Parkinson’s disease (PD) is the second most common neurodegenerative disorder after Alzheimer’s disease, thus representing one of the chronic diseases with the major medical, psychological and economic burden.

The cognitive disturbances are frequently found in PD patients, sometimes already at the time of diagnosis and during the early stages of disease. These include a mild cognitive impairment (MCI) found at onset in approximately 26.7% of patients (Litvan et al., 2012), up to an established diagnosis of dementia (Emre et al., 2007). It has been estimated that at least 75% of PD patients will meet dementia criteria 10 years after the initial diagnosis (Aarsland and Kurz, 2010).

In addition to early pharmacological treatment of PD (van den Heuvel et al., 2021), cognitive stimulation training in the early stage of the disease has been proposed to improve cognitive function.

The methods can be divided into traditional cognitive training (paper –pencil tasks training), computerized cognitive training and training with applied virtual reality technology to improve the accessibility of cognitive training (Angelucci et al., 2015; Hindle et al., 2016; Kalron and Zeilig, 2015; Leung et al., 2015; Petrelli et al., 2014, 2015; Yu and Wu, 2022).

The rehabilitation program should be adapted to the different stages of the disease and to the transition of each state. Few studies have investigated cognitive training in patients with PD-dementia complex (PDD), when difficulties in performing the daily living activities (ADLs) arise (Sperens et al., 2020).

Recent research in the neuro-rehabilitation field has reported the potential efficacy of a new strategy, the Action Observation Treatment (AOT) in the recovery of impaired motor abilities for a vast number of clinical conditions, spanning from traumatological patients to brain injuries and neurodegenerative diseases (Rizzolatti et al., 2021).

This technique is an approach based on the paradigm of mirror neurons (Bonini et al., 2022), meaning that the observation of motor actions performed by other individuals activates the motor system of those who observe in a similar manner to the activation that occurs during the actual execution of the movement. There is indeed an overlap between the brain areas that are active during the observation of an action and the ones active during the execution (Molenberghs et al., 2012).

Moreover, there is evidence that the observation of actions conducted by others improves lexical recovery and can be used in the rehabilitation of aphasia to recover naming of verbs (Marangolo et al., 2012; Bonifazi et al., 2013).

A future application of AOT could be training interventions focused not only to a single motor act, disregarding their concatenation into a goal-directed action, but also to the whole chain of movements composing a complex action (Rizzolatti et al., 2021).

There are some specific studies in PD patients on the effects of neuro-cognitive rehabilitation (including AOT) based on mirror neurons paradigm (Abbruzzese et al., 2015; Buccino, 2014; Monticone et al., 2015; Poliakoff, 2013; Wang et al., 2016).

Pelosin et al. (2010, 2018) reported that the AOT may be proposed as an innovative and effective method for improving gait disturbances and mobility, suggesting that this technique could represent a new rehabilitation approach. More specifically, it has been observed that subjects submitted to videos with active movements improved their freezing of gait more than patients who watched “static” videos. A further meaningful result has been obtained for the reduction of bradykinesia after only one session of AOT in PD patients compared to subjects submitted to acoustic cue (Pelosin et al., 2013).

A systematic review by Giannakopoulos et al. (2022) concluded that AOT improves functional ability and motor control in PD patients, with the intervention dose and the characteristics of the stimulus playing a decisive role in its efficacy.

Furthermore, a pilot study by Buccino et al. (2011) in a small cohort of PD patients reported that AOT might improve autonomy in the ADLs, thus representing a promising rehabilitative tool also in this view.

Therefore, the first aim of the present randomized controlled trial was to confirm the benefits of AOT treatment in the ADLs of the patients with PDD, using instruments specifically oriented toward the assessment of ADLs.

A second aim was to test whether the observation of actions could have any effect also on the naming of the actions themselves. This research question arises by the observation that PD patients perform worse than healthy controls in action naming. This could be explained by the evidence that the neural connections between basal ganglia, which are typically impaired in PD, and several cortical areas have a crucial role in integrating motor-semantic information in action-related words (Cotelli et al., 2018).

As the observation of actions pertaining to the human motor repertoire was found to be an effective rehabilitation approach for verb recovery in aphasic patients (Marangolo et al., 2012; Murteira and Nickels, 2020), we thought it would be interesting to investigate whether a similar effect could be found also in PDD patients.

Demographic and clinical characteristics of 20 patients at baseline divided by group (C-group; AOT-group) are reported in Tables 1, 2.

The two groups did not differ in terms of age, gender, years of education, MMPT, MMSE and also for measures of interest as UPDRS, ADCS-ADL, BADA, DAFs.

From baseline to the end of treatment after 4 weeks, in AOT-group there were significant improvements for ADCS-ADL (from 44.3 to 50.4; p = 0.001), DAFS (from 65.3 to 72.9; p = 0.005) and Subtest Verb Naming of BADA (from 19.2 to 21.5; p = 0.011), while with respect to UPDRS part III no significant change was found (see Table 3).

The C-group did not show any significant improvement for all the measures of interest.

The results of this preliminary study suggest potential insights concerning effective strategies in the context of the rehabilitation of PDD persons. The first aims of the present study were to assess the benefits of AOT treatment in the ADLs; the secondary aims were to assess the efficacy in the verb naming and in motor of the PDD patients at mild–moderate stage.

We found that AOT treatment may play a role in the recovery of ADLs of PDD patients and that it improves their performances on action naming.

Similar results were reported by Buccino et al. (2011), who experimented AOT in a small number of PD patients (7 cases and 8 controls). However, this study had relevant limitations in the evaluation tools (Unified Parkinson’s disease Rating Scale-UPDRS and Functional Independence Measures-FIM), which are not specifically oriented toward ADL. Conversely, we carried out an in-depth assessment of ADLs by the ADCS-ADL scale, which is an informant-reported scale specifically designed to assess the competence of patients in basic and instrumental ADLs, and by DAFS, which is a direct assessment of ADLs.

It is therefore reasonable to think that AOT is effective because it targets the Mirror Neuron System and activates the central representation of the behavior, which enhances learning and neural traces of motor actions. In humans the observation of action stimulates neural circuits similar to those involved in action planning and execution and some findings demonstrated that PD patients can still be engaged in action observation to some extent (Poliakoff, 2013).

Furthermore, the same basic principles relevant to enhance motor performance may also be efficacious to improve language functions. Some papers (Marangolo et al., 2010, 2012) reported improvement in verb retrieval only by observing video-clips of human actions in patients with aphasia and confirm that AOT can be a useful strategy to promote long lasting recovery in verb production.

We found similar results in our PDD patients. They presented significant improvement in the verb naming task of the BADA test after the AOT.

In agreement with the embodied cognition viewpoint, it has been suggested that gesture and speech are strongly connected to the same conceptual representation and that gestures are more likely encoded in a multimodal representation including sensory-motor features (Gallese and Lakoff, 2005). If the word is grounded in sensory-motor features, such as a verb, the actual execution and observation of gestures may enhance the activation of the sensory-motor representations in the semantic system and facilitate its retrieval. Therefore, the observation of an action carried out by others is sufficient to activate in the semantic system its corresponding sensory-motor representation, which serves as an input to facilitate retrieval at the lexical level.

Also in this case it is reasonable to think that this effect is due to the involvement of the Mirror Neuron System, which is active both when an action is performed and when it is observed.

Faced to positive results on linguistic abilities and functional competence, we did not find a clear motor efficacy of AOT: indeed, differently from the systematic review by Giannakopoulos et al. (2022), our study did not show any statistically significant effect of this rehabilitation approach on UPDRS part III score, but just a trend toward an improvement in AOT group and a trend toward a worsening in C-group. This suggests that the replication of our study in a larger cohort of PD patients might show a significant efficacy of the here proposed AOT approach also in terms of motor improvement. Furthermore, it should be taken into account that the selection of videos representing mainly fine motor skills in our investigation was typically oriented to try to produce an effect on functional competence (which is the primary outcome of the study) and not on strictly understood motor abilities as assessed by UPDRS part III (which encompasses both fine and gross motors skills). This is a crucial point, because, as specified in the above-mentioned systematic review, the characteristics of stimulus play a relevant role on AOT motor efficacy and in literature there is a significant heterogeneity of intervention features among studies focused on AOT. In other terms, a different selection of motor videos could produce different effects compared to those found in the present investigation.

In summary, our findings provide preliminary evidence on the efficacy of AOT in the rehabilitation of ADLs and language in PDD persons. However, it is important to acknowledge that our study presents some significant limitations.

First of all, given the limited number of patients, the results cannot be considered conclusive and deserve further investigation.

The lack of follow-up evaluation after the end of the treatment is another relevant limitation that might be addressed in further studies.

A third aspect that should be considered is the fact that in our study, like in another analogous studies on AOT (Buccino et al., 2011; Pelosin et al., 2018) the control group underwent a naturalistic video and not a video with motor contents without rehabilitation approach. This aspect is both a strength of the study and a potential source of bias. Indeed, on one hand, our design investigation allowed to make a precise comparison between the rehabilitative effects of visive stimuli acting on neuron mirrors and visive stimuli without any effect of neuron mirrors, thus furnishing an estimation of the validity of neuron mirror paradigm in the context of rehabilitation of PDD patients (it would be impossible if the C-group had been subjected to motor video); but, on other hand, this implies that our study could be affected by a potential overestimation of AOT efficacy due to placebo effect.

Furthermore, the fact that we focused our attention only on PDD patients with H&Y 2–3 clearly limits the generalizability of the results.

Finally, some studies highlighted some factors potentially influencing ADLs, such as gender (Sperens et al., 2020) and the presence of behavioral disorders (Chen et al., 2022), that were not evaluated in the present investigation. Therefore, further studies should take into account these factors that might influence patient outcome.

Taking into account all these relevant limitations, our study should be considered only an exploratory investigation and not a definite demonstration of AOT efficacy in PDD context. Further larger studies indeed will be necessary to confirm these preliminary results.

In conclusion, we believe that these new promising findings might contribute to open future directions for the research about new rehabilitation approaches in PD. We hope that future studies will explore these research directions in order to verify the validity of AOT approach in the improvement of functional competence and linguistic abilities of PDD subjects.