Effects of Virtual Reality in the Rehabilitation of Parkinson’s Disease: A Systematic Review

Background: Parkinson’s disease is characterised by the loss of balance and the presence of walking difficulties. The inclusion of rehabilitation therapies to complement pharmacological therapy allows for comprehensive management of the disease. In recent years, virtual reality has been gaining importance in the treatment of neurological diseases and their associated symptoms. Therefore, the objective of this systematic review was to analyse the effectiveness of virtual reality on balance and gait in patients with Parkinson’s disease. Methods: This study is a systematic review conducted following PRISMA’s statements. An electronic search of the literature was carried out in the following databases: PubMed, Cochrane, Dialnet, Scopus, Web of Science, PsycINFO and Science Direct PEDro. The inclusion criteria were controlled and non-controlled clinical trials published in the last 12 years in English or Spanish, in which virtual reality was applied to treat balance and gait impairments in patients with Parkinson’s disease. Results: 20 studies were finally included in this review. A total of 480 patients participated in the included studies. All patients were diagnosed with Parkinson’s disease. Most of the investigations used the Nintendo Wii + Balance Board or the Microsoft Kinect TM combined with the Kinect Adventures games as a virtual reality device. Conclusions: According to the results of this literature review, virtual reality-based interventions achieve good adherence to treatment, bring innovation and motivation to rehabilitation, and provide feedback as well as cognitive and sensory stimulation in patients with Parkinson’s disease. Therefore, virtual reality can be considered an alternative for personalised rehabilitation and for home treatment.


Introduction
Parkinson's disease is the most common neurodegenerative movement disorder [1]. This condition is characterised by the presence of motor and non-motor symptoms which are related to the damage of multiple structures of the central and peripheral nervous system [2][3][4]. These symptoms have a negative impact on coordination and on mobility [5], which has a significantly negative effect on the person and impairs the patient's quality of life [5].
The most commonly used pharmacological treatment is the intake of Levodopa or an oral dopamine precursor [4][5][6][7]. Despite the current medical and pharmacological treatments, patients continue to progressively develop motor and non-motor impairments, such as sleep disturbances, cognitive impairment, and mood disorders [1]. This progression of the condition makes a comprehensive rehabilitation treatment essential, with the physiotherapist as part of the multidisciplinary team [8,9]. Physiotherapy in people with Parkinson's disease will focus on six specific areas: transfers; posture; balance (falls); upper limb function; gait; and physical capacity and activity [10,11].
Currently, thanks to research and therapeutic and technological innovations, there are other treatments that can be complementary to pharmacology and conventional therapy. These include those treatment approaches based on the use of neurorehabilitation programmes by means of electronic systems, which allow rehabilitation to be extended beyond the health centre. Moreover, virtual reality is an innovative approach that has been gaining importance in the treatment of neurological diseases in both motor and non-motor impairments in recent years [12].
The term virtual reality was first used by J. Lamier in 1986. Although the definition of this term has changed over time, one of the most widely accepted is 'the simulation of a real environment generated by a computer, in which a human-machine interface allows the user to interact with certain elements of the simulated scenario' [13][14][15][16][17].
Virtual reality allows a therapeutic intervention based on the use of technologies with an interactive interface that recreates in real time the representation of a perceptual reality generated by the computer, with the patient being able to act and participate in this virtual environment [12]. It is important to note that virtual reality is a technology that allows the input and output of information in the system. In addition, the motor performance is displayed in the virtual environment, and subsequently, the system provides multimodal feedback related to the execution of the movement. Through the external and internal senses (proprioception), the sensory feedback is integrated into the patient's mental representation [18,19]. The sensory feedback associated with the exercises in the virtual environment appears to activate the mirror neuron systems, which would be able to store in primary motor cortical areas a memory of the representation of the movement to be performed [20,21].
In addition, scientific research has evidenced that computational neuroscience, i.e., based on the use of computers and technologies, has demonstrated that the application of virtual reality offers a greater feedback service of the actions performed [22,23]. Likewise, this feedback allows greater improvements in motor learning and task performance compared to traditional training [23]. Both immersive and non-immersive virtual reality are currently used, although the coupling between perception and action in non-immersive virtual reality can be quite different than in the real world, which is why immersive virtual reality is used to achieve greater reality as patients may 'forget' that they are in a training situation [22,24,25].
On the other hand, exergaming programs based on entertainment platforms such as Nintendo Wii or adaptations thereof have been feasible for therapeutic use, improving abilities such as balance or quality of life and achieving high levels of satisfaction and adherence in people with Parkinson's disease [26].
Based on all this, the objective of this systematic review was to know the effectiveness of virtual reality on balance and gait in patients with Parkinson's disease.

Study Design
This systematic review was carried out following the PRISMA statement [27]. The review protocol is available in PROSPERO (registration number: CRD42021256172).
In order to identify relevant studies, the search was conducted in the following databases: PubMed; Cochrane; Dialnet; Scopus; Web of Science; PsycINFO; and Science Direct PEDro (Physiotherapy Evidence Database).

Search Strategy
The keywords used in the abstract and title fields were as follows: Parkinson's disease; virtual reality; gait; balance; Parkinson's disease; physical therapy; physiotherapy. These keywords were introduced in Spanish when the database required it. The Spanish terms used were as follows: Parkinson; realidad virtual; equilibrio; deambulación (Parkinson's disease; virtual reality; balance; mobility). The keywords were combined with the Boolean operators AND or OR. The syntaxes of combined descriptors in the scientific database search can be found in Table 1.

Inclusion and Exclusion Criteria
The inclusion criteria were as follows: (a) both controlled (C) and non-controlled (NC) clinical trials; (b) published within the last 12 years; (c) in English or Spanish; (d) individuals ≥ 65 years. The search was limited to the last 12 years in order to analyse the most recent advances in the use of virtual reality in the variables under study and to update the scientific evidence available in the literature on this topic [28,29]. The exclusion criteria were established following the PICO model (population, intervention, control, comparison, and outcomes). Exclusion criteria were established as follows: the literature reviews or any type of document that is not a clinical trial; the use of treatment techniques that are not based on virtual reality; and treatments carried out on patients under 65 years of age.

Study Selection
A pre-selection of papers was performed, considering that they were within the proposed subject of the study. This selection was carried out by reading the abstract of the studies and excluding those that did not meet the established criteria.
The full text of the studies that met the inclusion criteria was revised, analysed, and included in the systematic review. All potential full-text articles were retrieved and evaluated by the two reviewers independently. Although the level of agreement between the two reviewers was not specifically calculated, any disagreements on the inclusion/exclusion of full-text articles were resolved via discussion ( Figure 1). The following data were obtained from the studies included in this review: author and date; study sample (sex and mean age); inclusion and exclusion criteria; intervention; follow-up; assessment scales used; and results obtained in the study. This data were compiled in a standard table. The reviewers who selected the articles also independently obtained the data and assessed the methodological quality of the studies. If there were any disagreements, they were resolved via discussion.

Assessment of Methodological Quality
The analysis of the methodological quality of the studies was performed using the PEDro (Physiotherapy Evidence Database) scale [30]. This scale consists of 11 items that can have a 'yes' or 'no' as a reply. The total range of scores is from 0 to 10 according to a low to excellent methodological quality. The results obtained in the scale were considered as High quality if the score is over 5 (6-8: good, 9-10 excellent), Moderate quality if the score is between 4 and 5 (fair quality study), and Low quality if the score is under 4 (poor quality study). The first item is additional as it is related to external validity and is not used to calculate the score obtained. Therefore, the maximum score is 10. Items 2 to 9 aim to justify if the study has enough internal validity, and items 10 and 11 analyse if the statistical information is appropriate to understand the results. The assessment of the methodological quality of this study was calculated by one reviewer only.

Risk of Bias Analysis
The risk of bias [31] was calculated for each included study, referring to the following types of bias: selection bias; performance bias; detection bias; attrition bias; reporting bias; and other biases. In this assessment, 7 criteria were assessed: 1 = Random sequence generation (selection bias); 2 = Allocation concealment (selection bias); 3 = Blinding of participants and personnel (performance bias); 4 = Blinding of outcome assessment (detection bias); 5 = Incomplete outcome data (attrition bias); 6 = Selective reporting (reporting bias); 7 = Other biases. The risk of bias and the quality of this study were calculated by one reviewer only.

Results
The literature search was conducted in May 2023. A total of 399 studies were obtained from the search in all databases. The PRISMA flow chart ( Figure 1) shows the selection process of the studies. The records that were duplicated were excluded, and 305 records were screened. Finally, 20 studies were included in this review. Table 2 shows the main findings of this review.

Description of the results
The main characteristics of the studies are shown in Table 1. The most relevant aspects are the following: Sample: A total of 480 patients participated in the included studies. Regarding the number of participants in the different studies, it is worth noting that there was great variability. On the one hand, we observed a sample size of more than 30 patients in six articles [14,[37][38][39][40]43,[46][47][48] and on the other hand, in four articles [8,[33][34][35], the sample size was less than or equal to 7 patients.
We found a higher frequency of men than women [14,26,32,34,35,39,40,42,44,45,47] among the sample. Only four studies [8,37,41,43] had a sample where the female sex prevailed over the male sex but without a significant difference. In the study by Loureiro et al. [33], the sex of the patients was not specified.
Methodology of the studies: All the studies included in this review were clinical trials, but they differed significantly in the methodology applied. Not all the studies had a control group [8,[32][33][34][35]44], and those were the studies with smaller samples, so the research was carried out without being able to compare the intervention with other techniques or without a placebo.
In the study conducted by Yen et al. [14], the control group received no treatment at all. In other studies, [37,38] the control group received fall prevention education or the research was based on the comparison of home treatment with virtual reality with supervised treatment in the clinic [39] or routine physical therapy [45][46][47].
In addition, in the study of Calabró et al. [40], we could observe that the procedure was different from the other studies. Although they did not have a control group that did not receive any type of treatment, all the participants completed 20 weeks of conventional physiotherapy, and after three months of rest, they all completed the virtual reality treatment as well.
Virtual reality devices: Regarding the hardware or devices that were used in the different studies, it can be said that there is a predominance of the two most accessible, low-cost devices. On the one hand, there are those studies that used the 'Nintendo Wii' [47], the 'Nintendo Wii + Balance Board' [33,[37][38][39], and on the other hand, there are those that have used the Microsoft Kinect TM [45] in combination with the Kinect Adventures games [34,35,40]. Other studies used different devices, such as the Virtools 3.5 tool [44], a console created by the National Formosa University that applies the treatment through two virtual-reality-based games called Bang Bang Ball and Simulated Board Driving, 'CAREN' [42], which is a device composed of a motion capture system and a base platform which is hydraulically driven by the subject's movements. Moreover, the Interactive video game-based System was used [43], which is a modification of the XaviX Console that applies two games as treatment, one in which the patient performs multidirectional steps and another in which the patient performs steps towards a target. One of the studies [32] used a virtual realisation system created specifically for the study where participants had to process different stimuli and make decisions while walking on the treadmill. One of the most recent studies used the Tymo ® system [46], which is a wireless platform for balance and postural control training. The Tymo ® system is connected to a screen and provides virtual reality games, adaptable to the functional capacity of the patient. In contrast, two of the analysed studies did not describe the virtual reality device used [41,48].
Type of training and duration of studies: The characteristics of the different treatments vary considerably in the different studies in terms of training volume (number of weeks), frequency, and duration of the sessions. Regarding the volume and frequency of the training, most of the interventions involved about 5-6 weeks of treatment along with two-three sessions per week [14,26,32,37,38,43,44,46]. However, Tunur et al. [8] carried out 3 weeks of treatment but had the highest number of sessions accumulated throughout the week, as they performed daily sessions. Another aspect of this study that can be highlighted is the fact that this training took place in the patient's own home.
On the other hand, in one study [36], the intervention consisted of 5 weekly sessions over 12 weeks. This period of treatment was similar to the study by Kashif et al. [47], while Hong et al. [48] applied 8 weeks of treatment once a week.
Follow-up assessment: All studies conducted a pre-treatment and post-treatment assessment; however, not all of them conducted follow-up evaluations to evaluate whether or not the effectiveness of their intervention was sustained over time. The studies that conducted follow-up carried out the assessments four weeks after the end of the training [14,32,[37][38][39]47].
In contrast, Calabró et al. [42] did the follow-up measures at three months, while in the other two investigations [14,40], the follow-up was performed one week after the treatment was completed, and one study did not specify when they established to complete the follow-up assessment [8].
Effects obtained: Several research studies showed that balance improved after virtual reality treatment [26,34,36,37,39,[44][45][46]. However, many of them did not perform follow-up evaluations and, thus, did not show evidence of the benefits in the long term.
In the study of Calabró et al. [42], the results were maintained in the long term, even after three months post-intervention. They observed that performing virtual reality training (CAREN) four times a week led to a significant improvement in the gait cycle in terms of duration, speed, length, cadence and step width reduction.
Other research whose results were maintained at the four-week follow-up assessment showed an increase in gait speed, stride length and stride time and an improvement in the 6-m walking test and even in obstacle negotiation [32]. Furthermore, in the study by Kashif et al. [47], the experimental group showed statistically significant improvements in balance at follow-up, with more than 90% of patients showing improvements in this outcome measure.
Yan et al. [26] and Hong et al. [48] used the timed up-and-go test as a measuring tool, which allowed them to prove that the use of virtual reality improved the patient's functional mobility with consecutive movements (sitting, standing, walking, turning, etc.). However, this improvement was also achieved with conventional home training, as no significant differences were found between virtual reality balance training and conventional home balance training [26]. However, the use of a routine basis treatment combined with virtual reality and Jiao scalp acupuncture made the participants perform the timed up-and-go test in a shorter time than those who received routine basic treatment and virtual reality alone [48].
Most of the results of the studies that compared conventional therapy versus virtual reality showed improvements in gait and balance in both groups. However, the research by Ferraz et al. [40] had three different groups (functional training, exercise bike, and virtual reality training), and only the virtual reality group had a significant improvement in gait speed at the 10-m walking test.
It has also been shown that these improvements in gait occur more effectively in treatments using virtual reality because cognitive and sensory functions are also stimulated. Balance improved notably thanks to the inclusion of integrative function training, which shows the importance of not focusing solely on motor exercises in rehabilitation [44]. Therefore, the use of virtual reality facilitates this treatment approach [44].
The improvement in cardiovascular endurance, which, in turn, influences the improvement in gait, is evidenced in the study carried out by Pompeu et al. [34]. Nevertheless, the maintenance of this improvement at the follow-up was discussed by the authors, as well as the possibility that this improvement was also achieved with conventional physiotherapy.
Regarding balance, we found that most studies used the Berg balance scale to assess this outcome measure. This scale is considered to be the 'gold standard' for evaluating func-tional balance and fall prevention tests that assess the patient's balance and static abilities. Almost all studies show an improvement in balance in Parkinson's disease patients after a virtual reality training programme [26,33,36,39,[41][42][43][44]46]. However, two investigations found that this improvement in balance was only significant in the experimental group where virtual reality was used [41,43,46].
Methodological quality: The results of the assessment of the methodological quality are shown in Table 3. It should be noted that a negative response does not necessarily mean that the study does not have this characteristic but rather that the requirement was not found in the text even after a thorough review of the article. Mirelman et al., 2011 [32]  The scores obtained in the clinical trials indicated that their methodological quality was fair, with a score of 4-5 [34][35][36]42], and good, with a score of 6-7 [14,26,[37][38][39][40][41][45][46][47][48]. Furthermore, we found four articles with poor quality [8,32,33,43], three of which did not have a control group and had a small sample size, and one [43] that, although it had a control group, had a small sample size and presented differences in the baseline data between the groups. Group A had a total of 2 men and 10 women, and group B had 9 men and 3 women, an aspect that may have had a significant influence on the results of the study.
In terms of the study design, it is worth noting that only in three studies [37,38,40] the allocation was concealed. Characteristic 4, or baseline of comparability, was not met by three poorly rated articles [8,32,43]. Another aspect to be taken into account is that none of the articles complied with patient and therapist blinding. In contrast, the follow-up of 85% of the subjects was met in almost all the studies except for three [36,43,44]. The statistical comparability between groups was not met in four articles [8,32,33,35], and the last criterion was met in all the studies except in those carried out by Tunur et al. [8] and Loureiro et al. [33].

Discussion
The aim of this systematic review was to analyse the efficacy of virtual reality on balance and gait in patients with Parkinson's disease. Several important aspects are discussed hereafter.
In relation to the effects obtained after the interventions, most of the studies analysed in this review indicated that virtual reality improved gait speed, stride length, balance, gait, and postural control in patients with Parkinson's disease [14,32,34,37,42,[45][46][47]. Furthermore, in the study conducted by Yang et al. [26], the Dynamic Gait Index, Timed up-and-go test and Berg Balance Scale showed significant improvement in both groups, and these changes were maintained during follow-up. Mirelman et al. [32] also achieved improvements in stride length, stride time, gait speed, and obstacle crossing after the intervention with virtual reality, and these significant improvements were maintained at follow-up. Using the CAREN Virtual Reality device, Calabró et al. [42] found that significant improvements in both gait (10-Meter walk test, Timed up-and-Go test and instrumental gait analysis) and balance (Berg Balance Scale) were only obtained at the follow-up assessment in the group that received virtual reality training alone. In the study by Yen et al. [14], there were also improvements after the training and at the subsequent follow-ups. However, there were no significant differences between the virtual reality groups and those with conventional physiotherapy. We believe that in order to know the real effect of virtual reality applications, it is of great importance that all studies follow up on the results over time and not only after the end of the treatment period. Based on these results and those of other research, such as that of Lei et al. [29], virtual reality technology could be considered a rehabilitation approach which is as effective as traditional rehabilitation therapy. Even in outcome measures, such as gait (stride, speed, stride length), balance, and quality of life, the results have shown that virtual reality is better than conventional training.
The role of virtual reality in the rehabilitation of patients with Parkinson's disease significantly influences the brain's ability to perceive, process, and integrate information [28]. In this aspect, the study by Pazzaglia et al. [44] showed that after the intervention, there was a significant improvement in balance and gait outcomes in the virtual reality group compared to the control group due to the fact that more cognitive and sensory functions were stimulated than with conventional physiotherapy. Furthermore, in addition to the improvements in the walking ability of Parkinson's disease patients, progressive increases in muscle strength and sensory integration have been found [38]. A previous systematic review [49] had similar results as virtual reality showed positive effects on balance and gait, as well as other variables, such as activities of daily living function, quality of life, and cognitive function in patients with Parkinson's disease. They considered that it could be possible that virtual reality provides more comprehensive and accurate motor feedback, which would explain the improvements achieved.
Another benefit that Lei et al. [29] showed about virtual reality is the instantaneous feedback that occurs with these devices, which also improves compliance with rehabilitation training and patient motivation. This coincides with Pompeu et al. [34], who concluded that the main factor that led to improvements in the learning of different motor functions thanks to virtual reality was the presence of continuous visual and auditory feedback provided by the Kinect games throughout the sessions. This aspect was also accounted for by Yang et al. [26], who compared a virtual reality group that focused on visual and audio feedback and a control group that focused on verbal feedback from the therapist. In addition, Feng et al. [41] concluded that the advantage of virtual reality over conventional rehabilitation was that providing continuous feedback improved the patient's cognitive sensation, increased interest and continuously stimulated the patient's motivation. As Canning et al. [50] stated, the interaction with the virtual environment and the feedback about performance and success promotes adherence and the success of the treatment.
Motivation and adherence are also influenced by the degree of difficulty and individualisation of the games and tasks performed in the virtual reality intervention. For example, Palacios et al. [35] highlighted the importance of individualisation in the configuration of the parameters of each game, adapting the degrees of difficulty to the ability of each patient. This also occurred in the research conducted by Pompeu et al. [34], where the selection of games was individualised according to the motor and cognitive demands of each patient. Moreover, Domínguez-Ferraz et al. [40] established a gradual progression of the intensity in order to adapt to the different degrees of difficulty of the participants. In this sense, Howard [51] stated that the real impact of virtual reality programmes is achieved through the improvements obtained by patient motivation. In addition, this author supports the idea that the effectiveness of the interventions will depend on the degree of interest that the patients have in these programmes.
Regarding the technological devices that were used in the studies analysed, it is interesting to highlight that most of the virtual reality interventions were carried out through immersive games. The most widely used technology in the studies analysed was the Nintendo Wii Console with the Balance Board accessory. In relation to cost-effectiveness, the Microsoft Kinect TM [34,35,40,45] and the Wii Fit [33,[36][37][38][39]47] were the devices that provided the advantages of the use of virtual reality but also reduced the economic cost of the treatment in chronic patients. Therefore, it seems appropriate that those devices should be considered for treatment. The results of the interventions analysed in terms of the devices used coincide with those obtained by other authors who have used virtual reality as a treatment tool, so the use of the Kinect device in the recovery of other pathologies also provides benefits [45,52,53]. In other studies analysed, we can see that different devices were used for the application of virtual reality treatment, such as the balance training system created for the occasion to apply virtual reality treatment in Parkinson's Disease patients [26] or the Tymo-system [46], obtaining similar benefits to the use of the Kinect device [54].
In most of the studies, the tests were conducted in clinics and specialised centres, and only two studies took place at home [26,39]. The study by Yang et al. [26] showed no significant differences between virtual reality at home compared to conventional home treatment, while Gandolfi et al. [39] concluded that sensory integration performed in the clinic was more effective than virtual reality at home. In contrast, Brachman et al. [45] was the only study that combined one training session a week supervised by a physiotherapist with two sessions performed at home.
In the same way that remote assessment tools such as telemedicine via video calls have been made available in the last years to ensure optimal assessment and treatment monitoring [55], rehabilitation services that can be provided at home without on-site medical supervision should be available, for example, through the use of virtual reality which provides easily accessible and low-cost technological tools [17]. In addition, these instruments can collect a report of the activities performed, allowing for constant feedback and recording of the patient's progress [23]. Therefore, virtual reality offers the possibility of developing telerehabilitation platforms, where professionals can remotely follow the evolution of the patient from the data recorded during each of the therapy sessions and could apply more personalised interventions to each of the patients [55].
Rehabilitation through virtual reality offers the possibility to carry out the exercises at home, ensuring that the treatment is not interrupted for such reasons as closure of the centre, contagion, difficulty of mobility to the centre, or confinement. On the other hand, we believe that, as rehabilitation can be carried out at home, the patient can do it when he/she feels better (ON phase), allowing for better physical work and greater control of the medication and the disease.

Limitations of the Study and Further Research
We consider that there is great methodological variability in the research analysed. The inconsistency in the use of assessment tools for the same variable made it difficult to compare results and could lead to different interpretations of the results despite being adequate, current, and validated tools in all the studies included in this review. Therefore, we believe that in future research, it would be necessary to analyse and describe the effects on balance and gait achieved by the application of virtual reality, in addition to an in-depth study of the physiological and psychological effects produced by this type of therapy and the establishment of criteria for inclusion and methodological application that will provide us with the most reliable results [56]. This could be due to the fact that this research is focused on comparing the benefits of these interventions and not on their potential use as a means of assessment.
In terms of methodological quality, the studies included in this review scored between 3 and 7 on the PEDro scale. According to the PEDro interpretation guidelines, if studies scored at least 5 out of 10, they were considered to be of acceptable quality. Studies that scored around 4 did not include blinding of all patients, therapists, and evaluators. Due to the nature of virtual reality interventions, it is very difficult to have triple blinding, as a placebo cannot be used, and the treatment provided is clear to the therapists.
The fact that the studies did not compare virtual reality interventions with each other and only did so with conventional treatments, together with the heterogeneity in the frequency of application of the interventions and number of sessions, means that we cannot conclude which type of virtual reality training is the most appropriate for achieving the greatest benefits in this type of patients. Furthermore, the objective of this study was to analyse the effects of virtual reality in patients with Pakinson's disease regardless of the type of virtual reality used (immersive or non-immersive), but it would be interesting in future research to analyse this aspect.
Further research is needed to provide better methodological quality and a more solid basis on what effects are achieved by virtual reality, to establish which type of virtual reality training would be the most appropriate and its application in different degrees of the disease, in order to extrapolate the results.

Implications of the Use of Virtual Reality in Clinical Practice
The results of this systematic review can have positive implications for the clinical practice of professionals working in the rehabilitation field. Virtual reality is a computersimulated reality that allows the user's experience of the world he or she perceives to be modified [57].
The studies analysed showed that this technique improves gait and balance in patients with Parkinson's disease. Furthermore, new virtual reality technologies can provide an engaging and immersive environment for exergaming techniques, maximising goal-oriented training and increasing patients' self-efficacy during rehabilitation [58]. The results of this review, as well as those from previous ones [28], support that home-based virtual reality can be used as a prolongation to conventional post-clinical rehabilitation programs and help extend the rehabilitation period and favour clinical benefits for patients. Compared to conventional physiotherapy, virtual reality provides the advantage of more personalised training, the possibility of home-based rehabilitation where data can be uploaded in real time and recorded, and allows for greater accessibility, especially in areas with limited access to rehabilitation services [29].

Conclusions
According to the results of this literature review, virtual-reality-based interventions showed improvements, which are similar to conventional therapy, in the gait variables (gait speed, stride length, decrease in stride width) and balance in patients with Parkinson's disease. This type of therapy achieves positive results in relation to adherence to treatment, individualisation of the treatment, innovation, motivation, and feedback capacity, as well as great cognitive and sensory stimulation for these patients. Furthermore, thanks to the benefits of virtual therapy, together with the possibility of doing it at home, it allows its application in situations of mobility restrictions. Therefore, virtual reality interventions may be a suitable alternative to the home rehabilitation approach allowing for personalised treatment for these patients.
Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.