Postural instability, bradykinesia, hypokinesia/akinesia, rigidity, and tremors are clinical elements in the diagnosis of Parkinson’s disease (PD) (1). Reduced walking velocity, stride length, cadence, and proportion of swing time to stance time are common gait abnormalities, but motor deficits resulting from significant dopaminergic striatal denervation are present in both the upper and lower limbs (2).

The majority of research to date has been focused on the functioning of the lower extremities, while the movements of the upper and lower extremities are coordinated, any disruptions in inter-limb coordination may affect stability and lead to gait disturbances. As reported, arm swing is an essential component of locomotion, which may influence human gait stability and remain reduced for even several years prior to diagnosis of PD (3–6).

Studies revealed that normal arm swing increases global gait stability and may help the recovery of gait stability following perturbation. Elftman (5) reported that arm swing decreases angular momentum around the vertical and vertical ground reaction moment causing a minimization of energy consumption. In this manner, arm swings generate a horizontal torque to the upper trunk, and through the shoulder girdle, they work as a gravitational pendulum. The result of the functional division of the body into the passenger and locomotor proposed by this author, is that tilting the body becomes an important driving force during gait. Some studies tried to answer the question whether arm swing is actively controlled (driven by muscle activity) or rather passive (7, 8). This idea was confirmed by Hogue (9); however, later studies suggested largely passive swinging of the arms (6, 9, 10). The assumption of a partly passive arm swing arose from the idea of evolution adjustment after a change from quadrupedal to bipedal locomotion.

It has been reported that significant physiological differences between the left and right arm swing amplitudes exist in healthy subjects. The asymmetry can be larger in an arm affected by neurological pathology. Asymmetrical presentation of motor dysfunctions in PD occurs as a result of the asymmetric process of nigrostriatal dopaminergic denervation. Symptoms are more visible on the most affected side and deteriorate with the ongoing disease (5, 11). No compensation of the lesser affected side was observed in comparison to other studies (12). Decreased inter-limb coordination was observed in early stages and in advanced stages of the disease (2, 8, 13). As movements of the upper and lower limbs influence each other, rehabilitation of gait focused on improving amplitude of arm movements may be beneficial.

Under these circumstances, the aim of this study was to analyze the arm, head, and lumbar movement in locomotion patterns, according to the presence of symptoms of patients with PD, on the basis of three-dimensional analysis and to compare the effect of a training program on selected gait parameters.

Comparison of the angular movements of lumbar spine, cervical spine, shoulder joint, and elbow joint in the sagittal plane, during gait in patients with PD, before and after treatment.

Charts regarding the angular movement changes in the lumbar spine, cervical spine, shoulder joint, and elbow joint, before and after treatment in comparison to the control group are shown in Figures 1–4, respectively.

The study showed that during gait of PD patients, tilting of the thorax was clearly smaller than the mean value of control group’s tilting angle (difference of about 8°). Average values of chest tilting in the sagittal plane (Figure 1) did not differ significantly between the most affected side and the less affected side.

The course and setting angle of the curve changes in the cervical spine in patients with PD before treatment was flexion and differed from the control group (p < 0.05, Figure 2).

The significantly smaller range of angular changes in the most affected shoulder joint during the terminal stance (F4) phase of gait registered before therapy differed by about ±6°, compared with the healthy subjects (p < 0.05, Figure 3). Shoulder of the most affected upper limb had a tendency to maintain extended throughout the gait cycle.

The range of movement in the less affected elbow joint prior to therapy was similar to the range of double SD achieved by healthy subjects during every moment of the gait cycle (Figure 4). Angular changes among the analyzed curves did not differ significantly between groups for the elbow joint of the less affected limb and shoulder joint of the most affected limb in the first test (p < 0.05). Other curves for both the most affected and less affected limb differed significantly from the control group’s average (p < 0.001).

Analyzing the average angle of the thorax tilting before and after treatment, significant differences were found. During gait in PD patients, tilting of the thorax was clearly smaller than the mean value of control group’s tilting angle (p < 0.05). After therapeutic treatment, we observed improvement in the angular range of changes in thorax tilting. Average values of chest tilting in the sagittal plane (Figure 1) did not differ significantly between the most affected and less affected sides.

After therapeutic treatment, mean peak angular changes and standardized time referring to collected motion angles in the cervical spine were significantly different from the values before treatment (p < 0.05). Improvement in the results relates to the cervical spine setting and angular changes at different phases of the gait cycle (Figure 2). The curve for the test conducted after therapy concerning the most affected and less affected sides in the whole gait cycle, was close to the average achieved by a healthy person. The difference between the studies in terms of angle changes was about ±3°. A significant reduction in flexion during dual support was observed after therapeutic treatment.

The angular range of changes in the shoulder joint improved after therapeutic treatment, but the observed improvement was not statistically significant. The only significant changes were observed in the less affected shoulder during the initial contact (F1), terminal stance (F4), and terminal stance (F8) phases of gait (p < 0.05).

Significant angle and setting changes in the elbow joint occurred during initial contact and terminal swing phases (F1, F8) for the most affected limb before and after treatment (p < 0.05). In the terminal stance phase (F4), an increase in range of motion of about ±4° was observed after therapeutic treatment. The curve indicates that the nature of the movement was mainly flexion in all the phases of gait in patients with PD.

Angular changes among the analyzed curves did not differ significantly between groups for the shoulder joint of the most affected limb and the elbow joint of the less affected limb in the second test (p < 0.05). Other curves for both the most affected and less affected limb differed significantly from the control group’s average and between the first and second tests (p < 0.001).

Parkinson’s disease leads to motor dysfunctions in gait, speech, or writing. In recent years, particular attention was paid to gait abnormalities and the possibility of their improvement by pharmacological treatment and rehabilitation. The use of new, more accurate measuring devices, such as the Vicon system, allows to record and analyze motion in three-dimensional space (14, 15). Until now, there were few publications (16–18) related to the use of the Vicon motion analysis system in the assessment of therapeutic treatment effectiveness in patients with PD.

No studies were found on the upper extremities movement changes in connection to head and lumbar movements of patients with PD. Arm swing, which is an essential component of locomotion, may influence human gait stability (3–6). As the upper and lower extremities are coordinated, disruption in inter-limb coordination may lead to gait disturbances. Decreased inter-limb coordination can be observed at the early stages as well as advanced stages of the disease (2, 8, 13). Gait rehabilitation including improving the amplitude of arm movements may be beneficial.

Our study showed significantly reduced chest tilting during gait in patients with PD (Figure 1). This was probably associated with lower walking speed and resulted in reduction of body force driving developed from body ground reaction momentum. Such a situation may cause the need for greater muscle involvement in order to provide propulsion of the body. As studies revealed, decreased trunk mobility during walking may be present in individuals with PD (19). During locomotion, changes in coordination patterns have been observed for head and trunk motions. Flexible coupling of thoracolumbar and cervical segments allows locomotor task-dependent use of the upper limbs. This optimal coupling depends on walking speed, as arm swing compensates angular momentum with increasing gait velocity. The vertical moment to counter the torque from torso motion is produced by arm swing and is the only torsional effect (20, 21).

Our results show that the mean values of angular changes in the most affected shoulder joint did not differ significantly between patients and the control group. Results are within the range of double SD (Figure 3), but a greater range of extension was registered for the most affected shoulder joint at the terminal stance (F4) phase of gait. In the most affected elbow joint, statistically significant differences were registered during initial contact (F1), terminal stance (F4), and terminal swing (F8) in comparison to healthy subjects. This demonstrates the lack of balance on the most affected side. Similar parameters related to changes in the angle of the shoulder and elbow joints during gait were assessed by Knutsson (22) and Murray et al. (23), using the method of photokinemetric registration. The authors noted lack of coordination of the upper limbs, oversized shoulder extension, and greater elbow flexion in patients compared with the results obtained in the control group. In our study, following therapeutic treatment, an improvement in the range of shoulder and elbow flexion was observed. This points out the alternate work of upper limbs during gait, which did not exist prior to treatment.

There is no reference, in literature, to the results that we obtained in the present study, in which the impact of gait therapy treatment based on three-dimensional motion analysis on parkinsonian gait was evaluated, analyzing the angular changes corresponding to the recorded movements of the cervical spine. As follows from our analysis of the cervical spine movement curve during gait, limited range of extension prior to the treatment was affirmed. The conducted therapy had positive impact on changes in angular movements of the cervical spine, and the average course of the curve approached the course of the control group’s curve. There has been significant improvement in the obtained average values of reduction in peak flexion (extension) of cervical spine movement during the gait cycle. In control group, they occurred during the mid-stance (F3) phase of gait, and the average value of peak flexion occurred at the final (terminal) phase of gait. In patients with PD, these phases were delayed and occurred in terminal stance (F4) of cervical spine extension while flexion occurred during the initial swing (F6) phase.

Analysis of individual gait cycles of people with PD showed evident improvement in the angle of selected joint motion (24).

Studies by Wells et al. (25) revealed an increase in the range of motion of the upper limbs. In their study, the results of pathological walking patterns occurring prior to treatment improved after treatment and moved closer to the physiological gait pattern. Palmer et al. (26) claimed functional improvement in gait, coordination, and reduction of upper limb tremor after 3 months of therapeutic treatment; one focused on upper limb karate training and the other on trunk stretching exercises. Walking without arm swing is possible but requires greater effort of the legs due to greater foot moment reaction. As Park (27) referred, it is believed that arm swing in human walking minimizes torque loading on the joint and skeletal structure leading to optimization of lower limb motion. Whether motor learning affected by the neurodegenerative process in basal ganglia may be mediated by exercise was always a highlighted question, while studies have shown that rehabilitation positively affects structures “occupied” by disease (28, 29). According to the findings of Nieuwboer et al. (30) limits in motor learning caused by disease, extended rehabilitation, and learning processes are necessary to achieve automatization.

In studies on their own kinematic work patterns of the upper extremity joints, the lumbar and cervical spines are managed to be improved. In PD, the pathological gait pattern is fixed in the first years of the disease. Because of the already advanced changes in the musculoskeletal system, resulting from years of disease, it may be assumed that the duration of the therapeutic program should be longer to maintain favorable kinematic changes in the joint working pattern.

Assessing at exactly which point in time the gait phases occur, the corresponding values of movement ranges in the upper limbs joints, lumbar, and cervical spine in patients with PD rehabilitation program was elaborated taking into account all discussed abnormalities of gait pattern.

Patients with PD were characterized by reduced range of motion in the joints of the lumbar spine, cervical spine, shoulder, and elbow. Exercise therapy slightly increased the range of movement in the examined joints and also resulted in improved locomotion patterns. The results presented in the elaboration of their own research may be an example of very precise assessment of gait abnormalities, used for preparing gait rehabilitation and documenting the result progression after gait therapeutic treatment. It seems that the obtained results should contribute to further research penetration including a more representative group. Future studies should include a follow-up study to show whether the positive changes in locomotion last over a longer period of time.

EM: study design, data collection, data interpretation, manuscript preparation, and literature search. JK: data interpretation, manuscript preparation, and literature search. JS: statistical analysis, data interpretation, manuscript preparation, and literature search. SP: manuscript preparation and literature search. MR: data collection and medical supervision. WC: data collection and statistical analysis.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.