**The E**ff**ect of a Secondary Task on Kinematics during Turning in Parkinson's Disease with Mild to Moderate Impairment**

#### **Francesca Nardello, Emanuele Bertoli, Federica Bombieri, Matteo Bertucco \* and Andrea Monte**

Department of Neurosciences, Biomedicine and Movement Sciences University of Verona, 43, 37131 Verona, Italy; francesca.nardello@univr.it (F.N.); emanuele.bertoli@studenti.univr.it (E.B.); federica.bombieri@univr.it (F.B.); andrea.monte@univr.it (A.M.)

**\*** Correspondence: matteo.bertucco@univr.it; Tel.: +39-045-8425131

Received: 10 July 2020; Accepted: 29 July 2020; Published: 3 August 2020

**Abstract:** Patients with Parkinson's disease (PD) show typical gait asymmetries. These peculiar motor impairments are exacerbated by added cognitive and/or mechanical loading. However, there is scarce literature that chains these two stimuli. The aim of this study was to investigate the combined effects of a dual task (cognitive task) and turning (mechanical task) on the spatiotemporal parameters in mild to moderate PD. Participants (nine patients with PD and nine controls (CRs)) were evaluated while walking at their self-selected pace without a secondary task (single task), and while repeating the days of the week backwards (dual task) along a straight direction and a 60◦ and 120◦ turn. As speculated, in single tasking, PD patients preferred to walk with a shorter stride length (*p* < 0.05) but similar timing parameters, compared to the CR group; in dual tasking, both groups walked slower with shorter strides. As the turn angle increased, the speed will be reduced (*p* < 0.001), whereas the ground–foot contact will become greater (*p* < 0.001) in all the participants. We showed that the combination of a simple cognitive task and a mechanical task (especially at larger angles) could represent an important training stimulus in PD at the early stages of the pathology.

**Keywords:** Parkinson's disease; gait; dual task; turning; kinematics

#### **1. Introduction**

Gait functionality is accomplished with good health, whereas locomotion disturbances are associated with an advancing age or with the presence of pathologies. Within 3 years of diagnosis, over 85% of people with Parkinson's disease (PD) develop gait deteriorations [1,2], as well as a reduced walking symmetry [2–4]. Such motor dysfunctions could be exacerbated under various conditions, such as the dual task performance and the changing of the walking direction (i.e., turning).

Specifically, the execution of the two tasks simultaneously ("concurrent performance") presents a challenge for PD patients because of their disabled lower-level spinal centers and basal nuclei [5–7]. Studies have shown that, when compared to controls, subjects with PD show greater reductions in their stride length [8,9] and, therefore, walking speed, as well as an increased variability of the kinematic parameters [10–16]. These patterns are aggravated with the severity of the pathology [17,18] and with the complexity of the concurrent task [5,13,19,20]. Since these findings strengthen the idea that the cognitive function could contribute to gait regulation, it could be helpful to understand the effect of different cognitive tasks, even very simple ones, in this kind of pathology [21–23]. This knowledge could provide an insight on what is the best training stimulus to be administered in those patients.

Concerning the walking direction, it has been shown that subjects with PD report turning difficulties, which are associated with an augmented risk of falling [24–27]. These impairments are due to the central nervous system's involvement in body re-orientation when travelling in the new direction.

From a kinematical point of view, patients with PD highlighted a smaller step width [9,28], a reduction in the step length and a decrease in their walking velocity during a change of direction of 45◦ in the stepping tasks [29]. If the turning angle increased (i.e., to 90◦), patients approached turns with a slower step length and by performing the turn with a larger number of steps [30]. Walking with an auditory cue reduced the gait-timing variability, as well as the step length, and increased the radius of gyration during a turn of 180◦ [31]. Moreover, recent literature has shown that subjects with PD spent more time during the turn (between 30◦ and 180◦) and exhibit a reduction in the walking stability compared to controls [32]. Turning 360◦ in place seemed to be a more compromised condition when compared to turning while walking, with similar impairments with and without episodes of freezing [33,34]. As in the case of the cognitive additional demand, the turning conditions could also represent an important stimulus in such a population. Hence, investigating different turning conditions could provide important information about the optimal training strategy in people with PD.

To date, a lingering question in the literature regards the combination of mechanical and cognitive perturbations. Indeed, few studies have combined the dual task condition with the mechanical perturbations (e.g., change of direction) [35–37]. Furthermore, previous studies usually used turn angles larger than 90◦, which represents per se an important perturbation for PD patients [30].

Therefore, we used the most common turning angles during daily life (60◦ and 120◦) [38], in combination with a simple cognitive task (repeating the days of the week backwards), to better understand the effects of these perturbations on gait kinematics. This combined approach has involved participants with mild to moderate PD in order to provide additional information on how the typical Parkinson's disease walking pattern is modified under simultaneous cognitive and mechanical loads.

Hence, the primary aim was to investigate if the presence of a "dual task" condition can further alter the kinematics (i.e., spatiotemporal parameters) of walking, or, on the contrary, if it can represent a good training stimulus in mild and moderate Parkinson's disease. This cognitive task was studied during forward/linear walking and during turning.

Based on the previously mentioned literature, we hypothesized that the simple cognitive and the mechanical tasks will exhibit similar effects in both populations (patients and controls), whereas the combination of the two stimuli will show a higher impact on PD patients.

#### **2. Materials and Methods**

#### *2.1. Participants*

This study has a cross-sectional, analytical, observational design. Nine patients with mild to moderate PD (3 women and 6 men; mean ± SD, 68.2 ± 5.95 years old, 74.2 ± 11.8 kg body mass, 1.70 <sup>±</sup> 0.10 m height, and 25.6 <sup>±</sup> 3.07 kg·m−<sup>2</sup> body mass index (BMI)) and nine healthy age-matched controls (4 women and 5 men; 67.2 <sup>±</sup> 3.45 years old, 72.6 <sup>±</sup> 13.1 kg, 1.68 <sup>±</sup> 0.08 m, and 25.7 <sup>±</sup> 3.43 kg·m<sup>−</sup>2) were recruited in this study.

All participants received written and oral instructions before the study and gave their written informed consent for the experimental procedure. The study was conducted in accordance with the Declaration of Helsinki, and the experimental protocol was approved by the Institutional Review Board of the Department of Neurosciences, Biomedicine and Movement Sciences, University of Verona (protocol number 2018-UNVRCLE-0451799).

Participants were recruited from a sample of late adulthood people attending the adapted physical activity program at the School of Sport and Exercise Science at our University (see their characteristics in Table 1).


 **1.**Characteristics of subjects with Parkinson's disease.

**Table**

Patients were excluded if their medical condition proved unstable due to other neurological, orthopedic, metabolic or cardiovascular co-morbidity factors affecting gait. There was no type of rehabilitation in the month prior to recruitment or disease-modifying therapy that was not well defined. A diagnosis of idiopathic PD was carried out by a neurologist, in accordance with the guidelines established by the London Brain Bank [39]. The disease severity was classified according to the modified Hoehn and Yahr scale (H&Y) [40], whereas the assessment of the degree of motor and functional impairment was obtained using part III of the Unified Parkinson's Disease Rating Scale (UPDRS) [41]. Finally, cognitive function was assessed by using the Mini Mental State Examination (MMSE), with scores of up to 30 (higher scores correspond to greater cognitive function).

We conducted an a priori screening to evaluate whether participants were able to carry out the task instructions required for the experimental protocol (see the "Experimental procedures" later). Based on this a priori screening, only individuals with PD with an MMSE score of 24 or higher were able to perform the double stimuli, and therefore were recruited in the study.

#### *2.2. Experimental Procedures*

Tests were conducted in the Biomechanics Laboratory at the Department of Neurosciences, Biomedicine and Movement Sciences. The spatial (distance) and temporal (time) characteristics of the step pattern were measured using an eight-camera motion capture system (MX Ultranet, VICON, Oxfordshire, UK), sampling at 100 Hz. This apparatus recorded the position of markers positioned bilaterally on the feet: the calcaneus, lateral malleolus and 5th metatarsal-phalangeal joint.

Participants started from a standing position and walked barefoot at their self-selected speed over a 10 m walkway (Figure 1, panel a). Each participant performed three different conditions: *(i)* forward walking (WF (Figure 1, panel b)); *(ii)* walking, turning at 60◦, walking (T60); *(iii)* walking, turning at 120◦, walking (T120). Furthermore, T60 and T120 were conducted turning in both directions (left and right (Figure 1, panels c and d)). During all turning tasks, a cone was positioned in the center of the walkway (see the black hexagon in Figure 1) to identify the turning point. Furthermore, an operator positioned himself at the end of the corridor lane that the subjects had to move. All directions were tested in single and dual tasks, and the cognitive load consists of walking while repeating the days of the week backwards [17]. Six trials of each condition were performed and the order of gait conditions was randomly allocated (Figure 2). Subjects sat and rested in a chair for three minutes between trials. Patients were tested at the peak dose in the medication cycle.

**Figure 1.** Schematic representation of walking conditions (**a**), and steps during walking (forward walking (WF), (**b**) and turning (T60 in (**c**) and T120 in (**d**)).

**Figure 2.** Flow chart of the research design.

#### *2.3. Data Reduction*

Kinematic data were recorded from 3 m before and after the turning step. The quantitative gait assessment included both temporal and spatial parameters (Figure 3), that are: *(i)* stance time (s), the period of time when the foot is in contact with the ground; *(ii)* swing time (s), the period of time when the foot is not in contact with the ground; *(iii)* stride time (s), the interval of time to complete a gait stride; *(iv)* cadence (stride·min<sup>−</sup>1), the number of strides taken in a unit of time; *(v)* stride length (m), the distance from the initial contact of one foot to the following initial contact of the same foot. All these parameters were obtained using standard definitions according to an algorithm programmed in LabView (version 10, National Instruments, Austin, TX, USA). Walking speed (m·s<sup>−</sup>1) was appreciated as the distance travelled during a complete stride cycle.

**Figure 3.** Gait parameter assessment for walking/turning conditions.

The average value of all the performed steps in each condition has been utilized in further analyses. The number of errors (i.e., when telling a day backwards went wrong) was counted and collected by an operator during the walking trials as well. For controls, we pooled together the data from the right and left lower limbs (*n* = 9), whereas for PD patients, the data referring to the less affected side (PDNA) and to the more affected side (PDA) were considered separately (*n* = 9).

Finally, we calculated the locomotor rehabilitation index "LRI" as: LRI = 100 \* (self-selected speed (SSWS)/optimal speed (OWS)), where the self-selected speed (SSWS) has been directly measured and the optimal speed (OWS) has been estimated according to a previous study [42].

#### *2.4. Data Analysis*

The data analysis was conducted using SPSS 19.0 (SPSS Inc, Chicago, IL, USA). Descriptive statistics were used to compute the means and standard deviations for the outcome variables. All the spatiotemporal gait data were normally distributed (Kolmogorov–Smirnov and Shapiro–Wilk tests) and did not violate the assumptions of homogeneity. In order to test the main hypotheses, a series of 2 (tasks: single task vs. dual task) × 3 (gait directions: WF vs. T60 vs. T120) repeated measures ANOVAs, with groups (CR, PDNA, PDA) as the between factors, were used to analyze the gait data. When a significant main effect was found (critical *p* value < 0.05), a post-hoc *t*-test was performed. A Bonferroni correction was applied when needed.

#### **3. Results**

#### *3.1. Cognitive Task*

For the MMSE, controls (28.1 ± 1.66) showed similar scores to patients (27.2 ± 1.85 (p = ns)).

The majority of the dual task trails were successful. Whereas individuals with PD performed the cognitive task with 97.2 ± 6%, 94.9 ± 8% and 95.2 ± 10% accuracy during WF, T60 and T120, respectively, and controls performed the same trails with an accuracy of 96.7 ± 5%, 95.3 ± 6% and 94.9 ± 9%. Regarding the gait direction, more correct answers were given while walking forward as compared to turning (*p* < 0.01 for T60 and T120) similarly in both groups.

#### *3.2. Spatiotemporal Parameters*

There were no differences between groups for age, body mass and height, or body mass index.

The average ± SD values of the measured temporal and spatial variables have been reported in Tables 2 and 3, as well as the post-hoc results.

The stance time differed among tasks (F (1, 24) = 34.456, *p* < 0.001) and walking direction (F (2, 48) = 27.234, *p* < 0.001), whereas there were no differences between groups (F (2, 24) = 0.262, p = ns). In particular, the dual task showed an increased stance time compared to the single one. The highest stance phase was observed during turning at 120◦, while the lowest was during walking in the straight direction.

The swing phase showed a significant effect for direction (F (2, 48) = 4.648, *p* < 0.05), but not for task (F (1, 24) = 3.745, p = ns) and group (F (2, 24) = 0.883, p = ns). Particularly, the 120◦ turn took a longer swing time in comparison to the other directions.

The cycle time showed significant differences among tasks (F (1, 24) = 31.840, *p* < 0.001) and directions (F (2, 48) = 34.227, *p* < 0.001), whereas there were no group changes (F (2, 24) = 0.618, p = ns). Both groups increased the cycle time during the dual task condition, and the 120◦ turn took the longest time compared with the other conditions.

Therefore, the step frequency and cadence changed significantly with regard to task (F (1, 24) = 30.257 and 30.317, *p* < 0.001) and direction (F (2, 48) = 31.329 and 31.441, *p* < 0.001), but not with group (F (2, 24) = 0.458 and 0.457, p = ns).


0.88 ± 0.09 b,g,m

> T120\_DT

CR: control group; PDNA: Parkinson's disease not affected side; PDA: Parkinson's disease more affected side. WF: walking forward; T60: turning at 60◦; T120: turning at 120◦. ST: single

task; DT: dual task. a = *p* < 0.05, b = *p* < 0.01, c = *p* < 0.001: significant difference by comparing ST to DT. d = *p* < 0.05, e = *p* < 0.01, f = *p* < 0.001: significant difference by comparing the

WF to T60. g = *p* < 0.05; h = *p* < 0.01, i = *p* < 0.001: significant difference by comparing the T60 to T120. m = *p* < 0.01, n = *p* < 0.001: significant difference by comparing the WF to T120.
