Mandibular Movement Analysis To Assess Ef Ficacy of Oral Appliance Therapy in OSA
Mandibular Movement Analysis To Assess Ef Ficacy of Oral Appliance Therapy in OSA
RATIONALE: The respiratory effort index derived from vertical mandibular movements
(MM-REI) is a potential marker of increased respiratory effort during sleep. We eval-
uated the effectiveness of mandibular advancement splint therapy using MM-REI, in
comparison with the apnea-hypopnea index (AHI) and oxygen desaturation index
(ODI).
METHODS: Fifty-six subjects (median age, 47 years) with OSA treated with a custom
mandibular advancement splint (Herbst appliance) were evaluated at the end of the titration
procedure when snoring was reported absent by the sleep partner. We employed a magne-
tometer to capture mandibular movements (Brizzy; Nomics). Mandibular advancement
splint efficacy was assessed as the percent change from baseline, using Bayesian multilevel
models.
RESULTS: At the end of titration, all indices of OSA severity decreased compared with
baseline: AHI (–48.9% to –71.1%), ODI (–49.5% to –77.2%), with obstructive hypopnea
index and MM-REI showing the largest responses (–70.6% to –88.5% and –69.5% to –96.3%,
respectively). MM-REI normalization via reductions in both mandibular movement event
rate and duration accurately reflected efficacy of the appliance.
CONCLUSIONS: The reduction of vertical respiratory mandibular movements estimated by
MM-REI and sleep respiratory effort duration accompanied the decrease in obstructive
hypopneas, AHI, and ODI when snoring resolved in subjects with OSA treated with an
optimally titrated mandibular advancement splint. CHEST 2018; -(-):---
KEY WORDS: mandibular advancement device; mandibular movements; oral appliance; OSA
ABBREVIATIONS: AHI = apnea-hypopnea index; MM = mandibular Pépin), Université Grenoble Alpes, Saint-Martin-d’Hères, France; and
movement; MMAS = sequence of large mandibular movement (MML) Respiratory Medicine (Dr Silkoff), Drexel University School of Medi-
with a peak-to-peak displacement of at least 0.3 mm followed by a cine, Philadelphia, PA.
sharp and sudden MM of high amplitude (MMS) concomitant with FUNDING/SUPPORT: This work was supported by the French National
cortical arousal; MM-REI = respiratory effort index derived from MM Research Agency (ANR-12-TECS-0010), in the framework of the
characterization; MMS = sharp and sudden mandibular movement of “Investissements d’avenir” program (ANR-15-IDEX-02). https://life.
high amplitude; OAT = oral appliance therapy; ODI = oxygen desa- univ-grenoble-alpes.fr.
turation index; OHI = obstructive hypopnea index; PG = polygraphy; CORRESPONDENCE TO: Jean-Benoit Martinot, MD, Centre du Som-
PSG = polysomnography; TST = total sleep time estimated from MM
meil et de la Vigilance, CHU UCL Namur Site Ste Elisabeth, 15, Place
analysis; RE = respiratory effort; VAS = visual analog scale
Louise Godin, Namur 5000, Belgium; e-mail: [email protected]
AFFILIATIONS: From the Sleep Laboratory (Dr Martinot), CHU UCL
Copyright Ó 2018 American College of Chest Physicians. Published by
Namur Site Sainte-Elisabeth, Belgium; Research and Development Elsevier Inc. All rights reserved.
(Dr Le-Dong; and Messrs Crespeigne, Cuthbert, and Denison),
DOI: https://doi.org/10.1016/j.chest.2018.08.1027
RespiSom, Erpent, Belgium; HP2 INSERM U1042 (Drs Borel and
chestjournal.org 1
Oral dental appliances including mandibular prevention of mouth opening have been poorly studied
advancement splints (oral appliance therapy [OAT]) are under OAT.6-9 Indeed, mouth opening leads to a
a recommended treatment for mild to moderate OSA posterior displacement of the mandible, which decreases
syndrome and CPAP intolerance or patient the upper airway patency that might impact OAT
nonacceptance.1 Apnea-hypopnea index (AHI) efficacy.10
reduction is generally lower with oral appliances, but
The amplitude of vertical mandibular movements
adherence to treatment is usually higher compared with
(MMs) during sleep reflects the overall amount of
CPAP.2 This may explain the same magnitude of
respiratory effort (RE) as validated against diaphragm
therapeutic benefit regarding symptoms, quality of life,
EMG activity.11 MM has been shown to provide
and blood pressure control for OAT compared with
reliable indices characterizing RE during sleep-
CPAP.1,3
disordered breathing in adults and in children (MM-
Mandibular protrusion enlarges the pharyngeal lumen REI).12-15
and stabilizes the upper airway.4,5 Beyond the increase in
upper airway size, the precise mechanisms of action of The main objective was to document treatment response
oral appliances are not fully understood. The to OAT in OSA by using MM-REI in comparison with
modulation of upper airway muscle activity, the classical indices of OSA severity (ie, AHI and oxygen
influence of the degree of bite opening, and the desaturation index [ODI]).
Mandibular Movements: MMs were recorded both during the Procedure of Mandibular Advancement Device Fabrication and
diagnostic PSG and control PG with a midsagittal MM magnetic Titration: After dental and temporomandibular examination to
sensor (Brizzy; Nomics), which measures the distance between two exclude any contraindications, impressions of both dental arches were
parallel, coupled, resonant circuits placed on the forehead and on the taken with bite recording. A commercial dental laboratory fabricated
chin with a resolution of 0.05 mm.16 At the beginning of the night, the appliance, following the Herbst model with the capability of
adding rubber bands between the arches. Once fitted, two time periods Statistical Analysis
were programmed: (1) weeks 0-4 for acclimatization and (2) weeks 4-12 The study data included the following measurements: AHI, ODI,
for incremental advancement, that is, mandibular advancement of obstructive hypopnea index (OHI), MM-REI, TST (in minutes), and
1 mm at a time. the durations of MM-derived respiratory events (standardized as
% TST). Each index was determined at two time points: before the
The advancement of the mandible was performed in accordance with
application of OAT (baseline level) and at the final point of titration.
the recommendations of the American Academy of Dental Sleep
Medicine until absence of snoring as reported by the sleep partner Data analysis was performed with R statistical programing language.25
or the maximum comfortable limit of advancement was As most of the data were not normally distributed, baseline values were
achieved.22-24 Snoring was evaluated on a 10-cm visual analog scale summarized as median and interquartile range. First, treatment effects
(VAS) (0, no significant remaining snoring; 10, persistent loud on AHI, ODI, MM-REI, and TST were explored using violin plots and
snoring). nonparametric Wilcoxon signed-rank test. The changing tendencies
for the 12 standardized indices were aggregated in a radar plot.
The subjects completed a questionnaire to measure her/his self-
perceived quality of treatment (by quoting three items on a VAS Bayesian multilevel regression26 was used to estimate the OAT effect
[comfort, efficiency, and global satisfaction]) with the oral size. Further information on the regression analysis could be found
appliance.22 Residual daytime sleepiness evaluated by the Epworth in the online article.
Sleepiness Scale.
The criteria for treatment success were either a relative index reduction
Finally, data on subjective compliance with the OAT (number of hours below –50% from baseline and/or an absolute AHI score below 5/h.10
per night and nights per week) were collected via a diary completed in The statistical significance thresholds were fixed at 30 for Bayes factor27
the last week of the titration period, just before the control PG. and 10–6 for traditional null hypothesis testing (P value).
chestjournal.org 3
TABLE 1 ] Baseline Characteristics of the 56 Subjects determination [R2] were 0.65, 0.79, 0.85, and 0.67 for
Included in the Final Analysis MM-REI, AHI, ODI, and OHI, respectively). Overall,
Parameter Median IQR treatment benefits (mean % reduction from baseline)
Age, y 47.15 15.34 were as follows: AHI, –61.5% (95% CI, –48.9 to –71.1);
BMI, kg/m2 26.65 3.40 ODI, –66.1% (95% CI, –49.5 to –77.2); with OHI and
ESS score (0-24) 8.00 13.30 MM-REI showing the largest responses, with a mean
TST, min 444.50 101.25 reduction of –78.99% (95% CI, –63.4 to –88.2) for MM-
ODI, No./h 9.35 19.47 REI and –88.0% (95% CI, –69.5 to –96.3) for OHI.
AHI, No./h 17.06 19.33
Posterior distributions of these effects are presented in
MAI, No./h 25.00 16.92
Figure 3.
RERA, No./h 3.70 3.85 OAT Effects on MM-REI and AHI Subcomponents
MM-REI, No./h 8.14 8.90
Compartmentalized analysis was performed to
Hypopneas, No. 100 126.25
understand OAT partial effects on the constituent
AHI ¼ apnea-hypopnea index; ESS ¼ Epworth Sleepiness Scale; IQR ¼ components of MM-REI and AHI. Medians of
interquartile range (25%-75%); MAI ¼ microarousal index; MM-REI ¼ standardized variables measured at baseline and at home
mandibular movement-based respiratory effort index; ODI ¼ oxygen
desaturation index; RERA ¼ respiratory effort-related arousals; TST ¼ total control were integrated in a radar chart (Fig 4).
sleep time.
Visual comparisons indicate that the effect sizes of OAT
were comparable among AHI, ODI, MM-REI, and
score was 8.0 (95% CI, 2.0-13.3). The main duration of MM-derived respiratory events. The
characteristics of the nine excluded subjects (n ¼ 9) were treatment effect on MM-REI could be explained by
not different from the analyzed cohort (see e-Table 1). significant reductions in both MMS (P ¼ .0004) and
The mean duration between PSG (treatment initiation) sequence of MML with a peak-to-peak displacement of
and PG (OAT efficacy evaluation after titration) was 18 at least 0.3 mm followed by a MMS concomitant with
(4–36) weeks, and the median percentage of the cortical arousal (MMAS) event rates (P < 10–6). The
maximal active mandibular protrusion of the final treatment effect on AHI subcomponents is more
mandibular advancement was 75% (63.6-90.9). Mean complex, as responses were positive (P < 10–8) only for
adherence to OAT during the week preceding the hypopneas (total or obstructive), while neither
control of snoring at home was 5.5 (4.35-7) h/night and obstructive nor central apneas changed under OAT
7 (5-7) days. At the time of the home PG, the median from baseline (e-Table 2).
(95% CI) improvement in snoring on a 0-10 VAS was,
respectively, –8.0 (range, –6 to –10) and the mean global Discussion
satisfaction with treatment was 7 (4-10). Our study demonstrated for the first time that MM
monitoring represents a powerful tool for assessing the
Efficacy of OAT efficacy of OAT. The OAT success rate and range of
Figure 2 shows the distribution of AHI, OHI, ODI, MM- improvement were comparable in our study population
REI, and MM-measured TST at baseline and at the end to results previously reported in larger prospectively or
of OAT titration. retrospectively studied cohorts with similar levels of
OSA severity.29-31 Two-thirds of the subjects showed a
Individual responses are presented in the online article
reduction in AHI and ODI of at least 50% (e-Fig 1), but
(e-Fig 1). For nine of the 56 subjects, AHI and MM-REI
only 20% reached values of AHI < 5 when snoring
showed different patterns of evolution. This was related
disappeared.
to the emergence of central sleep apnea under OAT in
seven subjects (e-Fig 2), and by residual elevated MM- As demonstrated by the posterior distribution of the
REI despite decreased AHI in two subjects (e-Fig 1). OAT effects on the four OSA severity indices (Fig 3),
every metric (MM-REI, AHI, ODI, or OHI) informs
OAT Effects on AHI, ODI, OHI, and MM-REI about treatment success in the study population and
Four separate models were built for AHI, ODI, OHI, and supports complementary information.32 The ratios of
MM-REI to estimate their relative changes (%) from the likelihood probability in favor of a change below
baseline values. All four models successfully converged –50% baseline level vs the null hypothesis were 53.8 for
and fitted the data reasonably well (coefficients of AHI, 67.9 for ODI, 4,000 for MM-REI, and infinitely
Post
Baseline
0 30 60 90
OHI (n/h)
Post
Baseline
0 5 10 15 20
ODI (n/h)
Time point
Post
Baseline
Baseline Post
0 30 60 90
MM-REI (n/h)
Post
Baseline
0 25 50 75 100
TST (min)
Post
Baseline
Figure 2 – Distribution of the outcomes measured at baseline and at the final point of OAT titration. Combined violin plots (continuous trace) and
boxplots (dotted line) represent the distribution shape, median, third and fourth quartiles, minimum and maximum values of AHI (number per hour),
OHI (number per hour), ODI (number per hour), and MM-REI (number per hour), and MM-based TST (minutes) at baseline (red) vs end of OAT
titration (blue). All variables except TST were significantly different (P < 10–7). AHI ¼ apnea-hypopnea index; MM ¼ mandibular movement; MM-
REI ¼ respiratory effort index derived by MM characterization; OAT ¼ oral appliance therapy; ODI ¼ oxygen desaturation index; OHI ¼ obstructive
hypopnea index; TST ¼ total sleep time.
large for OHI, indicating a high plausibility of these proportion of apneas presumably reflect higher
effects. collapsibility of the upper airway, which might justify
primary CPAP application.
Effects of OAT on Obstructive Apneas
vs Obstructive Hypopneas Discordance in AHI and MM-REI for Assessment of
A weaker effect of OAT was observed on obstructive Response to OAT
apneas (OAs) compared with hypopneas. Although Seven subjects showed persistent high AHI with a significant
the vast majority of OA events occurred in seven reduction in MM-REI (e-Fig 1). This was due to treatment-
subjects, the Bayesian-modeled response to OAT did emergent central apneas (n ¼ 5) whereas RE was
not show individual effect for explaining less activity dramatically decreased and abolished MM.33 In two cases,
of OAT on OA vs hypopneas. Indeed, a high central hypopneas were scored where the decrease in flow
chestjournal.org 5
TST
MM-RE
MM-REI duration
OHI
Desat. Obs.HYP
Hyp index index
Outcomes
MM-REI
Central Total
Apnea index Obs.Apnea HYP
ODI index
Baseline At titration point
AHI
Figure 4 – General tendencies of changes in the main indices and their
constituent components under oral appliance therapy. The main out-
–100 –90 –80 –70 –60 –50 –40 –30 comes (ODI, AHI, MM-REI, and MM-RE duration [duration of MM-
Incidence rate change (% of baseline level) derived respiratory events]) and their constituent components (TST,
MMAS or MMS, Total HYP [total number of hypopneas], Desat.Hyp
Figure 3 – Posterior distribution of the incidence rate change in index [index of hypopneas with oxygen desaturation] and Obs.HYP
obstructive hypopnea events, MM-REI, AHI, and ODI in response to index [index of obstructive hypopneas] were standardized and then
oral appliance therapy (OAT). Shown is the posterior distribution of the stratified by the time point factor (baseline vs titration point) on a radar
incidence rate change (% of baseline) of four outcomes (AHI, ODI, MM- plot. Each radial axis represents the same z-score scale of [–0.7 to 0.2]
REI, and OHI, determined as 100 [exp(fixed effect) – 1]; where fixed from center to periphery. Except for TST, obstructive apnea index;
effect measures the decrease in logarithmic scale of the target outcomes central apnea index, all the components were reduced by treatment. See
under OAT. The vertical dashed line locates the threshold for treatment Figure 1 and 2 legends for expansion of other abbreviations.
success or a reduction greater than 50%. In all four outcomes, the highest
density interval (97.5%) of response falls below –50%, indicating a sig-
nificant response of those indices to OAT. See Figure 2 legend for
expansion of other abbreviations.
the subjects met the traditional criteria for successful
treatment (a 50% reduction in AHI), although not all
was due mainly to a decrease in the central drive rather than patients normalized AHI below 5 (e-Fig 1).
pharyngeal obstruction.
Of note, OAT nocturnal compliance was self-reported
Two subjects exhibited decreased AHI but no change in by subjects in a mailed questionnaire. No thermometric
MM-REI because of persistent MM during long periods sensor was added to the appliance to record compliance
of flow limitation and remaining RE.14 This confirms the more objectively. However, the efficacy of the appliance
ability of MM to identify increased RE, but one was verified by questioning the sleep partner for audible
limitation of the technique might be to underscore snoring before programming home PG control.
persistent central apneas or hypopneas when used
without a concomitant airflow signal. However, this Study Limitations
situation could be suspected by the analysis of recorded
The end point for treatment evaluation was not met and
MM. A decrease in duration of MM-derived respiratory
further advancement (possible in four of these nine
events associated with no improvement or an increase in
patients) was not well tolerated. However, the addition of
MMS signal, which reflects persistent sleep
rubber bands at this point completely normalized snoring
fragmentation, is highly suggestive of treatment-
in six of these nine patients. We initially decided to exclude
emergent central apneas (e-Fig 2). The emergence of
the patients requiring rubber bands for snoring control
central apneas in the context of oral appliance treatment
from the analysis in order to prevent any confounding
is underrecognized and has been described mainly in
from mechanical interference with the mandibular
case reports.34-36 We suspect the role of excessive loop
movement assessment. Indeed, previous observations
gain37,38 in these patients, which could be a predictor of
showed us that peak-to-peak MM amplitudes are
poor response to OAT.39
dampened in the presence of rubber bands, especially
There may be a place for mandibular advancement in when the number of these bands had to be doubled. The
some subjects beyond snoring control used as the end impact of rubber bands on the MM signal needs to be
point for titration in this study. Of note at this point, all further investigated (e-Figs 3-5, e-Table 3).
Acknowledgments obstructive sleep apnoea. Eur Respir J. sleep apnea. J Clin Sleep Med. 2015;11(5):
2010;35:836-842. 567-574.
Author contributions: J. B. M. is the
guarantor of the article. E. C., J. B. M., J. C. B., 5. Chan ASL, Sutherland K, Schwab R, 14. Martinot JB, Le-Dong NN, Denison S,
et al. The effect of mandibular et al. Persistent respiratory effort after
and J. L. P. designed the research; E. C., J. B.
advancement on upper airway structure adenotonsillectomy in children with sleep-
M., V. C., S. D., and J. C. B. performed data in obstructive sleep apnoea. Thorax. disordered breathing. Laryngoscope.
acquisition and interpretation; N. N. L.-D. 2010;65:726-732. 2018;128(5):1230-1237.
analyzed data; E. C. and J. B. M. validated
and interpreted the results; N. N. L.-D., E. C., 6. Johal A, Gill G, Ferman A, McLaughlin K. 15. Gray EL, Barnes DJ. Beyond the
and J. B. M. wrote the article; and P. E. S., J. The effect of mandibular advancement thermistor: novel technology for the
C. B., and J. L. P. revised the manuscript. The appliances on awake upper airway and ambulatory diagnosis of obstructive sleep
masticatory muscle activity in patients apnoea. Respirology. 2017;22(3):418-419.
final manuscript has been revised and
with obstructive sleep apnoea. Clin Physiol
approved by all authors. Funct Imaging. 2007;27(1):47-53. 16. Senny F, Destiné J, Poirrier R. Midsagittal
Financial/nonfinancial disclosures: None jaw movements analysis for the scoring of
7. Pitsis AJ, Darendeliler MA, sleep apneas and hypopneas. IEEE Trans
declared. Gotsopoulos H, Petocz P, Cistulli PA. Biomed Eng. 2008;55(1):87-95.
Effect of vertical dimension on efficacy of
Role of sponsors: The sponsor had no role in 17. Cheliout-Heraut F, Senny F, Djouadi F,
oral appliance therapy in obstructive sleep
the design of the study, the collection and apnea. Am J Respir Crit Care Med. Ouayoun M, Bour F. Obstructive sleep
analysis of the data, or the preparation of the 2002;166(6):860-864. apnoea syndrome: comparison between
manuscript. polysomnography and portable sleep
8. Ma SYL, Whittle T, Descallar J, et al. monitoring based on jaw recordings.
Other contributions: The authors thank Association between resting jaw muscle Neurophysiol Clin. 2011;41(4):191-198.
Elodie Ducarme and Liesbeth Orij, who electromyographic activity and
kindly spent innumerable hours providing mandibular advancement splint outcome 18. Senny F, Destiné J, Poirrier R. Midsagittal
secretarial assistance. in patients with obstructive sleep apnea. jaw movements as a sleep/wake marker.
Am J Orthod Dentofacial Orthop. IEEE Trans Biomed Eng. 2009;56(2):303-
Additional information: The e-Figures and 2013;144(3):357-367. 309.
e-Tables can be found in the Supplemental
Materials section of the online article. 9. Kurtulmus H, Cotert S, Bilgen C, On AY, 19. Maury G, Senny F, Cambron L, Albert A,
Boyacioglu H. The effect of a mandibular Seidel L, Poirrier R. Mandible behaviour
advancement splint on electromyographic interpretation during wakefulness, sleep
activity of the submental and masseter in and sleep-disordered breathing. J Sleep
References patients with obstructive sleep apnea. Int J Res. 2014;23(6):709-716.
1. de Almeida FR, Cistulli P, Fleury B, Prosthodont. 2009;22(6):586-593.
20. Chakar B, Senny F, Poirrier A-L,
Gagnadoux F. Mandibular advancement 10. Meurice J-C, Marc I, Carrier G, Sériès F. Cambron L, Fanielle J, Poirrier R.
devices. Eur Respir Monogr. 2015;67: Effects of mouth opening on upper airway Validation of midsagittal jaw movements
253-265. collapsibility in normal sleeping subjects. to measure sleep in healthy adults by
2. Almeida FR, Hamoda M, Kohzuka Y. Oral Am J Respir Crit Care Med. 1996;153(1): comparison with actigraphy and
appliances for the management of OSA: 255-259. polysomnography. Sleep Sci. 2017;10(3):
an updated review of the literature. Chest. 11. Martinot JB, Le Dong NN, Cuthbert V, 122-127.
2018;153(2):544-553. et al. Mandibular movements as 21. Berry RB, Budhiraja R, Gottlieb DJ, et al.
3. Phillips CL, Grunstein RR, accurate reporters of respiratory effort Rules for scoring respiratory events in
Darendeliler MA, et al. Health outcomes during sleep: validation against sleep: update of the 2007 AASM Manual
of continuous positive airway pressure diaphragmatic electromyography. Front for the Scoring of Sleep and Associated
versus oral appliance treatment for Neurol. 2017;8:353. Events. J Clin Sleep Med. 2012;8(5):597-
obstructive sleep apnea: a randomized 12. Martinot JB, Borel JC, Cuthbert V, et al. 619.
controlled trial. Am J Respir Crit Care Mandibular position and movements: 22. Fleury B, Rakotonanahary D, Petelle B,
Med. 2013;187(8):879-887. suitability for diagnosis of sleep apnoea. et al. Mandibular advancement titration
4. Chan ASL, Lee RWW, Srinivasan VK, Respirology. 2017;22(3):567-574. for obstructive sleep apnea: optimization
Darendeliler MA, Grunstein RR, 13. Martinot JB, Senny F, Denison S, et al. of the procedure by combining clinical
Cistulli PA. Nasopharyngoscopic Mandibular movements identify and oximetric parameters. Chest.
evaluation of oral appliance therapy for respiratory effort in pediatric obstructive 2004;125(5):1761-1767.
chestjournal.org 7
23. Ramar K, Dort LC, Katz SG, et al. Clinical appliance therapy in patients with daytime 35. Kuzniar TJ, Kovacevic-Ristanovic R,
practice guideline for the treatment of sleepiness and snoring or mild to Freedom T. Complex sleep apnea
obstructive sleep apnea and snoring with moderate sleep apnea: a randomized unmasked by the use of a mandibular
oral appliance therapy: an update for clinical trial. JAMA Intern Med. advancement device. Sleep Breath.
2015. J Clin Sleep Med. 2015;11(7):773- 2015;175(8):1278-1285. 2011;15(2):249-252.
827.
30. Sutherland K, Takaya H, Qian J, Petocz P, 36. Gindre L, Gagnadoux F, Meslier N,
24. Scherr SC, Dort LC, Almeida FR, et al. Ng AT, Cistulli PA. Oral appliance Fleury B, Gustin JM, Racineux JL. Central
Definition of an effective oral appliance treatment response and apnea developing during treatment with a
for the treatment of obstructive sleep polysomnographic phenotypes of mandibular advancement device. Rev Mal
apnea and snoring: a report of the obstructive sleep apnea. J Clin Sleep Med. Respir. 2006;23(5 pt 1):477-480.
American Academy of Dental Sleep 2015;11(8):861-868.
Medicine. J Dental Sleep Med. 2014;1(1): 37. Randerath W, Verbraecken J, Andreas S,
51. https://aadsm.org/docs/ 31. Vecchierini MF, Attali V, Collet JM, et al. Definition, discrimination, diagnosis
definitionoforalappliance.pdf. et al. A custom-made mandibular and treatment of central breathing
repositioning device for obstructive sleep disturbances during sleep. Eur Respir J.
25. R Core Team. A Language and apnoea-hypopnoea syndrome: the 2017;49(1). pii: 1600959.
Environment for Statistical Computing. ORCADES study. Sleep Med. 2016;19:
Vienna, Austria: R Foundation for 38. Sands SA, Bradley AE, Terrill PI, et al.
131-140.
Statistical Computing; 2013. https://www. Phenotyping pharyngeal
r-project.org/. 32. Sharples LD, Clutterbuck-James AL, pathophysiology using
Glover MJ, et al. Meta-analysis of polysomnography in patients with
26. Bürkner PC. brms: an R package for randomised controlled trials of oral obstructive sleep apnea. Am J Respir
Bayesian multilevel models using Stan. mandibular advancement devices and Crit Care Med. 2018;197(9):1187-1197.
J Stat Software. 2017;80(1):1-28. https:// continuous positive airway pressure for
cran.r-project.org/web/packages/brms/ 39. Bradley AE, Andara C, Landry S, et al.
obstructive sleep apnoea-hypopnoea. Sleep
index.html. Upper-airway collapsibility and loop gain
Med Rev. 2015;27:108-124.
predict the response to oral appliance
27. Kass RE, Raftery AE. Bayes factors. J Am
33. Mohan A, Henderson J, Mador MJ. therapy in patients with obstructive sleep
Stat Assoc. 1995;90(430):773-795.
Mandibular advancement device- apnea. Am J Respir Crit Care Med.
28. American Academy of Sleep Medicine. emergent central sleep apnea can resolve 2016;194(11):1413-1422.
International Classification of Sleep spontaneously: a case report. J Clin Sleep
Disorders. 3rd ed. Darien, IL: American 40. Vandenbussche NL, Overeem S, van
Med. 2016;12(1):137-138.
Academy of Sleep Medicine; 2014;146(5): Dijk JP, Simons PJ, Pevernagie DA.
1387-1394. 34. Avidan AY, Guilleminault C, Robinson A. Assessment of respiratory effort during
The development of central sleep apnea sleep: esophageal pressure versus
29. Marklund M, Carlberg B, Forsgren L, with an oral appliance. Sleep Med. noninvasive monitoring techniques. Sleep
Olsson T, Stenlund H, Franklin KA. Oral 2006;7(2):187-191. Med Rev. 2015;24:28-36.