Introduction
Balance disorders are a frequent complication of head trauma irrespective of its severity.Reference Davies and Luxon1, Reference Hoffer, Gottshall, Moore, Balough and Wester2
Such patients often report dizziness; however, this vague, imprecise symptom is difficult to identify, quantify or monitor in an objective manner. It is even more difficult to diagnose patients who complain of chronic, nonspecific dizziness long after a head injury. In such cases, differentiating between physiological and ‘non-physiological’ dizziness can be a significant clinical and medico-legal problem.Reference Hart and Rubin3–Reference Staab and Ruckenstein5 Differentiation should ideally be based on objective, quantifiable data. Moreover, there is controversy over the primary site of vestibular dysfunction following head trauma.Reference Davies and Luxon1
In this setting, the ideal diagnostic tool would be capable of delineating the site of dysfunction as peripheral, central or a combination of both, and also of facilitating the planning and follow up of balance disorder treatment.
Traditionally, electronystagmography which relies on the corneo-retinal potential to record the eye movements has been considered the ‘gold standard’ for testing dizzy patients.
Recently, the videonystagmograph has been introduced, and offers a multitude of advantages over traditional electronystagmography protocols. In contrast to electronystagmography, videonystagmography records the eye movements using digital video image technology, utilising infrared illumination to determine eye position. The use of videonystagmography enables simultaneous subjective observation of eye movements together with objective collection and analysis of eye movement waveforms via computer algorithms.Reference McCaslin and Jacobson6
This study aimed to characterise balance disorders occurring after head trauma, according to a number of parameters, using videonystagmography. Moreover, we aimed to test the efficiency of videonystagmography as a tool for both diagnosis and treatment outcome monitoring.
Materials and methods
This study was designed as a prospective, cohort analysis of head trauma cases referred to our university-based, tertiary care balance unit over a two-year period.
The study included 126 patients: 89 (71 per cent) males and 37 (29 per cent) females. Their ages ranged from 16–56 years, with a mean age of 28.5 years (standard deviation ±7.2 years).
In order to improve study reliability, we excluded patients aged less than 16 years. Cases with cervical or serious intracranial injuries were also excluded for the patient's safety. All included patients had no previous complaints of balance disorders of otological or other cause, and had taken no sedating drugs, tranquilisers or vestibular suppressants for at least 48 hours prior to vestibular testing.
Head trauma was classified using the Glasgow coma scale, into severe (Glasgow coma scale ≤ 8), moderate (9–12), mild (13–14) or minor (15). Accordingly, trauma was severe in 16 cases (12.7 per cent), moderate in 24 (19 per cent), mild in 46 (36.6 per cent) and minor in 40 (31.7 per cent). Head trauma was penetrating in 10 (8 per cent) cases and blunt in 116 (92 per cent) cases.
Each patient included in the study underwent the following protocol: detailed otolaryngological history-taking, thorough otolaryngological clinical examination, standard neurological examination, computed tomography (CT) scanning of the skull and brain, pure tone audiometry, and videonystagmography (using a medical videonystagmography computerised analyser; ICS, Schaumburg, Illinois, USA). Videonystagmography testing was carried out within 24 hours of the head trauma, and then repeated at 48 hours, one week, two weeks, one month and then every two months until recovery. Repetition of videonystagmography testing depended on the persistence of subjective dizziness, objective abnormal videonystagmographic findings and clinically detected nystagmus.
The videonystagmography test battery consisted of three groups of tests performed in a systematic fashion and in the same chronological order, to assess the oculomotor and vestibular systems as well as their interactions.Reference Bojrab, Maya Kato, Glasscock and Gulya7
The first group of tests investigated visual oculomotor functions and evaluated nonvestibular eye movements. The saccade test detected disorders of the saccadic control system, while the tracking test and the optokinetic test detected disorders of the pursuit control system.
The second group of tests recorded abnormal eye movements in relation to altered head position. The gaze test evaluated limitations of eye movement gaze stability, ocular flutter, spontaneous nystagmus and latent nystagmus. The positional test assessed the effect of altered head position on the production or modification of nystagmus. The Dix–Hallpike positioning manoeuvre test assessed the presence of benign paroxysmal positional vertigo.Reference Dix and Hallpike8
The third test group assessed vestibulo-oculomotor function by means of a bithermal caloric test to detect dysfunction of the labyrinth or vestibular nerve. We utilised a closed loop circuit for ear irrigation, to prevent the possibility of spread of intracranial infection.
The test battery was interrupted if the patient was unable to tolerate it further. In this case, the patient was allowed to rest for two hours before resuming the test protocol.
Normative data for saccade, smooth pursuit and bithermal caloric testing were obtained from preprogrammed software within the videonystagmographic system.
All patients included in this study had a full range of eye movements, with no oculomotor palsy. Testing utilised a stimulation light bar with a light-emitting diode, used in dim lighting, in order to minimise the effect of visual acuity.
Peripheral vestibular lesions were defined based on the following criteria: (1) unidirectional, spontaneous nystagmus which enhanced after covering the eye goggles (to abolish optic fixation), and/or (2) unilateral canal paresis upon caloric testing, with normal suppression in the presence of optic fixation, with or without a directional preponderance.Reference Davies and Luxon1, Reference Bojrab, Maya Kato, Glasscock and Gulya7
Central vestibular lesions were defined on the basis of: bidirectional or directional changing nystagmus; nystagmus enhanced or unaltered by optic fixation; abnormal saccades; abnormal smooth pursuit; abnormal optokinetic nystagmus; and/or failure of suppression of caloric responses.Reference Davies and Luxon1, Reference Bojrab, Maya Kato, Glasscock and Gulya7
Mixed vestibular lesions were diagnosed based on the presence of a combination of videonystagmographic patterns defining both central and peripheral lesions.
Benign paroxysmal positional vertigo was diagnosed based on previously described criteria, i.e. vertigo and nystagmus preceded by a latent period; rotatory nystagmus towards the undermost ear; and a fatiguable and adaptable response upon repetition of the provocative position.Reference Davies and Luxon1, Reference Dix and Hallpike8
Patients with a diagnosed balance disorder, based on videonystagmography, underwent a standard vestibular rehabilitation programme.Reference Telian9 Each patient received an individualised programme which included habituation exercises (to facilitate central nervous system compensation by negating pathological responses to head movement), postural control exercises and general conditioning activities. These exercises were conducted in the physiotherapy department and were performed in an open area twice a day, six days a week, with a rest on the seventh day. Videonystagmography was used to monitor patients' rehabilitation progress.
Benign paroxysmal positional vertigo was treated using the Epley canalith repositioning manoeuvre.Reference Epley10
No labyrinthine sedatives were given to the patients enrolled in this study.
No treatment was given to patients who had subjective symptoms in the absence of objective videonystagmographic findings.
Results
The videonystagmography test battery, performed during the acute phase of head trauma, was tolerated by 84 per cent of enrolled patients. In the remaining 16 per cent, the test battery had to be interrupted due to severe dizziness with or without vomiting. In almost all cases, test intolerance was experienced after the second group of tests, which involved altering the head position in order to record any related abnormal eye movements. Patients who became intolerant of testing were allowed two hours of rest in which to recover, with the aim of completing the third test group (which included bithermal caloric testing). If severe dizziness precluded the performance of the third test group, this was undertaken the following day. There was no relationship between patients' head trauma severity and their intolerance of videonystagmographic testing.
Following head trauma, the following subjective symptoms of balance disorder were reported: vertigo in 54 (42.9 per cent) cases, unsteadiness in 20 (15.9 per cent) and dizziness in 12 (9.5 per cent). Forty (31.7 per cent) patients reported no balance disorder symptoms following head trauma. Other symptoms unrelated to balance disorder were also reported, most commonly headache and vomiting, reported by 48 (38.1 per cent) and 43 (34.1 per cent) patients, respectively, and also nausea (26 patients; 20.6 per cent) and blurred vision (21 patients; 16.7 per cent). Fifty (39.7 per cent) patients reported no associated symptoms of head trauma.
Videonystagmography enabled differentiation of balance disorders into the following types: peripheral vestibular in 30 (23.8 per cent) cases, central vestibular in 10 (7.9 per cent), mixed vestibular in 16 (12.7 per cent) and benign paroxysmal positional vertigo in six (4.8 per cent).
In this study, temporal bone fractures without evidence of brain damage were only seen in patients with moderate or severe head trauma. Computed tomography detected a transverse temporal bone fracture in two (1.6 per cent) cases and a longitudinal fracture in seven (5.6 per cent) cases. The two transverse fracture cases both showed a peripheral vestibular disorder on videonystagmography. Of the seven longitudinal fracture cases, three showed a peripheral vestibular disorder while the other four showed a mixed vestibular disorder. There was no statistically significant relationship between the temporal bone fracture type and the resulting balance disorder type.
In 64 (50.8 per cent) cases, videonystagmography findings were either within the normal reference range or were not diagnostic of any organic lesion. Of these 64 patients, 40 (31.7 per cent of the total group) were free of symptoms while 24 (19.1 per cent of the total group) had subjective symptoms.
Figure 1 shows the time of onset of the balance disorder after head trauma. While 47.6 per cent of patients experienced an immediate onset of their balance disorder, 31.7 per cent of patients had none at all. The remaining patients showed a delayed onset of balance disorder, varying from one week to more than two months after the initial head trauma.
Fig. 1 Onset of balance disorder following head trauma.
The relationship between head trauma severity and the resulting balance disorder is shown in Figure 2. Those cases with no balance disorder symptoms or only subjective symptoms had all suffered either minor or mild head trauma. All cases of benign paroxysmal positional vertigo followed mild or moderate trauma.
Fig. 2 Grade of head trauma severity in patients with different balance disorders. BPPV = benign paroxysmal positional vertigo
The relationship between the time of onset of the balance disorder (after head trauma) and the severity of head trauma is shown in Figure 3. A delayed onset of symptoms was only seen for patients with mild head trauma.
Fig. 3 Onset of balance disorder in patients with different grades of head trauma severity.
The relationship between the time of balance disorder onset (after head trauma) and the type of balance disorder (on videonystagmographic diagnosis) is shown in Figure 4. The balance disorder type with the longest mean time of onset was benign paroxysmal positional vertigo; one-third of patients with this balance disorder type had a time of onset of more than two months.
Fig. 4 Onset of balance disorder in patients with different balance disorders.
The recovery time required for patients with various balance disorder types is shown in Figure 5. Patients suffering only subjective symptoms reported resolution of their symptoms by the second week following head trauma. Patients with other types of balance disorder required a longer recovery period.
Fig. 5 Recovery time in patients with different balance disorders.
The relationship between balance disorder recovery time and head trauma severity is shown in Figure 6. There was a significant difference in recovery time, comparing moderate and severe trauma versus minor and mild trauma.
Fig. 6 Recovery time in patients with different grades of head trauma severity.
Discussion
Analysis of our study data indicates that several factors affect the nature of balance disorders caused by head trauma.
One of these factors is the severity of the head trauma. In our study, all patients with either no post-trauma balance disorder symptoms or only subjective symptoms had suffered minor or mild head trauma. Two-thirds of the benign paroxysmal positional vertigo cases had suffered mild head trauma, while one-third had suffered moderate head trauma. In contrast, peripheral, central and mixed vestibular lesions (as diagnosed by videonystagmography) occurred after mild, moderate or severe head trauma, but not after minor head trauma. Interestingly, 60 per cent of central vestibular lesions occurred after mild head trauma (rather than moderate or severe trauma). This agrees with reports which state that mild head injury can have severe sequelae.Reference Hsiang, Yeung, Yu and Poons11–Reference Servadei, Vergoni, Pasini, Fagioli, Arista and Zappi13 This is why we differentiated less severe head trauma into minor head injury (i.e. Glasgow coma scale of 15) and mild head injury (i.e. Glasgow coma scale of 13 or 14), and excluded cases with serious intracranial injuries. As previously reported, we too observed a different pattern of balance disorder presentation in patients subjected to minor versus mild head trauma.Reference Hsiang, Yeung, Yu and Poons11 In the current study, patients suffering minor head trauma complained only of subjective symptoms, and had no videonystagmographic findings.
The severity of initial head trauma also affected the time of onset of any subsequent balance disorder. Moderate and severe head trauma led to an immediate onset of balance disorder. After minor head trauma, two-thirds of cases experienced an immediate onset of balance disorder, while the remaining one-third developed a balance disorder within the first week.
Protracted development of post-traumatic balance disorder was seen only in mild head trauma cases. While more than half of the patients suffering mild head trauma developed a balance disorder either immediately or within one week, almost one-third of such cases developed a balance disorder between two weeks and two months after trauma. Only six per cent of mild head trauma cases developed a balance disorder during the second week post-trauma, and a similar percentage developed a balance disorder more than two months post-trauma.
Of patients diagnosed with benign paroxysmal positional vertigo, symptoms occurred either immediately or within the first week post-trauma in two-thirds of cases. However, one-third of benign paroxysmal positional vertigo cases developed their balance disorder after the second month post-trauma.
This pattern of a delayed onset of balance disorders following mild head trauma, in the absence of demonstrable CT and videonystagmographic findings, may provide some insight into the pathophysiological mechanisms of balance disorders occurring after head trauma, particularly mild head trauma. Post-traumatic neurological manifestations of mild head trauma have been classified into acute and delayed types.Reference Sakas, Whittaker, Whitwell and Singounas14 Delayed neurological manifestations are postulated to involve a variety of pathological causes, from early neuronal reflex phenomena to late cerebral oedema (due to hyperaemia or ischaemia, probably based on disturbed autoregulation of cerebral vasoreactivity).Reference Urasaki, Yasukouchi, Yokota and Aragaki12, Reference Sakas, Whittaker, Whitwell and Singounas14 Hyperaemia in such cases is termed ‘benign’ as it develops in structurally intact brain tissue and does not exhibit the features of congestive swelling, and therefore has only a minimal effect on intracranial pressure and clinical outcome.Reference Sakas, Bullock, Patterson, Hadley, Wyper and Teasdale15 However, such noncongestive cerebral hyperaemia has been shown (by single photon emission CT) to be associated with temporary neurological deficits.Reference Lewis, Longstreth, Wilkus and Copass16, Reference Aminian, Strashun and Rose17 Such cerebral hyperaemia has been suggested to lead to ischaemia via a decrease in the vasomotor tone, with slowing of the capillary circulation, arteriovenous shunting, and reduction in the capillary gas exchange.Reference Sakas, Whittaker, Whitwell and Singounas14
Therefore, we presume that the delayed onset of balance disorders following mild head trauma, as observed in this study, may represent a more delayed or prolonged hyperaemia phase related to a favourable outcome.
In patients with benign paroxysmal positional vertigo, there is compelling evidence that most cases are related to the presence of free-floating endolymph particles within the lumen of the semicircular canal.Reference Welling, Parnes, O'Brien, Bakaletz, Brackmann and Hinojosa18 It has been suggested that these particles are otoconia displaced from the otolithic membrane in the utricle, which settle in the dependent posterior canal and render it sensitive to gravity.Reference Dix and Hallpike8, Reference Schuknecht and Ruby19, Reference Brandt and Steddin20
While head trauma has been claimed to be the most common cause of otoconia displacement, there is also a strong suggested association between benign paroxysmal positional vertigo and migraine, which is presumed to cause inner ear damage secondary to vasospasm.Reference Ishiyama, Jacobson and Baloh21 In the current study, headache was the most common non-balance-disorder symptom following head trauma, being reported by 38 per cent of patients, in agreement with other investigators.Reference Hoffer, Gottshall, Moore, Balough and Wester2 Accordingly, it may be plausible to consider the delayed presentation of benign paroxysmal positional vertigo seen in the current study as reflecting the indirect, slower mechanism of otolith damage attributed to migraine, as compared with a direct, faster mechanism caused by head trauma and resulting in earlier presentation.
In our patients, we found variation in the duration of symptoms following head trauma, and this adds to the ongoing clinical debate. Some authors believe that head trauma symptoms only last weeks to months; however, many believe that symptoms can persist for years, and may even remain indefinitely.Reference Alexander22, Reference Hugenholtz, Stuss, Stethem and Richard23
Two-thirds of our patients with benign paroxysmal positional vertigo experienced recovery during the first two months post-trauma, while the remainder recovered during the third month. All our cases of benign paroxysmal positional vertigo were managed using the Epley canalith repositioning manoeuvre, which successfully and permanently alleviated patients' symptoms.Reference Epley10
The identification of patients with benign paroxysmal positional vertigo was aided by the unique ability of videonystagmography to record the torsional eye movements characteristic of this condition during Dix–Hallpike testing.Reference Dix and Hallpike8
Patients diagnosed with peripheral, mixed and central vestibular lesions recovered within the first three months following head trauma. We adopted an early rehabilitation programme for this group of patients, with the aim of promoting central nervous system compensation by habituation to the noxious effects of head rotation. The results of this study support our strategy of early enrolment of post-traumatic acute balance disorder patients into rehabilitation programmes, in accordance with previous studies.Reference Hoffer, Gottshall, Moore, Balough and Wester2, Reference Gottshall, Gray and Drake24, Reference Herdman, Clendaniel, Mattox, Holliday and Niparko25 Our results also support the effectiveness of rehabilitation for patients with central and peripheral vestibular disorders, regardless of differences in recovery time. This protocol differs from that of centres which recommend rehabilitation only for chronic peripheral vestibular lesions.
A minority (7 per cent) of our patients with a videonystagmographic diagnosis of peripheral vestibular lesion experienced a delay in recovery of more than three months, despite receiving rehabilitation. This group of patients had suffered severe head trauma. However, even in this group of patients, rehabilitation appeared to abolish their deconditioning caused by activity restriction and anxiety, thus improving their outcome.
In this study, 19 per cent of patients had subjective symptoms following head trauma, compared with 49 per cent who had objective symptoms. In the former patients, videonystagmographic results were in the normal reference range or were nondiagnostic. Dizziness in such patients has been termed ‘psychogenic’, and has been associated with both primary or secondary anxiety disorders (the latter precipitated by medical events caused by traumatic brain injury).Reference Hoffer, Gottshall, Moore, Balough and Wester2, Reference Staab and Ruckenstein26
The subjective symptoms of balance disorders following head trauma can represent a medico-legal problem, as the subjectivity of the complaints complicates assessment and make it difficult to determine whether symptoms are genuine or not. The possibility of malingering should be considered in such cases, especially if the probability of financial gain exists, or if the subjective complaints cannot be related to objective findings. The minor and mild head trauma patients in this study were found to meet one or both of these criteria. In the absence of objective videonystagmographic findings, cases were considered functional and managed as such.
However, the suggestion that post-traumatic subjective dizziness is solely of psychogenic origin has been challenged by reports of cases with evidence of a possible organic origin.Reference Davies and Luxon1
In the current study, we did not attempt to investigate the possibility of an underlying organic lesion in patients with only subjective symptoms and no videonystagmographic findings. This would have necessitated comprehensive evaluation of the multisensory integration of vision, somesthesia and vestibular system function, via the use of computed dynamic posturography, as has been previously suggested.Reference Goebel, Sataloff, Hanson, Nashner, Hirshout and Sokolow4, Reference Cevette, Puetz, Marion, Wertz and Muenter27, Reference Krempl and Dobie28
In the current study, patients with subjective symptoms recovered within the first two weeks of head trauma, without medical or physical therapy. Instead, this group of patients was managed with reassurance and psychiatric consultation. This may be taken as an indicator of the sensitivity of videonystagmography in detecting functional cases, and strongly supports the existence of a psychological mechanism for such balance disorder cases.
• Balance disorders are a frequent complication of head trauma; patients often report dizziness
• Head trauma severity affects the type and timing of balance disorder
• Acute post-trauma balance disorder cases should receive early rehabilitation
• Videonystagmography is a practical diagnostic and monitoring tool for dizzy post-trauma patients
• In this study, 19 per cent of post-trauma patients had only subjective dizziness, with no videonystagmographic findings
The results of the present study suggest that videonystagmography may be more clinically applicable and less time-consuming when employed during vestibular testing, with more reliable results, compared with conventional electronystagmography. Videonystagmography has the benefit of not requiring skin preparation and electrode application and wiring (with concerns about poor connections); this is especially advantageous when investigating head trauma patients. Adjustments are seldom required as videonystagmography does not depend on changes in corneo-retinal potential over time, as does electronystagmography. There is also the additional advantage of video-recording of eye movements; in the present study, this allowed us not only to conduct vestibular testing relatively quickly and easily, but also to repeat and compare results for each patient in an objective manner.
However, videonystagmography is unable to record the movements of closed eyes. Some of our patients closed their eyes as a reaction to feeling dizzy during vestibular testing. We managed this by reassuring the patient, together with frequent instruction to keep their eyes open during testing. At times, we had to delay completion of testing until the following day, especially in the first days following head trauma.
Conclusion
This study used videonystagmography to diagnose and monitor the treatment of balance disorders developing after head trauma of various grades of severity. It is currently agreed that simpler, more precise diagnostic techniques for post-traumatic balance disorders will facilitate better care and more favourable outcomes for patients. This study showed that videonystagmography is one such diagnostic technique.Reference Hoffer, Gottshall, Moore, Balough and Wester2 Furthermore, videonystagmography is cost-effective and easy to use.
The addition of computed dynamic posturography to the test protocol for patients with post-traumatic balance disorders would aid the detection of nonorganic balance disorders. However, further studies would be needed to justify the expense of such an addition, by demonstrating improved diagnostic capability and better patient outcomes.
