Effects of Gait Rehabilitation With Dual Task in Patients With Parkinson's Disease

BACKGROUND AND OBJECTIVES

In functional environments, dual-tasks (DT) are common and require a correct motor and cognitive performance to be carried out successfully. In people with neurodegenerative diseases such as Parkinson's disease (PD), to walk with a secondary task affects gait causing an alteration of velocity, stride length, cadence and double support time and therefore, greater risk of falls. In fact, patients report gait impairments as the most disabling motor symptoms of the disease. The inclusion of DT to the assessment and physical rehabilitation of patients allows to simulate day-to-day contexts in a controlled and safe environment and consequently, extrapolate more easily the advances of rehabilitation to daily life. Advantages of this kind of training are not only functional, but it could induce neuroplasticity because it is a highly challenging activity for people with PD. While the neurophysiological changes of dual-task training are not yet known, the effects of training with dual-tasks on gait seem relatively clear. Studies so far have shown that improvements of spatiotemporal parameters of DT gait can be observed in integrated and consecutive DT training, which means that the addition of cognitive tasks into gait rehabilitation, either simultaneous or consecutively, is effective. However, there is not enough evidence that DT training is better than traditional EP training, i.e., including only motor training in single-task (ST) mode. The rest of the published studies include a passive control group (i.e. group that does not perform any type of control rehabilitation) or they correspond to repeated measures studies.

Additionally, there are other methodological issues that hinder the external validity of the experiments referenced so far. Firstly, the use of the same DT, both in training and evaluation of these effects training, could prevent to observe the transfers of these effects in untrained contexts. Secondly, the use of complex secondary tasks during the assessment such as mathematical tasks or recite alphabet letters may not be representative of the DT that occur in daily life. Thirdly, the different studies included the follow-up usually one month after the training, only one study evaluated the effects beyond this time. Finally, the lack of training done in a group, it prevents knowing if training with DT is possible in a group format, as well as individually at home.

Thus, this study aims:

1. To compare the short and mid-term effectiveness of a group dual-task program versus a group single-task program on biomechanics gait in people with PD.
2. To analyze the effect of dual-task therapy on clinical gait variables (standardized scales and tests), cognitive performance and quality of life.
3. To analyze the impact of secondary cognitive and motor tasks on single-gait before and after both physiotherapy programs.
4. To compare the parkinsonian gait before and after both therapies with normal gait from matched healthy subjects.

DESIGN

This was a single-blind, Randomized Control Trial (RCT) with 8-week follow-up and a convenience sample (specifically, modal instance sampling) in two parallel-group (experimental and active control group [1:1]) and one matched healthy control group and factorial design study.

BLINDING

Due to the nature of physical rehabilitation interventions, patients cannot be totally blind to the treatment that were performed and neither the treating physiotherapists, but the hypothesis and objectives of the study were hidden. At the same time, all participants were instructed to not divulge information regarding their intervention to the raters physiotherapists, who were blinded to group allocation in all assessment times. To control expectancy effects for patients it was explained that it is not yet determined which therapy is more effective. On the other hand, a code of consecutive numbers were assigned to the patients without pointing to the treatment group to which they belong.

RANDOMIZATION

The randomization process was performed by an extern researcher who is not involved in the development of the study. Stratified randomization will be performed by the severity stages of the disease (Hoehn \& Yahr stages I, II and III) and participants were accordingly allocated to one of the two groups of physiotherapy treatment. Randomization was performed with a Microsoft Excel software with which a random sampling of half of the total of patients were extracted by the strata set for the dual-task program. After this, the treating physiotherapists will be responsible for the management and organization of the appointments of participants.

POWER CALCULATION

The sample size power calculation has been conducted based on the changes in the main outcomes of gait velocity evidenced in the literature in a verbal dual-task after a dual-gait training effect with PD patients. Previously, it has been shown that verbal tasks have great interference on Parkinsonian gait. For the sample size analysis, the Gpower software (Universität Kiel, Germany, Version 3.1) has been used [25]. A type I statistical error of 5% and a power of 80% was considered, which has resulted in 24 participants per group of treatment.

INTERVENTION

To ensure that the therapy sessions were carried out with a maximum of 10 people, the development of each program was repeated twice (DTP and STP), until completing the training of all recruited participants.

Both groups of the study (experimental and active comparator) performed 20 one-hour rehabilitation sessions, twice a week (10 weeks in total). All sessions of both groups were developed in a place provided by the researchers with the necessary elements for rehabilitation and in each of the sessions, there always were two physiotherapists. In each session, there were no more than ten patients.

1. Dual-task group Patients with Parkinson's disease carry out the rehabilitation gait with a dual-task program (DUALGAIT) with secondary cognitive and upper limb motor tasks. In this group, the training of walking and functional secondary tasks (cognitive or motor) first was separately (consecutive task training), to be performed later together in a dual-task (integrated dual-task training) with a progression system. Cognitive/motor secondary tasks were different from those used in the assessment of gait. Each training session consisted of three parts: the initial warm-up, dual-training, and back-to-calm. The objectives to walking improve and exercises were: i) Step/stride length with objective distance from 0.4 a 0.8 m in a progressive way and adapted to the heights of the participants, ii) Cadence/velocity training by digital metronome from 60 to 120 bpm, which in combination with the step lengths allowed the progressive training of walking velocity, iii) Lateral gait with lengths from 0.4 to 0.8 m., iv) Lower limb kinematics in support and swing phase through standardized practice of ankle, knee and hip movements during both phase, v) Arm swing practice during gait through elements, coordination with a partner and free, vi) Control of the flexor postural pattern by the practice of postural exercises and active movement that favor upright posture (neck extension, scapular adduction, shoulder retroversion, pelvic anteversion) and vii) Joint mobility and dissociation scapular-pelvic waist. In all of them, the use of external visual, auditory and verbal cues was included to provide the most objective feedback possible during motor learning. Additionally, the cognitive tasks used during the training were rehearsed in different positions (1st seated, 2nd biped and 3rd static gait) prior to the dual realization with the forward gait. The types of secondary tasks trained were: 1) verbal fluency, 2) auditory recognition, 3) visual recognition, 4) mathematical calculation, 5) memory, 6) visuospatial planning, 7) fine motor skills, 8) motor task of money manipulation, 9) motor transport task and 10) motor task of transferring an object to another person. Each task category had a level of complexity (basic, medium and high).
2. Single-task group Patients with Parkinson's disease carry out the rehabilitation gait without a dual-task program (physical and walking exercises without an additional load of cognitive or upper limb motor tasks). Each training session consisted of three parts: initial warm-up, physical exercise in single-task condition, and back-to-calm. The objectives and walking exercises were the same as those performed in the experimental group.

SUBJECTS

Forty participants completed the study, 23 performed the dual-task physiotherapy program and composed the Dual-task group (DTG), while the other 17 formed the Single-task group (STG). All patients came from the La Fe Clinical and Polytechnic Hospital (Valencia, Spain) completed all phases of the study. The patients were recruited with the diagnosis made by a neurological doctor. All participants signed informed consent prior to their participation in the study, then they were randomly assigned to one of the rehabilitation groups (DTG and STG).

TEST PROCEDURE

Before conducting the biomechanics gait assessments, a previous evaluation was carried out to verify the eligibility criteria. This valuation session was held a few days before the biomechanics gait assessment prior to the intervention and the following was verified:

• Registration of sociodemographic data and neurological history
• Anthropometric evaluation
• Clinical evaluation of Parkinson's disease (Disease duration, Hemibody where signs predominate, Hoehn \& Yahr stage, New Freezing of Gait Questionnaire)
• Cognitive state (Mini-mental test adapted for Parkinson's disease).

After determining the eligibility criteria, three assessments sessions were carried out at the University of Valencia laboratories. These were: 1) immediately before starting the rehabilitation program (Baseline), 2) immediately after completing the rehabilitation program (Postintervention) and 3) eight weeks after completing the rehabilitation program (8-week follow-up). Between Postintervention and 8-week follow-up tests, participants did not perform any kind of physical therapy, sports or physical activity. Participants were evaluated in the on-medication state. In each of the assessment sessions biomechanics of gait, the clinical performance of mobility, balance and gait, and cognition were evaluated.

The biomechanics gait evaluation was carried out in a corridor 10m long and data were registered using 3D photogrammetry with 12 smart cams (Kinescan/IBV software, Biomechanical Institute of Valencia, Valencia, Spain, version 5.3.0.1) and two force platforms (Dinascan/IBV Biomechanical Institute of Valencia, Valencia, Spain). All the participants walked at a self-selected comfortable speed, barefoot and under five conditions: i) single task (ST): walking without secondary tasks with the attention focused only on walking performance, ii) visual dual-task (viDT): walking while watching the time on an analog clock projected at the end of the walkway, iii) verbal dual-task (veDT): walking while telling the evaluator the activities they had performed the previous day in a chronological order, iv) auditory dual-task (aDT): walking while listening and recognizing different daily noises and, v) motor dual-task (mDT): walking while carrying one glass in each hand and repeatedly transferring their contents from one to the other. During DT gait, participants were urged to focus attention on the secondary task. For each of the five evaluated walking conditions, 10 repetitions were performed, five with each foot, to later use their average. Before recording gait, participants were allowed to walk in the corridor (ST condition) to familiarize themselves with the test.

The biomechanics outcomes were:

1. Spatio-temporal outcomes: velocity (m/s), stride length (m), cadence (steps/min), double support time (% of gait cycle) and step width (m).
2. kinematics outcomes: range of motion of the ankle (°), maximum hip extension (°), maximum hip flexion (°).
3. Kinetic outcomes: vertical reaction force 1 (N), vertical reaction force 2 (N), anterior-posterior reaction force 1 (N).

On the other hand, to know if the effects of the treatments would be detectable with the clinical tests and scales, a Dynamic Parkinson Gait Scale (DYPAGS), Tinetti Mobility Test, both total score (TinettiT) and scores from subtest gait (TinettiG) and balance (TinettiB) and Time up and go test (TUG) were used. Finally, we used the Trail making test, parts A (TMTa) and B (TMTb) to evaluate executive cognitive performance and the Parkinson's disease questionnaire-39 (PDQ-39) for quality of life assessment.

STATISTIC ANALYSIS

Statistical analyses were performed using SPSS v.24 (SPSS Inc., Chicago, IL, USA). Mean and standard deviations (SD) were calculated. A three-factor mixed Multivariate Analysis of Variance (MANOVA) was conducted to analyze their effects on the biomechanical dependent variables previously described in table 3. The within-subject factors are rehabilitation with three categories (Baseline, Postintervention and Follow-up) and conditions, with five categories (ST, viDT, veDT, aDT and mDT). The between-subject factor is group with two categories (DTG and STG). Stride length was standardized by the lower limb length of participants. In addition, a two-factor mixed MANOVA was conducted to analyze the effects of within-subject factors rehabilitation and the between-subject factor group on the clinical tests. When significant effects are found, Bonferroni was used for post hoc comparisons. Differences was declared statistically significant if p<0.05.

To check for differences between the demographic outcomes between groups, multivariate one-way analysis with the between-subject factor group was conducted. Furthermore, to test for sex differences between groups, a chi-square test was used.

Before the statistical analysis, the following assumptions were checked: 1) Normality, the distribution of the residuals was tested using the Shapiro-wilk test, 2) Sphericity, the homogeneity between variances of the differences between pairs of measures was checked with the Mauchly sphericity test, 3) Homoscedasticity, homogeneity of variances was evaluated using the Levene test, 4) Equality of covariance matrices, the equality of variance-covariance matrices across the cells formed by the between-subjects effects was tested with the Box's M test, 5) Absence of multicollinearity, the correlation level between dependent variables were observed with the Pearson correlation test. When any of these assumptions were not met, the necessary statistical corrections were used.

TIMING OF THE STUDY

• Resolution Ethics Committee of the University of Valencia (Spain): May 14, 2014.
• Resolution Ethics Committee of the Biomedical Research of La Fe University and Polytechnic Hospital (Valencia, Spain): September 9, 2014.
• Start of recruitment of participants: June 14, 2014.
• Completion of participant evaluations: June 30, 2016.
• Data pretreatment, statistical analysis and dissemination of results: from september 2016 to present.