8	ns (.08)

BMI: body mass index; FFMI: fat-free mass index; FEV₁: forced expiratory volume in 1 second; FVC: forced vital capacity; FAB: frontal assessment battery; %pred: percentage of the predicted value; ns: no significant difference between healthy control and COPD patients. Values are mean ± SD.

Cognitive performance during single-task and dual-task walking

The results are presented in Figure 1A. The RCR did not differ between groups in either the ST or DT condition (F_interaction = 0.33, p = 0.57; F_group = 2.01; p = 0.16). However, RCR was significantly lower in DT than ST for both patients and controls (F_condition = 15.1, p < 0.001).

Figure 1. (A) cognitive performance (rate of correct responses expressed in number per second), (B) gait speed (in km.h⁻¹) and (C) stride time variability (in %) in single-task and dual-task walking, in patients with COPD and healthy controls. ^***: different from single-tasking (p < 0.001). ^###: different from controls (p < 0.001). Error bars represent SD.

Gait performance during single-task and dual-task walking

Gait speed (Figure 1B) did not differ between groups in either ST or DT (F_interaction = 1.58, p = 0.22; F_group = 1.39; p = 0.25). However, it was significantly decreased in DT compared with ST for both groups (F_condition = 88.9, p < 0.001).

The interaction for stride time variability was significant (F_interaction = 15.03, p < 0.001). Indeed, while stride time variability was comparable in ST for patients and controls (p = 0.33), it was significantly higher in DT for patients (p < 0.001). In addition, it was increased only for patients in DT compared with ST (p < 0.001). When stride time variability was adjusted for the FAB score, the ANOVA F interaction ratio and the aforementioned post hoc results remained significant (F_interaction = 9.69, p < 0.01).

It should be noted that strictly similar results were obtained by retaining only the patients without chronic respiratory failure in the analyses (gait speed F_condition = 80.7, p < 0.001; stride time variability F_interaction = 11.05, p = 0.002).

Effects of pulmonary rehabilitation (n = 10)

On cognitive performance

The RCR was significantly modified by pulmonary rehabilitation (F_interaction = 11.2, p = 0.004; Figure 2A). Indeed, after rehabilitation, ST performance was significantly improved (p = 0.005). However, there was no significant difference between DT before and DT after rehabilitation (p = 0.44).

Figure 2. (A) cognitive performance (rate of correct responses expressed in number per second), B) gait speed (in km.h⁻¹) and (C) stride time variability (in %) in single-task and dual-task walking, in patients with COPD before and after pulmonary rehabilitation. ^** and ^***: different from single-tasking (p < 0.01 and 0.001, respectively). ^§§: different from “before PR” (p < 0.01). Error bars represent SD.

On gait performance (n = 10)

Gait speed was not modified by pulmonary rehabilitation (Figure 2B, F_interaction = 0.41, p = 0.54; F_time = 2.09, p = 0.18). Moreover, it exhibited a comparable decrease from ST to DT before and after pulmonary rehabilitation (F_condition = 22.56, p = 0.001).

Similarly, there was no effect of pulmonary rehabilitation on stride time variability (Figure 2C, F_interaction = 0.03, p = 0.87; F_time = 0.07, p = 0.79). In addition, the stride time variability exhibited a comparable increase in DT compared with ST both before and after rehabilitation (F_condition = 15.75, p = 0.003).

Discussion

The present study assessed the impact of dual-task walking on cognitive and gait performances in patients with COPD compared with healthy controls. Although the impact on cognitive performance and gait speed was comparable in the patients and controls, stride time variability was greater in the patients. For the subset of patients enrolled in a pulmonary rehabilitation program, the cognitive and gait performances in dual-task walking did not improve after rehabilitation, although cognitive function improved in the single-task.

As expected, the dual-task paradigm decreased the performance of the cognitive and gait parameters in both patients with COPD and controls. However, stride time variability, the most common temporal indicator of gait variability (22,23), increased only in the patients, providing evidence of their impaired neural control during dual-task walking (24). Our results are consistent with those of Sheridan et al., who reported that gait variability is the most greatly impaired gait characteristic in patients with dual-task deficit (25). Our finding of higher stride time variability in the patients supports our hypothesis that COPD is associated with greater interference during dual-task walking, with a specific impact on gait variability. Furthermore, this result is consistent with previous results from our team reporting motor control abnormalities during visuo-guided quadriceps contractions in this population (9). These results have major implications for COPD care management, as gait variability during dual-task activities has been closely associated with the increased risk of falls and disability in other populations (13–15).

We found increased cognitive-motor interference (higher gait variability) in the COPD patients during dual-task walking regardless of the frontal lobe dysfunction, which was assessed here by the FAB, a highly sensitive and reliable tool for assessing the major functions of the frontal lobe including attention, memory and executive functions (18,26). The level of cognitive interference has typically been associated with these capacities, with attention being the most important factor (5). In addition, a recent study reported an association between cognitive-motor interference and FAB score in patients with mild cognitive impairment or Alzheimer's disease (27). In this study, the FAB scores were not different between COPD patients and healthy controls. Moreover, the higher cognitive-motor interference in the patients persisted after adjusting for the FAB score. Therefore, these data suggest that the FAB may not be appropriate for detecting the attentional deficit indicated by impaired dual-task performance in COPD.

The increased gait variability in the patients with COPD was observed only in the dual-task, although previous studies have also indicated increased gait variability during single-task walking (3-minute and 6-minute walking tests) in these patients (28–29). However, for walking distances of 15 m or less, as in our study, the results have been inconsistent, with some authors reporting increased gait variability (12) and others not (30). This discrepancy might be explained by the greater disease severity or the higher prevalence of chronic respiratory failure (1/3 of the study population) in the study of Morlino et al. (12), as these factors are known to be associated with higher gait impairment in COPD (31). In our study, only four patients had chronic respiratory failure. Importantly, we observed the increased gait variability in dual-task walking even after these patients were removed from the analyses. Therefore, these results provide evidence that, even in less severe COPD patients free from chronic respiratory failure, higher gait variability may be displayed during dual-task activities, although not in a simple walking task.

On a subset of patients (n = 10), we investigated the effects of pulmonary rehabilitation on cognitive and gait performance during single- and dual-task walking. Although there was no difference in the subtraction scores of patients and controls at baseline, the patients' subtraction scores improved after rehabilitation. This result is consistent with the reports of previous studies that cognitive function improved in COPD patients after pulmonary rehabilitation (16,17). Although cognitive training is not usually part of pulmonary rehabilitation programs (32), this finding might be explained by the well-known positive effects of physical activity on cognition (33). However, these beneficial effects were not obtained in the dual-task condition, indicating that the improvement in single-task performance was not sufficient to increase cognitive performance in the dual-task. Furthermore, it has been shown that performance in dual-task walking is associated with patterns of brain activation (different areas and/or different activation levels) that differ from those observed for single-task performance (34,35). Our results provide further indirect evidence that single and dual tasks involve different mechanisms, which classic pulmonary rehabilitation programs do not seem to fully address. Moreover, similarly to cognitive performance, gait speed and stride time variability did not improve during dual-tasking after pulmonary rehabilitation. Taken together, these data support the view that the current recommendations for pulmonary rehabilitation care do not take into account the pathways involved in the increased cost of dual-tasks in COPD. Specific interventions should be added to pulmonary rehabilitation (such as dual-task training (36)) to address the reversibility of the impaired gait variability during dual-task walking.

Limitations

We found no differences in the mean FAB scores between COPD patients and healthy controls, the FAB being a sensitive tool to detect frontal lobe dysfunction. Although frontal lobe dysfunction has been clearly established in patients with COPD (7,8), the incidence of cognitive impairment varies in studies from 12% to 88% (37). Using the FAB cut-off (19), 36% of the patients included in our study showed evidence of cognitive impairment. Thus, the prevalence of cognitive impairment in our study population was within the range in the literature and supported the good reliability of the FAB demonstrated elsewhere (18,26).

The effects of pulmonary rehabilitation on dual-task walking were only assessed in a subset of 10 patients. Although we did not observe any significant effects, the statistical power of the ANOVA interaction for rate of cognitive responses, gait speed and stride time variability were all below 10%. Thus, this finding that classic pulmonary rehabilitation programs have no effect of on dual-task walking performance should be cautiously interpreted.

Conclusion

This study addressed the effects of COPD on cognitive and motor dual-task performances. The major result was that an arithmetic task during walking deteriorated the neural control of walking in the COPD patients. Indeed, the patients exhibited an exaggerated cognitive interference during dual-task walking, characterized by an increase in gait variability, a parameter closely related to fall risk and fall history. In a subset of patients, we found that a classic pulmonary rehabilitation program did not reduce the dual-task-induced interference, despite improved cognitive function in the single task. These results should be tested in larger cohort, and the addition of specific dual-task training should be considered for pulmonary rehabilitation programs to manage the increased gait variability in COPD and prevent the potential associated risk of falls.

Declaration of interest statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Acknowledgments

The authors thank Dr Pierre-Louis Bernard for the loan of the Locomètre®. The authors also thank Mrs Catherine Stott-Carmeni for English revision of the manuscript.