Parkinson's disease (PD) is a progressive and neurodegenerative disorder characterized clinically by bradykinesia, rigidity, tremor, and postural instability. Disability occurs at all stages of the disease and the severity of disability usually increases with disease duration. Patients frequently experience gait disturbance, difficulty in linking movements together smoothly, and episodes of freezing [1].

Of these cardinal motor features, gait disturbance is one of the most disabling characteristics of the disease since it not only reduces the patient's mobility, but also leads to increased fall frequency; the overall quality of life diminishes with aggravating gait disturbance. It is estimated that more than 70% of patients fall during the course of the disease and such events often result in fractures, often leading to another morbidity. To improve the quality of life for patients with PD, many research studies have evaluated the underlying mechanisms that cause gait disturbance and have focused on evaluating different intervention strategies [2].

Patients with PD often walk with a reduced gait speed, shorter stride length, reduced arm swing, and decreased postural stability [12]. Previous study on the gait evaluation of the PD patients suggested that reductions in gait velocity result from diminished stride length in persons with PD, whereas cadence is typically unaffected; the PD patients often develop a strategy to increase dynamic stability known as festinating gait; it is characterized by rapid, hypometric steps that minimize displacement of the center of gravity relative to the base of support and increase double-limb support time [23].

Parkinsonism diseases include the atypical Parkinsonian disorders of multiple system atrophy (MSA), dementia with Lewy bodies (DLB), progressive supranuclear palsy (PSP), and corticobasal syndrome, as well as secondary causes of Parkinsonism [4]. It is a challenge to distinguish PD from other Parkinsonism diseases because they share similar symptoms, despite of differences in etiology and treatment plans. Although the classic PD has the clinical features most commonly associated with Parkinsonism, there is a broader spectrum of diseases represented by a collection of phenotypically-similar neurodegenerative conditions that mimic many of the core features of PD, including gait pattern. Because of the similar clinical gait presentations, it is necessary to consider relevant disorders when diagnosing PD [5].

There is growing interest in having an objective assessment of these diseases using technology-based devices capable of providing unbiased measurements in clinical practice. Treadmill gait analysis is an attractive example, as it requires little space and allows for the use of harness support systems; several studies have found that the use of treadmill in training the patients with PD may improve clinically relevant gait parameters such as gait speed and stride length [16].

The gait parameters were often studied and analyzed in order to evaluate the gait patterns of specific disease group or patient at specific age. In the PD patients, Blin et al. [7] analyzed previous gait parameter studies and showed that kinematic gait parameters are different, yet the relationships between the parameters are unchanged. Gait parameters were used to describe the patterns or the effect of specific treatment so far, yet there was no attempt to make comparison between similar, but different diseases so far in our knowledge. Therefore, we hypothesized that the gait parameters are different in the PD and other Parkinsonism diseases, especially the parameters of MSA; the MSA has several subtypes in which MSA-Parkinson type (MSA-P) has the similar balance and gait disturbance shown in the PD.

The aims of this study were to compare the spatiotemporal parameters of gait in PD patients and other Parkinsonism patients, using such gait analysis treadmill platform.

The gait pattern in PD can be described characteristically and has been well studied in previous studies [78]. The steps are short and variable whereas the cadence is mostly normal, except for in later stages with severe akinesia [9]. The step width and foot rotation angles are normal [10]. In later stages of the disease the foot to floor clearance is reduced and the patient begins to shuffle over the ground. In our study, all 3 groups had decreased step lengths, approximately 15 cm, and increased step width, approximately by 5 cm, compared to the values found in control (normal) group in a study conducted by Bello et al. [11].

PD refers to a clinical condition: progressive Parkinsonism of undetermined cause without features suggestive of an alternative diagnosis responding to dopaminergic treatment. Although a diagnosis of PD as defined above can be a straightforward clinical exercise in patients with typical presentation of cardinal signs and excellent response to levodopa treatment, the differential diagnosis versus other forms of Parkinsonism can be challenging, especially early in the disease when the signs and symptoms of the different forms of Parkinsonism have the greatest overlap. The correct diagnosis of PD is important for prognostic and therapeutic reasons and is also essential in clinical research. Investigations of the diagnostic accuracy for PD and other forms of Parkinsonism in community-based samples of patients taking antiparkinsonian medication confirmed a diagnosis of Parkinsonism in only 74% of patients and clinically probable specific PD in only 53% of patients [12]. The gold standard for definite diagnosis for other Parkinsonism diseases, such as MSA, PSP, CBD, and DLB remains to be pathology [13].

For this reason, various researches are actively being conducted to distinguish PD from other types of Parkinsonism. Typically, studies that distinguish PD from other Parkinsonism use imaging techniques [1415]. One of these studies analyzed exhaled volatile organic compounds with a breath test to distinguish between PD and other Parkinsonian syndromes [16]. Still, to the best of our knowledge, no previous studies have used foot pressure analysis, especially the spatiotemporal parameters obtained from the analysis, to compare PD with other Parkinsonism, though one study has tried to compare the gait disturbance in PD and normal pressure hydrocephalus; according to Stolze et al., [17] normal pressure hydrocephalus patients displays hypokinetic gait pattern similar to that of the PD patients, but with enlarged step width and increased step length compared to the PD patients.

The freezing of gait, commonly seen in advanced PD, has been classified as its fifth cardinal feature. However, its presence frequently leads to a misdiagnosis of PD, as the freezing is in fact more common in atypical Parkinsonism [1819]. We examined PD and other Parkinsonism diseases in terms of spatiotemporal parameters from foot pressure analysis. In this study, we found that there was a statistically significant difference in the lateral symmetry during the gait (Table 2, Fig. 2); the median value of the lateral symmetry was higher in the PD group than in the MSA group, indicating that the PD patients tend to show more lateral deviation than the MSA patients. Freezing of gait in the PD is known to be associated with the disease severity [18], yet the PD patients had more lateral deviation during the gait compared to other disease groups. Such difference is may be due to different disease pathophysiologic characters. Due to the dopaminergic pathway dysfunction, the PD develops progressive loss of the decreased ability to internally control the center of mass during self-directed activities [2]. The decreased control leads to deficits in gait initiating, which requires a voluntary shift in the COP from a 2-leg stance to an alternating single-leg stance. Along with such character, the PD often shows unilateral symptom onset on either the dominant or non-dominant of the body [4]; it may also affect the tendency of having more lateral deviation during the gait.

Other gait parameters did not have significant differences among the 3 groups. Although the gait velocity seemed to have certain degree of clinically meaningful difference, the results did not show statistically meaningful difference within 95% CI. It implies that distinguishing PD from other Parkinsonism by the spatiotemporal parameters we obtained from the foot pressure analysis is not reliable for clinical use yet.

There were several limitations in this study. First, there might be differences in the measured parameters between treadmill walking and ground walking. There are several studies had compared the differences in gait kinematic parameters between treadmill walking and ground walking in PD, and they reported that a reduced gait variability and an increased step length were observed during treadmill walking [202122]. Second, there was a relatively small sample size of 3 groups; we could not assume for normal distribution among the groups, which forced our team to use non-parametric tests for data analyses. Third, the Parkinsonism other than PD group was made of heterogeneous diseases, including vascular Parkinsonism and drug-induced Parkinsonism. The heterogeneity in one of the groups might distorted the data analyses. Fourth, the onset of disease symptom duration was unclear for our patients. We evaluated the patients who visited the neurology and rehabilitation departments of the hospital, yet the patients did not have exact memory of the disease onset. Our evaluation was intended for the early stage patients, yet our patient pool might have the patients in various stages. Lastly, our analyses were made based on the first diagnosis given to the patients, which might have been inaccurate; the future study should be conducted with the longer period clinical follow-up along with the gait spatiotemporal parameter studies.

In our knowledge, there is no previous study that have analyzed the gait parameter differences between the PD and the MSA group using foot pressure analysis during the gait. Our study showed possibility that there may be the meaningful difference in terms of the gait parameters in the early stage of the diseases, such as lateral symmetry of the center of the pressure, which have shown a meaningful difference between the PD and the MSA groups.

This study showed the differences in spatiotemporal parameters among PD, Parkinsonism diseases, and MSA using foot pressure analysis during the gait; there was a statically significant difference in lateral symmetry of gait patterns between the PD and the MSA groups in the early stage of the diseases. The PD patients showed more lateral deviation during the gait compared to the MSA group. The data also suggest that there might be other spatiotemporal parameters with meaningful difference among the groups, such as gait velocity. These results support the approach of conducting foot pressure analysis, or other gait analysis programs capable of analyzing the spatiotemporal parameters, to aid distinguishing PD from other Parkinsonism diseases.