Benign paroxysmal positional vertigo (BPPV) is characterized by brief, recurrent episodes of vertigo accompanied by distinctive nystagmus triggered by changes in position [1]. BPPV is the most prevalent cause of peripheral vertigo, accounting for 17%–42% of cases [2]. The most common form is posterior canal BPPV, followed by lateral canal BPPV (LC-BPPV), accounting for 3.6%–22% of BPPV cases [2-5]. LC-BPPV typically manifests as vertigo and nystagmus, which are often induced by specific head movements, such as looking up or rolling over in bed. The supine roll test (SRT) is the standard diagnostic test for LC-BPPV [6]. This test involves rotating the patient’s head to the left or right from a supine position (head roll test [HRT]) and observing whether dizziness or nystagmus occurs in each position. A diagnosis of LC-BPPV can be made if either geotropic or apogeotropic nystagmus is observed during the SRT. Unlike posterior canal BPPV, which determines the affected side by analyzing the direction of nystagmus, LC-BPPV assesses the affected side by comparing the intensity of nystagmus induced by head rotations in both directions. Accurate diagnosis thus requires symmetric rotation angles and speeds during these movements. If there is a difference in rotational speed or angle between the two sides, the uneven force exerted on the otolith within the lateral semicircular canal (SCC) can alter the nystagmus pattern, potentially leading to a misdiagnosis. The body roll test (BRT), which involves simultaneous rolling of the head and body, ensures precise 90° rotations in both directions. It facilitates safe and adequate head rotations, even in patients with neck stiffness or cervical spine issues. This study aims to assess the diagnostic efficacy of the BRT compared to the HRT.

This study was approved by the Institutional Review Board of Incheon St. Mary’s Hospital (No. XC16OIMI0012O). All patients provided written informed consent to participate.

The average age of participants was 52.0±14.4 years, with a male-to-female ratio of 13:30. The time from onset to clinic visit averaged 12.5±18.1 days. The types of nystagmus observed were distributed as 18 geotropic and 25 apogeotropic. No significant differences were found in age, sex, or type of nystagmus between the two groups (Table 1). In patients suspected of having LC-BPPV, an effective diagnostic method must meet three criteria: (1) confirmation of LC-BPPV, (2) identification of the type of nystagmus (geotropic vs. apogeotropic), and (3) determination of the affected side (left vs. right). The first two criteria are met if the left and right roll tests consistently induce either geotropic or apogeotropic nystagmus. The lateralization of LC-BPPV can then be diagnosed by comparing the intensity of the nystagmus elicited on each side.

To account for the potential influence of fatigability on diagnostic results, the study participants were divided into two cohorts: group A and group B. Each group alternately underwent the HRT and BRT. During the initial BRT, the characteristic geotropic/apogeotropic nystagmus indicative of LC-BPPV was not elicited in one patient. However, during the subsequent HRT, this patient displayed characteristic apogeotropic nystagmus, leading to a diagnosis of LC-BPPV. For all other participants, the characteristic geotropic/apogeotropic nystagmus of LC-BPPV was consistently triggered in both testing sequences (HRT-BRT or BRT-HRT), confirming LC-BPPV as the underlying condition (Table 2). Determining the affected side in LC-BPPV requires a noticeable difference in nystagmus intensity between the bilateral roll tests. The analysis of nystagmus intensity dominance for determining the affected side showed similar results in both the first (81.4%, 35/43) and second (83.7%, 36/43) tests (Table 2).

Since there was no difference in outcomes based on the test sequence, the diagnostic rates were analyzed by combining the results of HRT and BRT, regardless of the order in which they were performed. Among the cases studied, 27 showed concordant results between HRT and BRT in diagnosing the affected side. There were nine cases with “single positive” results, where only one of the tests provided a side-specific diagnosis: BRT identified seven cases, and HRT identified two. In three patients, the intensity of nystagmus was bilaterally indistinguishable, preventing the determination of the affected side by either HRT or BRT. Additionally, four cases exhibited discordant diagnoses between HRT and BRT; in these instances, the bow and lying down position nystagmus (BLDN) patterns were used to determine the affected side. In these four cases, the BLDN was consistent with HRT in two cases, with BRT in one case, and was inconclusive in one case (Fig. 3). Overall, there were four cases where HRT identified the affected side, but BRT did not, eight cases where BRT identified the affected side but HRT did not, and four cases where neither test could identify the side.

Table 3 presents the diagnostic results for the affected side in 43 patients who were evaluated using both HRT and BRT. Of these patients, BRT successfully identified the affected side in 35 cases (81.4%), while HRT identified it in 31 cases (72.1%). These findings suggest that BRT has a slightly higher detection rate than HRT. Additionally, BRT resulted in fewer inconclusive outcomes, with only eight patients (18.6%) remaining undiagnosed. In contrast, HRT had 12 undiagnosed cases (27.9%), indicating a higher likelihood of obtaining a definitive diagnosis with BRT compared to HRT. However, the associated P-value was 0.189, indicating that the difference in diagnostic efficiency between BRT and HRT was not statistically significant within this study sample.

In the evaluation of the BRT, the confusion matrix analysis yielded an accuracy of 0.686, demonstrating the test’s general effectiveness. Additionally, both the recall and precision were notably high at 0.814, indicating a strong ability to correctly identify true positives with minimal false positives. The F1-score, which balances precision and recall, was equally high at 0.813953, confirming the test’s robust performance in accurately diagnosing conditions. The confusion matrix for the HRT showed an accuracy of 0.564, reflecting moderate effectiveness in overall test performance. Both recall and precision were 0.721, indicating that the test was reasonably effective at correctly identifying true positives while avoiding false positives. The F1-score, which was also 0.721, suggested a balanced measure of precision and recall, affirming the test’s reliability in cases where accurate detection is critical.

In the questionnaire completed after each test, both pain scores (HRT, 0.8±1.6; BRT, 0.7±1.4) and discomfort scores (HRT, 1.8±2.2; BRT, 1.8±1.9) were comparable across the types of tests, with no statistically significant differences observed (P>0.05). As for the order in which the tests were administered, there were no significant differences in pain scores (first, 1.7±1.8; second, 1.9±2.1) or discomfort scores (first, 0.7±1.4; second, 0.8±1.6) (Table 4). The results of the linear mixed model analysis for discomfort and pain scores are presented in Tables 5 and 6. For discomfort scores, among the fixed effects, age did not show a significant relationship (estimate=0; 95% CI, –0.04 to 0.04; P=0.878). However, sex was significantly associated with lower discomfort scores (estimate=–1.36; 95% CI, –2.64 to –0.09; P=0.036), indicating that men reported less discomfort than women. The period of symptom onset also had no significant effect on discomfort (estimate=–0.01; 95% CI, –0.04 to 0.03; P=0.718). In terms of random effects, the residual variance (σ²) for pain scores was 0.63, reflecting moderate within-group variability. The variance attributed to individual patients was 2.98, suggesting variability in scores among patients. However, the variance for test type was zero, indicating no differences in discomfort scores across the various testing methods.

Age did not significantly affect pain scores (estimate=0; 95% CI, –0.03 to 0.03; P=0.812). The effect of sex approached significance, with men generally reporting slightly lower pain scores than women (estimate=–0.96; 95% CI, –1.94 to 0.02; P=0.054). Similarly, the period of symptom onset did not significantly influence pain scores (estimate=–0.02; 95% CI, –0.04 to 0.00; P=0.118). In terms of random effects, the residual variance (σ²) for pain scores was 0.07, suggesting minimal variability within groups. The variance attributed to individual patients was 1.91, indicating variability in scores among patients. However, the variance for test type was zero, showing no differences in pain scores across different testing methods.

LC-BPPV is diagnosed when the geotropic or apogeotropic type of horizontal nystagmus is induced by the SRT in patients who report short episodes of positional vertigo [9,10]. The direction of the nystagmus, whether geotropic or apogeotropic, depends on the location of the detached otoliths within the lateral SCC. Therefore, the type of canalith repositioning procedure (CRP) selected for treatment is based on the observed nystagmus pattern.

LC-BPPV most commonly manifests as a canalithiasis type, where the detached otolith resides within the SCC. However, it can also present as a cupulolithiasis type, with the otolith adhering to the cupula. In cases of canalithiasis, during an SRT, turning the head towards the lesion side causes the otoliths to move towards the ampulla. This movement drags the endolymph towards the ampulla, leading to a deflection of the cupula in the utriculopetal direction. This action excites the lateral canal, triggering a fast phase of nystagmus towards the same side. As the patient is in a supine position, this nystagmus presents as geotropic, directed towards the ground. Conversely, when the head is turned away from the lesion side, the otolith on the lesion side moves away from the ampulla. This movement drags the endolymph in an ampullofugal direction, causing the cupula to deflect in the utriculofugal direction. This deflection suppresses the lateral SCC on the lesion side. The resulting inhibitory nystagmus, with its fast phase facing away from the lesion, appears geotropic (in the direction the head is turned, i.e., towards the ground) and is typically less intense than when the head is turned towards the lesion. Rotating the head away from the lesion side leads to a similar movement of the otoliths away from the ampulla, dragging the endolymph ampullofugal, and causing a utriculofugal deflection of the cupula, which inhibits the lateral SCC on the lesion side.

Cupulolithiasis is a condition in which the cupula becomes weighted down due to the attachment of otoliths. During the SRT, when the head is rotated toward the lesion side, the cupula deflects in the utriculofugal direction. This inhibits the lateral SCC and induces inhibitory nystagmus with the fast phase directed toward the opposite side, known as apogeotropic nystagmus, meaning it moves away from the ground. Conversely, when the head is turned away from the lesion side, the cupula deflects in the utriculopetal direction, which excites the lateral SCC and induces a stronger nystagmus that beats toward the lesion side. Thus, the direction of the nystagmus (geotropic vs. apogeotropic) helps determine the location of the detached otolith within the lateral SCC, and the intensity of the nystagmus on each side is compared to ascertain the affected side.

In clinical practice, distinguishing between geotropic and apogeotropic nystagmus is generally straightforward. However, the variations in nystagmus intensity can be subtle, complicating the identification of the affected side. In this study, the direction of nystagmus was consistently identifiable in both the HRT and BRT for all but one patient. For this individual, the nystagmus direction was indeterminate during the initial BRT but was clearly identified in the subsequent HRT (Table 2).

In assessing the direction of nystagmus, both HRT and BRT proved similarly effective. However, HRT identified the affected side in 31 out of 43 patients (72.1%), whereas BRT identified it in 35 out of 43 patients (81.4%). Although BRT appeared slightly more effective in identifying the affected side based on the intensity of nystagmus, this difference was not statistically significant (P=0.189) (Table 3). When comparing the performance metrics of BRT and HRT, BRT showed slightly higher values across all key indicators, including accuracy, recall, precision, and F1-score. Specifically, the accuracy of BRT, at 0.686, exceeded that of HRT, which was 0.564. Similarly, both the recall and precision of BRT were 0.814, compared to the value of 0.721 for HRT, with the same pattern reflected in the F1-scores. These results indicate that BRT is not only a viable alternative to HRT but also offers superior diagnostic performance, making it an excellent choice for accuracy and reliability in diagnosing LC-BPPV.

The intensity of nystagmus is expected to be significantly affected by the angle of head rotation in either direction [11]. Identifying the lesion side involves comparing the intensity of nystagmus triggered by bilateral head rotational stimuli. For an accurate diagnosis in the SRT, it is crucial that the rotational stimuli —both angle and speed—are equivalent on both sides. In the traditional supine “head”-roll test, head rotation can be either passive (performed by the examiner) or active (performed by the patient), with no standardized criteria or equipment. This lack of standardization can lead to unequal angles of head rotation on each side, potentially resulting in nystagmus intensity that does not accurately reflect the side of the lesion.

Another situation where the HRT may be inaccurate is when the angle of rotation is insufficient, even though it appears symmetrical. This can occur when neck rotation is limited by factors such as old age, neck stiffness, cervical spine disorders, or obesity. Active head-cervical range of motion (ROM) tends to decrease progressively after adolescence. Axial rotation decreased from 160° in teenagers (average age, 16) to 155° in young adults (average age, 23) and to 153° in middle-aged men (average age, 37) [12]. A systematic review on cervical ROM reported a decline in axial rotation with age, beginning notably in the 30s and continuing through the 60s. The review analyzed extensive datasets from Korea and Japan, with 642 individuals from Korea and 616 from Japan, making these countries the largest contributors to the study. The Korean study found no statistically significant differences between the sexes; however, it did show that cervical ROM significantly decreased with age. In contrast, the Japanese study highlighted variations in the aging processes of the cervical spine among populations by confirming a significant age-related decline in cervical ROM and identifying clear sex differences. This robust data supports the observed trend of a progressive age-related reduction in cervical spine flexibility, particularly affecting the ability to rotate the neck [13].

Inadequate head rotation due to decreased cervical ROM results in minimal movement of the otolith and cupula deflection, leading to a subtle intensity of induced nystagmus. This subtlety makes it challenging to differentiate between sides. Another significant limitation of the supine “head” roll test is the potential for misdiagnosing vertebral artery compression syndrome (VACS) as LC-BPPV. In patients with VACS, neck rotation compresses the vertebral artery by the cervical spine, reducing blood flow in the brain’s posterior circulation and causing nystagmus and vertigo [14,15]. Both conditions provoke vertigo and nystagmus upon head rotation. However, the difference between VACS and LC-BPPV lies in the triggering mechanisms and positional influences. VACS induces nystagmus and vertigo when turning the head, regardless of the patient’s position, and symptoms typically manifest only in the direction where the vessel is compressed. During the SRT, symptoms are triggered with the HRT but not with the BRT. In contrast, LC-BPPV symptoms are triggered by the movement of otoliths under the influence of gravity, causing nystagmus and vertigo during head rotation in both directions while lying down (SRT). Both the HRT and the BRT induce nystagmus and vertigo in patients with LC-BPPV, but these symptoms are not induced when turning the head while sitting.

In clinical practice, auxiliary diagnostic methods such as BLDN are used to clarify ambiguous cases when determining the lesion side with the SRT. The SRT involves rotating the patient’s head along the yaw axis while they are in a supine position, leveraging gravity’s effect on the lateral SCC to provoke a response. In contrast, BLDN tests involve rotating around the pitch axis of the lateral SCC during forward and backward head flexion. In the geotropic nystagmus form of LC-BPPV, the fast phase of nystagmus in the bow position and the opposite direction in the LDN test indicate the lesion side. For the apogeotropic nystagmus type of LC-BPPV, the opposite direction of nystagmus during the bow position test and the direction observed in the LDN test suggest the affected side. In BLDN, the lesion side is determined by the direction of the nystagmus rather than the intensity, offering a more straightforward approach than the SRT. Yetiser and Ince [16] suggested that the SRT holds the greatest diagnostic value among various positioning tests for LC-BPPV, and the lying down test also contributes to the diagnosis. Lee et al. [17] found that 76.8% of participants in the bow and lean test group exhibited either bowing or leaning nystagmus, aiding in accurately identifying the affected ear in LC-BPPV. Koo et al. [18] also noted that lateralizing the affected ear in LC-BPPV based on the direction of LDN may be beneficial. Their analysis of 54 patients with LC-BPPV showed that nystagmus predominantly beats ipsilesionally (toward the involved ear) in 80% of apogeotropic cases, while it typically beats contralesionally (toward the healthy ear) in 75% of geotropic cases. These findings indicate that the direction of lying-down nystagmus is a useful indicator for determining the affected ear in patients with LC-BPPV [18]. Another retrospective study assessed the lateralizing value of LDN and head-bending nystagmus (HBN) in 50 patients with LC-BPPV, using clinical and laboratory findings. LDN and HBN were observed in 68% and 76% of the patients, respectively, with the nystagmus direction induced by LDN and HBN consistent with the lesion side in 55.9% and 73.7% of the cases, respectively [19].

The other prospective study involved 211 patients and compared outcomes between those diagnosed exclusively with HRT and those diagnosed with both BLT and HRT. The findings revealed that the BLT group experienced higher remission rates after two CRPs, achieving 83.1% for canalolithiasis and 74.7% for cupulolithiasis, in contrast to 67.4% and 61.1% in the HRT group, respectively. Additionally, 76.8% of patients in the BLT group displayed bowing and/or leaning nystagmus, with 35.7% finding the BLT particularly advantageous when the HRT either failed to definitively identify the affected ear or produced inconsistent results [17]. In this study, there were four cases with discrepancies between the affected sides as indicated by HRT and BRT. The BLDN test was employed to ascertain the side of the lesion. The lesion side identified by the BLDN test results aligned with the HRT findings in two of the four patients and with the BRT findings in one patient. In the remaining patient, the lesion side could not be determined, even with the BLDN test (Fig. 3).

In our study, BRT successfully identified the lesion side in 35 of 43 patients (81.4%), demonstrating a higher diagnostic rate than HRT, although the difference was not statistically significant (Table 3). The reason the diagnostic rate of BRT was not significantly higher than expected may be due to the exclusion criteria used in the study, specifically excluding patients with cervical spine disease, severe obesity, and those over 70 years of age. It is likely that excluding these conditions for safety reasons, where BRT might have shown greater benefits, could have influenced the study results. In actual clinical practice, it is believed that BRT can be beneficial for safely and effectively diagnosing the affected side in patients with advanced age, obesity, or cervical spine disease.

The diagnostic rates for HRT (72.1%) and BRT (81.4%) align with those reported by other researchers. However, when HRT and BRT were combined, the diagnostic rate for the lesion side significantly increased to 90.7% (39/43). In our study, BRT outperformed HRT across all key metrics, including accuracy, recall, precision, and F1 score. Additionally, the generalized linear mixed model revealed no significant differences in pain and discomfort levels between the two tests. When diagnosing the lesion side in LC-BPPV is unclear, using both tests together can effectively determine the lesion side, thus improving the accuracy of CRP application and improving treatment success rates.

The HRT, which requires rapid head turning to induce nystagmus, was anticipated to cause discomfort and pain. Conversely, the BRT, involving the natural motion of rolling onto one’s side, was expected to be less painful and uncomfortable. However, according to the questionnaire analysis, the level of discomfort or pain reported by patients for both tests was almost negligible, with an average pain score of 1 out of 10 on the numeric pain scale. Notably, there was no significant difference in pain scores between the two tests (Table 4).

This study has some limitations. First, the relatively small sample size limits the statistical power, thereby constraining our ability to definitively conclude that BRT is superior to HRT. This limitation also affects the generalizability of our results. Future studies with larger cohorts are necessary to strengthen the robustness of our findings and provide more conclusive evidence regarding the efficacy of the diagnostic tests compared. Second, our study excluded certain patient populations, including those with cervical spine issues, obesity, or advanced age. This exclusion restricts the clinical applicability of our findings, as these groups might benefit most from an alternative diagnostic method such as BRT. To address this, future research should include a broader demographic to evaluate the utility of BRT across various clinical scenarios. Third, we assumed—but without verification—that head rotation in BRT could be achieved symmetrically and sufficiently as compared to HRT. Future research should directly measure and compare head rotation angles in both BRT and HRT, and correlate these angles with diagnostic rates. Additionally, clinical trials are necessary to compare the effectiveness of HRT and BRT in patients with advanced age, obesity, and cervical spine disorders.

While both the HRT and BRT tests are useful, the BRT can improve the reliability of diagnosing the affected side based on the intensity of nystagmus. The BRT is particularly beneficial when the HRT is not feasible due to neck stiffness, cervical spine disease, severe obesity, or advanced age. Additionally, combining the results of the HRT and BRT achieves a high diagnostic accuracy of 90% in identifying the lesion side. Therefore, when the results of a single test are inconclusive, conducting both tests together can significantly enhance diagnostic precision.