Effect of Body Posture on Spirometric Values and Upper Airway Obstruction Indices Derived From the Flow-Volume Loop in Young Nonobese Subjects*

Transcription

1 Effect of Body Posture on Spirometric Values and Upper Airway Obstruction Indices Derived From the Flow-Volume Loop in Young Nonobese Subjects* Marc Meysman, MD; and Walter Vincken, MD, PhD, FCCP Study objective: To define the effect of changes in body posture on flow-volume loops (FVLs) and four commonly used indices of upper airway obstruction (UAO) in young, nonobese normal subjects. Design: Prospective comparative study. Setting: Pulmonary function laboratory at an academic hospital. Participants: Thirty-one normal volunteers. Intervention: At least three FVLs per posture were obtained in the sitting, supine, and left and right lateral recumbent postures while maintaining a constant position of the head and neck in relation to the trunk. In each body posture, the largest observed flow rates were used to calculate the UAO indices. Results: When subjects changed from the sitting to each of the three recumbent postures, all spirometric values decreased significantly (p<o.oooi). However, among the four UAO indices, only the FEV /peak expiratory flow ratio increased significantly (although only slightly, by 2.9 and 4.4%, respectively) in both the right and left lateral recumbent postures (p<o.oooi), but not in the supine posture. None of the subjects developed an inspiratory or expi.-atory plateau on the FVL in any of the three recumbent postures. Conclusions: In young, nonobese normal subjects, recumbency does not induce UAO, at least not detectable by changes in the FVL configuration or in UAO indices derived from the FVL. Furthermore, the study provides the upper limits of recumbency-related changes in the various UAO indices for young, nonobese normal subjects. (CHEST 1998; 114: ) Key words: body posture; How-volume loop; upper airway obstruction; young normal subjects Abbreviations: ANOVA =analysis of variance; FEF 50 =forced expiratory flow measured at 50% of expiratmy vital capacity; FEV 0 _ 5 = forced expiratmy volume in 0.5 s; FIF 50 = forced inspiratory flow measured at 50% of inspiratory vital capacity; FVL= flow-volume loop; MEPS=maximal static expiratory mouth pressure; MIPS=maximal static inspiratory mouth pressure; OSAS=obstructive sleep apnea syndrome; PEF=peak expiratory flow; PIF= peak inspiratory flow; RL = pulmonaty flow resistance; UAO=upper airway obstruction In their classical papers, Miller and Hyatt 1,2 recognized the flow-volume loop (FVL) as a valuable test for the detection and assessment of upper airway obstruction (UAO). The value of the FVL, as opposed to static imaging techniques, lies in the fact that it takes into account the dynamic characteristics of the upper ai1way, ie, the changes in caliber that *From the Respiratory Division, Department of Medicine, Academic Hospital, University of Brussels (AZ-VUB ), Brussels, Belgium. Supported by the Belgian National Fund for Scientific Research (NFWO, Levenslijn grant ). Manuscript received April 7, 1997; revision accepted March 23, Correspondence to: Walter Vincken, MD, FCCP, Head, Respiratory Division, Academic Hospital, University of Bmssels (AZ VUB ), Laarheeklaan 101, 1090 Brussels, Belgium the collapsible parts of the upper airway undergo as a result of the changes in transmural pressure induced by forced inspiratory and forced expiratmy maneuvers. The effect of these maneuvers, used for the generation of a FVL, on the caliber of the upper airway and, hence, on the effort-dependent inspiratory and (early) expiratory flow rates is exaggerated in the presence of an upper airway lesion that affects the structural or functional integrity of the upper airway. As a result, the forced FVL is a useful test for the detection of upper airway instability by producing flow oscillations, 3 and for detecting frank UAO, as reflected by a flattened FVL contour, reduced effort-dependent inspiratory and expiratory flow rates, and abnormal values for so-called UAO indices, such as the FEY /forced expiratmy volume in 1042 Clinical Investigations

2 0.5 s (FEV 05 ) ratio, the FEV/peak expiratory flow (PEF) ratio, the forced expiratory flow measured at 50% of expiratory vital capacity (FEF, 50 )/forced inspiratory flow measured at 50% of inspiratory vital capacity (FIF 50 ) ratio, and the PEF/peak inspiratmy flow (PIF) ratio. 4 5 The forced FVL may even help localize flow-limiting lesions into the intrathoracic or extrathoracic portion of the upper airway (on the condition that these lesions have not become fixed but remain variable and still respond to the changes in transmural pressure induced by forced inspiration and expiration). In a previous paper published in CHEST, 6 we showed that changes in body posture may be used to further increase the capability of the FVL to detect UAO. Indeed, in two patients with a thyroid goiter, the FVL obtained in the supine posture revealed functional evidence of UAO not present on the FVL obtained in the traditional upright posture. However, before recommending recumbent FVLs in patients in whom UAO is suspected, it is necessa1y to determine the effect of the recumbent posture on the FVL-derived parameters and UAO indices in normal subjects, as reported in this paper. MATERIALS AND METHODS FVLs and spirometric values were obtained in 31 normal, healthy volunteers. There were 1.5 women and 16 men, with a mean age (±SD) of2.5.6±3.8 years (range, 19 to 36 years). Of the 31 subjects, only two were smokers, albeit very light ( <.5 cigarettes/day, <.5 packyears). Mean body weight (±SD) was 77.3±9.6 kg for the men and.56±4.3 kg for the women. Body mass index was 23.7±2.1 kglm 2 for the men and 20.9± 1.2 kglm 2 for the women. None of the subjects had a history of respiratory or cardiac disease, an upper airway disorder, or thoracic or upper abdominal surgery. None had suffered from an upper respiratory tract infection in the 4 weeks prior to the time of testing. All subjects gave informed consent and the study protocol was approved by the Ethics Committee of our hospital. Flow-volume loops were measured using the SensorMedics 2130 d1y-seal spirometer (SensorMedics; Bilthoven, The Netherlands) in four different body postures: sitting (erect), supine (dorsal recumbency), and right and left lateral recumbency. For every subject, the order of the four body postures was randomized. In each body posture, FVLs were obtained repeatedly until the sum of FEV 1 and FVC for at least three maneuvers was within.5% of the best sum, and until the PEF and PIF of at least three maneuvers were within 10% of the best PEF and PIF. Special care was taken to maintain a constant and similar position of the neck (flexion/extension) and of the head (rotation) in relation to the trunk in the various body postures: in each of the four body postures, the neck was held in neutral position (neither in flexion nor in extension) and the subject was looking straight ahead. Each subject rested for 10 min between body postures to avoid fatigue, but all measurements were obtained in one session lasting 60 to 90 min. Although some subjects were not naive to spirometry, all were carefully instructed in the correct execution of forced inspiratory and forced expiratory maneuvers and all were given a training tjial. The spirometric and FVL parameters (FVC, FEV 1, FEV PEF, FEF.; 0, PIF, and FIFs 0 ) used for analysis were not obtained from the single best overall FVL. Instead, for each parameter in each body posture we used the highest value obtained dming any of the various efforts in that body posture. From these parameters, the following UAO indices were calculated: the FEV/ FEV 0. 5 ratio (unitless), the FEV/PEF ratio (in muumin), the PEF/PIF ratio (unitless), and the FEF s</fif 50 ratio (unitless). In ll subjects, respiratory muscle strength was also evaluated in the four different body postures by measuring the maximal static inspiratmy mouth pressure (MIPS) and the maximal static expiratmy mouth pressure (MEPS) with a uniflow disposable pressure transducer (ALL.5.5 UF model; Baxter Healthcare Corp, Edwards Critical Care; Irvine, CA) equipped with a Hewlett-Packard carrier amplifier (H P; Brussels, Belgium) according to a technique described earlier 7 MIPS was measured during a Mueller maneuver at residual volume and MEPS was measured during a Valsalva maneuver at total lung capacity. Statistical l v l e t h o d ~ Parameters are expressed as mean values±sd. Analysis of variance (AN OVA) for repeated measures was used to determine the effect of body posture on the various parameters and UAO indices. The Bonferroni t test (multiple comparison against a single control) was used to compare the differences between body postures if ANOVA reached statistical significance. A difference was considered to be statistically significant when p< Linear regression analysis was used to examine correlations between the changes in PEF and MEPS, or between the changes in PIF and MIPS, when subjects assumed a new body posture. RESULTS Spirometric and FVL Parameters Table 1 shows that the mean FVC, FEVI> FEV 05, PEF, FEF 50, FIF 50, and PIF all decreased slightly but statistically significantly in the three recumbent postures (supine, left lateral, right lateral) compared with the sitting (erect) posture (p<0.05). The recumbency-induced reductions in these parameters varied between -4.6 and -10.4% compared to the sitting posture, and were similar in magnitude for the three recumbent postures. UAO Indices (Flow Ratios) No significant differences in mean FEV /FEV 05, PEF/PIF, and FEF. 5 offif 50 ratios were observed among any of the four body postures (Table 2). The only statistically significant change occurred in the mean FEV/PEF ratio (expressed in muumin), which increased in both the right and left lateral recumbent postures compared with the sitting posture (ANOVA: p<0.0001; Bonferroni test: p<0.05). The mean changes of the UAO indices in the three recumbent postures compared with the sitting posture were small, ranging from -2.8 to +4.4% (95% confidence intervals, to +8.48%) (Table 3). None of the 31 subjects developed an inspiratory CHEST I 114 I 4 I OCTOBER,

3 Table!-Forced Inspiratory and Expiratory Parameters and Respiratory Muscle Strength in Four Body Postures* Right Left Lateral Lateral Parameters Sitting Supine Recumbent Recumbent Forced vital capacity, L.5.23 (115) 4.99 (l.07)t 4.99 (l.09)t 4.87 (l.08)t -4.6% -4.6% -6.9% FEV 1, L 4.42 (0.87) 4.13 (0.80)t 4.15 (0.79)t 4.13 (0.80)t -6.6% -6.1 % -6.6% FEV 0 5, L 3.38 (0.67) 3.15 (0.64)t (0.63)t 3.14 (0 64)t -6.8% -6.8% % PEF, Us (2 99) 9.32 (222)\ 9.19 (2. 18)t 8.98 (2.08)t -70% -8.3% -10.4% FEF. 51, Us.5.7 (1.53) 5.16 (l.55)t 5.20 (l.49)t 5.18 (l.44)t -9.5% -8.8% -9.1% FIF 50, Us 707 (1.85) 6.61 (l73)t 6.40 (l.69)t 6.39 (l.52)t -6.5% -9.5% -9.6% PIF, Us 7.34 (1.80) 7.0 (l.64)t 6.82 (l.61)t 6.75 (l.5)t -4.6% -7.1% -8 0% MIPS, -em H (27 7) 97.2 (25.6)t 99.6 (31. 4) 99.6 (296) -8.3% -6.0% -6 0% MEPS, em H (40.9) (37) (30.8) (33.3) -2.0% -1.7% -5.1% *Data expressed as mean (SD ) and as percent change (in italics) fi om the sitting posture. MIPS and MEPS were measured in 11 subjects, all other variables in 31 subjects. fsi gnificantly decreased (p< ) compared with the sitting posture. or an expiratory flow plateau on their FVLs in any of the three recumbent postures. Small sex differences were observed for some UAO indices. In men, the only statistically significant change with recumbency was an increase of the FEV /PEF ratio in the right lateral recumbent posture compared with the supine (not sitting) posture (repeated measures ANOVA, p=0.01 ), whereas in women the FEY/ PEF ratio increased and the PEF!PIF ratio decreased in the right lateral recumbent posture compared with the sitting posture (repeated measures ANOVA, p=0.008 and p=o.ol5, respectively). MIPS and MEPS MEPS did not change significantly with any of the recumbent postures compared with the sitting posture (Table 1). MIPS decreased significantly only in the supine posture compared with the sitting posture (ANOVA: p<0.0001; Bonferroni test: p=0.04). Linear regression analysis did not reveal a statistically significant correlation between the posture-related changes in MIPS and PIF, nor between the posturerelated changes in MEPS and PEF. DISCUSSION The results of this study show that in normal subjects each of the measured maximal inspiratory and expiratory flow rates decreased by 4.6 to 10.4% when they assumed any of the recumbent postures compared with measures taken when they were in the sitting posture. However, the UAO indices did riot change significantly with recumbency, except for a slight increase of the FEV /PEF ratio in both lateral recumbent postures. Furthermore, upon recumbency none of the subjects developed a flow plateau on the maximal inspiratory and expiratory flow volume curves. Table 2-UAO Indices in Four Body Postures* Parameters Sitting Supine Left Lateral Recumbent Right Lateral Recumbent FEV/FEV 0 5 FEV/PEF, muumin FEF 5 ifif 50 PEF/PIF 1.31 (0.08) 7.43 (0.71) 0.84 (0.24) 1.39 (0.23) 1.31 (0.07) 7.48 (0.77) 0.81 (0.23) 1.35 (0.21) 1.32 (0 06) 1.32 (0.06) 7.64 (0 77)t 7.74 (0.66)t 0.85 (0.28) 0.84 (0.26) 1.37 (0.28) 1.35 (0.27) *Data expressed as mean (SD); n=3l. tp< vs sitting posture Clinical Investigations

4 ~ ~ ~ ~ Table 3-Changes in UAO Indices From the Sitting Posture to the Three Recumbent Postures* Sitting to Supine Sitting to Left Lateral Sitting to Right Lateral FEV/ FEV 0, (3.83) +0.8 (3.88) +0.8 (3.54) -1.07to to to FEV / PEF +0.8 (529) +2.9 (5.73) +4.4 (5.95) to to to FEF 5 </ FIF (13.22) +2 (1773) +2.1 (16.54) to to to PEF/PIF -2.3 (997) (12.47) (12.48) to to to *Data expressed as mean percentage change (SD); 95% confidence intervals a re give n in italics. Few data are available on the effect of recumbency on the spirometric and FVL parameters in young, nonobese normal subjects such as ours. Masumi et al 8 found a significant decrease in the forced inspiratory flow measured between 25% and 75% of vital capacity (the only parameter they measured, a rather unusual one) upon recumbency in nine nonobese normal subjects. More, albeit conflicting, literature data are available for obese patients who either snore or have obstructive sleep apnea syndrome (OSAS). Shepard and Burger, 9 for example, found a modest but significant reduction in the forced expiratory flow rates but no change in the forced inspiratory flow rates of 14 obese patients with OSAS when changing from the upright to the supine posture. Although Shepard and Burger 9 also studied 14 nonobese control subjects, their article does not provide the data for these control subjects when recumbent. Nahmias and Karetzy, 10 by contrast, did not find significant changes in FEV 1 or FVC (the only two reported parameters) in 20 obese snorers and 28 obese OSAS patients studied in the upright and supine postures. They did, however, report that an increased proportion of OSAS patients showed flattening of the maximal inspiratory flow volume curve (ie, a reduction in forced inspiratory flow rates) when tested in the supine posture. In accordance with Shore and Millman,ll they concluded that addition of supine FVLs increases the sensitivity of the FVL for detecting UAO in patients with the OSAS. In our group of young, nonobese normal subjects, the forced inspiratory and expiratory flow rates were lower in the three recumbent postures than in the sitting posture. This finding cannot be attributed to a systematic effect of fatiguing since, by design, the order in which the four body postures were tested was randomized. Furthermore, the subjects were allowed to rest between the four body postures. Could posture-induced changes in available respiratory muscle strength have caused the observed reduction in the effort-dependent flow rates (ie, PEF, PIF, and FIF 50 ) upon recumbency? Indeed, respiratory muscle strength, which can be estimated by measuring maximal static mouth pressures (MIPS and MEPS), is an important determinant of these effort-dependent flow rates, as shown by Fry and Hyatt's isovolume pressure-flow curves. 12 Both MIPS and MEPS have been shown to decrease in normal and obese subjects in the supine posture vs the sitting posture, 13 and normal subjects generate significantly less transdiaphragmatic pressure in the supine posture than in the sitting posture, 14 suggesting that at least the diaphragm's strength is affected by the supine posture. To examine the possibility that posture-induced changes in respiratory muscle strength were responsible for the observed posturerelated changes in effort-dependent flow rates, we measured MIPS and MEPS in 11 of the 31 subjects in the sitting and the three recumbent postures. Both MIPS and MEPS decreased in the recumbent postures (MIPS by 6.0 to 8.3% and MEPS by 1.7 to 5.1 %) compared with the sitting posture. However, this change was significant only for MIPS in the supine posture, and no significant correlation was found between the changes in MIPS and the changes in the effort -dependent maximal inspiratory flow rates (PIF or FIF 50 ). Similarly, no correlation was found between the changes in MEPS and the changes in PEF. Another mechanism that could explain the reduction in the effort-dependent maximal inspiratory and expiratory flow rates is an increase in the resistance of the upper airway on assuming any of the recumbent postures. It has indeed been shov.rn. that in normal subjects as well as in patients with OSAS,l7 pharyngeal size (but not the glottic or tracheal size) 15 decreases significantly in the supine posture compared with the upright posture. This occurs independently of posture-related changes in lung volume (functional residual capacity) 15 and is attributed to gravitational forces acting on the tongue and soft palate.l 6 Although this extrathoracic mechanism could contribute to the observed reduction in forced inspiratory flow rates, it is less likely to cause CHEST I 114 I 4 I OCTOBER,

5 Table 4-Upper Limit of the Posture-Related Changes in the UAO Indices for Young, Nonobese Normal Subjects* Sitting to Sitting to Sitting to Supine Left Lateral Right Lateral ll FEV/FEV % +12.2% +6.5% Ll FEV/PEF +10.1% % % ll FEF.sJFIF % +38.5% +30.1% ll PEF/PIF % +21% +24.3% *The upper limit of the expected change between the sitting posture and each of the three recumbent postures is expressed as the 95th percentile of the measured changes. the observed reduction in the forced expiratory flow rates, in particular the less or non-effort-dependent FEV 1 and FEF 50. Differences in the position of the neck may, by altering the longitudinal tension on the trachea and hence its stiffness, affect maximal expiratory flow rates.l 8 However, we avoided this by carefully keeping the relation of the. head and neck to the trunk with regards to both neck flexion-extension and head rotation, constant in the four body postures. Finally, the most probable explanation for the recumbency-induced changes in flow rates is a decrease in lung volume. Recumbency produces a significant reduction in total lung capacity, residual volume, and vital capacity; these changes are small (in the order of 10% or less) and are related mainly to an increase in intrathoracic blood volume Since airway caliber is clearly dependent on lung volume, it is to be expected that flovv rates also decrease with decreasing lung volume in the recumbent postures. In our study, the vital capacity decreased by 4.6 to 6.9% (an order of magnitude in accordance with data in the literature); accordingly, the PEF, for example, decreased by 7.0 to 10.4%, depending on the various recumbent postures. Recently, Elliot et al 21 demonstrated in well-trained astronauts that there is a correlation between the reduction in forced vital capacity and the reduction of PEF and FEF. 50 in the supine posture. The mean values of the UAO indices (flow ratios) observed in the sitting posture in our 31 normal subjects compare well to those previously reported. 4..'5 Although all measured inspiratory and expiratory flow rates decreased significantly in the three recumbent postures compared with the sitting posture, none of the UAO indices derived from these flow rates changed significantly in any of the three recumbent postures, except for the FEV/PEF ratio, which increased significantly (but only slightly) in the left (2.9%) and right (4.4%) lateral postures but not in the supine posture. This indicates that in both lateral recumbent postures, the PEF decreased pro portionally more than the FEVI> suggesting that upper airway resistance is higher in lateral postures than in the supine posture (since respiratory muscle strength was shown to be similar). In support of this speculation, Behrakis et al 22 found pulmonary flow resistance (RJ to be increased in 10 healthy young adults when shifting from a sitting to lateral posture, without further increase of RL in the supine posture in spite of a 35% decrease in expiratory reserve volume from the lateral to the supine posture. Since Vincent et al 23 showed that lower airways resistance increases with decreasing lung volume, Behrakis et al 22 reasoned that the lack of change in RL when shifting from the lateral to the supine posture suggested greater upper airways resistance in the lateral posture than in the supine posture. This greater upper ainvay resistance, then, would affect the flow rates at high lung volume (such as the PEF) to a larger extent than the less effort-dependent flow rates (such as the FEV 1 ). Hence, the FEV/PEF ratio can be expected to be higher in the lateral posture than in the supine posture. Indeed, flow rates at high lung volume, such as the PEF, are more strongly affected by changes in upper ainvay resistance than is the FEV 1 as shown by Miller and Hyatt.l 2 All other UAO indices (including the FEV/PEF ratio ), did not change with recumbency, indicating that the measured flow rates constituting these various flow ratios decreased proportionally with recumbency. Hence, the indices commonly used to detect U AO in the erect posture can also be used to indicate the absence or presence of UAO in the supine and both lateral recumbent postures. Although the mean posture-related changes of the various UAO indices were small (95% confidence intervals ranged between and +8.48%), the standard deviations in Table 3 show that the individual posture-related changes can be quite substantial, more so for the FEF 50 /FIF 50 and the PEF/PIF ratios than for the FEV/FEV and FEV/PEF ratios. In Table 4, the upper limit of the change (in %) of each UAO index between the sitting posture and each of the three recumbent postures in our young, normal subjects is shown. When an individual subject of comparable age and body size as our study population is found to reveal posture-related changes greater than these expected changes (as represented by the 95th percentile in Table 4), the presence of a recumbency-induced UAO, and hence the presence of an upper airway problem (unsuspected in the sitting posture) might be suspected. In conclusion, in this study we have shown that measurement of FVLs is feasible in various recumbent postures. Although recumbency decreased the maximal flow rates (presumably as a result of a Clinical Investigations

6 decrease in lung volume), the mean values of the commonly used UAO indices (flow ratios) did not change, except for a slight increase of the mean FEY /PEF ratio in the lateral postures. Taking into account the upper limits of individual recumbencyrelated changes in U AO indices reported here, measurement of the FVL in different postures may increase its sensitivity for detecting an upper airway lesion or assessing its effect on upper airway patency. Such an assessment might be useful in decisionmaking, eg, in deciding whether a patient with a goiter should undergo surgery, to document the return of integrity of the upper airway following goiter surgery, 6 24 after weight loss in obese patients with OSAS, 25 or after treatment of patients with pseudo-asthma or laryngeal dyskinesia. 26 ACKNOWLEDGMENT: The authors wish to thank the technical staff of the pulmonary function laboratmy and Mrs. Hilde De Smedt and Christine Van Hauwermeiren for preparing the manuscript. REFERENCES 1 Miller RD, Hyatt RE. Obstructing lesions of the larynx and trachea: clinical and physiologic characteristics. Mayo Clin Proc 1969; 44: Miller RD, Hyatt RE. Evaluation of obstructing lesions of the trachea and larynx by flow-volume loops. Am Rev Respir Dis 1973; 108: Vincken W, Cosio MG. Flew oscillations on the flow-volume loop: a non-specific indicator of upper airway obstruction. Bull Eur Physiopathol Respir 1985; 21: Rotman HH, Liss HP, Weg JG. Diagnosis of upper aiiway obstruction by pulmonary function testing. Chest 1975; 68: Empey DW. Assessment of upper airways obstruction. Br Med J 1972; 3: Meysman M, Noppen M, Vincken W. Effect of posture on the flow-volume loop in two patients with euthyroid goiter. Chest 1996; 110: Vincken W, Ghezzo H, Cosio MG. Maximal static respiratory pressures in adults: normal values and their relationship to determinants of respiratory function. Bull Eur Physiopathol Respir 1987; 23: Masumi S, Nishigawa K, Williams AJ, et al. Effect of jaw position and posture on forced inspiratory airflow in normal subjects and patients v. ~ obstructive t h sleep apnea. Chest 1996; 109: Shepard JW Jr, Burger CD. Nasal and oral flow-volume loops in normal subjects and patients with obstructive sleep apnea. Am Rev Respir Dis 1990; 142: Nahmias J, Karetzy MS. Upright and supine flow-volume curves in patients with OSA. N J Med 1989; 86: Shore ET, Millman RP. Abnormalities in the flow-volume loop in obstructive sleep apnea sitting and supine. Thorax 1984; 39: Fry DL, Hyatt RE. Pulmonary mechanics: a unified analysis of the relationship between pressure, volume and gasflow in the lungs of normal and diseased human subjects. Am J Med 1960; 29: Fiz JA, Aguilar X, Carreres A, et al. Postural variation of the maximum inspiratory and expiratory pressures in obese patients. Int J Obes 1991; 15: Koulouris N, Mulvey DA, Laroche CM, et al. The effect of posture and abdominal binding on respiratory pressures. Eur Respir J 1989; 2: Fouke JM, Strohl KP. Effect of position and lung volume on upper airway geometry. J Appl Physiol 1987; 63: Jan MA, Marshall I, Douglas NJ. Effect of posture on upper airway dimensions in normal humans. Am J Respir Crit Care Med 1994; 149: Brown IB, McClean PA, Boucher R, et al. Changes in pharyngeal cross-sectional area with posture and application of continuous positive airway pressure in patients with obstructive sleep apnea. Am Rev Respir Dis 1987; 136: Melissinos CG, Mead J. Maximum expiratory flow changes induced by longitudinal tension on trachea in normal subjects. J Appl Physiol: Respir Environ Exerc Physiol 1977; 43: Agostoni E, D'Angelo E. Statics of the chest wall. In: Roussos C, Macklem PT, eds. The thorax. New York: Marcel Dekker, 1985; Agostoni E, Hyatt RE. Static behavior of the respiratory system. In: Fishman AP, Macklem PT, Mead J, eds. Handbook of physiology. Section 3: The respiratory system. Volume III. Mechanics of breathing, part l. Bethesda, MD: Ame1ican Physiological Society, 1986; Elliot AR, Prisk GK, Guy HJB, e t al. Forced expirations and maximum expiratory flow-volume curves during sustained microgravity on SLS-1. J Appl Physiol 1996; 81: Behrakis PK, Baydur A, Jaeger MJ, et al. Lung mechanics in sitting and horizontal body positions. Chest 1983; 83: Vincent MJ, Knudson R, Leith DE, e t al. Factors influencing pulmonary resistance. J Appl Physiol 1970; 29: O'Donnell T, Karetzky M, Brief DK, et al. Treatment of upper airway obstruction associated with goiter. N J Med 1993; 90: Suratt PM, McTier RF, Findley LJ, et al. Changes in breathing and the pharynx after weight loss in obstructive sleep apnea. Chest 1987; 92: Martin RJ, Blager FG, Gay ML, et al. Paradoxic vocal cord motion in presumed asthmatics. Semin Respir Med 1987; 8: CHEST /114 I 4/ OCTOBER,