Physical Therapy for Migraine-Related Vestibulopathy and Vestibular Dysfunction With History of Migraine†
Presented at the Association for Research in Otolaryngology Meeting, St. Petersburg, Florida, February 21, 2000.
Objectives/Hypothesis To assess the efficacy of physical therapy for patients with a diagnosis of migraine-related vestibulopathy (MRV) or vestibular dysfunction with a history of migraine headache.
Study Design Retrospective case series
Methods Thirty-nine patients were identified through a retrospective chart review, 14 with a diagnosis of MRV and 25 with migraine headache. The patients were treated with a custom-designed physical therapy e-ercise program for a mean of 4.9 visits over a mean duration of 4 months. Patients completed the Dizziness Handicap Inventory (DHI), the Activities-Specific Balance Confidence Scale (ABC), and the Dynamic Gait Inde- (DGI), reported the number of falls they had e-perienced in the past 4 weeks, and rated the severity of their dizziness on an analogue scale of 0 to 100 at initial evaluation and at discharge.
Results Significant differences were seen before and after therapy in each of the outcome measures used. The average decrease in DHI score was 12 points (P < .01). ABC scores increased an average of 14 points (P < .01). Subjects increased their DGI scores an average of 4 points (P < .01). The number of patients reporting more than one fall decreased by 78% at discharge. (P < .05). Baseline symptoms of dizziness decreased an average of 11 points (P < .05).
Conclusions Patients with MRV and migraine headache demonstrated improvement in physical performance measures and self-perceived abilities after vestibular physical therapy.
Migraine headache is a relatively common neurologic disorder that affects 6% of men and 18% of women in the United States. 1 Neuro-otological symptoms have been reported by 33% to 72% of patients with migraines. 2,3 Migraine associated with vertigo has been reported to occur in 32% of the cases in a retrospective sample of 363 consecutive patients who presented to an otology practice. 4 Often the vestibular symptoms occur during the headache, 5 although there is evidence to support that the vestibular complaints can also occur before or after the actual migraine. 3,6–8 Depending on the type of migraine, the patient may complain of dizziness or vertigo. 6,9,10 If the migraine is very severe, along with vertigo, the individual may experience nausea, vomiting, 7 motion sensitivity, 9,11–13 and postural instability. 3,13
Persons who experience migraine-related vestibulopathies often have abnormal vestibular laboratory results. 3,4,9,13,14 Eighty percent of individuals with a diagnosis of common migraine, in a study by Toglia et al., 15 demonstrated abnormalities on vestibular testing. Kayan and Hood 3 reported that 77% of patients with a history of migraine and neuro-otological complaints demonstrated abnormalities in vestibular testing.
Migraine-related vestibulopathy has been identified as a disorder that is diagnosed based on a history of migraine or family history of migraine, complaints of space and motion sensitivity, and vestibular complaints that do not fit any other vestibular conditions. 9 Although the diagnosis of MRV is often a diagnosis of exclusion after other vestibular and central nervous system diseases have been ruled out, it is still controversial because of the various definitions used to describe the condition.
Use of medication and control of dietary triggers is often helpful in the control of MRV. 9,16 Vestibular rehabilitation has been suggested as an intervention for persons with MRV. 9,17 Seven of 10 subjects reported benefit from vestibular rehabilitation. 9 Johnson 17 reports that 92% (22 of 24) of patients who experienced motion sensitivity or postural instability reported subjective improvements when treated with a combination of vestibular rehabilitation and pharmacological therapy. Johnson 17 suggested that the medication allowed the subjects to be more tolerant of the motion induced with the vestibular rehabilitation program. No data were available on whether these patients improved their functional capabilities after vestibular rehabilitation intervention.
The purpose of this retrospective chart review was to determine the efficacy of physical therapy for patients with a diagnosis of MRV and migraine headache. There is no evidence in the literature that persons with MRV and migraine headache improve functionally with physical therapy intervention.
MATERIALS AND METHODS
Patients were identified through a retrospective chart review of patients seen for vestibular physical therapy between January 1997 and August 1999 at the Jordan Balance Center at the University of Pittsburgh Medical Center. Patients were included if they were given a diagnosis of MRV by their referring physician or had a history of migraine headache by patient report. Sixty-six patients met the inclusion criteria. Twenty-seven of the patients were excluded from the study because of either inadequate follow-up (n = 20) or only a remote history of migraine (n = 7). Of the remaining 39 patients (Table I), 14 had a diagnosis of MRV and 25 had a history of migraine headache. The mean age of the patients was 54 ± 16 years (range, 20–89 y) with 34 women and five men. Patients described an average duration of symptoms of dizziness and/or imbalance of 20.9 months (SD ± 26.7 mo). The median duration of symptoms was 7 months. This study met the criteria for an exempt review by the University of Pittsburgh Institutional Review Board.
Table Table 1.. Patient Characteristics (Mean ± SD).
*Scores missing because all outcome measures were not completed at every visit.
MRV = migraine-related vestibulopathy; MH = history of migraine headache.
Patients were referred from 12 physicians with either a diagnosis of MRV or a diagnosis of peripheral or central vestibular dysfunction and a history of migraine headache. Examples of vestibular diagnoses included benign paroxysmal positional vertigo (BPPV), dysequilibrium, labyrinthine concussion, vestibular neuritis, and recurrent labyrinthitis. All patients had undergone neurological or otological examination before referral. Ninety-five percent of the patients underwent vestibular function testing as part of the diagnostic process.
All patients received vestibular physical therapy provided by one of five physical therapists with specialized training in the assessment and treatment of vestibular disorders who worked at the Jordan Balance Center. A customized treatment plan was developed for each patient according to the results of the physical therapy evaluation and included the following components, as indicated: general strengthening and stretching exercises, habituation exercises, exercises to promote vestibular compensation, balance and gait training, or exercise to enhance the use of specific sensory inputs for balance control.
Patients were asked to rate their symptoms on a scale of zero (no symptoms) to 100 (the most severe symptoms that the patient could imagine) during each visit to physical therapy. In addition, they were asked to complete the Dizziness Handicap Inventory (DHI), which is a scale used to rate a patient's self-perception of handicap from their dizziness. 18 Higher scores on the DHI indicate greater handicap. Emotional, physical, and functional aspects of dizziness are quantified in three subscales of the DHI. 18 The DHI has been shown to be reliable and valid for use with patients with vestibular disorders. 18
Patients completed the Activities-Specific Balance Confidence Scale (ABC) during each visit to the clinic to assess balance confidence. 19 The ABC is a 16-item questionnaire and patients rate their confidence performing 16 activities of daily living. The ABC has been used previously with persons with vestibular disorders and was shown to be moderately correlated with the DHI in individuals with complaints of dizziness. 20 In addition, all patients were asked if they had fallen in the past 4 weeks at the onset of physical therapy and if they had fallen between return visits.
During patients' physical therapy evaluation and each subsequent re-evaluation, the Dynamic Gait Index (DGI) was calculated based on the their ability to perform eight gait tasks 21 including walking, walking with head turns, pivoting, walking over and around objects, and going up and down steps. The maximum point value on the DGI is 24. Scores of 19 or less indicate increased risk of falling in older adults. 21
To allow for determination of clinically significant improvement, a composite score was calculated using the following equation incorporating the ABC, DHI, and DGI:MATH
The composite score was developed to provide the clinician with an overall score of outcome. It combined three measures—two subjective and one that was a gait rating by the physical therapist. The maximum point value on the composite score was 296. From our clinical experience we defined composite scores greater than 270 as no impairment, scores between 240 and 270 as minimal impairment, scores between 120 and 240 as moderate impairment, and scores below 120 as severe impairment.
Using the criteria in Table II, we retrospectively assigned each patient a disability score 22 based on subjective information in the discharge summary. The correlation between the disability score and the composite score was used to validate the composite score.
Table Table 2.. Shepard and Telian's Disability Score Criteria. 22
For persons who were retired or students, work refers to usual household chores and academic activities, respectively.
From Shepard et al., 22 with permission.
Scores before and after therapy on the DHI, ABC, reported number of falls, dizziness symptoms severity rating, composite score, and the DGI were compared using the Wilcoxon matched-pair signed rank test, because all of our outcome measures were ordinal. Alpha was set at 0.05. Differences in raw scores were calculated to determine whether changes in scores were within clinically relevant ranges. The Mann-Whitney U test was used to determine if there were significant differences between the MRV and migraine headache subgroups in demographic or outcome data. 23 The Spearman correlation coefficient was used to compare the composite score with the disability score and symptom score. 23 The influence of medication in both subgroups and the total group were analyzed using the Mann-Whitney U test.
Descriptive patient information concerning age, sex, diagnosis, duration of symptoms, number and duration of treatments, pretreatment and posttreatment scores, and overall change in composite score is included in Table I. All patients had undergone neurological or otological examination and portions of the vestibular function testing before referral for vestibular physical therapy (Table III). Eighty-one percent of the patients presented with abnormal results in at least one of the vestibular function tests. The frequency of abnormal responses was similar to previously reported results. Toglia et al. 15 reported abnormal vestibular findings in 80% of patients with a history of migraine, whereas Cass et al. 9 reported 73% of their patients with a diagnosis of MRV had abnormal results in at least one vestibular function test. Abnormal caloric responses were demonstrated by 55% of the patients, rotational vestibular test results were abnormal in 42% of the patients, oculomotor test results were abnormal in 29% of the patients, and positional test results were abnormal in 19% of the patients. None of the patients with abnormal results on positional testing tested positive for BPPV. However, two patients tested positive for BPPV during their last physical therapy session and underwent a particle-repositioning maneuver. Overall, 59% of the patients described symptoms of space and motion discomfort. Sixty-four percent of the patients with MRV and 56% of the patients with migraine headache complained of symptoms consistent with space and motion disorder. Although not statistically significant, there was a trend toward significant differences in vestibular function testing between the MRV and the migraine headache groups. A greater proportion of the MRV group (92%) demonstrated abnormal responses on ocular motor, rotational, caloric, and positional testing than in the MH group (75%). These results are similar to those of Kayan and Hood, 3 who found that 70% of their patients who presented with neuro-otological symptoms temporally associated with migraine demonstrated abnormal results on vestibular testing.
Table Table 3.. Results (in Percentages) of Vestibular Laboratory Testing.
*Sensory Organization Testing only.
†40% reduced vestibular response; 12% directional preponderance.
Patients were treated for a mean of 4.9 visits (range, 2–21 visits; SD ± 3.44) over a mean duration of 4 months (range, 1–17 mo; SD ±3.44) with a customized physical treatment program. Fourteen patients received a diagnosis of MRV and 25 patients had migraine headache. The mean age of the patients with a diagnosis of MRV was 55.2 years (SD ± 10.1 y). The mean age of the patients with migraine headache was 54.2 years (SD ± 18.7 y). No statistically significant difference was seen between the patients with MRV and the patients with migraine headache on age, sex, symptom duration, treatment duration, number of physical therapy visits, vestibular function tests, or outcome measures.
The Wilcoxon matched pair signed rank test identified significant differences before and after therapy in each of the outcome measures recorded for the entire group. Mean outcome measure values at initial evaluation and discharge, and the mean amount of change from pretreatment to posttreatment values are identified in Table IV.
Table Table 4.. Total Group Mean Outcome Measure Values at Initial Evaluation and Discharge, and the Mean Amount of Change for the Dizziness Handicap Inventory (DHI), Activities-Specific Balance Confidence Scale (ABC), Perception of Dizziness Symptoms (PDS), the Dynamic Gait Index (DGI), and the Composite Score (CS).
Twenty-eight of the 39 patients demonstrated an increase (improvement) in the composite score. Scores increased from 3 to 156 points. Four patients demonstrated a decrease in their composite score. Scores decreased from 4 to 24 points. The average change in composite score from before therapy to after therapy was 41.9 (SD ± 44.7). The correlation between the composite score and the disability score using the Spearman correlation coefficient was r = −0.74 and between the composite score and the symptom score was r = −0.73.
Of the five subjects who were only mildly impaired before treatment (composite score 240–270), one was slightly worse and two improved only slightly (from 264 to 277 and 266.8 to 272). However, in the larger number of patients who were moderately or severely impaired, three were slightly worse and four had score changes of less than 5 composite score units, but 20 had sizeable composite score improvements. Seventy-four percent of the patients who were classified as moderately or severely impaired based on their pretherapy composite score improved.
Twenty-eight patients had a decrease in their self-perceived handicap from dizziness through their DHI score, nine patients had an increase, one patient's score was unchanged, and one patient failed to complete the DHI at discharge. The average decrease in DHI was 12 points (P < .01); 30% of the patients had a decrease equal to or greater than the previously defined clinically significant level of 18 points. 18 In contrast, only one patient had a score increase of greater than 18 points. Significant differences were also demonstrated in the emotional (P < .01), physical (P < .01), and functional (P < .01) subscales of the DHI.
Twenty-four patients indicated an improved perception of balance abilities with increased ABC scores. Eleven patients reported lower scores on the ABC, suggesting that they felt their balance abilities had worsened. Two of the patients did not complete the assessment at discharge and two had no change in ABC scores (one patient had a score of 100 at initial evaluation). ABC scores increased an average of 14 points (P < .01).
Thirty patients did not report any falls in the 4 weeks before the initial evaluation. Eight patients reported falling in the 4 weeks before the start of physical therapy, yet at discharge only three of these patients reported any falls in the previous 4 weeks. Six of these initial eight patients reported a decrease in their number of falls. Two patients who did not report falls during the initial evaluation reported an increase in the frequency of falls at discharge.
At the initial evaluation five of the patients reported more than one fall during the prior 4 weeks. At the last physical therapy visit, only two of the patients reported more than one fall (P < .05)
Eleven patients rated their baseline severity of symptoms as zero at the initial evaluation, and their symptoms did not increase at discharge. Of the remaining 28 patients, 15 rated their symptoms lower at discharge than at the initial evaluation, six rated their symptoms higher, and 1 was unchanged. Baseline severity of dizziness symptoms decreased an average of 11 points (P < .05).
Functional balance abilities increased in 25 of the patients as evidenced by increased DGI scores. Only one subject demonstrated a decrease in DGI score, six subjects did not complete the testing at discharge, and seven subjects' scores were unchanged. Patients increased their DGI scores by an average of four points (P < .01).
The entire group was separated for additional analysis by diagnosis of either MRV or migraine headache. Mean outcome measure values at initial evaluation, discharge, and the mean amount of change from pretreatment to posttreatment values for patients with a diagnosis of MRV are identified in Table V. In patients with a diagnosis of MRV, significant differences were demonstrated before and after therapy in DHI (P < .01), DGI (P < .01), and composite score (P < .01). The difference in ABC scores approached statistical significance and the difference in subjective rating of symptoms was not significant. Although the DHI as a whole demonstrated statistically significant changes, when considered individually, the physical and functional subscales demonstrated significant changes, but the emotional subscale did not.
Table Table 5.. Migraine-Related Vestibulopathy Group Mean Outcome Measure Values at Initial Evaluation and Discharge, and the Mean Amount of Change for the DHI, ABC, PDS, DGI, and CS.
Mean outcome measure values at initial evaluation, discharge, and the mean amount of change from pretreatment to and posttreatment values for patients with migraine headache are identified in Table VI. Patients with migraine headache demonstrated significant differences in pretreatment and posttreatment scores in all outcome measures. Statistically significant changes in pretreatment and posttreatment scores were demonstrated in all three DHI subscales and in the ABC scale.
Table Table 6.. History of Migraine Group Mean Outcome Measure Values at Initial Evaluation and Discharge, and the Mean Amount of Change for the DHI, ABC, PDS, DGI, and CS.
Twenty-two of the 39 patients were taking medications that might affect the severity and frequency of migraines and 17 patients were taking no antimigraine medications. The Mann-Whitney U test demonstrated no difference in demographic data between the two groups at the beginning of physical therapy intervention. Statistically significant differences were noted between the two groups in the initial composite score (P < .05), discharge composite score (P < .01), initial ABC score (P < .05), discharge DHI (P < .05), and discharge DGI (P < .05). The group taking medication demonstrated higher composite scores at both initial evaluation and discharge than did the group not taking medication. Subjects in the group taking medication demonstrated lower DHI scores and higher DGI scores at discharge (indicating less impairment) than did the nonmedicated group. The amount of change in the outcome measures before and after therapy was not statistically different between the two groups.
Of the patients with a diagnosis of MRV, seven were receiving medication and seven had not received medication. The MRV group that received medication demonstrated differences that approached statistical significance in discharge composite score (P = .06), DGI (P = .06), and DHI (P = .09) from the group not taking medication. In the migraine headache group, 15 were receiving medication and 10 had not received medication. No significant difference or trends in outcome measures were observed between the patients receiving or not receiving medication in the group with a history of migraine.
After physical therapy total DHI scores of both patients with MRV and patients with migraine improved. It is interesting that statistically significant improvements were demonstrated on the physical and functional subscales in the entire sample, yet the MRV did not improve significantly on the emotional subscale. That is, although the patients in the MRV group improved their physical performance during gait, they still felt emotionally handicapped. The MRV group may have had more severe or more frequent headaches that affected their scores on the emotional subscale of the DHI. If the differences on the emotional subscale between the two groups represent differences in how patients approach their disabilities, then perhaps the interventions should be tailored to address these differences.
The ABC indicates how confident the patients are about their balance. Patients' ABC scores improved from initial evaluation to discharge. When separated into diagnostic groups however, the MRV group did not improve as much as the migraine headache group on the ABC. This may be related to how the two groups perceive their dysfunction. The ABC is a tool that reflects fear of falling; fear and anxiety are key elements of the emotional component of the DHI.
The DGI scores for the total group improved from a mean of 17 to a mean of 20. Shumway-Cook et al. 21 stated that scores of 19 or less indicate high risk for falling in institutionalized older adults. The improvement of the DGI scores of the total group indicates a decrease in the risk of falling. It has been shown that both the risk of falling and the fear of falling lead to decreased activity, decreased functional status, and greater morbidity. 24 This phenomenon appears to be cyclic: falling and a loss of balance increase fear of falling, which leads to inactivity, which contributes to balance impairments. Vestibular physical therapy breaks this cycle by improving balance abilities, decreasing risk of falling, and increasing confidence, which increases activity, thereby promoting improved balance abilities.
Because many of these patients were taking medications, we attempted to determine if patients who were taking antimigraine medications improved more with physical therapy intervention. The patients who received antimigraine medications compared with the nonmedicated patients improved significantly more on the composite score, DHI, and DGI. The medicated group improved an average of 49 points and the nonmedicated group an average of 33 points on the composite score. In outcome in the MRV group there was a statistically significant difference between those who received medication and those who did not. The use of medication did not influence outcome in the migraine headache group.
Significant proportions of patients with migraine complain of space and motion discomfort. 9,11 Fifty-nine percent of the 39 subjects in this study had space and motion discomfort. A greater proportion of the subjects with MRV complained of space and motion discomfort. The use of medication may control space and motion discomfort and allow patients to tolerate the increased activity required in vestibular physical therapy. 17 Collaboration between the physical therapist providing vestibular rehabilitation and the physician improves the process of determining when medication is needed. The therapist can communicate to the physician how well the patient is tolerating the exercise and can recommend initiating or adjusting medication. In our clinical experience, we have found this collaborative relationship very beneficial to patient care and outcome.
After reviewing all of the data from these 39 patients, it is clear that a clinical randomized trial is needed to answer many of the questions that developed. The question of when and how medication should be used in conjunction with or in the absence of physical therapy intervention has not been answered. In addition, controlling the type of medication during intervention would have helped to clarify which medications are most effective in decreasing both the space and motion symptoms plus headache. A study comparing the improvements seen in patients with migraine headache with improvements in patients with similar vestibular diagnoses who have no history of migraine would allow us to determine the influence of migraine on patient outcome.
Because our only measure of balance was the DGI, questions remain about how patients with migraine perform on other measures of balance. It has been documented that patients with migraine experience anxiety, which can contribute to dizziness. 9,25 Could the improvements seen with vestibular rehabilitation in our population have been due to time spent by the physical therapist on education and reassurance, thereby decreasing anxiety and possibly decreasing dizziness, or was it the physical intervention itself? It is recommended that further studies address this and other important clinical issues.
Patients who had MRV or migraine headache improved with physical therapy intervention. There appears to be an improved outcome if a patient is taking an antimigraine medication in conjunction with physical therapy intervention. Only four of 39 patients referred for physical therapy were worse after intervention. After performing this retrospective study, the authors believe that physical therapy should be considered an efficacious treatment for patients with MRV. Also, a history of migraine should not be considered a contraindication to a trial of physical therapy.