跳到正文
原文
Nature 健康护理主题·· 3 天前AI 评分42

老年人对重复虚拟现实刺激的快速姿势控制适应

Rapid postural control adaptations to repeated virtual reality stimulation in older adults

AI 导读

老年人重复观看同一VR影片5次后,需跌倒防护支持的参与比例从首次的45%降至第5次的10%(p=0.013),且前后及侧向姿势能量在各频谱范围内逐次下降(均p≤0.011/0.004)。该结果提示重复沉浸式视觉扰动可诱导中枢感觉重加权与姿势稳定性适应。

正文

Introduction

Postural control during standing is a complex process that depends on the integration of sensory input, central nervous system (CNS) processing, and motor output1. The CNS plays a central role in maintaining balance by utilizing feedback and feedforward mechanisms, as well as stiffening strategies to stabilize the body in response to both expected and unexpected perturbations2. In older adults, postural control is often impaired due to age-related decline in sensory input, CNS integration, and motor function3. This decline can lead to instability and a significantly increased risk of falls, a major health concern in aging populations4. Sensory input, particularly from visual, vestibular, and proprioceptive systems, and its accurate integration within the CNS are essential for effective balance regulation5. When these sensory signals become distorted or inconsistent, such as in acute vestibular loss6 or motion sickness7, the CNS must reweight sensory information to maintain stability. Conditions such as peripheral vestibular lesions and central disorders can impair spatial orientation, especially in visually distorted environments8,9. Furthermore, disorders like multisensory dizziness10,11 and persistent postural-perceptual dizziness (PPPD)12 may result in excessive visual dependency for balance. In these cases, it becomes crucial for the CNS to identify sensory mismatches and adapt by extracting more reliable information from alternative sources.

Recent technological advancements, particularly in immersive Virtual Reality (VR), have enabled more controlled manipulation of visual input to assess its impact on postural control, also outside the laboratory setting. Moreover, novel research suggests that VR-based rehabilitation programs may help address sensory mismatch, especially in cases where visual input plays a disproportionate role13. Repeated VR exposure has been shown to induce central sensory reweighting, with increased resilience to visual perturbations observed after multiple sessions14,15,16.

In younger adults, a study using posturography and electromyography (EMG) has demonstrated significant changes in postural stability and muscle activity in response to visual challenges via VR17. In that recent study we found significantly reduced stability (increased energy use) and increased co-contraction of postural muscles due to visual stimulation (a rollercoaster video) using VR-headset. Moreover, we also found significant adaptations over repeated VR sessions, showing both increased stability and reduced muscle co-contractions. However, corresponding data on older adults remains limited. Understanding how older people respond to immersive visual perturbations is crucial for developing assessment tools and fall-prevention strategies tailored for an aging population.

The primary aim of this study was to investigate how older adults respond to challenging visual environments presented through immersive VR, in terms of postural stability and muscle activity. A secondary aim was to examine whether repeated exposures induce adaptive changes, whether muscle activity patterns, such as co-contraction, correlate with postural stability during the different conditions and to describe the occurrence of incidences requiring fall-preventing support.

Results

Occurrence of incidents requiring fall-preventing support

Fig. 1

Participants needing fall-preventive support during each test condition [%]. None of the participants needed support during the control tests, n = 31.

During VR session 1, 45% of the participants needed support from the operator to avoid falling (Fig. 1). The number of incidents significantly declined during the repeated sessions (p = 0.013). During VR session 5, 10% of the participants needed support to avoid falling. None of the participants needed support during the control tests.

Posturography assessments

Effects on postural stability from repeated VR sessions

Watching the same VR movie in repeated sessions caused the postural energy to decline by each session in both anteroposterior (p ≤ 0.011) and lateral (p ≤ 0.004) directions within all spectral ranges (Table 1).

Table 1 Effects on postural stability by repeated VR sessions.

Full size table

Initial effects of VR and effects of adaptation to VR on postural stability

Significantly more energy was used during VR session 1 compared to during the control test with eyes open (within-subjects analyses), in both anteroposterior (p < 0.001) and lateral (p < 0.001) directions (Fig. 2; Table 2). In the anteroposterior direction, total energy use was 1098% higher, with low-frequency energy increasing by 848% and high-frequency energy by 1694%. In the lateral direction, total energy use was 1962% higher, with low-frequency energy increasing by 596% and high-frequency energy by 5833%.

Table 2 Initial effects of VR and effects of adaptation to VR on postural stability.

Full size table

Fig. 2

Postural stability during repeated VR sessions and during the control tests. The performance recorded during the five repeated VR sessions are presented as bars (mean) and whiskers (SEM) values of normalized total (A), low frequency (C) and high frequency (E) torque variance. The performance during the eyes closed and eyes open control tests are presented as normalized total (B), low frequency (D) and high frequency (F) torque variance.

The energy used across the five VR sessions (i.e., energy used during VR session 1 vs. during VR session 5) decreased significantly in both anteroposterior (p ≤ 0.002) and lateral (p < 0.001) directions, reflecting the effect of adaptation (Table 2). In anteroposterior direction, total energy use was reduced by 61%; low frequency energy was reduced by 67%, and high frequency energy was reduced by 55%. In lateral direction, total energy use was reduced by 64%; low frequency energy was reduced by 46%, and high frequency energy was reduced by 71%.

The energy used during VR session 5 was significantly higher compared to the control test with eyes open in both anteroposterior (p ≤ 0.003) and lateral (p < 0.001) directions (Table 2). In anteroposterior direction, total energy use was 372% higher during VR; low frequency energy was 215% higher and high frequency energy was 711% higher. In lateral direction, total energy use was 636% higher during VR; low frequency energy was 274% higher, and high frequency energy was 1647% higher.

Effects on postural stability of vision during the control tests

The high frequency energy used in anteroposterior and lateral direction was significantly higher (within-subjects analyses) with eyes closed than with eyes open by 91% and 82%, respectively, (p < 0.001) (Fig. 2; Table 3).

Table 3 Effects on postural stability of vision during the control tests.

Full size table

EMG assessments

Effects on CCI from repeated VR sessions

Mean Co-Contraction Index (CCI) (p < 0.001) and standard deviation (SD) CCI (p < 0.001) decreased significantly across the five VR sessions in the three muscle compositions investigated: TibAnt vs. MedGas, TibAnt vs. PerLong and GlutMed L vs. GlutMed R (Table 4).

Table 4 Effects on CCI from repeated VR sessions.

Full size table

Initial effects of VR and effects of adaptation to VR on CCI

Mean CCI (p ≤ 0.001) and SD CCI (p < 0.001) increased significantly (within-subjects analyses) during VR session 1 compared to during the control test with eyes open in the three muscle compositions investigated (Fig. 3; Table 5). Mean CCI increased 229% for TibAnt vs. MedGas; increased 377% for TibAnt vs. PerLong and increased 72% for GlutMed L vs. GlutMed R. Standard deviation of the CCI increased 376% for TibAnt vs. MedGas; increased 570% for TibAnt vs. PerLong, and increased 166% for GlutMed L vs. GlutMed R.

Fig. 3

CCIs during repeated VR sessions and during the control tests. The CCIs between three muscle compositions during the five repeated VR sessions are presented as bars (mean) and whiskers (SEM) values of mean CCIs (A) and SD CCIs (C). The performance during the control tests with eyes closed and eyes open are presented as mean CCIs (B) and SD CCIs (D).

Table 5 Initial effects of VR and effects of adaptation to VR on CCI.

Full size table

Mean CCI (p < 0.001) and SD CCI (p < 0.001) were significantly reduced across the five VR sessions for the three muscle compositions investigated, reflecting the effect of adaptation (Table 5). Mean CCI decreased 31% for TibAnt vs. MedGas; decreased 39% for TibAnt vs. PerLong, and decreased 26% for GlutMed L vs. GlutMed R. Standard deviations of the CCIs decreased 34% for TibAnt vs. MedGas; decreased 41% for TibAnt vs. PerLong, and decreased 37% for GlutMed L vs. GlutMed R.

Mean CCI (p < 0.001) and SD CCI (p < 0.001) were significantly higher during VR session 5 compared to during the control test with eyes open in the three muscle compositions investigated (Table 5). Mean CCIs were 128% higher for TibAnt vs. MedGas; 189% higher for TibAnt vs. PerLong, and 28% higher for GlutMed L vs. GlutMed R. Standard deviations of the CCIs were 216% higher for TibAnt vs. MedGas; 293% higher for TibAnt vs. PerLong, and 68% higher for GlutMed L vs. GlutMed R.

Effects on CCI of vision during the control tests

Within-subjects analyses revealed that when the participants stood with eyes closed during the control tests mean CCI (p ≤ 0.037) and SD CCI (p ≤ 0.029) significantly increased in the three muscle compositions investigated compared with when the participants stood with eyes open (Fig. 3; Table 6). Mean CCIs increased 44% for TibAnt vs. MedGas; increased 58% for TibAnt vs. PerLong and increased 20% for GlutMed L vs. GlutMed R. Standard deviations of the CCIs increased 44% for TibAnt vs. MedGas; increased 58% for TibAnt vs. PerLong and increased 11% for GlutMed L vs. GlutMed R.

Table 6 Effects on CCI of vision during the control tests.

Full size table

Correlation analysis

Correlation between postural stability and CCI during VR

The analyses identified 21 significant (p ≤ 0.006) moderate to strong positive correlation relationships during VR in the total of 24 correlations performed between postural stability and mean CCI parameters (Table 7). Positive correlation relationships represent that increases in postural energy use were related to increased use of muscle co-contractions. The evaluations revealed 24 significant (p < 0.001) moderate to strong positive relationships during VR between postural stability and SD CCI in the 24 correlations performed.

Table 7 Correlation between postural stability and CCI during VR

Full size table

Correlation between postural stability and CCI during control tests

The analyses identified 5 significant (p ≤ 0.034) moderate positive correlation relationships during the control tests in the total of 24 correlations performed between postural stability and mean CCI parameters (Table 8). The evaluations revealed 9 significant (p ≤ 0.036) moderate positive relationships during the control tests between postural stability and SD CCI in the 24 correlations performed.

Table 8 Correlation between postural stability and CCI during control tests

Full size table

Discussion

The primary aim of this study was to investigate how older adults respond to visually complex environments presented via immersive virtual reality (VR), with a focus on postural stability and muscle activation patterns. A secondary aim was to evaluate whether repeated exposures to the same VR stimulus would induce sensorimotor adaptation, and whether these changes would manifest in both posturographic and muscle co-contraction indices (CCI) as well as change the occurrence of incidences requiring fall-preventing support.

The findings demonstrated that the initial exposure to VR significantly disturbed postural control and increased muscle co-contractions. However, over the course of five repeated sessions, participants exhibited substantial adaptation, as evidenced by fewer fall-preventing interventions by the operator, reduced use of energy for postural stability across all spectral bands, and decreased muscle co-contractions of all muscle pairs accessed. Significant moderate to strong positive correlations emerged between the postural stability measures and the CCIs, particularly during VR sessions. Thus, improved postural stability measures and lower CCI for the muscle pairs indicate that with less energy use, less co-contraction was needed to maintain balance.

These findings are generally in line with a similar study from our research group, on young adults performing a similar roller coaster virtual reality stimulation task17. However, some important differences need to be highlighted. Despite being exposed to a less challenging roller coaster ride, almost half of the older adults required fall-preventive support, whereas none of the younger participants did. Also, the increased energy used in the stability measures and the increase in CCI during VR compared to control tests were much higher in older adults, compared with the younger adults17. Moreover, the correlations between the stability measures and the CCI measures showed higher correlations compared to the younger adults, both during the VR-stimulation and the control tests17. This indicates that visual perturbation is more challenging to postural stability among older adults, and that they use more neuromuscular co-contraction, indicating a higher dependence on a stiffening strategy compared to younger adults. Similar results have been reported in previous research showing higher postural challenges and muscle co-contractions among older adults compared to young adults in various postural control tasks18,19.

During the first VR session, nearly half of the participants (45%) required assistance from the test leader to prevent falling. This high incidence underscores the destabilizing effect of immersive, dynamic visual stimuli in adults of older age. Quantitatively, torque variance values indicated a dramatic increase in energy expenditure across all frequency domains and directions compared to the eyes-open control condition. Concurrently, muscle co-contraction index increased significantly in all measured muscle pairs, reflecting a compensatory stiffening strategy in response to the destabilizing visual environment18. Consistent with these results, increased co-contraction has also been reported among older people with reduced postural control during a surface perturbation balance task20.

Despite the substantial initial destabilization, repeated VR exposure led to clear adaptation effects. In session five, only 10% of participants required support, and both postural sway and CCI measures had significantly decreased. These improvements were consistent across both anteroposterior and lateral directions, and across both low- and high-frequency energy use components. Nevertheless, even after five sessions, energy use and co-contraction levels remained significantly higher than in the baseline control test, suggesting that full adaptation was not yet achieved. Importantly, the rate of adaptation was relatively rapid, indicating that short-term exposure can be effective in inducing improvements of postural control strategies in older adults. Similar rapid adaptation has been shown in a repeated surface perturbation task among a majority, but not all participants20. The non-adapters in that study were of older age, had less muscle strength and longer reaction time; also, they used more co-contraction during the last perturbation trial. Future research should further investigate specific postural control adaptations to various sensory and physical perturbations among older adults, and factors facilitating and inhibiting these adaptations. This knowledge is of major importance for fall risk prevention among people of older age. Moreover, the amount of training needed for retention and transfer effects should be further investigated. A systematic review points out that a total dose of at least 4.5 h of training is necessary to achieve improvements in balance21.

The data showed that lateral postural control was more disrupted than anteroposterior control during initial VR exposure, particularly in the high-frequency range. This suggests that rapid corrective movements, likely required for lateral balance adjustments, were especially challenged by the dynamic VR content (i.e., the rapid left and right 180 degree turns in the roller-coaster movie the participants watched). This interpretation is supported by the muscle co-contractions around the ankles showing higher increase for TibAnt vs. PerLong, which are considered important for mediolateral stability, compared to TibAnt vs. MedGas, which are considered more important for anteroposterior stability. There was, moreover, a significant increase in GlutMed L vs. GlutMed R, which contributes important mediolateral stability around the hips.

A key finding was the strong positive correlation between postural energy use and muscle co-contraction during VR. This relationship was consistent across all spectral bands and in both movement directions, with the strongest correlations observed for the muscle pairs controlling the ankles, i.e., TibAnt vs. PerLong and TibAnt vs. MedGas. The mean CCIs of the GlutMed L vs. GlutMed R were significantly correlated only with the lateral stability measures, which is in line with the direction specific effects mentioned above. Notably, correlation coefficients were higher for SD CCI than for mean CCI for all muscle pairs, suggesting that variability in muscle activation may be more closely tied to instability than the overall co-contraction levels. These findings highlight the tight coupling between neuromuscular and postural responses in complex sensory environments.

Interestingly, VR-induced instability and muscle activity exceeded that observed during eyes-closed control tests. For example, high-frequency lateral energy use increased by over 5800% during the initial VR session, compared to an 82% increase in the eyes-closed condition. Similarly, TibAnt vs. PerLong co-contraction increased by 377% in VR but only 58% with eyes closed. This suggests that inconsistent or misleading visual input, such as that in VR, may pose a greater challenge to postural control than the complete absence of visual input, which is in line with previous results15,22,23,24.

The pronounced adaptation observed in this study supports the use of VR as a potential tool in balance rehabilitation for older adults. Importantly, the specific VR content used (a roller-coaster stimulation) produced much greater destabilization than calmer content, such as a virtual supermarket environment tested in prior studies15,22. This indicates that the level of difficulty of the VR-based interventions can and should be adjusted to the functional capacity of the individual. Gradual progression and repeated exposure may help the central nervous system to reweigh reliance between different sensory inputs and improve balance, even in the presence of misleading visual cues. However, more research is needed clarifying whether VR affects the reliance on information from different sensory systems.

One important limitation of this study is its focus on the short-term effects of VR exposure, specifically, the effects achievable through repeated training performed over a single hour within one day. This approach does not reflect conventional rehabilitation protocols, which are typically conducted over several weeks, and therefore cannot be expected to capture the full potential of repeated training. Despite this important limitation, our findings suggest that this intensive, repeated training approach produced a substantial improvement in postural stability during VR exposure, as indicated by a reduction in energy use of approximately 55% in the anteroposterior direction and 71% in the lateral direction within the most affected spectral bandwidths.

It should be acknowledged, however, that when the values do not return to baseline levels, this may indicate a sustained effect within the timeframe assessed. Thus, five repeated VR training sessions were not sufficient to render older adults fully resilient to immersive visual environments. For rehabilitation strategies to be effective in real-world settings, improvements in sensorimotor control strategies need to be consolidated over time and transferred to everyday activities, allowing the benefits to be retained during subsequent exposure to similar postural challenges21,22. Future studies should therefore investigate whether VR-based postural control training can produce sustained adaptations that are retained over time and transferred to everyday activities, particularly in clinical populations with vestibular, neurological, or multisensory impairments, who may have greater difficulty coping with situations in which visual information is conflicting or unreliable due to their underlying conditions.

Another limitation of the study is that five participants (16%) were excluded from the posturography/EMG analyses because they were unable to remain on the force platform during the sessions or discontinued participation due to discomfort. Although these instability events gradually decreased in frequency and severity across the VR sessions, indicating that even participants with poorer postural stability appeared to respond positively to the training, the recordings were sufficiently distorted by the near-fall events to preclude analysis. Consequently, the findings may primarily reflect the performance of a subset of older adults with better postural stability or a lower propensity to exhibit pronounced startle responses to immersive visual stimuli.

Finally, we acknowledge that our use of the terminology “adaptation” may not have sufficiently reflected the distinctions between the underlying processes. To our understanding, “adaptation” is commonly used in the literature to describe relatively swift systematic changes in responses that can be objectively quantified. However, the mechanisms underlying such changes are not always fully understood, making it difficult to determine unequivocally whether the observed effects reflect habituation, sensory adaptation, or other forms of sensorimotor adjustment. As our study was not designed to distinguish between these underlying mechanisms, we have used the terminology “adaptation” to describe the observed changes more neutrally, without attributing them to a specific process.

Conclusion

Older adults exhibit clear impairments in postural control and increased neuromuscular activation when exposed to visually complex VR environments. However, rapid adaptation occurs with repeated exposure, leading to improved stability and reduced co-contraction. These findings suggest that immersive VR can serve both as a diagnostic tool and a rehabilitative intervention challenge, with careful consideration to stimulus design and individual capacity. Further research is needed to explore the feasibility and the long-term and clinical applications of VR-based balance training among older adults.

Materials and methods

Ethical consideration

The study was performed in accordance with the latest Declaration of Helsinki and received ethical approval from the Ethical Review Authority in Sweden (Dnr:2023-00353-01). All participants gave their written consent prior to participation.

Study design

This cross-sectional experimental study used a repeated measures design to evaluate the participants’ reaction as well as adaptation to VR video stimulus on postural stability and muscle activity in postural muscles. The method is the same used in earlier studies17,25 and is described in detail below.

Participants

The recruitment of participants was conducted through social media platform advertisements (Facebook) and through contacts with colleagues, friends and family. Information about the study, inclusion criteria, and time/place for data collection were published on Lund University’s webpage. Those who accepted to participate in the study were referred to the Movement and Reality Laboratory (MoRe-lab) at Lund University for the testing.

Thirty-two subjects were recruited to participate in the study. The inclusion criteria were age above 65 years, being able to read and write in Swedish and to consider themselves healthy and had no prior history of diseases that could affect their current balance, e.g., a known history of vertigo or neurological diseases. The participants were screened for health disorders by using a custom-made questionnaire, e.g., whether they had reduced balance, prior injury in the lower extremity the last year. The questionnaire has been used in previous studies22.

Data from thirty-one participants of the original 32 participants were included when investigating the occurrence of incidences requiring fall-preventing support (35% woman, mean age 75.4 years (SD 5.0 years), mean height 1.74 m (SD 0.09 m), mean weight 74.0 kg (SD 10.8 kg), mean BMI 24.6 (SD 3.5)). One participant from the original group was excluded due to inability to perform VR tests.

Data from twenty-five participants of the original 32 participants were included in the posturography/EMG investigation (44% woman, mean age 74.8 years (SD 3.5 years), mean height 1.72 m (SD 0.09 m), mean weight 73.4 kg (SD 11.3 kg), mean BMI 24.8 (SD 3.6)). Two participants from the original group were excluded due to technical issues and five participants due to their inability to properly perform the tests, e.g., they stepped off the force platform during one or more assessments.

Electromyography and postural stability assessment equipment and EMG normalization data collection

Synchronized data from posturography and electromyography (EMG) systems were sampled at 1000 Hz (posturography) and 2000 Hz (EMG), respectively. The postural stability was recorded using an AMTI force platform (Model HPS 464508, AMTI Europe GmbH, Helmstadt-Bargen, Germany). The EMG system used was a wireless Noraxon Ultium system (USA. Inc, Scottdale, Arizona, USA) equipped with 16 channels for recording muscle activity. When making the EMG recordings, sampled muscle activity was lowpass filtered by the device at 500 Hz before the collected data was analyzed.

Prior to electrode application, the skin over target muscles was shaved and abraded using medical abrasion gel (Nuprep, Weaver and Company, Aurora, Colorado, USA). Following Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) guidelines26, bipolar surface electrodes (Noraxon, USA, Inc) with a 20 mm inter-electrode distance were placed bilaterally over the gluteus medius (GlutMed), medial gastrocnemius (MedGas), tibialis anterior (TibAnt), and peroneus longus (PerLong).

To normalize EMG data, Maximum Voluntary Contractions (MVCs) were assessed for each target muscle prior to the main experimental tasks, according to SENIAM recommendations27 with a modification for gluteus medius MVC, which was performed in supine position. Specific tasks included: hip abduction (GlutMed), plantar flexion (MedGas), dorsiflexion (TibAnt), and ankle eversion/pronation (PerLong). The MVC assessments were repeated three times for each muscle with 30 s pause between contractions, the highest value was used as MVC. During each trial, subjects were instructed to gradually build up force and perform the contraction with maximal voluntary effort.

Test Procedure

The protocol involved seven postural stability tests. Two control tests (quiet stance with eyes open [EO], quiet stance with eyes closed [EC]) were conducted in a randomized order before VR exposure. Participants stood without shoes (socks permitted) on the force platform, adopting a standardized stance with feet slightly externally rotated (~ 30°, guided by platform markings), arms folded across the chest, and head facing forward15,27,28. For the EO condition, participants focused on a target placed about 5 m away at eye level. These tests established baseline postural sway without VR influence, serving as a reference. Participants did not wear the VR headset during control tests.

Before the VR sessions, participants viewed a 360° grocery shopping video while seated, allowing for adjustment of image sharpness in the VR headset, using an Oculus Quest 2 VR headset (Facebook Technologies, LLC, Menlo Park, California, USA). A Quest 2 VR headset weighs about 503 g, which corresponds to the weight of a standard construction helmet. The horizontal field of view is about 95°, and the vertical field of view about 93° in a Quest 2 VR headset. However, when making rotational head or body movement, the visual environment is updated throughout the entire 360° − 3D rotation range. Participants then stood on the force platform in the same standardized posture used during control tests, while wearing the Quest 2 VR headset. To ensure standardized conditions across participants and minimize potential strain from excessive head motion within VR, participants were instructed to maintain a forward gaze throughout these sessions. Five postural assessments were conducted while participants viewed a 120-second, 360° photographic video of a roller coaster ride with rapid directional changes and elevation shifts. The same video was used for all five sessions, with a about 10 min seated rest period provided between each session. Illustrations of the kind of postural control responses evoked by watching an immersive VR movie can be found in the paper Fransson et al. 2019 15.

The operators were experienced clinicians, well acquainted with identifying and applying fall-preventive measures. For safety reasons, during all posturography tests, the operator stood close to the participant and was prepared to intervene, if necessary, while continuously monitoring for signs of severe instability. If the operator judged that a fall was likely to occur without external intervention, they supported the participants by briefly placing their hands on the participant to stop the ongoing movement and prevent a fall. The operator removed their hands as soon as the participant had regained stability. Any occasion on which a participant required such support was documented for subsequent analysis of the occurrence of severe instability. Five participants could not be included in the detailed EMG and posturography analyses because they stepped off the force platform during one or more assessments.

Posturography data analysis

Postural stability in the anteroposterior and mediolateral directions during control and VR tests was quantified by calculating the variance of the anteroposterior and mediolateral torque signals derived from the force platform data. Torque variance reflects the energy used towards the force platform to maintain stability28,29. Biomechanical models of the human postural control system reveal that force platform recordings are directly influenced by the anthropometric factors test subject’s weight and height28,30. To account for anthropometric influences on posturography recordings15,28, torque variance values were normalized using the square of the participant’s height and weight, ensuring unit agreement. Furthermore, spectral analysis using a fifth-order digital Finite Impulse Response (FIR) filter separated the torque signals into low-frequency (< 0.1 Hz, representing smooth postural corrections) and high-frequency (> 0.1 Hz, representing rapid balance adjustments) components31.

EMG data analysis

EMG data for the MVC assessments of GlutMed, MedGas, TibAnt and PerLong muscles were first processed by using a bandpass filter with cutoffs at 20 Hz and 500 Hz and then calculating the mean root square (RMS) envelope using a sliding window of 75 ms. The MVC values were thereafter obtained by filtering the EMG recordings, using a moving average filter with a sliding window of 1 s, and then selecting the highest value in the resulting data. The EMG muscle activity during the different test conditions was attained for the GlutMed, MedGas, TibAnt and PerLong muscles by first using the same initial data processing procedure with a 20–500 Hz bandpass filter and then calculating the RMS using a sliding window of 75 ms. Thereafter, the normalized EMG muscle activity for each muscle were calculated by dividing the filtered and RMS averaged data with the corresponding MVC values. The co-contraction index (CCI) was calculated on sample level for the three muscle pairs: (1) gluteus medius right and left side (GlutMed L vs. GlutMed R), (2) tibialis anterior and medial gastrocnemius (TibAnt vs. MedGas), and (3) tibialis anterior and peroneus longus (TibAnt vs. PerLong). For each muscle pair, the muscle with the highest normalized EMG muscle activity was regarded as agonist. CCI was computed using Rudolph’s method33 with the normalized EMG amplitude values:

$$\:CCI=\:\frac{antagonist}{agonist}\left(antagonist+agonist\right)$$

In the statistical analysis, the mean and SD values for the CCI values were evaluated. These values were first calculated individually for each leg. Thereafter, the mean values using data for both legs were calculated and analyzed. All signal processing and mathematical analyses were performed using MATLAB (Version 2023a, The Mathworks Inc, Natick, Massachusetts, USA).

Statistical analysis

A repeated measures General Linear Model (GLM) Analysis of Variance (ANOVA) analysis was used for evaluating the performance across repeated VR sessions. To validate the appropriateness of using the repeated measured GLM ANOVA method when analyzing the datasets in this study, it was first checked that all dataset combinations analyzed with this statistical method produced model residuals that had normal or close to normal distribution32,33. The primary independent variable in the analysis was Repetition, representing the VR session number (Sessions 1 to 5; degrees of freedom (d.f.) = 4). This factor was treated as a within-subjects variable in the analyses. The dependent variables evaluated included the log-transformed, normalized total, low-frequency, and high-frequency torque variance values derived from posturography data, as well as the log-transformed mean CCI and SD CCI obtained from EMG measurements.

Post hoc within-subject comparisons were conducted using the Wilcoxon matched-pairs signed-rank test (Exact significance, two-tailed). Specifically, comparisons were made between VR Session 1 and VR Session 5 to evaluate the cumulative effects of repeated VR exposure; between VR Session 1 and the eyes open control test to assess the relative initial impact of VR, and between VR Session 5 and the eyes-open control condition to determine the relative final performance after repeated VR sessions. Post hoc within-subject comparisons were also performed to determine the influence of visual input during the control tests.

Spearman’s rank correlation coefficients were calculated to assess potential associations between postural stability measures and muscle CCIs. Correlation analyses for VR conditions were based on pooled data from all five VR sessions, while analyses for control tests were based on pooled data from both control tests. Statistical significance for correlations was set at p < 0.05. The strength of the correlations was interpreted using Cohen’s criteria: r > 0.1 indicating a small, r > 0.3 a moderate, r > 0.5 a strong, and r > 0.7 a very strong relationship34.

To evaluate whether repeated VR exposure reduced the need for fall-preventing support, Fisher’s Exact Test was employed.

For the repeated measures GLM ANOVA analysis significance level was set at p < 0.05. For the post hoc within-subjects comparisons, significance was adjusted to p < 0.025 based on Bonferroni correction. For within-subjects comparisons involving the control tests, the significance threshold remained at p < 0.05. Non-parametric statistical methods were selected for post hoc analyses due to some dataset violations of normality assumptions, as assessed by the Shapiro-Wilk test. Log-transformation was insufficient to normalize these datasets35.

A sample size estimation based on posturography data indicated an effect size of 0.95. Given a significance level of p < 0.05 (two-tailed), a minimum of 11 participants was required to achieve a statistical power of 0.80.

All statistical analyses were performed using SPSS software (Version 28, IBM Corp., Armonk, NY, USA). The Power analysis was conducted using G*Power (Version 3.1.9.7).

Data availability

Data generated in this study is available upon request to the authors. However, restrictions apply for this data, according to the Swedish Ethical Review Authority.

References

  1. Horak, F. B. Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls? Age Ageing. 35 (Suppl 2), ii7-ii11. https://doi.org/10.1093/ageing/afl077 (2006). 

  2. Shumway-Cook, A. & Woollacott, M. H. Motor control: translating research into clinical practice. Fifth edition. ednWolters Kluwer, (2017).

  3. Lord, S. R., Menz, H. B. & Tiedemann, A. A physiological profile approach to falls risk assessment and prevention. Phys. Ther. 83, 237–252 (2003).

    Article  PubMed  Google Scholar 

  4. Rubenstein, L. Z. Falls in older people: epidemiology, risk factors and strategies for prevention. Age Ageing. 35 (Suppl 2), ii37–ii41. https://doi.org/10.1093/ageing/afl084 (2006).

    Article  PubMed  Google Scholar 

  5. Peterka, R. J. Sensorimotor integration in human postural control. J. Neurophysiol. 88, 1097–1118. https://doi.org/10.1152/jn.2002.88.3.1097 (2002).

    Article  CAS  PubMed  Google Scholar 

  6. Cullen, K. E., Minor, L. B., Beraneck, M. & Sadeghi, S. G. Neural substrates underlying vestibular compensation: contribution of peripheral versus central processing. J. Vestib. Res. 19, 171–182. https://doi.org/10.3233/ves-2009-0357 (2009).

    Article  PubMed  PubMed Central  Google Scholar 

  7. Laessoe, U., Abrahamsen, S., Zepernick, S., Raunsbaek, A. & Stensen, C. Motion sickness and cybersickness – Sensory mismatch. Physiol. Behav. 258, 114015. https://doi.org/10.1016/j.physbeh.2022.114015 (2023).

    Article  CAS  PubMed  Google Scholar 

  8. Zwergal, A., Grabova, D. & Schoberl, F. Vestibular contribution to spatial orientation and navigation. Curr. Opin. Neurol. 37, 52–58. https://doi.org/10.1097/WCO.0000000000001230 (2024).

    Article  PubMed  Google Scholar 

  9. Bronstein, A. M. Visual vertigo syndrome: clinical and posturography findings. J. Neurol. Neurosurg. Psychiatry. 59, 472–476. https://doi.org/10.1136/jnnp.59.5.472 (1995).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Kao, A. C., Nanda, A., Williams, C. S. & Tinetti, M. E. Validation of dizziness as a possible geriatric syndrome. J. Am. Geriatr. Soc. 49, 72–75. https://doi.org/10.1046/j.1532-5415.2001.49012.x (2001).

    Article  CAS  PubMed  Google Scholar 

  11. Strupp, M., Brandt, T. & Dieterich, M. in in Vertigo and Dizziness: Common Complaints. 1–8 (eds Strupp, M., Brandt, T. & Dieterich, M.) (Springer International Publishing, 2023).

  12. Staab, J. P. et al. Diagnostic criteria for persistent postural-perceptual dizziness (PPPD): Consensus document of the committee for the Classification of Vestibular Disorders of the Barany Society. J. Vestib. Res. 27, 191–208. https://doi.org/10.3233/VES-170622 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  13. Truijen, S. et al. Effect of home-based virtual reality training and telerehabilitation on balance in individuals with Parkinson disease, multiple sclerosis, and stroke: a systematic review and meta-analysis. Neurol. Sci. 43, 2995–3006. https://doi.org/10.1007/s10072-021-05855-2 (2022).

    Article  PubMed  PubMed Central  Google Scholar 

  14. Sulway, S. & Whitney, S. L. Advances in Vestibular Rehabilitation. Adv. Otorhinolaryngol. 82, 164–169. https://doi.org/10.1159/000490285 (2019).

    Article  PubMed  Google Scholar 

  15. Fransson, P. A. et al. Postural instability in an immersive Virtual Reality adapts with repetition and includes directional and gender specific effects. Sci. Rep. 9, 3168. https://doi.org/10.1038/s41598-019-39104-6 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Sana, V. et al. Effects of vestibular rehabilitation therapy versus virtual reality on balance, dizziness, and gait in patients with subacute stroke: A randomized controlled trial. Med. (Baltim). 102, e33203. https://doi.org/10.1097/MD.0000000000033203 (2023).

    Article  Google Scholar 

  17. Röijezon, U., Hansson, E. E., Nae, J. Ä., Östlind, E. & Fransson, P. A. Postural stability and muscle co-activation among younger adults adapting to an immersive virtual reality. Eur. J. Appl. Physiol. https://doi.org/10.1007/s00421-026-06177-x (2026).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Nagai, K. et al. Differences in muscle coactivation during postural control between healthy older and young adults. Arch. Gerontol. Geriatr. 53, 338–343. https://doi.org/10.1016/j.archger.2011.01.003 (2011).

    Article  PubMed  Google Scholar 

  19. Vette, A. H. et al. Ankle muscle co-contractions during quiet standing are associated with decreased postural steadiness in the elderly. Gait Posture. 55, 31–36. https://doi.org/10.1016/j.gaitpost.2017.03.032 (2017).

    Article  PubMed  Google Scholar 

  20. Falk, J. et al. Increased co-contraction reaction during a surface perturbation is associated with unsuccessful postural control among older adults. BMC Geriatr. 22 https://doi.org/10.1186/s12877-022-03123-2 (2022).

  21. Lee, J., Phu, S., Lord, S. R. & Okubo, Y. Effects of immersive virtual reality training on balance, gait and mobility in older adults: A systematic review and meta-analysis. Gait Posture. 110, 129–137. https://doi.org/10.1016/j.gaitpost.2024.03.009 (2024).

    Article  CAS  PubMed  Google Scholar 

  22. Almqvist Nae, J., Nystrom, A., Luccini, F. & Magnusson, M. Ekvall Hansson, E. Video exposure through virtual reality can improve older people’s ability to manage postural instability caused by distortive visual environments. PLoS One. 19, e0306834. https://doi.org/10.1371/journal.pone.0306834 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Nishiike, S. et al. The effect of visual-vestibulosomatosensory conflict induced by virtual reality on postural stability in humans. J. Med. Invest. 60, 236–239. https://doi.org/10.2152/jmi.60.236 (2013).

    Article  PubMed  Google Scholar 

  24. Saraiva, M., Vilas-Boas, J. P. & Castro, M. A. Postural Control and Muscle Activity during Dual-Task in Young Adults. Behav. Sci. (Basel). 14 https://doi.org/10.3390/bs14050403 (2024).

  25. Tjernstrom, F., Fransson, P. A., Hafstrom, A. & Magnusson, M. Adaptation of postural control to perturbations–a process that initiates long-term motor memory. Gait Posture. 15, 75–82. https://doi.org/10.1016/s0966-6362(01)00175-8 (2002).

    Article  CAS  PubMed  Google Scholar 

  26. Tjernstrom, F., Fransson, P. A. & Magnusson, M. Improved postural control through repetition and consolidation. J. Vestib. Res. 15, 31–39 (2005).

    Article  CAS  PubMed  Google Scholar 

  27. Hermens, H. J., Freriks, B., Disselhorst-Klug, C. & Rau, G. Development of recommendations for SEMG sensors and sensor placement procedures. J. Electromyogr. Kinesiol. 10, 361–374. https://doi.org/10.1016/s1050-6411(00)00027-4 (2000).

    Article  CAS  PubMed  Google Scholar 

  28. Johansson, R., Fransson, P. A. & Magnusson, M. Optimal coordination and control of posture and movements. J. Physiol. Paris. 103, 159–177. https://doi.org/10.1016/j.jphysparis.2009.08.013 (2009).

    Article  PubMed  Google Scholar 

  29. Riccio, G. E. & Stoffregen, T. A. Affordances as constraints on the control of stance. Hum. Mov. Sci. 7, 265–300 (1988).

    Article  Google Scholar 

  30. Fransson, P. A. Adaptation of Human Postural Control: Learning, sensorimotor and analysis aspects (VDM, 2009).

  31. Kristinsdottir, E. K., Fransson, P. A. & Magnusson, M. Changes in postural control in healthy elderly subjects are related to vibration sensation, vision and vestibular asymmetry. Acta Otolaryngol. 121, 700–706. https://doi.org/10.1080/00016480152583647 (2001).

    Article  CAS  PubMed  Google Scholar 

  32. Midway, S. & White, J. W. Testing for normality in regression models: mistakes abound (but may not matter). R Soc. Open. Sci. 12, 241904. https://doi.org/10.1098/rsos.241904 (2025).

    Article  PubMed  PubMed Central  Google Scholar 

  33. Knief, U. & Forstmeier, W. Violating the normality assumption may be the lesser of two evils. Behav. Res. Methods. 53, 2576–2590. https://doi.org/10.3758/s13428-021-01587-5 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  34. Cohen, J. Statistical power analysis for the behavioral sciences. edn (Lawrence Erlbaum Associates, Inc., 1988).

  35. Altman, D. Practical statistics for medical research (Chapman & Hall, 1991).

Download references

Acknowledgements

This study was conducted within the context of the Movement and Reality Lab (MoRe-Lab) at the Faculty of Medicine, Lund University. The authors want to acknowledge Malin E Gustafsson, Sonja Andersson Marforio and Jeannette Unge for their help with the data collection. Special thanks to Johan Jirlén for data processing and calculation of EMG data.

Funding

Open access funding provided by Lund University. The study was funded by Lions Skane research foundation. Special thanks to Johan Jirlén for data processing and calculation of EMG data.

Author information

Authors and Affiliations

  1. Department of Health Sciences, Lund University, Lund, Sweden

    Eva Ekvall Hansson, Jenny Älmqvist Nae, Elin Östlind & Ulrik Röijezon

  2. Department of Otorhinolaryngology Head and Neck Surgery, Skåne University Hospital, Lund, Sweden

    Eva Ekvall Hansson & Per-Anders Fransson

  3. Department of Clinical Sciences Lund, Lund University, Lund, Sweden

    Per-Anders Fransson

  4. Department of Health, Education and Technology, Luleå University of Technology, Luleå, Sweden

    Ulrik Röijezon

Authors

  1. Eva Ekvall Hansson
  2. Jenny Älmqvist Nae
  3. Elin Östlind
  4. Ulrik Röijezon
  5. Per-Anders Fransson

Contributions

EEH, UR and PAF planned the study and wrote the main manuscript text. PAF prepared Figs. 1, 2 and 3. EÖ and JÄN performed data collection and wrote parts of the manuscript text. UR and PAF performed statistical analyzes. All authors participated in preparing the manuscript for submission and in interpreting the results. All authors have approved the submitted version. All authors has also participated in the revision of the manuscript, after feedback from reviewers. All authors have approved the resubmitted version.

Corresponding author

Correspondence to Eva Ekvall Hansson.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hansson, E.E., Nae, J.Ä., Östlind, E. et al. Rapid postural control adaptations to repeated virtual reality stimulation in older adults. Sci Rep 16, 38327 (2026). https://doi.org/10.1038/s41598-026-80453-w

Download citation

  • Received: 23 March 2026

  • Accepted: 09 November 2026

  • Published: 17 November 2026

  • Version of record: 17 November 2026

  • DOI: https://doi.org/10.1038/s41598-026-80453-w

Keywords

来源:Nature 健康护理主题 · nature.com