Content uploaded by John A Fleetham
Author content
All content in this area was uploaded by John A Fleetham on Sep 26, 2014
Content may be subject to copyright.
181 Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
The Relationship Between Mouth Opening and Sleep Stage-
Related Sleep Disordered Breathing
Hiroko Tsuda, Ph.D.1; Alan A. Lowe, Ph.D.1; Hui Chen, Ph.D.1; John A. Fleetham, M.D.3,4; Najib T. Ayas, M.D.3,4; Fernanda R. Almeida, Ph.D.2
1Department of Oral Health Sciences,The University of British Columbia, Vancouver, BC, Canda; 2Department of Oral Biological
and Medical Sciences,The University of British Columbia, Vancouver, BC, Canada; 3UBC Hospital Sleep Disorders Program,
Respiratory Division, The University of British Columbia, Vancouver, BC, Canada; 4Respiratory Division, Department of Medicine,
The University of British Columbia, Vancouver, BC, Canada
SCIENTIFIC INVESTIGATIONS
A
variety of factors contribute to the development of ob-
structive sleep apnea (OSA). A better understanding of
these factors would help future therapeutic advances in OSA.1
Some physiological studies suggest that mouth opening in-
creases upper airway collapsibility.2-4 Miyamoto et al. reported
that the vertical mandibular posture is more open during sleep
in patients with OSA than in healthy adults. Furthermore,
sleep stage was a signicant factor in mandibular opening in
the supine position but not in the lateral position in patients
with OSA, as determined by using magnetic sensors attached
inside the subject’s mouth.5,6 One study has recommended that
jaw opening should be kept to a minimum for better treat-
ment success with oral appliance (OA) therapy in patients
with OSA.7 It has also been reported that patients with a large
amount of mouth breathing, which is only possible when both
the soft palatal seal and mouth are opened, are less adherent
to nasal continuous positive airway pressure (nCPAP) ther-
apy than patients with a small amount of mouth breathing.8
Furthermore, some patients with OSA have a higher percent-
age of obstructive events during REM compared to NREM
sleep.9,10 In these studies, REM-dependent subjects were re-
ported as being young and female, and this REM dependency
was suggested as a factor in understanding the difference in
the mechanism of obstruction between different types of pa-
tients with OSA. The clinical meaning of REM dependency
has not been fully investigated. However, since it has been
reported that sleep disordered breathing (SDB) during NREM
sleep, but not REM sleep, is associated with a risk of daytime
sleepiness,11 sleep staging should be considered to be a vari-
able that requires understanding.
Previous physiological studies have investigated the
amount of mouth opening and the subsequent collapse of
the upper airway; however, REM dependency or supine de-
pendency determined for the entire sleep duration was not
Study Objectives: To evaluate mouth opening during sleep
and the possible correlations between mouth opening and spe-
cic patient characteristics.
Methods: A total of 55 patients consecutively referred to as-
sess snoring and suspected obstructive sleep apnea (OSA)
were included. Sensors to record mouth opening were at-
tached to each patient’s face and synchronized with a standard
polysomnogram. Mouth opening data were evaluated for each
sleep stage as a percentage of maximum mouth opening. The
patients were divided into 2 groups: patients with REM apnea
hypopnea index (AHI) > NREM AHI (REM-dependent group =
RD group), and patients with NREM AHI > REM AHI (NREM-
dependent group = ND group).
Results: A total of 42 patients (male 69.0%, mean age 51.4 ±
12.9 years) underwent successful data collection. The amount
of mouth opening during stage 1 (18.8% ± 14.6%) was sig-
nicantly smaller than stage 2 (23.7% ± 16.4%, p < 0.01) and
REM (29.2% ± 20.3%, p < 0.01). Age, body mass index (BMI),
Epworth Sleepiness Scale (ESS) score, and AHI exhibited no
correlation with mouth opening. The RD and the ND groups
exhibited similar age, BMI, ESS, and AHI variables, but the
ND group opened their mouths signicantly more than the
RD group during total sleep time (28.3% ± 13.6% vs 17.8% ±
17.3%, p < 0.01), stage 1 (23.2% ± 13.5% vs 12.9% ± 14.3%,
p < 0.01), stage 2 (28.1% ± 17.9% vs 17.9% ± 17.4%, p < 0.01),
and REM (34.7% ± 19.2% vs 21.9% ± 19.8%, p < 0.05).
Conclusions: The ND patients opened their mouths wider
than the RD patients during most sleep stages. The relation-
ship between REM-dependent AHI and the amount of mouth
opening may be a factor in the pathogenesis of OSA.
Keywords: Obstructive sleep apnea, REM-related obstructive
event, mouth opening
Citation: Tsuda H; Lowe AA; Chen H; Fleetham JA; Ayas NT;
Almeida FR. The relationship between mouth opening and
sleep stage-related sleep disordered breathing. J Clin Sleep
Med 2011;7(2):181-186.
BRIEF SUMMARY
Current Knowledge/Study Rationale: This study evaluated mouth
opening during sleep and correlations between mouth opening and
patients’ characteristics. Patients with NREM-dependent OSA opened
their mouth wider than REM-dependent patients. Disease severity,
BMI, age, and body position were not found to affect mouth opening
during sleep.
Study Impact: Mouth opening was not a causative factor or associ-Mouth opening was not a causative factor or associ-
ated with sleep apnea. However, considering the physiological mech-
anism of sleep, NREM-dependent OSA may be partially caused by
mouth opening.
182
Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
H Tsuda, AA Lowe, H Chen et al
METHODS
Subjects
Fifty-ve consecutive patients referred to the Sleep Disorder
Program at the University of British Columbia Hospital to as-
sess snoring and suspected OSA were included in this study.
The protocol had the prior approval of the clinical research eth-
ics board, UBC Ofce of Research Services. The aim of the
study was explained to all patients, and written informed con-
sent was obtained.
Sensor for Jaw Movement
Sensors for recording jaw movement, which use resonating
magnetic eld transducers (JAWSENS, nomics, Belgium), were
attached to each patient’s forehead and chin (Figure 1). The
data were evaluated simultaneously with a standard polysom-
nogram (PSG). The data from the magnet sensor were sampled
every 0.03 sec throughout the night. Before each patient went to
sleep, maximum open, grind, lateral movement, swallow, chew,
and mouth breathing were checked and calibrated in a supine
position in bed (Figure 2). After this initial data collection, the
data, expressed as a percentage of the maximum opening level,
were evaluated for each sleep stage. One hundred percent cor-
responds to a maximum opening of the mouth, and 0% corre-
sponds to a closed mouth position (teeth touching lightly).
Polysomnogram (PSG)
PSG data were analyzed manually according to the American
Academy of Sleep Medicine (AASM) criteria by polysomno-
graphic technologists blinded to the JAWSEN data. Each stage
was dened in 30-sec epochs, and the JAWSEN data were also
calculated as the average of each epoch. NREM stages 3 and 4
were combined and referred to as N3 in the analysis.
assessed in these studies. To better understand the causes of
each SDB event and the predictors of treatment success, it
may be useful to estimate how often, how long, and how wide
patients with OSA open their mouths during sleep, and the
correlation between mouth opening and patient characteristics
such as age, sex, body mass index (BMI), and apnea hypop-
nea index (AHI). The aim of this study was to evaluate mouth
opening during sleep and the possible correlations between
mouth opening and specic patient characteristics.
100% Max open
0% Closed mouth
Clench lips apart
Lateral shift
Grind
Chew Swallow Mouth breathing
Figure 2—Sample of PSG and JAWSENS data at the calibration
Lower line denes 0% as the closed mouth position, upper line denes 100% as the maximum opening position
2cm
2cm
Figure 1—JAWSEN sensors
Picture comes from instruction manual
183 Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
Mouth Opening and Sleep Disordered Breathing
mine the correlation between demographic variables and mouth
opening variables. The independent student t-test, the χ2 test,
and the Mann-Whitney test were used for comparisons between
the groups.
RESULTS
A total of 42 patients (male 69.1%, mean age 51.4 ± 12.9
years) successfully completed both the JAWSENS and PSG
data collection (Table 1, 2). Seven patients did not have com-
plete data on JAWSEN, and 6 patients had AHIs < 5/h. The
methods error test was performed 10 times for each distance
and was 0.17-0.44 according to different measures. The results
from the ruler and the measurements from the sensors had a
high correlation. The ICC value was 0.9995. Percent averages
of maximum mouth opening were calculated for specic sleep
stages. Stage 1 (N1) (18.8% ± 14.6%) had the smallest mouth
Data Analysis
Based on the PSG results, patients with an AHI < 5/h were
excluded from the analysis in order to exclude normal subjects.
Each subject’s demographic data such as age, sex, BMI, Epworth
Sleepiness Scale (ESS) score, Mallampati score, and medical
history were collected from the patient’s chart. After scoring
the data, the patients were divided into 2 groups: patients with
REM AHI > NREM AHI (REM-dependent group: RD group),
and patients with NREM AHI > REM AHI (NREM-dependent
group: ND group).12 Demographic, sleep-related, and mouth
opening data were determined for the 2 groups. The difference
between patients according to body position tendency (supine-
dependent or non-supine-dependent sleep apnea) was also esti-
mated according to a criteria previously described.13
Methods Error
To assess the methods error, 8 different distances on the ruler
(every 1 cm from 9-16 cm) were measured. The methods er-
ror (ME) was determined by the formula: ME = √Σd2/2(n-1)
where d is the difference between measurement pairs and n is
the number of pairs. To assess whether and how the measure-
ments from the sensor matched the measurements on the ruler,
we also measured 8 different distances (every 1 cm from 9-16
cm) and calculated an intraclass correlation coefcient (ICC).
Statistical Analysis
Statistical analyses were performed, and p values < 0.05
were considered signicant. Correlation between measure-
ments from the ruler and the sensor were tested by using the
ICC. The Friedman test, followed by the Wilcoxon signed ranks
test, was applied between the mouth opening variables and
sleep stages. A Pearson correlation test was applied to deter-
Table 1—Demographic and polysomnographic data of the study patients
Total RD group ND group p value
N 42 18 24
Male/female 29 / 13 10 / 8 19 / 5 NS
Age (years) 51.4 ± 12.9 52.9 ± 11.8 50.2 ± 13.7 NS
BMI (kg/m2) 30.4 ± 6.4 31.2 ± 7.3 29.7 ± 5.7 NS
ESS 11.2 ± 4.3 11.4 ± 4.8 11.2 ± 3.9 NS
Neck circumference (cm) 40.6 ± 4.5 40.2 ± 4.1 40.9 ± 4.8 NS
Mallampati score: I 0 0 0 NS
II 14 4 10
III 12 7 5
IV 14 7 7
Previous ENT surgery 9 (21.4%) 3 (16.7%) 6 (25.0%) NS
Nasal congestion 15 (35.7%) 6 (33.3%) 9 (37.5%) NS
AHI (/hours) 26.2 ± 18.8 22.3 ± 11.7 29.1 ± 22.6 NS
Min O2 (%) 84.5 ± 8.8 82.4 ± 10.5 86.0 ± 7.0 NS
Total sleep time (min) 367.6 ± 64.2 353.2 ± 57.2 378.4 ± 68.1 NS
Time of stage N1 (%) 10.8 ± 8.6 11.8 ± 9.6 9.7 ± 7.8 NS
Time of stage N2 (%) 70.7 ± 9.2 71.7 ± 9.3 70.0 ± 9.1 NS
Time of stage N3 (%) 2.2 ± 4.4 1.2 ± 3.0 3.0 ± 5.2 NS
Time of stage REM (%) 16.3 ± 5.5 15.3 ± 5.3 17.1 ± 5.7 NS
NS, no signicance (p > 0.05)
Table 2—Percentage of mouth opening in different body
positions and sleep stages
RD group ND group p value
Mouth opening in
Supine (%) 16.0 ± 19.1 28.2 ± 11.7 < 0.01
Side (%) 18.9 ± 17.9 28.1 ± 13.9 < 0.05
REM-supine (%) 19.6 ± 20.2 35.7 ± 17.7 < 0.01
NREM-supine (%) 16.4 ± 18.0 28.2 ± 10.9 < 0.01
REM-side (%) 24.2 ± 19.9 29.6 ± 19.9 NS
NREM-side (%) 17.5 ± 16.8 27.8 ± 16.1 < 0.05
p values mean a signicant mouth opening difference between RD group
and ND group; NS, no signicance (p > 0.05).
184
Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
H Tsuda, AA Lowe, H Chen et al
in most sleep stages. Severity of disease (AHI), BMI, age, and
body position were not found to affect mouth opening. Stage 1
(N1) had the smallest mouth opening when compared to stage
2 (N2) and REM. Koo and colleagues reported that OSA oc-
curring during REM is more frequent in women,9 and younger
women may be protected from OSA during NREM sleep, even
obese patients.10 Although there was no signicant difference
between the groups, in this study 44.4% of the RD group were
female compared to 20.8% of the ND group, which is similar
to what has been reported previously. In the RD group, patients
did not open their mouths much in any sleep stages including
REM, although most of their obstructive events occurred dur-
ing REM sleep. As only 20 patients (9 in the RD group and 11
in the ND group) had N3, this may be the reason that no signi-
cant difference was found between the two groups in this stage.
ND group subjects opened their mouths wider than did the
RD group patients, but the average amount of mouth opening
was not greater than 30% of the maximum opening. If it is hy-
pothesized that if a patient can open his or her mouth 50 mm
(normal range 40-60 mm),14 then for such patient 30% corre-
sponds to 15 mm. This is almost the same value as the mouth
opening used in previous physiological studies.2,3 In this study
design, we cannot conclude that mouth opening was a causative
factor for sleep apnea. However, considering the physiologi-
cal mechanism of sleep as previously reported in OSA patients,
the ND group’s obstructive events may be partially caused by
mouth opening.
When comparing our results to the previous study,6 this
report is in agreement in that sleep stage was found to be a
signicant factor in mouth opening. In the previous study, this
factor was only signicant when patients were in the supine
position. The differing results between the two studies may
be due to the differing methods of measurement and analysis
used. Also, the sensors were attached differently in each study.
In the study by Miyamoto and associates,6 the sensors were at-
tached on the molar area in the mouth and were calibrated on
opening when compared to stage 2 (N2) (23.7% ± 16.4%) and
REM (29.2% ± 20.3%) (Figure 3). The demographic variables
such as age, BMI, neck circumference, Mallampati score, ESS,
and AHI were found to have no correlation with the mouth
opening variables. A possible concern was whether it is pos-
sible that a patient would more likely be assigned to the RD
group than the ND group on the basis of chance alone due to
potentially small differences between REM AHI and NREM
AHI. We excluded patients whose REM AHI and NREM AHI
differences were < 5/h and performed the same analysis. There
was no difference in the ndings, so the results are presented for
all patients under the denition of previous criteria.12 The RD
group (n = 18) and the ND group (n = 24) exhibited similar age,
sex, BMI, ESS, AHI, neck circumference, Mallampati score,
previous upper airway surgery, nose congestion, and sleep-re-
lated variables (Table 1). The ND group patients opened their
mouths signicantly wider than the RD group during various
sleep stages (Figure 4): total sleep time (RD: 17.8% ± 17.3%,
ND: 28.3% ± 13.6%, p < 0.01), N1 (RD: 12.9% ± 14.3%, ND:
23.2% ± 13.5%, p < 0.01), N2 (RD: 17.9% ± 17.4%, ND: 28.1%
± 14.4%, p < 0.01), and REM (RD: 21.9 ± 19.8%, ND: 34.7%
± 19.2%, p < 0.05). Signicant mouth opening differences in
various body positions (Table 2), including time in the supine
and lateral position, REM in the supine position, and NREM
in the supine and side positions were identied between ND
and RD groups. The supine-dependent group (n = 17) and the
non-supine-dependent group (n = 25) also had similar age, sex,
BMI, ESS, AHI, neck circumference, Mallampati score, previ-
ous upper airway surgery, nasal congestion, sleep-related, and
mouth opening variables. In addition, it was difcult to visually
detect any specic pattern of the obstructive event from jaw
movement data.
DISCUSSION
In this study, NREM-dependent patients with OSA opened
their mouths wider than did REM-dependent patients with OSA
Percentage of wakeful maximum
mouth opening %
40
35
30
25
20
15
10
5
0
-5
-10
**
**
**
*
Sleep Stages
N1 N2 N3 REMAverage
RD group ND group
Figure 4—Amount of mouth opening in each sleep stage
between REM-dependent (RD) and NREM-dependent (ND)
groups
Statistical signicance *p < 0.05,**p < 0.01
Percentage of wakeful maximum
mouth opening %
60
50
40
30
20
10
0
Sleep Stages
N1 N2 N3 REM
**
**
Figure 3—Mouth opening in each sleep stage
Statistical signicance **p < 0.01
185 Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
Mouth Opening and Sleep Disordered Breathing
to draw conclusions about these relationships without analyz-
ing the breathing route directly. An interesting report related to
mouth breathing found that men breathe a signicantly higher
percentage of total ventilation through their mouths compared
to women, and that this increases with age.17 We could specu-
late that the typical patient who has NREM-dependent char-
acteristics is likely to be a male who suffers from OSA caused
by mouth breathing and exhibits wide mouth opening. Patients
with REM dependency are more likely to be female with few
episodes of mouth opening and mouth breathing. This might
be a factor in understanding the pathology of OSA in women
or the reason why there is a low percentage of OSA among
young women. In addition, it has been reported that female pa-
tients and supine-dependent male patients were more likely to
experience treatment success with OA therapy.18 We could not
nd any typical mouth opening followed by an apnea event,
as there were many kinds of mouth opening patterns in each
subject. Also it was difcult to dene the start or end points
of mouth opening. This study used the percentage of awake
maximal opening as a variable. Further studies are required to
explore and compare different denitions for mouth opening
during sleep. As mentioned in the limitations of this study, the
amount or type of mouth opening could not fully explain spe-
cic characteristics of the disease. However, REM dependency
was the only variable that exhibited a correlation with mouth
opening, and we suggest that REM dependency be explored
as an important variable in future studies. Future research is
required to determine the relationships between mouth open-
ing, mouth breathing, gender, and REM-dependent OSA. Such
studies could further determine the mechanism of the obstruc-
tive event in REM-dependent patients and assess whether REM
dependency is a useful indicator for the diagnosis or prediction
of treatment success with nCPAP, OA, or surgery.
CONCLUSION
For our patients, the amount of mouth opening during N1
sleep was signicantly less than during the other sleep stages.
The ND patients opened their mouths wider than did the RD
patients during most sleep stages. This study demonstrates
the relationship between REM dependency and the amount
of mouth opening, and this could explain the pathogenesis of
REM-dependent OSA.
REFERENCES
1. Owens R, Wellman A, Malhotra A. The chicken-or-egg debate in OSA pathogen-
esis. Sleep 2009;32:1255-6.
2. Meurice JC, Marc I, Carrier G, Series F. Effects of mouth opening on upper
airway collapsibility in normal sleeping subjects. Am J Respir Crit Care Med
1996;153:255-9.
3. Isono S, Tanaka A, Tagaito Y, Ishikawa T, Nishino T. Inuences of head posi-
tions and bite opening on collapsibility of the passive pharynx. J Appl Physiol
2004;97:339-46.
4. Ayuse T, Inazawa T, Kurata S, et al. Mouth-opening increases upper-airway col-
lapsibility without changing resistance during midazolam sedation. J Dent Res
2004;83:718-22.
5. Miyamoto K, Ozbek MM, Lowe AA, et al. Mandibular posture during sleep in
healthy adults. Arch Oral Biol 1998;43:269-75.
6. Miyamoto K, Ozbek MM, Lowe AA, et al. Mandibular posture during sleep in
patients with obstructive sleep apnoea. Arch Oral Biol 1999;44:657-64.
an articulated dental cast with vertical opening movement at
the incisors. In our study, sensors were attached on the forehead
and chin, which results in a greater distance between the sen-
sors than those in the previous study. In our study, the sensors
were calibrated with the patient lying on the bed. In addition,
an advantage of the sensors used in our study is that sensors do
not prevent the mouth from sealing. As patients therefore do
not need to put anything in the mouth, they are able to breathe
normally.6 In the lateral position, if the direction of the grav-
ity shifted sideways and this was followed by patients opening
their mouths with a side shift, the sensor in this study may have
been able to detect these movements more accurately than the
sensors used in the previous study.
The subjects who were excluded from analysis had similar
demographic characteristics to subjects who were included. As
the calibration and sensor position are sensitive and important,
it is impossible to get correct data without recalibration if the
sensor gets displaced. Only four of the seven excluded patients
had sensor displacement, while the others wanted to stop be-
cause of discomfort of the sensors and nasal cannula.
The limitations of the study may be related to the denition
of classication between REM- and NREM-dependent groups,
the characteristics of the sensors used, absence of electromyo-
gram (EMG), and lack of information regarding the preferred
breathing route of the patient.
In this study, RD and ND groups were divided by using
previous criteria12; however, it might have been possible to as-
sign someone to the RD group rather than the ND group on the
basis of chance alone due to small difference between REM
AHI and NREM AHI. As subjects in the previous report12 had
≥ 5 events/h differences between REM AHI and NREM AHI,
we tried to analyze the results excluding the patients whose
AHI differences between REM AHI and NREM AHI were <
5 events/h, in order to prevent the effects of these borderline
subjects. Both the results with and those without borderline pa-
tients exhibited similar signicant results; therefore, we have
considered that the criteria introduced in the previous report is
adequate for grouping REM dependency in this study.
In relation to the type of sensor, mouth opening included the
vertical and anterior-posterior movements, and these ratios are
different for each patient. Even if the movements were similar
for two patients, it would be unlikely that the proportions of
vertical/anterior-posterior would be similar, as we just mea-
sured the distance between two sensors.
Since we did not collect EMG data in this study, it is also
difcult to determine the mechanism of mouth opening. To un-
derstand the mechanism or reason for a patient to open his or
her mouth, the muscle function and the relationship between
the muscle activation and other physiological variables should
be investigated.
Breathing route, especially mouth breathing, could be the
reason for a patient to open his or her mouth. Although some
patients in this study reported nasal congestion, there was no
signicant correlation between this nasal congestion and mouth
opening variables. It has been reported that some patients with
OSA use their mouths for breathing while asleep, even if they
are able to breathe through their noses when awake,15 and
that an older population uses the oronasal route for breathing
more often than does a younger population.16 It is impossible
186
Journal of Clinical Sleep Medicine, Vol. 7, No. 2, 2011
H Tsuda, AA Lowe, H Chen et al
18. Marklund M, Stenlund H, Franklin KA. Mandibular advancement devices in 630
men and women with obstructive sleep apnea and snoring: tolerability and pre-
dictors of treatment success. Chest 2004;125:1270-8.
ACKNOWLEDGMENTS
The authors would like to thank Mrs. Ingrid Ellis for her editorial assistance in the
nal preparation of this manuscript. A MITACS Accelerate Internship Grant supported
in part the post-doctoral fellowship salary of H. Tsuda.
SUBMISSION & CORRESPONDENCE INFORMATION
Submitted for publication May, 2009
Submitted in nal revised form October, 2010
Accepted for publication October, 2010
Address correspondence to: Fernanda R. Almeida, 2199 Wesbrook Mall, Vancouver,
BC, CANADA V6T 1Z3; Tel: (604) 822-3623; Fax: (604) 822-3562; E-mail: falmeida@
interchange.ubc.ca
DISCLOSURE STATEMENT
This was not an industry supported study. The authors have indicated no nancial
conicts of interest.
7. L’Estrange PR, Battagel JM, Harkness B, Spratley MH, Nolan PJ, Jorgensen
GI. A method of studying adaptive changes of the oropharynx to variation in
mandibular position in patients with obstructive sleep apnoea. J Oral Rehabil
1996;23:699-711.
8. Bachour A, Maasilta P. Mouth breathing compromises adherence to nasal con-
tinuous positive airway pressure therapy. Chest 2004;126:1248-54.
9. Koo BB, Dostal J, Ioachimescu O, Budur K. The effects of gender and age on
REM-related sleep-disordered breathing. Sleep Breath 2008;12:259-64.
10. Koo BB, Patel SR, Strohl K, Hoffstein V. Rapid eye movement-related sleep-
disordered breathing: inuence of age and gender. Chest 2008;134:1156-61.
11. Punjabi NM, Bandeen-Roche K, Marx JJ, Neubauer DN, Smith PL, Schwartz
AR. The association between daytime sleepiness and sleep-disordered breath-
ing in NREM and REM sleep. Sleep 2002;25:307-14.
12. Siddiqui F, Walters AS, Goldstein D, Lahey M, Desai H. Half of patients with
obstructive sleep apnea have a higher NREM AHI than REM AHI. Sleep Med
2006;7:281-5.
13. Marklund M, Persson M, Franklin KA. Treatment success with a mandibu-
lar advancement device is related to supine-dependent sleep apnea. Chest
1998;114:1630-5.
14. Okeson JP. Management of temporomandibular disorders and occlusion. 6th ed.
St. Louis, MO: Mosby Elsvier, 2008.
15. Koutsourelakis I, Vagiakis E, Roussos C, Zakynthinos S. Obstructive sleep
apnoea and oral breathing in patients free of nasal obstruction. Eur Respir J
2006;28:1222-8.
16. Madronio MR, Di Somma E, Stavrinou R, et al. Older individuals have increased
oro-nasal breathing during sleep. Eur Respir J 2004;24:71-7.
17. Gleeson K, Zwillich CW, Braier K, White DP. Breathing route during sleep. Am
Rev Respir Dis 1986;134:115-20.