SEARCH

SEARCH BY CITATION

Keywords:

  • closing pressure;
  • collapsibility;
  • fat distribution;
  • obstruction;
  • pharynx

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Epidemiological evidence suggests there are significant links between obesity and obstructive sleep apnoea (OSA), with a particular emphasis on the importance of fat distribution in the development of OSA. In patients with OSA, the structure of the pharyngeal airway collapses. A collapsible tube within a rigid box collapses either due to decreased intraluminal pressure or increased external tissue pressure (i.e. reduction in transmural pressure), or due to reduction in the longitudinal tension of the tube. Accordingly, obesity should structurally increase the collapsibility of the pharyngeal airway due to excessive fat deposition at two distinct locations. In the pharyngeal airway region, excessive soft tissue for a given maxillomandibular enclosure size (upper airway anatomical imbalance) can increase tissue pressure surrounding the pharyngeal airway, thereby narrowing the airway. Even mild obesity may cause anatomical imbalance in individuals with a small maxilla and mandible. Lung volume reduction due to excessive central fat deposition may decrease longitudinal tracheal traction forces and pharyngeal wall tension, changing the ‘tube law’ in the pharyngeal airway (lung volume dependence of the upper airway). The lung volume dependence of pharyngeal airway patency appears to contribute more significantly to the development of OSA in morbidly obese, apnoeic patients. Neurostructural interactions required for stable breathing may be influenced by obesity-related hormones and cytokines. Accumulating evidence strongly supports these speculations, but further intensive research is needed.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Obstructive sleep apnoea (OSA) is characterized by repetitive pharyngeal narrowing and closure during sleep, with obesity being an important feature associated with the presence and development of this disorder. Longitudinal studies in the general population consistently demonstrate that weight changes are directly associated with increases or decreases in the severity of OSA in both genders, although the association is less robust in women than in men1,2 (Fig. 1). Large cross-sectional studies in the general population have also identified BMI as an independent predictor of OSA. However, other factors impact on the association between obesity and OSA, with gender, age and neck circumference also being significant predictors of sleep disordered breathing.3 Notably, there are similar associations between weight and OSA across different races, despite a wide range of BMI, suggesting an interaction between obesity and craniofacial dimensions in the development of OSA.4–7 In clinical populations, BMI is not a good predictor of OSA, with neck and waist circumferences being more sensitive parameters for prediction of OSA.8–10 Furthermore, the amounts of adipose tissue adjacent to the pharyngeal airway and in the intraperitoneal space are directly associated with AHI but not with BMI.11,12 Differences in fat distribution between men and women, and the resulting mechanical and functional influences, may partly explain gender differences in the prevalence of OSA.13,14

image

Figure 1. Five year follow-up of changes in weight and respiratory disturbance index (RDI) indicates a clear direct relationship between the two, which was more marked in males than in females. (From Newman et al.1 with permission)

Download figure to PowerPoint

Epidemiological and clinical evidence strongly suggests that there is a complex interaction between obesity and OSA. This review discusses the possible contributions of obesity, and in particular, the role of fat deposition in the neck, surrounding the pharyngeal airway, as well as intra-abdominal fat deposition, in the pathogenesis of OSA, based on accumulating knowledge of the pathophysiology of upper airway (UA) obstruction in humans.

COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

A collapsible tube as a model for understanding the behaviour of the human pharyngeal airway

The pharynx, which is the site where OSA develops, is inherently collapsible during sleep when the activity of muscles dilating the UA decreases.15 The pharyngeal airway is structurally surrounded by soft tissues such as the tongue, soft palate, tonsils and pharyngeal fat pads, while it is also enclosed by bony structures, such as the mandible, maxilla and cervical spine.16 This structure is mechanically analogous to an artificial collapsible tube in a rigid chamber (Fig. 2), and the behaviour of an atonic or hypotonic pharyngeal airway (passive pharynx) can be demonstrated by a simple mechanical model.17

image

Figure 2. Most pharyngeal airway behaviour can be modelled by a collapsible tube in a rigid box. The cross-sectional area of the tube is determined by a ‘tube law’ representing the intrinsic mechanical properties of the tube. Two distinct mechanisms result in collapse of the tube: reduction in transmural pressure (Ptm) due either to a decrease in intraluminal pressure (Plumen) or an increase in surrounding tissue pressure (Ptissue) (A to B), or a reduction in longitudinal tension of the tube (A to C), making the tube mechanically more compliant and therefore the slope of the ‘tube law’ more steep (‘tube law’ 1 to ‘tube law’ 2).

Download figure to PowerPoint

In this mechanical model, the cross-sectional area of the tube is determined by a ‘tube law’, representing the intrinsic mechanical properties of the tube and transmural pressure (Ptm), the difference between intraluminal and tissue pressures (Ptm = Plumen − Ptissue). Plumen is defined as the lateral wall pressure acting on the luminal surface of the tube, while Ptissue is defined as the tissue pressure acting on the outside surface of the tube. Two distinct mechanisms result in collapse of the tube: reduction of Ptm due either to a decrease in Plumen or an increase in Ptissue (A to B in Fig. 2), or reduction in the longitudinal tension of the tube (A to C in Fig. 2), making the tube more compliant and therefore the slope of the ‘tube law’ more steep.18 In the human pharyngeal airway, Ptissue may be increased by anatomical imbalance of the UA due to the deposition of fat surrounding the pharyngeal airway within the maxillomandibular bony enclosure,16 while longitudinal traction of the pharyngeal airway may be decreased by lung volume reduction as discussed below.

How can we assess pharyngeal airway collapsibility in humans?

Because of the difficulty in measuring the Ptissue surrounding the pharyngeal airway in humans, static Plumen-area relationships have been determined by measuring the pharyngeal cross-sectional area during step reductions of Plumen in cadavers of infants19 and in anaesthetized humans20 as substitute models for the ‘tube law’. The static Plumen-area relationship is not linear and is described by an exponential function, irrespective of age, gender and the coexistence of OSA, indicating that pharyngeal wall compliance, defined as the slope of the curve, varies with changes in Plumen.21 Over the steep portion in the lower range of Plumen, significant pharyngeal narrowing occurs even for a small reduction in Plumen, inducing inspiratory flow limitation.22 In contrast, the pharyngeal airway is stable along the flat portion in the higher range of Plumen. The closing pressure of the pharyngeal airway (Pclose), as determined by the intercept of the curve on the Plumen axis, is an important mechanical parameter representing collapsibility of the pharyngeal airway.

Although not a direct measure of Ptissue, the critical closing pressure (Pcrit), which is determined by assuming the UA to be a Starling resistor that collapses when Ptissue exceeds Plumen, is considered to reflect Ptissue and is measurable in sleeping humans.23 However, caution is required when interpreting Pcrit values because Pcrit can be measured while the UA dilating muscles are actively contracting (active Pcrit) or when they are suppressed (passive Pcrit).24 Recently, Kairaitis et al. developed a technique for measuring Ptissue surrounding the pharyngeal airway in anaesthetized rabbits and demonstrated changes in Ptissue and UA resistance in response to various mechanical interventions, thereby providing support for the ‘tube law’ concept and the increase in Ptissue as a mechanism for increasing Pclose of the pharyngeal airway.25–29Table 1 shows the agreement among Pclose,30–33 Pcrit34–36 and Ptissue25–29 in response to various mechanical interventions.

Table 1.  Agreements among three different experimental approaches for exploring pharyngeal airway collapsibility in responses to various mechanical interventions. (summarized based on the references)
Conditions (references)PtissuePcritPclose
Neck flexion21,25,34[UPWARDS ARROW][UPWARDS ARROW][UPWARDS ARROW]
Mandibular advancement22,26,35[DOWNWARDS ARROW][DOWNWARDS ARROW][DOWNWARDS ARROW]
Tracheal traction or lung volume23,27,34[DOWNWARDS ARROW][DOWNWARDS ARROW][DOWNWARDS ARROW]
Mass loading28[UPWARDS ARROW]N/A([UPWARDS ARROW])
UA muscle contraction24,29,36[DOWNWARDS ARROW][DOWNWARDS ARROW][DOWNWARDS ARROW]

Is the pharynx structurally more collapsible in individuals with OSA than in those without OSA?

Isono et al. assessed the collapsibility of the passive pharynx under general anaesthesia and muscle paralysis by eliminating neuromuscular control mechanisms.20 They showed that Pclose of the passive pharynx in OSA patients (0.6 ± 1.5 cm H2O for mild OSA, 2.2 ± 3.0 cm H2O for severe OSA) was significantly greater than that in age- and BMI-matched subjects without OSA (-4.4 ± 4.2 cm H2O), providing evidence for abnormally collapsible pharyngeal structures in OSA patients (Fig. 3). Interestingly, the reported passive Pcrit values determined during suppression of the UA dilating muscles agreed well with the Pclose determined during anaesthesia and paralysis, suggesting that the higher Pclose in OSA patients is at least partly due to a higher Ptissue.37

image

Figure 3. Relationship between closing pressure (Pclose) at the velopharynx (retropalatal airway) and Pclose at the oropharynx (retroglossal airway), as determined in anaesthetized and paralyzed age- and BMI-matched individuals without OSA (closed circles), or with mild OSA (open squares), and in patients with severe OSA (closed triangles). Note that the patients with OSA had higher Pclose at both segments and that retropalatal Pclose values were greater than retroglossal Pclose. (From Isono et al.20: this is an original work of the author)

Download figure to PowerPoint

Is Ptissue increased in obese patients with OSA?

In the collapsible tube model, an increase in Ptissue narrows the cross-sectional area of the tube. The pharyngeal airway shares space with soft tissues within the maxillomandibular bony enclosure. The balance between the size of the bony enclosure and the amount of soft tissue may determine the airway space and Ptissue.16,38 An increase in the amount of soft tissue within the bony enclosure or a decrease in the size of the bony enclosure would result in limitation of the space available for the airway and consequently, a narrowing of the airway and increase in Ptissue. Winter et al. inflated a balloon inside the bony enclosure at the level of the pharynx in anaesthetized pigs and showed that UA resistance increased during balloon inflation.39 Kairaitis et al. confirmed that an increase in the mass of tissue surrounding the pharyngeal airway resulted in increases in both Ptissue and UA resistance in anaesthetized rabbits.28 Tsuiki et al. found that OSA patients had significantly larger tongues for any given maxillomandibular size compared with BMI-matched subjects without OSA, suggesting that increased Pclose and Ptissue were due to an anatomical imbalance in OSA patients.40 The dependence of the passive Pcrit on BMI in both men and women is clear37 (Fig. 4). A 10 kg/m2 difference in BMI has been estimated to produce 1.67 and 0.95 cm H2O differences in passive Pcrit in men and women, respectively.37 Interestingly, weight loss equivalent to a 10 kg/m2 reduction in BMI resulted in a more significant change in active Pcrit from 3.1 ± 4.2 to −2.4 ± 4.4 cm H2O, suggesting improvement in neural control mechanisms in addition to anatomical balance.41,42

image

Figure 4. Relationship between passive Pcrit and BMI in male and female patients with OSA. In both genders, there was a significant correlation between passive Pcrit and BMI (adjusted for age) for the entire sample. Note that the distribution of subjects in the subset that was matched for respiratory disturbance index, and BMI (closed symbols) did not differ from the distribution of the entire subject group (open and closed symbols). Dashed lines represent 95% CIs. (From Kirkness et al.37 with permission)

Download figure to PowerPoint

In which pharyngeal segments does obesity increase collapsibility?

Various techniques, including multi-sensor catheter, pharyngeal endoscopy, fluoroscopy, dynamic MRI and fast CT scanning have been used to determine the sites of pharyngeal obstruction during sleep in patients with OSA. Although obstruction may occur at any segment along the pharyngeal airway, it is most frequently observed in either the retropalatal and/or retroglossal regions of the airway. The pattern and severity of pharyngeal occlusion are significantly influenced by a number of factors, including dynamic changes in respiratory effort, varying activity of the UA dilator muscles and the level of consciousness during natural OSA.43–45 Therefore, in order to determine the airway segment(s) primarily responsible for OSA, Plumen should ideally be controlled and the activity of UA muscles suppressed.

Only two techniques can successfully achieve both these conditions. Remmers and colleagues suppressed UA muscle activity in patients with OSA by applying optimal CPAP and manipulating Plumen to induce pharyngeal obstruction during either natural or diazepam-induced sleep. They then determined segmental Pclose by endoscopic observation during measurements of airflow and pharyngeal pressures.21,46,47 As described previously, Isono et al. eliminated UA muscle activity under general anaesthesia and total paralysis and measured segmental Pclose during step Plumen reduction in OSA patients and age- and BMI-matched control subjects.20 Both techniques identified two distinct patterns of primary site(s) of pharyngeal closure. Approximately half the OSA patients showed a Pclose greater than atmospheric pressure exclusively in the retropalatal segment, while the other half showed a positive Pclose in both the retropalatal and retroglossal segments. Furthermore, obesity appeared to be more frequently associated with the former type of airway closure, whereas craniofacial abnormalities were more common in the latter type.16 Structurally, the tongue is located anterior to the soft palate and the retropalatal airway is narrower than the retroglossal airway. In anaesthetized and paralyzed humans, Pclose at the retropalatal airway has been shown to be greater than Pclose at the retroglossal airway20 (Fig. 3). It is likely that as a consequence of these structural arrangements, the retropalatal airway narrows earlier than the retroglossal airway when there is accumulation of fat in both the lateral pharyngeal wall and the tongue.

MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Is fat deposited excessively within the maxillomandibular enclosure in patients with OSA?

Shelton et al. demonstrated a correlation between the amount of fat enclosed by the mandibular ramus and AHI, but not BMI.11 Horner et al. showed that there was significantly more fat deposition at the level of the soft palate in OSA patients compared with weight-matched control subjects without OSA.48 Even relatively non-obese OSA patients show excess fat deposition, especially anterolateral to the UA, as compared with individuals without OSA with similar BMI and neck circumferences.49

In order to better understand the anatomical risk factors for OSA, Schwab et al. performed detailed three dimensional volumetric soft tissue measurements at various UA regions in a group of subjects with OSA (44 ± 25 events/h) and compared this with data from a group of control subjects who were matched for gender, age and ethnicity.50 While enlargement of the soft tissues at all UA regions in OSA patients was observed, multivariate analysis identified tongue volume and lateral pharyngeal wall volume as independent risk factors for sleep apnoea. Although fat deposits are easily identifiable in the lateral pharyngeal wall, it is difficult to accurately assess fat deposition within the tongue musculature. However, using consecutive autopsy specimens from the general population, Nashi et al. showed that fat content in the posterior tongue (30%) was significantly higher than that in the anterior tongue (11%) or other somatic muscles (3%), and was correlated with BMI51 (Fig. 5).

image

Figure 5. (A) Standardized digital image of sagittal section of the tongue, demonstrating extramyocellular fat within the posterior third of the tongue and in the sublingual region just below the intrinsic tongue musculature. (B) Average percentage fat content in the anterior, posterior,and sublingual regions (95% CI of mean). (From Nashi et al.51 with permission)

Download figure to PowerPoint

Saito et al. recently demonstrated that in rats that became obese after being fed a high-fat diet, the percentages of oil droplet areas increased in the genioglossus and geniohyoid muscles but not in the masseter muscle. This resulted in an increase in muscle fibre diameter, particularly in the slow-type myofibres, but there was no change in the composition of the fibre types in these muscles.52 However, intermittent hypoxaemia, which frequently occurs as a consequence of upper airway obstruction, appears to induce a transition from endurance type IIa fibres to more fatigable type IIb fibres.53 In patients with OSA, this transition has been shown to be reversible with CPAP treatment.54

Tongue volume, however, is not always greater55–59 and does not necessarily decrease after moderate weight reduction in OSA patients.60 It should be noted that measurements of soft tissue volume alone, without controlling for craniofacial dimensions, do not reflect the anatomical balance of the UA and Ptissue. In fact, in two studies in which tongue volumes were not greater in patients with OSA,58,59 the anatomical balance of the UA, as assessed by the ratio of relative soft tissue volume to maxillomandibular dimensions, differed between OSA patients and control subjects without OSA. Furthermore, patients with position dependent OSA, defined as a ratio of AHI in the lateral position to AHI in the supine position of 0.5 or less, had a smaller lateral pharyngeal wall volume than patients with non-positional OSA, indicating the importance of regional anatomical balance.59

In summary, fat may be deposited in any part of the soft tissue surrounding the UA, thereby increasing the total volume of soft tissue within the maxillomandibular enclosure. However, fat accumulation in the lateral pharyngeal wall and posterior tongue appears to play an important role in the development of OSA.

How much of an increase in UA soft tissue volume is necessary for the development of OSA?

As discussed previously, during weight gain fat may accumulate within the UA soft tissues, leading to an anatomical imbalance. However, it remains unclear how much of an increase in soft tissue volume is necessary before OSA develops. Schwab et al. reported that total UA soft tissue volumes were approximately 30 cm3 greater in obese patients with OSA than in non-obese control subjects, although BMI and craniofacial dimensions were not controlled.50 Interestingly, the difference in soft tissue volumes within the maxillomandibular enclosure were unexpectedly small, considering the significant differences in BMI between the groups (36 vs 26 kg/m2). It should be noted that the participants in that study were primarily Caucasians and African-Americans, and that according to the UA anatomical balance concept, the amount of excess soft tissue underlying the development of OSA may differ between races and in those with craniofacial abnormalities.40 In fact, three dimensional soft tissue measurements performed on Japanese males showed that the difference in UA soft tissue volumes between OSA patients and subjects without OSA, who were matched for craniofacial size, was approximately 20 cm3, which is smaller than the difference reported by Schwab et al..59 Sutherland et al. showed that a decrease in total UA fat volume of approximately 17 cm3 after moderate weight reduction (7.8 ± 4.2 kg) resulted in a 31% reduction in AHI in Caucasian patients with OSA.61 To my knowledge, no other study has measured changes in UA soft tissue volume before and after successful weight loss and resolution of OSA. Upper airway imaging studies in adults suggest that OSA would presumably develop following a 20–30 cm3 increase in UA soft tissue volume, depending on the size of the maxillomandibular enclosure.

Why do patients with OSA have excessive UA soft tissue in the submandibular region?

Unlike the cranial space, the maxillomandibular bony enclosure is not a closed space. Excessive soft tissue surrounding the pharynx may be displaced outside the bony enclosure through the submandibular space, possibly offsetting its impact on Ptissue and pharyngeal airway patency. This excess submandibular soft tissue may be detected in patients with OSA as an increased neck circumference, a ‘double chin’, or reduced cricomental space, which have high negative predictive values for exclusion of OSA.62 Although caudal displacement of the hyoid bone is a common finding in lateral cephalograms of patients with OSA, this may arise as a result of UA anatomical imbalance rather than being a cause of OSA.63 A significantly greater distance between the hyoid and mandibular planes (MP-H) (26 mm vs 16 mm) has been measured in patients with OSA compared to control subjects without OSA, who were matched for BMI and craniofacial dimensions.40 Mandibular advancement, with either an oral appliance or by surgery, which improves the anatomical imbalance, was reported to decrease the MP-H.64,65 The caudal displacement of excessive soft tissue not only compensates for the UA anatomical imbalance but, at the same time, elongates the pharyngeal airway. Some researchers consider that lengthening of the pharyngeal airway by 15–30% may contribute to the development of OSA. As the airway is longer in males than in females, even after normalization for body height, this may partly explain the gender difference in the prevalence of OSA, despite the relatively small impact of airway length on airway resistance, as compared with that of cross-sectional area.61,66,67 Alternatively, a longer UA in patients with OSA and in males could be considered beneficial in increasing the longitudinal tension of the pharyngeal airway. Shortening of the long UA in patients with OSA, without changes in other structural dimensions, would further increase UA collapsibility. In the mechanical collapsible tube model, shortening of the tube will increase its collapsibility by decreasing longitudinal tension, and therefore increasing the slope of the ‘tube law’. Caudal displacement of the trachea has been demonstrated to decrease Ptissue, Pcrit and UA resistance in experimental animal models.27,34 Further studies are required, but a thick neck, double chin and lower hyoid bone are clinical markers of UA anatomical imbalance and caudal displacement of excessive UA soft tissue. These changes may serve to compensate for UA anatomical imbalance by decreasing Ptissue and increasing longitudinal pharyngeal wall tension, although further investigations are necessary to confirm this.

POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Do patients with OSA have excessive visceral fat?

Although the ratio of visceral fat volume to subcutaneous fat volume is small, metabolic consequences such as metabolic syndrome and type II diabetes are predominantly caused by intra-abdominal adipose tissue. This may also hold true for the pathogenesis of OSA and associated consequences. Visceral fat volume, but not subcutaneous fat volume, is directly associated with AHI in patients with OSA.12 Despite having a similar BMI and total fat weight, patients with OSA accumulate more visceral adipose tissue than individuals without OSA.68,69 Short-term nasal CPAP reduces visceral fat and serum leptin levels even without changes in BMI.70 Interestingly, in women, both the prevalence of OSA and increases in visceral fat are accelerated by menopause.71,72 Furthermore, the prevalence of sleep disordered breathing is lower in postmenopausal women receiving hormone replacement therapy72,73 and higher in women with polycystic ovarian syndrome.74,75 Despite the potential importance of visceral fat in the pathogenesis of OSA, we still have little understanding of the link between OSA and visceral fat volume. As discussed below, both mechanical and functional influences of increased visceral fat on OSA need to be considered, although these associations have yet to be proven.

Does obesity-related reduction in lung volume influence UA patency?

A reduction in FRC due to reduced expiratory reserve volume without changes in residual volume is the most significant and consistent effect of obesity on pulmonary function76–79 (Fig. 6). Clinically significant impairment of dynamic pulmonary function parameters such as FVC and FEV1 only occurs in extremely obese subjects. Contrary to general belief, Babb et al. failed to demonstrate that abdominal fat distribution was a predominant factor accounting for the reduction in end expiratory lung volume (EELV) either in obese men or obese women. They concluded that the reduction in EELV most likely arose from the cumulative effect of increased chest wall fat rather than any specific regional distribution of chest wall fat.80 In morbidly obese patients, weight loss significantly improved expiratory reserve volume as well as resting blood gas parameters.81

image

Figure 6. Relationship between BMI and FRC with an exponential regression line, in 373 individuals of both genders with normal FEV1/FVC. (From Jones and Nzekwu78 with permission)

Download figure to PowerPoint

While a reduction in FRC underlies the development of severe hypoxaemia during OSA, decreased lung volume per se, is thought to contribute to pharyngeal airway obstruction; this is based on the landmark work of Hoffstein et al., who demonstrated the dependence of pharyngeal airway patency on augmented lung volume in obese, awake patients with OSA.82 Using acoustic reflection, this group showed that pharyngeal cross-sectional area decreased significantly more during slow exhalation from total lung capacity to residual volume in obese patients with OSA than in weight-matched control subjects without OSA. Notably, the FRC of obese, apnoeic subjects was significantly lower than that of obese non-apnoeic subjects, suggesting the possible importance of the type of obesity. Interestingly, the same investigators also reported significant differences in lung volume dependence between females with or without OSA, despite similar reductions in FRC, suggesting an interaction between reduced lung volume and increased pharyngeal collapsibility.83

How do tonic and phasic tracheal traction forces influence pharyngeal airway collapsibility?

Although the lung is located at a distance from the pharyngeal airway, the trachea connects the two mechanically. Static and dynamic inflation of the lung produces tonic and phasic tracheal traction forces, respectively. The tonic force may be reduced in obese patients with lower FRC through alterations in the intrinsic mechanical properties of the pharyngeal airway, that is, ‘tube law’ or static Plumen-area curve. The phasic tracheal traction force varies during the respiratory cycle and may serve to stabilize the pharyngeal airway and airflow during inspiration, as predicted by the behaviour of an artificial collapsible tube.

Does reduction in lung volume increase pharyngeal airway collapsibility?

In both animal and human studies, UA resistance and collapsibility decreases in response to either direct changes in tonic tracheal traction force or lung inflation.27,32,34,84–88 In sleeping individuals without OSA, genioglossus activity decreases in response to lung inflation, despite a reduction in UA resistance,85,86 indicating that the lung volume dependence of pharyngeal airway patency is primarily due to structural mechanisms, such as increased tonic tracheal traction forces. In anaesthetized cats, the impact of direct tracheal traction on UA patency was greater during tongue displacement84 and neck extension,34 suggesting an interaction between UA anatomical balance and lung volume dependence. Kairaitis and colleagues showed that there were small but significant reductions in Ptissue in response to tracheal traction in anaesthetized rabbits.27

Tagaito et al. showed that following an increase in lung volume of 0.7 litre, there was a small but significant reduction in retropalatal Pclose of 1.2 cm H2O, in anaesthetized, paralyzed patients with OSA (BMI 23.1–30.8 kg/m2).32 Although this study included few obese patients with OSA, the direct association between changes in retropalatal Pclose and BMI suggests there are differences in the lung volume dependence of pharyngeal collapsibility between obese and non-obese patients with OSA.32 A greater reduction in Pcrit (2.2 cm H2O to −1.0 cm H2O) was observed in response to a 0.6 litre lung inflation during non-rapid eye movement sleep in more obese patients with OSA (BMI 25.5–45.5 kg/m2).87 Squier et al. elegantly demonstrated an indirect relationship between Pcrit and EELV in non-obese men and women88 (Fig. 7). These authors estimated that there was approximately a 2 cm H2O reduction in Pcrit in response to an increase in lung volume of one litre in these subjects. Interestingly, the Pcrit value for a given lung volume was more negative in women than in men, suggesting that protective factors other than lung volume may be operating in women.

image

Figure 7. Relationships between absolute end expiratory lung volume (EELV) and upper airway critical closing pressure (Pcrit), as determined by the isovolumetric technique in men (n = 11) and women (n = 7) without OSA. There was an inverse relationship between EELV and Pcrit in both genders with the upper airway being more collapsible in men than in women for the same EELV. (From Squier et al.88 with permission)

Download figure to PowerPoint

If a reduction in lung volume is involved in the development of OSA, lung inflation during sleep should decrease the frequency of OSA. Such a finding has been documented in a case report89 and in a clinical trial of obese patients with OSA90 but could not be confirmed in another trial involving less obese patients with OSA.91 However, these conflicting results based on just 22 individuals with OSA do not provide conclusive evidence for the lung volume hypothesis. Nevertheless, the difference in the severity of obesity between these studies is worthy of note, and the findings are likely to be explained by differences in study protocols and populations.

What is the role of tracheal tug during OSA?

In humans, the trachea is reported to move caudally, even during tidal breathing.92 Van de Graaff demonstrated that a phasic inspiratory increase in tracheal traction forces was evident in anaesthetized dogs breathing through a tracheostomy, but this disappeared after mechanical disconnection of the trachea from the UA.93 Furthermore, as illustrated in Figure 8, he showed that augmentation of oesophageal pressure and a decrease in caudal movement of the carina during inspiratory occlusion increased tracheal traction, but maintenance of the carina in position during expiratory obstruction failed to maintain the tracheal traction forces.94 This suggests that lung inflation may not be essential for producing and maintaining phasic tracheal traction forces. A ‘tracheal tug’ is often evident in patients with OSA during progressively increasing respiratory drive in OSA. This has two contradictory influences on pharyngeal obstruction; it serves to terminate obstruction by a progressive increase in tracheal traction forces and also maintains obstruction by increasing the negative Plumen. The phasic tracheal traction force is believed to operate more efficiently during OSA in obese patients with reduced lung volumes because of the increased diaphragmatic curvature at lower lung volumes.95

image

Figure 8. Respiratory changes in tracheal traction force (Ttx) in a spontaneously breathing, anaesthetized and tracheotomized dog. Note that inspiratory occlusion (4th breath) resulted in augmentation of oesophageal pressure (Pes) and Ttx but decreased caudal movement of the carina (Dcar). Maintenance of the carina in position during expiratory obstruction (9th breath) failed to maintain Ttx. (From Van de Graaff94 with permission)

Download figure to PowerPoint

Visceral fat and stability of UA neural control mechanisms

While the main focus of this paper has been the mechanisms promoting increased pharyngeal airway collapsibility, recent evidence suggests possible contribution of increased visceral fat, and particularly circulating leptin, to breathing instability during sleep in some patients with OSA. In short, obesity is associated with increased levels of circulating leptin, which is possibly a respiratory stimulant.96 In obese patients with OSA, any increase in the loop gain of negative feedback to the respiratory system will promote respiratory instability and consequently produce more frequent OSA oscillations. While this is speculative, patients with OSA have been shown to have a higher respiratory loop gain during sleep,97 and this loop gain was significantly associated with AHI in OSA patients with a Pclose near atmospheric pressure.98 Increased gas exchange efficacy because of a low lung volume, low dead space, low metabolic rate, low cardiac output and high arterial carbon dioxide, all of which are common in obese patients with OSA, may also contribute to the increased respiratory loop gain.99 Patients with OSA who are eucapnic but not those with hypercapnia, have higher serum leptin levels and hypercapnic ventilatory responses than BMI-matched individuals without OSA.100 Interestingly, higher leptin levels in obese patients with OSA were independent of obesity and were reversed by short term use of CPAP.70,101. Evidence that serum leptin levels are higher in obese individuals and those with OSA suggests the potential for leptin resistance to contribute to the development of obesity and resistance to weight loss. Excess visceral adipose tissue secretes a number of other hormones and pro-inflammatory cytokines that may also influence breathing in obese patients with OSA.102 While this line of investigation may be fruitful, further intensive research is still required to better understand the pathogenesis of OSA.

SUMMARY

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Obese individuals may develop OSA due to excessive deposition of fat at two distinct locations, as summarized in Figure 9. In the pharyngeal airway region, excessive soft tissue within the maxillomandibular enclosure may increase tissue pressure, leading to pharyngeal airway narrowing. Even mild obesity may cause an anatomical imbalance in patients with a small maxilla and mandible. Excessive submandibular soft tissue indicates and serves to compensate for this anatomical imbalance. Excessive central fat deposition may decrease lung volumes and longitudinal pharyngeal wall tension. The lung volume dependence of pharyngeal airway patency may be more significant in obese patients with OSA. Obesity-related hormones and cytokines such as leptin may impair the neuro-anatomical interactions necessary for stable breathing, thereby increasing the frequency of OSA. Although current evidence strongly supports these speculations, further intensive research is required.

image

Figure 9. Hypothetical neuroanatomical relationships between obesity and OSA. See text for detailed explanation.

Download figure to PowerPoint

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. COLLAPSIBILITY OF THE PHARYNGEAL AIRWAY IN PATIENTS WITH OSA
  5. MECHANICAL CONSEQUENCES OF FAT DEPOSITION WITHIN THE MAXILLOMANDIBULAR ENCLOSURE ON PHARYNGEAL COLLAPSIBILITY
  6. POSSIBLE INFLUENCES OF VISCERAL FAT ON OSA
  7. SUMMARY
  8. ACKNOWLEDGEMENTS
  9. REFERENCES