Novel Clinical Assessment of Visual,Vestibular,Somatosensory,and Autonomic Function: Establishing Test Re-Test Reliability in a Healthy Population

Abstract

Introduction

Maintaining balance is a complex process involving the integration of information from the visual, vestibular, and somatosensory (VVS) systems, along with autonomic nervous system (ANS) function. Traditional assessments of these systems are often expensive, limited to specialised settings or focus on overall balance outcomes, potentially overlooking deficits in the individual systems. This study aimed to evaluate the test-retest reliability of a novel, fully portable clinical assessment tool designed to provide objective measures for individual components of the VVS system and ANS function.

Methods

Twenty-eight participants (aged 20–88 years), with no comorbidities and meeting Australian physical activity guidelines, completed the protocol twice. The novel clinical assessment tool comprised of two systems: (1) a virtual reality-type headset incorporating eye-tracking to evaluate visual-vestibular function (smooth pursuit and voluntary saccades), and autonomic function (pupil light reflex); and (2) the Active Movement Extent Discrimination Apparatus (AMEDA) for somatosensory function. Reliability was assessed using two-way mixed-effects model (consistency type, single rater) Intraclass Correlation Coefficients (ICC 3,1) calculated in R-studio. The standard error of measurement (SEM) and minimal detectable change (MDC) was also calculated. Bland-Altman plots were utilised to visualise the agreement between two test repeats.

Results

Each metric demonstrated at least moderate to good test re-test reliability: left and right AMEDA (ICC = 0.69 and 0.75), smooth pursuit (ICC = 0.67), voluntary saccades (ICC = 0.53), autonomic response delay (ICC = 0.80), parasympathetic function (ICC = 0.86), and sympathetic function (ICC = 0.89).

Discussion

This study supports the reliability of a new, fully portable clinical assessment tool to assess VVS and ANS function. By demonstrating the reliability of this new streamlined tool for evaluating the VVS and ANS systems, the findings of this study has the potential to enhance clinical practice and research in falls prevention and balance rehabilitation.

Keywords

balance assessment eye-tracking AMEDA fall prevention older adults sensory integration clinical protocol

Introduction

Maintaining posture and balance control is essential for daily life, underpinning a wide range of activities from walking and playing to competitive sports (Winter, 1995). Balance is maintained by keeping the body’s centre of gravity aligned vertically over the base of support, relying on continuous and rapid information about body posture and the environment from the visual, vestibular, and somatosensory (VVS) systems (Zatsiorsky & Duarte, 1999). Disruptions to balance, whether due to injury, illness, or age-related decline, can significantly impact quality of life (Winter, 1995). Therefore, understanding the mechanisms that support balance, including the quality of the information available from the VVS systems, is essential for improving movement and reducing the risk of injury.

Each of the components of the VVS complex are crucial for maintaining balance and postural control (Horak, 2006). The visual system informs the Central Nervous System (CNS) about the surrounding environment and the body’s position relative to external objects (Davids et al., 2005). Given the reliance on visual input for these tasks, a significant portion of the human brain is dedicated to processing eye movements, involving cortical and subcortical regions such as the brainstem, cerebellum, and basal ganglia (Lal & Truong, 2019; Leigh & Zee, 2015). These areas are critical for visual processing and motor control.

The vestibular system, located in the inner ear, plays a vital role in detecting head movements and changes in position relative to gravity (Gaerlan, 2010). It works synergistically with the visual system through the vestibulo-ocular reflex (VOR), which stabilises vision by coordinating eye movements in response to head motion (Fetter, 2007). Beyond the VOR, the vestibular system works centrally to assist with eye movements, therefore, evaluating eye movements provides valuable insights into the integrity of the central vestibular and visual systems, given their close interaction (Lal & Truong, 2019; Termsarasab et al., 2015; Van Der Torn & Van Dijk, 2000). Voluntary eye movements, such as saccades, are generated in the frontal eye fields of the cerebral cortex, while smooth pursuit movements, which occur in response to a moving stimulus, involve a network of brain regions, including the visual and parietal cortices and the cerebellum (Lal & Truong, 2019). It follows, that assessing smooth pursuit and voluntary saccades provides valuable information into the integrity of the visual-vestibular pathways, offering an accessible and accurate means of identifying disruptions in these pathways (Van Der Torn & Van Dijk, 2000). Pearce et al. (2023) demonstrated that similar assessments of smooth pursuit, using virtual reality headsets, demonstrate good test re-test reliability.

Somatosensation, the final pillar of the VVS complex, encompasses the senses of touch, proprioception, pressure, and thermoreception, which collectively provide information about the body’s position within its surrounding environment (Goodman & Bensmaia, 2020). Three primary methods are commonly used to measure proprioception: (1) Threshold to Detection of Passive Motion, (2) Joint Position Reproduction, and (3) Active Movement Extent Discrimination Assessment (AMEDA). A 2016 review highlighted that while Threshold to Detection of Passive Motion offers high conceptual purity by isolating proprioceptive input from various mechanoreceptors, its ecological validity is lower (Han et al., 2016). Joint position reproduction is noted for efficiently assessing hemispheric asymmetries in sensorimotor abilities, despite its lesser validity in testing. AMEDA, on the other hand, provides a comprehensive assessment of proprioception during functional movements and boasts higher ecological and test validity, though it may be less intuitive than the other two methods (Han et al., 2016). AMEDA’s requirement for participants to be in a weight-bearing position integrates proprioception, touch, and pressure into the assessment, making it potentially the most valid method for evaluating somatosensation in functional contexts. Earlier designs of the AMEDA exhibit moderate to excellent test re-test reliability, with differences in results likely attributable to variations in protocol and population tested (Antcliff et al., 2021; Pan et al., 2022; Shi et al., 2023).

Beyond the VVS systems, the autonomic nervous system (ANS) plays a role in movement control by regulating involuntary physiological functions and preparing the body for movement (Gordon et al., 2023). The ANS’s involvement in movement control, while not initially apparent, is important, as it regulates involuntary physiological processes like heart rate, blood pressure, and respiration, that support postural stability (Bär et al., 2009; Gordon et al., 2023). ANS function is commonly quantified using Heart Rate Variability (HRV) and while this is an accessible method for assessing ANS activity, relying on it as a comprehensive measure of the entire ANS function is misleading (Hayano & Yuda, 2019). Using the Pupil Light Reflex (PLR) offers a solution to this critique, as it involves the integrated function of both the parasympathetic and sympathetic nervous systems (Hall & Chilcott, 2018) for pupil constriction and dilation, respectively (Figure 1). The parasympathetic arc for pupil constriction begins with photoreceptor cells in the retina sending a signal to the Edinger–Westphal nuclei. There, parasympathetic neurons are stimulated to carry a signal which causes contraction of the iris sphincter muscles through release of the neurotransmitter, acetylcholine (Hall & Chilcott, 2018). Meanwhile, the sympathetic influence for pupil dilation follows a three-neuron arc originating in the hypothalamus, traversing the spinal cord, and culminating in the contraction of the iris dilator muscle via noradrenaline release (Hall & Chilcott, 2018). It follows, that this complex, centrally mediated pathway arguably provides a more direct assessment of ANS function.

Figure 1.

Simplified schematic of mechanism of pupil light reflex adapted from Hall & Chilcott (2018).

The Systems Model emphasises the adaptability of the balance control system; when one sensory system is compromised, the remaining systems compensate to preserve overall balance (Peterka, 2002; Shumway-Cook & Woollacott, 2012). While compensatory mechanisms are beneficial in maintaining balance and preventing falls, they may impose additional demands on intact sensory systems and the body, potentially resulting in inefficiencies or long-term vulnerabilities (Horak, 2006; Merla & Spaulding, 1997). Traditional clinical assessments of balance often focus on whole body outcomes, such as sway during standing or performance on functional tests (Pollock et al., 2000). While such metrics are valuable for assessing overall balance ability, they may obscure underlying compensatory mechanisms (Horak, 2006; Peterka, 2002). Without evaluating the individual contributions of the VVS systems, specific deficits may be overlooked. This limitation underscores the need for refined assessments that can isolate and evaluate each component of the VVS and ANS complex independently.

While other assessments have been developed that measure somatosensation, smooth pursuit, voluntary saccades, and the PLR, these tools have previously been expensive, non-portable, and limited to specialised settings—making them difficult to translate into clinical practice. Alternative methods for assessing components of the VVS systems – such as dynamic posturography for visual-vestibular function (costing up to $200 000 (Wix, 2024)), or the original AMEDA for somatosensation – are impractical for routine use. Prism Neuro (Canberra, Australia) addresses these limitations by designing a new, portable and user-friendly system capable of evaluating VVS and ANS function quickly. This includes a compact, transportable version of the AMEDA device and a virtual reality headset that integrates smooth pursuit, voluntary saccades, and PLR into a streamlined, efficient assessment—offering a more accessible alternative to traditional lab-based tools. This system also enables the reliability of multiple measures to be assessed concurrently within a single population.

Reliability, referring to the consistency of a measurement across repeated trials, is a critical first step in validating any new assessment tool (Portney, 2020). For clinicians and researchers to adopt a new tool with confidence, it must demonstrate good test-retest reliability, ensuring stable and reproducible results over time. This study aims to establish the test re-test reliability of a newly designed tool designed to evaluate VVS and ANS function, in a general healthy population. It is hypothesised that this tool will demonstrate good reliability, offering a valuable resource for improving balance assessment in clinical settings.

Method

Participants and Ethical Considerations

This study was a prospective, observational study with repeated measurements. Ethics approval was obtained from the university’s Human Research Ethics Committee prior to any data collection occurring (HREC – 13,197). Participants aged 18 years or older were recruited through flyers on the University campus, various gyms around the region, and social media. Using a hypothesis-testing approach the sample size was estimated with a significance level of 0.05, a power of 80%, an anticipated true Intraclass Correlation Coefficients (ICC) of 0.82 (Witchalls et al., 2014), and a null ICC of 0.65 (Antcliff et al., 2021), resulting in a required sample size of 28 participants (Borg et al., 2022). Participants met the following inclusion criteria: (1) aged 18 years or older, (2) consistently meet the Australian physical activity and exercise guidelines (AusGov, 2021), (3) had no COVID-19 infection, concussion, or lower limb injury within the previous 12 months, (4) no diagnosis or history of conditions affecting the oculomotor or vestibular systems (e.g., vestibular disorders, apraxia, diplopia, post-viral conditions or seizures), and (5) not currently using medications or substances that influence nervous system responses. Participant demographics are shown in Table 1.

Table 1.

Participant Numbers and Demographics (Mean ± SD).

Variable	Young adults (n = 10, age 18–39)	Middle-aged adults (n = 13, age 40–64)	Older adults (n = 5, age 65+)	All participants (n = 28, age 18–88)
Demographic
Age (years)	25.7 ± 5.8	53.1 ± 8.6	71.8 ± 10.0	46.6 ± 18.9
Sex (F%)	60.0%	69.2%	40.0%	60.7%
Height (m)	1.73 ± 0.07	1.71 ± 0.09	1.71 ± 0.13	1.72 ± 0.09
Weight (kg)	72.48 ± 9.37	71.20 ± 10.81	77.98 ± 12.61	72.87 ± 10.53

	Test 1	Test 2	Test 1	Test 2	Test 1	Test 2	Test 1	Test 2
Autonomic
Response delay (msec)	272.10 ± 20.37	270.00 ± 20.16	278.23 ± 17.84	267.85 ± 15.77	275.60 ± 15.57	278.20 ± 11.10	275.57 ± 17.99	270.46 ± 16.69
Parasympathetic (mm/sec)	3.13 ± 0.63	3.38 ± 0.65	3.43 ± 0.64	3.54 ± 0.57	2.89 ± 0.81	2.64 ± 0.76	3.22 ± 0.67	3.32 ± 0.69
Sympathetic (mm/sec)	0.61 ± 0.11	0.59 ± 0.09	0.57 ± 0.13	0.57 ± 0.13	0.48 ± 0.18	0.53 ± 0.21	0.57 ± 0.14	0.57 ± 0.13
Somatosensory
AMEDA left (AUC)	73.40 ± 4.12	75.70 ± 6.15	71.92 ± 3.12	74.77 ± 5.45	65.60 ± 5.68	67.80 ± 9.50	71.32 ± 4.75	73.86 ± 6.90
AMEDA right (AUC)	76.30 ± 2.54	74.90 ± 3.28	71.77 ± 5.39	73.31 ± 4.79	63.40 ± 7.73	67.20 ± 8.93	71.89 ± 6.67	72.79 ± 5.79
Visual-vestibular
Smooth pursuit (AU)	10.89 ± 1.62	11.4 ± 2.15	11.63 ± 3.46	11.88 ± 3.41	11.76 ± 1.01	12.43 ± 4.88	11.39 ± 2.55	11.83 ± 3.22
Voluntary saccades (time on target/ time to target)	1.60 ± 0.37	1.63 ± 0.51	1.14 ± 0.34	1.22 ± 0.40	1.44 ± 0.72	1.31 ± 0.59	1.36 ± 0.47	1.38 ± 0.49

Note. F = female, AMEDA = Active Movement Extent Discrimination Assessment, AUC = Area under the curve, AU = arbitrary units.

Procedure

Participants attended a single laboratory-based testing session. Upon arrival, participants provided written informed consent, completed the ESSA Pre-Exercise Screening tool, and had their height and weight collected. Participants then completed the VVS protocol, which consisted of two 5-min assessments: an eye-tracking assessment and the AMEDA. To assess test-retest reliability, each participant underwent the VVS protocol twice, with the second assessment conducted following a short rest period of approximately 5-min. Participants remained in a quiet temperature controlled (22°C) laboratory, with consistent lighting. Noise and environmental distractions were removed by conducting tests in this, private, closed-door lab.

The first half of the VVS protocol used a virtual reality headset (Prism Neuro^®, Figure 2) to measure both pupillometry and eye-tracking metrics. The headset contained two organic light emitting diode displays with a resolution of 2560 × 1440 pixels each, a frame rate of 70 FPS, a 100° field of view, and a 120 Hz infrared eye-tracking system with a reported accuracy of 1.15° median across the screen.

Figure 2.

Prism neuro eye tracking assessment.

The eye-tracking assessment comprised three tests, which were administered in the same order for all participants to account for the impact of light adaptation on pupil reflex responses (Kalloniatis & Luu, 2005). Prior to testing, participants underwent a calibration within the headset. This process involved capturing images of the eye from various angles, which were then used to map gaze positions across multiple points within the field of view, sufficient to track the full scope of movement that occurs during testing. Testing commenced immediately after calibration. The first test involved tracking a small red dot as it moved in a circular motion (Figure 3). The second test required participants to focus on a small red ‘x’ while lights flashed, and data on constriction and dilation velocities were recorded (Figure 4). In the final test, participants tracked a small red ‘x’ making rapid movements across the screen (Figure 5). Standardised instructions were provided to all participants to minimise bias and any subsequent possible impact on reliability, as detailed in Appendix 1. Data collected from the headset underwent proprietary automatic blink filtering and signal smoothing using a low-pass filter. These preprocessing steps were applied uniformly across all participants and sessions and therefore not expected to bias the reliability outcomes. The smooth pursuit metric was calculated as the root means square error of the participants’ gaze position from the provided stimulus during the first test. A greater error was indicated by a larger test score and were reported in arbitrary units (AU) (Pearce et al., 2023). The voluntary saccade metric was calculated as the time spent accurately on the target divided by the time taken to reach the target, during the third test. Finally, autonomic function was assessed during the second test, where response delay was calculated as time taken for the pupil to respond to the light. Parasympathetic data was collected as peak constriction velocity of the pupil and sympathetic data was collected as average dilation velocity of the pupil (Hall & Chilcott, 2018). Similar to smooth pursuit error, smaller numbers indicate better performance for the response delay measure.

Figure 3.

Smooth pursuit where the red dot represents the stimulus, blue dot represents the participant’s gaze position and black line represents stimulus path.

Figure 4.

Pupil light reflex where the red ‘X’ represents the stimulus, the blue dot represents the participant’s gaze position.

Figure 5.

Voluntary saccades where the red ‘X’ represents the stimulus, the blue dot represents the participant’s gaze position and black line represents stimulus path.

The second half of the VVS protocol involved measuring somatosensory system performance using the AMEDA, Figure 6. The AMEDA tests a participant’s ability to discriminate between different angles of inversion of a footplate (Waddington & Adams, 1999a, 1999b; Witchalls et al., 2014). Participants stood with one foot on a stable platform and the other on a tilting platform designed to permit controlled ankle inversion movements at five different one-degree intervals, ranging from 10 to 14°. All participants were tested on their left side first, followed by their right. Previous research in healthy groups has shown no significant difference in somatosensory ability scores due to test order, as measured by the AMEDA, between dominant and non-dominant sides (Pan et al., 2022). Participants were familiarised with the five different tilt ranges three times in a shallow-to-deep order before the formal testing began. During the test, the five tilt levels were presented in a random order ten times each, resulting in a total of 50 trials per side. Participants were instructed to tilt the platform, return it to the start position, and immediately report the perceived tilt level. Standardised instructions were used to ensure consistency, as outlined in Appendix 1.

Figure 6.

Prism Neuro^® Active movement extent discrimination apparatus (AMEDA).

To calculate a participant’s ability to distinguish between different levels, the accuracy of distinguishing between each adjacent pair of angles (1 vs. 2, 2 vs. 3, 3 vs. 4, and 4 vs. 5) is determined using the Area Under the Curve (AUC) of the Receiver Operating Characteristic curve for each pair. The total score is the average of the AUC scores across the four degrees of freedom between the five stop positions. This score is then converted into a percentage for readability. A score of 100% represents perfect joint position sense discrimination between all positions, while a score of 50% indicates performance equivalent to chance.

Statistical Analysis

Statistical analyses were conducted using RStudio (Version 2024.09.0 + 375). Descriptive statistics, mean and standard deviation (SD) were calculated. Two-way mixed-effects model, consistency type, single rater ICC (3,1), with 95% confidence intervals (95% CI), were calculated to assess the test-retest reliability of the measurements (Koo & Li, 2016). This method tests variability due to participant performance or testing conditions, not rater differences. Portney (2020) suggested that “Values less than 0.5 are indicative of poor reliability, values between 0.5 and 0.75 indicate moderate reliability, values between 0.75 and 0.9 indicate good reliability, and values greater than 0.90 indicate excellent reliability” (Portney, 2020). The standard error of measurement (SEM) was calculated using the formula SEM = SD^*√(1-ICC3), where SD represents the pooled standard deviation (Pan et al., 2022). The minimal detectable change at the 95% confidence level (MDC₉₅) was calculated using the formula: MDC₉₅ = 1.96 × √2 × SEM, where 1.96 is the Z-score for 95% CI, representing the smallest change that exceeds measurement error with 95% confidence (Pan et al., 2022).

Bland-Altman plots were used to visually assess the relative agreement between the two measurement sessions (Test 1 and Test 2) for each variable (Giavarina, 2015). These plots facilitate the evaluation of bias between the mean differences of the two sessions and provide an estimate of the agreement interval, within which 95% of the differences between the two measurements are expected to fall (Giavarina, 2015).

To assess the differences between Test 1 and Test 2 scores across the variables, paired-samples t-tests and Wilcoxon signed-rank tests were conducted, depending on the normality of the data (Kirkwood & Sterne, 2010). Normality was assessed for each variable using the Shapiro-Wilk test, and if the data were normally distributed, a paired-samples t test was performed to compare the means of Test 1 and Test 2 scores, and Cohen’s D was calculated to assess the effect size (Fritz et al., 2012). According to Cohen (1988), d values of 0.2, 0.5, and 0.8 represent small, medium, and large effects, respectively (Cohen, 1988). These can be interpreted more meaningfully as merely statistical, subtle, and obvious effects (Fritz et al., 2012). For variables that did not meet the assumption of normality (p < .05 from the Shapiro-Wilk test), a Wilcoxon signed-rank test was used to compare the median scores between the two test sessions. For non-parametric comparisons, effect size was calculated using the rank-biserial correlation (r), derived from the Wilcoxon test statistic (Fritz et al., 2012). This approach allowed identification of any potential learning effect and the magnitude of the effect, if any, between test sessions, ensuring that changes in performance were not due to familiarity but rather true variability in measurement (Pearce et al., 2023). Statistical significance was set at p < .05.

Results

Participants’ demographic descriptors and VVS scores are provided in Table 1. ICC values for each variable are summarised in Table 2. The ICC results for the visual autonomic metrics were all good, approaching excellent. The response delay exhibited an ICC of 0.80 (95% CI = 0.61–0.90, SEM = 7.78, MDC₉₅ = 21.56), while parasympathetic and sympathetic responses recorded ICC values of 0.86 (95% CI = 0.71–0.93, SEM = 0.25, MDC₉₅ = 0.71) and 0.89 (95% CI = 0.78–0.95, SEM = 0.05, MDC₉₅ = 0.12) respectively. The left and right AMEDA scores demonstrated an ICC of 0.69 (95% CI = 0.44–0.84, SEM = 3.35, MDC₉₅ = 9.28) and 0.75 (95% CI = 0.53–0.88, SEM = 3.1, MDC₉₅ = 8.58), respectively, indicating moderate to good reliability. The ICC for smooth pursuit was 0.67 (95% CI = 0.40–0.83, SEM = 1.65, MDC₉₅ = 4.59), suggesting moderate reliability. Similarly, voluntary saccades had an ICC of 0.53 reflecting a moderate reliability (95% CI = 0.20–0.75, SEM = 0.33, MDC₉₅ = 0.91).

Table 2.

Intraclass Correlation Coefficients (ICC) for Test-Retest Reliability of Visual-Vestibular, Somatosensory, and Autonomic Function.

Variable	95% confidence interval
Variable	ICC (3,1)	F	p value	Lower limit	Upper limit	Standard error of measurement	Minimal detectable change
Autonomic
Response delay	0.80	9.1	<0.001	0.61	0.90	7.78	21.56
Parasympathetic	0.86	13	<0.001	0.71	0.93	0.25	0.71
Sympathetic	0.89	16	<0.001	0.78	0.95	0.05	0.12
Somatosensation
AMEDA left	0.69	5.5	<0.001	0.44	0.84	3.35	9.28
AMEDA right	0.75	7	<0.001	0.53	0.88	3.1	8.58
Visual-vestibular
Smooth pursuit	0.67	5	<0.001	0.40	0.83	1.65	4.59
Voluntary saccades	0.53	3.2	0.001	0.20	0.75	0.33	0.91

Note. AMEDA = Active Movement Extent Discrimination Assessment.

The Bland-Altman analysis is presented in Table 3 and the plots generated for each variable to visualise the agreement between the two assessment sessions are illustrated in Appendix 2. The analysis revealed only small deviations of the average difference from zero, random distribution of the difference measures and fairly narrow limits of agreement. As expected, due to the lower ICCs, the analysis for smooth pursuit and voluntary saccades showed the largest spread of differences. For this reason, these measures were further explored using age stratification, see Appendix 2. Smooth pursuit appeared to have an age effect with smaller spread of differences in the young age group while voluntary saccades was random across age groups.

Table 3.

Bland Altman Analysis.

Variable	Mean bias	Std Dev of differences	Lower limit of agreement	Upper limit of agreement
Autonomic
Response delay	5.1	10.9	−16.3	26.5
Parasympathetic	−0.10	0.37	0.62	−0.82
Sympathetic	0	0.07	−0.13	0.13
Somatosensation
AMEDA left	−2.5	4.6	−11.6	6.6
AMEDA right	−0.9	4.4	−9.5	7.7
Visual-vestibular
Smooth pursuit	0.4	2.4	−5.1	4.2
Voluntary saccade	−0.03	0.47	−0.95	0.90

Note. AMEDA = Active Movement Extent Discrimination Assessment.

Results from the paired-samples t test or Wilcoxon signed-rank test, depending on normality assessment, showed a significant difference between Test 1 and Test 2 for left AMEDA (Wilcoxon test, p = .014, r = 0.457) and response delay (Paired t test, p = .0201, Cohen’s d = −0.294). In contrast, no significant differences were found for right AMEDA (Wilcoxon test, p = .414, r = 0.140), parasympathetic function (Paired t test, p = .173, d = 0.147), sympathetic function (Paired t test, p = .819, d = 0.0), smooth pursuit (Paired t test, p = .334, d = 0.153) or voluntary saccades (Paired t test, p = .776, d = 0.042), indicating stability in performance across the two test sessions (Table 4).

Table 4.

Paired t-Tests and Wilcoxon Signed-Rank Tests, With Effect Size, for Test 1 and Test 2 Scores.

Variable	Normality p value	Test type	p value	Effect size
Autonomic
Response delay	0.547	Paired t test	0.020*	Cohen’s d = −0.294
Parasympathetic	0.334	Paired t test	0.173	Cohen’s d = 0.147
Sympathetic	0.800	Paired t test	0.819	Cohen’s d = 0
Somatosensation
AMEDA left	0.032*	Wilcoxon test	0.014*	r = 0.457
AMEDA right	0.007*	Wilcoxon test	0.414	r = 0.140
Visual-vestibular
Smooth pursuit	0.699	Paired t test	0.334	Cohen’s d = 0.153
Voluntary saccade	0.605	Paired t test	0.776	Cohen’s d = 0.042

Note. * denotes statistically significant result. AMEDA = Active Movement Extent Discrimination Assessment.

Discussion

This study was the first to establish the reliability of a novel clinical protocol assessing VVS systems, along with ANS function, using the Prism Neuro^® AMEDA and eye-tracking technology. Consistent with our hypothesis, the protocol demonstrated moderate to good reliability across the various measures, confirming its potential utility in clinical settings. Additionally, there were no adverse outcomes using this protocol, highlighting its safety and usability even in participants aged up to 88 years old.

The highest reliability was observed in the autonomic metrics, which demonstrated superior consistency compared to other test measures. This result can be attributed to the reflexive nature of these metrics, as they are governed by autonomic processes and are not influenced by cognitive input from participants (McDougal & Gamlin, 2015). Reflexive responses, such as the PLR, are mediated by well-defined neural pathways, which are less susceptible to variability (Barbur, 2004; Hall & Chilcott, 2018). Additionally, there were no reported external factors between Test 1 and Test 2, as these tests occurred within the same session, that could have significantly altered the autonomic state of participants, further supporting the stability of these measures. Consequently, the lack of cognitive or external interference likely contributed to the high reliability observed in the autonomic metrics, underscoring their robustness as objective indicators of autonomic function (Bradley et al., 2008). Administering the test within the same session was deemed the most appropriate approach to minimise influence of external factors on autonomic function. Autonomic function, as measured by PLR, is highly sensitive to acute physiological and psychological stressors, including but not limited to changes in fatigue (Różanowski et al., 2015), substance use such as caffeine (Hartmann et al., 2024), and exposure to painful stimulus (Plouffe & Stelmack, 1979). Conducting assessments on separate days could introduce variability due to these factors, potentially compromising the reliability of the results. Therefore, performing the tests in a single session ensured a more controlled assessment of autonomic function. Although a statistically significant difference was observed between Test 1 and Test 2 for the response delay metric (paired t test, p = .0201), the calculated effect size (Cohen’s d = −0.294) suggests this difference was minimal and likely to be of limited practical significance. This is further supported by the MDC for this metric (MDC = 21.56msec), compared with a mean difference between Tests of only 5 milliseconds.

The AMEDA test evaluates participants’ perception of ankle position, relying on cognitive input as an integral component of the assessment. In this study, the test demonstrated moderate to good reliability, consistent with findings from other test-retest reliability studies using different iterations of the AMEDA (Pan et al., 2022; Shi et al., 2023). Comparable research has reported test-retest ICCs within the moderate to good range, indicating that the reliability of psychometric proprioception assessments like the AMEDA test is generally acceptable (Shi et al., 2023). Cognitive involvement in proprioceptive tasks inherently introduces variability, as these tests assess both sensory processing and conscious perception of joint position (Han et al., 2016). Despite this, the observed reliability suggests that the AMEDA test provides a robust method for evaluating proprioceptive accuracy in functional contexts.

While previous studies have shown no significant difference between left and right AMEDA scores in healthy populations (Pan et al., 2022), our results indicated greater reliability on the right side, which was consistently assessed second. This is potentially explained with the apparent learning effect demonstrated on the left side (Wilcoxon test, p = .014). However, the AUC change between Test 1 and Test 2 on the left AMEDA was only 2.54. Given that the minimal detectable change (MDC) for left and right AMEDA was calculated to be 9.28 and 8.58 AUC respectively, this small change does not meet the threshold for a meaningful learning effect. Future research could explore why reliability improved on the right side, while no corresponding change in ability was observed. While the mean AUC score on the left AMEDA showed a statistically significant increase from Test 1 to Test 2, the percentage change from the left to the right side on Test 1 was only 0.80%, indicating no substantial order effect.

Similar to AMEDA testing, smooth pursuit and voluntary saccades require cognitive effort from the participant and are not reflexive tests like the PLR assessment. Previous research has reported good intraclass correlation coefficients (ICCs) of 0.79 (95% CI: 0.70, 0.86) for smooth pursuit; however, the participants in that study were younger, with a mean age of 22.7 ± 7.4 years (Pearce et al., 2023). In contrast, the current study included participants up to 88 years of age, with a mean age of 46.6 ± 18.9 years, which may partly explain the moderate ICC observed for smooth pursuit. The mean age of the participants in Pearce et al.'s study is more comparable to the ‘young adult’ group in the present study, which had a mean age of 25.7 ± 5.8 years. The Bland-Altman plot with age stratification revealed a smaller spread of differences in the young adult group, prompting a recalculation of the ICC for this subset. The recalculated ICC(3,1) for young adult participants was 0.819, indicating good reliability between the two trials. This finding is consistent with the good test-retest reliability observed by Pearce et al. (2023), supporting the robustness of smooth pursuit measures.

Each measure, including left and right AMEDA, smooth pursuit, voluntary saccades, and autonomic function (assessed through response delay, parasympathetic, and sympathetic measures), exhibited acceptable levels of reliability. These findings align with prior research indicating the reliability of individual assessments of these systems. However, our study’s novelty lies in its integration of these assessments into a single, time-efficient protocol, that is also fully portable, making it more accessible for routine clinical use.

As with all research, the findings of this study must be interpreted within the context of its limitations. One limitation is that the study was not powered to detect between-group differences, particularly across different age groups. Understanding age-related declines in VVS function is crucial for developing targeted balance interventions, as deficits in these systems are linked to increased fall risk. While the study sample was intentionally broad to reflect a general adult population, this demographic heterogeneity may have influence reliability estimates. Although our primary aim was to assess test-retest reliability across a general population, we acknowledge that age- and gender-related physiological difference may have affected measurement consistency. Future research should include larger, stratified samples to allow for a more detailed analysis of age- and gender-related differences in VVS and ANS function and test-retest reliability.

Conclusion

This study provides evidence supporting the reliability of a comprehensive, integrated assessment of VVS and ANS function. This protocol, due to its brevity and ease of administration, holds promise for enhancing the assessment of balance-related impairments in clinical practice. Future research should build on these findings by exploring the application of this protocol across diverse populations and settings to fully realise its potential in improving balance interventions and reducing fall risk.

Supplemental Material

Supplemental Material - Novel Clinical Assessment of Visual, Vestibular, Somatosensory, and Autonomic Function: Establishing Test Re-Test Reliability in a Healthy Population

Supplemental Material for Novel Clinical Assessment of Visual, Vestibular, Somatosensory, and Autonomic Function: Establishing Test Re-Test Reliability in a Healthy Population by Michaela McGrath, Gordon Waddington, Bernie Bissett, Angie Fearon, and Jeremy Witchalls in Perceptual and Motor Skills.

Footnotes

Acknowledgements

We would like to thank all participants for their time and contribution to this study. Your involvement has been invaluable to our research.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interest

Professor Gordon Waddington is a founding shareholder in Prism Neuro Pty Ltd. The lead author, Michaela McGrath, shares the surname of founders of Prism Neuro Pty Ltd however receives no financial support from Prism Neuro Pty Ltd nor holds shares in this entity.

Ethical Statement

ORCID iDs

Michaela McGrath

Gordon Waddington

Jeremy Witchalls

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.*

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Michaela McGrath is a physiotherapist and PhD candidate in the Research Institute for Sport and Exercise at the University of Canberra. Her research is focused on the impact of long COVID on sensorimotor and autonomic nervous system function.

Gordon Waddington is the Australian Institute of Sport Research Professor of Sports Medicine, at the University of Canberra Research Institute for Sport and Exercise. His major research interest is in the assessment of human somatosensory function and its relationship to movement performance and the impact of injury and rehabilitation.

Bernie Bissett is the Joint Professot of Allied Health Research at the University of Canberra and a physiotherapist specialising in intensive care and acute care rehabilitation. Her primary research interest lies in improving recovery and quality of life for ICU patients and individuals recovering from COVID-19.

Angie Fearon is a physiotherapist, academic lecturer, and researcher. Her major research interest is in lower limb musculoskeletal conditions, particularly the assessment and management of greater trochanteric pain syndrome (GTPS), with the goal of providing clinicians with practical, evidence-based tools for everyday challenges.

Jeremy Witchalls is a physiotherapist, academic lecturer, and researcher. He completed a PhD focussed on proprioception in ankle and lower limb, and conducts ongoing research. His primary focus of research is musculoskeletal injury, and the role of somatosensory perception, in injury prediction, prevention and recovery.

References

Antcliff

Welvaert

Witchalls

Wallwork

S. B.

Waddington

(2021). Assessing proprioception in an older population: Reliability of a protocol based on active movement extent discrimination. Perceptual and Motor Skills, 128(5), 2075–2096. https://doi.org/10.1177/00315125211029906

AusGov . (2021). Physical activity and exercise guidelines for all Australians. https://www.health.gov.au/topics/physical-activity-and-exercise/physical-activity-and-exercise-guidelines-for-all-australians

Bär

K.-J.

Schulz

Koschke

Harzendorf

Gayde

Berg

Voss

Yeragani

V. K.

Boettger

M. K.

(2009). Correlations between the autonomic modulation of heart rate, blood pressure and the pupillary light reflex in healthy subjects. Journal of the Neurological Sciences, 279(1-2), 9–13. https://doi.org/10.1016/j.jns.2009.01.010

Barbur

J. L.

(2004). Learning from the pupil: Studies of basic mechanisms and clinical applications. In Chalupa

L. M.

Werner

J. S.

(Eds.), The visual neurosciences (Vol. 1, pp. 641–656). MIT Press.

Borg

D. N.

Bach

A. J.

O’Brien

J. L.

Sainani

K. L.

(2022). Calculating sample size for reliability studies. PM&R, 14(8), 1018–1025. https://doi.org/10.1002/pmrj.12850

Bradley

M. M.

Miccoli

Escrig

M. A.

Lang

P. J.

(2008). The pupil as a measure of emotional arousal and autonomic activation. Psychophysiology, 45(4), 602–607. https://doi.org/10.1111/j.1469-8986.2008.00654.x

Cohen

. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Erlbaum.

Davids

Williams

A. M.

Williams

J. G.

(2005). Visual perception and action in sport. Routledge.

Fetter

(2007). Vestibulo-ocular reflex. Developments in Ophthalmology, 40, 35–51. https://doi.org/10.1159/000100348

10.

Fritz

C. O.

Morris

P. E.

Richler

J. J.

(2012). Effect size estimates: Current use, calculations, and interpretation. Journal of Experimental Psychology: General, 141(1), 2–18. https://doi.org/10.1037/a0024338

11.

Gaerlan

M. G.

(2010). The role of visual, vestibular, and somatosensory systems in postural balance. University of Nevada, Las Vegas.

12.

Giavarina

(2015). Understanding bland altman analysis. Biochemia Medica, 25(2), 141–151. https://doi.org/10.11613/BM.2015.015

13.

Goodman

J. M.

Bensmaia

S. J.

(2020). The neural mechanisms of touch and proprioception at the somatosensory periphery. In Fritzsch

(Ed.), The senses: A comprehensive reference (2nd ed., pp. 2–27). Elsevier. https://doi.org/10.1016/B978-0-12-805408-6.00014-2

14.

Gordon

E. M.

Chauvin

R. J.

Van

A. N.

Rajesh

Nielsen

Newbold

D. J.

Lynch

C. J.

Seider

N. A.

Krimmel

S. R.

Scheidter

K. M.

Monk

Miller

R. L.

Metoki

Montez

D. F.

Zheng

Elbau

Madison

Nishino

Myers

M. J.

… Dosenbach

N. U. F.

(2023). A somato-cognitive action network alternates with effector regions in motor cortex. Nature, 617(7960), 351–359. https://doi.org/10.1038/s41586-023-05964-2

15.

Hall

C. A.

Chilcott

R. P.

(2018). Eyeing up the future of the pupillary light reflex in neurodiagnostics. Diagnostics (Basel), 8(1), 19. https://doi.org/10.3390/diagnostics8010019

16.

Han

Waddington

Adams

Anson

Liu

(2016). Assessing proprioception: A critical review of methods. JSHS, 5(1), 80–90. https://doi.org/10.1016/j.jshs.2014.10.004

17.

Hartmann

E. V.

Reichert

C. F.

Spitschan

(2024). Effects of caffeine intake on pupillary parameters in humans: A systematic review and meta-analysis. Behavioral and Brain Functions, 20(1), 19. https://doi.org/10.1186/s12993-024-00245-y

18.

Hayano

Yuda

(2019). Pitfalls of assessment of autonomic function by heart rate variability. Journal of Physiological Anthropology, 38(1), 3–8. https://doi.org/10.1186/s40101-019-0193-2

19.

Horak

F. B.

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

20.

Kalloniatis

Luu

(2005). Light and dark adaptation. University of Utah Health Sciences Center.

21.

Kirkwood

B. R.

Sterne

J. A.

(2010). Essential medical statistics. John Wiley & Sons.

22.

Koo

T. K.

M. Y.

(2016). A guideline of selecting and reporting intraclass correlation coefficients for reliability research. Journal of Chiropractic Medicine, 15(2), 155–163. https://doi.org/10.1016/j.jcm.2016.02.012

23.

Lal

Truong

(2019). Eye movement abnormalities in movement disorders. Clin Parkinsonism Related Disorders, 1(1), 54–63. https://doi.org/10.1016/j.prdoa.2019.08.004

24.

Leigh

R. J.

Zee

D. S.

(2015). The neurology of eye movements. Oxford University Press.

25.

McDougal

D. H.

Gamlin

P. D.

(2015). Autonomic control of the eye. Comprehensive Physiology, 5(1), 439–473. https://doi.org/10.1002/cphy.c140014

26.

Merla

J. L.

Spaulding

S. J.

(1997). The balance system: Implications for occupational therapy intervention. Physical & Occupational Therapy in Geriatrics, 15(1), 21–36.

27.

Pan

Wang

Rong

Ferland

Adams

Han

Zhu

(2022). Assessing bilateral ankle proprioceptive acuity in stroke survivors: An exploratory study. Frontiers in Neurology, 13, Article 929310. https://doi.org/10.3389/fneur.2022.929310

28.

Pearce

A. J.

Daly

Ryan

King

(2023). Reliability of a smooth pursuit eye-tracking system (EyeGuide focus) in healthy adolescents and adults. J Funct Morphol Kinesiol, 8(2), 83. https://doi.org/10.3390/jfmk8020083

29.

Peterka

R. J.

(2002). Sensorimotor integration in human postural control. Journal of Neurophysiology, 88(3), 1097–1118. https://doi.org/10.1152/jn.2002.88.3.1097

30.

Plouffe

Stelmack

R. M.

(1979). Neuroticism and the effect of stress on the pupillary light reflex. Perceptual and Motor Skills, 49(2), 635–642. https://doi.org/10.2466/pms.1979.49.2.635

31.

Pollock

A. S.

Durward

B. R.

Rowe

P. J.

Paul

J. P.

(2000). What is balance? Clinical Rehabilitation, 14(4), 402–406. https://doi.org/10.1191/0269215500cr342oa

32.

Portney

L. G.

(2020). Foundations of clinical research: Applications to evidence-based practice. F. A. Davis.

33.

Różanowski

Bernat

Kamińska

(2015). Estimation of operators’ fatigue using optical methods for determination of pupil activity. International Journal of Occupational Medicine & Environmental Health, 28(2), 263–281. https://doi.org/10.13075/ijomeh.1896.00274

34.

Shi

Ganderton

Tirosh

Adams

Ei-Ansary

Han

(2023). Test-retest reliability of ankle range of motion, proprioception, and balance for symptom and gender effects in individuals with chronic ankle instability. Musculoskeletal Science and Practice, 66, Article 102809. https://doi.org/10.1016/j.msksp.2023.102809

35.

Shumway-Cook

Woollacott

M. H.

(2012). Motor control: Translating research into clinical practice (4th ed.). Lippincott Williams & Wilkins.

36.

Termsarasab

Thammongkolchai

Rucker

J. C.

Frucht

S. J.

(2015). The diagnostic value of saccades in movement disorder patients: A practical guide and review. Journal of clinical movement disorders, 2, Article 14. https://doi.org/10.1186/s40734-015-0025-4

37.

Van Der Torn

Van Dijk

(2000). Testing the central vestibular functions: A clinical survey. Clinical Otolaryngology and Allied Sciences, 25(4), 298–304. https://doi.org/10.1046/j.1365-2273.2000.00371.x

38.

Waddington

Adams

(1999a). Ability to discriminate movements at the ankle and knee is joint specific. Perceptual and Motor Skills, 89(3 Pt 1), 1037–1041. https://doi.org/10.2466/pms.1999.89.3.1037

39.

Waddington

Adams

(1999b). Discrimination of active plantarflexion and inversion movements after ankle injury. Australian Journal of Physiotherapy, 45(1), 7–13. https://doi.org/10.1016/s0004-9514(14)60335-4

40.

Winter

D. A.

(1995). Human balance and posture control during standing and walking. Gait & Posture, 3(4), 193–214. https://doi.org/10.1016/0966-6362(96)82849-9

41.

Witchalls

J. B.

Waddington

Adams

Blanch

(2014). Chronic ankle instability affects learning rate during repeated proprioception testing. Physical Therapy in Sport, 15(2), 106–111. https://doi.org/10.1016/j.ptsp.2013.04.002

42.

Wix

(2024). $200k virtual reality system enabling cutting-edge balance research at US university. Audiology Worldnews. https://www.audiology-worldnews.com/world-news/research/200k-virtual-reality-system-enabling-cutting-edge-balance-research-at-us-university/

43.

Zatsiorsky

V. M.

Duarte

(1999). Instant equilibrium point and its migration in standing tasks: Rambling and trembling components of the stabilogram. Motor Control, 3(1), 28–38. https://doi.org/10.1123/mcj.3.1.28

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.94 MB

0.00 MB