Objective: 

To evaluate the influence of craniofacial morphology on the upper airway dimensions in healthy adult subjects.

Materials and Methods: 

The records of 276 healthy 17- to 27-year-old patients were extracted from the cone-beam computed tomography image database of the Institute of Stomatology, Riga Stradins University. Dolphin 11.7 software was used to evaluate craniofacial anatomy and semiautomatic segmentation of the upper airway. Measurements of oropharyngeal airway volume (OPV), minimal cross-sectional area (CSAmin), and nasopharyngeal airway volume (NPV) were obtained. The presence of adenoid tissues was recorded. Associations between variables were analyzed by Spearman's correlation coefficients, and multivariate linear regression analysis was used to identify factors that had a possible influence on upper airway dimensions.

Results: 

The following factors were identified as influencing the variability of NPV (23%): SNA angle, gender, and presence of adenoids. Statistically significant, although weak, correlations were found between SNB angle and OPV (r  =  0.144, P < .05) and CSAmin (r  =  0.182, P < .01).

Conclusion: 

The results suggest that craniofacial morphology alone does not have a significant influence on upper airway dimensions.

Respiratory function and upper airway morphology are greatly relevant to orthodontic diagnosis and treatment planning as altered breathing function could influence facial growth and morphology.1  More importantly, breathing disturbances could lead to increased morbidity and mortality in a condition like obstructive sleep apnea (OSA).2  The posterior airway space (PAS) of patients with OSA has been shown to be smaller than that of healthy persons.3  Evidence suggests that one of the reasons for this could be related to craniofacial morphology features, such as the retro position of the mandible and increased upper and/or lower face heights.4  Because a close relationship is found between the pharyngeal airway patency and craniofacial structures in patients with OSA,5  an association could be expected to exist between the upper airway dimensions and the craniofacial pattern.

Although many studies have reported changes in the dimensions of the upper airway after surgical repositioning of the mandible and the maxilla, estimates about the changes in the upper airway after orthognathic surgery operations remain controversial.6  Despite a few case reports in which mandibular setback surgery in patients with skeletal Class III malocclusion induced OSA associated with airway narrowing,7,8  prospective studies9,10  have failed to demonstrate disturbances of respiration during sleep after mandibular setback even though retro palatal airway size was reduced. These disagreements might be explained by the suggestion that preoperative airway size in patients with Class III malocclusion is larger than that of the clinically normal population.10,11 

Several studies have used lateral cephalometry (LC) to explore whether upper airway size is associated with a specific craniofacial pattern.1220  Although some of these studies have representative samples and a sophisticated design, LC is limited by its two-dimensional nature. Since the advent of cone-beam computed tomography (CBCT), more accurate evaluation of the airway has become possible, giving more complete information than LC21,22  with a significantly reduced radiation dose compared with conventional computed tomography.23  Nevertheless, three-dimensional (3D) studies that have aimed to find pharyngeal size differences among various skeletal patterns have provided inconsistent results.2431  Therefore, the aim of the present study was to investigate whether upper airway dimensions in healthy adult subjects are influenced by craniofacial morphology.

In this retrospective study, the CBCT image database collected at the Institute of Stomatology, Riga Stradins University, from December 2008 to December 2012 was inspected. The study was reviewed and approved by the Ethics Committee of Riga Stradins University, Riga, Latvia. All CBCT scans were taken with an ICAT Scanner (Imaging Sciences International, Hatfield, Pa) as part of the initial diagnostic records for the patients according to orthodontic, prosthetic, or surgical indications. The CBCT images were taken according to a standard protocol with the subject seated in a chair with the following parameters: 120 kV, 5 mA, 0.4 mm voxel, and scan time of 20 seconds.

The following criteria were formulated to select the images for inclusion in this investigation: scans of 17- to 27-year-old patients, full field of view images (13 cm × 17 cm) that allowed clear visualization of the craniofacial structures from the cranial base to the upper border of the epiglottis, and images with the patients' teeth in the central occlusion. The inclusion criteria were met by 396 scans. Full records of these subjects were inspected and the subsequent exclusion criteria were applied: medically compromised patients (OSA, syndromes and arthritis), previous orthognathic surgery, patients with facial clefts and severe craniofacial asymmetries, and subjects with altered craniocervical inclination. As the measurements of airway size are influenced by head posture,32  craniocervical inclinations of all subjects were examined to ensure that they were between 90° and 110°. After applying the exclusion criteria, 276 patients (100 men and 176 women) remained and were further investigated.

The Dolphin 11.7 software (Patterson Dental Supply Inc, Chatsworth, Calif) was used to analyze CBCT images. Conventional hard-tissue landmarks were identified for the cephalometric analysis, and six angular measurements were collected (SNA, SNB, ANB, MP-SN, MM, FMA) for the craniofacial relationships (Table 1). The same software was used to obtain the airway measurements. Once the CBCT data sets were imported in the format of single-file DICOM (Digital Imaging and Communications in Medicine) files, the reconstructions were reoriented as close as possible to the guidelines suggested by Haskel et al.33  The airway analysis tool was used to obtain the desired portions of the nasopharyngeal (NP) and the oropharyngeal (OP) airway. The area of interest was defined by a clipping box and seeds in the airway space. The limits for segmentation of the upper airway and examples of the virtual surface models are presented in Figure 1. Measurements of the oropharyngeal airway volume (OPV), minimal cross-sectional area (CSAmin), and nasopharyngeal airway volume (NPV) were calculated by the software. The presence of adenoid tissues was recorded. All data were collected by one operator.

Figure 1.

Limits for upper airway segmentation and examples of virtual surface models. Oropharyngeal airway: superior, the edge of the soft palate to the posterior of the pharynx (parallel to FH); inferior, the plane from the tip of the epiglottis to the posterior of the pharynx (parallel to FH); lateral, anterior and posterior, the internal walls of the pharynx, including the full extensions of the lateral projections, and limited by the posterior surfaces of the tongue and the soft palate. Nasopharyngeal airway: superior, the highest point of the nasopharynx, coinciding and consistent with the anterior limit; inferior, the superior border of the oropharyngeal airway; anterior, a coronal plane through the posterior nasal spine perpendicular to the sagittal plane at the lowest border of the vomer; posterior, the posterior wall of the pharynx; lateral, the lateral walls of the pharynx, including the full extensions of the lateral projections.

Figure 1.

Limits for upper airway segmentation and examples of virtual surface models. Oropharyngeal airway: superior, the edge of the soft palate to the posterior of the pharynx (parallel to FH); inferior, the plane from the tip of the epiglottis to the posterior of the pharynx (parallel to FH); lateral, anterior and posterior, the internal walls of the pharynx, including the full extensions of the lateral projections, and limited by the posterior surfaces of the tongue and the soft palate. Nasopharyngeal airway: superior, the highest point of the nasopharynx, coinciding and consistent with the anterior limit; inferior, the superior border of the oropharyngeal airway; anterior, a coronal plane through the posterior nasal spine perpendicular to the sagittal plane at the lowest border of the vomer; posterior, the posterior wall of the pharynx; lateral, the lateral walls of the pharynx, including the full extensions of the lateral projections.

Close modal
Table 1.

Cephalometric Measurements

Cephalometric Measurements
Cephalometric Measurements

Statistical analyses were performed with SPSS 20.0 for Windows (SPSS, Chicago, Ill). The Kolmogorov-Smirnov test was used to check the normality of distribution. For analysis of method error all measurements were repeated for 30 randomly selected subjects by the same operator on two separate occasions at least 2 weeks apart. The intraclass correlation coefficient (ICC) for the volumes and CSA and the Dahlberg formula for the angular measurements were calculated. The statistical significance was set at 0.05. Associations between the variables were analyzed by Spearman correlation coefficients; multivariate linear regression analysis was used to identify the factors that possibly influence upper airway dimensions. The following nine variables were analyzed as possible predictors for pharyngeal airway size: gender, ANB, SNA, SNB, MM, MP-SN, FMA, sum of angles SNA and SNB, and presence of adenoids.

The method error ranged from 0.45° to 0.79° for angular measurements, and ICC values for airway variables varied from 0.97 to 0.99. Demographic and cephalometric characteristics of the sample are summarized in Table 2. This sample had a diversity of craniofacial patterns as revealed by the range of values of cephalometric variables. Of the 276 subjects, 48 (17%) had increased nasopharyngeal adenoid tissues of various degrees.

Table 2.

Descriptive Statistics of the Sample

Descriptive Statistics of the Sample
Descriptive Statistics of the Sample

Bivariate correlations between the variables and the airway measurements are given in Table 3. The NPV correlated negatively with gender, presence of adenoid tissues, and MP-SN and SNB angles; in addition, there were positive correlations with SNA angle and the sum of angles SNA and SNB. Both OPV and CSAmin correlated significantly with SNB angle, sum of angles SNA and SNB, and presence of adenoids. Additionally, OPV had a significant negative correlation with gender, and CSAmin correlated negatively with ANB angle.

Table 3.

Correlations Between Cephalometric and Anatomic Variables and Volume and Cross-sectional Area Measurementsa

Correlations Between Cephalometric and Anatomic Variables and Volume and Cross-sectional Area Measurementsa
Correlations Between Cephalometric and Anatomic Variables and Volume and Cross-sectional Area Measurementsa

Of the nine variables included in the regression analysis, five were significantly associated with NP volume, four with OP volume, and four with minimal CSA. Because of the strong inter-correlations, one variable for the volumes and CSAmin were excluded from the multivariate regression analysis. Thus, the remaining four variables were included in the regression model for NPV (gender, MP-SN, SNA, presence of adenoids), and three were included for OPV (gender, SNB, presence of the adenoids). In the final model, gender, SNA angle, and presence of adenoids explained 23% of the variation in NPV. Only 11% of the OP volume variation was explained by gender, SNB angle, and presence of adenoids. It was possible to explain only 6% of CSAmin variation by two variables (SNB and presence of adenoids).

To our knowledge the present 3D study on upper airway dimensions and craniofacial morphology is the most complete in terms of the number of subjects included. However, the main weakness of the study is its retrospective nature. The records of the patients were included in the investigation based on availability, and important information, such as data on body mass index, smoking status, and assessment of the soft tissues and neuromotor factors, is missing, although these could have a significant influence on upper airway size.34,35  The influence of aging on upper airway size is well recognized36 ; however, it was assumed that between the ages of 17 and 27 only minor changes would have taken place; therefore, this age range was set as an inclusion criterion.

Several studies using two-dimensional and 3D imaging techniques have tried to find relationships between pharyngeal dimensions and craniofacial morphology.1220,2431  However, the inconsistencies of the findings are found among lateral cephalometric studies and 3D investigations (Table 4). The inconsistencies could be explained by variations in sites of the airway measurements and by the diversity of the study samples. Most of the previous investigations used vertical discrepancies as an exclusion criteria,12,15,24,2931  and in other studies vertical skeletal relationships were not evaluated.17,25,27  Nevertheless, the influence of vertical morphology should not be underestimated. If only the ANB angle is used to measure the relative position of the maxilla and the mandible to each other, the location of points A and B in the vertical plane can influence the value of the angle.37  Therefore, misleading conclusions could be made without incorporating the vertical growth type in the evaluation of the upper airway.

Table 4.

Summarized Data from Previous Studies on Upper Airway Dimensions and Craniofacial Morphology

Summarized Data from Previous Studies on Upper Airway Dimensions and Craniofacial Morphology
Summarized Data from Previous Studies on Upper Airway Dimensions and Craniofacial Morphology
Table 4.

Continued

Continued
Continued

Several 3D investigations have not found any association between the volume of the nasopharynx and the horizontal position of the maxilla24,26,27 ; however, in some studies significant differences in the nasopharyngeal volume among sagittal malocclusions were reported.28,29  We found a weak correlation between the value of SNA and the nasopharyngeal volume, although the presence of increased adenoid tissues had a more significant influence. The nasopharyngeal adenoid tissues, which are often present during the childhood, usually spontaneously atrophy by puberty. We found enlarged adenoids in 17% of the patients in our sample. This is an unexpected finding, but Wolford et al.38  also pointed out the presence of adenoids in many patients undergoing orthognathic surgery.

Retrognathic position of the mandible has been frequently linked with the narrowing of the upper airway in patients with obstructive sleep apnea (OSA).4,5  Similarly, several studies have found smaller oropharyngeal volumes for subjects with Class II malocclusion compared with subjects with Class I26,2931  or Class III26,29,31  malocclusion. In our study, only a very weak influence of craniofacial morphology was found on the oropharyngeal dimension. This is in agreement with other studies that have reported that OPV24,25,27,28  or oropharyngeal CSA24  do not differ among persons with various skeletal patterns, and either weak or insignificant correlations were reported between craniofacial variables and oropharyngeal airway dimensions.14,15,2629 

Since the effective dose of CBCT is substantial, the use of this method as a part of the diagnostic records should be fully justified. The ideal way of investigating the upper airways is magnetic resonance imagining (MRI), as it allows the pharyngeal soft tissues to be examined without any radiation exposure39 ; however, because of its affordability, CBCT has overtaken MRI in dentistry. Although CBCT does not show clear delineations between soft tissues, it clearly demonstrates the airway space and related skeletal structures and has been shown to provide precise and clinically relevant information on upper airway dimensions.40  We used the Dolphin software for airway segmentation, which has been found to be accurate and reliable.41 

The vastly different findings of previous investigations using CBCT scans2531  suggest that the associations between the upper airways and craniofacial morphology are ambiguous. The representativeness of our study sample suggests that the isolated influence of craniofacial morphology on the upper airway dimensions is minor, and clear differences among craniofacial patterns in pharyngeal measurements could only be found in selected extremities. Therefore, orthodontic treatment planning per se should not be an indication for extensive upper airway investigation with CBCT. However, for borderline patients, if clinical investigation has given a reason for it, CBCT investigation could provide additional and valuable information to assess the precise site of obstruction and could justify different treatment options. These reasons include orthodontic patients suffering from OSA and evaluation of the need for maxillomandibular advancement surgery42  or for orthognathic surgery involving mandibular setback.6 

  • Craniofacial morphology has a minor influence on the dimensions of the upper airways.

1.
Solow
B
,
Siersbaek-Nielsen
S
,
Greve
E
.
Airway adequacy, head posture, and craniofacial morphology
.
Am J Orthod
.
1984
;
86
:
214
223
.
2.
Banno
K
,
Kryger
MH
.
Sleep apnea: clinical investigations in humans
.
Sleep Med
.
2007
;
8
:
400
426
.
3.
Enciso
R
,
Nguyen
M
,
Shigeta
Y
,
Ogawa
T
,
Clark
GT
.
Comparison of cone-beam CT parameters and sleep questionnaires in sleep apnea patients and control subjects
.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
.
2010
;
109
:
285
293
.
4.
Tangugsorn
V
,
Krogstad
O
,
Espeland
L
,
Lyberg
T
.
Obstructive sleep apnoea: multiple comparisons of cephalometric variables of obese and non-obese patients
.
J Craniomaxillofac Surg
.
2000
;
28
:
204
212
.
5.
Lowe
AA
,
Fleetham
JA
,
Adachi
S
,
Ryan
CF
.
Cephalometric and computed tomographic predictors of obstructive sleep apnea severity
.
Am J Orthod Dentofacial Orthop
.
1995
;
107
:
589
595
.
6.
Mattos
CT
,
Vilani
GN
,
Sant'Anna
EF
,
Ruellas
AC
,
Maia
LC
.
Effects of orthognathic surgery on oropharyngeal airway: a meta-analysis
.
Int J Oral Maxillofac Surg
.
2011
;
40
:
1347
1356
.
7.
Guilleminault
C
,
Riley
R
,
Powell
N
.
Sleep apnea in normal subjects following mandibular osteotomy with retrusion
.
Chest
.
1985
;
88
:
776
778
.
8.
Liukkonen
M
,
Vahatalo
K
,
Peltomaki
T
,
Tiekso
J
,
Happonen
RP
.
Effect of mandibular setback surgery on the posterior airway size
.
Int J Adult Orthodon Orthognath Surg
.
2002
;
17
:
41
46
.
9.
Turnbull
NR
,
Battagel
JM
.
The effects of orthognathic surgery on pharyngeal airway dimensions and quality of sleep
.
J Orthod
.
2000
;
27
:
235
247
.
10.
Hochban
W
,
Schurmann
R
,
Brandenburg
U
,
Conradt
R
.
Mandibular setback for surgical correction of mandibular hyperplasia—does it provoke sleep-related breathing disorders
?
Int J Oral Maxillofac Surg
.
1996
;
25
:
333
338
.
11.
Degerliyurt
K
,
Ueki
K
,
Hashiba
Y
,
Marukawa
K
,
Nakagawa
K
,
Yamamoto
E
.
A comparative CT evaluation of pharyngeal airway changes in class III patients receiving bimaxillary surgery or mandibular setback surgery
.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
.
2008
;
10
:
495
502
.
12.
Ceylan
I
,
Oktay
H
.
A study on the pharyngeal size in different skeletal patterns
.
Am J Orthod Dentofacial Orthop
.
1995
;
10
:
69
75
.
13.
Joseph
AA
,
Elbaum
J
,
Cisneros
GJ
,
Eisig
SB
.
A cephalometric comparative study of the soft tissue airway dimensions in persons with hyperdivergent and normodivergent facial patterns
.
J Oral Maxillofac Surg
.
1998
;
56
:
135
139
.
discussion 39–40
.
14.
Trenouth
MJ
,
Timms
DJ
.
Relationship of the functional oropharynx to craniofacial morphology
.
Angle Orthod
.
1999
;
69
:
419
423
.
15.
Abu Allhaija
ES
,
Al-Khateeb
SN
.
Uvulo-glosso-pharyngeal dimensions in different anteroposterior skeletal patterns
.
Angle Orthod
.
2005
;
75
:
1012
1018
.
16.
de Freitas
MR
,
Alcazar
NM
,
Janson
G
,
de Freitas
KM
,
Henriques
JF
.
Upper and lower pharyngeal airways in subjects with Class I and Class II malocclusions and different growth patterns
.
Am J Orthod Dentofacial Orthop
.
2006
;
130
:
742
745
.
17.
Muto
T
,
Yamazaki
A
,
Takeda
S
.
A cephalometric evaluation of the pharyngeal airway space in patients with mandibular retrognathia and prognathia, and normal subjects
.
Int J Oral Maxillofac Surg
.
2008
;
37
:
228
231
.
18.
Zhong
Z
,
Tang
Z
,
Gao
X
,
Zeng
XL
.
A comparison study of upper airway among different skeletal craniofacial patterns in nonsnoring Chinese children
.
Angle Orthod
.
2010
;
80
:
267
274
.
19.
Ucar
FI
,
Uysal
T
.
Orofacial airway dimensions in subjects with Class I malocclusion and different growth patterns
.
Angle Orthod
.
2011
;
81
:
460
468
.
20.
Memon
S
,
Fida
M
,
Shaikh
A
.
Comparison of different craniofacial patterns with pharyngeal widths
.
J Coll Physicians Surg Pak
.
2012
;
22
:
302
306
.
21.
Aboudara
C
,
Nielsen
I
,
Huang
JC
,
Maki
K
,
Miller
AJ
,
Hatcher
D
.
Comparison of airway space with conventional lateral headfilms and 3-dimensional reconstruction from cone-beam computed tomography
.
Am J Orthod Dentofacial Orthop
.
2009
;
135
:
468
479
.
22.
Lenza
MG
,
Lenza
MM
,
Dalstra
M
,
Melsen
B
,
Cattaneo
PM
.
An analysis of different approaches to the assessment of upper airway morphology: a CBCT study
.
Orthod Craniofac Res
.
2010
;
13
:
96
105
.
23.
Ludlow
JB
,
Ivanovic
M
.
Comparative dosimetry of dental CBCT devices and 64-slice CT for oral and maxillofacial radiology
.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
.
2008
;
106
:
106
114
.
24.
Alves
PV
,
Zhao
L
,
O'Gara
M
,
Patel
PK
,
Bolognese
AM
.
Three-dimensional cephalometric study of upper airway space in skeletal class II and III healthy patients
.
J Craniofac Surg
.
2008
;
19
:
1497
1507
.
25.
Iwasaki
T
,
Hayasaki
H
,
Takemoto
Y
,
Kanomi
R
,
Yamasaki
Y
.
Oropharyngeal airway in children with Class III malocclusion evaluated by cone-beam computed tomography
.
Am J Orthod Dentofacial Orthop
.
2009
;
136
:
318 e1
e9
.
discussion 18–19
.
26.
Grauer
D
,
Cevidanes
LS
,
Styner
MA
,
Ackerman
JL
,
Proffit
WR
.
Pharyngeal airway volume and shape from cone-beam computed tomography: relationship to facial morphology
.
Am J Orthod Dentofacial Orthop
.
2009
;
136
:
805
814
.
27.
Oh
KM
,
Hong
JS
,
Kim
YJ
,
Cevidanes
LS
,
Park
YH
.
Three-dimensional analysis of pharyngeal airway form in children with anteroposterior facial patterns
.
Angle Orthod
.
2011
;
81
:
1075
1082
.
28.
Hong
JS
,
Oh
KM
,
Kim
BR
,
Kim
YJ
,
Park
YH
.
Three-dimensional analysis of pharyngeal airway volume in adults with anterior position of the mandible
.
Am J Orthod Dentofacial Orthop
.
2011
;
140
:
e161
e169
.
29.
El
H
,
Palomo
JM
.
Airway volume for different dentofacial skeletal patterns
.
Am J Orthod Dentofacial Orthop
.
2011
;
139
:
e511
e521
.
30.
Alves
M
Jr,
Franzotti
ES
,
Baratieri
C
,
Nunes
LK
,
Nojima
LI
,
Ruellas
AC
.
Evaluation of pharyngeal airway space amongst different skeletal patterns
.
Int J Oral Maxillofac Surg
.
2012
;
41
:
814
819
.
31.
Abdelkarim
A
.
A cone beam CT evaluation of oropharyngeal airway space and its relationship to mandibular position and dentocraniofacial morphology
.
J World Fed Orthod
.
2012
;
1
:
e55
e59
.
32.
Muto
T
,
Takeda
S
,
Kanazawa
M
,
Yamazaki
A
,
Fujiwara
Y
,
Mizoguchi
I
.
The effect of head posture on the pharyngeal airway space (PAS)
.
Int J Oral Maxillofac Surg
.
2002
;
31
:
579
583
.
33.
Haskell
JA
,
McCrillis
J
,
Haskell
BS
,
Scheetz
JP
,
Scarfe
WC
,
Farman
AG
.
Effects of mandibular advancement device (MAD) on airway dimensions assessed with cone-beam computed tomography
.
Semin Orthod
.
2009
;
15
:
132
158
.
34.
Martin
SE
,
Mathur
R
,
Marshall
I
,
Douglas
NJ
.
The effect of age, sex, obesity and posture on upper airway size
.
Eur Respir J
.
1997
;
10
:
2087
2090
.
35.
Arens
R
,
Marcus
CL
.
Patophysiology of upper airway obstruction: a developmental perspective
.
Sleep
.
2014
;
27
:
997
1019
.
36.
Kollias
I
,
Krogstad
O
.
Adult craniocervical and pharyngeal changes—a longitudinal cephalometric study between 22 and 42 years of age. Part II: Morphological uvulo-glossopharyngeal changes
.
Eur J Orthod
.
1999
;
21
:
345
355
.
37.
Hussels
W
,
Nanda
RS
.
Analysis of factors affecting angle ANB
.
Am J Orthod
.
1984
;
85
:
411
423
.
38.
Wolford
LM
,
Perez
D
,
Stevao
E
,
Perez
E
.
Airway space changes after nasopharyngeal adenoidectomy in conjunction with Le Fort I osteotomy
.
J Oral Maxillofac Surg
.
2012
;
70
:
665
671
.
39.
Schwab
RJ
,
Pasirstein
M
,
Pierson
R
,
et al
.
Identification of upper airway anatomic risk factors for obstructive sleep apnea with volumetric magnetic resonance imaging
.
Am J Respir Crit Care Med
.
2003
;
168
:
522
530
.
40.
Ghoneima
A
,
Kula
K
.
Accuracy and reliability of cone-beam computed tomography for airway volume analysis
.
Eur J Orthod
.
2013
;
35
:
256
261
.
41.
Weissheimer
A
,
Menezes
LM
,
Sameshima
GT
,
Enciso
R
,
Pham
J
,
Grauer
D
.
Imaging software accuracy for 3-dimensional analysis of the upper airway
.
Am J Orthod Dentofacial Orthop
.
2012
;
142
:
801
813
.
42.
Holty
JE
,
Guilleminault
C
.
Maxillomandibular advancement for the treatment of obstructive sleep apnea: a systematic review and meta-analysis
.
Sleep Med Rev
.
2010
;
14
:
287
297
.