Comparison of the oxygen cost of breathing exercises and spontaneous breathing in patients with stable chronic obstructive pulmonary disease. (Research Report).

Due to their increased ventilatory demands, people with chronic obstructive pulmonary disease (COPD) have a higher resting oxygen consumption ([Vo.sub.2]) than do people without pulmonary disease. (1,2) This higher resting [Vo.sub.2] may be explained by increased mechanical work of breathing or reduced ventilatory muscle efficiency, or both, in patients with severe COPD. (3) Mechanical work in biological systems is achieved when a force applied to a structure results in movement. The amount of work produced is the product of the force applied and the distance moved. With respect to spontaneous ventilation, the mean pressure generated by the ventilatory muscles is equivalent to the force, and the tidal volume is equivalent to the distance. Ventilatory muscle work is primarily dependent on minute ventilation, airway resistance, and lung compliance, which, in turn, determine ventilatory muscle efficiency and the pattern of breathing. (4) Efficient ventilatory muscle function is equivalent to the attainment of the requisite minute ventilation with the least energy cost.

In COPD, the loss of alveolar tethering and elastic recoil contributes to increased lung compliance and impaired lung perfusion. (5) Typically, these changes result from prolonged inhalation of cigarette smoke, which irritates the airways, thus increasing mucous production and airway resistance. Over time, these pathophysiologic changes contribute to increased anatomical dead space in the lungs and overall total lung capacity. (6) Consequently, hyperinflation of the lungs and the chest wall causes the hemidiaphragms to become depressed, which contributes further to breathing inefficiency and increased metabolic cost. In combination, these changes lead to the abnormal blood gases characteristic of COPD, specifically, hypoxemia and hypercapnia. (6)

To help relieve the symptoms and physical limitations of patients with COPD, physical therapists have taught breathing exercises in the form of diaphragmatic breathing (DB), pursed-lip breathing (PLB), or a combination of these 2 patterns (CB). (1,2,7) The effects of conventional breathing exercises reported in the literature, however, have been inconsistent, and the benefit of this type of breathing thus is unclear. For example, PLB has been associated with increased tidal volume and arterial oxygenation, as well as a reduction in respiratory rate (RR), (8-12) yet PLB has been reported to increase alveolar air trapping and the work of breathing. (13) Some investigators (12,14,15) have attributed benefits of breathing control exercises to reduced RR, increased tidal volume, and improved alveolar ventilation. Gosselink and colleagues (7) provided compelling evidence that DB reduces rather enhances breathing efficiency in people with severe COPD. These investigators concluded that DB contributed to inappropriate chest wall motion and decreased mechanical efficiency while increasing dyspnea. Furthermore, DB has been reported to provoke post-hyperventilation hypoxemia. (16) A recent review of the literature by Cahalin and colleagues (17) has confirmed that reports on the effects of DB are equivocal. Despite the varying findings in the literature with respect to the effects of breathing exercises, controlled breathing maneuvers appear to be widely used in the physical therapy management of patients with COPD, both individually and in pulmonary rehabilitation programs. (18)

The fact that imposed breathing patterns are not main, rained by patients is an interesting question that has not been studied. A better understanding of breathing patterns in people without pathology and in patients who are prone to ventilatory distress, such as those with COPD, could yield information about enhancing noninvasive physical therapist interventions used to reduce the mechanical work of breathing and to improve ventilatory efficiency.

A better physiological understanding of spontaneous breathing in patients with COPD, in relation to responses to breathing exercises could help elucidate what role, if any, the teaching of breathing exercises has in the management of this prevalent condition. To describe the physiological responses of spontaneous breathing and breathing exercises, we chose to examine changes in [Vo.sub.2] to assess the effect on oxygen demands and to examine changes in RR to assess the impact on breathing pattern. Therefore, the purpose of this study was to compare [Vo.sub.2] and RR during 3 commonly prescribed breathing exercises (DB, PLB, and CB) with [Vo.sub.2] and RR during spontaneous breathing at rest (SB).

Method

Subjects

The primary inclusion criteria for the study were that participants had to meet the inclusion criteria of an established pulmonary rehabilitation program and that they had to be medically stable at the time of the study. Patients with COPD who had completed a 6-week pulmonary rehabilitation program (based on the standards of the American Thoracic Society) (18) affiliated with a regional hospital were invited to participate in the study and constituted a sample of convenience. All participants had learned and were experienced in performing 3 breathing exercises (ie, DB, PLB, and CB). The participants had been taught these maneuvers by the same physical therapist (CC). Proficiency in their performance was based on the judgment of that physical therapist, who had 6 years of experience in respiratory medicine. All participants as assessed by the physical therapist based on the inclusion and exclusion criteria, were medically stable for at least 4 weeks, were able to walk independently without aids, and had no comorbidity that would interfere with the dependent variables of interest or their ability to adhere to the study procedures. None of the participants smoked at the time of the study.

Spirometry, including measurement of forced vital capacity (FVC) and forced expiratory volume in 1 second (FE[V.sub.1]), was undertaken prior to the study. These measurements were obtained using a hand-held spirometer and the procedures of the American Thoracic Society. (19) Patients with cardiac, metabolic, or endocrine disorders; an acute exacerbation of COPD; acute chest infection; or surgery within the last 2 months were excluded from the study, as were patients with a diagnosis of cancer or active gastrointestinal problems and those requiring supplemental oxygen. Patients also were excluded if they reported that the enclosure of the canopy apparatus used for the metabolic measures would contribute to claustrophobia. To eliminate the effects of varying medications, we also selected participants who would not be harmed by delaying any scheduled morning medication. Based on the study design, no participant would have been denied required medication.

The participants' descriptive data are presented in Table 1. Thirty patients (6 women and 24 men) with stable, moderately severe COPD (mean age=68.5 years, SD=7.83, range=46-84) participated in the study. Two patients declined participation in the study due to difficulty in arranging transportation from their home to the medical center. All subjects were informed of the purpose of the study, and written consent was obtained.

Procedure

Measurement of [Vo.sub.2] was performed at the same time in the morning and in the same quiet laboratory setting under controlled lighting and temperature conditions for all subjects. Subjects fasted after the previous evening meal until after the measurements were obtained the following morning. Subjects were asked not to perform any vigorous exercise the day before testing. They also were asked to use transportation to the laboratory to reduce ambulatory effort and fatigue and to wear comfortable, loose clothing. After arrival at the laboratory, subjects rested in the supine position for 30 minutes before we obtained baseline measurements of variables during SB (ie, [Vo.sub.2], RR, heart rate [HR], and pulse oximetry [[Spo.sub.2]]). Subjects were instructed to remain relaxed and not to talk or sleep while the measurements were being taken. They also were instructed to indicate to the investigator (CC) if they felt cold or anxious. Subjects remained supine throughout the data collection.

Oxygen consumption was measured with the MedGraphics Critical Care Management System (MCCMS) * using expired gas analysis. (20) The system consists of a 30-L plexiglass ventilated hood (canopy), gas analyzers, and a pneumotachograph. We selected the use of the canopy method to measure [Vo.sub.2] given that the use of alternate methods requiring a mouthpiece and noseclip or a mask would alter both SB and the performance of the learned breathing patterns. Use of the canopy method necessitated that subjects remained in the supine position. The MCCMS was calibrated before measurements of [Vo.sub.2] and carbon dioxide production were taken from each participant. We used the manufacturer's recommended procedures.

The investigator (CC) was responsible for the calibration of the metabolic measurement cart and measurements. Based on the calibration for gas volume and concentration, which, in turn, was based on known gas volumes and gas concentrations, we believed our measurements were valid. Standardized calibration procedures and reference gases were used for analysis. The calibration gas concentration of 4% to 7% C[O.sub.2] and 12% to 16% [O.sub.2] were used, with an accuracy of [+ or -] 0.03%. The reference gas also was specified to 0.03% accuracy, with 21% [O.sub.2] and [N.sub.2] as the balance gas. The reliability of the measurements was based on a sample of an additional 9 patients with COPD for whom we obtained an intraclass correlation coefficient (1,1) of .97 (95% confidence interval=.88-.99). These data were collected at the beginning of the study. As a safety precaution and to establish that the subjects were not aroused, HR and [Spo.sub.2] were continuously monitored during the interventions using a pulse oximeter (Nellcor N-20PA) ([dagger]) with the low [Spo.sub.2] alarm set at 90%.

The canopy was positioned over the subject's head. The subject was asked to breathe normally. The [Vo.sub.2] and RR were continuously displayed on the screen of the MCCMS. When expired C[O.sub.2] stabilized ([+ or -] 2%), data for the following 10 minutes were collected. The order of the breathing exercises was randomly allocated for each subject. The procedures used for each breathing exercise and the instructions that were given to the subjects are as follows.

Diaphragmatic breathing (21) occurs when there is a conscious appreciation of inspiring air to the lung bases with slight forward abdominal displacement and passive relaxed expiration. The instruction given to the participants was "breathe in slowly through your nose and aim at getting the air to the lower part of your lungs; remember to relax your tummy and allow the air to go under here [the investigator put his hand on the subject's epigastric/subcostal region]. Then relax and let all air out through your mouth, allowing your tummy to sink gently."

Pursed-lip breathing (8) in our study consisted of each subject's normal pattern of inspiration, but expiration was performed by gently blowing through "pursed" lips. The instruction we gave was "breathe in through your nose and exhale by blowing gently against your loosely closed lips, like blowing a candle flame so that it bends but doesn't blow out."

The CB pattern (8,21) required subjects to have some slight forward abdominal displacement during inspiration and during expiration through pursed lips. The instruction given was "breathe in through your nose, aim at getting the air to the lower part of your lungs; allow the air to go under here [investigator's hand over the subject's epigastric/subcostal region] and breathe out by gently blowing against your loosely closed lips, as if you are blowing a candle flame so that it bends but doesn't blow out."

Participants were observed while performing the breathing exercises to ensure that each exercise was performed as instructed. No other verbal instruction or encouragement was given during the recording. After the [Vo.sub.2] stabilized following adoption of the first randomly allocated breathing exercise, data were collected for 10 minutes. The subjects then resumed the breathing pattern they normally used until [Vo.sub.2] returned to stable baseline levels during SB. We monitored the variation in C[O.sub.2] production during the resumption of spontaneous breathing and considered it to be stable when the curve returned to the x-axis for 60 seconds and the [Vo.sub.2] fluctuated less than 1%. This procedure was then repeated for each of the 2 remaining breathing exercises sequenced according to their randomized order.

Data Analysis

A pilot study of an additional 9 subjects was used to determine effect size ([treatment mean--control mean]/SD). With 30 patients in the group, this study had at least 80% power to detect a [Vo.sub.2] effect size of 0.7 between the breathing pattern data and baseline spontaneous breathing data at the 5% significance level (Power Analysis and Sample Size for Windows, version 6.0 ([double dagger])). We selected a high effect approximating 0.8 based on the recommendations of Cohen (22) to maximize the potential clinical importance of the findings.

The [Vo.sub.2] and RR measurements recorded during the 3 breathing exercises were analyzed using a one-way analysis of variance for repeated measures. Linear contrast analysis was then applied to identify differences among means, (23) and individual alpha levels were adjusted using the modified Bonferroni method to minimize Type I error due to multiple testing. (24) SPSS for Windows (version 9.0.0) ([section]) was used for the statistical analysis, and a probability value of less then .05 was considered statistically significant.

([double dagger]) NCSS Statistical Software, 329 North 1000 East, Kaysville, UT 84037.

([section]) SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

Results

The results of spirometry are summarized in Table 1. The mean ratio of FE[V.sub.1]/FVC, an index of airway obstruction and thus disease severity, was 42.6%. This ratio falls within the range of 44% down to 35%, which is consistent with moderate disease severity. (25) The HR remained stable (ie, remained with [+ or -] 2 beats per minute) after application of the canopy over the subjects' head, indicating that the canopy did not elicit arousal. The [Spo.sub.2] for all subjects remained above 90% throughout the experimental procedure.

The mean RR ([+ or -] SD) recorded during SB (17.3 [+ or -] 4.23 breaths/min) was higher than that recorded when a breathing exercise was adopted. Specifically, the mean RR was 15.0 [+ or -] 4.32 breaths/min during DB, 12.8 [+ or -] 3.53 breaths/min during PLB, and 11.2 [+ or -] 2.7 breaths/min during CB. Furthermore, CB was associated with the least RR, followed by PLB and DB (Tab. 2).

The mean [Vo.sub.2] measurements recorded during SB and each of the breathing exercises also are displayed in Table 2. The mean [Vo.sub.2] was consistently lower during the breathing exercises than during SB. Specifically, the mean [Vo.sub.2] during the breathing exercises was 165.8 [+ or -] 22.3 mL [O.sub.2]/min for DB, 164.8 [+ or -] 20.9 mL [O.sub.2]/min for PLB, and 167.7 [+ or -] 20.7 mL O2/min for CB compared with 174.5 [+ or -] 25.2 mL [O.sub.2]/min for SB (Tab. 2). There were no differences in mean [Vo.sub.2] among the 3 breathing exercises.

Discussion

Patients who had completed an established pulmonary rehabilitation program were selected for the study so that we could examine the effects of imposed breathing patterns in patients who might be considered by an experienced physical therapist to be well practiced and proficient in these exercises. We believe that the breathing patterns learned in our sample were representative of those patterns taught by physical therapists to patients with COPD.

The finding that RR can be reduced in patients with stable COPD when performing established breathing exercises was consistent with earlier reports, (11,26) and this reduced rate was associated with a commensurate reduction in [Vo.sub.2]. The normally increased oxygen cost of spontaneous breathing in patients with COPD reflects increased airway resistance and increased lung compliance due to loss of elastic recoil of the lung parenchyma. With progressive impairment, ventilatory [Vo.sub.2] increases Correspondingly. (27) As dead space ventilation increases with air trapping and airway obstruction, minute ventilation increases correspondingly to support gas exchange and in an attempt to normalize [O.sub.2] and C[O.sub.2] levels. To support increased minute ventilation, V[O.sub.2] can increase several-fold depending on the severity of disease. (26,27) Our data further demonstrated an increase in resting [Vo.sub.2] in our subjects with COPD as compared with individuals without COPD. (28)

Breathing at rest in people without pathology requires less than 5% of total [Vo.sub.2]. (29-31) This metabolic cost reflects the energy required to support the work of the ventilatory muscles to generate negative pressure in the thoracic cavity for a given combination of airway resistance and lung compliance. Normally, an individual adopts an optimal breathing pattern with the least mechanical stress and work. (32) The finding that both RR and [Vo.sub.2] return to baseline levels after a period of 10 minutes of adopting any 1 of the 3 breathing patterns is consistent with the preservation of mechanical efficiency at the expense of energetic efficiency. Physical therapists have focused on reducing RR, and therefore [Vo.sub.2], as opposed to focusing primarily on enhancing mechanical efficiency, a function of biomechanical work and [Vo.sub.2], which is more consistent with the outcome of this study.

Our findings indicate that people with COPD adopt a breathing pattern at rest that is not associated with the least [Vo.sub.2]. (31-33) The components of respiratory mechanics and respiratory muscle efficiency may play a greater role in supporting spontaneous breathing than simply minimizing [Vo.sub.2], which now appears to be an overly simplistic approach. These components may include optimizing ventilatory muscle fiber length, ventilatory muscle mechanics, and chest wall configuration. Such an approach also is consistent with the conclusions of Barach, (34) who reported reduced pulmonary ventilation and dyspnea secondary to increased diaphragmatic excursion as a result of body tilting. Further, blood gases and pH were favorably changed.

The work of breathing can be assessed with electromyography, with measures of pressure and flow, or with measures of [Vo.sub.2]. Of these, [Vo.sub.2] provides the most comprehensive picture and is easiest to measure. Thus, we selected [Vo.sub.2] as the dependent variable of interest in our study. Detrimental work of breathing results when there is a discrepancy among the inspiratory work, the expiratory work in the case of air trapping, and the capacity of the ventilatory muscle pump. This work of breathing is detrimental because it exceeds normal energy demands and thereby contributes to the clinical consequences of shortness of breath and reduced capacity to perform physical activity. Therefore, we believe that physical therapists should aim to reduce inspiratory and expiratory work, improve ventilatory muscle efficiency, or both, as a means of minimizing detrimental work of breathing. These variables may be preferentially manipulated with body position, or with body positioning coupled with breathing control as opposed to breathing exercises. Almost 30 years ago, Barachs (34) demonstrated that body position alone could be used to optimize breathing mechanics such that minute ventilation was reduced along with accessory muscle activity and dyspnea. According to Barachs (34) and Dean, (35) the position or intervention that optimizes ventilation and perfusion matching, thus reducing hypoxemia, is the position or intervention of choice.

Although imposing a learned pattern of breathing by means of teaching breathing exercises may reduce [Vo.sub.2] by reducing RR, neither is this improvement sustained nor is the pattern adopted by patients with respiratory insufficiency. Thus, optimizing respiratory muscle mechanics to reduce biomechanical work and improving ventilatory muscle efficiency may reduce oxygen cost, thus decreasing the work of breathing and potentially resulting in a sustained improvement. Manipulating RR does not take into account the role of the respiratory muscles in stabilizing ventilation (36) and vertebral support. (37)

The discordance between airflow limitation and gas exchange has been reported previouslyss, (38,39) and lends further support for focusing on the determinants of breathing pattern rather than attempting to manipulate breathing pattern. Further studies are needed to examine the effect of altering ventilatory muscle efficiency with and without altering RR. We contend studies are needed that examine factors that alter ventilatory muscle efficiency through their effects on optimizing respiratory muscle fiber length, lung volumes, chest wall configuration and movement, thoracoabdomical interaction, and lung function in people with COPD in addition to associated [Vo.sub.2]. The apparent discrepancy between optimal respiratory mechanics and the metabolic cost of breathing in patients with COPD warrants reconciliation and thereby elucidation of which factors in particular will yield the greatest benefit on respiratory mechanics.

Physiologically, the absolute changes in RR could be considered clinically important, that is, up to a 35% reduction; however, the corresponding decrement in [Vo.sub.2] was less than 10%. The patients we studied were medically stable and not experiencing an acute exacerbation of their COPD. Thus, we could expect that these effects would be accentuated during an exacerbation. Furthermore, these imposed breathing patterns associated with apparently lower oxygen costs were not sustained, and we believe this should be of considerable interest. We did not correlate dyspnea with our dependent variables given that the sense of effort that dyspnea represents is multifactorial, primarily reflecting mechanical as well as metabolic factors. Dyspnea, in part, reflects (1) afferent activity from muscle spindles and tendon organs associated with length tension and force velocity characteristics of the respiratory muscles and (2) chemoreceptor stimulation; this afferent activity and chemoreceptor stimulation contribute to an uncomfortable urge to breathe. (27,40)

Studies are needed to examine the determinants of the spontaneous breathing pattern of people with COPD and how these determinants change with progression of the disease. The results of such studies could shed light on those factors that can be influenced by physical therapy interventions to maximize overall ventilatory efficiency. Manipulating the determinants of respiratory mechanics in terms of respiratory work and metabolic cost might be possible.

Clinical Implications

Our results indicate that the rationale for teaching conventional breathing exercises is equivocal. The fact that breathing patterns associated with the lowest RR and oxygen cost compared with SB are not the preferred breathing patterns of people with moderately severe COPD is of considerable physiologic interest and potentially of considerable clinical importance. Our results indicate that an understanding of respiratory mechanics in conjunction with energetics and work of spontaneous breathing is needed to provide a basis for interventions designed to optimize breathing patterns in people with COPD. Variables other than RR and metabolic cost that are related to respiratory biomechanics and mechanical work play a role in enhancing the breathing efficiency of patients susceptible to respiratory insufficiency. Attention to variables that influence hypoxemia, such as ventilation and perfusion matching, could lead to the least ventilatory distress by directing attention to optimizing respiratory mechanics and thereby a sustained reduction in oxygen cost. Physical therapists, we believe, have tended to focus on the abnormal breathing pattern as the cause of a patient's breathing distress rather than viewing the breathing pattern as the effect of an impairment in respiratory mechanics. Viewed in this way, we contend that focusing on optimizing respiratory mechanics could result in a sustained, more normal breathing pattern, whereas altering the breathing pattern has not been associated with sustained reduction in oxygen cost of breathing and symptoms.

Although our subjects were medically stable and not in respiratory distress, the supine position in which they were studied often contributes to orthopnea in patients with COPD. (33) The relatively small changes in RR and [Vo.sub.2] observed in this clinical trial may be accentuated with increasing degrees of respiratory distress.

Our results indicate that a better understanding of the normal determinants of spontaneous breathing in people with lung disease could yield greater insight into how ventilatory distress can be minimized in people with COPD. Contrary to conventional approaches in which breathing pattern manipulation is believed to reduce breathing distress, our data support the findings of more recent studies suggesting that optimizing respiratory mechanics, in turn, will optimize the breathing pattern, the work of breathing, and the associated metabolic cost. Clinically, attention warrants being directed at the determinants of the abnormal breathing pattern, which is the effect of the underlying problem, rather than viewing the breathing pattern as the cause and attempting to manipulate it.

Conclusions

The effectiveness of conventional breathing exercises including DB and PLB is increasingly being called into question. In addition, the negative effects of these procedures have been reported. Our results do not indicate that the breathing frequency in people with COPD is selected to minimize the work needed for respiration.

Our study confirmed that DB, PLB, and CB resulted in lower oxygen cost, which can be explained by a commensurate reduction in RR. These effects, however, persist only while these exercises are being performed. Humans are thought to adopt a breathing pattern that is optimal in terms of biomechanical efficiency and gas exchange, a function of both mechanical work and ventilatory oxygen cost. Our results indicate that absolute oxygen cost is not the sole factor responsible for determining a person's SB pattern. Other variables related to respiratory biomechanics and mechanical work also play an important role in determining a person's SB pattern, and these variables need to be the focus of therapeutic intervention in addition to metabolic cost in enhancing the breathing efficiency of people who are susceptible to respiratory insufficiency. Studies are needed that focus on understanding the mechanics and the energetics of the spontaneous breathing pattern of people with COPD, and its determinants, to provide a rational basis for physical therapy management. Interventions that focus on optimizing respiratory mechanics may result in a better therapeutic outcome rather than a focus on breathing patterns that primarily may be the result of impaired respiratory mechanics.

Table 1.
Descriptive Data on Subjects (N=30) (a)

Subject Descriptors

Age (y)
  [bar]X                                                     68.5
  SD                                                          7.83
  Range                                                      46-84

Sex (male:female)                                             6:24

Body mass index (kg*[m.sup.2])
  [bar]X                                                     19.6
  SD                                                          3.28
  Range                                                      13.7-28.1

Pulmonary Function

FVC (L)
  [bar]X                                                      1.79
  SD                                                          0.54
  Range                                                       0.9-2.9
FVC (% predicted)
  [bar]X                                                     75.90
  SD                                                         22.84
  Range                                                      31.8-128.4
FE[V.sub.1] (L/s)
  [bar]X                                                      0.72
  SD                                                          0.22
  Range                                                       0.3-1.2
FE[V.sub.1] (% predicted)
  [bar]X                                                     39.0
  SD                                                         12.53
  Range                                                      15.5.-65.2
FE[V.sub.l]/FVC
  [bar]X                                                     42.6
  SD                                                         14.00
  Range                                                       0.20-0.75
FE[V.sub.1]/FVC (% predicted)
  [bar]X                                                     81.1
  SD                                                         15.49
  Range                                                      66.5-105.2

(a) FVC=forced vital capacity, FE[V.sub.1] =forced expiratory
volume in 1 second.

Table 2.

Descriptive Statistics for Subjects' Steady-Rate Oxygen Cost (Oxygen
Consumption [V[O.sub.2]]) and Respiratory Rate (RR) During Selected
Breathing Exercises (a)

Breathing                      RR              V[O.sub.2]
Exercise                       (Breaths/Min)   (mL [O.sub.2]/min)

Spontaneous breathing (SB)
  [bar]X                       17.3            174.50
  SD                            4.23            25.20
  Range                        11-29           127.0-247.0
Diaphragmatic breathing (DB)
  [bar]X                       15.0            165.80
  SD                            4.32            22.30
  Range                         8-24           112.0-205.0
Pursed-lip breathing (PLB)
  [bar]X                       12.8            164.80
  SD                            3.53            20.90
  Range                         7-20           125.0-211.0
Combined breathing
    exercises (CB)
  [bar]X                       11.2            167.70
  SD                            2.7             20.70
  Range                         5-17           126.0-218.0

(a) Mean V[O.sub.2]: PLB, DB, and CB not significantly different;
PLB, DB, and CB were less than SB; P<.05). Mean RR: CB, PLB, DB, and
SB all significantly different from each other, P<.05.

* Medical Graphics Corp, 350 Oak Grove Pkwy, St Paul, MN 55127.

([dagger]) Nellcor, 4280 Hacienda Dr, Pleasanton, CA 94588.

References

(1) Donahoe M, Rogers RM, Wilson DO, Pennock BE. Oxygen consumption of the respiratory muscles in normal and in malnourished patients with chronic obstructive pulmonary disease. Am Rev Respir Dis. 1989; 140:385-391.

(2) Schols AMWJ, Fredrix EWHM, Soeters PB, et al. Resting energy expenditure in patients with chronic obstructive pulmonary disease. Am J Clin Nutr. 1991;54:983-987.

(3) Rodarte JR, Hyatt RE. Respiratory mechanics: basics of respiratory disease. Am Thor Soc. 1976;4:1-6.

(4) Marini JJ. Work of breathing. In: Kacmarek RM, Stoller JK, eds. Current Respiratory Care. Philadelphia, pa: BC Decker Inc; 1988:188-194.

(5) West JB. Pulmonary Pathophysiology: The Essentials. 4th ed. Baltimore, Md: Williams & Wilkins; 1992.

(6) Juzar A, Summer WR, Levitzky MG, eds. Pulmonary Pathophysiology. New York, NY: McGraw-Hill Inc; 1999.

(7) Gosselink RAAM, Wagenaar RC, Rijswijk H, et al. Diaphragmatic breathing reduces efficiency of breathing in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 1995;151: 1136-1142.

(8) Ito M, Kakizaki F, Tsuzura Y, Yamada M. Immediate effect of respiratory muscle stretch gymnastics and diaphragmatic breathing on respiratory pattern. Intern Med. 1999;38:126-132.

(9) McConnell EA. Teaching pursed-lip breathing. Nursing. 1999;29:18.

(10) Mueller RE, Petty TL, Filley GF. Ventilation and arterial blood gas changes induced by pursed lip breathing. J Appl Physiol. 1970;28: 784-789.

(11) Sharp J, Drutz W, Moisan T, et al. Postural relief of dyspnea in severe chronic obstructive pulmonary disease. Am Rev Respir Dis. 1980;122:201-211.

(12) Thoman R, Stoker G, Ross J. The efficacy of pursed lips breathing in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis. 1966;93:100-106.

(13) Tiep BL, Burns M, Kao D, et al. Pursed lips breathing training using oximetry. Chest. 1986;90:218-221.

(14) Vitacca M, Clini E, Bianchi L, Ambrosino N. Acute effects of deep diaphragmatic breathing in COPD patients with chronic respiratory insufficiency. Eur Respir J. 1998; 11:408-415.

(15) Ingram R, Schielder DP. Effect of pursed-lips expiration on the pulmonary pressure flow relationship in obstructive lung disease. Am Rev Respir Dis. 1967;96:381-387.

(16) Casciari RJ, Fairshter RD, Harrison A, et al. Effects of breathing retraining in patients with chronic obstructive lung disease. Chest. 1981;79:393-398.

(17) Cahalin LP, Braga M, Matsuo Y, Hernandez ED. Efficacy of diaphragmatic breathing in persons with chronic obstructive pulmonary disease: a review of the literature. J Cardiopulm Rehabil. 2002;22:7-21.

(18) Lareau SC, Wallack RZ, Carlin B, et al. Pulmonary rehabilitation: the official statement of the American Thoracic Society. Am J Respir Crit Care Med. 1999;159:1666-1682.

(19) American Thoracic Society. Standardization of spirometry: 1987 update. Am Respir Dis. 1987;136:1285-1298.

(20) MedGraphics Critical Care Management System: Operator's Manual St Paul, Minn: Medical Graphics Corp.

(21) Pryor JA, Webber BA, Bethune D, et al. Physiotherapy techniques. In: Pryor JA, Prasad SA, eds. Physiotherapy for Respiratory and Cardiac Problems. 3rd ed. Edinburgh, Scotland: Churchill Livingstone; 2002: 182-186.

(22) Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.

(23) Steel RGD, Torrie JH, Dickey DA. Principles and Procedures of Statistics: A Biometrical Approach. 3rd ed. New York, NY: McGraw-Hill Inc; 1997:183-187.

(24) Hochberg Y. A sharper Bonferroni procedure for multiple tests of significance. Biometrika. 1988;75:800-803.

(25) Oakes D. Clinical Practitioner's Pocket Guide to Respiratory Care. Old Town, Minn: Health Education Publications Inc; 1991.

(26) Faling LJ. Pulmonary rehabilitation: physical modalities. Clin Chest Med. 1986:7:599-618.

(27) Field S, Kelly SM, Macklem PT. The oxygen cost of breathing in patients with cardiorespiratory disease. Am Rev Respir Dis. 1982;126:9-13.

(28) Casan P, Villafranca CC, Kearon C, etal. Contribution of respiratory muscle oxygen consumption to breathing limitation and dyspnea. Can Respir J. 1997;4:101-107.

(29) Cherniack RM. The oxygen consumption and efficiency of the respiratory muscles in health and emphysema. J Clin Invest. 1959;38: 494-499.

(30) Campbell EJM, Westlake EK, Cherniak RM. The oxygen consumption of the respiratory muscles of young male subjects. Clin Sci. 1959;18:55-64.

(31) Bartlett R. The oxygen cost of breathing. J Appl Physiol. 1958;12: 413-417.

(32) Shapiro BA, Harrison RA, Kacmarek RM, et al. Clinical Application of Respiratory Care. 3rd ed. Chicago, Ill: Year Book Medical Publishers Inc; 1985.

(33) McIllroy MB, Eldridge FL, Thomas JP, Christie RV. The effect of added elastic and non-elastic resistances on the pattern of breathing in normal subjects. Clin Sci. 1956;15:337-342.

(34) Barach AL. Chronic obstructive lung disease: postural relief of dyspnea. Arch Phys Med Rehabil. 1974;55:494-504.

(35) Dean E. Effect of body position on pulmonary function. Phys Ther. 1985;65:613-618.

(36) Sears TA, Mead J, Leith D, et al. The role of the diaphragm in stabilizing ventilation. Proc Int Union Physiol Sci. 1968;7:394-398.

(37) Hodges PW, Gurfinkel VS, Brumagne S, et al. Coexistence of stability and mobility in postural control: evidence from postural compensation for respiration. Exp Brain Res. 2002;144:293-302.

(38) Ross J, Bates DV, Dean E, Abboud RT. Discordance of airflow limitation and ventilatory homogeneity in asthma and cystic fibrosis. Clin Invest Med. 1992;15:97-102.

(39) Bates DV. Revisiting "respiratory function in emphysema in relation to prognosis." Can Respir J. 2000;7:271-279.

(40) Leblanc P, Bowie DM, Summers E, et al. Breathlessness and exercise in patients with cardiorespiratory disease. Am Rev Respir Dis. 1986;133:21-25.

AYM Jones, PT, PhD, FACP, is Associate Professor, Department of Rehabilitation Science, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong ([e-mail]). Address all correspondence to Dr Jones.

E Dean, PT, PhD, is Professor and Coordinator of the Advanced Graduate Programs, School of Rehabilitation Sciences, T-325, 2211 Wesbrook Mall, Vancouver, British Columbia, Canada V6T 2B5, and Coordinator of the Post-Polio Clinic, University of British Columbia. She was Visiting Professor, Department of Rehabilitation Science, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, at the time of this study.

CCS Chow, PT, MSc, is Physical Therapist, Caritas Medical Centre, Kowloon, Hong Kong.

All authors provided concept/research design, writing, project management, and consultation (including review of manuscript before submission). Dr Jones and Mr Chow provided data collection and analysis. Dr Jones provided fund procurement, and Mr Chow provided subjects and facilities/equipment. Dr Jones and Dr Dean provided institutional liaisons and clerical support.

Ethical approval was obtained from the ethics review committees of The Hong Kong Polytechnic University and the Caritas Medical Centre, Hong Kong.

This article was submitted May 15, 2002, and was accepted December 3, 2002.