WEN-LING LIAO,
PT1,2 • TZU-PU CHANG, MD3HSUAN-JU CHEN, MSc4,5 • CHIA-HUNG KAO, MD6-8
Benign Paroxysmal Positional Vertigo Is
Associated With an Increased Risk of
Fracture: A Population-Based Cohort
Study
B
enign paroxysmal positional vertigo (BPPV) is a
brief rotational vertigo induced by head position change.
It is the most common disorder causing dizziness, with a
lifetime prevalence of 2.4%.
24In spite of its benign
course, BPPV significantly increases the risk of falls.
14Benign paroxysmal positional vertigo is caused by dislodged otolithic debris entering a semicircular canal and subsequently inducing endolymph flow during position change.8 Although the mechanism is
clearly defined, the possible underlying causes are poorly understood. Benign paroxysmal positional vertigo occasionally directly results from head trauma or otologic disorders, but most cases are idiopathic.1
Idiopathic BPPV may be associated with old age and female sex. Previous studies have demonstrated a relationship between BPPV and osteoporosis by investigating the bone mass of patients with BPPV.12,23
Fracture is an important public health issue, particularly in the elderly, that increases mortality and impacts quality of life.7,10,13
Fractures in the elderly often arise from the combination of falls and osteoporosis.17 Elderly people with osteoporosis are
susceptible to low-energy fractures when falling from standing height or on stairs. Other risk factors for fractures include body mass index, smoking, and alcohol consumption.5 Vascular
diseases and dementia also increase risk of fracture.20,25
There is an increased risk of falls in those with BPPV, and the high ratio of osteoporosis in patients with BPPV may further increase the risk of fracture secondary to falls. Therefore, we hypothesized that individuals with BPPV would have higher risk of fractures. The primary purpose of this study was to investigate the association between BPPV and the risk of fracture, which has not been studied previously. The incidence of BPPV increases with age, as does that of fracture. As the global population ages, a better understanding of the potential association between the occurrence of fractures and BPPV may have crucial clinical implications.
METHODS
Data Sources
T
his population-based cohort study analyzed data from the National Health Insurance Research Database (NHIRD), managed by the Taiwan National Health Research Institutes, which contains comprehensive health care data on 1 million people randomly selected from all beneficiaries of the National Health Insurance program. Implemented in Taiwan in March 1995, the universal National Health Insurance program covers more than 99% of the entire population of Taiwan (23 million people) and has contracts with 97% of health care institutions in the country.2The present study included NHIRD claims data recorded between 1996 and 2011. The NHIRD encrypts patients’ personal information to protect privacy and provides researchers with anonymous identification numbers associated with relevant claim information, including patients’ sex, date of birth, medical services utilized, and prescriptions. Patient consent is not required to access the NHIRD. This study was approved by the Ethics Review Board of China Medical University (CMU-REC-101-012).
Study Participants
The study cohort consisted of patients aged 20 years or older who were newly diagnosed with BPPV, according to the Inter-national Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9- CM code 386.11), between January 1, 2000 and December 31, 2006. Patients with a history of fracture (ICD-9-CM codes 800- 829) before the date of BPPV diagnosis were excluded from the study.
The comparison cohort consisted of patients randomly selected from the database who had no history of vertigo-related diagnoses (ICD-9-CM codes 078.81, 386.0-386.9, 780.4) or fracture (ICD-9-CM codes 800-829). Those in the comparison group were matched to case patients at a control-to-case ratio of 4:1 by age (every 5 years), sex, and index year (year of BPPV diagnosis).
Criteria and Definitions
Each patient was followed to assess
events of fracture (ICD-9-CM codes 800- 829), which were classified according to body region: skull fractures (ICD-9-CM codes 800-804), trunk fractures (vertebra, ribs, and pelvis; ICD-9-CM codes 805-809), upper-limb fractures (ICD- 9-CM codes 810-819), and lower-limb fractures (ICD-9-CM codes 820-829). Follow-up was discontinued only when the patients died, on the dates of outcome (ie, fracture) incidence, or on December 31, 2011. The longest possible follow-up duration was 12 years. The medical records were updated annually.
Several variables, such as age; sex; and the presence/history of diabetes mellitus (ICD-9-CM code 250), hypertension (ICD-9-CM codes 401-405), hyperlipidemia (ICD-9-CM codes 272.0-272.4), osteoporosis (ICD-9-CM code 733.0), dementia (ICD-9-CM codes 290, 294.1, 331.0-331.2), and ischemic or hemorrhagic stroke (ICD-9-CM codes 430- 438), were included in our analytical model to examine the relationship between fracture and comorbidities.
Statistical Analysis
The SAS statistical package (Version 9.2 for Windows; SAS Institute Inc, Cary, NC) was used for data analysis. A 2-sam ple t test was used to compare age between cohorts, and chi-square tests were used to compare demographic characteristics and comorbidities between those with and without BPPV. We used a Cox proportional hazard regression model to calculate the hazard ratio (HR) and 95% confidence interval (CI) of the risk of fracture in patients with BPPV compared to controls, after adjusting for age, sex, and comorbidities. The cumulative incidence of fracture in patients with BPPV and those without BPPV during the study period was analyzed using Kaplan-Meier curves, whereas the log-rank test was used to determine the differences between the 2 cohorts. We adopted a 2-tailed significance level of .05 for all statistical tests.
RESULTS
T
he BPPV cohort consisted of 3796 patients, and the non-BPPV cohort consisted of 15 184 patients. The mean SD longitudinal duration of follow-up was 7.07 2.89 years in the BPPV cohort and 7.08 2.85 years in the comparison cohort.TABLE 1 shows the baseline demographics and incidence of comorbidities for both groups. Both cohorts were matched by age and sex, and consequently no significant differences in age and sex were observed. The mean SD age was 57.13 16.01 years for the BPPV cohort and 56.72 16.12 years for the comparison cohort. Women accounted for 63.54% of the patients for both cohorts. For both cohorts, 35.91% were aged 65 years or older. The prevalence of comorbidities, namely, diabetes mellitus, hypertension, hyperlipidemia, osteoporosis, dementia, and stroke, was higher among patients with BPPV than among those without BPPV (chi-square test; P<.001).
In the FIGURE, the cumulative incidence curves of fracture, determined using Kaplan-Meier analysis, show a significantly higher risk of fracture in patients with BPPV than in those without BPPV (log-rank test; P<.001).
age, sex, or comorbidities. The overall incidence rate of fracture was 23.99 per 1000 person-years in the BPPV cohort and 20.09 per 1000 person-years in the comparison cohort. Compared to the patients without BPPV, the patients with BPPV exhibited a 1.14-fold (95% CI: 1.04, 1.25; P<.01) higher risk of fracture, adjusting for age, sex, and comorbidities.
An analysis of patients stratified according to sex revealed that men in the BPPV cohort had a 1.43-fold (95% CI: 1.22, 1.66) higher risk of fracture than did those in the non-BPPV cohort. Patients with BPPV aged 65 years or older exhibited a 1.17-fold (95% CI: 1.03, 1.33) higher risk of fracture compared to patients without BPPV. Patients were also stratified by comorbidity (diabetes, hypertension, hyperlipidemia, osteoporosis, dementia, or stroke) and, compared to patients with no comorbidities, the risk of fracture was significantly higher in the BPPV cohort (P<.05).
TABLE 3 shows the incidence rates and HRs of fracture according to body region. The incidence rates of fracture in the trunk (adjusted HR = 1.24; 95% CI: 1.06, 1.45) were significantly higher in the BPPV cohort than in the non-BPPV cohort. No significant difference was observed between the BPPV cohort and the non-BPPV cohort for skull, upper-limb, or lower-limb fractures. Additional analysis of the data revealed that the rate of hospitalization for all fractures was 34.47% for the BPPV cohort and 35.88% in the comparison cohort. The mean SD hospitalization cost was 52 370 54 001 New Taiwan dollars per person for the BPPV cohort and 54 717 73 312 New Taiwan dollars per person in the comparison cohort. No
significant difference was observed in hospitalization cost between cohorts (P = .66).
DISCUSSION
T
he primary finding of this study was that patients with BPPV exhibited a significantly higher risk of fracture than those without BPPV, even after adjusting for age, sex, and comorbidities. To our knowledge, this is the first longitudinal population-based study to explore this relationship. The findings of this study have clinical implications, specifically, the need to assess individuals with BPPV and provide proper interventions to prevent fractures.Although the magnitude of overall fracture risk was not high, our analyses showed that the positive association between BPPV and a higher risk of fracture may fluctuate, based on various demographic factors and sites of fracture. In addition, a previous study indicated that 9% of elderly participants had unrecognized BPPV,18 implying that the number of BPPV cases in the general
population may be higher than that identified in the current study; thus, the relative risk of fracture for those with BPPV compared to those without BPPV might have been underestimated.
In contrast to the lack of difference between groups in women, the current study revealed that men with BPPV exhibited a 1.43-fold higher risk of fracture than those without BPPV (TABLE 2). It is also noteworthy that the incidence rate of fractures for men with BPPV was higher
than that for women with BPPV, which is in contrast to the data between sexes for those without BPPV, for whom women had a higher incidence rate of fractures (TABLE 2). These results identifying an increased risk of fracture secondary to BPPV in men only may be related to sex
than differences in regular daily activities in Taiwan.
The current results show that only patients aged 65 years or older exhibited a significantly higher risk of fracture related to the presence of BPPV. Compared to younger individuals, elderly individu
als may have additional balance deficits attributable to multiple sensory-organ
degeneration and slower motor response, which make them more susceptible to fall, and the higher prevalence of osteoporosis among elders may further increase the risk of fall-related fractures.
Several factors may account for the greater occurrence of fracture in patients with BPPV. First, BPPV increases the risk of falls. Based on the data from a previous study, 44% of patients with positional nystagmus had fallen, compared to 28% of those without positional nystagmus.14 Furthermore, even after performance of the particle repositioning maneuver,
residual disequilibrium is often still present, as reported by Seok et al,21 who observed
residual dizziness after repositioning in two thirds of patients with BPPV. Dizziness is regarded as a major risk factor for falls, and falling is the leading cause of fracture.5,6 Data
have indicated that men and women with a fall history after the age of 45 years exhibit unadjusted HRs of fracture of 7.31 and 8.56, respectively. A study revealed that falls were the main cause of fracture in 89% of men and 93% of women.13
Second, osteoporosis may play a role in the association between fracture and BPPV. Although the result of our study showed that an association between BPPV and fracture was not observed when considering underlying osteoporosis, the database we used was linked to reimbursement of health insurance. In the database, coding for osteoporosis often reflects poor bone status, requiring medication or other medical measures. Therefore, in these individuals, osteoporosis itself may be the most important risk factor for fracture. In our study, when considering only those without the diagnosis of osteoporosis, the HR of fracture in those with BPPV was significantly higher than it was for those in the comparison group. However, osteoporosis is known as a silent disease, and is often detected only after a fracture has occurred. Low bone mineral density should not be ignored in individuals with BPPV, even
in the absence of a diagnosis of osteoporosis. There is growing evidence of the association between BPPV and osteoporosis.12,16,23 Vibert et al23 proposed 2 plausible explanations for
the association between BPPV and osteopenia and osteoporosis: (1) the internal structure of the otoconia or their connection and attachment to the gelatinous matrix may be disturbed by a decrease in estrogen, and (2) an increased concentration of free calcium in
the endolymph, resulting from increased calcium resorption, may reduce the capacity to dissolve the dislodged otoconia. Jeong et al12 reported that the prevalence of osteopenia
and osteoporosis was higher in women and men with BPPV than in participants with no history of vertigo. Recent studies suggest that the frequency of BPPV recurrence increases as bone mineral density decreases, and that a correlation may exist between BPPV and serum levels of biochemical markers of bone turnover.19,26
Third, we observed that the prevalences of diabetes mellitus, hypertension, hyperlipidemia, and stroke were higher among patients with BPPV. Previous studies have suggested that vascular risk factors (ie, diabetes mellitus, hypertension) lead to vascular damage to the inner ear and, thus, lead to BPPV.3,24 Furthermore, stroke and other complications
associated with these factors, such as diabetic neuropathy and peripheral artery disease, directly cause disequilibrium and claudication, and subsequently increase the risk of falling. These vascular risk factors are also risk factors for low bone mineral density. Therefore, comorbid factors may induce unfavorable physiological changes that predispose patients with BPPV to experience a fracture event.
A fracture puts a heavy economic burden on individuals and society.4,15 Based on our
estimate, the mean cost of hospitalization for fracture is more than 50 000 New Taiwan dollars per person, which is twice the average monthly salary in Taiwan.
Though both BPPV and fractures are more common in women, men seem to have worse outcomes after fractures. For example, after hip fracture, the relative hazards for all-cause mortality in the first 3 months were 5.75 in women compared to 7.95 in men,7 and the
12-month mortality rates were 32% in men compared to 17% in women.13 In our study
popula-tion, the trunk region exhibited the highest HR for fracture (crude HR = 1.35). A previous study reported pronounced reduction in most domains of the Medical Outcomes Study 36-Item Short-Form Health Survey in patients with vertebral fractures.9
To reduce the adverse outcomes of fractures in patients with BPPV, we suggest vestibular rehabilitation to reduce falls caused by BPPV itself. Vestibular rehabilitation used to treat BPPV includes (1) canalith repositioning therapy to flush the dislodged otolithic debris back into the utricle; and (2) balance retraining therapy to enhance central compensation, improve balance, and eliminate residual dizziness.11
The strength of this study is the use of a representative data source to identify a high number of BPPV cases and controls matched according to age and sex. We identified new cases of BPPV and followed them longitudinally for the event of fracture by using a cohort study design. The established time sequence strengthened inferences that an examined factor was the cause of an outcome. The ICD-9- CM contains a specific diagnostic code for BPPV, which increased the validity of the diagnosis. This study had several limitations that must be considered. First, potential risk factors for fracture, such as body mass index, smoking, and alcohol consumption, could not be obtained in the claims data. Sunami et al22
reported that smoking and alcohol consumption were significantly more common in comparison groups than in patients with BPPV. Based on previous findings, smoking and alcohol consumption may not be major risk factors for fractures in patients with BPPV. Further studies are needed to investigate these risk factors. Second, because some ICD-9 codes are based on vertigo symptom descriptions and unrecognized BPPV may be coded as another vertigo-related condition, patients with a history of vertigo-related diagnoses were excluded, which might have caused selection bias when establishing the comparison cohort.
CONCLUSION
P
atients with BPPV had a higher risk of fracture than those without BPPV. In clinical practice, the risk of fracture in patients with BPPV cannot be ignored, because fractures affect long-term quality of life considerably more than BPPV. To prevent fracture, vestibular rehabilitation and balance retraining are recommended.
