This prospective study shows the prevalence of BPPV in an older population with an increased risk of falling and demonstrates some effect of treatment on fall incidents. Fall incidents are multifactorial, however the contribution of BPPV on fall incidents has barely been investigated. For that reason, our dizziness centre collaborated with the geriatric department in our hospital to determine the prevalence of BPPV in an older population referred to the geriatric consultant because of an increased risk of falls.
First, we found that 19% of our study population suffered from BPPV. This percentage is comparable to that reported by Lawson et al. who analysed older patients in a specialised falls service complaining of dizziness (n = 850). Of these patients 14.5% had positional nystagmus.28 Two other studies found exceptionally high BPPV frequencies. Abbott et al. studied 37 patients ≥ 65 years of age who had been admitted to hospital.29 Forty-five percent of this study population was diagnosed with BPPV. However, both patients admitted with dizziness and patients admitted with a collapse or fall incident were included in this study which could increase the prevalence of BPPV. Hyland et al. found a prevalence of 39% suffering from BPPV in a cohort of more than 600 patients.30 However, this study had a retrospective design and it was conducted in a falls and balance clinic. This could have caused an overrepresentation of the true prevalence by possibly studying a different population since a falls and balance clinic implies to be a clinic for both people with a high falls risk and for people with balance disorders.
Second, we aimed to examine whether there was a reduction in the number and severity of fall incidents six months after a successful repositioning manoeuvre in older subjects with BPPV. Before treatment we counted a median of two fall incidents per person. Six months after treatment the median number was reduced to 1.5 (p = 0.04), which suggests a preventive effect of the treatment. Comparing our result with the studies of Jumani et al. and Ganança et al., presenting a reduction of median fall incidents per person of 1.7 and 2.3, respectively, the reduction in our study seems small.18, 19 However, these two studies had a retrospective design eliciting a possible recall or information bias, they had a longer follow-up period, and larger groups to include in their statistical tests. Although the magnitude of our result seems small, we think this outcome is clinically relevant nonetheless, since even a small reduction in fall incidents could be meaningful in the prevention of morbidity. Furthermore, this is the first study to prospectively demonstrate this effect on fall incidents.
Before treatment for BPPV, two patients were treated in the hospital because of the severity of their fall incident. Even though our small group size prevented us to statistically test for differences before and after treatment regarding the severity of fall incidents, after treatment none of the patients reported a fall incident with the necessity of in-hospital treatment.
We found no statistical significant differences between patients with and without BPPV regarding age, sex, cognitive impairment, smoking, alcohol consumption, a specific medical history or the use of specific medication increasing the risk of falling. Not surprisingly, patients with BPPV experienced positional dizziness more often than patients without BPPV. An interesting finding is that eight patients with BPPV reported no complaints of positional dizziness which is almost half of the patients (47%) diagnosed with BPPV. This is a higher percentage than the 26% found by Hyland et al. which corresponds with the possible inclusion of patients with prior balance complaints.30 Also, our result is confirming results of former studies claiming that older patients can have BPPV without experiencing the typical symptoms such as positional dizziness.21 22 23
Patients with BPPV had a significantly lower BMI than patients without BPPV. This could be an incidental finding since no other studies have reported this relation. One retrospective case-control study specifically analysed whether cardiovascular risk factors, including BMI, were associated with the risk of BPPV and found no relation.31 However, a recently published retrospective study reported an association between BPPV and poor nutritional status using a serum marker rather than BMI.32
A strong point of this study was the inclusion of an older, vulnerable and dependent population. However, the choice of population also resulted in difficulties in inclusion. Almost half of the referred patients were unwilling to participate, mostly because of logistical limitation to visit our dizziness clinic or because of their physical condition. Moreover, the geriatric consultants decided whether a patient was physically able to participate. This could have caused bias in selection. Another limitation is the possible information bias caused by the fact that we asked patients to recollect the number and severity of fall incidents until six months prior to the start of the study. To reduce this bias during the study, we asked patients to document every fall incident and the corresponding severity. Lastly, our study sample was rather small. This was partly due to the high number of patients declining participation. Another reason was that our study was partly conducted during the COVID-19 pandemic. For several months only urgent care was provided in the hospital and it was temporarily prohibited to include patients for medical research. After a while, medical research with human subjects was allowed again but patients were still reluctant to participate out of fear of infection by the SARS-CoV-2 corona virus in the hospital.
For future research, we recommend a more comprehensive, multi-centre study with easier access to diagnostic and therapeutic manoeuvres to obtain larger sample sizes and to reduce barriers for patients to refuse participation due to logistical reasons. Ideally, an independent researcher should be in charge of selecting participants to avoid selection bias.