JCRP2802_128-141 7/3/08 20:23 Page 128

Inspiratory Muscle Training

Compared With Other Rehabilitation
Interventions in Chronic Obstructive
Pulmonary Disease
A SYSTEMATIC REVIEW UPDATE

Kelly O’Brien, BScPT, E. Lynne Geddes, MRE, W. Darlene Reid, PhD, Dina Brooks, PhD, and Jean Crowe, MHSc

■ PURPOSE: To determine the effect of inspiratory muscle training (IMT) K E Y W O R D S

(alone or combined with exercise and/or pulmonary rehabilitation)
and compare with other rehabilitation interventions among adults breathing exercises
with chronic obstructive pulmonary disease (COPD).
chronic obstructive pulmonary disease
■ METHODS: We conducted a systematic review, using Cochrane
Collaboration protocol. We included randomized controlled trials, lung diseases
published in English, comparing IMT or combined IMT and
exercise/pulmonary rehabilitation with other rehabilitation interven- obstructive
tions among adults with COPD. Abstracts were reviewed independently
by 2 investigators to determine study eligibility up to December 2005. respiratory muscles
Data were abstracted and methodological quality of included studies
was assessed. respiratory muscle training
■ RESULTS: A total of 156 additional articles were retrieved. Two new stud-
ies met the inclusion criteria and were included with 16 studies in the
original review. Results highlight updated subgroup analyses compar- This article is an update of an original sys-
ing (1) IMT versus exercise and (2) combined IMT and exercise versus tematic review: Crowe J, Reid WD,
Geddes L, O’Brien K, Brooks D. Inspiratory
exercise alone. Fourteen meta-analyses were performed for outcomes
muscle training compared with other
of inspiratory muscle strength, exercise tolerance, and quality of life.
rehabilitation interventions in adults with
Results showed significant improvements in maximum inspiratory chronic obstructive pulmonary disease.
pressure and maximum exercise tidal volume favoring combined IMT COPD: J COPD. 2005;3:319–329.
and exercise compared with exercise alone.
Author Affiliations: Department of Physical
■ CONCLUSIONS: Performing a combination of IMT plus exercise may Therapy, University of Toronto, Toronto
lead to significant improvements in inspiratory muscle strength and (Ms O’Brien and Dr Brooks), and School of
one outcome of exercise tolerance for individuals with COPD. Rehabilitation Science, McMaster University,
Hamilton (Ms Geddes and Ms Crowe),
Ontario, Canada; and Department of
Physical Therapy, University of British
Columbia, Vancouver, British Columbia,
Canada (Dr Reid).

Corresponding Author: Kelly O’Brien,

BScPT, Department of Physical Therapy,
University of Toronto, 500 University Ave,
Room 160, Toronto, Ontario, Canada M5G
1V7 ([email protected]).

In 1999, an estimated 750,000 Canadians were living likely an underestimation as individuals often do not
with chronic obstructive pulmonary disease (COPD).1 seek medical attention until the disease is moderate or
During this time, approximately 3.6% of women and severe. COPD is a common and costly condition.3,4 In
2.8% of men were affected.2 However, these figures are Canada, it is the fourth leading cause of death in men

128 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141 www.jcrpjournal.com

JCRP2802_128-141 7/3/08 20:23 Page 129

and seventh leading cause of death in women and its cation alone.13 While these results suggested
prevalence continues to rise.5 Hence, the care of indi- improvements with IMT for inspiratory muscle
viduals with COPD is becoming more complex and strength and endurance, results for other outcomes,
costly.5 along with comparisons of IMT with other interven-
COPD is a progressive, irreversible lung disease tions (such as exercise) or cointerventions (IMT com-
characterized by airflow limitation, worse on expira- bined with exercise) were less clear. Furthermore,
tion, which leads to compression of small airways, this last review included a small number of studies
resulting in air trapping in the lungs and subsequent (n  16) resulting in 7 meta-analyses that combined
hyperinflation.1 Individuals with COPD may experi- data from only 4 of these studies.13 Thus, the purpose
ence inspiratory muscle dysfunction due to increased of this article is to report on an update of this sys-
residual volumes, hyperinflation, and reduced range tematic review,13 which included 2 additional studies,
of motion of the diaphragm.6,7 Malnutrition, hyper- to determine the effect of IMT (alone or in combina-
capnia, hypoxemia, and corticosteroid use may fur- tion with exercise and/or pulmonary rehabilitation)
ther reduce inspiratory muscle strength and compared to other rehabilitation interventions such
endurance.6,7 These impairments may additionally as exercise, education, other breathing techniques, or
lead to dyspnea, reduced exercise tolerance, and exercise and/or pulmonary rehabilitation on out-
reduced quality of life (QOL). comes of inspiratory muscle strength and endurance,
Pulmonary rehabilitation consisting of exercise, exercise tolerance, dyspnea, and QOL among adults
education, and psychosocial support has been shown with COPD. This article specifically focuses on 2 sub-
to decrease dyspnea and improve exercise tolerance group analyses that were performed for this update;
and mastery of symptoms among persons with the effect of (1) IMT versus exercise, and (2) com-
COPD.8 Evidence has proposed the use of inspiratory bined IMT and exercise versus exercise alone.
muscle training (IMT) as a way to enhance outcomes
of inspiratory muscle strength and endurance, exer-
cise capacity, dyspnea, and QOL for this population.9 METHODS
Four meta-analyses have assessed the use of IMT
alone or in combination with other rehabilitation inter- We performed an update of an original systematic
ventions among adults with COPD. Smith et al10 review,13 using methods of the Cochrane
reviewed 17 studies and demonstrated little evidence to Collaboration.14 In the original review, electronic data-
support the use of IMT on inspiratory muscle strength bases and reference lists from appropriate texts and
and endurance among IMT versus control. A more articles were searched up to August 2003. We con-
recent review by Lotters et al11 reviewed 10 studies not tacted authors for additional data, and hand-searched
part of the earlier review, and showed significant targeted journals to locate articles for inclusion. In this
improvements in inspiratory muscle strength and update, we conducted an additional search of the lit-
endurance, and dyspnea during exercise and rest erature from September 2003 to December 2005,
among participants undergoing IMT versus control. using similar methods to the original review.
Two other systematic reviews determined the
effect of IMT for adults with COPD.12,13 The first Selection of Articles and Data Abstraction
review focused on subgroup analyses comparing IMT Copies of the articles retrieved by the search were
with sham or no intervention, low versus high inten- reviewed independently by 2 reviewers to identify
sities of IMT, and different modes of IMT.12 Results those studies that met the 5 inclusion criteria: (1) adult
showed improvements in some outcomes of inspira- participants (18 years of age or older), (2) COPD,
tory muscle strength and endurance, exercise capaci- (3) an IMT intervention, (4) a randomized comparison
ty, and dyspnea among adults with COPD who group that received an intervention other than sham,
engaged in targeted resistive or threshold IMT com- and (5) published in English. When there was a lack
pared with sham or no intervention.12 The second of agreement between reviewers, a third reviewer
review focused on comparing IMT (alone or com- independently read the article and determined study
bined with exercise and/or pulmonary rehabilitation) inclusion.
with other rehabilitation interventions including exer- Inspiratory muscle training was defined as any
cise, education, other breathing techniques, or exer- intervention(s) aimed at improving strength and/or
cise/pulmonary rehabilitation among adults with endurance of the inspiratory muscles. Modes of IMT
COPD.13 Results from this review showed significant were classified as either targeted and/or threshold
improvements in inspiratory muscle strength and IMT, normocapnic hyperventilation IMT, or “other”
endurance, and the dyspnea scale of a QOL ques- inspiratory resistance training IMT. Rehabilitation
tionnaire for participants engaged in IMT versus edu- interventions included but were not limited to

www.jcrpjournal.com IMT and Other Rehabilitation Interventions in COPD / 129

JCRP2802_128-141 7/3/08 20:23 Page 130

exercise and/or pulmonary rehabilitation, education, tion with education alone35 and the other study com-
or other breathing techniques. Pulmonary rehabilita- pared combined IMT plus exercise with exercise
tion was defined as exercise training for at least 4 alone.36 Characteristics of all 18 included studies are
weeks, with or without education, and with or with- provided in Table 1.
out psychological support.15 Exercise was defined as
any upper or lower extremity strength or aerobic Methodological Quality of Included Studies
training that may or may not include other compo- Methodological quality of all included studies is pro-
nents of a pulmonary rehabilitation program. vided in Table 2. We assumed that an intention-to-
Education was defined as patient teaching that treat analysis was performed in 13 of the studies
included the pathophysiology of COPD, along with whereby participants at study completion were ana-
strategies for its management. Other breathing tech- lyzed on the basis of the groups to which they were
niques were defined as any type of breathing tech- originally randomized.19–22,24–26,28,30,32,34–36 The remain-
nique or exercise other than IMT. ing 5 studies reported a per-protocol analysis, in
Two reviewers abstracted the relevant data from which participants who were nonadherent with the
each included article and entered the information onto intervention were excluded from the analy-
standardized data abstraction forms. Methodological sis.23,27,29,31,33 Four studies reported no withdrawals of
quality of included studies was assessed using Jadad participants19,20,26,28 whereas withdrawal rates in the
criteria.16 We also assessed trials for similarity of groups other studies ranged from 7% to 59% (Table 2). The
of participants upon entry to the studies and deter- 2 most common reasons for withdrawal included
mined whether an intention-to-treat analysis was con- health reasons (acute exacerbations of COPD and
ducted. Data abstraction was confirmed between the 2 other conditions), and lack of interest, motivation, or
reviewers. If needed, a third reviewer confirmed any other social reasons that resulted in incompletion of
discrepancies or uncertainties related to the data the study.
abstraction process. Authors were contacted if addi-
tional data were required. Characteristics of Participants
All studies included participants with stable COPD.
Meta-Analysis Fourteen of the 18 included studies enrolled more
Where studies were comparable and used similar par- men than women. One included only men22 and
ticipants, similar modes of IMT, similar training pro- 1 included predominantly women.23 The mean age of
tocols, and measurement of outcomes, we performed participants in most studies was between 56 and 72;
meta-analyses, using RevMan 4.2.2 computer soft- however, 3 studies included some participants
ware.17 Given that both targeted and/or threshold and younger than 50 years.20,28,31 Participants were
normocapnic hyperventilation modes of IMT include described as moderately to severely affected by their
a baseline maximum inspiratory pressure and involve COPD except for 1 study that included some partici-
working toward a targeted workload, we considered pants who were mildly affected.29 The forced expira-
them comparable for this update. tory volume in 1 second (FEV1) was less than 65%
Outcomes were analyzed as continuous outcomes predicted20,24–36 or less than 1.3 liters19,21–23 in all stud-
using a random effects model to calculate a weighted ies. Eleven studies included data on FEV1/FVC
mean difference and 95% confidence interval (CI). (forced vital capacity) ratio indicating mean values
None of the outcomes were dichotomous. A P  .05 below 0.7.22–25,27–32,35
indicated a statistical significance for an overall effect,
and a P  .1 indicated statistical significance for het- Meta-Analyses
erogeneity between studies.18 In situations where The addition of 2 new studies enabled us to update
meta-analyses were not possible, we report qualita- 2 sub-group analyses that compared (1) IMT with
tive descriptions of the studies. exercise and (2) combined IMT and exercise with
exercise alone. Results focus on these analyses.

RESULTS IMT Compared With Exercise

Five of the 18 studies compared IMT with exercise
The search from the initial review revealed 274 arti- alone.19,20,22,27,31 The exercise in all 5 studies included
cles, of which 16 met the inclusion criteria.19–34 The lower extremity aerobic exercise; however, Jones and
search from this update revealed an additional 156 colleagues22 also included upper extremity training.
articles, of which 2 additional studies met the inclu- Six meta-analyses were performed with a combi-
sion criteria.35,36 One study compared a home-based nation of 4 studies (Table 3).19,20,27,31 Jones and
program of combined IMT and pulmonary rehabilita- colleagues22 used a nontargeted inspiratory resistance

130 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141 www.jcrpjournal.com

 JCRP2802_128-141
www.jcrpjournal.com
T a b l e 1 • CHARACTERISTICS OF INCLUDED STUDIES (ORIGINAL REVIEW AND UPDATE) (n  18)
Study and Interventions Monitoring of Frequency
Number of Compared Type of Breathing Rate Time and and Duration Progression Supervision
Participants by Group IMT Device and Rhythm Intensity of IMT of IMT of IMT of IMT Outcomes Measured

7/3/08
Reid (1984) IMT Inspiratory resistance Yes, metronome At maximum intensity 1d Yes Supervised by PT Insp muscle strength
n  12 Treadmill training with target rate control tolerable 30 min/ 5 d/wk Exercise tolerance
No intervention Targeted and/or Threshold (16/min) session 6 wk PFT

20:23
Chen (1985) IMT  PR Inspiratory resistance Not monitored Unable to calculate 2d No Diary Insp muscle strength
n  13 Sham IMT  PR training—no target intensity 15 min/ 7 d/wk and endurance
Other Inspiratory session 4 wk Exercise tolerance

Page 131
Resistance Training
Jones (1985) IMT Inspiratory resistance NR Unable to calculate 2d Yes Supervised Exercise tolerance
n  30 General exs, training—no target intensity 15 min/ 7 d/wk every 2 wk at Dyspnea
walking Other Inspiratory session 10 wk laboratory PFT
Sham IMT Resistance Training
Levine (1986) IMT IPPB Normocapnic Yes Target flow rate @ 1d Yes Supervised in Insp muscle strength
n  48 IPPB hyperventilation 56L/min (maximum 5 d/wk laboratory and endurance
sustained ventila- 6 wk Exercise tolerance
tory capacity at Quality of life
baseline) 15 min/ PFT
session
Ries (1986) IMT  PR Normocapnic NR 15 min/session 3d Yes Diary Insp muscle strength
IMT and Other Rehabilitation Interventions in COPD / 131

n  18 Walking  PR hyperventilation based on maxi- 7 d/wk Exercise tolerance

mum sustainable 6 wk
ventilation for
15 min
Noseda IMT Inspiratory resistance Not monitored Unable to calculate 2d Yes Supervised by PT Insp muscle endurance
(1987) Respiratory muscle training—no target or controlled intensity (highest 7 d/wk for first 2 wk Exercise tolerance
n  25 stretch and Other Inspiratory intensity able to 8 wk
breathing exs Resistance Training tolerate) 15 min/
session
Goldstein IMT PR Threshold Yes, noseclips Unable to calculate 2d Yes Supervised by Insp muscle strength
(1989) Sham IMT  PR Targeted and/or Threshold intensity 10 to 5 d/wk respiratory Exercise tolerance
n  12 20 min/session 4 wk therapist PFT
Dekhuijzen IMT  PR Inspiratory resistance Yes, metronome 70% PImax 15 min/ 2d Yes Supervised by PT Insp muscle strength
(1991) PR training—with target session 5 d/wk and endurance
n  40 Targeted and/or Threshold 10 wk Exercise tolerance
QOL
Nosworthy IMT ACBT Threshold NR 70% PImax or less, 2d Yes Supervised by PT Insp muscle strength
(1992) Treadmill  ACBT Targeted and/or Threshold until able to sustain 3 d/wk 3  wk, and endurance
n  46 PD  ACBT for 15 min 15 min/ 6 wk encouraged Exercise tolerance
session to do daily (continues)
 JCRP2802_128-141
132 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141
T a b l e 1 • Continued
Interventions Monitoring of Frequency
Compared Type of Breathing Rate Time and and Duration Progression Supervision
Study by Group IMT Device and Rhythm Intensity of IMT of IMT of IMT of IMT Outcomes Measured

7/3/08
Weiner (1992) IMT  general exs Threshold NR 15% PImax, increased 1d Yes Supervised by Insp muscle strength
n  36 Sham IMT  exs Targeted and/or Threshold to 80% PImax 3 d/wk PT and endurance
No intervention 30 min/session 24 wk Exercise tolerance

20:23
(cointervention during cycle and PFT
IMT during cycle rowing
and rowing)

Page 132
Wanke (1994) IMT  cycle Threshold Yes, metronome 80% PImax (strength) 1d Yes Supervised by Insp muscle strength
n  60 Cycle Targeted and/or Threshold and 70% PImax 7 d/wk PT and endurance
(endurance) 8 wk Exercise tolerance
16 min/session
Berry (1996) IMT  general exs Threshold NR 15% PImax, increased 2d Yes Diary Insp muscle strength
n  27 Sham IMT  exs Targeted and/or Threshold to 80% PImax. 7 d/wk Exercise tolerance
Sham IMT  15 min/session 12 wk Dyspnea
breathing and PFT
flexibility/stretch
exs
Larson (1999) IMT Threshold NR 30% PImax, increased 1d Yes Home-based Insp muscle strength
n  130 Cycle Targeted and/or Threshold to 60% PImax 5 d/wk program with and endurance
IMT  cycle 30 min/session 16 wk diary and Exercise tolerance
Education weekly visits Dyspnea
by nurse QOL—dyspnea and
fatigue
Covey (2001) IMT Threshold NR 30% PImax, increased 1d Yes Diary and weekly Insp muscle strength
n  37 Education Targeted and/or Threshold to 60% PImax 5 d/wk visits by nurse and endurance
30 min/session 16 wk who super- QOL—dyspnea
vised training PFT

Minoguchi IMT Threshold NR 30% PImax 2d No Not supervised Insp muscle strength
(2002) Respiratory muscle Targeted and/or Threshold 10 min/session 7 d/wk Exercise tolerance
n  16 stretch and 4 wk PFT
breathing exs
www.jcrpjournal.com

Scherer IMT Normocapnic Yes, metronome 50%–60% of partici- 2d Yes Diary Insp muscle strength
(2002) IS with minimal hyperventilation pant’s vital capacity 5 d/wk and endurance
n  34 resistance 15 min/session 8 wk Exercise tolerance
Dyspnea
QOL
PFT
(continues)
JCRP2802_128-141 7/3/08 20:23 Page 133

training mode of IMT and thus was not combined

Abbreviations: ACBT, active cycle of breathing training; IMT, inspiratory muscle training; insp, inspiratory; exs, exercises; IPPB, intermittent positive pressure breathing; IS, incentive spirometry, L/min, liters per minute;
MVV, maximum voluntary ventilation; NR, not reported; PD, postural drainage; PImax, maximal inspiratory pressure; PFT, pulmonary function testing, PT, physical therapist; PR, pulmonary rehabilitation; QOL, quality
Outcomes Measured
with the other 4 studies in the meta-analyses.

Insp muscle strength

Insp muscle strength

Exercise tolerance

Exercise tolerance
and endurance

and endurance

QOL—dyspnea
Inspiratory Muscle Strength

and fatigue
One meta-analysis was performed for inspiratory

Dyspnea
muscle strength and found no difference in the

QOL
change of maximum inspiratory pressure (PImax)

PFT

PFT
among participants in the IMT group compared with
the exercise group but indicated a favorable trend

ment to reach
program with
Supervision

Yes and verbal

encourage-
visits twice
towards a greater positive influence of IMT (Table 3).
Home-based

telephone
of IMT

weekly

Inspiratory Muscle Endurance

target
No meta-analyses updates could be performed for
inspiratory muscle endurance. The 1 study that mea-
Progression

sured this outcome reported a significant increase in

of IMT

inspiratory muscle endurance (measured by discon-

Yes

Yes

tinuous incremental threshold loading (DC-ITL))

among participants in the IMT group, and reported
no change in the exercise group.31
and Duration
Frequency

Exercise Tolerance
of IMT

7 d/wk

3 d/wk
5d

1d
8 wk

8 wk

Five meta-analyses were performed for outcomes of

exercise tolerance. None yielded statistical signifi-
cance for an effect with IMT compared with exercise
(Table 3). Individual study results varied. Larson and
participants’ MVV
Intensity of IMT
Unable to calculate

1–20 min/session
at last 3 sessions
intensity 5 min/

colleagues31 showed greater increases in maximum

Averaged 72% of
Time and

oxygen consumption (VO2max), heart rate maximum

(HRmax), and maximum minute ventilation (VEmax), in
session

the exercise group compared with the IMT group.31

Ries and Moser20 reported increases in VO2max and
VEmax in the IMT group and improvements in the 12-
Breathing Rate
Monitoring of

minute walking distance test (12MWD) for the exer-

and Rhythm

Yes, noseclips

cise group.20 Nosworthy and colleagues27 (who used

IMT with active cycle breathing retraining) reported
improvements in 12MWD in both groups, but signifi-
NR

cantly greater improvements in the exercise (tread-

mill) group.27 Reid and Warren19 found no significant
differences between the IMT and exercise groups
Resistance Training
Inspiratory resistance
training—no target

Normocapnic hyper-

for HRmax, VEmax, maximum carbon dioxide uptake

IMT Device

Other Inspiratory
Type of

(VCo2max), and VO2max.19

ventilation

Dyspnea
No meta-analysis updates could be performed for
dyspnea outcomes. Larson and colleagues31 was the
only study of 4 to measure dyspnea and found simi-
Education (at initial
T a b l e 1 • Continued

IMT  PR (includ-

lar decreases in rating of perceived breathlessness

Interventions

ing education)
Compared
by Group

during a DC-ITL test among both the combined IMT

IMT  cycle
interview)

Studies included in this update.

and exercise and exercise-only groups.

Cycle

Quality of Life
No meta-analyses could be performed for QOL in this
of life; #, number.

subgroup analysis. Larson and colleagues31 was the

a
Oh (2003)
n  34

n  38
(2005)a

only 1 of the 4 studies to measure QOL and found

Mador

improvements in the Dyspnea and Fatigue Scales of

Study

the Chronic Respiratory Questionnaire (CRQ) in both

a

the IMT and exercise groups.

www.jcrpjournal.com IMT and Other Rehabilitation Interventions in COPD / 133

 JCRP2802_128-141
134 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141
T a b l e 2 • METHODOLOGICAL QUALITY OF INCLUDED STUDIES (n = 18)
Description of Reasons % of Was
Randomization Groups Study Withdrawals for Withdrawals/ Intention-to-

7/3/08
Randomized Method Similar at Double- Description and Withdrawals Dropouts Lost treat Analysis
Study Study? Described? Baseline? Blinded? of Blinding Dropouts? Provided? to Follow-up Conducted?
Reid (1984) Yes NR NR No NR No NR 0% (assumed) Assumed

20:23
Chen (1985) Yes NR No–Control No (single-blind) Participants— Yes NA 0% Assumed
group sham
weighed

Page 134
more; had
more males
Jones (1985) Yes Drew names No—Sham No (single-blind) Participants— Yes Yes 30% (9/30) Assumed
from box, for group had sham
allocation more females
Levine Yes NR Yes No (single-blind) Participants— Yes Yes 33% (16/48) Assumed
(1986) sham
Ries (1986) Yes NR No—IMT group No NR Yes Yes 33% (6/18) No—data
had higher excluded
walking from non-
endurance compliant
time participants
(per-proto-
col analysis)
Noseda Yes NR Yes No NR Yes Yes 20% (5/25) Assumed
(1987)
Goldstein Yes NR Groups similar No (single-blind) Participants— Yes Yes 8% (1/12) Assumed
(1989) for 6-min sham
walk test and
submaximal
exs test; not
similar for
FEV1, weight
Dekhuijzen Yes NR Yes No NR Yes NA 0% Assumed
(1991)
Nosworthy Yes NR Yes No NR No Yes 15% (7/46) No—data
(1992) excluded
www.jcrpjournal.com

from non-
compliant
participants
(per-protocol
analysis)
Weiner (1992) Yes NR Yes No (single-blind) Participants— Yes NA 0% Assumed
sham
(continues)
 JCRP2802_128-141
www.jcrpjournal.com

T a b l e 2 • Continued

7/3/08
Description of Reasons % of Was
Randomization Groups Study Withdrawals for Withdrawals/ Intention-to-
Randomized Method Similar at Double- Description and Withdrawals Dropouts Lost treat Analysis
Study Study? Described? Baseline? Blinded? of Blinding Dropouts? Provided? to Follow-up Conducted?

20:23
Wanke Yes NR Yes No NR Yes Yes 30% (18/60) No—data exclud-
(1994) ed from non-
compliant par-

Page 135
ticipants (per-
protocol analy-
sis)

Berry Yes NR Yes No (single-blind) Participants— Yes Yes 7% (2/27) Assumed

(1996) sham
Larson Yes NR Yes No (single-blind) Assessors of Yes Yes 59% (77/130) No—data exclud-
(1999) outcomes ed from non-
compliant par-
ticipants (per-
protocol analy-
sis)
IMT and Other Rehabilitation Interventions in COPD / 135

Covey Yes NR Yes No (single-blind) Assessors of Yes Yes 27% (10/37) Assumed
(2001) outcomes
Minoguchi Yes Crossover with Yes No NR Yes Yes 25% (4/10) No—data exclud-
(2002) 4-wk washout ed from non-
compliant par-
ticipants (per-
protocol analy-
sis)

Scherer Yes Yes—computer Yes No (single-blind) Participants— Yes Yes 12% (4/34) Assumed
(2002) generated sham
random table
Oh Yes Random assign- Yes No NR Yes Yes 32% (11/32) Assumed
(2003)a ment based
on order of
referral
Mador Yes NR Yes No (single-blind) Assessors of Yes Yes 24% (9/38) Assumed
(2005)a outcomes

Abbreviations: exs  exercises, FEV1  forced expired volume in 1 second, IMT  inspiratory muscle training, min  minutes, NA  not applicable, NR  not reported.
a
Studies included in this update.
JCRP2802_128-141 7/3/08 20:23 Page 136

T a b l e 3 • RESULTS OF SUBGROUP META-ANALYSES—IMT VERSUS EXERCISE

Individual Number of
Studies Participants Weighted 95% P value of
Included in Included in Mean Confidence Overall
Outcomes Meta-analysis Meta-analysis Difference Interval Effect Interpretation
Inspiratory muscle strength
PImax, cmH2O Reid (1984) 36 6.56 cmH2O 1.00, 14.13 .09 No difference in change
Larson (1999) in PImax among the IMT
group compared with the
exercise group
Confidence interval
indicates a trend toward
an increase in PImax among
IMT group compared
with exercise group
Exercise tolerance
VEmax, L/min Reid (1984) 47 4.01 L/min 8.01, 0.02 .05 No difference in change
Ries (1986) in VEmax among the IMT
Larson (1999) group compared with
the exercise group
VO2max, L/min Reid (1984) 47 0.05 L/min 0.26, 0.17 .67 No difference in change
Ries (1986) in VO2max among the IMT
Larson (1999) group compared with
the exercise group
12MWD, m Ries (1986) 37 12.33 m 145.74, .86 No difference in change
Nosworthy 121.08 in 12MWD among the
(1992) IMT group compared
with the exercise group
HRmax, beats/min Reid (1984)a 62 4.31 beats/ 9.36, .54 No difference in change
Nosworthy min 17.98 in HRmax among the IMT
(1992)b group compared with
Larson (1999)b the exercise group
VCO2max, L/min Reid (1984) 36 0.11 L/min 0.22, 0.00 .05 No difference in change
Larson (1999) in VCO2max among the IMT
group compared with
the exercise group
Abbreviations: cmH2O  centimeters of water; HRmax  heart rate maximum; IMT  inspiratory resistance training; L/min  liters per minute, max  maxi-
mal; min  minute; 12MWD  12-minute walking distance test; PImax  maximal inspiratory pressure; VCO2max  maximum carbon dioxide uptake;
VEmax  maximum minute ventilation; VO2max  maximum oxygen uptake.
Statistical significance P  .05.
a
Maximal exercise test.
b
Submaximal exercise test.

Combined IMT and Exercise Compared Intensity of cycle exercise was initiated at 50% work
With Exercise Alone rate maximum31,36 or within 10 beats of 60% heart rate
maximum29 and progressed as tolerated. Two studies
Three of the 18 included studies compared combined used a targeted and/or threshold mode of IMT29,31
IMT and exercise with exercise alone.29,31,36 Exercise and 1 study used a normocapnic hyperventilation
in all 3 studies included aerobic training on a station- mode of IMT.36 Eight meta-analyses were performed
ary cycle ergometer for at least 20 minutes, at least with a combination of 3 studies (Table 4).29,31,36
3 days per week, for at least 8 weeks. One study
included treadmill training, stretching, and calisthen- Inspiratory Muscle Strength
ics with and without small weights in addition to the One meta-analysis was performed for inspiratory mus-
cycle ergometer training.36 Larson and colleagues31 cle strength. Results showed a significant improvement
included interval training whereas the other 2 studies in PImax of 8.60 cmH2O (95% CI: 2.55, 14.65) favoring
included continuous aerobic cycle training.29,36 participants in the combined IMT and exercise group

136 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141 www.jcrpjournal.com

JCRP2802_128-141 7/3/08 20:23 Page 137

T a b l e 4 • RESULTS OF SUBGROUP META-ANALYSES—COMBINED IMT AND EXERCISE

VERSUS EXERCISE ALONE
Individual Number of
Studies Participants Weighted 95% P value of
Included in Included in Mean Confidence Overall
Outcomes Meta-analysis Meta-analysis Difference Interval Effect Interpretation
Inspiratory muscle strength
PImax, cmH2O Larson (1999) 57 8.60 cmH2O 2.55, 14.65 .005 Statistically significant increase in
Mador (2005) PImax favoring the combined IMT
and exercise group compared
with the exercise alone group
Exercise tolerance
VEmax, L/min Wanke (1994) 99 1.05 L/min –5.95, 8.05 .77 No difference in change in VEmax
Larson (1999) among the combined IMT and
Mador (2005) exercise group compared with the
exercise alone group
VO2max, L/min Wanke (1994) 99 0.03 L/min –0.09, 0.14 .42 No difference in change in VO2max
Larson (1999) among the combined IMT and
Mador (2005) exercise group compared with the
exercise alone group
Maximum Wanke (1994) 99 0.14 L 0.08, 0.19 .00001 Statistically significant increase in TV
exercise TV, L Larson (1999) favoring the combined IMT and
Mador (2005) exercise group compared with the
exercise alone group
WRmax, W Wanke (1994) 99 2.48 W –4.02, 8.98 .46 No difference in change in WRmax
Larson (1999) among the combined IMT and
Mador (2005) exercise group compared with the
exercise alone group
HRmax, Wanke (1994)a 99 5.02 beats/min –3.31, 13.35 .24 No difference in change in HRmax
beats/min Larson (1999)b among the combined IMT and
Mador (2005)a exercise group compared with the
exercise alone group
Quality of life
CRQ—dyspnea Larson (1999) 57 –1.94 –2.88, –1.01 .0001 Statistically significant greater
Mador (2005) improvement in the Dyspnea
Scale score of the CRQ among the
exercise only group compared
with the combined IMT and exer-
cise group
CRQ—fatigue Larson (1999 57 0.23 3.85, 3.40 .90 No difference in change in Fatigue
Mador (2005) Scale score of the CRQ among the
combined IMT and exercise group
compared with the exercise alone
group

Abbreviations: CRQ  Chronic Respiratory Questionnaire; cmH2O  centimeters of water; IMT  inspiratory resistance training; HRmax  heart rate
maximum; L/min  liters/minute; max  maximal; min  minute; PImax  maximal inspiratory pressure; TV  tidal volume; VEmax  maximum minute
ventilation; VO2max  maximum oxygen uptake; W  watts; WRmax  work rate maximum.
Statistical significance P  .05.
a
Maximal exercise test.
b
Submaximal exercise test.

compared with participants in the exercise-alone group cle strength, reporting an increase in maximal transdi-
(Table 4). Wanke and colleagues29 were not included in aphragmatic pressure (a measure of diaphragmatic
the meta-analysis, because of the different outcomes of rather than inspiratory muscle force) among participants
inspiratory muscle strength used. However, this study in the combined IMT and cycling group compared to
showed similar directional changes of inspiratory mus- those in the cycling only group.29

www.jcrpjournal.com IMT and Other Rehabilitation Interventions in COPD / 137

JCRP2802_128-141 7/3/08 20:23 Page 138

Inspiratory Muscle Endurance (2) combined IMT and exercise with exercise alone.
No meta-analyses updates could be performed for Results of meta-analyses found no significant differ-
inspiratory muscle endurance due to the variability of ence in effect for outcomes of inspiratory muscle
outcomes measured. Two individual studies reported strength and exercise tolerance among patients with
significant increases in inspiratory muscle endurance COPD who engage in IMT compared with exercise,
time in the combined IMT and exercise group com- but suggest potential improvements in inspiratory
pared with the exercise-only group.29,36 The other muscle strength and one outcome of exercise toler-
study found increases among both groups posttrain- ance for patients with COPD who engage in com-
ing with no significant difference between groups.31 bined IMT and exercise compared with exercise
alone. This implies that for some individuals with
Exercise Tolerance COPD, supplementing an aerobic exercise program
Five meta-analyses were performed for outcomes of with IMT may improve inspiratory muscle strength.
exercise tolerance. Results showed significant increas- Added benefits of IMT with exercise pertaining to
es in maximum exercise tidal volume (TV) of 0.14 L other outcomes of inspiratory muscle endurance, dys-
(95% CI: 0.08, 0.19) favoring participants in the com- pnea, and QOL were not supported and hence
bined IMT and exercise group compared with partici- remain unclear.
pants in the exercise-alone group (Table 4). Results This update of the IMT versus exercise subgroup
for the other 4 meta-analyses were not significant analysis concurs with results of the original review,
(Table 4). Individual study results showed greater which also comprised meta-analyses for outcomes of
increases in VO2max,29 VEmax,29 work rate maximum inspiratory muscle strength and endurance.13 Similar
(WRmax),29 and endurance time36 among the combined to the original review, results found no differences in
IMT and exercise compared with exercise alone, change in PImax between the IMT and exercise
whereas Larson and colleagues31 did not find signifi- groups. While the original review found small but sig-
cant differences between groups for these outcomes, nificant increases for VEmax (6.00 L/min) and VO2max
nor for HRmax and TV. Similarly, Mador and col- favoring the exercise group compared with the IMT
leagues36 found significant posttraining increases in group,13 when the study by Ries and Moser20 was
both groups for the 6-minute walk distance (6MWD), included in the updated meta-analyses, the differ-
VO2max, and WRmax but no differences in improvements ences between groups for these outcomes became
achieved after training between groups. nonsignificant.
Dyspnea The second subgroup analysis comparing com-
No meta-analysis updates could be performed for bined IMT and exercise with exercise alone was new
dyspnea outcomes. Refer to individual study results to this update. Hence, we focus the majority of our
from Larson and colleges.31 discussion on these findings.
Results showed a significant improvement in PImax
Quality of life—Dyspnea and Fatigue of 8.60 cmH2O favoring participants who engaged in
Scales of the CRQ combined IMT and exercise compared with exercise
Two meta-analyses were performed on scales of the alone. This positive result concurs with 2 subgroup
CRQ. Results showed significantly greater improve- analyses reported in our original review that found an
ments in the CRQ Dyspnea Scale of 1.94 points (95% increase in PImax of 14.00 cmH2O favoring targeted
CI: 2.88, 1.01) favoring exercise alone compared with and/or threshold IMT compared with education, and
combined IMT and exercise (Table 4). No difference an increase of 6.53 cmH2O favoring targeted and/or
in change the CRQ Fatigue Scale was found between threshold IMT compared with exercise.13
the 2 groups (Table 4). Individual study results found One of the meta-analyses for exercise tolerance
significant improvements within groups after training, (TV) suggested added benefits of IMT in combination
but not between the combined IMT and exercise with exercise in comparison to exercise alone. Both
group compared with the exercise-alone group for comparison groups received cycling as a component
the Fatigue and Dyspnea Scales31,36 and the Emotion of the exercise training, which likely explains the sim-
and Mastery Scales.36 No other QOL outcomes were ilar improvements for exercise tolerance found in
measured within these 3 studies. both groups for the other meta-analyses. One indi-
vidual study demonstrated significantly greater
improvements for outcomes of exercise tolerance in
DISCUSSION the combined IMT and exercise group; however, it
included a higher intensity and frequency of IMT
The update of this systematic review focused on 2 sub- (70%–80% PImax, 7 days per week)29 compared with
group analyses comparing (1) IMT with exercise and the other 2 studies that included lower intensities and

138 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141 www.jcrpjournal.com

JCRP2802_128-141 7/3/08 20:23 Page 139

frequencies (30–60% PImax 5 days per week31 and endurance with combined IMT and exercise.29,36
72% of maximum voluntary ventilation (MVV) for 3 Lotters and colleagues11 found no additional effect of
days per week.36 One study involved a home-based IMT on exercise capacity. This was similarly found
program31 with weekly nurse visits whereas the other with all 5 meta-analyses for outcomes of exercise tol-
2 studies involved supervision at each session by a erance in the IMT versus exercise comparison
physical therapist29 and verbal encouragement.36 This (Table 3) and 4 of 5 of the meta-analyses in the com-
may explain why Wanke and colleagues29 and Mador bined IMT and exercise versus exercise-only compar-
and colleagues36 reported significantly greater ison (Table 4). We agree with the position papers that
improvements in inspiratory muscle endurance IMT is of potential benefit to inspiratory muscle
whereas Larson and colleagues31 did not find any strength and endurance, but more studies are required
greater improvements with IMT combined with exer- to specifically examine the effects on exercise toler-
cise for inspiratory muscle endurance nor dyspnea. ance and QOL6,37,38 and whether a group of COPD
These results indicate that a higher training intensity patients with similar severity might demonstrate
and frequency of IMT with supervision at each ses- greater benefits from this treatment.
sion may be needed to elicit added benefits of IMT This review continues to be based on a small num-
combined with exercise for outcomes of exercise tol- ber of trials (n  18), only 5 of which were able to
erance, inspiratory muscle endurance, and dyspnea. be included in the IMT versus exercise subgroup
Results pertaining to QOL were less evident. analysis19,20,22,27,31 and 3 of which were able to be
Despite results from individual studies that reported included in the combined IMT and exercise versus
nonsignificant differences between the combined IMT exercise-alone subgroup analysis.29,31,36 The ability to
and exercise group and exercise-alone group for the perform meta-analyses was limited because of differ-
CRQ Dyspnea and Fatigue Scales, meta-analysis yield- ences in modes of IMT, types of interventions among
ed a significant result favoring exercise alone for the the comparison groups and types of outcomes
CRQ Dyspnea Scale. Results from our previous sys- assessed in the included studies. Seven meta-analyses
tematic review demonstrated a significant improve- were performed in the original review, each of which
ment in the CRQ Dyspnea Scale among participants included a combination of only 2 studies with a sam-
training with IMT compared to those who received ple size of 35 or 52 participants.13 Thirteen new meta-
education alone.13 Hence, further research is required analyses were performed in this update, each of
to explore the effect of IMT in combination with exer- which included 2 or 3 studies with a slightly larger
cise for outcomes of QOL for adults with COPD. sample size ranging from 36 to 99 participants. While
In this update, we combined targeted and/or still fraught with limited studies and small sample
threshold and normocapnic hyperventilation modes sizes, this systematic review provides a rationale for
of IMT in the subgroup analyses given both modes future studies and serves as the foundation for future
involve working toward a targeted workload. This updates to incorporate studies with additional out-
enabled us to perform a new subgroup analysis for comes of inspiratory muscle strength and endurance,
combined IMT and exercise versus exercise alone by exercise tolerance, dyspnea and QOL.
including the newly identified article by Mador and Results of these 2 subgroup analyses presented
colleagues,30 and update the subgroup analysis for here should be interpreted cautiously. Four of the
IMT versus exercise in the original review by includ- 8 studies conducted a per-protocol analysis by
ing the article by Ries and Moser.20 Combining these excluding participants from the analysis who were
2 modes did not appear to yield problems with het- nonadherent with the intervention.20,27,29,31 Hence,
erogeneity. Nevertheless, when interpreting the results these studies assessed efficacy rather than effective-
of this review, one should consider the different phys- ness of IMT. The studies included in the subgroup
iologic effects associated with these 2 modes of IMT. analyses consisted of small sample sizes ranging from
This review is in agreement with the meta-analysis 12 to 130 participants in the total study at baseline.
conducted by Lotters and colleagues,11 who reported Furthermore, these studies demonstrated a range of
greater benefits in inspiratory muscle strength (PImax) withdrawal rates from 0% to 59%, suggesting that the
and inspiratory muscle endurance time (in seconds) validity of these studies may be questioned.
among combined IMT and general exercise recondi- In summary, the update of this systematic review
tioning compared with a general exercise recondi- suggests that combining IMT with other exercise
tioning alone.11 While we were unable to perform interventions at least 3 times per week for at least 8
meta-analyses for inspiratory muscle endurance, 2 of weeks may be beneficial for increasing inspiratory
the 3 individual studies in the combined IMT and muscle strength and one outcome of exercise toler-
exercise versus exercise-alone subgroup analysis sug- ance for individuals with COPD. Individual study
gested similar improvements in inspiratory muscle results suggest potential added benefits with IMT for

www.jcrpjournal.com IMT and Other Rehabilitation Interventions in COPD / 139

JCRP2802_128-141 7/3/08 20:23 Page 140

outcomes of inspiratory muscle endurance. Results 12. Geddes EL, Reid WD, Crowe J, O’Brien K, Brooks D.
pertaining to dyspnea and QOL outcomes are less Inspiratory muscle training in adults with chronic obstructive
pulmonary disease: a systematic review. Respir Med.
clear. Further trials are required to investigate the 2005;99:1440–1458.
effect of IMT or combined IMT plus other rehabilita- 13. Crowe J, Reid WD, Geddes L, O’Brien K, Brooks D. Inspiratory
tion interventions compared to other rehabilitation muscle training compared with other rehabilitation interven-
interventions without IMT for outcomes of inspirato- tions in adults with chronic obstructive pulmonary disease.
ry muscle endurance, exercise tolerance, dyspnea, COPD: J COPD. 2005;3:319–329.
14. Higgins JPT, Green S, eds. Cochrane handbook for systematic
and QOL. Further research also is needed to deter- reviews of interventions 4.2.6 [updated September 2006]. In:
mine the optimal IMT frequency, intensity, and time The Cochrane Library, Issue 4, Chichester, UK: John Wiley and
to maximize outcomes for individuals with COPD, Sons Ltd; 2006:1–265.
and to determine whether a specific subgroup might 15. Druz WS, Danon J, Fishman HC, Goldberg NB, Moisan TC,
demonstrate greater benefits from this intervention. Sharp JT. Approaches to assessing respiratory muscle function
in respiratory disease. Am Rev Respir Dis. 1979;119(2, Pt
2):145–149.
—Acknowledgments— 16. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ,
This study was supported by the Canadian Lung McQuay HJ. Assessing the quality of randomized controlled
Association and Ontario Respiratory Care Society trials: is blinding necessary? Control Clin Trials. 1996;
(Ontario Lung Association). Kelly O’Brien is support- 17:1–12.
17. Review Manager (RevMan) [Computer Program]. Version 4.2
ed by a Fellowship from the Canadian Institutes of for Windows. Oxford, England: The Cochrane Collaboration;
Health Research (CIHR). Dina Brooks was supported 2002.
by New Investigator Award during the completion of 18. Lau J, Ioannidis JPA, Schmid CH. Quantitative synthesis in sys-
this project. tematic reviews. Ann Intern Med. 1997;127:820–826.
The authors thank Joanna Gorny, research assis- 19. Reid WD, Warren CPW. Ventilatory muscle strength and
endurance training in elderly subjects and patients with chron-
tant, for her assistance during the early stages of this ic airflow limitation. A pilot study. Physiother Can.
study. 1984;36:305–311.
20. Ries AL, Moser KM. Comparison of isocapnic hyperventilation
and walking exercise training at home in pulmonary rehabili-
References tation. Chest. 1986;90:285–289.
1. Lacasse Y, Brooks D, Goldstein RS. Trends in the epidemiolo- 21. Chen H-I, Dukes R, Martin BJ. Inspiratory muscle training in
gy of COPD in Canada. 1980–1995. Chest. 1999;116:306–313. patients with chronic obstructive pulmonary disease. Am Rev
2. Canadian Institute for Health Information, Canadian Lung Respir Dis. 1985;131:251–255.
Association, Health Canada, Statistics Canada. Respiratory 22. Jones DT, Thomson RJ, Sears MR. Physical exercise and resis-
Disease in Canada. Ottawa, Canada. http://www.phac- tive breathing in severe chronic airways limitation–-are they
aspc.gc.ca/publicat/rdc-mrc01/pdf/rdc0901e.pdf. Published effective? Eur J Respir Dis. 1985;67:159–166.
2001. Accessed January 5, 2007. 23. Levine S, Weiser P, Gillen J. Evaluation of a ventilatory muscle
3. Gosker HR, Wouters EF, van der Vusse GJ, Schols AM. Skeletal endurance training program in the rehabilitation of patients
muscle dysfunction in chronic obstructive pulmonary disease with chronic obstructive pulmonary disease. Am Rev Resp Dis.
and chronic heart failure: underlying mechanisms and therapy 1986;133:400–406.
perspectives. Am J Clin Nutr. 2000;71:1033–1047. 24. Noseda A, Carpiaux JP, Vandeput W, Prigogine T, Schmerber J.
4. American Thoracic Society and European Respiratory Society. Resistive inspiratory muscle training and exercise performance
ATS/ERS Statement on Pulmonary Rehabilitation. Am J Respir in COPD patients. A comparative study with conventional
Crit Care. 2006;173:1390–1413. breathing retraining. Bull Eur Pathophysiol Respir.
5. Canadian Thoracic Society Review Panel. State of the art com- 1987;23:457–463.
pendium: Canadian thoracic society recommendations for the 25. Goldstein R, De Rosie J, Long S, Dolmage T, Avendano MA.
management for chronic obstructive pulmonary disease. Can Applicability of a threshold loading device for inspiratory mus-
Resp J. 2004;11:7B–59B. cle testing and training in patients with COPD. Chest.
6. Laghi F, Tobin MJ. Disorders of the respiratory muscles. Am J 1989;96:564–571.
Respir Crit Care Med. 2003;168:10–48. 26. Dekhuijzen PNR, Folfering HTM, van Herwaarden CLA. Target-
7. Soicher J, Dechman G. Inspiratory muscle function in chronic flow inspiratory muscle training during pulmonary rehabilita-
obstructive pulmonary disease. Phys Ther Rev. 1998;3:31–39. tion in patients with COPD. Chest. 1991;99:128–133.
8. Lacasse Y, Brosseau L, Milne S, et al. Pulmonary rehabilitation 27. Nosworthy J, Barter CE, Thomas SA, Flynn MG. An evaluation
for chronic obstructive pulmonary disease. Cochrane Database of three elements of pulmonary rehabilitation. Aust J
Syst Rev. 2002;3:CD003793. Physiother. 1992;38:189–193.
9. Reid WD, Geddes EL, Brooks D, O’Brien K, Crowe J. 28. Weiner P, Azgad Y, Ganam R. Inspiratory muscle training com-
Inspiratory muscle training in chronic obstructive pulmonary bined with general exercise reconditioning in patients with
disease. Physiother Can. 2004;56:128–142. COPD. Chest. 1992;102:1351–1356.
10. Smith K, Cook D, Guyatt GH, Madhavan J, Oxman AD. 29. Wanke T, Formanek D, Lahrmann H, et al. Effects of combined
Respiratory muscle training in chronic airflow limitation: a inspiratory muscle and cycle ergometer training on exercise
meta-analysis. Am Rev Respir Dis. 1992;145:533–539. performance in patients with COPD. Eur Respir J.
11. Lotters F, van Tol B, Kwakkel G, Gosselink R. Effects of con- 1994;7:2205–2211.
trolled inspiratory muscle training in patients with COPD: a 30. Berry MJ, Adair NE, Sevensky KS, Quinly A, Lever HM.
meta-analysis. Eur Respir J. 2002;20:570–576. Inspiratory muscle training and whole-body reconditioning in

140 / Journal of Cardiopulmonary Rehabilitation and Prevention 2008;28:128–141 www.jcrpjournal.com

JCRP2802_128-141 7/3/08 20:23 Page 141

chronic obstructive pulmonary disease. A controlled random- and quality of life. Am J Respir Crit Care Med. 2000;162:
ized trial. Am J Respir Crit Care Med. 1996;153:1812– 1709–1714.
1816. 35. Oh EG. The effects of home-based pulmonary rehabilitation in
31. Larson JL, Covey MK, Wirtz SE, et al. Cycle ergometer and patients with chronic lung disease. Int J Nurs Stud.
inspiratory muscle training in chronic obstructive pulmonary 2003;40:873–879.
disease. Am J Respir Crit Care Med. 1999;160:500–507. 36. Mador JM, Deniz O, Aggarwal A, Shaffer M, Kufel TJ, Spengler
32. Covey MK, Larson JL, Wirtz SE, et al. High intensity inspirato- CM. Effect of respiratory muscle endurance training in patients
ry muscle training in patients with chronic obstructive pul- with COPD undergoing pulmonary rehabilitation. Chest.
monary disease and severely reduced function. J Cardiopulm 2005;128:1216–1224.
Rehabil. 2001;21:231–240. 37. ACCP/AACVPR Pulmonary Rehabilitation Guidelines Panel.
33. Minoguchi H, Shibuya M, Miyagawa T, et al. Cross-over com- Pulmonary rehabilitation. Joint ACCP/AACVPR Evidence-based
parison between respiratory muscle stretch gymnastics and guidelines. Chest. 1997;112:1363–1396.
inspiratory muscle training. Int Med. 2002;41:805–812. 38. Pauwels RA, Buist AS, Claverley PMA, Jenkins CR, Hurd SS.
34. Scherer TA, Spengler CM, Owassapian D, Imhof E, Boutellier Global strategy for the diagnosis, management, and prevention
U. Respiratory muscle endurance training in chronic obstruc- of chronic obstructive pulmonary disease. Am J Respir Crit
tive pulmonary disease. Impact on exercise capacity, dyspnea Care Med. 2001;163:1256–1276.

www.jcrpjournal.com IMT and Other Rehabilitation Interventions in COPD / 141