Visual Vestibular Mismatch

A poorly understood presentation of balance system disease

Download PDF version (3.4MB)

Chapter 5

Dizziness from whiplash and head injury: differences between whiplash and head injury

Mallinson, AI, Longridge NS.

Am J Otol 1998;20(4):814-818.


Objective: Large discrepancies exist in the literature regarding incidence and types of symptomatology in whiplash. This is because of the evolution of whiplash injury over the years with the advent of head rests and seat belts. Previous authors have regarded symptoms of dizziness as a result of brainstem or cerebellar injury or both. It has been difficult in those studies to ascribe a mechanism of injury, as patients with whiplash injury only have been grouped with those who have incurred mild traumatic brain injury as a result of a significant blow to the head. The authors saw the need to delineate patients who had suffered whiplash injury from those who also had suffered mild head injury, as defined in the rehabilitation-neurosurgical literature, to attempt to define differences in symptoms, abnormalities, and mechanisms of recovery in these two groups.

Study Design: The study design was a retrospective case review.

Setting: The study was conducted at a tertiary-quaternary referral clinic.

Patients: The records of 36 patients were reviewed. Nineteen of these patients suffered a whiplash-associated disorder and 17 suffered a mild head injury as well. These patients were referred for assessment of symptoms persisting for at least 2 years after their injury. Patients were excluded if they had not completed clinical assessment, including electronystagmography (ENG) and Computerized Dynamic Posturography (CDP).

Interventions: A full history, otolaryngologic examination, including assessment of eye movements, corneal reflexes and gait, as well as an investigation, including ENG and CDP, and history taking and detailed recording of related complaints immediately before diagnostic work-up were performed.

Main Outcome Measures: Symptoms reported by patients who had received either whiplash alone or whiplash plus mild head trauma as defined in the literature were measured. Patients were classified according to type of accident, type of injury suffered, and degree and nature of posturographic abnormalities.

Results: Patients often have similar complaints regardless of whether or not they had suffered a brain injury. Although CDP showed abnormalities in both groups, standard ENG assessment, including caloric testing, showed abnormalities only in the head-injured group. The posturographic abnormalities also were analyzed in both groups, and it was found that there was a correlation between the type of posturographic abnormality and the type of injury suffered. Although ENG testing is done routinely, posturography is shown to be more sensitive in picking up abnormalities. In addition, the authors have shown that posturography can delineate the type of injury suffered by exhibiting the compensation strategy used as well as the efficacy of that compensation strategy.

Conclusions: Because ENG abnormalities are limited to patients who have suffered a head injury, the inference is that these two groups of patients have suffered damage at different sites along the balance system pathways, but both of these lesions can lead to similar symptoms. Although the mechanisms of whiplash injury and how they affect the vestibular system are poorly understood, posturography testing is essential in inferring how a patient is recovering by measuring how and how well the patient is overcoming his or her deficit. This has important medical legal implications regarding legitimizing a patient’s problem, prognostic factors, as well as rehabilitation plans, measures, and outcomes.

KEYWORDS: posturography, compensation, whiplash, head injury, dizziness

Dizziness as a whiplash-associated disorder due to flexion-extension injury is common. Numerous articles have described this disorder. These go back to the classic article by MacNab (1), whose interpretation of the term “whiplash” differed from that of the current day. Patients in his study suffered gross movement of the neck to the point at which the head could touch the thoracic spine during the hyperextension process. In his study (published in 1964), symptoms resulting from a lateral collision appeared to be milder than most of those occurring after a rear-end collision. Modern-day whiplash results from hyperextension of the neck, which is arrested by the impact of the head on the head restraint, followed by a flexion process, which is altered by the seat belt that serves to restrain the body. As a result, lateral accidents, in which the head movement is not damped by a restraint, may now inflict more severe injury than rear-ending collisions.

It is not surprising that over the years, there have been large discrepancies in the incidence and types of symptomatology related to whiplash injury. Hinoki (2) describes the T-bone accident, in addition to the rear-ending accident, and classified symptoms suffered in both types of accidents as having the same pathologic origin. However, it is unclear whether his patients had head rests. He reported that 85% of the patients suffered “dizzy” symptoms, without defining his use of this term. No mention is made of direct head trauma in this study. The reader is left to assume that none of the patients suffered a head blow in the accident other than from the head rest. Oosterveld et al. (3) describe 85% of the patients with whiplash as reporting “some type of vertigo.” These symptoms included floating feelings and lightheadedness, but they did not ascribe a site to the lesion causing these symptoms. Oosterveld et al. inferred pathology to the cervical muscular system in 64% of the patients studied and “proved the presence of both brainstem and cerebellar pathology” in 43% of the patients. Sturzenegger et al. (4) indicated that a short latency before onset of symptoms was prognostically bad and regarded vertigo and imbalance as caused by brainstem injury. Although Oosterveld offered pursuit abnormalities as evidence of both brainstem and cerebellar pathology, Sturzenegger gave no clear evidence as to how he reached his conclusion. Sturzenegger also noted that head restraints might reduce neck symptomatology somewhat but not dramatically and also commented that the speed of the cars at the time of the accident appeared to bear little relevance to the patient’s symptoms.

Probably another reason for the discrepancies in symptomatology seen in the literature is that patients with whiplash and those who incur a significant blow to the head are lumped together. Rubin et al. (5) studied a group of patients who had suffered a closed head injury or a whiplash injury. They did not delineate between these two seemingly differing traumas. No study that we have found has particularly delineated patients suffering from whiplash only and those suffering head injury at the same time.

As discussed in our companion article, both groups of patients often report similar symptoms, and the characteristic symptoms they describe are attributable to disruption of one or more of the vestibular reflexes. However, despite the striking similarities in their stories, it still follows that they have suffered injuries to different structures and, in addition, it seems that they might use different mechanisms of compensation or recovery from their injuries.

Compensation for balance system damage has been investigated extensively both from the point of view of the physiologic mechanisms involved as well as from the point of view of the practical aspects of vestibular rehabilitation. A basic tenet of compensation developed by Bronstein (6) is that there is a dominant role of vision in the maintenance of posture. In the absence of the vestibular “template” proposed by Nashner and McCollum (7) (against which all other sensory information is compared and either used or disregarded), there is a dogmatic dependence on visual information to keep balance. This “visual preference strategy” can be delineated with Computerized Dynamic Posturography (CDP). Unfortunately, in everyday life, there are some patients in whom this preference has its drawbacks. Some patients have an intolerance for any discongruency between visual and vestibular signals. This “visual-vestibular mismatch” (VVM) was originally described by Paige (8), and Mallinson et al. (9) regard the development of VVM as indicative of vestibular pathology. Patients who have VVM have symptoms develop that are annoying in everyday life. Compensation for vestibular injury classically entails enlisting visual information, which usually is given hierarchical preference over proprioceptive input unless the visual information is deficient or unreliable or is seen to have been compromised in some way (6). This study set out to determine whether patients with whiplash-associated disorder only had findings different from those who had mild head injury.


A retrospective study of our 36 patients with dizziness after a motor vehicle accident (MVA) was undertaken. This included analysis of CDP and standard electronystagmography (ENG) assessment. These patients were long-term sufferers (>2 years) of complaints characteristic of a disorder of the balance system. During patient assessment, it was determined whether, in addition to the whiplash injury, the patient had incurred a blow to the head apart from impact with the head rest. Patients were excluded if their head injury was too severe to meet the definition of mild traumatic brain injury (10,11).


Dizziness developed in one patient 2 years after his MVA. It came on after temporomandibular joint surgery for jaw pain induced by his whiplash injury. He was excluded from the study.

Our study divided accidents into six types (Table 1). The type of accident (rear end, T-bone, or head-on) made little difference to the patient’s presenting complaints, but the presence or absence of mild head injury (criteria for mild head injury are outlined in our companion article) was discovered to be important with respect to abnormalities found on standard investigation. As a result, patients were separated into “head injured” and “whiplash only” groups. Twenty-seven patients in all had abnormal CDP patterns. The CDP abnormality patterns in the two groups were examined (Table 2).

Of 21 patients who did not suffer a head injury in their MVA, 19 suffered whiplash only (Type A). In none of the 19 was a standard, recognized ENG abnormality detected, and none had a caloric reduction. However, CDP testing showed abnormalities in 15 (79%) of 19 patients.

TABLE 1. Legend of motor vehicle accident types

  1. Rear end accident with whiplash and no head injury (19)
  2. Rear-end accident with whiplash and head injury (5)
  3. T-Bone or sideswipe accident with whiplash and no head injury (-)
  4. T-bone or sideswipe accident with whiplash and head injury (7)
  5. Head-on accident with whiplash and no head injury (2)
  6. Head-on accident with whiplash and head injury (3)

In 10 of the 15 patients, the abnormality was a standard, recognized vestibular pattern, whereas in 5 patients, there were nonspecific abnormalities (12).

Twelve patients had a T-bone or head-on MVA, and only 2 of these patients did not suffer a head injury. Both of these had normal ENG and CDP results. Fifteen patients studied suffered a mild head injury. Five (33%) of the 15 patients had a caloric abnormality and 3 (20%) had central ENG findings. Twelve (80%) of the 15 patients had CDP abnormalities. Nine of these abnormalities were characteristically vestibular. To study the nature of the CDP abnormalities, these abnormalities were broken down into four groups, all of which are delineated by Nashner in the interpretation manual (12):

  • Group 1: Classic vestibular abnormalities (sensory organization test 5 or 6 down pattern or both). Ten of the patients fell into this group, 6 of whom were in the head-injured group. This CDP pattern suggests the presence of some type of vestibular deficiency.
  • Group 2: Somatosensory-preferenced abnormalities (sensory organization tests 4, 5, and 6 down patterns). Six patients fell into this group, all of whom had whiplash only. These patients exhibit a strong somatosensory dependence and cannot make effective use of either vestibular or visual inputs in the absence of a stable surface (12).
  • Group 3: Visually preferenced abnormalities (sensory organization tests 3, 5, and 6 down patterns). Three patients fell into this group, all of whom had had head injuries. These patients are destabilized by orientationally inaccurate visual stimuli.
  • Group 4: Nonspecific abnormalities (none of the above patterns but abnormal CDP performance). Eight of our patients fell into this group, three of whom had had head injuries.

Table 2. Computerized dynamic posturography (CDP) abnormality patterns

Patient Age/Sex CDP Abnormality Abnormality Group
1 54 M Condition 6 down 1
2 36 M Condition 6 down 1
3 18 F Normal n/a
4 28 M Condition 3 and 6 down 3
5 25 F Normal n/a
6 37 M Condition 3 and 6 down 3
7 15 F Condition 5 and 6 down 1
8 34 F Normal n/a
9 58 M Normal n/a
10 19 F Condition 3, 5 and 6 down 3
11 53 M Condition 5 and 6 down 1
12 50 F Condition 5 and 6 down 1
13 33 F Condition 1, 2 and 4 down 4
14 42 F Condition 5 and 6 down 1
15 42 F Normal n/a
16 19 F Condition 1-6 down 4
17 40 M Condition 3 down 4
18 38 F Condition 5 and 6 down 1
19 54 M Condition 1-6 down 4
20 33 M Condition 2, 3 and 5 down 4
21 18 F Normal n/a
22 42 F Condition 4-6 down 2
23 60 M Condition 4-6 down 2
24 28 F Condition 6 down 1
25 46 M Normal n/a
26 29 F Condition 4-6 down 2
27 28 F Condition 4-6 down 2
28 45 F Condition 5 and 6 down 1
29 48 M Condition 5 and 6 down 1
30 39 F Condition 5 and 6 down 1
31 30 F Condition 2-6 down 4
32 53 F Normal n/a
33 49 M Condition 4-6 down 2
34 40 M Condition 2 down 4
35 33 F Normal n/a
36 37 M Condition 3 and 5 down 4


Although closed head injury can occur as a result of whiplash (13), there is a marked difference in the threshold of force required to injure the brain and to injure the neck. The lowest threshold to cause brain trauma in animals is recognized to be approximately 60 g of force (13) (and the threshold for humans is suspected to be at least 70-80 g) (14), but the limitation of tolerance to whiplash is a gravitational acceleration of approximately 14 g (15). That is, a whiplash-type motion resulting in injury causes the head to be exposed to accelerating forces far below those required to be injurious to the brain. The whiplash motion in itself does not necessarily cause traumatic brain injury, and it is extremely important in any study to delineate patients who have suffered whiplash only from those who also have incurred a blow to the head.

In patients suffering a rear-end accident in a modern motor vehicle, head injury apart from minor head rest contact is unusual. In our group of nonhead-injured patients, no caloric reduction and no central abnormality on ENG were detected. However, 79% of these patients showed a CDP abnormality. Computerized dynamic posturography is the most effective investigation in patients who have not suffered head injury, although it is ENG that routinely is done.

The ENG abnormalities are limited to patients who have suffered a head injury, suggesting the presence of both peripheral as well as central disturbance. The inference is that despite their similar symptoms, patients who have suffered pure whiplash versus those with a mild head injury may have a different etiology and mechanism of damage.

The small number of patients in accident types B through F limits statistically significant conclusions. However, it generally is accepted that CDP abnormalities with reductions on conditions 5 and 6 indicate a disturbance in the vestibular system (12).

In patients who suffered a head injury, a significant number had abnormalities detectable on ENG. However, the CDP results were abnormal in approximately the same percentage of patients as in those who did not suffer head injury. It was not possible to differentiate between the head-injured and the nonhead- injured groups using only the criterion of normal or abnormal CDP results.

Both of these groups have to compensate for their damage. It is expected in the presence of normal visual input and intact cerebellar function that our patients would have a visual preference develop, the “regular” mechanism of compensation as outlined by Bronstein (6). However, we have identified a group of patients suffering whiplash only who show a preference for somatosensory input. We postulate that these patients represent a separate group with a different lesion site and that both groups require a specifically tailored mechanism of compensation.

It is difficult to compare our findings with those of other investigators because they have not grouped the patients as we have. There are other reports that support our findings showing that patients suffering from head injury and whiplash use a visually preferenced method of compensation (5). Chester (16) also found different groups of CDP abnormalities in his 43 patients. He did not delineate head-injured from nonhead-injured patients. We believe that the differing CDP findings in our groups of patients can be broken down with respect to site of lesion and mechanism of injury and that our head-injured and nonhead-injured patients represent two distinct groups of patients having “picked” different methods of compensation for their injuries.

The “vision preference” pattern of abnormality occurs in patients with balance disorders secondary to traumatic head injury. Patients with a vision preference are more likely to show either normal ENG results or subtle losses of peripheral vestibular function (12). This supports the premise that the visual system is the primary choice of compensation for subtle vestibular losses. Clinical evidence for this premise is advanced by the many patients with vestibular disorders or caloric reductions or both who show a newly developed failure to integrate visual and vestibular function as a result of their newly developed dogmatic reliance on visual information for balance maintenance. This results in development of post-traumatic motion sickness or VVM or both. This is confirmed by Bronstein (17). He has shown that the efficiency of the visuopostural loop increases when there is a conflict between the visual and proprioceptive cues. Perhaps evidence is provided to support this by two of our patients who reported a worsening of symptoms after being prescribed standard visual rehabilitation exercises.

The vestibular system, as outlined by Nashner, serves as a template against which other conflicting information is suppressed. Thus, a person with normal vestibular function can ignore the ordinarily powerful visual-optokinetic input in everyday life, because if unable to do so, he or she would be destabilized by any conflict such as watching a bus pull away from the curb.

Our group of head-injured patients had developed a strategy of compensation with visual preference abnormality patterns showing on CDP, suggesting that they had compensated in a fashion in keeping with Nashner’s explanation. In these patients, visual input dominates in the maintenance of balance.

We postulate that the concept of “compensation” (i.e., recovery from permanent vestibular damage) involves development of the ability to suppress actively an inappropriate response. In patients who initially develop visually preferenced CDP abnormalities, eventual compensation-conditioning allows them to suppress appropriately visual stimuli that are inappropriate, so that they eventually “learn” (i.e., condition to) which visual stimuli are appropriate to balance maintenance and which visual stimuli are inappropriate. The compensation process in this situation is not the development of new pathways but the development of appropriate suppression mechanisms.

All six patients who showed a somatosensory-referenced abnormality were whiplash-only patients. They had had rear-end MVAs, which are the most likely to cause neck injury at very low velocities because of the design of seats, head rests, and seat belts. The central nervous system may achieve head stability in different ways, depending on the nature of the movement task. Both vestibulocollic and cervicocollic reflexes may contribute to stabilization of the head, depending on the goal of the subject performing the task (18). Patients with cervical root compression also manifest impaired CDP performance before surgical treatment of their spinal cord compression, which again returns to normal postsurgically, along with diminished complaints of balance difficulties (19). These findings were confirmed by Persson et al. (20) in 1996. They suggest that such surgery may reduce cervical muscle tension and subsequent “normalization” of the proprioceptive signals from the neck and a reduction of the sensory mismatch when proprioceptive signals converge with vestibular information in the central nervous system.

We believe that this belief supports evidence advanced by Persson et al. (20) that the CDP abnormalities are the result of disruption of tonic neck input into the linear vestibulo-ocular reflexes and are not caused by vestibular system damage. We postulate that in the same way that the vestibular system acts as a template to suppress inappropriate visual information (which takes over in the presence of vestibular system damage), the tonic neck reflexes serve to suppress inappropriate proprioceptive information. Loss or damage to these reflexes causes a loss of this suppression inability, with resultant “taking over” by the proprioceptive information. In the same way that the vestibular system works hand in hand with vision to suppress inappropriate visual information, the vestibulocollic and cervicocollic systems work hand in hand with the proprioceptive system, suppressing inappropriate somatosensory information. The resulting somatosensory-preferenced CDP abnormality arises as a result of loss of this suppression and not as a result of vestibular damage. This explains the observation in the Equitest manual that “this pattern (i.e., somatosensory dependence) is not commonly observed in patients with pathology limited to the vestibular system” (12).

We have speculated that compensation for vestibular damage involves a relearning of the suppression techniques for inappropriate information. We also hypothesize that in the same way, somatosensory-dependent individuals can do the same. Once patients are well on the way to compensating, they are better trained at suppressing the inappropriate information, although they still do not perform as well in the absence of that information. This gives rise to the “nonspecific” patterns of CDP abnormalities. Patients with these CDP abnormalities arise from both groups of patients (whiplash-only and head injured), and we propose that they represent patients who, for want of a better term, have “compensated” as far as they can for their difficulties. This does not mean to infer that these patients are symptomatic or even that they are functioning well. Of 10 patients who had 5 or 6 down patterns or both, 6 had head injuries and 4 had no head injuries. The 5 or 6 down patterns or both can be exhibited by a compensated patient from either group. The same holds true in our group with nonspecific CDP abnormalities. Three of the eight had had head injuries and five of the eight had had whiplash only.

We have identified a group of patients on CDP investigation suffering from vestibular symptoms who have a somatosensory rather than a visual preference. The group of patients we have identified also seems to be delineated by the type of injury suffered.


We have long wondered what influences the central nervous system when it “picks a strategy of compensation” as shown on CDP assessment in recovering patients. Perhaps it is not a matter of picking a strategy to compensate for a common lesion, but the CDP abnormality reflects the pathway that has been least injured and also shows the stage of recovery that the patient has attained at the time of assessment.

Continued follow-up may show strategy changes if the compensation process is evolving or lack of change if it has been completed. Ongoing accumulation of more patients should allow us to confirm or refute the postulate that we have delineated a group using a somatosensory rather than a visual template in the compensation process.


  1. MacNab I. Acceleration injuries of the cervical spine. J Bone Joint Surg 1964;46A:1797-9.
  2. Hinoki M. Vertigo due to whiplash injury: a neurotological approach. Acta Otolaryngol (Stockh) 1985;419 (Suppl):9-29.
  3. Oosterveld WJ, Kortschot HW, Kingma GG, et al. Electronystagmographic findings following cervical whiplash injuries. Acta Otolaryngol (Stockh) 1991;111:201-5.
  4. Sturzenegger M, DiStefano G, Radanov BP, et al. Presenting symptoms and signs after whiplash injury: the influence of accident mechanisms. Neurology 1994;44:688-93.
  5. Rubin AM, Woolley SM, Dailey VM, et al. Postural stability following mild head or whiplash injuries. Am J Otol 1995;16:216-21.
  6. Bronstein AM. Suppression of visually evoked postural responses. Exp Brain Res 1986;63:655-8.
  7. Nashner LM, McCollum G. The organization of human postural movements: a formal basis and experimental synthesis. Behav Brain Sci 1985;8:135-72.
  8. Paige GD. Senescence of human visual-vestibular interactions. J Vestib Res 1992;2:133-51.
  9. Mallinson AI, Longridge NS, Peacock C. Dizziness, imbalance and whiplash associated disorder. J Musculoskeletal Pain 1996;4:105-12.
  10. Jane JA. Mild to moderate head injury. Definitions. In: Hoff JT, Anderson TE, Cole TM, eds. Mild to moderate head injury. Contemporary issues in neurological surgery. Boston: Blackwell Scientific Publications. 1989:1-8.
  11. Kay T, Harrington DE, Adams R, et al. Definition of mild traumatic brain injury. J Head Trauma Rehabil 1993;8:86-8.
  12. Neurocom International Inc. EquiTest Interpretation Manual. Clackamas, OR: Neurocom International, 1992.
  13. Ommaya AK. Biomechanics of head injury. In: Nahum AM, Melvin J, eds. The biomechanics of trauma. Norwalk: Appelton Century Croft, 1985:245-69.
  14. Reid SE. Radio-telemetry studies of head impacts on the football field. In: Head and neck injuries in sports. C. Thomas, 1984:24-79.
  15. Mertz HJ, Patrick LM. Investigation of the kinematics and kinetics of whiplash. SAE Transactions 1967;76:Paper #670919:2952-80.
  16. Chester JB. Whiplash, postural control and the inner ear. Spine 1991;16:716-20.
  17. Bronstein AM. Visual control of balance in cerebellar and Parkinsons syndromes. Brain 1990;113:767-79.
  18. Shupert CL, Horak FB. Effects of vestibular loss on head stabilization in response to head and body perturbations. J Vestib Res 1996;6:423-37.
  19. Vitte E, Lazennec JY, Pharaboz C, et al. Induced perturbations of Equitest in cervical spine pathologies. In: Woollacott M, Horak F, eds. Posture and gait: control mechanisms (Vol 2). Eugene, OR: University of Oregon Books, 1992:176-9.
  20. Persson L, Karlberg M, Magnusson M. Effects of different treatment on postural performance in patients with cervical root compression. J Vestib Res 1996;6:439-53.