European Archives of Oto-Rhino-Laryngology

https://doi.org/10.1007/s00405-018-5165-3

OTOLOGY

Recovery of positional nystagmus after benign paroxysmal positional

vertigo fatigue
Takao Imai1   · Tomoko Okumura1 · Suetaka Nishiike2 · Noriaki Takeda3 · Yumi Ohta1 · Yasuhiro Osaki4 ·
Takashi Sato1 · Hidenori Inohara1

Received: 7 July 2018 / Accepted: 9 October 2018

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Purpose  In benign paroxysmal positional vertigo (BPPV), positional nystagmus is generally weaker when the Dix–Hallpike
test is repeated. This phenomenon is known as BPPV fatigue. The positional nystagmus induced by the Dix–Hallpike test
can be observed again when time has passed. There has been no study regarding the length of time required to recover the
positional nystagmus. The purpose of this study was to examine whether positional nystagmus recovers within 30 min after
the disappearance of the nystagmus by BPPV fatigue.
Methods  This was a prospective observational study. Twenty patients with posterior canal type of BPPV (canalolithiasis of
the posterior canal) were included. Dix–Hallpike tests were performed three times for each patient. A second Dix–Hallpike
test was performed immediately after the first Dix–Hallpike test. A third Dix–Hallpike test was performed 30 min after the
second Dix–Hallpike test. We recorded positional nystagmus induced by the Dix–Hallpike tests and analyzed maximum
slow-phase eye velocity (SPEV) of the positional nystagmus.
Results  The average maximum SPEV of positional nystagmus induced by the second Dix–Hallpike test (4.8°/s) was statisti-
cally lower than that induced by the first Dix–Hallpike test (48.0°/s); this decrease was caused by BPPV fatigue. There was
no statistical difference between average maximum SPEV of positional nystagmus induced by the first Dix–Hallpike test and
that induced by the third Dix–Hallpike test (41.6°/s); this indicates that the effect of BPPV fatigue disappeared. The effect
of BPPV fatigue disappears within 30 min.
Conclusions  A second Dix–Hallpike test should be performed at least 30 min after the first.

Keywords  Canalolithiasis · Dix–Hallpike test · Positional nystagmus · Three-dimensional · BPPV fatigue

Introduction

Benign paroxysmal positional vertigo (BPPV) is diagnosed

* Takao Imai
timai@ent.med.osaka‑u.ac.jp
1
Department of Otorhinolaryngology‑Head and Neck
Surgery, Osaka University Graduate School of Medicine, 2‑2
Yamadaoka, Suita‑shi, Osaka 565‑0871, Japan
2
Department of Otorhinolaryngology‑Head and Neck Surgery,
Osaka Rosai Hospital, Osaka, Japan
3
Department of Otorhinolaryngology‑Head and Neck Surgery,
Tokushima University Graduate School of Medicine,
Tokushima, Japan
4
Department of Otolaryngology, Kinki University Faculty
of Medicine, Osaka, Japan
by the observation of typical positional torsional nystag-
mus induced by the Dix–Hallpike test, where the patient
is brought from the upright to supine head-hanging posi-
tion with the head turned 45° to the affected ear [1]. In
BPPV, positional nystagmus is generally weaker when the
Dix–Hallpike test is repeated. This phenomenon is known
as BPPV fatigue [2]. The fatigability is a specific character-
istic of BPPV and is important in discriminating between
BPPV and central positional vertigo [3]. The disappearance
of positional nystagmus by BPPV fatigue does not indicate
that BPPV is cured, because head movement during the
Dix–Hallpike test is not intended to transfer free otoconial
debris from the semicircular canal back to the utricle (i.e., as
in the Epley maneuver, the treatment for BPPV) [4]. Thus,

13
Vol.:(0123456789)

the positional nystagmus induced by the Dix–Hallpike test
can be observed again when time has passed. There has been
no study regarding the length of time required to recover
the positional nystagmus. The purpose of this study was to
examine whether the positional nystagmus recovers within
30 min after the disappearance of the nystagmus by BPPV
fatigue.
Patients and methods
The current study was conducted with the approval of the
ethical committee at our hospitals and was performed in
accordance with the Declaration of Helsinki. Before the
experiment, written informed consent was obtained from
all participants. This study was a prospective observa-
tional study. The study included 20 patients (seven male,
13 female; 51–70 years, median 70 years; right ear affected
in 11, left ear affected in nine) with canalolithiasis of the
posterior canal (pc-BPPV) [1] at the Department of Oto-
rhinolaryngology—Head and Neck Surgery, Osaka Rosai
Hospital or Osaka University Hospital. They underwent
two consecutive Dix–Hallpike tests on the affected ear side.
Subsequently, they exited the consultation room. 30 minutes
later, they returned to the room and underwent a Dix–Hall-
pike test on the affected ear side. They were instructed not to
take a head position that might induce a vertigo attack, such
as bowing and leaning, during the 30-min waiting period.
For the Dix–Hallpike tests, patients wore goggles with an
infrared charge-coupled device (CCD) camera (RealEyes;
Micromedical Technologies, IL, USA) and the positional
nystagmus was recorded on a Windows computer with USB-
connected video capture, GV-USB2 (I-O DATA, Ishikawa,
Japan).
Fig. 1  Extraction of the two-dimensional coordinates of the center of
the pupil and an iris freckle. a Original eye image of a patient. In the
patient, movement of the coordinates of an iris freckle (indicated by
white dotted triangle) was tracked. b Two-dimensional coordinates of
13
European Archives of Oto-Rhino-Laryngology
Three‑dimensional analysis of eye movements
The 30-Hz eye movement images were converted to
720 × 480-pixel JPEG images and analyzed with an algo-
rithm developed in our laboratory [5, 6]. The two-dimen-
sional coordinates of the center of the pupil and an iris
freckle in the image were determined, as shown in Fig. 1
[6]. The head coordinates were reconstructed in three
dimensions and defined as follows: the X-axis was paral-
lel to the naso-occipital axis (positive forward), the Y-axis
was parallel to the interaural axis (positive left), and the
Z-axis was normal to the X–Y plane (positive upwards)
(inserted figure in Fig. 2a). The relationship between the
three-dimensional coordinates of the center of pupil and
the iris freckle in the head, compared between test and
reference positions, was used to calculate the axis angle
of the eye position [7, 8]. The reference position was
defined as the eye position when the participant was look-
ing “straight ahead” with their head in an upright position.
Looking “straight ahead” was defined as looking at a target
located horizontally in front of the eyes [9]. The X, Y, and
Z components of the axis angle of the eye position primar-
ily reflected the roll, pitch, and yaw components, respec-
tively. The direction of rotation was described from the
participants’ point of view. The accuracy of this method
for analyzing eye rotation vectors has been described else-
where [5, 10]. We calculated the axis angle of eye velocity
around the X-, Y-, and Z-axes [9]. We then extracted the
slow-phase eye velocity (SPEV) of the nystagmus using a
fuzzy set-based approach [11, 12] and found the maximum
SPEV of positional nystagmus.
the center of the pupil and an iris freckle. Two-dimensional coordi-
nates of the center of the pupil are indicated by a white cross; two-
dimensional coordinates of an iris freckle are indicated by white X
mark
European Archives of Oto-Rhino-Laryngology

Fig. 2  Axis angle of positional

nystagmus during the first Dix–
Hallpike test in a representative
patient A. a Axis angle of eye
position. Nystagmus could be
observed during head move-
ment, as well as after the head
reached the right head-hanging
position. b Axis angle of slow-
phase eye velocity (SPEV). Two
peaks were visible. The first
peak was induced by vestibulo-
ocular reflex during head
movement, while the second
peak was induced by positional
nystagmus. c Value of axis
angle of SPEV. The maximum
SPEV of positional nystagmus
was 113.7°/s

Statistical analysis
Statistical analysis was performed with the GraphPad Prism
7.0c (GraphPad Software, CA, USA). Differences in maxi-
mum SPEVs of positional nystagmus were compared for
statistical significance using Friedman test, a nonparamet-
ric test for testing the difference between several related
samples. In addition, Wilcoxon signed-ranks test was used
for post hoc test. Statistical significance was defined as a p
value < 0.05.
Results
Three-dimensional axis angles of positional nystagmus
induced by Dix–Hallpike tests in a representative patient A
(male, 67 years) are shown in Figs. 2, 3, and 4. In Fig. 2a, axis
angle of eye position of positional nystagmus is shown; in
Fig. 2b, axis angle of SPEV of positional nystagmus induced
by the first Dix–Hallpike test is shown. As shown in the initial
√
(X component of axis angle of SPEV)2 + (Y component of axis angle of SPEV)2 + (Z component of axis angle of SPEV)2 .
part of Fig. 2a, during head movement in the first Dix–Hall-
pike test, weak nystagmus could be observed. This nystag-
mus resulted from the vestibulo-ocular reflex (VOR) induced
by head movement. After the head reached the right head-
hanging position, right torsional (X component), upward (Y
component), and leftward (Z component) nystagmus could be
observed. As shown in Fig. 2b, in the X and Y components,
two peaks could be observed. The first peak was caused by
VOR, while the second was caused by abnormal positional
nystagmus. After the second peak, SPEVs in X, Y, and Z com-
ponents were reduced; after 12 s, the SPEV became almost
0. Thus, 12 s after the peak SPEV, the positional nystagmus
disappeared. These results indicate that after reaching the
affected side head-hanging position, the right torsional and
upward nystagmus appeared with short latency; furthermore,
the intensity of the nystagmus increased and then decreased,
lasting for 12 s. These are typical characteristics observed in
the positional nystagmus of patients with right-side affected
BPPV [1, 13, 14]. In Fig. 2c, the value of the axis angle of
SPEV (i.e., SPEV around the eye rotational axis) is shown.
The value was calculated as the following formula:

13
 European Archives of Oto-Rhino-Laryngology

Fig. 3  Axis angle of positional

nystagmus during the second
Dix–Hallpike test in patient A. a
Axis angle of eye position. Nys-
tagmus could be observed dur-
ing head movement, as well as
after the head reached the right
head-hanging position. b Axis
angle of slow-phase eye velocity
(SPEV). Only one peak was vis-
ible. This peak was induced by
vestibulo-ocular reflex during
head movement. c Value of axis
angle of SPEV. The maximum
SPEV of positional nystagmus
was 4.7°/s

Fig. 4  Axis angle of positional

nystagmus when third Dix–
Hallpike test in patient A. a
Axis angle of eye position. Nys-
tagmus could be observed dur-
ing head movement, as well as
after the head reached the right
head-hanging position. b Axis
angle of slow-phase eye velocity
(SPEV). Two peaks were vis-
ible. The first peak was induced
by vestibulo-ocular reflex
during head movement, while
the second peak was induced by
positional nystagmus. c Value
of axis angle of SPEV. The
maximum SPEV of positional
nystagmus was 108.0°/s

13
European Archives of Oto-Rhino-Laryngology
The peak SPEV of positional nystagmus, 113.7°/s, could
be observed. In this paper, the peak value was designated
as the maximum value of SPEV of positional nystagmus.
Figure 3a, b shows respective axis angles of the eye
position of positional nystagmus and of SPEV of positional
nystagmus, both induced by a second Dix–Hallpike test that
was performed immediately after the first Dix–Hallpike
test. As shown in the initial part of Fig. 3a, during head
movement in the second Dix–Hallpike test, VOR-induced
nystagmus could be observed. After the head reached the
right head-hanging position, weak right torsional and
upward positional nystagmus could be observed. The inten-
sity of the positional nystagmus at this point was much
weaker than that of the positional nystagmus observed in
Fig. 2a. As shown in Fig. 3b, a single peak was observed,
which was caused by VOR during head movement in the
second Dix–Hallpike test. The SPEV of the positional nys-
tagmus was very low; the maximum SPEV of the positional
nystagmus was 4.7°/s (Fig. 3c).
Figure 4a, b shows respective axis angle of eye position
of positional nystagmus and axis angle of SPEV of posi-
tional nystagmus, both induced by a third Dix–Hallpike test
that was performed 30 min after the second Dix–Hallpike
test. As shown in the initial part of Fig. 4a, during head
movement in the third Dix–Hallpike test, VOR-induced
nystagmus could be observed. After the head reached the
right head-hanging position, strong right torsional, upward,
and leftward positional nystagmus could be observed. The
intensity of the positional nystagmus at this point was much
stronger than that of the positional nystagmus, as shown in
Fig. 3a. The maximum SPEV of the positional nystagmus
was 108.0°/s (Fig. 4c), almost identical to the maximum
SPEV of the positional nystagmus induced by the first
Dix–Hallpike test (shown in Fig. 2c).
All maximum SPEVs of positional nystagmus in all 20
patients are shown in Fig. 5. The average maximum SPEVs
during the first, second, and third Dix–Hallpike tests were
48.0°/s, 4.8°/s, and 41.6°/s. Friedman’s test showed a sig-
nificant p value, p < 0.0001. Maximum SPEVs between the
first and second Dix–Hallpike tests were significantly dif-
ferent (p < 0.001, Wilcoxon signed-ranks test). In addition,
maximum SPEVs between the second and third Dix–Hall-
pike tests were significantly different (p < 0.001, Wilcoxon
signed-ranks test). However, there was no statistical dif-
ference in maximum SPEV between the first and third
Dix–Hallpike tests (p = 0.6575, Wilcoxon signed-ranks
test). These results indicate that the intensity of positional
nystagmus during the second Dix–Hallpike test became
weaker than during the first Dix–Hallpike test; however,
when the third Dix–Hallpike test was performed 30 min
later, the intensity returned to that observed during the
first Dix–Hallpike test. In two patients, whose data are
shown by rhombuses in Fig. 5, the intensity of positional
Fig. 5  Maximum slow-phase eye velocities (SPEVs) of positional
nystagmus in all patients. DH Dix–Hallpike test. The maximum
SPEV during the second Dix–Hallpike test was significantly lower
than that of the first Dix–Hallpike test. However, the maximum SPEV
during the third Dix–Hallpike test was not significantly different from
that of the first Dix–Hallpike test. The three maximum SPEVs for an
individual patient were connected by dotted lines. Rhombuses show
the data of two patients, whose maximum SPEV during the third
Dix–Hallpike test was less than one-tenth of the maximum SPEV
during the first Dix–Hallpike test. The gray squares show the aver-
aged maximum SPEV; error bars show the 95% confidence interval
nystagmus during the third Dix–Hallpike test was very
weak, because the maximum SPEVs during the third
Dix–Hallpike test were < 10% of the maximum SPEVs
during the first Dix–Hallpike test.
Discussion
In patients with pc-BPPV, the average maximum SPEV of
positional nystagmus induced by the second Dix–Hallpike
test was significantly smaller than that induced by the first
Dix–Hallpike test. This reduction of maximum SPEV was
caused by BPPV fatigue. However, there was no statistical
difference in the average maximum SPEV of positional nys-
tagmus between the first and third Dix–Hallpike tests. These
results indicate that the effect of BPPV fatigue was lost and
that the maximum SPEV of positional nystagmus recovered
30 min later. However, not all patients showed recovery
30 min later. In two patients, whose data are shown by rhom-
buses in Fig. 5, the maximum SPEV of positional nystagmus
remained small at the time of the third Dix–Hallpike test.
Although the mechanism causing BPPV fatigue remains
unknown, two hypotheses have been proposed. In one,
BPPV fatigue is attributed to the disintegration of lumps
of otoconial debris into smaller parts [15–17]. The other
hypothesis is as follows: during the first Dix–Hallpike test,
otoconial debris located close to the ampulla moves into
the posterior canal. When head position is returned to a sit-
ting position, the otoconial debris returns, but may not fully
regain its initial position. Thus, head movement during the
13

second Dix–Hallpike maneuver leads to differences in the
trajectories of otoconial debris that cause smaller cupula dis-
placements [2]. According to the first hypothesis, otoconial
debris that were disintegrated by the first Dix–Hallpike test
might have coagulated into lumps 30 min later. According
to the second hypothesis, otoconial debris located far from
the ampulla of the posterior canal after the first and second
Dix–Hallpike tests returned to the position near the ampulla
30 min later. Thus, when positional nystagmus is induced
by the first Dix–Hallpike test, some transformation occurs,
which may be either the disintegration of otoconial debris,
a change in the position of otoconial debris in the posterior
canal, or another mechanism that has not yet been identified.
The transformation is restored to its original form 30 min
later. Notably, in a few patients (10%, 2/20), the restoration
did not occur.
When pc-BPPV is diagnosed, if there is no observation
of positional nystagmus induced by an initial Dix–Hallpike
test, it is ineffective to perform a second Dix–Hallpike test
soon after the first, because positional nystagmus cannot
be clearly observed, due to BPPV fatigue. Thus, a second
Dix–Hallpike test should be performed at least 30 min after
the first.
Conclusions
The intensity of positional nystagmus induced by the
Dix–Hallpike test was weakened by BPPV fatigue. Thirty
minutes later, in most patients (90%, 18/20), the nystagmus
recovered to its original intensity upon disappearance of
BPPV fatigue.
Acknowledgements  Approval for the study was obtained from the eth-
ics committees of Osaka Rosai Hospital and Osaka University Hospital
(No. 16165), and the study was included in the University Hospital
Medical Information Network (UMIN) (UMIN000025291). We thank
Ryan Chastain-Gross, Ph.D., from Edanz Group (http://www.edanz​
editi​ng.com/ac) for editing a draft of this manuscript.
Author contributions  TI substantial contributions to conception, acqui-
sition of data, analysis and interpretation of data, and drafting of the
article. TO substantial contributions to analysis of data. SN substantial
contributions to acquisition of data. NT substantial contributions to
interpretation of data. YO substantial contributions to acquisition of
data. YO substantial contributions to acquisition of data. TS substantial
contributions to acquisition of data. HI substantial contributions to
study supervision.
Funding  This work was supported by JSPS KAKENHI Grant (no.
17K11327) and a Grant-in-Aid from the Japan Agency for Medical
Research and Development.
Compliance with ethical standards  new operative finding during posterior semicircular canal occlu-
Conflict of interest  The authors declare that they have no conflict of
interest.
13
European Archives of Oto-Rhino-Laryngology
Ethical approval  All procedures performed in studies involving human
participants were in accordance with the ethical standards of the insti-
tutional and/or national research committee and with the 1964 Helsinki
Declaration and its later amendments or comparable ethical standards.
Informed consent  Informed consent was obtained from all individual
participants included in the study.
References
1. von Brevern M, Bertholon P, Brandt T, Fife T, Imai T, Nuti D,
Newman-Toker D (2015) Benign paroxysmal positional vertigo:
diagnostic criteria. J Vestib Res 25(3–4):105–117
2. Boselli F, Kleiser L, Bockisch C, Hegemann S, Obrist D
(2014) Quantitative analysis of benign paroxysmal positional
vertigo fatigue under canalithiasis conditions. J Biomech
47(8):1853–1860
3. Brandt T (1990) Positional and positioning vertigo and nystagmus.
J Neurol Sci 95(1):3–28
4. Epley JM (1992) The canalithi repositioning procedure: for treat-
ment of benign paroxysmal positional vertigo. Otolaryngol Head
Neck Surg 107(3):399–404
5. Imai T, Takeda N, Morita M, Koizuka I, Kubo T, Miura K,
Nakamae K, Fujioka H (1999) Rotation vector analysis of eye
movement in three dimensions with an infrared CCD camera.
Acta Otolaryngol 119(1):24–28
6. Imai T, Takimoto Y, Takeda N, Uno A, Inohara H, Shimada S
(2016) High-speed video-oculography for measuring three-
dimensional rotation vectors of eye movements in mice. PLoS
One 11(3):e0152307
7. Schnabolk C, Raphan T (1994) Modeling three dimensional
velocity-to-position transformation in oculomotor control. J Neu-
rophysiol 71(2):623–638
8. Raphan T (1998) Modeling control of eye orientation in three
dimensions. I. Role of muscle pulleys in determining saccadic
trajectory. J Neurophysiol 79(5):2653–2667
9. Haslwanter T (1995) Mathematics of three-dimensional eye rota-
tions. Vis Res 35(12):1727–1739
10. Imai T, Sekine K, Hattori K, Takeda N, Koizuka I, Nakamae K,
Miura K, Fujioka H, Kubo T (2005) Comparing the accuracy of
video-oculography and the scleral search coil system in human
eye movement analysis. Auris Nasus Larynx 32(1):3–9
11. Arzi M, Mignin M (1987) A fuzzy set theoretical approach to
automatic analysis of nystagmic eye movements. IEEE Trans
Biomed Eng 36(9):954–963
12. Naoi K, Nakamae K, Fujioka H, Imai T, Sekine K, Takeda N,
Kubo T (2003) Three-dimensional eye movement simulator
extracting instantaneous eye movement rotation axes, the plane
formed by rotation axes, and innervations for eye muscles. IEICE
Trans Inf Syst 86(11):2452–2462
13. Imai T, Takeda N, Uno A, Morita M, Koizuka I, Kubo T (2002)
Three-dimensional eye rotation axis analysis of benign paroxys-
mal positioning nystagmus. ORL J Otorhinolaryngol Relat Spec
64(6):417–423
14. Imai T, Takeda N, Ikezono T, Shigeno K, Asai M, Watanabe Y,
Suzuki M, Committee for Standards in Diagnosis of Japan Society
for Equilibrium Research (2017) Classification, diagnostic criteria
and management of benign paroxysmal positional vertigo. Auris
Nasus Larynx 44(1):1–6
15. Parnes L, McClure J (1992) Free-floating endolymph particles: a
sion. Laryngoscope 102(9):988–992
European Archives of Oto-Rhino-Laryngology

16. Parnes L, Price-Jones R (1993) Particle repositioning maneuver 17. Parnes L, Agrawal S, Atlas J (2003) Diagnosis and manage-
for benign paroxysmal positional vertigo. Ann Otol Rhinol Lar- ment of benign paroxysmal positional vertigo (BPPV). CMAJ
yngol 102(5):325–331 169(7):681–693

13