Patient Education and Counseling 98 (2015) 182–190

Contents lists available at ScienceDirect

Patient Education and Counseling

journal homepage: www.elsevier.com/locate/pateducou

Intervention

Effects of guided deep breathing on breathlessness and the breathing

pattern in chronic obstructive pulmonary disease: A double-blind
randomized control study
Christine R. Borge a,b,1,*, Anne Marit Mengshoel a, Ernst Omenaas c, Torbjørn Moum d,
Inger Ekman e, Martha P. Lein b, Ulrich Mack b, Astrid K. Wahl a
a
Department of Health Sciences, University of Oslo, Norway
b
Department of Medicine, Lovisenberg Diaconal Hospital, Oslo, Norway
c
Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway
d
Department of Behavioral Sciences in Medicine, University of Oslo, Norway
e
Institute of Health and Care Sciences, the Sahlgrenska Academy, University of Gothenburg Sweden and Centre for Person-centered care,
University of Gothenburg, Sweden

A R T I C L E I N F O A B S T R A C T

Article history: Objective: To investigate whether guided deep breathing using a device improves breathlessness, quality
Received 26 April 2014 of life, and breathing pattern in moderate and severe stage of chronic obstructive pulmonary disease
Received in revised form 15 August 2014 (COPD).
Accepted 19 October 2014
Methods: In total, 150 patients participated in a double-blind randomized controlled trial in a four-week
intervention and a four-month follow-up. Participants were randomized into a guided deep breathing
Keywords: group (GDBG), music listening group (MLG), or sitting still group (SSG). The patients’ symptom score
COPD
using the St George’s Respiratory Questionnaire (SGRQ), and a Global Rating Change scale (GRC) was
Breathlessness
Quality of life
applied to measure breathlessness as primary outcome. The activity score and impact score of SRGQ, and
Randomized control trial breathing pattern were secondary outcomes.
Self-management Results: Positive effects of the GDBG were detected in GRC scale in breathlessness at four weeks
(p = 0.03) with remaining effect compared to MLG (p = 0.04), but not to SSG at four months follow-up.
GDBG showed positive effect for respiratory rate (p < 0.001) at four weeks follow-up. A positive
significant change (p < 0.05–0.01) was found in all groups of SGRQ symptom score.
Conclusion: GDBG had a beneficial effect on respiratory pattern and breathlessness. MLG and SSG also
yielded significant improvements.
Practice implications: Guided deep breathing may be used as a self-management procedure.
ß 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Chronic obstructive pulmonary disease (COPD) is demanding for
the individual patient and the symptom breathlessness and impaired
quality of life (QOL) [1,2] increase during the more severe stages of
the disease [3]. Breathlessness is often the reason why patients with
COPD seek medical help. Thus, relieving symptoms and improving
* Corresponding author at: Institute of Health and Society, Department of Health
Sciences, PB 1089 Blindern, 0317 Oslo, Norway. Tel.: +47 22845372;
fax: +47 22845091.
E-mail addresses: c.r.borge@medisin.uio.no, christineraaheim.borge@lds.no
(C.R. Borge).
1 Diaphragmatic breathing, deep breathing, yoga breathing, and
Lovisenberg Diaconal Hospital, Medical Department, 0440 Oslo, Norway.
Tel.: +47 23225000.
QOL are important aims for patient care and treatment approaches
[3]. The use of pharmacological treatments alone is not effective;
thus, additional nonpharmacological approaches such as breath
retraining exercises have been suggested [4].
COPD is caused by a complex array of physical changes such as
chronic obstruction of the airways and destroyed alveoli. This may
lead to an overload or fatigue of respiratory muscles, disturbances of
gas exchange between the lung and blood and air trapping
[4]. Hence, people with COPD may develop an ineffective breathing
pattern in form of an insufficient ventilation, resulting in an
increased respiratory rate (RR), decreased vital capacity, and
decreased time on inspiration (TIN), and expiration (TEX), which
are also associated with symptoms of breathlessness/dyspnea [5,6].
pursed-lip breathing are breathing exercises that can affect the

http://dx.doi.org/10.1016/j.pec.2014.10.017
0738-3991/ß 2014 Elsevier Ireland Ltd. All rights reserved.
 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190
ability [7] to improve an ineffective breathing pattern, to breathe
deeply and, thereby reducing breathlessness.
Studies have investigated the effects of diaphragmatic, pursed-
lip breathing, and yoga breathing, but few were randomized
control trials (RCTs) [8–11]. For instance, a recent systematic
review by Holland et al. [10], based on pooled data from two RCTs
reports that practicing pursed lip breathing gave less breathless-
ness. In the same review, one RCT on diaphragmatic breathing and
one on yoga breathing showed effects on disease related QOL. All
the said studies were rated as being of low to moderate quality and
the results were interpreted as inconsistent, indicating that
additional high quality RCTs are required. In addition, other
interventions such as music listening alone and in combination
with breathing control exercises have been found to reduce
breathlessness in patients with COPD [12]. In order to refine
the evidence of BCEs it is important to design breathing control
trials that control for other interventions that might reduce
breathlessness.
Previous studies that employed a biofeedback device to guide
the user to breathe more deeply have reported positive effects on
symptoms [13], blood pressure [14], and QOL [13,14] in patients
with heart failure [13] and hypertension [14]. In the biofeedback
device a voice guides the users through air phones/plugs telling
them to follow music tones on inspiration and expiration based on
the respiratory movement, measured with a sensor belt, while
breathing patterns are stored in the device [15].
Such device-guided breathing control based on musical tones
and instruction has not been investigated in patients with COPD
and we hypothesize that it may guide patients to breathe more
deeply and slowly and have positive effects on their breathing
pattern and breathlessness.
In the present study, we performed a three-armed double-blind
RCT to investigate whether subjects in the moderate and severe
stages of COPD reported beneficial effects on breathlessness, QOL,
and their breathing pattern after a four weeks intervention
program using device-guided breathing control and at four months
follow-up compared with a group who listened to music and a
sham control group who sat still.
2. Methods
2.1. Research design
We performed a three-armed, parallel group, double-blind,
randomized RCT where the patients were in the moderate or
severe stages of COPD. The arms comprised a guided deep
breathing group (GDBG), a music listening group (MLG), and a
sham group sitting still (SSG).
In total, 150 patients with COPD who had been diagnosed with
moderate and severe stages of COPD according to the Global
Initiative for Chronic Obstructive Lung Disease criteria participated
in this study [3]. The patients were enrolled between July 2011 and
September 2013 from outpatient units at three different hospitals
in Oslo, Norway (Hospitals A, B, and C) following similar treatment
procedures. All the participants were offered free transportation to
attend individual appointments at the main hospital A.
2.2. Recruitment
At each hospital nurses and doctors were instructed to screen
for and recruit eligible patients with moderate and severe stage
COPD. Patients were given oral and written information at each
hospital before they consented to be contacted by telephone by the
researcher or a study nurse to confirm their willingness to
participate in the study as well as fulfilling the exclusion and
inclusion criteria (Table 1).
183
Table 1
Inclusion and exclusion criteria.
Inclusion criteria Exclusion criteria
Diagnosis of COPD in medical records Change in medication during the last
with moderate stage or severe stagea four weeks.
at inclusion.
MRC dyspnea scaleb1 at inclusion. Diagnosis of cancer.
Able to read, write, and speak Presently attending a pulmonary
Norwegian rehabilitation course or other similar
COPD education course.
Attending a competing study.
Diagnosis of neuromuscular disease
or dementia.
Present drug abuse or alcohol abuse.
Receiving help from a pulmonary
physiotherapist.
a
Moderate stage and severe stage = FVC < 70% and FEV1%  30% and  80% of
predicted values (FVC = forced volume capacity, FEV1 = forced expiratory volume
in 1 s).
b
MRC (Medical Research Council) dyspnea scale measures disabilities associated
with breathlessness using five levels (0–4), where a higher score represents more
disabilities associated with breathlessness.
2.3. Ethics
The study was approved by the data protection supervisor of
the three hospitals and by the Regional Committee for Medical
Research Ethics for Southern Norway (reference number: 2010/
1521), as well as reported to the Clinical Trials government registry
(ClinicalTrials.gov identifier: NCT015120043). The protocol com-
plied with the Declaration of Helsinki. All participants signed a
written consent before entering the study.
2.4. Intervention
The participants were instructed orally and in writing by a
study nurse on how to use the device (RespeRate, InterCure Inc.,
New York, USA; www.Resperate.com/MD) [13,15]. The device was
used for the first time at the hospital for 15 min and later at home,
twice a day (i.e. morning and evening) for four weeks. When using
the device the participants were told to sit down in a comfortable
chair with few interrupting elements around them.
During the intervention, the patients used the same device to
measure their breathing pattern, but different instructions were
given via earphones/earplugs.
The GDBG group members used the device and received
instructions about how to breathe slowly based on their RR, which
was measured by a sensor belt that was placed around the waist.
The device played soft, non-rhythmic music in the background, and
a voice instructed the participants to breathe out while the musical
note lasted during expiration, whereas a new musical note
followed during inspiration.
The MLG group members listened to the same music being
played in the background as the GDBG group members, but they
were not given any instructions to breathe slowly.
The SSG members only received an introductory instruction to
sit down and listen to the same music for 1–2 min, but without any
instructions about breathing or music during the at the rest of the
session.
2.5. Data collection
A questionnaire booklet covering self-reported outcomes
related to sociodemographic variables, breathlessness, and QOL
was sent by postal mail. Participants were asked to complete the
questionnaire at home 1–2 days before attending the main project
hospital at baseline (T1), in the follow up after four weeks (T2) and
four months after baseline (T3).
184 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190
At T1 the questionnaire was checked for items with non-response
and physical tests (i.e. pulmonary function test, blood gas sample)
were performed by the main researcher before the participants were
allocated to the assigned group. At T2 and T3 the same physical tests
as baseline were performed by the main researcher who was blinded
for group assignment and the questionnaire was checked for items
with non-response by the study nurse. The study nurse called the
participants after two days and then subsequently once a week
during the four weeks intervention period to obtain information
about the RR and the quality of the measured RR. This information
was obtained from the device guided with detailed instruction by
the study nurse. The same data collection was performed for all
groups (i.e. GDBG, MLG, SSG).
2.6. Outcomes
The primary outcome was measured in two ways: relief of
symptoms using the St George’s Respiratory Questionnaire (SGRQ)
and reported changes in breathlessness using the Global Rating of
Change scale (GRC). The secondary outcomes were scores for
diseases related to QOL (i.e., activity and impact according to the
SGRQ) and the breathing pattern (i.e., TIN, TEX, and RR).
The SGRQ (four-week version) [16] comprises 50 items that yield
one total score, but it can also be divided into three subcategories:
symptoms (i.e., out of breath, coughing, wheeze, and sputum),
activity limitation (i.e., breathlessness during various tasks and
physical activity), and impact of the disease (i.e., how illness and
breathlessness affect social, emotional, and family life). The ranges of
the scores were 0–100, where a high score indicated more
limitations. The SGRQ has been tested thoroughly to confirm its
reliability and validity in previous studies [17]. In the present study,
Cronbach’s alpha was 0.79 for the symptom score, 0.82 for the
activity score, 0.89 for the impact score, and 0.90 for the total score.
The GRC scale was applied to assess change in breathlessness.
The GRC scale was chosen because it is simple, easy to understand
for patients and gives responders the opportunity to quantify their
evaluation of improvement over time [18] with regard to
breathlessness. We asked whether the participants experienced
a positive change in breathlessness after using the device for four
weeks and at the four months follow-up, on a numeric rating scale
anchored 0–10, where 0 = no change and 10 = a high degree of
change. A similar instrument called the vertical visual analogue
scale (VAS) is known to be valid for measuring dyspnea in asthma
and COPD patients [19]. The NRS and VAS are highly correlated, but
NRS is recommended because it is easier to use [20].
The RespeRate breathing device used in the study [15]
comprised a belt-type respiration sensor and a computerized
control unit. The system analyzed and stored detailed data for the
breathing pattern variables, i.e., RR, TIN, and TEX (per minute
during each 15-minute session), the number of sessions, and the
breath detection results for the members of all the groups.
Specialized software was prepared for the MLG and the SSG groups.
The measures of breathing pattern variables were available during
the period between T1 and T2.
The lung function values (i.e., forced vital capacity (FVC), forced
expiratory volume per second (FEV1), the ratio of FEV1/
FVC = FEV%, and the vital capacity) [21] were measured by
spirometry (MasterScreen Bodyplethysmography, Jaeger,
Germany) and the blood gases were sampled (i.e., pressure of
oxygen (PO2) and pressure of carbon dioxide tension (PCO2) [22];
Roche OMNI C, Mannheim, Germany).
2.7. Randomization
A person who was not involved in the project was responsible
for randomizing the participants into the GDBG, MLG, and SSG
groups. Envelopes were prepared and the patients were numbered
(1–120). The study was randomized with 14 blocks, with 14–16
patients in each block and four to six patients in each group. First,
the block was drawn. Next, a coded name from the group was
drawn in each block and put into a sealed envelope.
2.8. Blinding
The study nurse made sure that the participants and the main
researcher received no information about whether the participants
had been allocated to guided breathing, music, or sitting still.
However, the participants were informed that the aim of the
project was to test if relaxation, silence, or breathing could change
the experience of breathlessness in people with COPD. The
participants were told not to discuss the content of their device
with the main researcher and such contact with the main
researcher was also minimized after they had received the device.
An external person entered all the data in the statistical software
and data was not analyzed before the data collection was
completed.
2.9. Power calculation
We calculated the sample size using the main outcome variable,
i.e., the symptom score according to the SGRQ, based on the effects
found in a COPD rehabilitation study [23] where the threshold of
change >4 [24]. We aimed for an expected effect of a difference of
eight points between groups, standard deviation (SD) of 16,
significance level of 5%, and a power of 80%; thus, we estimated a
necessary sample size of 150 in total, with 50 in each group. We
oversampled each group by 20 patients (total n = 210), but stopped
inclusion at n = 150 because of difficulties recruiting more patients.
2.10. Analysis
SPSS statistical package version 22 (SPSS Inc, Chicago, IL, USA)
was used to perform the analyses. Descriptive analyses were
performed for sociodemographic and clinical variables, and for the
scores obtained using the questionnaires. Analysis of covariance
(ANCOVA) was used to assess differential changes between groups
from T1 to T2 and T1 to T3, while controlling for the T1 values
(General Linear Model procedure in SPSS). Paired Student’s t-tests
were used to assess changes within groups from T1 to T2 and T1 to
T3. Wilcoxon signed ranks test was performed when criteria of
normally distributed variables were not met. ANOVAs were used to
assess possible differences in the means between groups at T1. We
used the first measured variables from the first session and last
session for TIN, TEX and RR in the analyses. An intention-to-treat
analysis (last carry forward) was performed (i.e. n = 8 at T2 and
n = 16 at T3 for the SGRQ symptoms score, n = 12 at T3 for GRC
scale). These analyses included x–y additional subjects (depending
on outcome) and revealed only marginal differences in the mean
outcomes, with identical conclusions regarding p-values to the
analyses performed as standard protocol. Thus standard protocol
results have been reported in our tables. Interaction analyses were
performed for the main outcomes to test whether the effect of the
intervention differed significantly between specific subgroups of
patients as defined by age, gender, infection, smoking status, COPD
stage, co-morbidities and living status. The interaction terms were
entered one at a time in the ANCOVAs while retaining the main
effects in the model. We calculated and interpreted effect sizes
according to Cohen [25]. The differences within groups were
calculated by dividing the mean changes by the average SD at T1
and T2, and T1 and T3, respectively, and between groups by
dividing the difference of the mean changes in the experimental
and the control groups by their average SDs.
 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190
3. Results
3.1. Patient enrollment
Among 341 patients who were eligible, 150 patients (44%
response rate) with moderate or severe stages of COPD agreed to
participate and were randomized.
The randomization procedure allocated 51 patients to the GDBG
group, 50 patients to the MLG groups, and 49 patients to the SSG
group. During the four-week intervention, two patients dropped out
from each of the GDBG and MLG groups, whereas three patients
dropped out from the SSG group. Four of these patients turned in
their device, but data from these were not considered in the analysis.
Seven patients in the GDBG group, eight patients in the MLG
group, and five in the SSG group were lost to follow-up at four months
after baseline. Four of these patients turned in their questionnaire by
postal mail. These patients received the intervention and are
therefore included in the analysis of subjective outcomes.
Fig. 1 shows further information.
All the three groups used the device an equal number of times in
terms of the mean number of sessions conducted during the four
weeks: the GDBG group with a mean of 50.3 times (SD = 10.5), the
MLG group with a mean of 49.2 times (SD = 11.6), and the SSG
group with a mean of 50.8 times (SD = 10.2).
3.2. Demographic and clinical characteristics
Table 2 shows the clinical characteristics of the participants at
baseline. Predicted lung function is significantly higher in the SSG
compared to GDBG and MLG (p < 0.05). Further number of
Fig. 1. Flow diagram of enrollment.
185
comorbidities was significantly lower in the MLG compared to
GDBG and SSG at T1 (p < 0.05).
3.3. Effect of deep breathing
3.3.1. Primary outcomes
There was statistically significant within-group decrease in the
SGRQ symptom score (p  0.05–0.001) in all the groups from T1 to
T2, and for MLG and SSG (p  0.05–0.001) from T1 to T3. There
were no significant changes in the SGRQ symptom scores between
groups from T1 to T2 and from T1 to T3.
The GRC scale for breathlessness detected a significant positive
difference with GDBG (p = 0.03) at T2. At T3 the GRC scale GDBG
was significantly different from MLG (p = 0.04), whereas there
were no significant differences when GDBG was compared to SSG
(not reported in table).
No strong systematic interaction effects were found between
groups on the primary outcome (i.e. symptom score T2 and gender
(eta square = 5%), GRC scale and gender T3 (eta square = 7%)).
3.4. Secondary outcomes
For the activity score and the total SGRQ score, there were no
significant differences between groups at T1 and no significant
changes between groups at T2 and T3. Further information is
provided in Table 3.
For the whole GDBG group, the mean percentage following the
instructions given by the device was 53.3 (SD = 23). For all groups
there were no significant differences between groups of the sensor
recognizing the breathing pattern.
186 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190

Table 2
Demographic and clinical variables at baseline n = 150.

Variables All participants n = 150 Guided deep breathing Music listening group Sitting still group (SSG)
group (GDBG) n = 51 (MLG) n = 50 n = 49

n (%) Mean (SD) n (%) Mean (SD) n (%) Mean (SD) n (%) Mean (SD)

Age 67.4 (9.0) 67.2 (9.9) 68.4 (8.2) 66.6 (8.9)

Gender
Male 75 (50) 23 (45.1) 30 (60) 22 (44.9)
Female 75 (50) 28 (54.9) 20 (40) 27 (55.1)
Living status
Living alone 81 (57) 30 (63.8) 26 (56.5) 25 (51)
Living with someone 61 (43) 17 (36.2) 20 (43.5) 24 (49)
Education
Primary school 34 (25.6) 13 (30.2) 10 (21.3) 11 (25.6)
Vocational school 18 (13.5) 7 (16.3) 5 (10.6) 6 (14)
Secondary school 38 (28.6) 12 (27.9) 16 (34) 10 (23.3)
University < 4 years 27 (20.3) 6 (14) 11 (23.4) 10 (23.3)
University > 4 years 16 (12) 5 (11.6) 5 (10.6) 6 (14)
Smoking
Yes 60 (40.3) 23 (45.1) 15 (30.6) 22 (44.9)
No 89 (59.7) 28 (54.9) 34 (69.4) 27 (55.1)

Infection in the last four weeks

Yes 26 (17.3) 7(13.7) 9 (18) 10 (20.4)
No 124 (82.7) 44 (86.3) 41 (82) 39 (79.6)
Use of breathing exercises
Yes 55 (37.7) 21 (43.8) 17 (34) 17 (35.4)
No 91 (62.3) 27 (56.3) 33 (66) 31 (64.6)
Pulmonary function
FEV1 predicted 57.6 (17.0) 55.4 (15.1) 54.6 (18.6) 62.7 (16.2)*
RV predicted 178.4 (45.5) 182.2 (49.2) 186.4 (45.8) 166.7 (39.2)
VC predicted 84 (28.2) 82.5 (16.7) 83.2 (20.1) 86.1 (18.1)
COPD stage
Moderate stage 97 (65.1) 31 (60.8) 26 (53.1) 40 (81.6)
**
Severe stage 52 (34.9) 20 (39.2) 23 (46.9) 9 (18.4)
Blood gas analysis
PO2 9.6 (1.3) 9.4 (1) 9.6 (1.4) 9.7 (1.5)
PCO2 5.3 (0.8) 5.2 (0.5) 5.4 (0.8) 5.3 (0.9)
Years with COPD 6.1 (6.1) 6.8 (6.9) 6.3 (6.4) 4.3 (1.8)
Number with comorbidity 3.9 (1.9) 4.3 (1.9) 3.1 (1.6)* 4.3 (1.8)*
*
Significant difference between groups, p < 0.05 according to ANOVA.
**
Significant difference between groups, p < 0.01 according to Pearson Chi-Square.

At T1, there were significant differences between groups in
terms of RR, TIN, and TEX (p  0.001), with positive effects in the
GDBG group.
There were also significant changes in RR (p  0.001) and TEX
(p  0.001) between groups from T1 to T2 at the start and at the end
of sessions, with positive effects in the GDBG group. There was a
significant change for TIN end of session (p  0.001) in favor of
GDBG, but not for TIN start of session between groups from T1 to T2.
Further information is provided in Table 4.
4. Discussion, limitations and conclusion
To the best of our knowledge, this is the first three-armed
double-blind RCT to evaluate the relief of breathlessness using
nonpharmacological interventions of device guided breathing for
patients in the moderate or severe stages of COPD. We found that
all the groups (i.e., GDBG, MLG, and SSG) exhibited changes in their
symptom scores during the four-week intervention but there were
no differences in the effects on the symptom scores between the
groups. We also detected a significant positive effect in terms of the
question that evaluated changes in breathlessness after using the
device and there was a positive effect on the breathing pattern in
the guided deep breathing group.
4.1. Discussion
Although, there was no significant effect for the SRGQ symptom
score, our study demonstrated that the patients in the DGBG group
were able to change their breathing pattern by breathing more
slowly, thereby decreasing their RR and they reported a perceived
change in breathlessness for the GRC scale. An improved breathing
pattern may lead to reduced sympathetic activity [26] and
strengthen the baroreflex to improve breathlessness [11]. The
breathing patterns were improved in the GDBG members from their
first session until their last session, which showed that the patients
appeared to have learned how to breathe slowly. Our results may
justify recommending guided deep breathing by use of a biofeed-
back system as an appropriate self-management support procedure
that could help people with moderate and severe stage of COPD to
take more control over their ineffective breathing pattern and
impairment due to breathlessness. However, we found limited
effects on breathlessness in the GRC scale at four months follow-up,
indicating that continued long-term instructions are needed in order
to achieve an optimal performance of the exercise. A device might
then be more practical to use in a home situation and pulmonary
training programs than guided instruction by a health person.
Although the global rating scale showed effect between
groups, there were no differences between the groups in terms
of the SGRQ symptom score. Other studies have reported
differences between groups using the same outcome measure
in breathing control studies. For example, Katiyar et al. [27]
reported a positive effect of yoga breathing according to all the
SGRQ scores. In addition, Yamaguti et al. [28] detected a difference
in the total SGRQ score where there was a positive effect of
diaphragmatic breathing compared with a usual care group. We
found that breathlessness was reduced in all of our study groups in
Table 3
Mean scores according to St George’s Respiratory Questionnaire (SGRQ) and a numerical rating scale (NRS). Differences between GDBG MLG, and SSG were tested by ANCOVA using the baseline as a covariate. Reported changes in
breathlessness, symptoms, and quality of life (QOL) were estimated at two time points.

Outcomes Group n Baseline After four Cohen Cohen between p-value n Baseline After four Cohen between Cohen between p-value T1–T3
(T1)Mean weeks (T2) between GDBG and MLG, T1–T2 (T1)Mean months (T3) T1 and T3 GDBG and MLG, (whole model)
(SD) Mean (SD) T1 and T2 GDBG and SSG (whole (SD) Mean (SD) and GDBG and SSG
model)

C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190

Main outcome variables
Reported change in GDBG 46 NM 3.2 (2.9) 45 NM 2.8 (3.1)
breathlessness after MLG 48 NM 1.8 (2.4) 0.53 42 NM 1.5 (2.4) 0.48
using the device SSG 45 NM 1.9 (2.9) 0.45 0.03a 43 NM 2.4 (3.2) 0.13 0.1b
(GRC scale)#

Symptom score## (SRGQ) GDBG 48 57.4 (21.8) 52.1 (21.9)* 0.24 45 57.3 (22.3) 56.1 (26) 0.05
MLG 47 56.4 (21.8) 48.4 (21.1)** 0.37 0.13 45 54.9 (24.5) 50.3 (24.5)** 0.19 0.13
SSG 45 50.2 (24.7) 45.3 (26.6)* 0.19 0.02 0.6 42 49.8 (25.4) 46.0 (27.3)* 0.14 0.10 0.5

Secondary outcome variables:

Activity score## (SRGQ) GDBG 47 63.5 (19.3) 63.8 (20.2) 0.02 0.2 46 63.1 (19) 63 (21) 0.01 0.1
MLG 48 64 (18.8) 62.2 (21.3) 0.09 0.10 43 64.3 (18.7) 60.9 (21.9) 0.17 0.15
SSG 46 56.3 (21.6) 52.8 (24.4) 0.15 0.17 42 55.8 (21.4) 50.4 (22.9)** 0.24 0.24

Impact score## (SRGQ) GDBG 48 40.4 (22.7)c 39.6 (23.2) 0.04 46 39.5 (22)c 39.9 (24.1) 0.02
MLG 48 37.9 (20.0) 35.4 (18.3) 0.13 0.08 44 37.1 (19.9) 33.7 (19.5) 0.17 0.17
SSG 45 29.3 (20.5) 28.4 (17.6) 0.05 0.01 0.5 42 28.4 (20.4) 28.7 (17.2) 0.02 0.01 0.2

##
Total score (SRGQ) GDBG 47 49.8 (19.7) 48.5 (20.3) 0.07 46 49.4 (19.3) 49.7 (21.6) 0.00
MLG 47 48.9 (18.3) 46 (17.6)* 0.16 0.08 43 48.4 (18.1) 44.9 (18.7)* 0.19 0.18
SSG 46 40.3 (19.6) 38.3 (19.6)d 0.10 0.04 0.5 41 39.8 (19.7) 37.6 (18.9) 0.11 0.11 0.1

GDBG = guided deep breathing group; GRC = global rating change MLG = music listening group; NM = not measured; RR = respiratory rate; SSG = sitting still group; SGRQ = St George’s Respiratory Questionnaire.
*
Significant difference within group from baseline, p  0.05.
**
Significant difference within group from baseline, p  0.001.
#
Higher score = higher degree of positive change. Possible score from 0–10.
##
Higher score = higher degree of problems. Possible score from 0–100.
a
t-test and ANOVA between groups at T2.
b
t-test and ANOVA between groups at T3.
c
Significant difference at baseline between GDBG and SSG, p < 0.05.
d
Non-parametric test p < 0.05, t-test non significant.

187
188 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190

Table 4
Mean scores for the breathing pattern variables (respiratory rate, RR; time on inspiration, TIN; and time on expiration, TEX). Differences between GDBG, MLG, and SSG were
tested by ANCOVA using baseline as a covariate. Changes in the breathing patterns were measured at one time point.

Respiratory pattern Group n Baseline (T1) After four Cohen between Cohen between p-value T1–T2
obtained using the device mean (SD) weeks (T2) T1 and T2 GDBG and MLG, (whole model)
mean (SD) and GDBG and SSG

RR start of session GDBG 49 16 (5.1) 13.4 (5.6)** 0.48

MLG 48 18.5 (5.4) 18.3 (5.6) 0.04 0.43
SSG 46 16.8 (4.9) 16.6 (4.9) 0.04 0.46 0.00a
RR end of session GDBG 49 11.8 (5.1)# 9 (5)** 0.55
MLG 48 17.9 (5.3) 17.6 (4.9) 0.06 0.50
SSG 46 16.6 (4.9) 17.3 (5) 0.14 0.70 0.00a
TIN start of session GDBG 49 1.5 (0.5) 1.9 (1.1)** 0.50
MLG 48 1.4 (0.8) 1.4 (0.6) 0 0.47
SSG 46 1.4 (0.7) 1.7 (1.3) 0.3 0.08 0.06a
TIN end of session GDBG 49 2.3 (1.4)# 2.9 (1.9)** 0.36
MLG 48 1.3 (0.4) 1.4 (0.6) 0.20 0.40
SSG 46 1.4 (0.4) 1.3 (0.4) 0.25 0.61 0.00a
TEX start of session GDBG 49 2.7 (1.3) 3.4 (1.6)* 0.48
MLG 48 2.2 (1.0) 2.2 (0.9) 0 0.56
SSG 46 2.5 (0.8) 2.5 (1.4) 0 0.47 0.001a
TEX end of session GDBG 49 4.3(2.4)# 5.5 (2.5)** 0.49
MLG 48 2.4 (0.9) 2.3 (1) 0.11 0.74
SSG 46 2.6 (1.2) 2.6 (1.2) 0 0.65 0.00a

GDBG = guided deep breathing group, MLG = music listening group, SSG = sitting still group.
#
Significant difference between groups at baseline, p  0.001.
*
Significant difference within group from baseline, p  0.05.
**
Significant difference within group from baseline p  0.01.
a
Significant positive effect of GDBG.

terms of the symptom score. There could be several explanations
for this result.
Firstly, our results showed that the participants were only able
to synchronize with a mean of 53.3% (recommended value 50–99%)
[29] with a large SD of 23. This demonstrates that some of the
participants probably could not perform the exercise. We may
speculate that this may be due to physical barriers such as
physiological changes in the diaphragmatic muscles resulting in an
abnormal asynchronous thoracic abdominal motion, which may
limit the possibility of deep breathing [26]. Further may this give
difficulties to synchronize the breathing with the instruction given
through the device. Another barrier for synchronizing might be a
lack of understanding [30] of the instruction given from the device.
Secondly, breathlessness is a complex subjective experience
[31], which is extremely difficult to measure in terms of its
characteristics and intensity, thus numerous instruments have
been developed trying to describe this phenomenon [20]. The
SGRQ might not have been the best instrument to capture the
differences between groups in our study. Unidimensional tools
such as the NRS have been shown to be easy to answer and
sensitive to changes, thus they are considered to be more suitable
for use [20]. In our study, we focused explicitly on the performance
of the intervention and we obtained information about how the
participants felt about the changes in their breathlessness during
the study. GRC scales have been shown to obtain different results
where the minimal clinically important difference ranges from
1.3 to 2 points [18]. Our study detected a difference of 1.3–1.4
between the GDBG and the other two groups (MLG and SSG). We
did not determine the minimal clinically important difference in
patients with COPD, but we may conclude that the patients in the
GDBG reported the relief of breathlessness. The global rating
question also detected a moderate effect size [25].
Thirdly, there were also positive changes in the symptom
scores in the groups listening to music and sitting still, which
might have been attributable to a distracting stimulus directing the
attention away from negative symptoms, as described in reviews
[11,12]. Thus, the SSG with a sham device can also be interpreted as
a meditation intervention. Sitting still for 15 min twice per day
may have allowed patients time to become more aware of their
breathing and how to relax. This awareness may also have affected
their responses to symptoms [32]. Biofeedback devices such as
those used in our study are known as mind–body therapies, which
affect the awareness of changing bodily functions and this might
lead to the cognition of subjective change [33]. Deep breathing
reduced breathlessness in the GDBG group but it also stimulated
mucus mobilization and resulted in less coughing according to the
symptom score. Deep breathing combined with ‘‘huffing’’ is used
often in techniques such as active breathing cycles [34]. We found
that the effect sizes between T1 to T2 and T1 to T3 were low to
moderate [25], thereby indicating that more research is needed.
Fourthly, we controlled for attention by calling the patients
once each week, thus empathetic support may have affected their
awareness and changed the symptom scores. Empathetic support
may predict adjustment to chronic illness and it also is associated
with fewer symptoms in COPD [35,36]. The ‘‘meaning of response’’
has been proposed as a mechanism for explaining the placebo
effect that arises in the relationship between patients and health
professionals in RCTs [37]. RCTs are viewed widely as providing the
strongest evidence when assessing the efficacy of treatment and
care interventions, and they are treated as the gold standard for
evaluation in the context of evidence-based medicine. However,
the status of RCTs has been disputed, partly because of problems
related to the establishment of suitable control groups [38].
All the four groups may have been subjected to mind–body
therapies that involved conscious mental processes, which may
have affected bodily functions and possibly the symptom scores.
This means that not only GDBG may be a self-management
procedure, but also listening to music and relaxation may be used
as self-care to give improvement of COPD related symptoms such
as breathlessness in people with moderate and severe stage of
COPD.
4.2. Limitations
There are several limitations to this study. For example, the
use of a global rating change scale that ranged from 5 to 5 rather
 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190
than 0–10 to rate breathlessness might have yielded more
information regarding possible negative changes [18]. Although,
the participants were not informed of the group to which they had
been assigned, they may have suspected having been randomized
to a control group (i.e. MLG, SSG). However, since all the groups
improved significantly on the symptoms score at the first follow-
up, we believe that most of the patients were not aware of their
actual group assignment.
The study nurse checked the questionnaires for missing
responses at T2 and T3. She was also the nurse who handed the
device to the participants at T1, thus she could have influenced the
participants’ responses in the questionnaires. However, the
questionnaires were completed at home by the patients before
they attended their hospital appointments and the information
given to the participants was stored safely. Thus, we consider that
the study nurse could only have had a limited effect on the
responses in the questionnaires. The study nurse also called the
patients each week to investigate compliance by asking questions
about breathing pattern read out from the device. This may have
motivated the users and clarified some questions, but it might also
have affected the results of our study.
Out of the total of 150 patients, 7 (4.7%) failed to follow up at
four weeks and 20 (13%) patients failed to follow up at four
months. This percentage dropping out is less than 15%, which is
recommended as the maximum drop-out rate [39]. We are
therefore rather confident that our results are not due to a
drop-out bias. However, overall our study only had a response rate
of 44% which may raise questions regarding the representativeness
of our sample. We have no information about non-responders, but
it is likely that the reason for not participating in the study might
concern the extra effort it would entail. The significantly higher
lung function in the SSG and the lower number of comorbidities in
the MG at baseline may have marginally affected our result. In the
ANOVAs, those with severe stage of COPD showed slightly greater
improvement on the GRC scale at T3 (p = 0.02, not shown in tables),
but severity yielded no significant interaction terms with groups.
We therefore believe that the slightly uneven distribution of
severity across groups has not biased our results.
4.3. Conclusions
The device-guided breathing appeared to have a positive effect
on the breathing patterns and perceived change of breathlessness
in COPD patients compared with a group listening to music and a
sham control group sitting still. However, all three approaches
produced positive changes in the symptom scores from baseline to
the post intervention assessment.
4.4. Practice implications
Our results indicate that self-care support is effective to some
extent. Deep breathing with a biofeedback device may be used as a
self-management procedure by COPD patients in both at home and
in combination with other non-pharmacological interventions.
However, further RCTs of breathing control exercises should
include comparisons with other non-pharmacological and atten-
tion-based interventions.
Author contributions
CRB, AKW, AMM, TM, and IE were responsible for the
conception and design of the study. CRB, MPL, and UM were
responsible for data collection. CRB, AKW, and TM performed the
analysis. CRB, AKW, AMM, MPL, EO, TM, UM, and IE were
responsible for writing and revising the manuscript. All authors
have approved the final article.
189
Funding
This project was supported financially by the Norwegian Extra
Foundation for Health and Rehabilitation through EXTRA funds, the
Norwegian Nurses’ Organisation (NNO) for the pulmonary nurses in
the NNO, the Norwegian Heart and Lung Patient Organization, the
University of Oslo Department of Health Sciences, and the main
project hospital; Lovisenberg Diaconal Hospital.
Competing interest
None declared.
Acknowledgments
We acknowledge the participants in the study.
References
[1] Bentsen SB, Henriksen AH, Wentzel-Larsen T, Hanestad BR, Wahl AK. What
determines subjective health status in patients with chronic obstructive
pulmonary disease: importance of symptoms in subjective health status of
COPD patients. Health Qual Life Outcomes 2008;6:115.
[2] Tsiligianni I, Kocks J, Tzanakis N, Siafakas N, van der Molen T. Factors that
influence disease-specific quality of life or health status in patients with COPD:
a review and meta-analysis of Pearson correlations. Prim Care Respir J
2011;20:257–68.
[3] GOLD. Global Initiative for Chronic Obstructive Lung Disease, Global
strategy for the diagnosis, management, and prevention of chronic obstructive
pulmonary disease; 2013. Available from: hhttp://www.goldcopd.orgi
[11.01.13].
[4] American Thoracic Society.. Dyspnea. Mechanisms, assessment, and manage-
ment: a consensus statement. Am J Respir Crit Care Med 1999;159:321–40.
[5] Loveridge B, West P, Anthonisen NR, Kryger MH. Breathing patterns in patients
with chronic obstructive pulmonary disease. Am Rev Respir Dis
1984;130:730–3.
[6] Siela D. Ineffective breathing pattern. In: Ackley BJ, Ladwig GB, editors.
Nursing diagnosis handbook: an evidence-based guide to planning care. St.
Louis, MO: Mosby Elsevier; 2013. p. 175–8.
[7] Gosselink R. Controlled breathing and dyspnea in patients with chronic
obstructive pulmonary disease (COPD). J Rehabil Res Dev 2003;40:25–33.
[8] Dechman G, Wilson CR. Evidence underlying breathing retraining in people with
stable chronic obstructive pulmonary disease. Phys Ther 2004;84:1189–97.
[9] 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.
[10] Holland AE, Hill CJ, Jones AY, McDonald CF. Breathing exercises for chronic
obstructive pulmonary disease. Cochrane Database Syst Rev 2012;10:
CD008250.
[11] Norweg A, Collins EG. Evidence for cognitive-behavioral strategies improving
dyspnea and related distress in COPD. Int J Chronic Obstr Pulm Dis 2013;8:
439–451.
[12] Panigrahi A, Sohani S, Amadi C, Joshi A. Role of music in the management of
chronic obstructive pulmonary disease (COPD): a literature review. Technol
Health Care 2014;22:53–61.
[13] Ekman I, Kjellstrom B, Falk K, Norman J, Swedberg K. Impact of device-guided
slow breathing on symptoms of chronic heart failure: a randomized, con-
trolled feasibility study. Eur J Heart Fail 2011;13:1000–5.
[14] Mahtani KR, Nunan D, Heneghan CJ. Device-guided breathing exercises in the
control of human blood pressure: systematic review and meta-analysis. J
Hypertens 2012;30:852–60.
[15] Gavish B. Device-guided breathing in the home setting: technology, perfor-
mance and clinical outcomes. Biol Psychol 2010;84:150–6.
[16] Jones PW, Quirk FH, Baveystock CM, Littlejohns P. A self-complete measure of
health status for chronic airflow limitation. The St. George’s Respiratory
Questionnaire. Am Rev Respir Dis 1992;145:1321–7.
[17] ATS. (American Thoracic Society): St. George’s Respiratory Questionnaire
(SGRQ). hhttp://www.thoracic.org/assemblies/srn/questionaires/sgrq.phpi
[04.26.14].
[18] Kamper SJ, Maher CG, Mackay G. Global rating of change scales: a review of
strengths and weaknesses and considerations for design. J Man Manip Ther
2009;17:163–70.
[19] Gift AG. Validation of a vertical visual analogue scale as a measure of clinical
dyspnea. Rehabil Nurs 1989;14:323–5.
[20] Bausewein C, Farquhar M, Booth S, Gysels M, Higginson IJ. Measurement of
breathlessness in advanced disease: a systematic review. Respir Med 2007;101:
399–410.
[21] Pellegrino R, Viegi G, Brusasco V, Crapo RO, Burgos F, Casaburi R, et al.
Interpretative strategies for lung function tests. Eur Respir J 2005;26:948–68.
190 C.R. Borge et al. / Patient Education and Counseling 98 (2015) 182–190
[22] Hyatt RE, Scanlon PD, Nakamura M. Interpretation of pulmonary function
tests: a practical guide. third ed, Philadelphia, PA: Lippincott Williams &
Wilkins; 2009.
[23] Griffiths TL, Burr ML, Campbell IA, Lewis-Jenkins V, Mullins J, Shiels K, et al.
Results at 1 year of outpatient multidisciplinary pulmonary rehabilitation: a
randomised controlled trial. Lancet 2000;355:362–8.
[24] Jones PW. Interpreting thresholds for a clinically significant change in health
status in asthma and COPD. Eur Respir J 2002;19:398–404.
[25] Cohen J. Statistical power analysis for the behavioral sciences. second ed,
Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
[26] Raupach T, Bahr F, Herrmann P, Luethje L, Heusser K, Hasenfuss G, et al.
Slow breathing reduces sympathoexcitation in COPD. Eur Respir J 2008;32:
387–392.
[27] Katiyar SK, Bihari S. Role of pranayama in rehabilitation of COPD patients—a
randomized controlled study. Indian J Allergy Asthma Immunol 2006;20:
98–104.
[28] Yamaguti WP, Claudino RC, Neto AP, Chammas MC, Gomes AC, Salge JM,
et al. Diaphragmatic breathing training program improves abdominal
motion during natural breathing in patients with chronic obstructive
pulmonary disease: a randomized controlled trial. Arch Phys Med Rehabil
2012;93:571–7.
[29] InterCure Ltd. RespeRate User Manual. hhttp://www.resperate.com/wpi. con-
tent/uploads/2014/04/UserManual-english-orange.pdf. [08.14.14].
[30] Roberts NJ, Younis I, Kidd L, Partidge MR. Barriers to the implmentation of self
management support in long term lung conditions. Lond J Prim Care 2012;13–25.
[31] Hayen A, Herigstad M, Pattinson KT. Understanding dyspnea as a complex
individual experience. Maturitas 2013;76:45–50.
[32] Lansing RW, Gracely RH, Banzett RB. The multiple dimensions of dyspnea:
review and hypotheses. Respir Physiol Neurobiol 2009;167:53–60.
[33] Park C. Mind-body CAM interventions: current status and considerations for
integration into clinical health psychology. J Clin Psychol 2013;69:45–63.
[34] Dean E, Frownfelter D. Cardiovascular and pulmonary physical therapy evi-
dence and practice. fourth ed. St. Louis, MO: Mosby Elsevier; 2006.
[35] Graydon JE, Ross E. Influence of symptoms, lung function, mood, and social
support on level of functioning of patients with COPD. Res Nurs Health
1995;18:525–33.
[36] Ketelaars CA, Huyer Abu-Saad H, Halfens RJ, Schlosser MA, Mostert R, Wouters
EF. Effects of specialized community nursing care in patients with chronic
obstructive pulmonary disease. Heart Lung 1998;27:109–20.
[37] Miller FG, Colloca L, Kaptchuk TJ. The placebo effect: illness and interpersonal
healing. Perspect Biol Med 2009;52:518–39.
[38] Grossman J, Mackenzie FJ. The randomized controlled trial: gold standard, or
merely standard? Perspect Biol Med 2005;48:516–34.
[39] Herbert R, Jamtvedt G, Hagen KB, Mead J. Practical evidence-based physio-
therapy. Second ed, Eidinburgh, London, New York, Oxford, Philadelphia, St
Louis, Sydney, Torontor: Chrichill Livingstone Elsevier; 2014.