Dysphagia
DOI 10.1007/s00455-009-9265-5
ORIGINAL ARTICLE
Longitudinal Changes of the Swallowing Process in Subacute
Stroke Patients with Aspiration
Han Gil Seo • Byung-Mo Oh • Tai Ryoon Han
Received: 5 July 2009 / Accepted: 24 November 2009
Ó Springer Science+Business Media, LLC 2010
Abstract The purpose of this study was to evaluate
longitudinal changes of the swallowing process in stroke
patients with aspiration using kinematic analysis. Twenty-
eight subacute stroke patients with aspiration on fluid at
initial videofluoroscopic swallowing studies (VFSS) were
included. Follow-up VFSS was performed at 2-4 weeks
after initial studies that were conducted at an average of
26 days after stroke. Temporal and spatial variables were
calculated by two-dimensional motion analysis of hyoid
bone and epiglottic movements. Swallowing process delays
associated with hyoid bone and epiglottic movements were
improved at follow-up studies, whereas spatial variables
were not. Fourteen patients had recovered from aspiration
at follow-up. Time to the start of the hyoid movement was
longer in the nonrecovered patient group at initial studies
(1.76 ± 1.07 s) than in the recovered group (0.90 ±
0.82 s, P = 0.024). Although time-associated differences
between the nonrecovered and recovered groups disap-
peared at follow-up studies, aspiration persisted in the
nonrecovered group. This study shows that recovery from
delays in the swallowing process is a conspicuous change
during the subacute stage in stroke patients with aspiration.
Our findings suggest that delayed swallowing triggering at
initial VFSS is a useful predictor of poor recovery from
aspiration in stroke patients.
Keywords Stroke
Deglutition
Deglutition disorders

Respiratory aspiration
Recovery of function
H. G. Seo
B.-M. Oh
T. R. Han (&)
Department of Rehabilitation Medicine, Seoul National
University Hospital, 101, Daehang-ro, Jongno-gu,
Seoul 110-744, Republic of Korea
e-mail: tairyoon@snu.ac.kr
Introduction
Dysphagia after stroke is a common and disabling problem
that can cause significant clinical complications such as
asphyxia, aspiration pneumonia, and malnutrition. The
incidence of dysphagia has been reported to be as high as
80% in stroke patients in videofluoroscopic swallowing
studies (VFSS) [1]. Aspiration during videofluoroscopic
evaluations is an important finding with respect to an ele-
vated risk of pneumonia during the subacute stage after
stroke [2]. Teasell et al. [3] reported that 18 of 42 patients
(43%) who demonstrated evidence of aspiration on thin
liquids by VFSS, performed an average of 5 weeks fol-
lowing the onset of stroke, clinically stopped aspirating an
average of 6 weeks after the initial VFSS. On the other
hand, aspiration by VFSS can persist for over 6 months
after stroke in some patients [4].
The pharyngeal stage of swallowing starts with a verti-
cal and a horizontal movement of the hyoid bone caused by
contraction of the suprahyoid muscles [5]. The whole lar-
ynx is pulled upward and forward by passive and active
mechanisms with thyrohyoid muscle contraction. This
displacement tucks the larynx under the base of the tongue
[6]. Furthermore, a downward tilt of the epiglottis occurs
with this laryngeal elevation and tongue retraction occurs
to seal the laryngeal vestibule [7]. These movements of the
hyoid bone and epiglottis importantly prevent aspiration
during swallowing.
Several studies have analyzed the motions of pharyngeal
and laryngeal structures using kinematic methods. These
studies investigated vertical and horizontal displacements
of the hyoid bone in normal subjects [8–10] and the tem-
poral properties of pharyngeal swallowing in normal sub-
jects [11–13] and in stroke patients [14]. In a previous
study, we examined the use of two-dimensional kinematic
123

swallowing-motion analysis in patients with dysphagia due
to different etiologies [15]. However, few reports have
been issued on longitudinal changes in the swallowing
process after stroke, especially changes related to recovery
from aspiration using a kinematic analysis method.
The purpose of this study was to investigate longitudinal
changes in the kinematic and temporal characteristics of
the swallowing process in subacute stroke patients with
aspiration. We hoped to identify factors related to recovery
from aspiration.
Materials and Methods
Subjects
In this retrospective study, we reviewed our hospital’s VFSS
database covering the period from May 2005 to May 2008.
Only patients diagnosed by a neurologist as having experi-
enced acute stroke that was confirmed by computed tomog-
raphy (CT) or magnetic resonance imaging (MRI) were
selected. In the 337 stroke patients who were initially selected,
28 cases that met the following inclusion criteria were inclu-
ded in this study: (1) initial VFSS (T1) performed during the
subacute stage from 1 to 12 weeks after stroke onset, (2) at
least one follow-up VFSS (T2) performed at 2-4 weeks after
T1, and (3) definite subglottic aspiration on 2 or 5 ml of liquid
at T1. Patients with a history of previous stroke or some other
disease that could affect swallowing function, such as Par-
kinson’s disease, bulbar palsy, brain tumor, or head and neck
cancer, were excluded prior to the initial selection. Fifty-one
patients were excluded because their initial VFSS were per-
formed before 1 week or after 12 weeks poststroke. In the
remaining 286 patients, definite aspiration was noted in only
71 patients. Twenty-eight patients who underwent follow-up
VFSS at an adequate period of time without any significant
medical aggravation, which was potentially hazardous to
functional recovery, were finally selected.
All patients were treated for dysphagia after an initial
swallowing assessment using a routine protocol that
included diet modification and a daily treatment session
with a therapist, which involved posture education, exer-
cise, and pharyngeal stimulation for 30 min. Patients with
persistent aspiration were allocated to the nonrecovered
group, and those that recovered from aspiration at T2 were
allocated to the recovered group.
The study protocol was approved by the Institutional
Review Board at our hospital.
Videofluoroscopic Swallowing Study (VFSS)
VFSS was performed using a modification of the protocol
initially described by Logemann [16]. Subjects were seated
123
H. G. Seo et al.: Swallowing Process Changes in Stroke Patients
upright in a chair for the duration of the study and were given
2 and 5 ml of diluted barium (35% w/), pudding, curd-type
yogurt, and boiled rice twice in a spoon. A 24-mm-diameter
coin was taped under the chin at midline to serve as a
dimensional reference. Frame-by-frame images were
acquired to digital imaging files using a computer-based
image processing system equipped with a digital computer
frame grabber board (Pinnacle Studio MoviBox DV;
Pinnacle System, Inc., Mountain View, CA) and image
processing software (Pinnacle Studio 9.0). The X-ray volt-
age was set at a 40-kV peak, which allowed the soft tissues of
pharyngeal structures to be visualized.
The images were analyzed by two physiatrists who had at
least 2 years’ experience with VFSS and dysphagia
management under the supervision of a senior physician.
Conclusions were drawn by consensus. The videofluoro-
scopic dysphagia scale (VDS) [17] and the American
Speech-Language-Hearing Association National Outcome
Measure System (ASHA NOMS) swallowing scale [18]
were used to produce result scores. The VDS includes 14
items that represent oral and pharyngeal swallowing func-
tions in the VFSS, such as oral transit time, mastication,
premature bolus loss, laryngeal elevation, aspiration, and
vallecular and pyriform sinus residue. The VDS is a negative
rating system and provides a maximum possible score of
100, that is, a score of zero means a completely normal
finding. The ASHA NOMS scale represents the current state
of diet in dysphagic patients.
Two-Dimensional Motion Analysis
Videofluoroscopic studies were recorded at a resolution of
720 9 480 pixels at 30 frames per second. The images of
the first attempt to swallow 2 ml of liquid were analyzed as
we previously described [15]. In brief, locations of the
anterior-superior margin of the hyoid bone, the base-to-tip
margin of the epiglottis, and the head of the bolus were
digitized for each frame using a motion analysis software
system (Ariel Performance Analysis System; Ariel
Dynamics, Inc., Trabuco Canyon, CA). To calculate the
coordinates of points, we defined the y axis as the straight
line connecting the anterior margins of the inferior end-
plates of the fourth (the zero point) and second cervical
vertebral bodies, and we defined the x axis as the line
perpendicular to the y axis passing through the zero point
(Fig. 1). Data were transformed into actual distances and
angles using the dimensions of the above-mentioned coin.
The initial epiglottic angle was defined as 0°. The time
reference point (time 0) was defined as the moment when
the head of the food bolus passed the angle of the man-
dibular ramus at the base of the tongue [16]. The vertical
and horizontal motions of the hyoid bone and angles of
epiglottic downward tilt were plotted (Fig. 2a).
H. G. Seo et al.: Swallowing Process Changes in Stroke Patients
Fig. 1 Digital coordination of the hyoid bone and epiglottis and the
control frame for the x and y axes. The zero point (0, 0) is
operationally defined as the anterior-inferior border of the fourth
cervical vertebra. The y axis is the line drawn from the zero point to
the anterior-inferior border of the second cervical vertebra, and the x
axis is the line perpendicular to the y axis. A coin was used as a
dimensional reference
Measurements
Latencies (sec) from time 0 to initiation, peak, and termi-
nation of hyoid excursion and epiglottic downward tilt
were measured. The durations (seconds, s) of hyoid and
epiglottic motion from initiation to termination were also
recorded (Fig. 2b). Maximal excursions (mm) of the hyoid
bone in vertical and horizontal directions and maximal
angles (°) of epiglottic tilt were measured (Fig. 2b). All
measurements were made using MATLAB 7.4 (The
MathWorks, Inc., Natick, MA).
The reliabilities of the measurements were confirmed by
reliability testing using the VFSS images of ten patients.
Briefly, the analysis was performed by two well-trained
researchers using the same images, and 2 weeks later these
images were reanalyzed by the same researchers for the
variables analyzed in this study. Intrarater reliability testing
showed excellent agreement with intraclass correlation
coefficient (ICC) values ranging from 0.861 to 0.996.
Interrater reliability testing also produced good to excellent
results, with ICC values ranging from 0.752 to 0.995.
Statistical Analysis
The paired t test was performed to compare temporal and
spatial variables between T1 and T2 to evaluate longitu-
dinal changes of the swallowing process. Subjects were
divided into two groups according to the occurrence of
aspiration on liquid at T2. The v2 test or Fisher’s exact test
was used to compare baseline characteristics between the
recovered and nonrecovered groups. Student’s t test was
used to determine the significance of differences between
group variables. Clinical scores were compared using a
nonparametric test such as Wilcoxon’s signed ranks test or
the Mann-Whitney test. All statistical analyses were per-
formed using SPSS 12.0 (SPSS Inc., Chicago, IL). Values
given are mean ± standard deviations(SD). P values of
less than 0.05 were deemed significant.
Results
Patient characteristics are given in Table 1. All patients
showed aspiration on diluted barium at T1: 15 patients on
2 ml and 13 patients on 5 ml of liquid. Aspiration was
noted before, during, and after swallowing in 5, 22, and 1
patient, respectively. Mean VDS scores of all 28 patients
showed no significant difference between T1 and T2
(36.0 ± 15.0 and 31.1 ± 15.8, P = 0.202), whereas the
mean ASHA NOMS score was significantly improved at
T2 (3.3 ± 1.9 and 4.5 ± 1.9, P \ 0.001).
Table 2 gives the differences between temporal and
spatial variables at T1 and T2. All temporal variables of
hyoid and epiglottic motion were significantly delayed at
T1 as compared with T2. However, spatial variables and
durations of movement showed no significant differences at
T1 and T2.
Maximal angles of epiglottic tilt showed different dis-
tributions at T1 and T2. The distribution was definitely a
dichotomous pattern at T1 and no patient had a tilt angle of
50-80°. However, considerable overlap occurred between
these two initially dichotomous groups at T2 (Fig. 3).
Of the 28 patients, 14 recovered from aspiration at T2.
Table 3 gives patient characteristics of recovered and
nonrecovered groups. Patients that experienced aspiration
on only the 5 ml of liquid at T1 tended to recover from
aspiration at T2 better than patients with aspiration on the
2 ml of liquid (P = 0.058). VDS scores showed significant
differences between the recovered and nonrecovered
groups at T1 (31.6 ± 17.3 and 40.3 ± 11.3, P = 0.048)
and at T2 (21.2 ± 12.8 and 40.9 ± 12.2, P \ 0.001),
whereas AHSA NOMS scores showed a significant dif-
ference between the two groups at T2 only (5.6 ± 1.2 and
3.4 ± 1.9, P = 0.002).
Differences between temporal and spatial variables in
the recovered and nonrecovered groups are presented in
Table 4. Almost all temporal points at T1 were signifi-
cantly delayed in the nonrecovered group with the excep-
tion of the initiation of epiglottic tilt. Delayed swallowing
in the nonrecovered group was improved at T2 and no
123
 H. G. Seo et al.: Swallowing Process Changes in Stroke Patients

Fig. 2 a Example graphs of the vertical and horizontal motions of the
hyoid bone and angles of epiglottic downward tilt. b A superimposed
graph of the vertical motion of the hyoid bone, the angle of epiglottic
downward tilt, and bolus location. Temporal and spatial variables are
represented. MaxHV, maximal vertical excursion of hyoid bone;
MaxEA, maximal angle of epiglottic downward tilt; Hi, time to
initiation of hyoid motion; Hm, time to maximal excursion of hyoid
motion; Ht, time to termination of hyoid motion; Hd, duration of
hyoid motion; Ei, time to initiation of epiglottic tilt; Em, time to
maximal angle of epiglottic tilt; Et, time to termination of epiglottic
tilt; Ed, duration of epiglottic tilt

Table 1 Patient characteristics (n = 28) difference was observed between the two groups. At both
a T1 and T2, the maximal angle of epiglottic tilt was smaller
Age (years) 66.9 ± 11.0
in the nonrecovered group than in the recovered group, but
Gender (M/F) 19/9
these differences were of marginal significance (P = 0.174
Type of stroke (n)
and P = 0.056, respectively).
Ischemic stroke 23
Hemorrhagic stroke 5
Location of lesion (n)
Supratentorial lesion 15 Discussion
Infratentorial lesion 13
Side of lesion in the brain (n) The results of this study show that significant improve-
Right side 16 ments in swallowing occur with respect to the temporal
Left side 12 aspects of the hyoid bone and epiglottic movements in
Timing of aspiration (n) subacute stroke patients with dysphagia. Furthermore, the
Preswallowing 5
majority of temporal variables of the hyoid bone and epi-
During swallowing 22
glottic movements in the nonrecovered group were delayed
Postswallowing 1
significantly compared to those in the recovered group at
T1. Although temporal differences between the two study
Time from onset to initial study (days)a 26.0 ± 13.9
groups disappeared at T2, aspiration persisted in the non-
Time from onset to follow-up study (days)a 49 ± 17.7
recovered group.
Interval between 2 studies (days)a 23 ± 7.6
The latency to the initiation of hyoid motion in this
All other values are the number of patients study is the same as stage transition duration (STD), which
a
Values are means ± standard deviation was defined by Robbins et al. [19] as the time from when

123
H. G. Seo et al.: Swallowing Process Changes in Stroke Patients

Table 2 Comparisons of temporal and spatial variables at initial and aspirators on 5 ml was 1.28 s in this previous study, which
follow-up studies concurs with our T1 data (1.33 s). However, they did not
Variables Initial Follow-up P value* present changes over time. On the other hand, we studied
longitudinal changes between initial and follow-up studies
Mean SD Mean SD
in stroke patients. Although all patients in our study
MaxHV (mm) 15.51 5.35 14.80 5.83 0.543 showed greater delayed swallowing initiation at T1 com-
MaxHH (mm) 12.04 2.45 11.73 3.91 0.669 pared with normal data in previous studies, a significant
MaxEA (°) 80.64 35.33 81.63 32.93 0.795 temporal difference at T1 was found between the recovered
Hi (s) 1.33 1.03 0.68 0.82 0.002 and nonrecovered groups, and aspiration persisted after a
Hm (s) 1.88 1.11 1.25 0.85 0.003 significant improvement in swallowing initiation in the
Ht (s) 2.69 1.17 1.88 0.92 0.001 nonrecovered group. The above findings cast doubt on the
Hd (s) 1.37 0.60 1.24 0.42 0.270 notion that delayed swallowing initiation per se elevates an
Ei (s) 1.38 1.09 0.81 0.92 0.006 aspiration risk. Instead, these results suggest that delayed
Em (s) 1.98 1.13 1.28 0.97 0.004 swallowing triggering at initial VFSS is likely to portend
Et (s) 2.61 1.18 1.85 0.97 0.004 poor recovery from aspiration in subacute stroke patients.
Ed (s) 1.26 0.77 1.04 0.42 0.090 A recent study also showed that pharyngeal delay time at
an average of 3 months poststroke (range = 1–5 months)
MaxHV maximal vertical excursion of hyoid bone, MaxHH maximal
horizontal excursion of hyoid bonem, MaxEA maximal angle of
is a prognostic factor of persistent aspiration at 1 year in
epiglottic downward tilt, Hi time to initiation of hyoid motion, patients with stroke-related oropharyngeal dysphagia [20].
Hm time to maximal excursion of hyoid motion, Ht time to termi- The epiglottic downward tilt to seal the laryngeal ves-
nation of hyoid motion, Hd duration of hyoid motion, Ei time to tibule is an important protective mechanism against aspi-
initiation of epiglottic tilt, Em time to maximal angle of epiglottic tilt,
Et time to termination of epiglottic tilt, Ed duration of epiglottic tilt
ration, and as was shown by a previous study, a lack of
* P value by paired t test
functional epiglottic movement and abnormal epiglottic tilt
are highly associated with a risk of aspiration [21].
Although the maximal angle of epiglottic tilt did not
barium first passes the ramus of the mandible to the change during the follow-up period in the present study, the
beginning of maximum hyoid excursion. They also tested angle distribution among patients changed (Fig. 3). In
participants with 2-ml boluses as was used in the present particular, the dichotomous distribution of the angles
study, and STD for older normal subjects was found to observed at T1 suggested a decrease in epiglottic tilt after
range from 0.3 to 0.4 s. Kim et al. [14] reported that stroke stroke occurred in an ‘‘all or none’’ manner, which subse-
patients who aspirated had longer STD than nonaspirating quently gradually improved. However, few patients
stroke patients and normal subjects. The average STD of showed a definite change in epiglottic tilt between the two

Fig. 3 Distribution of maximal angles of epiglottic tilt at initial and distribution observed at initial and follow-up studies. The definite
follow-up studies. Left Scatterplot shows change in angle on a per- dichotomous pattern observed for initial studies was not evident for
subject basis. Right Histograms show the difference between the follow-up studies

123
 H. G. Seo et al.: Swallowing Process Changes in Stroke Patients

Table 3 Comparisons of patient characteristics in the recovered and Table 4 Comparisons of temporal and spatial variables in the
nonrecovered groups recovered and nonrecovered groups
Characteristics Recovered Nonrecovered P value* Variables Recovered Nonrecovered P value*

Age (years) 70.4 ± 8.9 63.4 ± 12.2 0.091 Mean SD Mean SD

Gender (n) 0.225 Initial study
Male 8 11 MaxHV (mm) 15.22 5.48 15.79 5.40 0.787
Female 6 3 MaxHH (mm) 11.84 2.49 12.24 2.49 0.680
Type of stroke (n) 1.000 MaxEA (°) 89.83 29.11 71.45 39.53 0.174
Ischemic 12 11 Hi (s) 0.90 0.82 1.76a 1.07 0.024
Hemorrhagic 2 3 Hm (s) 1.46 1.01 2.29a
1.07 0.044
Location of lesion (n) 0.705 Ht (s) 2.15 1.04 3.24a 1.05 0.011
Supratentorial lesion 8 7 Hd (s) 1.26 0.40 1.48 0.76 0.347
Infratentorial lesion 6 7 Ei (s) 1.05 0.95 1.71a 1.16 0.110
Side of lesion in 0.127 Em (s) 1.49 1.03 2.48a 1.04 0.018
the brain (n)
Et (s) 2.12 1.12 3.09a 1.06 0.026
Right side 10 6
Ed (s) 1.07 0.32 1.45 1.02 0.200
Left side 4 8
Follow-up study
Amount of liquid 0.058
caused aspiration at MaxHV (mm) 15.36 4.91 14.25 6.77 0.623
initial study (n) MaxHH (mm) 12.08 5.04 11.38 2.48 0.646
2 ml 5 10 MaxEA (°) 93.45 27.67 69.81 34.44 0.056
5 ml 9 4 Hi (s) 0.63 0.84 0.72a 0.83 0.771
Initial MMSE a
19.4 ± 8.6 21.8 ± 7.9 0.427 Hm (s) 1.22 0.85 1.29a 0.88 0.834
Initial MBIa 36.2 ± 22.9 46.8 ± 22.1 0.228 Ht (s) 1.76 0.97 2.00a 0.89 0.492
Time from onset to 26.9 ± 13.5 25.1 ± 13.5 0.731 Hd (s) 1.20 0.34 1.28 0.49 0.616
initial study (days) Ei (s) 0.80 0.95 0.83a 0.92 0.918
Interval between 23.1 ± 7.9 22.9 ± 7.5 0.961 Em (s) 1.26 0.99 1.29a 0.99 0.937
2 studies (days)
Et (s) 1.85 0.98 1.86a 1.00 0.977
MMSE mini mental state examination, MBI modified Barthel index Ed (s) 1.05 0.36 1.02 0.49 0.873
a
Initial MMSE and MBI scores were checked in 20 patients (12 in
MaxHV maximal vertical excursion of hyoid bone, MaxHH maximal
the recovered group and 8 in the nonrecovered group) and in 24
horizontal excursion of hyoid bone, MaxEA maximal angle of epi-
patients (13 in the recovered group and 11 in the nonrecovered
glottic downward tilt, Hi time to initiation of hyoid motion, Hm time
group), respectively
to maximal excursion of hyoid motion, Ht time to termination of
* P value by the Student’s t test, the v2 test, or Fisher’s exact test hyoid motion, Hd duration of hyoid motion, Ei time to initiation of
epiglottic tilt, Em time to maximal angle of epiglottic tilt, Et time to
studies. One patient with a decreased angle at T1 (33.46°) termination of epiglottic tilt, Ed duration of epiglottic tilt
a
showed a definite improvement at T2 (82.7°) and recovered P \ 0.05 by the paired t test for comparisons between initial and
follow-up studies
from aspiration. Two patients with reduced angles at T1
* P value by the Student’s t test
(28.14° and 48.31°) had intermediate angles at T2 (71.7°
and 66.97°, respectively) but no recovery from aspiration.
Furthermore, one patient with a normal angle at T1 reported no significant difference between aspirators and
(108.85°) showed an intermediate angle at T2 (54.41°) and nonaspirators in terms of maximal hyoid displacement
failed to recover from aspiration. The maximal angle of [22]. However, both our study and this recent study had
epiglottic tilt at T2 was greater in the recovered group than relatively small sample sizes. Furthermore, maximal ver-
in the nonrecovered group with a borderline significance tical displacement differed by more than 1 mm in the
(P = 0.056). Improved epiglottic tilt might contribute to recovered and nonrecovered groups at T2. A further
recovery from aspiration after stroke but appears rarely investigation is warranted in a larger cohort if we are to
over a relatively short period of 2–4 weeks. understand the influence of hyoid displacement on aspira-
The present analysis of vertical and horizontal dis- tion in stroke patients.
placement of the hyoid bone produced no significant This study focused on the temporal and kinematic
finding. Although movement of the hyoid bone is important properties of the hyoid bone and epiglottic movements.
for swallowing, its influence on the occurrence of aspira- However, some other factors could influence the recovery
tion is questionable. A recent study in poststroke patients of aspiration. The timing of aspiration, usually classified as

123
H. G. Seo et al.: Swallowing Process Changes in Stroke Patients
pre-, during, and post-swallowing aspiration, is one possible
factor [20]. The severity of aspiration could also play a role
in recovery. Further analysis should incorporate these
important features of aspiration, which can be checked by
VFSS. In addition, previous studies have suggested other
mechanisms related to airway protection, such as pharyn-
geal constriction [23], coordination between breathing and
swallowing [24], and the sensory regulation of swallowing
[25]. In the present study, persistent aspiration after the
recovery of swallowing initiation in the nonrecovered group
could have resulted from impairments of these mechanisms.
Future studies should address the influence of these mech-
anisms on the recovery of aspiration in stroke patients.
The VDS represents oral and pharyngeal swallowing
functions in the VFSS. It has been reported to be useful for
predicting the long-term prognosis of dysphagia after
stroke [17]. The mean VDS score at T1 was significantly
lower in the recovered group than in the nonrecovered
group, and the group VDS score difference increased at T2.
Although aspiration is one of the most important focuses of
the VFSS, it can be useful for predicting prognosis to
consider other oral and pharyngeal swallowing functions
together in stroke patients.
This study was designed as a retrospective study, and the
heterogeneity of time from stroke onset to study initiation
limits the interpretation and generalization of our results.
However, we did try to select patients in similar clinical
situations. Initial VFSS after acute stroke is usually per-
formed when a patient can cooperate and maintain a sitting
position for several minutes without mechanical ventilation
or intensive monitoring. It might have been more reasonable
to have included patients at a similar recovery state after
stroke rather than patients at a similar time from stroke
onset. If obvious dysphagia such as aspiration is confirmed
at initial VFSS, a follow-up study is usually performed in 2–
4 weeks at our hospital. This reassessment protocol is sup-
ported by a recent study on stroke survivors [26]. Therefore,
the inclusion criteria of this study represent normal clinical
evaluations after acute stroke. A prospective study with a
long-term follow-up period is needed to confirm our find-
ings and to acquire more information about the recovery
process of dysphagia after stroke.
In conclusion, this study shows that the recovery of
delayed hyoid bone and epiglottic movements is one of the
most conspicuous changes that occur during the subacute
phase in stroke patients with aspiration. Furthermore, our
findings suggest that delayed swallowing triggering at
initial VFSS might predict poor recovery from aspiration in
subacute stroke patients. A decrease in epiglottic tilt after
acute stroke revealed an ‘‘all or none’’ pattern initially; on
the other hand, its improvement appears to be a gradual
process. We believe that the findings of this investigation
aid our understanding of recovery from aspiration after
stroke and suggest that the VFSS may enable recovery to
be predicted.
References
1. Perry L, Love CP. Screening for dysphagia and aspiration in
acute stroke: a systematic review. Dysphagia. 2001;16:7–18.
2. Holas MA, DePippo KL, Reding MJ. Aspiration and relative risk
of medical complications following stroke. Arch Neurol. 1994;
51:1051–3.
3. Teasell RW, Bach D, McRae M. Prevalence and recovery of
aspiration poststroke: a retrospective analysis. Dysphagia. 1994;
9:35–9.
4. Mann G, Hankey GJ, Cameron D. Swallowing function after
stroke: prognosis and prognostic factors at 6 months. Stroke. 1999;
30:744–8.
5. Ertekin C, Aydogdu I. Neurophysiology of swallowing. Clin
Neurophysiol. 2003;114:2226–44.
6. Matsuo K, Palmer JB. Anatomy and physiology of feeding and
swallowing: normal and abnormal. Phys Med Rehabil Clin N Am
2008;19:691–707, vii.
7. Logemann JA. Swallowing disorders. Best Pract Res Clin Gas-
troenterol. 2007;21:563–73.
8. Logemann JA, Pauloski BR, Rademaker AW, Colangelo LA,
Kahrilas PJ, Smith CH. Temporal and biomechanical character-
istics of oropharyngeal swallow in younger and older men.
J Speech Lang Hear Res. 2000;43:1264–74.
9. Logemann JA, Pauloski BR, Rademaker AW, Kahrilas PJ.
Oropharyngeal swallow in younger and older women: videoflu-
oroscopic analysis. J Speech Lang Hear Res. 2002;45:434–45.
10. Kim Y, McCullough GH. Maximum hyoid displacement in nor-
mal swallowing. Dysphagia. 2008;23:274–9.
11. Ohmae Y, Logemann JA, Kaiser P, Hanson DG, Kahrilas PJ.
Timing of glottic closure during normal swallow. Head Neck.
1995;17:394–402.
12. Kim Y, McCullough GH, Asp CW. Temporal measurements of
pharyngeal swallowing in normal populations. Dysphagia. 2005;
20:290–6.
13. Leonard R, McKenzie S. Hyoid-bolus transit latencies in normal
swallow. Dysphagia. 2006;21:183–90.
14. Kim Y, McCullough GH. Stage transition duration in patients
poststroke. Dysphagia. 2007;22:299–305.
15. Paik NJ, Kim SJ, Lee HJ, Jeon JY, Lim JY, Han TR. Movement
of the hyoid bone and the epiglottis during swallowing in patients
with dysphagia from different etiologies. J Electromyogr Kine-
siol. 2008;18:329–35.
16. Logemann JA. Manual for the videofluoroscopic study of swal-
lowing. 2nd ed. Austin, TX: Pro-Ed; 1993.
17. Han TR, Paik NJ, Park JW, Kwon BS. The prediction of per-
sistent dysphagia beyond six months after stroke. Dysphagia.
2008;23:59–64.
18. Wesling M, Brady S, Jensen M, Nickell M, Statkus D, Escobar N.
Dysphagia outcomes in patients with brain tumors undergoing
inpatient rehabilitation. Dysphagia. 2003;18:203–10.
19. Robbins J, Hamilton JW, Lof GL, Kempster GB. Oropharyngeal
swallowing in normal adults of different ages. Gastroenterology.
1992;103:823–9.
20. Terre R, Mearin F. Resolution of tracheal aspiration after the acute
phase of stroke-related oropharyngeal dysphagia. Am J Gastroen-
terol. 2009;104:923–32.
21. Garon BR, Huang Z, Hommeyer M, Eckmann D, Stern GA,
Ormiston C. Epiglottic dysfunction: abnormal epiglottic move-
ment patterns. Dysphagia. 2002;17:57–68.
123
 H. G. Seo et al.: Swallowing Process Changes in Stroke Patients

22. Kim Y, McCullough GH. Maximal hyoid excursion in poststroke 26. Heckert KD, Komaroff E, Adler U, Barrett AM. Postacute
patients. Dysphagia 2009;24. doi:10.1007/s00455-009-9224-1. reevaluation may prevent dysphagia-associated morbidity. Stroke.
23. Yip H, Leonard R, Belafsky PC. Can a fluoroscopic estimation of 2009;40:1381–5.
pharyngeal constriction predict aspiration? Otolaryngol Head
Neck Surg. 2006;135:215–7.
24. Hardemark Cedborg AI, Sundman E, Boden K, Hedstrom HW, Han Gil Seo MD
Kuylenstierna R, Ekberg O, et al. Co-ordination of spontaneous Byung-Mo Oh MD
swallowing with respiratory airflow and diaphragmatic and
abdominal muscle activity in healthy adult humans. Exp Physiol. Tai Ryoon Han MD, PhD
2009;94:459–68.
25. Jafari S, Prince RA, Kim DY, Paydarfar D. Sensory regulation of
swallowing and airway protection: a role for the internal superior
laryngeal nerve in humans. J Physiol. 2003;550:287–304.

123