Abstract
BACKGROUND:
Machado Joseph Disease (MJD) is an autosomal dominant neurodegenerative disease. In previous studies, we described significant bilateral horizontal Vestibulo-Ocular Reflex (VOR) deficit within this population without any reference to the presence of vestibular symptomatology.
OBJECTIVE:
To evaluate whether, beyond cerebellar ataxia complaints, MJD patients have typical vestibular symptomatology corresponding to the accepted diagnostic criteria of Bilateral Vestibulopathy (BVP) according to the definition of the International Barany Society of Neuro-Otology.
METHODS:
Twenty-one MJD, 12 clinically stable chronic Unilateral Vestibulopathy (UVP), 15 clinically stable chronic BVP, and 22 healthy Controls underwent the video Head Impulse Test (vHIT) evaluating VOR gain and filled out the following questionnaires related to vestibular symptomatology: The Dizziness Handicap Inventory (DHI), the Activities-specific Balance Confidence Scale (ABC), the Vertigo Visual Scale (VVS) and the Beck Anxiety Inventory (BAI).
RESULTS:
The MJD group demonstrated significant bilateral vestibular impairment with horizontal gain less than 0.6 in 71% of patients (0.54±0.17). Similar to UVP and BVP, MJD patients reported a significantly higher level of symptoms than Controls in the DHI, ABC, VVS, and BAI questionnaires.
CONCLUSIONS:
MJD demonstrated significant VOR impairment and clinical symptoms typical of BVP. We suggest that in a future version of the International Classification of Vestibular Disorders (ICVD), MJD should be categorized under a separate section of central vestibulopathy with the heading of bilateral vestibulopathy. The present findings are of importance regarding the clinical diagnosis process and possible treatment based on vestibular rehabilitation.
Keywords
Introduction
Machado Joseph Disease (MJD), also known as Spinocerebellar Ataxia type 3 (SCA3), is an autosomal dominant neurodegenerative disease with slowly progressive cerebellar signs, including poor coordination of gait, dysarthria, and abnormal eye movements such as saccadic smooth pursuit, gaze evoked nystagmus and saccadic dysfunction [3, 44]. The diagnosis is confirmed by genetic testing, which detects the disease-specific expansion of CAG trinucleotide repeats in the ATXN3 gene located on chromosome 14q32.12. While normal alleles have between 12 to 44 repeats, MJD patients have 51 to 87 repeats [3]. The mean age at onset (AO) is 34–40 years [7, 48] and the mean survival is 21.2 years after the onset of symptoms [25, 46].
In previous studies using the clinical Head Impulse Test (HIT) and the laboratory video HIT (vHIT), we described significant bilateral horizontal Vestibulo-Ocular Reflex (VOR) deficit in MJD patients [12, 29] and postulated that a selective degeneration of the vestibular nuclei could account for this finding [14]. An extensive clinical study corroborated that the horizontal VOR deficit detected by the bedside HIT is present in 90% of the 53 examined MJD patients, suggesting that it is part of the MJD population phenotype [46]. We also recently reported that MJD patients have low VOR gain (eye velocity/head velocity) in the horizontal and vertical planes and that the horizontal VOR gain is an excellent classifier of MJD patients compared to healthy controls [14]. All the above findings raised the possibility that the horizontal VOR gain measured by the vHIT can be a potential neurophysiological biomarker of MJD [12], with the advantage of being an easy, simple, and quick test that requires little cooperation from the patient [30]. Despite the clear evidence of bilateral vestibular impairment, MJD is still not included in the classification of disorders with chronic central Bilateral Vestibulopathy BVP.
According to the Consensus document of the International Barany Society of Neuro-Otology [41], chronic BVP is characterized by bilateral impairment of the VOR (such as horizontal canal gain less than 0.6 as measured by the vHIT) and symptoms of instability in walking, dizziness, and blurred vision. Since the prominent symptoms of MJD are signs of cerebellar ataxia, previous studies that examined the vestibular function in MJD patients focused on the VOR measurement without any reference to the possibility that there may also be vestibular symptomatology, raising the possibility of incorporating the principles of vestibular rehabilitation in MJD and classifying MJD as a neurodegenerative/genetic disease with central BVP.
The purpose of the current work is to evaluate whether beyond cerebellar ataxia, MJD patients have typical vestibular symptomatology which corresponds to the accepted diagnostic criteria of BVP according to the definition of the International Barany Society of Neuro-Otology [41].
Materials and methods
Participants were 21 Jews of Yemenite origin with clinically and genetically confirmed MJD patients (15 women, 6 men, aged 59±16, duration of the disease 8±5 years, Scale for the Assessment and Rating of Ataxia (SARA) score 12.6±4.5); 12 clinically stable chronic Unilateral Vestibulopathy (UVP) patients (8 women, 4 men, aged 64±14) with a previous clinical diagnosis of vestibular neuritis (evaluated at least 6 months after the acute event); 15 clinically stable chronic BVP patients (9 women, 6 men, aged 60±13; 4 due to ototoxicity and 11 idiopathic) based on the Barany Society criteria and evaluated after more than 9 months of final diagnosis and 22 healthy Controls (9 women, 13 men, aged 56±14), the later with no history of neurological, sensory or balance problems. None of the participants had hearing complaints, and bedside hearing testing was normal in all subjects. All participants underwent a detailed clinical neuro-otological evaluation at the Dizziness and Eye Movements Clinic of the Meir Medical Center. The protocol of the study was approved by the Ethics Committee (Institutional Review Board) of the Meir Medical Center, Kfar-Saba, Israel), and followed the tenets of the Declaration of Helsinki. Following a detailed explanation of the study, all patients signed the written informed consent, underwent the vHIT, and filled out a series of questionnaires related to vestibular symptomatology.
Video head impulse test (vHIT)
Angular horizontal and vertical VOR gains (eye velocity/head velocity) were recorded using the video Head Impulse test (vHIT; Otometrics, Natus, USA). This system comprises goggles with a high-speed camera and inertial head movement sensors with a total weight of 60 g. The examiner executed rapid and unpredictable angular rotations of the subject’s head in the three spatial planes stimulating the six semicircular inner ear canals (three in each inner ear) while the subject was asked to fixate on a central target projected on a screen located 150 cm ahead. Head and monocular eye movements were recorded, and the system detected gain deficits together with catch-up re-fixation saccades that appear during (covert saccade) or following (overt saccade) the head impulse. Normative gain is considered at least 0.8 for the lateral canals and 0.7 for the anterior and posterior canals [19]. The test at each plane was repeated until obtaining at least ten valid trials for each canal.
Dizziness handicap inventory (DHI)
A self-report questionnaire that evaluates perceived handicapping effects imposed by dizziness and imbalance symptoms [22] and is widely accepted in the assessment of vestibular symptoms. This test contains 25 questions rated on a 3-point Likert scale with values 0 (No), 2 (Sometimes) and 4 (Yes) and takes about 3 minutes to be completed. The items in the test are grouped by Functional (9 questions), Emotional (9 questions), and Physical (7 questions) subscales. For vestibular dysfunction, at least one study defines Mild (0–30), Moderate (31–60), and Severe (61–100) handicap [43]. We used the validated Hebrew version [23].
The activities-specific balance confidence scale (ABC)
A self-report questionnaire that evaluates confidence in performing certain ambulatory activities. This test contains 16 questions rated from 0 (No confidence) to 10 (Complete confidence) and takes about 5 minutes to complete [6]. We used the validated Hebrew version [33].
Vertigo visual scale (VVS)
On this test, the subject is instructed to indicate on a scale the number that represents the intensity of his/her dizziness ranging from 0 (No dizziness at all) to 10 (Unbearable dizziness). It is important to note that even though the English name of the test refers explicitly to vertigo, the Hebrew word for vertigo is more commonly used for dizziness. We refer to the complaint measured by the VVS as dizziness following the most common understanding of the Hebrew word (“Scharhoret”).
Beck anxiety inventory (BAI)
The BAI test is rated on a 4-point Likert scale with values 0 (Not at all), 1 (Mildly, but it didn’t bother me much), 2 (Moderately, it wasn’t pleasant at the time), and 3 (Severally, it bothered me a lot), used to assess the severity of 21 anxiety symptoms. Its items are descriptive of subjective, somatic, or panic-related anxiety symptoms [2]. The scoring system determines anxiety levels as Minimal (0–7), Mild (8–15), Moderate (16–25), and Severe (26–63).
Statistical analysis
Descriptive statistics for vHIT results were used to characterize each group. MANOVA was used to compare the groups on the three subscales of DHI with Bonferroni as a post-hoc test. ANOVA was used to compare the groups on the horizontal VOR gain, total DHI, ABC, VVS, and BAI tests with Bonferroni as a post-hoc test. Possible correlation between horizontal VOR and the different questionnaires scores were investigated using a two-tailed Pearson coefficient test.
Results
Table 1 shows the averaged VOR gain of the two ears and the weakest and strongest gain for each canal measured in all groups. Because the definition of BVP refers to the horizontal VOR impairment [39], we present the statistical analysis of the horizontal VOR only. Fifteen out of the 21 MJD patients (71%) demonstrated horizontal VOR gain lower than 0.6 in both left and right canals, and the other 6 patients demonstrated VOR gain between 0.6 to 0.8. Analysis of the horizontal VOR gain found a significant effect of groups [F(3,66) = 70.03, p < 0.001]. Post-hoc comparison between groups showed that MJD as well as UVP and BVP had significantly lower horizontal VOR gain compared to the healthy Controls (p < 0.001) with significantly lower VOR gain in the BVP compared to the MJD group (p < 0.001), with no other significant differences between the three patient groups. Figure 1a–d shows the vHIT results for MJD, BVP, UVP, and healthy Control individual participants.
VOR Gain as measured by vHIT in MJD, UVP, BVP and Control groups. Avg: Average of two ears. For the UVP group Weak means affected side; Strong means unaffected side, and Avg is not presented (NP) because it is no relevant
VOR Gain as measured by vHIT in MJD, UVP, BVP and Control groups. Avg: Average of two ears. For the UVP group Weak means affected side; Strong means unaffected side, and Avg is not presented (NP) because it is no relevant

The video Head Impulse Test (vHIT). Example of recordings of an (a) healthy Control (b) UVP patient (c) BVP patient and (d) MJD patient on the vHIT: Velocity in degree/sec of the head; (light blue for the left side and orange for the right side), the eye (green) and the saccade when present (red) are shown for each semicircular canal together with the mean gain. The central hexagon is a graphical representation of the VOR gain relative to the age-normative data, showing normal range in green and lower range values in red.
Figure 2 shows the results of the Dizziness Handicap Inventory (DHI). Analysis for the DHI total score found a significant effect of groups [F(3,66) = 34.88, p < 0.001]. Post-hoc comparison between groups found that MJD as well as UVP and BVP scored significantly higher vs the healthy Controls (p < 0.001) with no significant differences between the three patient groups (MJD vs UVP vs BVP; p>0.05). Analysis of the three DHI’s subscales found a significant effect of groups [Wilks Lambda F(3,64) = 77.65, p < 0.001] on the Emotional [F(3,66) = 16.83, p < 0.001] Functional [F(3,70) = 28.71, p < 0.001] and Physical [F(3,70) = 28.25, p < 0.001] subscales. MJD as well as UVP and BVP demonstrated higher score on the three subscales of the DHI compared to healthy Controls (p < 0.001 in all cases). Except for finding a higher score on the Functional DHI in MJD compared to the UVP group (p < 0.01), no differences were found between the three patient’s groups in the three subscales of the DHI (p>0.05 for all other cases).

The Dizziness Handicap Inventory (DHI). A score of the degree of disability caused by dizziness (mean±SEM) in the MJD, BVP, UVP, and Control groups. The dotted line marks the cutoff level for the clinical cut-off. p < **0.001, p < *0.01.
Figure 3 shows the results of the Activities-specific Balance Confidence Scale (ABC). Analysis of the ABC scores showed a significant effect of groups [F(3,66) = 28.54, p < 0.001]. Post-hoc comparison of groups found that MJD as well as UVP and BVP groups reported a higher level of imbalance compared to healthy Controls (p < 0.001). The MJD group reported a higher level of imbalance compared to BVP and UVP (p < 0.05 in both cases). No significant differences were found between BVP and UVP groups (p>0.05).

The Activities-specific Balance Confidence Scale (ABC): A score of the degree of confidence in balance (mean±SEM) in the MJD, BVP, UVP, and Control groups. p < **0.001, p < *0.01.
Figure 4 shows the results of the Vertigo Visual Scale (VVS). Analysis for the VVS scores showed a significant effect of groups [F(3,65) = 5.6, p < 0.01]. Post hoc comparison between groups found that MJD and BVP report a higher level of vertigo than Controls (p>0.01 and p < 0.05, respectively). No significant differences were found when comparing UVP to the other groups and between MJD and BVP (p>0.05).

Vertigo Visual Scale (VVS): A score of the degree of Dizziness (mean±SEM) in the MJD, BVP, UVP, and Control groups. p < **0.001, p < *0.01.
Figure 5 shows the results of the Beck Anxiety Inventory (BAI). Analysis of the BAI score found a significant effect of groups [F(3,66) = 8.02, p < 0.001]. Post-hoc comparison between groups showed that MJD and BVP groups report higher anxiety levels than Controls (p < 0.01 and p < 0.001, respectively). No significant differences were found when comparing UVP to the other groups and between MJD and BVP (p>0.05).

Beck Anxiety Inventory (BAI): A score of the degree of anxiety (mean±SEM) in the MJD, BVP, UVP, and Control groups. p < **0.001, p < *0.01.
A weak significant correlation was found only between the VOR gain and the physical DHI scale score in the MJD group (r = 0.49, p < 0.05). The rest of the correlations did not reach the level of significance in any of the study groups.
Discussion
The present study evaluates whether MJD patients have vestibular symptomatology and VOR impairment that meet the diagnostic criteria of BVP according to the Consensus document of the Classification Committee of the International Barany Society of Neuro-Otology [41].We now confirm the presence of significant bilateral vestibular deficits in MJD, with 71% of patients having VOR gain lower than 0.60 and the other 29% having abnormal low VOR gain between 0.6 to 0.8. Previous studies report that MJD patients demonstrated a bilateral VOR impairment without referencing the Barany Society definition of BVP and without reporting on specific vestibular symptomatology or complaint [11, 29].
In the DHI, considered an accepted test for vestibular symptoms, MJD patients scored significantly higher than healthy Controls and similar to chronic peripheral UVP and BVP groups in the total score and all subscales (Functional, Emotional, Physical), except for a higher score on the Functional DHI in MJD as compared to the UVP group. The DHI score of the MJD group is also similar to those of UVP and BVP patients reported in various studies [43] and in a study of a heterogenous group of cerebellar patients [37]. Our data support the concept that MJD patients have indeed vestibular symptomatology.
In the Activities-specific Balance Confidence Scale (ABC), MJD, UVP, and BVP groups present a higher level of insecurity in balance than the healthy Controls as expected [6]. The level of insecurity in the MJD group was higher than these of the UVP and BVP groups, possibly due to the combination of cerebellar and vestibular deficits. Since the imbalance in MJD is mainly caused by cerebellar degeneration, it is very hard to estimate the degree of the vestibulopathic contribution to balance insecurity.
In the Vertigo Visual Scale (VVS), MJD patients reported mild to moderate dizziness levels being higher than healthy Controls and similar to UVP and BVP groups. The VVS scores of the MJD group are also similar to those of chronic peripheral vestibular patients, as reported in previous studies [8]. It is important to note that the Hebrew word “scharhoret” does not distinguish between “dizziness” and “vertigo”. Since dizziness is one of the most typical complaints of vestibular disorders, this finding is in line with our conclusion that MJD has vestibular symptomatology.
Using the Beck Anxiety Inventory (BAI), the MJD patients reported a mild to moderate level of anxiety that was higher compared to healthy Controls and similar to UVP and BVP groups. Although anxiety is not a diagnostic criterion for vestibular disorders, there is an extensive literature describing anxiety in conditions of vestibular impairment in humans and animal vestibular model [1, 40]. Elevated anxiety has also been reported previously in the MJD population [24] and is considered part of the “Cerebellar Cognitive Affective Syndrome” [5]. It is hard to estimate the degree of the vestibular contribution to elevated anxiety beyond the cerebellar ataxia and the psychological distress involved in everyday life with the disease.
There were no significant correlations between the impaired horizontal VOR gain and the different questionnaires evaluating vestibular symptomatology except for a weak correlation with the DHI physical subscale score of the MJD group. This is in line with a large prospective study of patients with chronic vestibular disorders reporting on the lack of correlation between the DHI and various vestibular function tests including the vHIT [45].
Interestingly, and in line with our previous study [17], MJD patients do not complain about oscillopsia. It is probable that the slow and long-term progression of the vestibular loss in MJD is compensated by different mechanisms such as suppression of the perception of oscillopsia, use of effort to localize a target in space, preprogramming of compensatory slow eye movements, and restriction of movement of the head.
Considering all the above findings, we conclude that MJD patients demonstrate an objective bilateral VOR gain impairment and typical symptomatology of vestibular disorder fulfilling the diagnostic criteria of BVP according to the Consensus document of the Classification Committee of the International Barany Society of Neuro-Otology [41].
Although the Barany classification committee document mentioned the possibility of BVP in various cerebellar disorders, the concept of central vestibulopathy has not been explicitly elaborated. Taking into account the central vestibular origin of the VOR deficit in MJD, we suggest that in a future version of the International Classification of Vestibular Disorders (ICVD) [4], MJD should be categorized under a separate section of central vestibulopathy with the heading of bilateral vestibulopathy and included in the list of BVP related to neurodegenerative and/or genetic diseases, together with other diseases such as Friedreich’s Ataxia [31, 47] SCA6 [20], SCA17 [28] and Cerebellar Ataxia with Neuropathy and Vestibular Areflexia Syndrome (CANVAS) [42].
The possibility of central vestibular degeneration in MJD was raised in a pioneering study in 2003 [17] in which we showed severe VOR impairment in a group of MJD patients but not in patients with other cerebellar disorders. This central origin assumption is based on the set of clinical and electrophysiological findings in the recordings of the VOR and eye movements using different techniques [17, 18] and other pathological works showing degeneration of the central vestibular nuclei and pathways in MJD [35, 38].
In previous studies, we and others [12, 49] suggested that the VOR impairment could precede the appearance of the accepted cerebellar clinical signs of MJD and therefore can be used as a biomarker for MJD disease onset. Our present findings support the idea that any dizziness or imbalance complaint in a patient with a family history of MJD but without cerebellar signs should be carefully examined for a vestibular deficit heralding the appearance of clear cerebellar signs. Since in MJD, it is easier to detect cerebellar features in the clinical examination compared to vestibular deficits, clinicians tend not to check the possible vestibular signs, and in fact, signs of cerebellar ataxia may well mask the vestibular signs and symptoms. This may explain the lack of reports of vestibular symptoms in MJD patients. Further studies are necessary to examine the prevalence of vestibular complaints among pre-ataxic MJD carriers.
The present study may also have clinical value concerning the treatment of balance impairment in MJD. Vestibular rehabilitation therapy, through exercises that improve VOR and balance with the help of visual focus during various head and body movements, is effective in the syndrome of BVP [34]. Since MJD patients present with both vestibular impairment and vestibular symptoms, follow-up studies applying vestibular rehabilitation techniques that help balance, improve VOR, and reduce dizziness and anxiety are needed to test the effectiveness of this treatment in MJD.
In conclusion, MJD patients have, in addition to the cerebellar presentation, a significant impairment in VOR and typical symptoms that meet the diagnostic criteria of BVP of the International Barany Society of Neuro-Otology. We suggest that in a future version of the International Classification of Vestibular Disorders (ICVD), MJD should be categorized under a separate section of central vestibulopathy with the heading of bilateral vestibulopathy. The present findings are also of importance regarding the clinical diagnostic process, and possible treatment based on vestibular rehabilitation.
Footnotes
Acknowledgments
The authors thank the patients and families that participated in the study.
Funding
This work was supported by the Israeli Chief Scientist Office, Ministry of Health (CSO MOH, IL) within the framework of the European-Latin America Consortium (EU-LAC) Health Joint Fund (grant # 3-000-14307).
