Combining immersive exergaming with physiotherapy in a specialized intensive Parkinson’s disease rehabilitation program: a randomized controlled trial

Abstract

Background

Exergaming is a new technology for implementing innovative rehabilitation interventions for neurological conditions. Our aim was to evaluate the feasibility, safety and efficacy of a novel exergaming experience called neuroanimation, in the form of MindPod^TM Dolphin, added to an intensive physiotherapy program for Parkinson’s disease (PD).

Methods

MindPod Dolphin provides a 2D oceanic environment where patients make high-amplitude 3D arm movements controlling an animated dolphin. We conducted a rater-blinded, randomized-controlled trial. Eligible people with PD were assigned into two groups: MindPod Dolphin over 12-weeks combined with physiotherapy (COM) or sequentially after 6-weeks of physiotherapy alone (SEQ). Sessions occurred three times/week. Assessments occurred at baseline, 6-week, 12-week, and 4-week post-intervention. The primary outcome was change from baseline to 6-week in the Timed Up and Go (TUG) test. Secondary outcomes were change, from baseline to each evaluation timepoint, in several motor, cognitive and quality of life measures. Participants’ safety was assessed. Adherence, system usability and participants’ satisfaction were used as measures of feasibility.

Results

Thirty individuals were recruited. Baseline characteristics were similar between groups. Our primary outcome (change in TUG from baseline to 6-week) did not significantly differ between groups [median TUG change in COM = 0.243 [-1.873, 1.176] vs. SEQ=-0.016 [-0.810, 0.350], estimated difference = 0.002 (95%CI -1.103; 1.273); p = 0.983]. Both groups improved in motor and cognitive measures with a trend favoring COM. When compared to SEQ, COM improved significantly in TUG cognitive from baseline to 16-week (p = 0.048). A main effect of time was observed for TUG cognitive in COM, and mini-BEST in SEQ. Adverse events (n = 309) were mostly mild (55%). Overall, 20 participants (67%) adhered to the study protocol, with the COM being more compliant than the SEQ group. MindPod Dolphin was considered easy to use and over 80% of participants were at least moderately satisfied with it as an intervention.

Conclusions

Neuroanimation-based exergaming in PD was feasible, safe and effective in improving multiple secondary measures. The advantages of the exergame became evident at 12-weeks and beyond, suggesting that it had cumulative and delayed beneficial effects on cognitive and motor outcomes when added to a lead-in phase of intense physiotherapy.

Trial registration

ClinicalTrials.gov registration: NCT04699617.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12984-025-01640-w.

Background

Parkinson’s disease (PD) is a neurodegenerative disease characterized by a combination of progressive motor and nonmotor symptoms that substantially affect functioning and quality of life of the person living with it [1]. The progressive degeneration of dopaminergic neuronal cells in the nigrostriatal pathways, causing dopamine depletion in the posterior sensorimotor striatum, is thought to be the main mechanism leading to the core motor symptoms of PD. However, neurodegeneration involving several other striatal-thalamic-cortical functional circuits, as well as aberrant and compensatory connectivity patterns, is also believed to contribute to the complex constellation of PD motor and non-motor symptoms, including impairment in motor control, cognition and other behaviors [2, 3]. Notably, it has been shown that posterior striatum pathology leads to an increased reliance on the relatively spared anterior associative putamen and, consequently, to more goal-directed control of movement [4, 5]. Thus, while dopamine replacement therapy has been pivotal in treating PD motor symptoms, higher motor control and cognitive deficits may be due to compensatory neural responses and therefore less influenced by dopaminergic therapy [6]. Instead, these pathways involved in external cueing and cognitive engagement may be better targeted by goal-directed non-pharmacological interventions [7-9].

Physiotherapy and exercise have gained increased relevance as non-pharmacological interventions for PD symptomatic management [10] with proved benefits in several domains, including severity of motor signs, muscle strength, walking speed, stride length, functional mobility, balance capacity and quality of life [8, 11, 12]. Therefore, the European Physiotherapy Guideline recommend practice incorporating motor learning, dual-tasks and motor strategy training [8]. However, most people with PD have a sedentary lifestyle [13, 14], with lack of motivation to comply with physical activity for long periods of time [8, 15]. Hence, approaches to promote motivation to maintain exercise are needed [8, 16].

In recent decades new technologies using gaming environments and virtual reality (VR) have been investigated as innovative cognitive-motor rehabilitation tools [17]. They provide multisensorial stimuli and enable a wide variety of complex movements within engaging and challenging environments, coupling the possibility to personalize training with motivational elements [17-20]. In PD, although evidence remains limited, VR seems to be a useful complement to conventional physiotherapy to improve motor, cognitive and quality of life outcomes [21].

Neuroanimation is a novel full-body exergaming approach that uses simulated creatures for neurorehabilitation. MindPod Dolphin is the first neuroanimation experience and consists of an oceanic environment within which patients control a simulated dolphin by performing large amplitude arm movements [22, 23]. While learning how to control the dolphin, patients encounter challenging tasks demanding cognitive engagement. In a focus group, therapists evaluated MindPod Dolphin as a holistic training approach with carry-over to functional tasks and high levels of motivation in stroke patients [24]. Therefore, we hypothesized that incorporating neuroanimation into a specialized physiotherapy program could serve as supplementary motor-cognitive training to improve functional mobility in people with PD.

We aimed to evaluate the feasibility, safety and efficacy of incorporating the MindPod Dolphin into the management of functional mobility in PD, as a complement to a specialized six-week physiotherapy program. In addition, we evaluated the effect of extending this combined therapy for 12 weeks and compared it with an intervention in which MindPod Dolphin was added for six weeks following a six-week period of physiotherapy alone.

Methods

Study design

We conducted a rater-blinded, randomized controlled trial at CNS-Campus Neurológico, a specialized movement disorder center in Portugal. This study was approved by the CNS Ethics Committee (Ref. 03-2020) and performed in accordance with national legislation and the Declaration of Helsinki. ClinicalTrials.gov registration number is NCT04699617.

Participants

Participants were recruited from the CNS outpatient clinic and by advertising on social media. Subjects willing to participate were eligible if they were diagnosed with Parkinson’s disease [25], with a Hoehn and Yahr (HY) stage between I and III on medication (MED ON), able to perform the Timed Up and Go (TUG) test at a normal pace and without assistance in less than 11.5 seconds in MED ON, on stable medication for the past month, able to communicate with the investigator, to understand and comply with study requirements, and to provide written informed consent to participate in the study. Participants were excluded if they had a history of falls (at least one fall in the previous three months), major psychiatric co-morbidities, a Montreal Cognitive Assessment (MoCA) score under 21, significant visual or visual-perceptual deficits or neuropsychological impairments, any other neurological or orthopedic disorders likely to affect gait or exercise capacity, any unstable medical condition including cardiovascular instability in the past 6 months, any interfering activities performed at a high level including sports, inability to correctly respond to the assessment protocol according to the clinician’s judgment or lack of support from a caregiver for enrolling. Changes in dopaminergic medication were allowed to optimally reflect common daily practice and to assure the best medical treatment during the study participation. Individuals who agreed to participate were asked to provide written informed consent.

Study procedures

Eligible people with PD were randomized (1:1) using computer-based allocation to receive either a combination of PD specialized physiotherapy with MindPod Dolphin training for 12 weeks (COM) or a sequential intervention starting with PD specialized physiotherapy for six weeks to which the MindPod Dolphin training was added for the following six weeks (SEQ). Figure 1 shows the study protocol. Outcome assessors were masked to study group assignment.

Fig. 1.

Trial design.Participants were randomly allocated to one of the study interventions: the combined therapy which entails a combination of MindPod Dolphin training with Parkinson’s disease (PD) specialized physiotherapy for 12 weeks, or the sequential therapy which comprises a sequential intervention starting with PD specialized physiotherapy for 6 weeks to which the MindPod Dolphin was added for the following 6 weeks. The primary analysis refers to the first 6 weeks of the trial where a combined therapy was compared to physiotherapy alone. The second 6-week interval allowed for the comparison between a 12-week combined therapy and a more sequential intervention. The last 4 weeks without intervention enabled the evaluation for retention effects. Figure created in BioRender. Pimenta Da Silva, D. (2025) https://BioRender.com/n26g266

Assessments were performed by the same trained raters throughout the study, a physiotherapist (RC) and a neurologist (DPS). Participants were evaluated at baseline, at the end of the 6 and 12-week, and after a 4-week washout period, while on their usual medication. At baseline, demographic and medical data was obtained by interview and confirmed by medical records when possible. Medication was registered and reviewed at each evaluation timepoint. Levodopa equivalent daily dose (LEDD) was calculated following reported recommendations [26].

Interventions

Both physiotherapy and MindPod Dolphin sessions lasted 60 minutes each and were delivered on the same day, three times per week. All sessions were supervised by two physiotherapists experienced in PD rehabilitation.

MindPod Dolphin, the neuroanimation experience, consists of an oceanic scene with a virtual dolphin (Bandit) swimming, as well as fish, sharks, a killer whale, and oceanic flora (Fig. 2). To enhance immersion, this environment was displayed on a large white wall, oceanic sounds and music were played and delivered from the ceiling over the patient and the lights were dimmed during the entire session. Additionally, the space between participant and the wall was free of equipment and objects to eliminate clues from the real surrounding. While playing, patients had to control the dolphin by performing arm movements in all planes of 3D active ranges of motion. They had to complete a variety of goals including chasing and eating fish, evading attacks, escaping and jumping above water level, often within a limited time. Patients could alternate the arm being used and, in some levels, a remote device could be used to increase the speed of the dolphin thus providing bimanual game control. Besides arms, the wide range of movements could also challenge balance. Progression could be personalized by the therapist, who would take into account feelings of motivation or frustration reported by participants. Number of completed levels and time spent on game were recorded. Physiotherapists were present during the entire session to give verbal feedback correcting movements, postures and exploration of the workspace, providing technical assistance and assuring patients’ safety. Before study initiation, physiotherapists underwent in-person and video training on the game software to guarantee that proper instruction techniques would be provided to participants.

Fig. 2.

Neuroanimation experience.Figure created in BioRender. Pimenta Da Silva, D. (2025) https://BioRender.com/y60m402

Physiotherapy was delivered in accordance with the European guidelines for rehabilitation in PD [8]. These sessions consisted of 20 minutes of cardiovascular training (cycling or treadmill), 30 minutes of balance and functional training and 10 minutes of cool down and stretch. The functional training consisted of exercises of mobility and transfers (sit-to-stand, power movements, bed mobility), strength (functional exercises with or without extra load or isometric) and gait (with or without dual-tasking and Nordic walk). Thus, physiotherapy was focused on physical capacity training, as well as adaptive strategies to enhance functionality at home. The specific type of exercises were adapted to patients’ needs.

Outcomes

The primary outcome was change in the TUG from baseline to 6 weeks. The TUG is widely used in PD to assess functional mobility and its sensitivity to change has made it a reliable test to measure treatment responsiveness in clinical trials [27, 28]. Absolute minimal detectable change (MDC) values vary from 3.5 to 11 seconds [27, 28].

Changes in the following secondary outcomes were assessed across different study periods: TUG, TUG with motor (TUG motor) and cognitive (TUG cognitive) dual-tasking, 10-meter walk (10MW) [27, 28], MDS-UPDRS [29], mini-Balance Evaluation System Test (mini-BEST) [30], SCales for Outcomes in PArkinson’s disease-COGnition (SCOPA-COG) [31], 39-item Parkinson’s Disease Questionnaire (PDQ-39) [32], Clinical Global Impression of Improvement (CGI-I) and Patient Global Impression of Change (PGI-C). Kinematic gait parameters were recorded using Physilog^TM sensors (Mindmaze SA, Lausanne, Switzerland) during 10 MW. These are wearable sensors that use body-attached inertial measurement units to assess spatiotemporal parameters of gait [33, 34]. Mindmaze SA analyzed data using a previously validated algorithm [35, 36] in MATLAB. To evaluate steady-state walking, the two initiation and termination cycles were discarded. We investigated change in gait cycle time (seconds), stride length (meters), and stride length variability [37], calculated using the within-person stride length standard deviation [38].

All adverse events (AE) were recorded by the physiotherapist, and sufficient information was pursued to allow a determination of the temporal and causal relationship between the AE and study interventions, as well as the outcome of the event.

Two additional scales were used to assess MindPod Dolphin usability and safety: the System Usability Scale (SUS) [39, 40] and the Simulator Sickness Questionnaire (SSQ) [41], respectively. Participants’ satisfaction was also recorded.

Adherence was calculated as participants’ compliance to the protocol. A participant was considered compliant if he/she attended at least 80% of the 36 sessions.

We used the clinical and kinematic outcomes as measures of efficacy, adverse events and SSQ as measures of safety, and adherence, SUS and participants’ satisfaction as measures of feasibility.

Statistical analysis

Sample size calculation was based on a previous study comparing the effect of different types of physiotherapy interventions on functional mobility [42]. It was estimated that a total of 26 participants, 13 per study group, would be necessary to provide 80% power to detect an effect size of 1.16 in TUG. Accounting for a 20% dropout rate, a sample of 30 participants (15 per group) was recruited.

Descriptive analyses were performed to characterize the sample using relevant summary statistics. Baseline characteristics were compared between groups. Since most variables were not normally distributed, non-parametric testing was used.

We used an intention-to-treat analysis. Any participant who was randomly allocated to a group intervention, had a full baseline assessment and at least one follow-up evaluation was included in the analysis. For our primary outcome, we compared the change from baseline to 6-week in the TUG between COM and SEQ by performing a Mann-Whitney U test. The same procedure was used regarding secondary outcomes. We used the Wilcoxon signed rank test to assess within-group differences, and the Friedman test for repeated measures to assess time-dependent effects. Effect sizes were calculated as Z statistic divided by square root of the sample size (N) [43] and were defined as small (0.20–0.49), medium (0.50–0.80), and large (> 0.80), according to Cohen’s classification [44]. The level of significance was set at 5%. All statistical analysis were performed using IBM SPSS version 29.0.

Results

Thirty-two individuals consented to participate and were assessed. Two met exclusion criteria, thus 30 were randomly assigned to one of the interventions. Two participants were lost to follow-up due to non-serious AE, one completed the 6-week and the other the 12-week assessment. Therefore, both were included in the intention-to-treat analysis. Figure 3 represents the flow diagram of participants in the trial.

Fig. 3.

CONSORT flow diagram of the trial

Baseline characteristics are summarized in Table 1. Median [IQR] age was 68.5 [58.25, 68.5] and median disease duration was 7 [5, 10.5] years. Median disease severity as measured by MDS-UPDRS part II and III was 26.5 [18.75, 39.25]. There were no statistically significant differences between groups regarding baseline characteristics.

Table 1.

Participants’ characteristics at baseline visit

	All sample (n = 30)	COM (n = 15)	SEQ (n = 15)	p value
Age (years), median [IQR]	68.5 [58.25, 68.50]	63 [54, 71]	70 [66, 76]	0.068
Male, n (%)	16 (53.3)	9 (60)	7 (46.7)	0.481
Education (years), median [IQR]	12 [4, 15]	12 [6, 15]	12 [4, 15]	0.458
Disease duration (years), median [IQR]	7 [5, 10.5]	7 [5, 8]	7 [5, 12]	0.707
LEDD, median [IQR]	795 [500, 1040]	758 [500, 970]	890 [500, 1200]	0.547
MoCA, median [IQR]	26 [25, 28.25]	27 [26, 29]	26 [25, 28]	0.254
TUG (seconds), median [IQR]	8.562 [7.324, 9.073]	8.637 [7.350, 9.293]	8.520 [7.247, 8.910]	0.885
mini-BEST test, median [IQR]	22 [21, 24]	23 [21, 25]	21 [20, 22]	0.090
HY, median [IQR]	2 [2, 3]	2 [2, 2]	2 [2, 3]	0.087
MDS-UPDRS part II and III, median [IQR]	26.5 [18.75, 39.25]	27 [22, 39]	26 [18, 40]	0.575
CGI-S, median [IQR]	3 [3, 4]	3 [3, 4]	3 [3, 4]	0.388
PGI-S, median [IQR]	3 [3, 4]	3 [3, 4]	3 [3, 4]	1

COM– Combined MindPod Dolphin with physiotherapy over 12 weeks; SEQ– Sequentially added MindPod Dolphin after 6-weeks of physiotherapy alone; IQR– interquartile range [25th, 75th percentile]; LEDD– Levodopa Daily Dose Equivalents; MDS-UPDRS– MDS-Unified Parkinson’s Disease Rating Scale; mini-BEST– mini-Balance Evaluation System Test; HY– Hoehn and Yahr; MoCA– Montreal Cognitive Assessment; CGI-S– Clinical Global Impression of Severity; PGI-S– Patient Global Impression of Severity

Primary outcome

We found no statistically significant differences between both groups in change from baseline to 6-week in TUG [estimated difference = 0.002 (95% CI -1.103; 1.273); p = 0.983] (Table 2).

Table 2.

Absolute median changes in clinical and kinematic outcomes during each study period (I to V), with results from between-group comparisons (estimated differences and confidence intervals), as well as within-group comparisons

Variables	Study period	Combined therapy (COM)			Sequential therapy (SEQ)			COM vs. SEQ
		Median change [IQR]	p-value	Effect size	Median change [IQR]	p-value	Effect size	Estimated difference	95% CI	p-value
TUG (seconds)	I	0.243 [-1.873, 1.176]	0.804	-	-0.016 [-0.810, 0.350]	0.762	-	0.002	-1.103, 1.273	0.983*
	II	-0.155 [-1.074, 0.659]	0.542	-	0.123 [-0.783, 0.650]	0.934	-	0.135	-0.671, 1.204	0.631
	III	-0.215 [-1.737, 0.906]	0.296	-	-0.333 [-1.137, 0.990]	0.720	-	0.229	-0.929, 1.377	0.541
	IV	-0.385 [-0.943, 0.729]	0.296	-	0.217 [-0.112, 0.713]	0.194	-	0.634	-0.104, 1.300	0.089
	V	-1.057 [-1.811, 0.158]	0.042	0.39	-0.180 [-0.725, 0.956]	0.839	-	0.950	-0.106, 1.763	0.081
TUG cognitive (seconds)	I	-1.176 [-4.370, 0.370]	0.107	-	-0.694 [-3.153, 0.210]	0.048	0.36	0.457	-1.700, 2.336	0.604
	II	-0.669 [-2.725, 0.277]	0.153	-	0.140 [-1.983, 0.987]	0.730	-	1.014	-0.911, 2.576	0.383
	III	-2.184 [-5.060, -0.800]	0.047	0.37	-1.733 [-2.230, 0.117]	0.110	-	1.577	-0.500, 3.700	0.176
	IV	-0.590 [-1.684, 0.553]	0.091	-	0.047 [-0.420, 1.187]	0.463	-	0.820	-0.150, 2.227	0.073
	V	-2.769 [-5.394, -1.395]	0.003	0.53	-1.268 [-2.349, 0.506]	0.078	-	2.218	0.093, 4.314	0.048
TUG motor (seconds)	I	-0.490 [-1.647, 1.116]	0.571	-	-0.607 [-1.027; 0.074]	0.083	-	-0.100	-1.593, 0.990	0.852
	II	-0.520 [-1.400, 0.338]	0.217	-	0.160 [-0.970, 1.096]	1.000	-	0.445	-0.490, 1.526	0.458
	III	-0.619 [-1.993, 0.954]	0.241	-	-0.487 [-1.740, 0.860]	0.330	-	0.171	-1.187, 1.497	0.861
	IV	-0.367 [-1.670, 0.643]	0.194	-	0.352 [-0.841, 0.732]	0.626	-	0.620	-0.196, 1.750	0.118
	V	-1.246 [-1.665, 0.629]	0.030	0.41	-0.562 [-1.252, 0.252]	0.268	-	0.562	-0.600, 1.733	0.270
mini- BEST	I	0 [-1, 1]	0.608	-	1 [1, 3]	0.002	0.54	1	0, 2	0.017
	II	0 [-1, 2.25]	0.348	-	1 [0, 2]	0.119	-	0	-1, 1	0.705
	III	1 [-0.25, 2]	0.100	-	2 [1, 4]	0.002	0.53	1	0, 3	0.125
	IV	1 [-1, 1]	0.652	-	0 [-0.25, 1]	0.672	-	0	-1, 1	0.665
	V	1.5 [0, 3]	0.020	0.44	3 [1, 3.25]	0.002	0.55	1	0, 2	0.114
MDS- UPDRS II and III	I	-1 [-6, 4]	0.682	-	-4 [-8, 2]	0.147	-	-2	-8, 6	0.533
	II	-4 [-10.25, 2.75]	0.139	-	-1 [-4, 3]	0.624	-	2	-3, 8	0.294
	III	-4 [-14.5, 4.25]	0.178	-	-2 [-12, 5]	0.140	-	1	-7, 9	0.600
	IV	-1 [-5.25, 6.5]	0.939	-	-1.5 [-6.25, 7]	0.797	-	-1	-7, 5	0.662
	V	-2 [-16.25, 4.5]	0.208	-	-2.5 [-11.5, 2]	0.065	-	-1	-9, 7	0.800
MDS- UPDRS I	I	0 [-3, 3]	0.982	-	0 [-2, 3]	0.755	-	0	-2, 3	0.835
	II	0 [-2, 2.25]	0.832	-	0 [-4, 3]	0.774	-	-1	-4, 2	0.660
	III	-0.5 [-3.25, 4.25]	0.773	-	1 [-2, 1]	1.000	-	0.5	-4, 3	0.843
	IV	-0.5 [-2.25, 1.25]	0.731	-	1 [-0.75, 2.75]	0.311	-	1	-2, 4	0.276
	V	-1 [-3.25, 4]	0.791	-	1 [-2, 3]	0.339	-	1	-2, 4	0.488
SCOPA- COG	I	2 [1, 4]	0.014	0.44	2 [-1, 3]	0.264	-	-1	-4, 1	0.337
	II	1 [0, 2.25]	0.153	-	2 [0, 3]	0.006	0.52	1	-1, 2	0.479
	III	4 [1.25, 6]	0.005	0.51	3 [1, 5]	0.004	0.50	-1	-3, 2	0.613
	IV	3 [-0.5, 4]	0.109	-	1.5 [-0.25, 4]	0.146	-	0	-3, 2	0.835
	V	5.5 [1, 7.5]	0.000	0.58	4 [1.75, 7.5]	0.005	0.50	-1	-5, 2	0.548
Gait cycle time (seconds)	I	0.007 [-0.047; 0.078]	0.359	-	-0.027 [-0.067; 0.022]	0.188	-	-0.036	-0.095, 0.013	0.165
	II	-0.039 [-0.093; 0.026]	0.119	-	-0.001 [-0.027; 0.073]	0.421	-	0.051	-0.005, 0.101	0.067
	III	-0.003 [-0.083; 0.083]	0.761	-	-0.017 [-0.086; 0.043]	0.421	-	-0.009	-0.068, 0.063	0.930
	IV	-0.006 [-0.065; 0.037]	0.391	-	0.010 [-0.032; 0.027]	0.626	-	0.019	-0.022, 0.072	0.312
	V	-0.008 [-0.093; 0.052]	0.502	-	-0.011 [-0.044; 0.054]	0.952	-	0.020	-0.043, 0.087	0.520
Stride length (meters)	I	0.045 [-0.071; 0.163]	0.489	-	0.035 [-0.036; 0.100]	0.151	-	0.025	-0.078, 0.114	0.724
	II	0.037 [-0.076; 0.161]	0.268	-	0.014 [-0.061; 0.043]	0.847	-	-0.038	-0.144, 0.066	0.383
	III	0.087 [-0.049; 0.124]	0.042	0.39	0.045 [-0.019; 0.116]	0.064	-	-0.022	-0.103, 0.077	0.793
	IV	-0.006 [-0.053; 0.113]	0.542	-	-0.040 [-0.080; 0.012]	0.153	-	-0.045	-0.122, 0.019	0.154
	V	0.104 [-0.019; 0.209]	0.020	0.43	0.009 [-0.046; 0.078]	0.426	-	-0.086	-0.196, 0.025	0.108
Stride length variability (%)	I	-1.416 [-2.570; 0.058]	0.007	0.48	0.809 [-0.625; 2.745]	0.064	-	2.739	1.082, 4.895	0.002
	II	-0.481 [-2.252; 0.418]	0.296	-	-1.695 [-4.295; -0.250]	0.000	0.59	-1.191	-3.091, 0.375	0.116
	III	-1.857 [-4.421; 0.553]	0.058	-	-0.861 [-2.193; 0.773]	0.107	-	1.307	-1.020, 3.559	0.239
	IV	0.900 [-1.613; 2.066]	0.426	-	0.493 [-0.156; 1.922]	0.241	-	-0.164	-1.812, 1.427	0.783
	V	-1.681 [-3.273; -0.626]	0.013	0.46	-0.880 [-2.036; 1.299]	0.502	-	1.327	-0.374, 3.471	0.168

COM– Combined MindPod Dolphin with physiotherapy over 12 weeks; SEQ– Sequentially added MindPod Dolphin after 6-weeks of physiotherapy alone; Study period I– first 6-week period; study period II– second 6-week period, study period III– 12-week intervention period; study period IV– “no intervention” period; study period V– 16-week study protocol. Significance set to p < 0.05 is marked in bold

TUG– Timed Up and Go test; mini-BEST– mini Balance Evaluation Systems Test; MDS-UPDRS– MDS-Unified Parkinson’s Disease Rating Scale; SCOPA-COG– Scales for Outcomes in Parkinson’s disease-COGnition

Secondary outcomes

Table 2 presents comparisons between and within-groups for primary and secondary outcomes.

From baseline to 6-week (study period I), there was a statistically significant difference between groups in change in mini-BEST favoring SEQ [estimated difference = 1 (95% CI 0; 2); p = 0.017] and stride length variability favoring COM [estimated difference = 2.739 (95% CI 1.082; 4.895); p = 0.002]. No other significant between-group differences were evident. Regarding within-group differences, COM significantly improved in SCOPA-COG (p = 0.014, r = 0.44) and stride length variability (p = 0.007, r = 0.48), whereas the SEQ showed a significant improvement in TUG cognitive (p = 0.048, r = 0.36) and mini-BEST (p = 0.002, r = 0.54).

From baseline to 12-week (study period III), we found no statistically significant between-group differences in changes. However, the COM significantly improved in TUG cognitive (p = 0.047, r = 0.37), SCOPA-COG (p = 0.005, r = 0.51) and stride length (p = 0.042, r = 0.39), while SEQ significantly improved in mini-BEST (p = 0.002, r = 0.53) and SCOPA-COG (p = 0.004, r = 0.50).

Between 12 and 16-week assessments (study period IV), there were no between or within-group significant differences.

From baseline to 16-week (study period V), there was a statistically significant difference in TUG cognitive with a greater improvement in COM when compared to SEQ [estimated difference = 2.218 (95% CI 0.093; 4.314); p = 0.048]. Regarding within-group differences in COM, a significant improvement was found in TUG (p = 0.042, r = 0.39), TUG cognitive (p = 0.003, r = 0.53), TUG motor (p = 0.030, r = 0.41), mini-BEST (p = 0.020, r = 0.44), SCOPA-COG (p = 0.000, r = 0.58), stride length (p = 0.020, r = 0.43), and stride length variability (p = 0.013, r = 0.46), whereas the SEQ presented a significant improvement in mini-BEST (p = 0.002, r = 0.55), and SCOPA-COG (p = 0.005, r = 0.50).

The ‘overall study effect’ for the pooled sample (n = 30) (Additional file, Table 1) was a significant improvement in TUG cognitive (p = 0.010, r = 0.33), mini-BEST (p = 0.004, r = 0.36) and SCOPA-COG (p = 0.010, r = 0.33), from baseline to 6-week. During the 12-week study intervention, we found a significant improvement in TUG cognitive (p = 0.009, r = 0.34), mini-BEST (p = 0.000, r = 0.44), MDS-UPDRS part II and III (p = 0.043, r = 0.26), SCOPA-COG (p = 0.000, r = 0.49), as well as in stride length (p = 0.007, r = 0.35) and stride length variability (p = 0.011, r = 0.33).

Time-dependent effects

There was a significant main effect of time in COM for TUG cognitive (p = 0.001), SCOPA-COG (p = 0.000), patient and clinical global impression of improvement (p = 0.018, p = 0.011, respectively), as well as for stride length variability (p = 0.014) (Supplementary Table 2). In SEQ, there was a main effect of time for mini-BEST (p = 0.000), SCOPA-COG (p = 0.000), and stride length variability (p = 0.014).

Adverse events

In total, 309 AEs were reported over the 16-week study duration (Table 3), none of which were serious. Two AEs led to study discontinuation: one due to shoulder pain in the most affected side of PD possibly linked to the use of MindPod Dolphin, and another due to an unrelated fall.

Table 3.

Reported adverse events

Adverse event	N (% of attended sessions)	Combined therapy (COM) N (% of attended sessions)			Sequential therapy (SEQ) N (% of attended sessions)
		0-6week	6-12week	12-16week	0-6week	6-12week	12-16week
Discomfort/ pain	105 (6.5)	30 (5.9)	13 (3.0)	1	16 (7.3)	42 (9.3)	3
Fatigue/ tiredness	39 (2.4)	8 (1.6)	8 (1.8)	0	6 (2.7)	17 (3.7)	0
Frustration	23 (1.4)	13 (2.6)	5 (1.1)	0	0 (0)	5 (1.1)	0
Falls	16 (1.0)	4 (0.8)	6 (1.4)	0	2 (0.9)	3 (0.7)	1
Dyskinesias	14 (0.9)	2 (0.4)	5 (1.1)	0	2 (0.9)	5 (1.1)	0
Irritability/ crying	14 (0.9)	4 (0.8)	8 (1.8)	0	1 (0.5)	1 (0.2)	0
Motor fluctuations	14 (0.9)	2 (0.4)	2 (0.5)	0	3 (1.4)	7 (1.5)	0
Sleeping disturbances	13 (0.8)	5 (1.0)	3 (0.7)	0	3 (1.4)	2 (0.4)	0
Nausea/ vomiting	11 (0.7)	4 (0.8)	7 (1.6)	0	0 (0)	0 (0)	0
Tremor	11 (0.6)	5 (1.0)	1 (0.2)	0	0 (0)	5 (1.1)	0
Hypotension	10 (0.6)	4 (0.8)	4 (0.9)	0	1 (0.5)	1 (0.2)	0
Dystonia	7 (0.4)	4 (0.8)	2 (0.5)	0	0 (0)	1 (0.2)	0
Anxiety	5 (0.3)	2 (0.4)	0 (0)	0	0 (0)	3 (0.7)	0
Eye discomfort	4 (0.2)	3 (0.6)	0 (0)	0	0 (0)	1 (0.2)	0
Headache	4 (0.2)	1 (0.2)	0 (0)	0	0 (0)	3 (0.7)	0
Cramps	3 (0.2)	2 (0.4)	0 (0)	0	1 (0.5)	0 (0)	0
Dizziness	3 (0.2)	1 (0.2)	2 (0.5)	0	0 (0)	0 (0)	0
Dry mouth	2 (0.1)	1 (0.2)	1 (0.2)	0	0 (0)	0 (0)	0
Hypertension	2 (0.1)	0 (0)	1 (0.2)	0	0 (0)	1 (0.2)	0
Malaise	2 (0.1)	0 (0)	1 (0.2)	0	0 (0)	1 (0.2)	0
SARS-CoV-2 infection / vaccine reaction	2 (0.1)	0 (0)	0 (0)	0	1 (0.5)	0 (0)	1
Cold extremities	1 (0.06)	0 (0)	0 (0)	0	0 (0)	1 (0.2)	0
Injury	1 (0.06)	0 (0)	0 (0)	0	1 (0.5)	0 (0)	0
Urinary problems	1 (0.06)	0 (0)	0 (0)	0	1 (0.5)	0 (0)	0
Odynophagia	1 (0.06)	1	0 (0)	0	0 (0)	0 (0)	0
Syncope	1 (0.06)	1	0 (0)	0	0 (0)	0 (0)	0
Total	309 (19.1)	97 (19.2)	69 (15.8)	1	38 (17.4)	99 (21.8)	5

COM– Combined MindPod Dolphin with physiotherapy over 12 weeks; SEQ– Sequentially added MindPod Dolphin after 6-weeks of physiotherapy alone; Total attended sessions = 1617; Attended sessions in COM = 944 (0–6 week = 506 sessions, 6–12 week = 438 sessions); Attended sessions in SEQ = 673 (0–6 week = 219 sessions, 6–12 week = 454 sessions)

AEs occurred in 19.1% of total physiotherapy and MindPod Dolphin sessions. The most frequent AEs included pain/discomfort (n = 105, 6.5% of total sessions), fatigue/tiredness (n = 39, 2.4%) and feelings of frustration (n = 23, 1.4%). Among participants experiencing pain/discomfort, back pain was the most prevalent (n = 30). Of all AEs, 55% (n = 170) were mild, 36.6% (n = 113) were deemed unrelated or unlikely related to study intervention, and 62.1% (n = 192) were considered at least possibly related to study intervention.

Regarding symptoms of cybersickness, median SSQ after the 1st week using MindPod Dolphin was 7.48 [0, 14.96], and at the end of the 6 and 12-week was 14.85 [7.48, 27.12] and 16.83 [2.81, 35.53], respectively.

Adherence

Overall, 20 participants (67%) were compliant with the study protocol, 12 from the COM and 8 from the SEQ. Reasons for no attendance were transportation problems (n = 2), work duties (n = 1), AE unrelated to intervention and/or medical appointments (n = 3), AE possibly related to intervention (n = 1), unspecified reason (n = 2), and more than one reason (n = 1).

System-related measures

We measured performance by the number of completed levels. COM completed a median [IQR] of 184 [139, 343] while SEQ completed 115 [55, 145] levels (p = 0.001). Participants spent a mean of 45 ± 5.63 minutes (COM = 44.35 ± 6.19; SEQ = 47.36 ± 4.76; p = 0.147) on game. Six participants had to perform some sessions while seated due to fatigue (n = 3), back pain (n = 1), postural instability (n = 1) and somnolence (n = 1).

Regarding usability, the SUS median [IQR] score after the 1st week was 77.5 [65, 85] and at the end of the study was 75 [58.1, 89.4] (above average) [45]. We also used a Likert-scale to evaluate participants’ satisfaction with the Mindpod Dolphin system, which was rated from 1 (“Extremely satisfied”) to 7 (“Not satisfied”). After 1 week using the system, 29 (96.7%) participants were at least moderately satisfied, and after the end of the intervention, 25 (83.3%) were still at least moderately satisfied.

Discussion

In this randomized controlled trial, we investigated the feasibility, safety and efficacy of adding an immersive exergaming experience, neuroanimation in the form of MindPod Dolphin, to an intensive physiotherapy program for the management of functional mobility in people with PD. Our study shows that both interventions are effective in improving different measures of functional mobility, motor function, balance and cognition following six and 12 weeks of training, with a tendency for the neuroanimation to improve functional mobility more in dual-task conditions and to have more sustained long-term effects.

The MindPod Dolphin is an animation-based exergame which provides an immersive experience and a markerless motion capture. The user makes large amplitude 3D arm movements that control the 2D movements of the animated cetacean. This whole-body exergame combines gorgeous visuals, cognitive challenge, and cardiovascular exercise, along with reward-based learning of a new skill. Thus PD patients are encouraged to practice multiple cognitive and motor capacities simultanously. We are not aware of another full-body exergaming system that was specifically designed for neurological patients.

Our study design enabled us to investigate the effect of different combinations of specialized intensive physiotherapy with this neuroanimation-based exergaming. First, we compared the effect of a PD specialized physiotherapy alone to its combination with MindPod Dolphin in a high-dose program lasting six weeks. Second, we prolonged this high-dose program to 12 weeks and compared it with a more sequential training program. Third, we evaluated retention effects after a 4-week period without therapy.

We could not demonstrate efficacy by our primary outcome. In addition, at six weeks, PD specialized physiotherapy alone was superior in improving mini-BEST, which measures different aspects of balance and gait [28]. Differences in baseline scores could explain this result, since the group that got physiotherapy alone in the first six weeks was more impaired, providing more room for improvement. In addition, given this greater degree of gait and balance impairment in the SEQ group, there might have been more emphasis on the functional components measured by the mini-BEST. However, we can also speculate that adding one-hour of complex gaming on the same day participants were still learning the basics via high-dose physiotherapy might have overloaded them and prevented expression of immediate gains. Indeed, while still able to learn, people with PD may take time and additional practice to learn complex motor sequences [46], which in fact we observed in our participants– the benefits of a high-dose combination therapy were only expressed in the subsequent six weeks.

When considering the whole 12-week intervention period, we showed that neuroanimation-based exergaming combined with physiotherapy in a high-dose program is effective in people with PD for improving both cognition in a dual-task and stride length. Sustained improvements were observed in several measures of functional mobility under dual-task conditions, as well as balance, gait, cognition, and stride-length variability. In addition, the sequential training program led to sustained improvements in balance and cognition. When comparing both interventions, we found that neuroanimation-based exergaming over the whole 12-weeks was superior from baseline to 16-week in improving TUG cognitive. Altogether, these results demonstrate that both the combined and sequential interventions with MindPod Dolphin seem to be beneficial for people with PD with mild-to-moderate disease, with better outcomes compared to physiotherapy alone in the second six weeks of the program, as well as better long-term effects once the program had ended.

There is evidence showing that adding complex conditions that demand motor learning and dual-tasking may result in better executive function, reduced number of falls and modulation of prefrontal cortex activation [9, 47]. MindPod Dolphin may have added to the benefits of exercise in the long-term by cognitively engaging participants in tasks that demanded sustained attention, task shifting, visual-spatial skills, eye-hand coordination and implicit learning [48], while playing in an enjoyable oceanic environment. This strong cognitive component of neuroanimation-based exergaming may have contributed to the more prolonged effect on TUG cognitive. Thus, tests requiring dual-tasking may be more sensitive to detect changes in people with PD undergoing exergaming-based rehabilitation. Future studies should consider TUG cognitive as a primary outcome.

Regarding kinematic results, stride length variability distinguished the two interventions at the 6-week evaluation, favoring combined MindPod Dolphin and physiotherapy. Gait variability has shown to be increased in people with PD from early stages, and to be correlated with measures of functional mobility [37], possibly reflecting loss of gait automaticity and compensatory strategies [33, 49]. Even though not specifically designed to train gait, we speculate that the cognitive engagement plus whole body movement promoted by MindPod Dolphin may contribute to an improvement in gait kinematics.

Since our interventions exposed participants to a specialized physiotherapy protocol for PD focused on physical capacity and functional training, delivered by two highly experienced physiotherapists, we expected that both groups would have positive results from physiotherapy [10], which was confirmed by the improvement in several motor and functional mobility measures. The protocols for the combined versus sequential MindPod Dolphin groups differed slightly, with the former having a two-hour program three times per week and the latter a more progressive exercise intensity program over 12 weeks. These differences in dose and intensity may have influenced our results and long-term outcomes. It would be interesting to consider a longer intervention program in future studies with a control group that got physiotherapy alone, in order to test the hypothesis that physiotherapy is focused on training specific functional tasks, while exergaming might improve indirect motor and cognitive capacities, with a cumulative effect on clinical and gait kinematic outcomes transferable to daily activities [50].

We demonstrated that MindPod Dolphin used in a combined program is feasible and safe to individuals with PD, as it did not result in serious AEs; AEs were mostly mild. Regarding cybersickness symptoms, we used the SSQ to quantify symptoms related to the neuroanimation experience. Based on reported classification scores to quantify cybersickness [51], our participants reported “minimal” to “concerning” symptoms. The SSQ has been increasingly used in VR literature, however current classification is still based on data from simulator military pilots, and therefore its interpretation is questionable [52]. Besides, the scale includes symptoms commonly present in people with PD (e.g. nausea, fatigue), which could be present before study initiation. Measuring the SSQ before intervention could mitigate this uncertainty.

The system was considered easy to use, most participants were at least moderately satisfied with the MindPod Dolphin throughout the duration of the study, and the observed progression through the levels further emphasize that they were able to positively engage in the game.

Overall compliance with the study protocol was modest. Perhaps most of the participants were still active in their work life, and a protocol including 6 h of exercise per week might be difficult to attain. Yet, COM was more compliant than the SEQ group, which is consistent with the notion that gamified therapeutic interventions are more attractive and engaging than conventional rehabilitation programs. It appears that innovative technologies are needed to promote increased adherence to multidisciplinary programs in PD.

Our study has some limitations. First, we studied individuals with mild to moderate PD, without reported falls and cognitive impairment, which may have limited the magnitude of effect in both motor and cognitive outcomes. Second, our included population was heterogeneous, as demonstrated by the distribution of age, disease duration and severity, and our analysis did not consider those covariates. This limitation is partially attenuated by the fact that baseline characteristics were not statistically different between groups. Third, there was modest overall compliance, which was lower in the SEQ, a possible source of bias. Fourth, the effect of the non-pharmacological intervention was against a background of dopaminergic medication. It might be the case that it is the combination that is required to see the effects reported here. Finally, we did not include a control group with ‘no intervention’, thus we can not isolate the effect of neuroanimation. As VR/serious games and physiotherapy programs seem to target different motor learning processes [9, 50], we believe that neuroanimation-based exergaming may be better suited as a complement to physiotherapy rather than as a substitute to it, which is why we first decided to compare adding it to physiotherapy with physiotherapy alone.

Conclusion

Our study showed that it is feasible and safe to use a neuroanimation-based exergaming experience in the form of MindPod Dolphin in people with mild-to-moderate PD. MindPod Dolphin, when added to a PD specialized intensive physiotherapy program, was effective in improving several cognitive, motor and gait kinematic parameters. Superiority of the combined therapy with MindPod Dolphin was only demonstrated in the second six weeks and in follow-up. This is of great interest as it suggests that intensive programs should go on longer than six weeks, and that more complex motor tasks, which challenge skill acquisition and cognition, may only start to have an impact after more basic capacities have first been addressed by physiotherapy alone. Future work will be needed to optimize high-dose location-based training programs for PD, but we can anticipate that combinations of immersive full-body gaming challenges with conventional physiotherapy will yield better results than either alone.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

DPS, RBM, FPF, JWK and JJF participated in the conception and study design. FPF and TL conducted training sessions. DPS and RC participated in data collection. DPS, RBM, JWK and JJF contributed to data processing, analysis and interpretation of results. RA conducted gait kinematic analysis. DPS drafted the manuscript and all authors made critical revisions and contributed to the writing of the manuscript.

Funding

MindMaze provided CNS-Campus Neurológico with the MindPod-Dolphin serious video gaming system, Physilog gait recording and analysis systems, and a part of the funding for the trial. The authors received no financial support for the research, authorship, and/or publication of this article.

Data availability

Raw data are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the CNS Ethics Committee (Ref. 03-2020) and performed in accordance with national legislation and the Declaration of Helsinki. Individuals who agreed to participate were asked to provide written informed consent.

Consent for publication

All participants consented to these data being published.

Competing interests

The authors declare no competing interests.DPS was granted by Santa Casa da Misericórdia de Lisboa (“Prémio João Lobo Antunes”) for a project unrelated to the present manuscript. RA was an employee and shareholder of MindMaze SA. JWK has consulted for and has options in MindMaze, Emerja, and Meta. JJF has provided consultancy for AbbVie, BIAL, Biogen, Lundbeck, and Sunovion, has received grants from Angelini, Novartis, Medtronic, AbbVie, Zambon, BIAL, Biogen, and Grunenthal, and has received speaker fees for BIAL, Ono, SK Chemical, Abbvie and Infucure. All other authors have no financial disclosures.