ARTICLE IN PRESS

Respiratory Medicine (2007) 101, 1412–1418

0954-6111/$ - see frdoi:10.1016/j.rmed.

�Corresponding au

E-mail address: u

Patterns of chest wall kinematics during volitionalpursed-lip breathing in COPD at rest

Roberto Bianchi, Francesco Gigliotti, Isabella Romagnoli, Barbara Lanini,Carla Castellani, Barbara Binazzi, Loredana Stendardi, Michela Grazzini,Giorgio Scano�

Fondazione Don C. Gnocchi (IRCCS), Via Imprunetana, 124, 50020 Pozzolatico, Firenze, Italy

Received 23 November 2006; accepted 29 January 2007Available online 12 March 2007

KEYWORDSChronic obstructivepulmonary disease(COPD);Rehabilitation;Chest wall;Dyspnea;Breathing exercises

ont matter & 20072007.01.021

thor. Tel.: +39 055

opneumo.pozzola

SummaryBackground: Analysis of chest wall kinematics can contribute to identifying the reasonswhy some patients benefit from pursed-lip breathing (PLB).Material and methods: We evaluated the displacement of the chest wall and itscompartments, the rib cage and abdomen, by optoelectronic plethysmography (OEP),during supervised PLB maneuver in 30 patients with mild to severe chronic obstructivepulmonary disease (COPD).Results: OEP showed two different patterns. A first pattern characterized the 19 mostseverely obstructed and hyperinflated patients in whom PLB decreased end-expiratoryvolumes of the chest wall and abdomen, and increased end-inspiratory volumes of thechest wall and rib cage. Deflation of the abdomen and inflation of the rib cage contributedto increasing tidal volume of the chest wall. The second pattern characterized 11 patientsin whom, compared to the former group, PLB resulted in the following: (i) increased end-expiratory volume of the rib cage and chest wall, (ii) greater increase in end-inspiratoryvolume of the rib cage and abdomen, and (iii) lower tidal volume of the chest wall. In thepatients as a whole changes in end-expiratory chest wall volume were related to change inBorg score (r2 ¼ 0.5, po0.00002).Conclusions: OEP helps identifying the reason why patients with COPD may benefit fromPLB at rest.& 2007 Elsevier Ltd. All rights reserved.

Elsevier Ltd. All rights reserved.

260 11; fax: +39 055 260 1272.

tico@dongnocchi.it (G. Scano).

ARTICLE IN PRESS

Table

1Anthrop

ometric,

clinical,an

dfunc

tion

aldata.

Age

BMI

Smok

eVC

FEV1

FEV1/V

CFR

CTLC

RVPa

CO2

PaO2

MRC

(yea

rs)

(kgm�2)

(py)

(%p.v.)

(%p.v.)

(%)

(%p.v.)

(%p.v.)

(%p.v.)

(Torr)

(Torr)

(a.u.)

Group

(30subj.)

7177

2673

46715

87721

45716

38711

1387

3811

1720

1567

5641

.777.2

73.2710

.52.97

0.8

Subgrou

p1(19subj.)

7277

2573

51712

84722

38713

3479

1527

3811

8719

1777

5442

.677.7

71.0710

.52.97

0.8

Subgrou

p2(11subj.)

7178

2873

38717

92719

56714

45712

1157

2710

0716

1197

3640

.276.4

76.979.9

2.97

0.7

po0.00

9po

0.05

po0.00

2po

0.00

5po

0.00

9po

0.02

po0.00

4

Values

aremea

n7SD

.VC:vitalca

pac

ity;

FEV1:forced

expiratoryvo

lumein

1s;

FEV1/V

C:Tiffen

eauratio;

FRC:func

tion

alresidu

alca

pac

ity;

RV:residu

alvo

lume;

MRC:Med

ical

Research

Cou

ncilDyspn

eaScale;

p.v.:

predictedva

lue;

a.u.:arbitraryun

its.

Thoracic kinematics and pursed lip breathing 1413

Introduction

Pursed-lip breathing (PLB) performed as nasal inspirationfollowed by expiratory blowing against partially closed lips isa breathing retraining strategy employed by patients withchronic obstructive pulmonary disease (COPD).1 Not allpatients, however, employ PLB, or report benefiting from it.Some apply this technique spontaneously, whereas otherpatients do not use it even when they are taught.

Studies on patients who naturally incorporate PLB intotheir breathing pattern may provide additional informationon whether changes in chest volume help identify thereasons why PLB relieves dyspnea in patients with COPD. Ithas been reported that the relief from dyspnea provided byPLB is related to its ability to promote a slower and deeperbreathing.2,3 In contrast, an increase in tidal volume withunchanged end-expiratory-lung-volume variably affectsdyspnea during exercise either by enhancing it or leavingit unmodified.4

An increased tidal volume, however, may be obtained intwo different ways: (i) by decreasing end-expiratory volumeof the abdomen and limiting the increase in end-inspiratoryvolume of both the rib cage and abdomen. This would limitpressure production of rib cage muscles and the diaphragmto a small fraction of their maximal pressure-generatingcapacity, therefore, attenuating the sensation of dyspnea insymptomatic patients3,5,6; (ii) in contrast, without theabdominal contribution, the tidal volume of the chest wallis the result of a higher rib cage inspiratory musclefractional pressure; this, conceivably, would not attenuatethe dyspnea.5,6

Thus, the question arises: do changes in operational chestwall volumes help identify the reason for dyspnea relief withPLB in patients with COPD?

We hypothesized that an increase in tidal volume of thechest wall promoted by PLB at the exclusive expense of tidalvolume of the rib cage would not be associated with dyspnearelief in COPD patients.

To assess whether the volume changes of chest wall canhelp identify the reason why some patients benefit from PLBwhile others do not, we applied a recently well-developedtechnique based on optoelectronic plethysmography (OEP)which allows the evaluation of volume changes of chest wallcompartments.7–9 We performed direct measurements ofend-expiratory and end-inspiratory chest wall volumesduring natural breathing while avoiding measurementsbased on wearing a mouthpiece and nose clip. Breathingthrough a mouthpiece with a nose clip in place is known toalter the pattern of breathing, primarily by increasing tidalvolume.10,11

Methods

Patients

Thirty COPD patients with moderate to severe airwayobstruction and mild to moderate hyperinflation andhypoxemia participated in the study (Table 1). They wereselected from a pulmonary rehabilitation program if theysatisfied each of three criteria: (i) long history of smokingand moderate to severe chronic dyspnea score (MRC4II),

ARTICLE IN PRESS

R. Bianchi et al.1414

(ii) clinically stable condition, with no exacerbation, orhospital admission in the preceding four weeks, (iii) free fromother significant disease potentially contributing to dyspnea.

Protocol

Routine lung function was measured first and then patientswere familiarized with procedures and scales for ratingsymptom intensity. Compartmental lung volumes wereevaluated with subjects in a seated position at rest duringboth quiet breathing (QB) defined as a patient’s habitualcomfortable breathing, and PLB. A physiotherapist gaveevery patient the same instructions about PLB execution.Patients were instructed to make a nasal inspirationfollowed by expiratory blowing against partially closed lips,avoiding forceful exhalation.1

No patient had difficulty in learning this technique. Threetrials of both QB and PLB maneuvers, performed correctlyand with care in random order, were recorded for at least6min with the exclusion of sighing and coughing, and thenaveraged. Dyspnea sensation was measured before, duringand after QB and PLB. The study was approved by the EthicsCommittee of the Institution and informed consent wasobtained from subjects.

Lung function

Routine spirometry obtained with subjects in a seatedposition was measured as previously reported.12 FRC(functional residual capacity) was measured by a constantvolume whole-body plethysmograph (Autobox DL 6200Sensor Medics; Yorba Linda, CA, USA.). The normal valuesfor lung volumes are those proposed by the EuropeanRespiratory Society.13

Chest wall kinematics and compartmental volumes

The volume of the chest wall (VCW) was modeled as the sumof the volumes of the rib cage (VRC), and abdomen (VAb). Thevolumes of the chest wall and its compartments wereassessed by applying a noninvasive OEP technique, used aspreviously described.14 Briefly, 89 reflecting markers wereplaced front and back over the trunk from the clavicles tothe anterior superior iliac spines along pre-defined verticaland horizontal lines. To measure the volume of chest wallcompartments from surface markers we defined the follow-ing: (1) the boundaries of rib cage as extending from theclavicles to the costal margin anteriorly down from thexiphisternum, and to the level of the lowest point of thelower costal margin posteriorly; and (2) the boundaries ofthe abdomen as extending caudally from the lower rib cageto a horizontal line at the level of the anterior superior iliacspine. The landmark coordinates were measured with asystem configuration of four infrared TV-cameras, twoplaced 4m behind and two 4m in front of the subject, ata sampling rate of 50Hz. Starting from these coordinates thevolume of the chest wall was computed by triangulating thesurface and then using Gauss’s theorem to convert thevolume integral to an integral over this surface, as describedpreviously.14 Flow signal was obtained by integrating volumetrack. The end-expiratory and end-inspiratory volume of

each compartment was measured at the beginning and endof inspiratory flow (zero-flow points). The differencebetween the end-inspiratory and end-expiratory volume ofeach compartment was calculated as the tidal volume (VT)contribution by each compartment. Thus, VCW ¼ VRC+VAb,and changes in VCW can be calculated as

DVCW ¼ DVRC þ DVAb,

assuming that the only factor causing chest wall volumechanges is gas movement. OEP calculates absolute volumesand the absolute volume of each compartment at FRC incontrol conditions was considered as the reference volume.Volumes are reported either in absolute values or as changesfrom the volume at FRC in control conditions.

Dyspnea

Subjects were asked to quantify the following: (i) chronicexertional dyspnea by MRC questionnaire15 and, (ii) thesensation of nonspecific discomfort associated with the actof breathing by pointing to a score on a modified Borg scalefrom 0 (none) to 10 (maximal) arbitrary units (a.u.).16

Data analysis

Values are means 7SD. Statistical procedure was used totest differences for paired and unpaired samples. Simpleregression analysis was performed using Pearson’s correla-tion coefficient. The level of significance was set at po0.05.All statistical procedures were carried out usingthe Statgraphics Plus 5.1 statistical package (Manugistics,Rockville, MD, USA).

Results

Chest wall kinematics

The analysis of the time course of volume changes in chestwall compartments allowed us to identify two different PLBpatterns. The first pattern (left panels of Fig. 1 and Table 2)characterized 19 patients we called euvolumics in whom thePLB maneuver decreased end-expiratory volumes of bothchest wall ðVCWee Þ and abdomen ðVAbee Þ, and increased, to alesser extent, end-inspiratory volumes of both chest wallðVCWei

Þ and rib cage ðVRCei Þ, but not abdomen ðVAbei Þ. Asshown in Fig. 2 expiratory deflation of chest wall ðVCWee Þ andabdominal compartment ðVAbee Þ down to the FRC line (thedotted line in the fig), and mild inspiratory inflation of ribcage compartment ðVRCei Þ contributed to increasing the tidalvolume of the chest wall ðVTRC þ VTAb ¼ VTCW ¼ þ

0:63� 0:52 LÞ. Also, VCWeiand VRCei , but not VAbei , increased

with PLB. This pattern was also associated with a decreasein Borg score (PLB vs QB, po0.0007) (Fig. 3).

The second pattern (right panels of Fig. 1 and left panelsof Table 3) characterized 11 patients, called hyperinflators,with lower VTRC (po0.01), VTAb (po0.05) and their sum VTCW(po0.01) than euvolumics at QB. In these patients, PLBincreased end-expiratory volumes of the chest wall and ribcage up to FRC (dotted line) (VCWee : þ0:15� 0:21 L;VRCee : þ0:11� 0:14 L), and increased, to a greater extentthan in the euvolumics, end-inspiratory volumes of the chest

ARTICLE IN PRESS

Time (s)

VC

W (

L)

29

30

31

Hyperinflator

VR

C (

L)

16

17

18

0 10 20 30 40 50 60 70

VA

b (

L)

12

13

14

Euvolumic

VC

W (

L)

29

30

31

VR

C (

L)

18

19

20

0 10 20 30 40 50 60 70

VA

b (

L)

10

11

12

PLBQB PLBQB

Figure 1 Time course of breathing pattern during QB and PLB in two representative patients. From top to bottom: volumes of chestwall (VCW), rib cage (VRC) and abdomen (VAb). QB is quiet breathing; PLB is pursed lip breathing.

Table 2 Effects of PLB on volumes of chest wall compartments and breathing pattern in euvolumic patients.

Euvolumics (19 subj.) QB PLB–QB p QB PLB–QB p

VCWeiðLÞ 30.2473.71 +0.2570.44 o0.05 VE (Lmin�1) 10.1272.78 +0.3072.75 n.s.

VRCei ðLÞ 19.0072.31 +0.2370.31 o0.005 Rf (min�1) 14.9174.91 �6.7874.40 o0.000003

VAbei ðLÞ 11.2472.19 +0.0270.22 n.s. VT (L) 0.7470.26 +0.6370.52 o0.00005

VCWee ðLÞ 29.5073.66 �0.3770.23 o0.00001 Ti (s) 1.6170.56 +0.9571.13 o0.002VRCee ðLÞ 18.7172.28 �0.0670.18 n.s. Te (s) 2.9171.03 +2.7271.57 o0.00001VAbee ðLÞ 10.7972.16 �0.3170.22 o0.00001 Ttot (s) 4.527 1.50 +3.6772.57 o0.00001VTCW ðLÞ 0.7470.26 +0.6370.52 o0.00005 VT/Ti (L s

�1) 0.4870.16 +0.1070.17 o0.03

VTRC ðLÞ 0.2870.16 +0.2970.31 o0.001 VT/Te L s�1) 0.2770.08 �0.0170.07 n.s.

VTAb ðLÞ 0.4670.16 +0.3370.25 o0.00005 Ti/Ttot 0.3670.06 �0.0570.05 o0.0002

Borg (a.u.) 1.6171.13 �0.4270.45 o0.0007

Values are mean7SD. VE: minute ventilation; Rf: respiratory frequency; VT: tidal volume; Ti: inspiratory time; Te: expiratory time; Ttot:total time of the respiratory cycle; VT/Ti: mean inspiratory flow; VT/Te: mean expiratory flow; Ti/Ttot: duty cycle. VCWei

: chest wallend-inspiratory volume; VRCei : rib cage end-inspiratory volume; VAbei : abdomen end-inspiratory volume; VCWee : chest wall end-expiratory volume; VRCee : rib cage end-expiratory volume; VAbee : abdomen end-expiratory volume; VTCW : tidal volume of the chestwall; VTRC : tidal volume of the rib cage; VTAb : tidal volume of the abdomen.

Thoracic kinematics and pursed lip breathing 1415

wall (VCWei, po0.000001) and its compartments

ðVAbei ;VRCei Þ. As shown in Fig. 4 the average increase inVTCW ðþ0:36� 0:30 LÞ was exclusively due to an increase inend-inspiratory compartmental volumes ðVAbei ;VRCei Þ. Nochange in Borg score was associated with this pattern (PLBvs QB, p ¼ ns) (Fig. 3).

Breathing pattern

Tables 2 and 3 (right panels) show the effects of PLB on QBpattern in euvolumics and hyperinflators, respectively.Analysis of the variance indicated greater PLB increase inVT (po0.0004), inspiratory time (Ti) (po0.005), expiratorytime (Te) (po0.0009) and total time of the respiratory cycle(Ttot) (po0.003) in euvolumics than in hyperinflators.

Important clinical differences were associated with the twopattern groups. Euvolumics were more severely obstructed(FEV1/VC and FEV1) and hyperinflated (FRC, TLC, RV) atbaseline (Table 1). They appeared to adopt PLB during commonactivities of the pulmonary rehabilitation program. In contrast,hyperinflators did not use PLB if not specifically requested.Moreover, a greater decrease in VCWee was correlated with agreater level of airway obstruction (r ¼ 0.45, po0.02). Finally,change in VCWee positively correlated with change in Borg score(r2 ¼ 0.50, po0.00002) (Fig. 5).

Discussion

The OEP analysis of chest wall kinematics shows why not allpatients with COPD obtain symptom relief from PLB at rest.

ARTICLE IN PRESS

QB QB

BO

RG

(a.u

.)

0

1

2

3

4

BO

RG

(a.u

.)

0

1

2

3

4

PLB PLB

Figure 3 Change in dyspnea sensation with PLB in euvolumics (left panel) and hyperinflators (right panel).

QB PLB

VCW(L)

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

QB PLB

VRC(L)

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

QB PLB

VAb (L)

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Figure 2 Changes in volumes of chest wall (CW) and its compartments with PLB in euvolumics. Left panel: changes in the volumes ofthe chest wall (VCW); middle panel: volumes of the rib cage (VRC); right panel: volumes of the abdomen (VAb). Closed symbols: end-expiratory volume, open symbols: end-inspiratory volume. The dotted line: functional respiratory capacity. Bars are means 7SEM.For explanation see Results.

R. Bianchi et al.1416

The most severely affected patients who deflate the chestwall during volitional PLB reported improvement in theirsensation of breathlessness. This was not the case in thegroup who hyperinflated during PLB.

The major difference between the two PLB patterns wasthe ability of euvolumics to decrease VCWee by decreasingVAbee , while limiting the increase in VRCei . In line withprevious data of ours,3 changes in VCWee and VAbee aredirectly related to increase in Te in that the greaterthe latter the greater is the reduction in the volume. Inparticular, we explain the decrease in VCWee in hyperinflatedpatients with lengthening Te and Ttot, not with increasingmean expiratory flow (VT/Te see Tables 2 and 3).This mechanism is similar to that expected to reducethoracic gas volume entrapment and exercise breathlessnessafter pulmonary rehabilitation programs in patients withCOPD.17

The increased VTCW was, therefore the result of both VTAb ,by exploiting the expiratory reserve volume, and VTRC , byexploiting the inspiratory reserve volume. This strategyexploits the stores of elastic energy of the most compliantchest wall compartment, the abdomen,18 and prevents VCWei

from reaching total chest wall capacity.In contrast, the increase in tidal volume of the chest wall

in hyperinflators was due to an increase in the tidal volumeof the rib cage without the contribution of abdominal tidalvolume. End-inspiratory volume of the chest wall mainlyresulted from rib cage inspiratory muscles; in thesecircumstances esophageal pressure may reach a higherfraction of maximal inspiratory muscle pressure capacityand generate a greater dyspnea score19 (see also fig. 34).Most importantly, in the conditions of the present study thelack of decrease in VAbee along with increase in VAbei suggesta lower abdominal muscle contribution, and higher

ARTICLE IN PRESS

Table 3 Effects of PLB on volumes of chest wall compartments and breathing pattern in hyperinflator patients

Hyperinflators (11 subj.) QB PLB–QB p QB PLB–QB p

VCWeiðLÞ 31.4274.37 +0.5170.15 o0.000001 VE (Lmin�1) 9.8572.87 +0.3072.88 n.s.

VRCei ðLÞ 18.4072.90 +0.3470.10 o0.000001 Rf (min�1) 22.2975.69 �9.0675.14 o0.0002

VAbei ðLÞ 13.0272.09 +0.1770.07 o0.00005 VT (L) 0.4770.22 +0.3670.30 o0.003

VCWee ðLÞ 30.957 4.28 +0.1570.21 o0.05 Ti (s) 1.1570.33 +0.6470.68 o0.02VRCee ðLÞ 18.2572.85 +0.1170.14 o0.05 Te (s) 1.8270.77 +1.4971.28 o0.00004VAbee ðLÞ 12.7072.02 +0.0470.10 n.s. Ttot (s) 2.9771.06 +2.1471.82 o0.003VTCW ðLÞ 0.4770.22 +0.3670.30 o0.005 VT/Ti (L s

�1) 0.4170.10 +0.0770.12 n.s.(o0.06)

VTRC ðLÞ 0.1570.08 +0.2370.18 o0.005 VT/Te L s�1) 0.2870.10 �0.0170.09 n.s.

VTAb ðLÞ 0.3270.19 +0.1370.14 o0.05 Ti/Ttot 0.4070.06 �0.0570.08 n.s.

Borg (a.u.) 0.9670.57 +0.0970.38 n.s.

Values are mean7SD. VE: minute ventilation; Rf: respiratory frequency; VT: tidal volume; Ti: inspiratory time; Te: expiratory time; Ttot:total time of the respiratory cycle; VT/Ti: mean inspiratory flow; VT/Te: mean expiratory flow; Ti/Ttot: duty cycle. VCWei

: chest wallend-inspiratory volume; VRCei : rib cage end-inspiratory volume; VAbei : abdomen end-inspiratory volume; VCWee : chest wall end-expiratory volume; VRCee : rib cage end-expiratory volume; VAbee : abdomen end-expiratory volume; VTCW : tidal volume of the chestwall; VTRC : tidal volume of the rib cage; VTAb : tidal volume of the abdomen.

QB PLB

VCW (L)

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

QB PLB

VRC (L)

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

QB PLB

VAb (L)

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Figure 4 Changes in volumes of chest wall (CW) and its compartments with PLB in hyperinflators. Left panel: changes in thevolumes of the chest wall (VCW); middle panel: volumes of the rib cage (VRC); right panel: volumes of the abdomen (VAb). Closedsymbols: end-expiratory volume; open symbols: end-inspiratory volume. The dotted line: functional respiratory capacity. Bars aremeans 7SEM. For explanation see Results.

ΔVCWee (L)

-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

ΔBO

RG

(a.u

.)

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

EUVOLUMICS

HYPERINFLATORSp<0.00002

r2=0.50

Figure 5 Relationship between change in end-expiratoryvolume of the chest wall ðVCWee Þ and change in Borg score.Open circles: euvolumics; closed circles: hyperinflators.

Thoracic kinematics and pursed lip breathing 1417

inspiratory diaphragmatic contribution to abdominal tidalvolume,20 respectively.

The mechanisms of apparent symptomatic improvementwith PLB have not been fully elucidated.2,21,22 Dynamicairway compression by itself may produce afferent sensoryinformation that contributes to the sense of dyspneaexperienced.23 Assumption of a slower, deeper-breathingpattern during PLB would reduce intrinsic end-expiratory-positive-alveolar-pressure (PEEPi) and, thereby, the inspira-tory work of breathing.24 By using PLB, COPD patientsbreathe larger tidal volumes2,22 which correlate withsymptomatic relief of dyspnea.2 Nonetheless, the issue ofwhether and to what extent PLB affects dyspnea is still amatter of debate, since the efficacy of PLB in relievingdyspnea varies greatly among COPD patients.4,25

Our present and previous findings3 of decreased VCWee

associated with a lower Borg score are in line with the effectof hyperinflation on dyspnea26 and, conversely, with theattenuation of the symptom with chest wall deflation.18,27

The net effect of decreased VCWee is an improvement in the

ARTICLE IN PRESS

R. Bianchi et al.1418

length–tension relationship of the inspiratory muscles sothat, for any given neural input, lung expansion (VT) iscommensurate with the degree of pressure generated whenthe muscle shortens. This is a physiological mechanism ofneuromuscular coupling of the ventilatory pump.26 Incontrast and for opposite reasons lung hyperinflation, whichpromotes the shortening of the inspiratory muscle alongwith a higher than normal neural motor output, results inneuromuscular discoupling and more breathlessness.

In this connection, the significant difference in increasedtidal volume we found in the two groups translates into adifferent PLB effect on dyspnea. The greater VTCW wasobtained through increased abdominal contribution ineuvolumics as compared with hyperinflators. The ability ofPLB to expand VTCW with the expiratory abdominal con-tribution (VTAb , see Fig. 2) limits the inspiratory work ofbreathing, thus modulating the sensation of breathlessness.In contrast, the null expiratory abdominal contribution toVTCW expansion, and the enhanced inspiratory contributionof both the rib cage and abdomen to end-inspiratory volumeof the chest wall ðVCWei

Þ could be the reason for theunmodified Borg score in hyperinflators.

In conclusion, OEP analysis of VCW helps identify thereason for the benefit of the PLB maneuver in severelyobstructed COPD patients.

References

1. American Thoracic Society. Pulmonary Rehabilitation-1999. AmJ Respir Crit Care Med 1999;159:1666–82.

2. Mueller RE, Petty TL, Filley GF. Ventilation and arterial bloodgas changes induced by pursed lips breathing. J Appl Physiol1970;28(6):784–9.

3. Bianchi R, Gigliotti F, Romagnoli I, et al. Chest wall kinematicsand breathlessness during pursed-lip breathing in patients withCOPD. Chest 2004;125(2):459–65.

4. Spahija J, de Marchie M, Grassino A. Effect of imposed pursed-lip breathing on respiratory mechanics and dyspnea at rest andduring exercise in COPD. Chest 2005;128(2):640–50.

5. Killian KJ, Gandevia SC, Summers E, Campbell EJ. Effect ofincreased lung volume on perception of breathlessness, effortand tension. J Appl Physiol 1984;57(3):686–91.

6. El-Manshawi A, Killian KJ, Summers E, Jones NL. Breathlessnessduring exercise with and without resistive loading. J ApplPhysiol 1986;61(3):896–905.

7. Sanna A, Bertoli F, Misuri G, et al. Chest wall kinematics andrespiratory muscle in walking healthy humans. J Appl Physiol1999;87(3):938–46.

8. Lanini B, Bianchi R, Romagnoli I, et al. Chest wall kinematics inpatients with hemiplegia. Am J Respir Crit Care Med2003;168(1):109–13.

9. Gorini M, Iandelli I, Misuri G, et al. Chest wall hyperinflationduring acute bronchoconstriction in asthma. Am J Respir CritCare Med 1999;160(3):808–16.

10. Gilbert R, Auchincloss Jr JH, Brodsky J, Boden W. Changes intidal volume, frequency, and ventilation induced by theirmeasurement. J Appl Physiol 1972;33(2):252–4.

11. McCool FD, Kelly KB, Loring SH, Greaves IA, Mead J. Estimatesof ventilation from body surface measurements in unrestrainedsubjects. J Appl Physiol 1986;61(3):1114–9.

12. Scano G, Garcia-Herreros P, Stendardi D, Degre S, De Coster A,Sergysels R. Cardiopulmonary adaptation to exercise in coalminers. Arch Environ Health 1980;35(6):360–6.

13. European Community for Steel and Coal. Standardiza-tion of lung function test. Official Statement of theEuropean Respiratory Society. Eur Resp J Suppl 1993;16:1–100.

14. Cala SJ, Kenyon CM, Ferrigno G, et al. Chest wall and lungvolume estimation by optical reflectance motion analysis. JAppl Physiol 1996;81(6):2680–9.

15. American Thoracic Society. Surveillance for respiratory hazardin the occupational setting: ATS statement. Am Rev Respir Dis1982;126(5):952–6.

16. Borg GA. Psychophysical bases of perceived exertion. Med SciSports Exercise 1982;14(5):377–81.

17. O’Donnell DE, McGuire M, Samis L, Webb KA. The impact ofexercise reconditioning on breathlessness in severe chronicairflow limitation. Am J Respir Crit Care Med 1995;152(6 pt1):2005–13.

18. Konno K, Mead J. Measurement of the separate volume changesof rib-cage and abdomen during breathing. J Appl Physiol 1967;22:407–22.

19. Gorini M, Misuri G, Corrado A, et al. Breathing pattern andcarbon dioxide retention in severe chronic obstructive pulmon-ary disease. Thorax 1996;51(7):677–83.

20. Romagnoli I, Gigliotti F, Lanini B, et al. Chest wall kinematicsand respiratory muscle coordinated action during hypercapniain healthy males. Eur J Appl Physiol 2004;91(5–6):525–33.

21. Breslin EH. The pattern of respiratory muscle recruitmentduring pursed-lip breathing. Chest 1992;101:75–8.

22. Thoman RL, Stoker GL, Ross JC. The efficacy of pursed-lipsbreathing in patients with chronic obstructive pulmonarydisease. Am Rev Respir Dis 1966;93:100–6.

23. O’Donnell DE, Sanii R, Younes M. Improvement in exerciseendurance in patients with chronic airflow limitation usingcontinuous positive airway pressure. Am Rev Respir Dis 1988;138:1510–4.

24. Haluszka J, Charthrand DA, Grassino AE, Milic-Emili J. IntrinsicPEEP and arterial PCO2 in stable patients with chronicobstructive pulmonary disease. Am Rev Respir Dis 1990;141:1194–7.

25. American Thoracic Society. Dyspnea. Mechanisms, assessment,and management: a Consensus Statement. Am J Respir CritCare Med 1999;159:321–40.

26. O’Donnell DE, Webb KA. Exertional breathlessness in patientswith chronic airflow limitation. The role of lung hyperinflation.Am Rev Respir Dis 1993;148:1351–7.

27. Gigliotti F, Coli C, Bianchi R, et al. Exercise training improvesexertional dyspnea in patients with COPD: evidence of role ofmechanical factors. Chest 2003;123:1794–802.