If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To examine the relationship of 25hydroxyvitamin D serum levels with BPPV incidence and recurrence rates.
Methods
A retrospective cross-sectional, case-controlled study with follow-up phone survey was performed on patients diagnosed with BPPV between 05/2017-05/2020, who had available 25hydroxyvitamin D serology. Patients were seen at a multidisciplinary, vestibular-focused, neurotology clinic at a tertiary referral center. Controls consisted of subjects from the National Health and Nutrition Examination Survey (NHANES), and a locoregional age, sex, and race-matched group of patients from our institution.
Results
Our BPPV cohort consisted of 173 patients (mean age 66.2 ± 11.8 years), who were predominately female (75.7%) and Caucasian (76.3%). Almost all age subgroups (BPPV, NHANES, and locoregional groups) ≤60 years old had insufficient levels of vitamin D. However, the overall BPPV cohort had a significantly higher vitamin D level than the NHANES control (31.4 ± 16.5 v. 26.0 ± 11.2 ng/mL, d=0.474 [0.323, 0.626]). There was no significant difference when compared to the overall locoregional control (31.4 ± 20.5 ng/mL). Migraines were significantly correlated to increased BPPV recurrence rates on univariate (beta=0.927, p=0.037, 95% CI: [0.057, 1.798]) and multiple regression analyses (beta=0.231, 95% CI: [0.024, 2.029], p=0.045). Furthermore, patients with BPPV recurrences had significantly lower levels of vitamin D at initial presentation when compared to patients with no recurrences (29.0 ± 12.0 v. 37.6 ± 18.3 ng/mL, d=0.571[0.139,1.001]).
Conclusion
Many BPPV patients in our cohort had insufficient vitamin D levels, and patients with BPPV recurrences had insufficient and significantly lower vitamin D levels than those without. As a readily available and affordable supplement, vitamin D may be used as an adjunct treatment but prospective studies should be done to confirm if it can prevent or reduce recurrence.
It is accepted that BPPV is caused by otoconia, or calcium carbonate crystals, that dislodge into the semicircular ducts, causing vertigo and nystagmus with changes in head position. However, the pathophysiology of this disease is unknown. BPPV has known risk factors of older age (>60), sex (female>male), osteoporosis, migraine, hypertension, hyperlipidemia, and smoking.[
Despite the burden of BPPV on patients’ medical and social well-being, it is considered a benign disorder. The American Academy of Otolaryngology-Head and Neck Surgery guidelines currently recommend canalith repositioning exercises, such as the Epley maneuver, or habituation exercises, such as the Brandt-Daroff exercises, as primary treatments for BPPV.[
] While effective at resolving BPPV episodes, physical therapy does not significantly reduce BPPV recurrence rates and it is unclear why some patients are more prone to recurrences than others.[
As otoconia are composed of calcium carbonate crystals, and vitamin D is responsible for calcium and phosphate homeostasis, vitamin D is speculated to have a role in the pathophysiology of BPPV.[
In this cross-sectional study, we explored incidence of vitamin D deficiency in a single institution cohort. Two control groups were used for assessment of patient demographics and regional influences on vitamin D levels: 1) the National Health and Nutrition Examination Survey (NHANES), representative of the general United States adult population, and 2) an age, gender, and race-matched locoregional cohort from our institution.
2. Materials and methods
This study was approved by our Institutional Review Board (Protocol #95413). First, a retrospective cross-sectional analysis was performed by collecting data from the medical records of patients diagnosed with BPPV in our multidisciplinary, vestibular-focused, neurotology clinic between 5/2017 and 5/2020. Patients were seen by a neurotologist and a vestibular therapist, who verified resolution of BPPV in follow-up visits.
Inclusion criteria for this study were: (i) upbeat torsional nystagmus on Dix-Hallpike maneuver (consistent with posterior canal BPPV), or horizontal geotropic/ apogeotropic nystagmus in the supine roll test for horizontal canalithiasis or cupulolithiasis (ii) 25hydroxyvitamin D serology result within six months of BPPV diagnosis, and (iii) ≥18 years of age. Patients with medical conditions affecting the ear or temporal bone (e.g. chronic otitis media) or comorbid conditions causing dizziness (e.g. Meniere's disease, especially since Meniere's disease is in itself a risk factor for BPPV) were excluded.[
] We excluded patients with vestibular migraine but not those with migraine headaches. Patient charts were reviewed for vitamin D supplementation prior to BPPV diagnosis. A patient flowchart is presented in Fig. 1.
Data relating to BPPV was extracted from medical records extra including: demographics, comorbid conditions, vitamin D levels, and BPPV recurrence. In clinical practice as well as in this paper, we used the Holick et al. definition of vitamin D sufficiency (>30 ng/mL), insufficiency (21-29 ng/mL), and deficiency (<21 ng/mL).[
Second, patients were telephoned (3.02-52.21 months following their initial visit for BPPV) to determine: (i) additional episodes of BPPV, (ii) duration of episodes, (iii) severity of episodes, (iv) length of the series of episodes, (v) vitamin D dosage, and (vi) duration of vitamin D supplementation. Of note, criteria for BPPV recurrence included: (i) episodes lasting <60 seconds, (ii) consistency with previously observed induced episodes in-office, (iii) episodes induced by specific positions of the head, (iv) no other neurological signs (i.e. cerebellar/oculomotor signs) and (v) no other pertinent medical conditions attributable to the episode of vertigo (Fig. 3). Three attempts were made by phone call to reach patients. Of the n=173 patients included in this study, 108 patients were successfully contacted. Patients were occasionally unable or unwilling to answer questions within the phone survey, resulting in an individual “n” per question different than the n=108 total patients. For example, only 95 were successfully contacted for confirmation of vitamin D supplementation taken prior to diagnosis.
To accurately analyze the influence of regional factors on the local patient cohort's 25hydroxyvitamin D levels, two control groups were utilized. The NHANES database served as the first control group, a comprehensive dataset of health and nutritional information collected and produced by the CDC. The aim of this dataset is to be representative of the general United States adult population (available at: https://www.cdc.gov/nchs/nhanes/index.htm). Of note, 25hydroxyvitamin D levels from our institution are reported in ng/mL. Thus, vitamin D levels from the NHANES database were converted from nmol/L to ng/mL using a conversion factor of 1 ng/mL=2.5 nmol/L.[
]The second control group was an age, sex, and race-matched locoregional cohort, curated by an institutional honest data broker (one of the BPPV patients was 91 years old and was not matched and later excluded from analysis when comparing the cohort to the locoregional control). These patients were treated at our institution, had a vitamin D serology in our system, were not on any vitamin D supplementation, and did not have any diagnoses pertaining to BPPV, dizziness or vitamin D deficiency, and were never seen for a main complaint of dizziness. For patients in the locoregional control group with vitamin D serology reported as <3.0 ng/mL, we used 0 ng/mL for statistical analyses.
All data analyses were performed using SPSS® or Excel® (2013) software. For descriptive analysis of demographics, mean and standard deviations were determined. For comparison of vitamin D levels among subgroups, effect sizes were calculated using Cohen's d. Univariate and multiple regression analyses were utilized to identify potential risk factors for BPPV recurrence. Results were considered significant at p<0.05.
3. Results
3.1 Patient characteristics
Of 680 BPPV patients seen in our clinic 05/2017-05/2020, 173 met inclusion criteria. A flowchart of patient selection is presented in Fig. 1. One hundred and sixty-nine patients (97.7%) had posterior canal BPPV, and 21 (12.1%) were diagnosed with bilateral BPPV. The most common comorbidities in the BPPV group were hyperlipidemia (n=67, 38.7%), hypertension (n=41, 23.7%), thyroid disease (n=33, 19.1%), asthma/allergic rhinitis (n=25, 14.5%), and diabetes mellitus (n=21, 12.1 %). Only 9.2% (n=16) of our patients had migraines despite a 13.5-15.3% prevalence of migraines in the general population in the US.[
] The mean age of this cohort was 66.2 ± 11.8 years (Fig. 2) and 75.7% (n=131) were women. The cohort was 76.3% (n=132) Caucasian and 16.8% (n=29) African American (Table 1).
The NHANES control group (n=5962) had a mean age of 46.7 ± 11.62 years; 42.5% (n=2535) were Caucasian and 19.9% (n=1184) were African American. Compared to the NHANES group, the BPPV group was significantly older (66.2 ± 12.3 vs. 46.7 ± 11.2 years, d=1.738 [1.584,1.892]), had a lower proportion of males (24.3% vs. 47.7%, d=0.471 [0.320, 0.622]), and had a higher proportion of Caucasian patients than African-American patients (see Table 1,Fig. 1).
3.2.1 Vitamin D subgroup analysis: NHANES control
The BPPV group had a significantly higher mean vitamin D level at the time of BPPV diagnosis compared to the NHANES population (31.4 ± 16.5 vs. 26.0 ± 11.2 ng/mL, d=0.474 [0.323, 0.626]). Vitamin D level subgroup analyses by gender, age, and race are summarized in Table 2. Male, female, and Caucasian patients with BPPV had greater vitamin D levels than those of the NHANES population. Age subgroups did not reveal any significant differences, but both BPPV and NHANES patients between 30-40, 40-50, and 50-60 years of age had vitamin D levels below the upper range of normal (30 ng/mL).
Table 2Vitamin D Levels in Subgroup Analyses. Comparison of 25hydroxyvitamin D levels between BPPV cohort and NHANES population subgroups.
3.2.2 Vitamin D subgroup analysis: Locoregional control
The locoregional control patients had similar vitamin D levels to that of BPPV patients overall (Table 3). However, males between 51-60 years of age with BPPV had a significantly lower vitamin D level than the locoregional control (22.1 ± 7.5 vs 33.9 ± 13.4 ng/mL). Furthermore, African-American males with BPPV had the lowest mean vitamin D levels of all subgroups (13.5 ± 1.6 ng/mL) which was significantly lower than that of the locoregional control (31.5 ± 2.3 ng/mL, d=8.950 [2.446, 15.454]).
Table 3Vitamin D Level Analysis of BPPV Cohort and Locoregional Control Subgroups.
NHANES and locoregional overall cohort numbers are not equivalent (n=173 v. 172) because we did not match the 91-year-old from the BPPV cohort, and this subject was not included in subsequent analyses.
BPPV Locoregional
172
31.5 (16.5) 32.4 (19.0)
0.051 [-0.161, 0.262]
Male
BPPV Locoregional
42
29.9 (13.0) 31.3 (18.3)
0.088 [-0.345, 0.521]
Female
BPPV Locoregional
131
32.0 (17.4) 32.8 (19.2)
0.044 [-0.199, 0.286]
30-40
BPPV Locoregional
5
21.7 (4.4) 22.6 (18.1)
0.068 [-1.172, 1.308]
41-50
BPPV Locoregional
11
24.5 (7.9) 32.9 (12.4)
0.808 [-0.061, 1.677
51-60
BPPV Locoregional
36
27.7 (17.3) 29.6 (13.7)
0.122 [-0.341, 0.584]
61-70
BPPV Locoregional
50
30.2 (18.1) 34.6 (19.4)
0.235 [-0.159, 0.628]
71-80
BPPV Locoregional
53
35.1 (14.9) 32.5 (22.1)
0.138 [-0.243, 0.519]
81-90
BPPV Locoregional
17
39.7 (16.4) 34.5 (21.0)
0.276 [-0.400, 0.952]
Male 30-40
BPPV Locoregional
1
n/a
n/a
Male 41-50
BPPV Locoregional
0
n/a
n/a
Male 51-60
BPPV Locoregional
10
22.1 (7.5) 33.9 (13.4)
1.087 [0.148, 2.026]
Male 61-70
BPPV Locoregional
8
24.2 (8.0) 28.8 (15.2)
0.378 [-0.610, 1.368]
Male 71-80
BPPV Locoregional
16
34.0 (11.2) 33.1 (18.8)
0.184 [-0.507, 0.882]
Male 81-90
BPPV Locoregional
6
40.0 (20.6) 28. 2 (29.4)
0.465 [-0.682, 1.611
Female 30-40
BPPV Locoregional
4
20.3 (3.5) 24.7 (20.1)
0.277 [-1.115, 1.670]
Female 41-50
BPPV Locoregional
11
24.5 (7.9) 32.9 (12.4)
0.808 [-0.061, 1.168]
Female 51-60
BPPV Locoregional
26
29.9 (19.5) 28.0 (13.7)
0.113 [-0.431, 0.657]
Female 61-70
BPPV Locoregional
42
31.3 (19.3) 35.7 (20.1)
0.223 [-0.206, 0.652]
Female 71-80
BPPV Locoregional
37
35.5 (16.4) 32.2 (23.6)
0.162 [-0.294, 0.619]
Female 81-90
BPPV Locoregional
11
39.5 (14.8) 38.0 (15.3)
0.100 [-0.737, 0.936]
Caucasian
BPPV Locoregional
132
34.8 (18.1) 33.3 (19.4)
0.080 [-0.223, 0.383]
African American
BPPV Locoregional
29
18.9 (8.5) 16.6 (8.6)
0.269 [-0.475, 1.013]
Caucasian male
BPPV Locoregional
37
31.8 (12.3) 30.8 (16.7)
0.068 [-0.388, 0.524]
Caucasian female
BPPV Locoregional
95
33.7 (17.6) 34.1 (20.9)
0.021 [-0.258, 0.299]
African American male
BPPV Locoregional
2
13.5 (1.6) 31.5 (2.30)
8.950 [2.446, 15.454]
African American female
BPPV Locoregional
27
25.5 (16.3) 18.2 (11.5)
0.518 [-0.025, 1.060]
NHANES and locoregional overall cohort numbers are not equivalent (n=173 v. 172) because we did not match the 91-year-old from the BPPV cohort, and this subject was not included in subsequent analyses.
Of 95 patients with available data, 32.6% (n=31) took vitamin D supplementation prior to BPPV diagnosis. Of these patients, 31 were able to confirm their vitamin D dose taken prior to BPPV diagnosis: 41.9% (n=13) <1000 IU, 45.2% (n=14) 1001-5000 IU, and 12.9% (n=4) >5000 IU. Of 110 patients who took vitamin D supplementation following their BPPV diagnosis (collected by phone calls (n=108) and chart review (n=2)) 70 (63.6%) had been on vitamin D for >1 year, 19 (17.3%) had been on vitamin D for <1 year, and 21 (19.1%) were no longer on vitamin D supplementation.
3.4 BPPV recurrence
Follow-up telephone calls made a mean of 17.3 ± 15.6 months after initial BPPV diagnosis revealed a mean recurrence rate of 1 ± 2.75 episodes every 2.5 months, with a mean severity of 5 ± 2.7 out of 10 (a subjective measure similar to a visual analog scale, with patients instructed to select 10 as the most severe) per episode. The majority of patients attended vestibular physical therapy after being referred during their initial BPPV diagnosis (69.5%, n=121) (Table 4), which we recommend routinely. Notably, patients with no recurrences had significantly different (p=0.012) vitamin D levels compared to individuals with ≥1 recurrence (38.0 vs 29.4 ng/mL). However, there was no difference in recurrence rates between those who were on vitamin D supplementation compared to those who were not.
Table 4Univariate Analyses of Comorbid Conditions of BPPV. The use of univariate analysis on common comorbidities within the cohort to determine their significance to BPPV incidence.
Cohort (n)% or Mean (SD)
Unstandardized coefficient
95.0% Confidence Interval for B
Sig.
Adjusted R2
B
[lower CI, upper CI]
Hypertension
41 23.6%
0.398
[-0.080, 0.876]
0.102
0.02
Hyperlipidemia
67 38.5%
-0.000
[-0.494, 0.493]
0.999
-0.012
Diabetes Mellitus
21 12.1%
0.351
[-0.539, 1.241]
0.435
-0.004
Asthma/allergic rhinitis
26 14.9%
-0.055
[-0.816, 0.707]
0.886
-0.012
Kidney disease
12 6.9%
-0.123
[-1.081, 0.836]
0.800
-0.011
Migraine
16 9.2%
0.927
[0.057, 1.798]
0.037
0.039
Heart disease
11 6.3%
-0.086
[-1.045, 0.872]
0.858
-0.011
Thyroid disease
33 19.0%
0.111
[-0.488, 0.710]
0.713
-0.01
Hearing loss
129 74.6%
-0.587
[-1.252, 0.079]
0.083
0.027
On vit. D prior to diagnosis
31 32.6%
0.300
[-0.579, 1.179]
0.495
-0.012
Bilateral BPPV
19 11.0%
-0.380
[-1.420, 0.661]
0.470
-0.006
Attended vestibular physical therapy
121 70.0%
0.078
[-0.429, 0.585]
0.762
-0.011
Currently on vit. D
88 50.9%
0.359
[-0.288, 1.006]
0.273
0.003
Severity of episodes (on a scale of 1-10, with 10 being most severe)
5.0 (2.7)
0.092
[-0.064, 0.248]
0.240
0.009
Duration of symptoms (how long before first visit had BPPV)
Regression analyses were performed to find patient characteristics associated with BPPV recurrence rate, with episodes per month as the dependent variable. We did not find any significance with respect to the factors of comorbid diseases, hearing loss, vitamin D supplementation prior to diagnosis, bilateral BPPV, attendance of vestibular therapy, vitamin D supplementation at the time of interview, duration and severity of episodes, or vitamin D serology (Table 4). Migraine was the only independent variable found to be significant on univariate regression (odds ratio=2.53, p=0.037) (Table 4). Hearing loss, hypertension, and migraine were included in the multiple regression analyses because they are common comorbidities with BPPV. [
] Multiple regression also found migraine to be a significant factor with an odds ratio of 2.79 for recurrence of BPPV (p=0.045), suggesting that those with migraines have a 2.79-fold increase in risk of recurrence of BPPV compared to those without. Hearing loss and hypertension were not significant factors on multiple regression analysis (p=0.056, p=0.158, respectively).
4. Discussion
Two control groups were utilized in this study. The 2017-2018 NHANES data, a large cohort with serum vitamin D data representative of the general US population, made up the first control group. Our BPPV cohort was significantly older (65.9 vs. 46.7 years) and had fewer males (23.8% vs. 47.7%) than the NHANES control. Our BPPV population demographics are consistent with those in literature, in which the female:male patient ratio can be 1.7-2.7, with mean ages of BPPV diagnosis ranging from 55-60 years.[
] The racial distribution was significantly different between the two groups (d=0.671[0.519, 0.823]). Notably, only 1.7% of our BPPV cohort were Hispanic, compared to 14.6% of the NHANES group. However, 16.8% of our cohort were African American versus 19.9% in the NHANES group. D'Silva et al. found that African Americans were 1.3 times more likely than Caucasians to have BPPV (p<0.02).[
Retrospective data suggests that the higher prevalence of benign paroxysmal positional vertigo in individuals with type 2 diabetes is mediated by hypertension.
] However, the majority of our BPPV patients were Caucasian (76.3%). Some of these differences are due to the regional variations in racial distribution but potentially to racial disparities in healthcare access.
Previous studies have shown migraine, hypertension, hyperlipidemia, and stroke to all be independently associated with BPPV development.[
Retrospective data suggests that the higher prevalence of benign paroxysmal positional vertigo in individuals with type 2 diabetes is mediated by hypertension.
] The common comorbidities present in our BPPV cohort, such as hyperlipidemia (38.5%), hypertension (23.6%), thyroid disease (19.0%), allergic rhinitis (14.9%), and diabetes mellitus (12.1%), were consistent with reports in literature.[
Retrospective data suggests that the higher prevalence of benign paroxysmal positional vertigo in individuals with type 2 diabetes is mediated by hypertension.
] Of our patients with migraine, 37.5% had recurrence of BPPV and our regression analysis shows an odds ratio of 2.79 for BPPV recurrence with migraine. Brevern et al. reported a higher odds ratio of 7.5 between migraines and occurence of BPPV.[
] There is no clear pathophysiologic link between migraines and BPPV. It has been proposed that migraines may induce repetitive vasospasm of labyrinthine arteries or alter the trigeminovascular system and affect the vestibulocochlear microcirculation. This leads to dislodging of the utricular otoconia into the semicircular canal, generating BPPV.[
Retrospective data suggests that the higher prevalence of benign paroxysmal positional vertigo in individuals with type 2 diabetes is mediated by hypertension.
] Of note, we excluded patients with comorbid vestibular migraine since vestibular migraine patients can have motion sensitivity and dizziness with movements of the head which would make the follow-up with the telephone survey less reliable for a true recurrence of BPPV. Migraine may be an even bigger factor of recurrence that is underestimated in our current paper, as previous studies have found a prevalence of BPPV of 42.3% among a cohort of patients with migraine. [
Our study found sufficient and higher vitamin D levels in the BPPV cohort compared to the NHANES population, which had insufficient levels (31.4 vs. 26.0 ng/mL). This effect (BPPV vitamin D levels > NHANES) persisted within male, female, and Caucasian (d=0.282 [0.107, 0.457]) subgroups. Interestingly males younger than 70 and females younger than 50 had insufficient levels of vitamin D and lower levels of vitamin D compared to NHANES controls but this did not reach significance. While our BPPV cohort had, on average, sufficient and higher levels of vitamin D compared to the NHANES cohort, subgroup analysis shows that male patients (< 70 years old) in our cohort had insufficient levels whereas females subjects were borderline sufficient starting at the 51-60 subgroup (29.9 ng/mL), with sufficiency in all subgroups >60 years.
This discrepancy in when gender subgroups achieve sufficient vitamin D levels could be due to females taking supplements at earlier ages than males, as the age of 50 may mark a peri-menopausal time which is when vitamin D supplementation is typically recommended to prevent osteoporosis. Previous studies of women have shown bone mineral density to remain relatively stable through 40s, with a rapid decline starting in a woman's 50s, which further supports the phenomenon seen herein.[
] Women below the age of 50 are less likely to be on supplements, however the BPPV cohort female subjects aged 30-40 and 41-50 had a more insufficient levels (20.3 ± 3.5 ng/mL and 24.5 ± 7.9 ng/mL, respectively) than their NHANES counterparts (23.8 ± 10.4 ng/mL and 25.5 ± 11.0 ng/mL, respectively). Also our BPPV cohort was made of 173 subjects compared to the 5962 subjects in the NHANES cohort, so it is possible significance could have been achieved if our cohort was a larger size. Differences in racial subgroup vitamin D levels were also noted. While Caucasians had sufficient levels (33 ng/mL), African American were in the insufficient range (24.6 ng/mL) but still higher than NHANES control.
Given the unclear relationship between vitamin D levels of the NHANES control and BPPV cohort, we repeated the analyses using a locoregional control group to reduce bias from geography, nutritional habits, and demographic variations. We did not find a significant difference in overall vitamin D level between the BPPV and locoregional control (BPPV: 31.5 ± 18.5 ng/mL, locoregional: 32.4 ± 19.0 ng/mL, d=0.051 [-0.161, 0.262]). It is noted that the vitamin D levels of the female age subgroups were increasing with age; starting as a deficient level of 20.3 ng/mL in the 30-40 age group and increasing to a sufficient level of 39.5 ng/mL in the 81-90 age group. As stated above, since women at the age of 50 usually take supplements, this can explain the increasing vitamin D levels. While the above findings were similar with this new control group, significant differences were found in the males ages 51-60 subgroup, with lower, insufficient vitamin D levels in the BPPV group compared to the control. The Institute of Medicine recommends adults up to 70 years to take 600 IU of vitamin D daily and 800 IU of vitamin D in adults older than 70 years.[
A global study of vitamin D status and parathyroid function in postmenopausal women with osteoporosis: baseline data from the multiple outcomes of raloxifene evaluation clinical trial.
] This, combined with previous studies showing decline of bone density beginning around 50-60, indicates that our findings may reflect either a period of bone mineral density loss exacerbation or under supplementation.[
] Low vitamin D levels within this age range can be attributed to factors not analyzed in this study, such as body habitus, level of physical activity, or diet. No matter how the vitamin D levels became insufficient, this can increase the likelihood of patients getting osteomalacia/osteoporosis or increasing the risk of falls, ultimately increasing the risk of fractures. Vitamin D could also be a specific risk factors in males whereas other factors related to osteoporosis in women may be independent risk factors diluting the effect of vitamin D deficiency in that subgroup. Finally, there is limited data on the turnover rate of otoconia in the inner ear and a vitamin D level at one point in time may not reflect previous deficiencies that were since corrected and could have contributed to BPPV pathology.
Although a causative link cannot be established through a retrospective study, if a low vitamin D level were related to BPPV in this subgroup, it is reasonable to assume that supplementation may play a role in preventing BPPV onset. For example, Buki et al.[
] showed a correlation between supplementation and a decrease in BPPV recurrence. A caveat to that is the fact that in our prospective follow-up, 31 out of 95 successfully contacted patients disclosed taking vitamin D supplements prior to the diagnosis of BPPV (Table 1). Forty percent of those subjects (13/31) were taking <1000 units but 13% (4/31) took > 5000 units and still developed BPPV. Oftentimes patients are unaware their over-the-counter supplements contain vitamin D and are undisclosed.
A significant difference was also found in the African American males with BPPV, who also had a lower vitamin D level than the locoregional cohort (BPPV: 13.5 ± 1.6 ng/mL, locoregional: 31.5 ± 2.3 ng/mL, d=8.950 [2.446, 15.454]). This could be attributed to the low number of African American males within the data set, but also African Americans are known to have vitamin D deficiency due to eumelanin production. However, the clinical significance of this low vitamin D level is similar to those of the 51-60 year old male BPPV subgroup and more severe deficiency may be an independent risk factor in these subgroups. African Americans and males below the age of 60 may not see the need for vitamin D supplementation. Some studies have found a positive correlation of vitamin D levels with testosterone levels.[
] As such, this could represent a period of the male aging process in which testosterone production begins to decrease. Another reason for the discrepancy between the two groups could be that there was a lower number of African American males within the cohort. Also, it must be noted that duration of vitamin D supplementation can't be accessed due to the cross-sectional analysis being a snapshot in time. Future studies could involve supplementing vitamin D in BPPV patients found to have low vitamin D serology, with subsequent follow-ups for a period of time to see how many recurrences they have compared to a control group with low vitamin D serology but no supplementation. Previous studies have shown that dose, frequency, and duration of Vitamin D supplementation are all factors influencing vitamin D serology and could be optimized in such a study.[
] It is also possible there are other confounders within these subgroups that are at play that we have yet to identify.
Theories have been proposed on how vitamin D deficiency may affect the risk of developing BPPV. Otoconia can degenerate due to age, medication exposures, and changes in pH or ionic concentration.[
] For example, the use of streptomycin was noted by Takumida et al. to contribute to otoconia decomposition and recrystallization in guinea pigs but streptomycin discontinuation led to otoconia restoration.[
] Unfortunately, a prospective longitudinal study comparing the prevalence of BPPV in subjects with longstanding vitamin D deficiency versus those without it is unrealistic for logistic and ethical reasons.
We also investigated the relationship of vitamin D levels with recurrence rates. We found that individuals with BPPV who had no recurrences had a greater vitamin D level than those with ≥1 recurrence (38.0 vs. 29.4 ng/mL). This suggests that vitamin D deficiency may play a role in the natural history/recurrence of BPPV, but we cannot conclude that vitamin D supplementation will decrease BPPV recurrences. It could be that chronic vitamin D deficiency is the major risk factor and if not detected early, supplementation can have minimal effect on otolith degeneration. Previous studies have found severe vitamin D deficiency (<10 ng/mL) to be associated with BPPV recurrence rates.[
]Talaat et al. determined that those with severe vitamin D deficiency had an odds ratio of 4.54 for BPPV recurrence, when compared to patients with vitamin D levels >10 ng/mL.[
] However, we found that patients compliant with supplementation did not have a significantly different rate of BPPV recurrence compared to those who did not take supplements. Furthermore, we did not find vitamin D supplementation to be associated with BPPV recurrence on regression analyses (Table 4). As the guidelines established by Holick et al.[
]are based on bone health, and we have found that patients with recurrences of BPPV have insufficient and significantly different (p=0.012) levels of vitamin D (29.4 ng/mL) vs. those with no recurrences (38.0 ng/mL), it is possible that further studies on ideal vitamin D levels specifically for BPPV are warranted. These levels may be less permissive than for bone health. Also, the duration of vitamin D deficiency may play a role as postulated above.
Within this study, we also analyzed the association of BPPV recurrence rate with the duration and severity of episodes, without any significant findings. Abdelmaksoud et al. found that the number of recurrent BPPV attacks had a negative correlation with vitamin D deficiency, with a significant decrease in recurrence rates 6 months after initiation of vitamin D supplementation with Epley maneuvers versus patients who were treated with the Epley maneuver alone.[
] However, we did not replicate these findings. It may reflect an inadequate period of supplementation in our patient population, or too wide a range of variation in supplementation duration, as Abdelmaksoud et al. assessed their patient's symptomatology at 6months follow-up. Or, this could represent a difference in environment as their study was conducted in Egypt and vitamin D levels are dependent on sun exposure. We also asked patients to report their perceived severity of symptoms on a scale of 1-10, with 10 being the most severe. This was also a non-significant factor in BPPV recurrence. Unfortunately, this is a poorly studied variable in BPPV literature. Sheikhzadeh et al. previously found that BPPV severity (called intensity in their study) significantly decreased after 2 months in treated (vitamin D supplemented) and non-treated vitamin D deficient groups, despite vitamin D sufficiency being reached in the groups receiving supplementation.[
] However, they did not analyze the correlation of BPPV recurrence rates with perceived severity of symptoms.
However, it can be assumed that an individual's perception of pain and disability widely varies, and is likely subject to recency bias. The cross-sectional nature of our study limits us from addressing this specific question in more depth.
This study is limited by the biases of a retrospective review and by a small sample size, which hindered analysis power. While it does not offer definitive answers, it further confirms that the relationship of BPPV with vitamin D levels is complex and not a direct causality. Given the low risk of toxicity with vitamin D supplementation, the prevalence of vitamin D deficiency in the general population, and the physical burden of BPPV, it is reasonable to check vitamin D levels and consider supplementation in patients who have insufficient/borderline insufficient levels. Further research should be done to determine if supplementation of vitamin D could not only increase serum concentrations but also decrease the rate of BPPV episodes; the present study is unable to do so due to limitations of the study design and the shortcomings of medication reconciliation to include vitamin D supplementation in the medication list of patients since these are often over the counter products. For example, on several occasions patients endorsed taking a daily multivitamin, but could not ascertain or did not acknowledge the dose vitamin D within their multivitamin as part of their reported daily vitamin D intake. In addition, screening patients for migraines and treating them could help reduce risk of recurrence since in this study as in others it seems to be a specific risk factor. Controlling for compliance notwithstanding, it remains unclear what threshold cut-off value of vitamin D levels would be considered adequate in this patient population.
Conflict of Interest
There are no conflicts of interest.
Acknowledgements
This publication was supported by the South Carolina Clinical & Translational Research (SCTR) Institute, with an academic home at the Medical University of South Carolina (NIH - NCATS Grant Number UL1 TR001450).
References
Agrawal Y
Ward BK
Minor LB.
Vestibular dysfunction: prevalence, impact and need for targeted treatment.
Retrospective data suggests that the higher prevalence of benign paroxysmal positional vertigo in individuals with type 2 diabetes is mediated by hypertension.
A global study of vitamin D status and parathyroid function in postmenopausal women with osteoporosis: baseline data from the multiple outcomes of raloxifene evaluation clinical trial.
NHANES and locoregional overall cohort numbers are not equivalent (n=173 v. 172) because we did not match the 91-year-old from the BPPV cohort, and this subject was not included in subsequent analyses.
00154-7&id=gr1.jpg){kind=link}
00154-7&id=gr2.jpg){kind=link}
00154-7&id=gr3.jpg){kind=link}