Abstract
Keywords
Introduction
Multiple sclerosis (MS) is a progressive, neurodegenerative disease of the central nervous system (CNS) that affects over 2.5 million people worldwide [5]. MS-related symptoms can vary tremendously [5], but people living with MS report that symptoms affecting visual function are among the most important [14]. Impaired visual acuity from optic neuritis is the most recognized visual impairment in MS, and typically occurs early, in up to 50% of people with MS [10]. As MS progresses to involve the brainstem and cerebellar structures, the ocular motor system is affected [28], with prevalence ranging from 44% [7] to 80% [33]. In people with MS, ocular motor impairments are associated with more advanced disability [7, 28].
Vestibulo-ocular reflex (VOR) is an important ocular motor process responsible for maintaining visual stability and acuity during head movement. Impaired VOR function causes retinal slip and oscillopsia, resulting in dizziness, and impaired balance and mobility [6, 11]. VOR function is primarily processed by the central vestibular nuclear complex of the brainstem and further modulated by the cerebellum [2, 34]. This is important for people with MS, because between 68–72% of people with MS may have brainstem lesions [24, 30], 87% have abnormal brainstem reflexes [24], and up to 70% present with cerebellar signs [41]. Central vestibular disorder in MS can also lead to problems with balance, dizziness, and fatigue [12, 13]. The relationship between these symptoms and VOR function in people with MS is unclear; however, balance, dizziness, and fatigue have been shown to improve in people with MS following vestibular rehabilitation that includes training the functional capacity of the VOR through gaze stabilization exercises [13]. Similar vestibular rehabilitation programs in people with peripheral vestibular disorders have been shown to improve the functional capacity of VOR as measured by gaze stabilization and dynamic visual acuity [1, 36]. However, it is not known if these rehabilitation strategies would also improve gaze stabilization and dynamic visual acuity in people with MS. In order to better understand the clinical manifestations of VOR function impairment in people with MS, and accurately quantify changes in the functional capacity of the VOR following intervention, valid and reliable clinical tests are needed.
The computerized Gaze Stabilization Test (GST) and Dynamic Visual Acuity Test (DVAT) are two novel clinical tests that measure the functional capacity of the VOR. Though these tests are not direct physiologic measures of VOR, both tests are able to clinically measure the functional capacity of VOR: the GST measures the ability to identify an optotype of a fixed size at maximal head velocity [9], and the DVAT measures the smallest optotype a person can identify at a fixed head velocity [17]. The psychometric properties of the GST [9, 42] and DVAT [17, 35] are favorable, but have been established almost exclusively in people with peripheral vestibular disorders and in healthy adults. Of note, one study included participants with unspecified central vestibular disorder and reported poor-to-fair reliability in both the GST and DVAT, but many of the participants were undergoing vestibular rehabilitation, and the data for final analysis was pooled with participants who had peripheral vestibular disorder [27]. To our knowledge, the psychometric properties of computerized GST and DVAT have not been reported in people with MS.
The objective of this study was to determine the validity and reliability of the GST and DVAT in people with MS. The primary aim was to determine the discriminant validity of the GST and DVAT between people with MS and healthy controls. The hypothesis was that people with MS would have significantly lower GST and DVAT scores compared to healthy controls. This study also examined the discriminant validity between participants with MS who had lower disability versus those with higher disability, the test-retest reliability of the GST and DVAT, and the concurrent validity with patient-reported outcomes (PROs) of balance, dizziness, and fatigue.
Methods
Participants
Thirty people with MS and 15 healthy controls participated in the study. Participants were recruited through advertisements at the University of Colorado Hospital and the University of Colorado Denver Anschutz Medical Campus. Eligibility for people with MS included: 18–60 years old, confirmed diagnosis of MS, ability to provide consent and follow simple directions, ability to walk at least 25 feet with bilateral support (Patient Determined Disease Steps – PDDS≤6). Age and sex comparable controls were included without neurological, muscular, or skeletal disorders. Exclusion criteria for all participants included: corrected visual acuity > 20/40, visual perception time > 80 msec as measured by computerized testing, complete or legal blindness in either eye, movement restriction or pain with 20 degrees of cervical rotation, participation in eye movement training in the last two months, medical history or other cause of vision or ocular motor dysfunction. People with MS were also excluded if they had an MS-related exacerbation within the past month. All participants provided signed informed consent approved by The Colorado Multiple Institution Review Board.
Equipment & setup
GST and DVAT were assessed with the Natus inVision™ SMART Balance Master® software version 9.0 (Clackamas, OR). A head-mounted sensor, the InterSense Inertia Cube2 3-axis integrating gyro, measured head velocity and direction. Participants sat eight feet from a 17-inch high-resolution monitor in a private room with minimal background distractions (Fig. 1 and 2). The ambient light in the dedicated balance performance laboratory was kept constant at 80 Lux.
Outcomes
Computerized GST and DVAT
GST, the primary outcome, measures maximum head velocity achieved while visually fixating on a static optotype, and is recorded in degrees per second (deg/s). The GST has been found to discriminate between people with and without peripheral vestibular disorders [9, 42], fallers and non-fallers [18], and older and younger healthy adults [19, 39]. Test-retest reliability has been reported as fair to good in healthy adults [40].
The DVAT measures the difference between static and dynamic visual acuity and is recorded in units of the logarithm of minimal angle of resolution (logMAR). The logMAR is a conversion of the Snellen eye chart, where 0.1 logMAR represents one line of change, and 0.00 logMAR is equivalent to 20/20 vision. The DVAT has been found to identify people with peripheral vestibular dysfunction [17], discriminate between fallers and non-fallers [20], and has been found to be reliable in both healthy adults and people with peripheral vestibular disorders [17, 35].
Patient-reported outcomes (PROs)
The PDDS, a valid surrogate for the Kurtzke Expanded Disability Status Scale (EDSS) [23], was used to quantify disability level in the participants with MS. Mild disability was defined as PDDS 0–3 and moderate-to-severe disability PDDS 4–6 [21, 25]. Activities-specific Balance Confidence Scale (ABC), Dizziness Handicap Inventory (DHI), and Modified Fatigue Impact Scale (MFIS) were also assessed. The ABC has been found to be a reliable [3] and valid self-report measure to assess balance confidence [4, 29] and fall risk in people with MS [29]. The DHI measures self-reported disability due to dizziness and has been found to be reliable [3] and valid in people with MS [4]. The MFIS measures the impact of fatigue on participation, and has been found to have high test-retest reliability [22] and good correlation with other fatigue scales in people with MS [37]. Mean ABC score and total scores for the DHI and MFIS were reported.
Experimental procedure
This cross-sectional study consisted of two visits 14 days apart for the participants with MS. Following consent, the sequence of experimental procedures for participants with MS included: completing the PROs, the GST, a 5-minute rest, and the DVAT. After a 20-minute rest, participants with MS repeated the GST and DVAT protocols again. Participants with MS then returned two weeks later to complete a single GST and DVAT protocol session. The healthy control participants (n = 15) performed a single session of the GST and DVAT (Fig. 3). Degree of dizziness and oscillopsia was assessed in all participants following each GST and DVAT test on a 100 mm visual analog scale (VAS), where 0 mm represented “no symptoms” and 100 mm represented “worst possible symptoms”.
Testing protocol
The standardized Natus inVision™ protocol generates an optotype, “E”, presented in 1 of 4 directions. Participants were asked to identify the optotype direction, if able, without guessing. With correct identification in 3 trials out of 5, the test difficulty was automatically increased (faster speed for the GST, smaller optotype for the DVAT); with incorrect identification in 3 trials out of 5, the test difficulty decreased (slower speed for the GST speed, larger optotype for the DVAT). The algorithms repeat until the highest head velocity is achieved during the GST and the smallest optotype is identified during the DVAT.
Because people with MS are often susceptible to symptom provocation (e.g., dizziness, oscillopsia, headache, nausea) and performance fatigue from repeated head movements, premature termination of GST and DVAT testing was possible. To minimize this possibility, additional protocol parameters were included for all participants: performing yaw plane movements only; limiting practice sessions to 2 minutes; and resting 10 seconds after each optotype presentation.
Static visual acuity and perception time test
In order to score the GST and DVAT, Static Visual Acuity (SVA) and Perception Time Test (PTT) are required. SVA, recorded in logMAR, is the smallest optotype the participant can identify with the head still. Poor SVA may confound GST and DVAT results, so participants were excluded with SVA≥0.3 logMAR, or 20/40 vision [27]. PTT is the shortest optotype presentation time, in milliseconds, the participant can identify with the head still. PTT optotype size was always fixed at 0.2 logMAR above SVA. GST and DVAT optotype presentation time was up to 35 msec longer than PTT score. Participants who were unable to achieve a PTT≤80 msec after two attempts were excluded from the study to prevent optotype identification by mechanisms other than VOR during the GST and DVAT [40].
Gaze stabilization and dynamic visual acuity tests
For the GST and DVAT the optotype was presented when a target head velocity and amplitude was achieved. The target amplitude for head rotation (yaw) was constant at 20 degrees bilaterally for both tests. Feedback bars on the screen assisted participants in reaching target velocity and amplitude (Fig. 1 and 2). The GST optotype size remained fixed at 0.2 logMAR above SVA while the head velocity requirements varied, from 10–150 deg/s, and got progressively faster or slower depending on performance. The DVAT optotype size started at 0.2 logMAR above SVA, and got progressively smaller or larger depending on performance. Minimum head target velocity was set at the lowest possible speed, 85 deg/s [26], in a continued effort to minimize the effect of motion-induced symptoms.
Statistical analysis
Power and sample size estimates for the study were based on the GST. To estimate the expected between group differences for the GST, preliminary data for people with MS (119±34 deg/s) was compared to published data for healthy controls (155±36 deg/s) [40]. To detect this 36-point mean difference, at a power of 80% using a two-sided significance level of 5%, 15 participants per group were needed. The sample was then doubled in the group of people with MS in order to also determine reliability.
Descriptive statistics were used to characterize the participants. For both the GST and DVAT, separate scores were recorded for left and right head movements, averaged together, and reported as a composite score. Two-sided independent t-tests were used to evaluate differences between initial GST and DVAT scores for participants with MS and controls, as well as between lower and higher disability levels in participants with MS. Discriminate validity was confirmed by statistically significant group differences on the GST (α= 0.05). Test-retest reliability was assessed with two-way random effects, single measurement intraclass correlation coefficient (ICC(2,1)) for both same day (Tests 1 and 2) and across 2 weeks testing (Tests 1 and 3, and Tests 2 and 3). ICC values were considered excellent from 0.75 to 1.0, fair-to-good from 0.40 to 0.74, and poor if under 0.40 [8]. Correlations were analyzed using the Pearson product–moment correlation coefficient (r), and interpretations of coefficient values were: good-to-excellent from 0.75 to 1.00, moderate-to-good from 0.50 to 0.74, fair from 0.25 to 0.49, and minimal-to-none if under 0.25 [31]. SPSS (Version 22, Chicago, Illinois) was used for all statistical analysis.
Results
Thirty-two people with MS and 16 age-comparable controls were consented. One participant with MS was excluded after experiencing neck pain with testing. Two participants, one with MS and one control, were excluded when corrected visual acuity was found to be > 20/40. Of the remaining 30 participants with MS, all returned for the second session (third GST and DVAT), with a mean follow up time of 15.4±3.5 days. Only 29 data points were recorded for the second GST and DVAT, as one participant with MS declined to perform the second set of tests because of fatigue. There were no significant differences in age, sex, or visual acuity between the participants with MS and the healthy controls (Table 1).
For the GST, there was a significant difference between the participants with MS and the healthy controls (43.6 deg/s; p < 0.001) (Table 2). There was also a significant difference in the GST between higher and lower disability groups in participants with MS (30.7 deg/s; p = 0.035) (Table 3). GST test-retest reliability ICC(2,1) values were 0.74 (tests 1 and 3), 0.58 (tests 1 and 2), and 0.48 (tests 2 and 3) (Table 4). There were fair, statistically significant (α= 0.01) correlations between the GST and the ABC (r = 0.48; p = 0.008), DHI (r = –0.47; p = 0.009), and MFIS (r = –0.48; p = 0.007). GST also correlated strongly with the DVAT (r = –0.77; p < 0.001) (Table 5).
For the DVAT, there was a significant difference between participants with MS and the control group (–0.095 logMAR; p = 0.001) (Table 2). There was no difference between lower and higher disability groups on the DVAT (–0.045 logMAR; p = 0.313) (Table 3). The DVAT had fair-to-good reliability for all comparisons (ICCs = 0.47–0.60) (Table 4). Lastly, non-significant correlations between DVAT and PROs were found (r = –0.31 to 0.33; p = 0.077 to 0.099) (Table 5).
The participants with MS who reported higher disability (n = 13) also reported worse scores on the ABC, DHI, and MFIS than those with lower disability (n = 17) (Table 3). Compared to the healthy controls, participants with MS reported significantly more symptoms during GST and DVAT performance: dizziness (GST: 13.6 mm; p = 0.037; DVAT: 15.0 mm; p = 0.045); and oscillopsia (GST: 17.1 mm; p = 0.012; DVAT: –18.4 mm; p = 0.018).
Discussion
The overall objective of this investigation was to better understand the functional capacity of the VOR in people with MS by determining the utility of the GST and DVAT. The GST discriminated between participants with MS and healthy controls. The mean yaw plane GST scores for participants with MS are comparable to prior findings other populations with abnormal GST scores, including adults older than 70 years (95.5 to 104.8 deg/s) [19, 39], and people with peripheral vestibular dysfunction (90 to 93 deg/s) [32]. Additionally, the mean GST scores in this study for the healthy controls are similar to prior studies involving comparable healthy adult samples (141 to 155 deg/s) [9, 40]. This study provides initial evidence that gaze stabilization is impaired in people with MS as measured by the computerized GST.
Further supporting the discriminant validity of the GST, there were significant differences between the participants with MS and higher disability compared to those with lower disability. The GST scores in the participants with MS who have higher and lower disability found in this study were also comparable to prior reports of GST scores in people with unilateral peripheral disorder towards the lesioned side (83.9 to 86.3 deg/s) and the non-lesioned side (105.7 to 112.5 deg/s) [9, 38]. These findings indicated that disability advancement in people with MS might be accompanied by decreased ability to maintain gaze stability.
For test-retest reliability across two weeks, the GST demonstrated good reliability. The findings of this study are higher compared to previous reports of between session reliability across 7–10 days in healthy adults (ICC = 0.59) [40], and in a study that included people with central vestibular disorders (ICCs = 0.00 to 0.33) [27]. The differences between the results in this study and prior results may be because the testing protocol parameters implemented in this study were effective at minimizing the effect of movement-related symptom provocation, leading to improved testing tolerance and performance. In contrast, the same-day test-retest reliability found in this study, though slightly higher than what has been previously reported for same-day reliability in a sample including people with central vestibular disorders (ICCs = 0.38 to 0.48) [27], was lower than same-day reliability for healthy adults (ICC = 0.75) [40]. The reason for this difference might be that healthy adults experience less test-related fatigue than people with MS or other vestibular disorders, which is consistent with findings in this study where participants with MS reported significantly more post-test dizziness and oscillopsia following GST than healthy controls. Nevertheless, only one participant with MS declined to perform the second same-day assessment because of test fatigue. The results of this study suggest that in people with MS a single GST test, using the protocol described in this study, is adequate to identify impairments in gaze stabilization and establish a baseline status of the functional capacity of the VOR.
In this study, GST scores correlated with ABC, DHI, and MFIS scores suggesting worse gaze stabilization is associated with worse self-reported balance, dizziness, and fatigue in people with MS. Additionally, ABC, DHI, and MFIS scores correlated significantly with each other (Table 5), which is similar to previous findings that reported relationships between dynamic balance performance, DHI, and MFIS in people with MS [12, 13]. Collectively, these findings support the need for future studies aimed at improving primary outcomes of both VOR function and balance as proximal outcomes, and perceived dizziness and fatigue as more distal outcomes.
Like the GST, the DVAT discriminated between participants with MS and healthy controls in this study, suggesting that dynamic visual acuity measured by the DVAT is impaired in people with MS. The mean DVAT values for participants with MS in this study were comparable to the lower end of the range of yaw plane DVAT values found in prior studies that included participants with unilateral peripheral vestibular hypofunction (0.21 – 0.36 logMAR) [9, 38], and participants with peripheral and central vestibular disease (0.21–0.29 logMAR) [27].
Same day reliability of the DVAT was comparable to the GST in this study, however, across 2-weeks reliability of the DVAT was lower compared to the GST. Nevertheless, the range of test-retest reliability between sessions found in this study for the DVAT is higher than prior findings in participants with peripheral and central vestibular disease (ICCs = 0.09 to 0.60) [27]. The slightly more favorable and consistent range of ICC values in this study might be attributed to the protocol, which was designed to minimize the effect of fatigue and other movement-induced symptoms.
Finally, the DVAT did not discriminate between participants with MS based on disability, nor was it significantly correlated with any of the PROs. There was a strong correlation between DVAT and GST scores for participants with MS, suggesting that both tests had similar capacity to identify the functional capacity of the VOR. Therefore, if deciding between the two tests, the GST may be more useful as it may identify differences between the higher and lower disability groups, and may be more strongly associated with common MS-related symptoms of balance-confidence, dizziness, and fatigue. More investigations are needed, however, to determine if one test is indeed more useful as an outcome measure in people with MS.
Limitations
The GST and DVAT do not measure physiologic VOR, and are complex tests whose performance may be affected by other impairments commonly found in MS such as coordination, motor learning, or cognition/attention. Therefore further investigation is needed to determine concurrent validity of the GST and DVAT with physiological measures of VOR and other vestibular functions. Secondly, this study did not include a gait-related measure, which might have provided additional important insight as GST scores have been shown to identify gait impairments in people with peripheral vestibular conditions [42]. Of note, however, participants with MS with higher PDDS scores (worse self-reported walking function) also had worse GST scores, lending support to a relationship between gait and GST scores. Thirdly, GST was always tested prior to DVAT, which could have contributed to the comparably lower psychometric properties for the DVAT. Lastly, in the current study the participants only performed yaw head movements. Future studies could also include pitch for a more complete determination of the utility of the tests.
Conclusions
This study found that a single GST and DVAT test each discriminated between participants with MS and healthy controls. The GST and DVAT had comparable within session test-retest reliability, indicating that both can be considered as outcome measures in people with MS. The GST, however, had slightly higher across 2-weeks reliability, discriminated between higher and lower disability groups, and correlated with self-reported balance confidence, dizziness, and fatigue. Future intervention trials are needed to determine the responsiveness of GST and DVAT to vestibular rehabilitation and to provide further support of the utility of the tests in the clinical setting. Meanwhile, clinicians should routinely examine the function capacity of VOR in patients with MS.