Action Observation in People with Parkinson’s Disease. A Motor–Cognitive Combined Approach for Motor Rehabilitation. A Preliminary Report

The aim of this study was to assess the role of Action Observation (AO) to improve balance, gait, reduce falls, and to investigate the changes in P300 pattern. Five cognitively intact People with Parkinson’s disease (PwP) were enrolled in this prospective, quasi-experimental study to undergo a rehabilitation program of AO for gait and balance recovery of 60 min, three times a week for four weeks. The statistical analysis showed significant improvements for Unified Parkinson’s Disease Rating Scale (UPDRS) motor section III p = 0.0082, Short form 12-items Healthy Survey (SF-12) Mental Composite Score (MCS) p = 0.0007, Freezing of gait Questionnaire (FOG-Q) p = 0.0030, The 39-items Parkinson’s Disease Questionnaire (PDQ-39) p = 0.100, and for P300ld p = 0.0077. In conclusion, AO reveals to be a safe and feasible paradigm of rehabilitative exercise in cognitively preserved PwP.

Together with pharmacological treatment, intensive physical therapy is used to improve motor performances in Parkinon's disease (PD) patients [1]. Unfortunately, patients with poor motor ability cannot easily take part in physical therapy programs due to motor or cognitive impairment. Therefore, experience-dependent neural plasticity often proves to be a hard task for neurorehabilitation, in means of repair and recovery. However, the systematic observation of significant motor actions closely linked to their execution (action observation treatment (AOT)) has been proposed to be a valuable and feasible practice for motor-impaired patients [2]. As reported by Abruzzese and colleagues, Action Observation (AO) therapy shows its effectiveness in learning or enhancing the quality of execution of specific motor skills [3], and it has been described as an effective cognitive tool for rehabilitation [2,[4][5][6], since it can shape neural circuit reorganization and promote neural plasticity and motor learning [7][8][9][10][11]. Despite this promising evidence, very few studies have been published about the rehabilitation of People with Parkinson's (PwP) using the AO paradigm [3].
Neural activity, reflecting the cognitive load, has usually been measured through electroencephalography recording, specifically by analyzing the event-related potentials (ERPs), which show the synchronous activity of large groups of neurons firing together [12]. Among ERP components, the late positive wave deflection component, known as P300, has been identified as an At the end of each sequence, the therapist encouraged the patient to perform the observed motor action movement over the subsequent 15 min, providing help as needed. The observed action was performed, indicating to patients to repeat it as many times as they could. Each session was repeated twice per day in two separate sessions, at least 60 min apart. All actions were object and goal-directed. The therapist maintained the patient's attention with verbal feedback (Figure 1). performed, indicating to patients to repeat it as many times as they could. Each session was repeated twice per day in two separate sessions, at least 60 min apart. All actions were object and goal-directed. The therapist maintained the patient's attention with verbal feedback (Figure 1). A preliminary descriptive analysis was performed to check the normal distribution of patients' clinical and instrumental data using Shapiro-Wilk test. Whereby the variables collected presented a normal distribution, parametric statistic tests were used. Non-parametric tests were instead used where the normality distribution criteria wasn't satisfied. Specifically, the Wilcoxon signed-ranks test was performed to determine the location of any significant differences between time points. The alpha level for significance were set at p < 0.05 for the first level of analysis. The distribution of the study sample (n = 5) by age and main clinical and demographical characteristics are reported in Table 1. The study didn't have any dropouts during the treatment and all subjects enrolled were able to complete the program. Moreover, no issues have been registered during the execution of the AO program, since every precaution required in a structured rehabilitation facility was taken. In Table 1, the descriptions of all the outcome variables are presented, both measured at baseline T0 (N = 5) and post-treatment T1 (N = 5) ( Table 1). The Wilcoxon test analysis showed statistically significant improvements for UPDRS motor section III p = 0.0082, SF-12 MCS p = 0.0007, FOG-Q p = 0.0030, PDQ-39 p = 0.100, and for P300LD p = 0.0077, results are also shown in Table 1. No statistically significant improvements for other scales were found. Furthermore, a correlation analysis using Spearman's r was performed in order to enlighten the exact role of the P300 change over clinical outcomes. The results did show strong coefficient value of Spearman r between ΔT1-T0 change of clinical outcomes (BBS, SFM-12 MCS, PDQ-39, and FIM scores), and the latency duration ΔT1-T0 change of P300. Despite the high strength value of correlation, no statistical significance was found for any of the reported outcomes, primarily due to the limited sample size included in this study. A preliminary descriptive analysis was performed to check the normal distribution of patients' clinical and instrumental data using Shapiro-Wilk test. Whereby the variables collected presented a normal distribution, parametric statistic tests were used. Non-parametric tests were instead used where the normality distribution criteria wasn't satisfied. Specifically, the Wilcoxon signed-ranks test was performed to determine the location of any significant differences between time points. The alpha level for significance were set at p < 0.05 for the first level of analysis.
The distribution of the study sample (n = 5) by age and main clinical and demographical characteristics are reported in Table 1. The study didn't have any dropouts during the treatment and all subjects enrolled were able to complete the program. Moreover, no issues have been registered during the execution of the AO program, since every precaution required in a structured rehabilitation facility was taken. In Table 1, the descriptions of all the outcome variables are presented, both measured at baseline T0 (N = 5) and post-treatment T1 (N = 5) ( Table 1). The Wilcoxon test analysis showed statistically significant improvements for UPDRS motor section III p = 0.0082, SF-12 MCS p = 0.0007, FOG-Q p = 0.0030, PDQ-39 p = 0.100, and for P300LD p = 0.0077, results are also shown in Table 1.
No statistically significant improvements for other scales were found. Furthermore, a correlation analysis using Spearman's r was performed in order to enlighten the exact role of the P300 change over clinical outcomes. The results did show strong coefficient value of Spearman r between ∆ T1-T0 change of clinical outcomes (BBS, SFM-12 MCS, PDQ-39, and FIM scores), and the latency duration ∆ T1-T0 change of P300. Despite the high strength value of correlation, no statistical significance was found for any of the reported outcomes, primarily due to the limited sample size included in this study. In the present paper, a non-invasive rehabilitation approach based on selected observation of action and imitation has been proposed in order to improve motor and cognitive performance [18]. To our knowledge, so far only two studies have shown a positive effect of Action Observation Therapy in Parkinson disease. In the first study, Buccino and co-workers have demonstrated that these original approaches, in addition to conventional rehabilitation, can significantly improve dexterity and independence in daily activities in PD [19]. In particular, the study investigated the effectiveness of rehabilitative treatment with AO in 15 subjects with PD. Individuals improved significantly more than controls in the UPDRS score and the Functional Independence Measure (FIM) scale [3,19]. In the second study, Pelosin and colleagues have shown positive additional effects on walking ability recovery and freezing of gait reduction; in particular, the authors investigated whether AO, combined with practicing the observed actions, was able to reduce the frequency Freezing of Gait (FOG) episodes in PD [20]. Our results are in accordance with Pelosin and colleagues about the recovery of motor performances (UPDRS III), SF-12 MCS, FOG-Q, and PDQ-39 [3,20]. The novelty of our protocol is related to the introduction of P300 assessment for the detection of changes in brain activity of subjects attending an AO program. As Abruzzese and colleagues highlight, it is widely accepted that the observation of actions performed by others can provoke in the brain the same activation of neural structures recruited for the actual execution of those actions [3]. Specifically, Buccino and colleagues say that Action Observation therapies are founded on the principles of movement "imitation", implying at the same time motor imagery, observation, and actual movement execution [3]. Previous studies have shown a positive correlation between the prolongation of P300 latencies in PD and cognitive deficits in memory, attention, executive function domains, and depression [21][22][23][24]. Moreover, prolonged P300 latencies are correlated with cognitive impairment and wider behavioral problems in PD patients [25]. Our results showed an improvement in cognitive and motor performances that seems to be related with a reduction of P300 LD signals by means of clinical scales score, related to motor, cognitive, and quality of life assessment, according to current literature [26][27][28][29][30][31]. In conclusion, AO appears to be a safe and feasible paradigm of rehabilitative exercise in cognitively preserved PwP. Moreover, our study suggests that AO practice produces a combined effect both on walking skills recovery and cortical activity changes in PwP. This approach could contribute to increase lower limb motor recovery in idiopathic PD patients. The results of this study, despite the small sample size of subjects enrolled and treated with AO, were also encouraging and supported by recent research works in Parkinson's disease cognitive and motor rehabilitation [18,32]. In general, the simplicity of treatment, the lack of side effects, and the positive results from patients support our theory. Although, AO exercises and training effects over gait and cognitive recovery need to be further investigated. Future research should include a larger sample size and analyze in depth the underlying central and peripheral neural mechanisms over time using long-term follow-up.

Conflicts of Interest:
The authors have no conflict of interest to report.