S Y S T E M A T I C R E V I E W

The Effect of Inspiratory Muscle Training on Duchenne Muscular Dystrophy:

A Meta-analysis
Elizabeth Williamson, PT, PhD; Natalie Pederson, PT, DPT; Hannah Rawson, PT, DPT; Todd Daniel, PhD
Department of Physical Therapy (Dr Williamson), Missouri State University, Springfield, Missouri; St Luke’s East Outpatient Rehabilitation (Dr Pederson), Lee
Summit, Missouri; Kansas City Orthopedic Institute (Dr Rawson), Kansas City, Missouri; Research by Design, LLC (Dr Daniel), Springfield, Missouri.

Purpose: This study reports the respiratory muscle training effect on strength and endurance in individuals with Duchenne
muscular dystrophy.
Methods: Articles published from 1984 to 2017 were reviewed. Six articles met the inclusion criteria that included
within-subject control or between-subject control group, participants with a diagnosis of only Duchenne muscular dystrophy,
participation in respiratory muscle training intervention, and outcome measures of endurance and strength. Effect sizes
were calculated for each study and overall, weighted mean effect sizes for strength and endurance outcome measures.
Results: There was a large effect for improving respiratory endurance and a moderate effect for muscle strength. However,
these effects were not significant.
Conclusion: Findings justify further exploration of the potential benefits of respiratory muscle training for individuals with
Duchenne muscular dystrophy. (Pediatr Phys Ther 2019;31:323–330)
Key words: Duchenne muscular dystrophy, muscular dystrophy, respiratory muscle training

Duchenne muscular dystrophy (DMD) is a progressive dis- tractility and right ventricular function. For the majority of indi-
order in which initial weakness of large proximal muscles of viduals with DMD, death occurs following a respiratory infec-
the shoulder and hip girdles results in an unsteady, waddling tion in early adulthood.1
gait by age 3 to 5 years, difficulty ascending and descending The estimated prevalence of DMD ranges from 1.7 to 4.2
stairs and rising from the floor by age 5 to 7 years, and loss of individuals per 100 000 in the population.2 DMD is a genetic
the ability to safely ambulate by age 7 to 12 years.1 Eventually, disorder in which deletions on the X chromosome prevents the
deformities of the thoracic spine and ribs and weakness of the production of dystrophin and related proteins, the dystrophin-
diaphragm and accessory respiratory muscles compromise pul- glycoprotein complex (DGC).3 The DGC, a complex of trans-
monary function. Total lung capacity and vital capacity decrease, membrane proteins, is associated with the inner surface of the
and residual volume increases. Effort-dependent components of sarcolemma. Dystrophin and dystrophin-associate proteins con-
respiration are affected such as peak expiration flow and forced tribute to the maintenance of fiber integrity and enable interac-
inspiratory flow.1 Late in the disease progression, individuals tion of the DGC with microtubulars, actin, and the intermediate
develop cardiomyopathy, which may affect left ventricular con- cytoskeleton essential for maintenance and repair.4 Dystrophin
is also expressed in cardiac muscle, the neuromuscular junction,
the brain, and the retina.3,4
0898-5669/110/3104-0323 Without dystrophin, dystrophin-associated proteins are
Pediatric Physical Therapy
reduced, which disrupts the linkage between the inner surface
Copyright © 2019 Academy of Pediatric Physical Therapy of the American
Physical Therapy Association
of the sarcolemma and the extracellular matrix during contrac-
tion and relaxation of the muscle. The cascade that follows
Correspondence: Elizabeth Williamson, PT, PhD, Department of Physical includes mechanical weakening of the sarcolemma, inappro-
Therapy, Missouri State University, 901 S. National Ave, Springfield, MO priate calcium influx, abnormal signaling, increased oxidative
65802 (ewilliamson@missouristate.edu).
stress, recurrent ischemia, and eventually cell death.5 Early in
Natalie Pederson and Hannah Rawson completed the initial work as stu-
the disease, necrotic cells are replaced by proliferation and dif-
dents in the Department of Physical Therapy at Missouri State University,
Springfield, Missouri. ferentiation of satellite cells. However, chronic degeneration and
Supplemental digital content is available for this article. Direct URL citation impaired regeneration of affected myofibers over time reduce
appears in the printed text and is provided in the HTML and PDF versions satellite cell capacity and skeletal myofibers are replaced with
of this article on the journal’s Web site (www.pedpt.com) noncontractile, fibrotic tissue.6,7
The authors declare no conflicts of interest. A similar cascade occurs with cardiac muscle. Cardiac
DOI: 10.1097/PEP.0000000000000648 myocytes show increased fragility to physiological stretch with
associated fibrosis, inflammation, and elevation of serum cardiac

Pediatric Physical Therapy The Effect of Inspiratory Muscle Training on Duchenne Muscular Dystrophy 323

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
troponin I.7 Dystrophin-deficient muscle fibers have reduced tension increased by approximately 30% in these mdx mice.
blood flow due to decreased levels of nitric oxide in vascular Zelikovich et al18 reported that treadmill walking at a rate of
smooth muscle.7 Excessive muscle activity and force lead to 4 m/min (low-intensity) and 8 m/min (moderate-intensity) for
deterioration of skeletal and cardiac muscle over time.6,7 30 minutes, 3 days per week over 6 months, improved respi-
High-intensity resistive training of skeletal muscle is dis- ratory capacity as measured by greater minute volume in mdx
couraged for individuals with DMD and the use of graded mice. Another study found that low-intensity-trained mdx mice
exercise in the treatment of DMD is controversial.8 However, had less necrosis and greater regeneration of the diaphragm than
these conclusions are based on a small number of studies with sedate mdx mice, but other findings indicated little benefit to
small sample sizes. Although high-intensity resistance training the training.19 These mixed findings suggest that the intensity,
detrimentally affects skeletal muscles in individuals with a dosage, and duration of exercise may be important factors.
neuromuscular disease, moderate exercise improves exercise Although the above studies provided some evidence that
capacity and reduces muscle fatigue in both dystrophin- aerobic exercise of a specific intensity, dosage, and duration
deficient (mdx) mice and in individuals with DMD.7 This may improve diaphragm function, these studies do not specifi-
study specifically examined the evidence for resistive training cally address the potential benefit or harm of respiratory muscle
of respiratory muscles. training. Only one animal study was found that addressed this
Expansion of the lungs depends primarily upon the specific question. Krupnick et al20 banded the trachea of mdx
rhythmic and repeated contraction of the diaphragm against mice and reduced the tracheal cross-sectional area by 80%,
elastic and resistive forces. The force of contraction of diaphrag- thus increasing respiratory workload. Compared with mdx mice
matic muscle fibers is dictated by 3 factors: the force-velocity without the tracheal banding, there was no significant difference
of the contraction, the initial length of the muscle, and the in tetanic force generated or fibrosis formation of the diaphragm
rate of muscle contractions.9 DMD affects each of these fac- and the accessory muscles of respiration over a 5-month period.
tors directly or indirectly. The weakening of shoulder girdle There was greater myofiber regeneration among tracheal banded
musculature encourages a slump sitting posture with asymmet- mdx mice, suggesting adaptation to the increased workload.20
rical rotation of the thoracic spine, which shortens the initial Excessive exercise has been used to exacerbate myofiber
length of the muscle fibers, thus reducing the force generated damage in mdx mice when assessing the effectiveness of various
during each breath.1 The muscle composition of the diaphragm medical interventions, but few studies have examined the con-
includes slow twitch oxidative (∼55% type I), fast twitch, oxida- sequences of moderate resistive exercise.21,22 Mild to moderate
tive glycolytic type (∼25% type IIa), and fast twitch, glycolytic exercise leads to increased regeneration of myofibers in many
type (∼20% type IIb) muscle fibers.9 Muscle destruction due of the animal studies we reviewed.17-20 Less clear is whether
to DMD primarily targets type IIb muscle fibers, with a gradual this regeneration is an adaptive process (delaying respiratory
shift to a greater predominance of type I fibers.10 Type IIb fibers compromise) or a pathological process (accelerating respiratory
generate the greatest force-velocity; therefore, progressive loss compromise).
of these fibers reduces the ability of the diaphragm to expand Improvements in respiratory muscle strength and
the chest wall as necessary for effective inspiration and reduces endurance have been noted in individuals with DMD. Unfortu-
the recoil of the chest wall necessary for effective expiration.10 nately, many of these studies provided no statistical analyses23,24
Diminished air exchange due to decreased force-velocity genera- or lacked a control group.23-25 Controlled studies used differing
tion is initially compensated by increasing the rate of respiration, training durations, different measurements, used small sample
but this compensatory mechanism is limited and eventually the sizes, and reported inconsistent strength findings, thereby
act of breathing approaches fatiguing threshold. When inspira- preventing generalization of findings.13,26-30
tory muscle strength falls below 30% of normal, hypercapnic Meta-analysis is ideal for addressing problems with low
respiratory failure is likely to occur.9 power and varying methodologies because it allows for the anal-
Respiratory muscle training improves respiratory function ysis of data from numerous studies. This process can better
in healthy adults and individuals with respiratory compromise define population effects of mean and variance by summarizing
secondary to cerebrovascular accident.11,12 Over the last 60 findings, estimating the variability, and seeking potential moder-
years, several studies examined respiratory muscle training in ators that explain variability across studies.31 The purpose of our
individuals with DMD. However, based on review of their own study was to gain a better understanding of potential improve-
work and those of 2 other researchers, Smith et al13 concluded ments in strength and endurance of respiratory muscle training
that the benefits to training “would be small” and inspira- in individuals with DMD through a meta-analysis of controlled
tory resistance training “potentially hazardous” since capacities studies. Respiratory compromise is the result of weakened inspi-
may be close to fatiguing thresholds.13 Stedman et al14 found ratory, upper airway and expiratory muscles. Inspiration is pri-
a similar pattern of progressive degeneration, weakness, and marily dependent on the contraction of the diaphragm, external
fibrosis of the diaphragm compared with limb musculature in intercostals, and the scalene muscles. During excessive exer-
an mdx mouse over time. Additional studies noted that after tion or with weakening of primary inspiratory muscles, the
completing voluntary exercises, such as wheel running or swim- sternocleidomastoid and trapezius may be recruited. Expira-
ming, the diaphragms of mdx mice generated less force and had tion is usually a passive maneuver. Elastic recoil of the lungs
increased areas of fibrosis formation.15,16 However, Dupont- pushes air out. During excessive exertion or coughing, the
Versteegden et al17 reported that daily, voluntary running activ- rectus abdominus, transverse abdominus, internal intercostals,
ities did not alter fatigability of diaphragmatic fibers and active and internal and external obliques may be recruited. Upper

324 Williamson et al Pediatric Physical Therapy

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
airway musculature helps to expand and hold open the airway. surgery, effectiveness of pharmacological interventions, or treat-
Strengthening of inspiratory muscles increases the volume of ment of other neuromuscular diseases. The method sections
air flowing into the lungs and strengthening of the expiratory of 10 relevant articles were reviewed using the following cri-
muscles increases the velocity of air flowing out of the lungs. teria: either within-subject control or a between-subject con-
Training may improve musculature strength and endurance. trol group, participants with a confirmed diagnosis of DMD
Strengthening of inspiratory musculature, directly or indirect only, and no participants with other neuromuscular diseases,
through training of expiratory musculature, may delay the need participation in a resistive respiratory muscle training inter-
for assisted ventilation due to hypoventilation.32 vention, and outcome measures of respiratory endurance and
Weakness of both inspiratory and expiratory muscles strength (Figure 1). We reviewed references from acquired arti-
reduces vital capacity; however, other factors may contribute to cles, seeking other articles not identified during the initial
reduce vital capacity. Vital capacity is a nonspecific and insen- searches. One article was identified but upon review did not
sitive measurement of respiratory muscle function.33 There- meet the criteria. In order to identify unpublished research,
fore, our review of resistive respiratory muscle training focused a Google Scholar search of unpublished dissertations using
on outcome measurements, which specifically measured res- the same keywords as the search for published research was
piratory muscle strength such as maximal static inspiratory completed and known researchers of DMD were contacted.
pressure (Pimax ) and transdiaphragmatic pressure (Pdimax ) and No unpublished research was found. Prior to submission of
endurance such as time limit (Tlim ), cumulative time of contin- the manuscript, the first author completed a second electronic
uous breathing cycles (TE), and time until maximal pressure fell search using the same keywords, databases, and procedure for
below 90% of Pimax . These noninvasive/volitional tests are con- articles published between January 2015 and December 2017.
sidered reliable among cooperative children 7 years or older.34 No additional studies were found (Figure 2). Six studies met
Therefore, our study reviewed studies in which individuals the inclusion criteria; 4 had sufficient data for meta-analysis of
with DMD completed respiratory muscle training. muscle strength, and 3 had sufficient data for meta-analysis of
endurance (Table 1). Supplemental Digital Content 1 (available
at: http://links.lww.com/PPT/A270) lists 10 relevant articles that
METHODS did not meet the inclusion criteria or lacked the data needed to
Literature Search complete a meta-analysis.
Using the keywords Duchenne’s muscular dystrophy, mus-
cular dystrophy, or DMD and respiratory, respiratory exercise,
or inspiratory muscle training, we sought published research Statistical Analysis
completed from January 1984 to December 2014 in the fol- For each study, effect sizes were calculated for strength and
lowing databases: MEDLINE, CINAHL, Cochrane, SPORT- endurance outcome measures comparing the difference between
Discus, Health Technology Assessments, NHS Economic Eval- the control group or period (if participants were acting as their
uation, and Abstracts Review of Effects. Upon identifying 1189 own control) and the intervention. Since the standard deviation
published articles, we narrowed the electronic search by limiting between control and treatment findings varied greatly in the 2
it to published articles written in English, included human par- between-group studies, effect sizes of strength and endurance
ticipants, had an available abstract, and were randomized con- outcomes for each study were calculated using Glass’s  in
trolled trials. which the standard deviation from the control group was used in
Two reviewers independently reviewed the remaining 564 place of the pooled standard deviation of the control and treat-
articles. A third reviewer acted as the arbitrator when disagree- ment findings. Use of Glass’s  is predicated on the assumption
ments between the 2 independent reviews occurred. There were that control findings represent the variance of the population
no disagreements between reviewers. Initially, the reviewers better than the pooled standard deviation.35 Overall weighted
reviewed the article abstract for relevance to respiratory muscle mean effect sizes for strength and endurance were calculated
training. The majority of articles examined respiratory care, using the Hunter-Schmidt method.31 This method assumes that
assessment of respiratory function, effectiveness of scoliosis each study is sampling from different populations.16 In order to

Fig. 1. Articles reviewed from January 1984 to December 2014 and January 2015 to December 2017.

Pediatric Physical Therapy The Effect of Inspiratory Muscle Training on Duchenne Muscular Dystrophy 325

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
 Fig. 2. Forest plot of effect sizes and standard deviations of strength findings.

correct for the possible heterogeneity of the samples, the effect consistent over 3 studies,26,28,30 the results in the meta-analysis
size estimates for each study were weighted by the sample size of were also nonsignificant. The nonsignificant results in this meta-
the particular study.31 An overall mean effect size for each out- analysis, other human and mdx mice studies, suggest the need
come measure for all studies theoretically reduces sample error for further research using a common protocol.
and creates estimates closer to the actual population. The use of Inconsistent findings between studies may be explained by
the broader calculation of the Hunter-Schmidt method allows differences in exercise method, intensity, and timing of the inter-
for generalizations of the findings beyond the studies included ventions. Topin et al26 used a standardized method of controlled
in the meta-analysis.31 We also calculated 95% confidence inter- breathing developed by Matecki et al37 to quantify respiratory
vals for each effect size as noncentrality parameters.36 muscle endurance while the other studies in the meta-analysis
allowed for modification of tidal volume, and inspiratory and
expiratory time during training. Topin et al26 demonstrated the
RESULTS greatest effect with a Glass  of 3.88 compared with 1.4430
Table 2 has respiratory muscle strength findings and and 0.64.28 Maximum voluntary ventilation (a measurement of
Figure 2 is a forest plot of means and confidence intervals for muscle strength) and duration of progressive isocapnic hyper-
all studies. Respiratory muscle training had a moderate effect on ventilation (a measurement of muscle endurance) improved
respiratory muscle strength (k = 4, n = 73, overall d = 0.79). for 15 individuals with DMD who completed a graded respi-
Table 2 has respiratory muscle endurance findings and Figure ratory exercise program using a series of video games over 5
3 has a forest plot graph of means and confidence intervals of weeks. Difficulty levels increased automatically when individ-
all studies. Respiratory muscle training had a large effect on res- uals achieved 80% or better success rate at current levels.25
piratory muscle endurance (k = 4, n = 52, overall d = 1.85). Winkler et al38 found a positive correlation between the inten-
Neither mean effect size was statistically significant. Although sity of training as measured by number of correctly performed
there was a positive moderate effect for strength and a positive exercises against resistance over a 9-month period and improve-
large effect for endurance, standard deviations for both strength ment in maximal inspiratory pressure among individuals with
and endurance were positive and negative indicating major dif- DMD. Similar findings were noted with aerobic training of
ferences between participants within individual studies. mdx mice with greater inspiratory function as measured by
minute volume (MVb) walking on a treadmill at 8 m/min (44%
increase in MVb) versus 4 m/min (13% increase in MVb) or
DISCUSSION 0 m/min (13% decrease in MVb).18 Wanke et al28 found that
Respiratory muscle training had a moderate effect on res- improvements in both strength and endurance occurred only in
piratory muscle strength and had a large effect on respiratory participants with vital capacity values equal or greater than 25%
muscle endurance. The moderate effect on respiratory muscle of predicted and/or partial pressure CO2 levels equal to or less
strength may be misleading since only one study28 had a sig- than 45 mm Hg. Winkler et al38 noted a similar finding among
nificant finding and the results in the meta-analysis were not participants with DMD or spinal muscle atrophy, with positive
significant. Although the improvement in respiratory muscle result only among individuals in which vital capacity declined
endurance appears more robust since the large effect size was less than 10% in the preceding year. Thus, future studies that

326 Williamson et al Pediatric Physical Therapy

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
 TABLE 1
Summary of Characteristics of Each Study

Measurements

Article Age Mean (Range) Sample Size Training Protocol Strength Endurance Study Duration

Topin et al26,a Rx: 14.7 (10.2-19.2) 16 Rx: Imposed inspiratory load of Pimax Tlim a Training daily for 6 wk
35% of Pimax
Ctrl: 12.6 (10.8-14.4) Rx: 8 Ctrl: Imposed inspiratory load of Measurements prior to
Ctrl: 8 5% of Pimax study and after
training
Wanke et al28 Rx: 13.5 (10-24) 30 Rx: 20 resistive breathing Pdimax a Pdimax × Tξ a Training 2 times a day
Ctrl: 14.5 (9-20) Rx: 15b maneuvers against variable for 6 mo
Ctrl: 15b resistance and 10 maximal Measurements 3 mo
static inspiratory efforts! prior to training,
Ctrl: No intervention before study, after
study, after 6 mo
Rodillo et al27 11.6 (9-14) 20c 9 individuals trained with Triflow Pimax None Training daily for 6 wk
II inspirometer, switched to Measurements prior to
mini-peak flow meter after study, after 18 d, and
3 wk after 36 d
11 individuals trained with
mini-peak meter and switched
to Triflow II after 3 wk
Martin et al30 14.2 (7-20) 18d 9 individuals performed Pimax Pe time Training 5 d/wk
maximum static inspiratory Pi timea Measurements prior to
and expiratory maneuvers and study, 2 wk, 4 mo,
ventilated to exhaustion 3 and 6 mo
times against resistance for 2
mo, rested for 2 mo, and
trained for 2 mo
9 individuals rested for 4 mo and
performed maximum static
inspiratory and expiratory
maneuvers and ventilated to
exhaustion 3 times against
resistance for the last 2 mo

Abbreviations: Ctrl, control group; Pdi, transdiaphragmatic pressure; Pimax , maximal inspiratory pressure determined at functional residual capacity; Rx,
treatment group; Tlim , time limit, which represents the maximal time a participant sustained breathing at functional residual capacity against a predetermined
load without fatigue.
a Indicative of significant findings at P < .05.
b Five training patients discontinued after 1 month due to lack of progress; all had vital capacity values less than 25% of predicted and/or a PaCO of more than
2
45 mm Hg. Two of the control were dropped too prior to the end of the study due to the development of acute respiratory failure.
c Twenty-two individuals began the study, but 2 withdrew due to illness.
d One patient died during the training period. Two participants did not complete the maneuvers necessary for the initial P
imax measurement (strength) and 3
participants did not complete the maneuvers necessary for the initial Pi time measurement (endurance).

TABLE 2
Means, Standard Deviations, P Value for Original Findings, and Glass’s  for All Findings

Respiratory Muscle Strengtha Respiratory Muscle Enduranceb

Mean Standard Mean Standard

Study N Difference Deviationc P Value Glass’s  Difference Deviationc P Value Glass’s 

Martin et al30,d 15 0.625 2.91 NS 0.21 8.53 5.92 .01 1.44

Rodillo et al27,b 11 1 29 NS 0.03
Rodillo et al27,b 9 −13 23 .02 − 0.57
Topin et al26 16 −1.3 20.7 NS − 0.06 176.7 45.6 .05 3.88
Wanke et al28,a 22 2.49 1.11 .001 2.24 3.4 5.3 .001 0.64

Abbreviation: NS, nonsignificant.

a Five training participants discontinued after 1 month due to lack of progress.
b Twenty-two participants began the study but 2 withdrew due to illness.
c Standard deviation of control condition. Significant P ≤ .05.
d One participant died during the training period. These participants are not included in the analyses.

Pediatric Physical Therapy The Effect of Inspiratory Muscle Training on Duchenne Muscular Dystrophy 327

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
 Fig. 3. Forest plot of effect sizes and standard deviations of endurance findings.

investigate the potential benefits of respiratory muscle training individual to gain strength and endurance without damaging
for individuals with DMD should use a standardized method of myofibers. Other potential medical interventions, which may
controlled breathing,27 define exercise parameters based on the be enhanced through concurrent respiratory muscle training,
individual’s current capabilities, and adjust as necessary these include sildenafil42 and various gene therapies.43 Therefore,
interventions as strength and endurance improves.18,25,38 Res- future studies should also assess the interaction between med-
piratory muscle training should be initiated early in the disease ical interventions and resistive respiratory muscle training.
progression.28,38 Major limitations to this meta-analysis are the limited
Sustainability of possible improvements over time should be number of available studies, overall sample size, and failure
addressed in future studies. In the Wanke et al28 study, assess- to identify possible unpublished data. The prevalence of DMD
ments were completed 3 months prior to training, immedi- is low (1.7 to 4.2 per 1000) in comparison to other progres-
ately prior to, and immediately following 3 months of training, sive neurological diseases such as multiple sclerosis (67.8 per
and 3 and 6 months after the cessation of training. Improve- 1000). Thus, a small population may prevent the identification
ments were retained 6 months after cessation of training. A of an adequate sample size in a particular geographical loca-
2-year study by Koessler et al39 also found that respiratory tion necessary for commonly used statistical tests. Small sample
endurance and strength improved with sustained inspiratory sizes (generally <30 participants) lack the power to confirm or
muscle training over a 2-year-period among 18 participants with reject the null hypothesis using a Mann-Whitney test, t test,
DMD and 9 participants with spinal muscular atrophy. Because or Wilcoxon Rank Sum test.44 The overall sample size for the
there is concern that respiratory muscle training may accelerate strength variable was 73 participants, but standard deviations
the approach to respiratory fatigue13 and the need for mechan- were large for 2 of the 4 studies.26,27 Overall sample size for the
ical ventilation, participants’ inspiratory muscle fatigue should endurance variable was 53 participants, with a large standard
be monitored using the tension time index or a similar nonin- deviation for 1 study.26 We attempted to identify nonpublished
vasive test and creatine kinase levels should be monitored over data through personal correspondence and a Google search of
time.34,38,39 Future studies should be longitudinal, examining dissertations. Registration of the study prior to beginning the
the sustainability of improvements over 12 to 24 months and analysis at Preferred Reporting Items for Systematic Reviews and
include measures of quality of life and longevity. Meta-Analyses (PRISMA) would have perhaps identified unpub-
There is evidence that several pharmacological treatments lished data. However, this was only considered after completion
may improve pulmonary function in individuals with DMD. of the meta-analysis. Statistical methods that estimate publica-
Corticosteroids have proven to increase patient survival by tion bias, such as a funnel plot, are better suited for larger meta-
5 to 15 years and to improve pulmonary function as mea- analyses.31 With such a limited number of studies, the number
sured by forced vital capacity and possibly maximal expira- of data points is too few to reasonably assess a shape to the plot,
tory pressure.40 Similarly, idebenone is a potent antioxidant, so we had no means to estimate statistical bias. It is possible that
which has reduced the loss of respiratory function among these findings are distorted by a publication bias.
individuals with DMD in a phase 3 trial.41 The reduction of A single case research design (ABAB) would allow for greater
inflammation through the use of corticosteroids34 or oxida- power with fewer participants since the participants would act
tive damage through the use of Idebenone3 may mitigate as their own control. This design also allows for normal distribu-
potential myofiber damage related to exercise and allow an tion and homogeneity of measurement, which provides strong

328 Williamson et al Pediatric Physical Therapy

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
internal validity. Using a standard protocol, data could be col- 17. Dupont-Versteegden EE, McCarter RJ, Katz MS. Voluntary exercise
lected at multiple sites across the globe. Individual participant’s decreases progression of muscular dystrophy in diaphragm of mdx
mice. J Appl Physiol. 1994;77(4):1736-1741.
findings could be synthesized through meta-analyses using
18. Zelikovich AS, Quattrocelli M, Salamone IM, Kuntz NL, McNally EM.
regression-based effect sizes to explore the external validity of Moderate exercise improves function and increases adiponectin in the
respiratory muscle training for individuals with DMD.45 mdx mouse model of muscular dystrophy. Sci Rep. 2019;9(1):5770. doi:
These findings are limited, further research is necessary 10.1038/s41598-019-42203-z.
prior to advocating respiratory muscle training for individuals 19. Morici G, Frinchi M, Pitruzzellla A, et al. Mild aerobic exercise
training hardly affects the diaphragm of mdx mice. J Cell Physiol.
with DMD. A large effect was noted for improving respiratory
2017;232(8):2044-2052. doi:10.1002/jcp25573.
endurance and a moderate effect was noted for improving res- 20. Krupnick AS, Jianliang Z, Nguyen T, et al. Inspiratory loading
piratory muscle strength. The results of these studies may justify does not accelerate dystrophy in mdx mouse diaphragm: implica-
revisiting the question of respiratory muscle training as a ther- tions for regenerative therapy. J Appl Physiol. 2003;94(2):411-419.
apeutic intervention for individuals with DMD, particularly in doi:10.1152/applphysiol.00689.2002.
21. Radley-Crabb H, Terrill J, Shavlakadze T, Tonkin J, Arthur P, Grounds
combination with medical interventions.
M. A single 30 min treadmill exercise session is suitable for ‘proof-of
concept studies’ in adult mdx mice: a comparison of the early con-
sequences of two different treadmill protocols. Neuromuscul Disord.
REFERENCES 2012;22(2):170-182. doi:10.1016/j.nmd.2011.07.008
22. Markert CD, Ambrosio F, Call JA, Grange RW. Exercise and Duchenne
1. Gozal D. Pulmonary manifestations of neuromuscular disease with spe- muscular dystrophy: toward evidence-based exercise prescription.
cial reference to Duchenne muscular dystrophy and spinal muscular Muscle Nerve. 2011;43(4):464-478. doi:10.1002/mus.21987.
atrophy. Pediatr Pulmo. 2000;29:141-150. 23. Adams MA, Chandler LS. Effects of physical therapy program
2. Theadom A, Rodrigues M, Roxburgh R, et al. Prevalence of mus- on vital capacity of patients with muscular dystrophy. Phys Ther.
cular dystrophies: a systematic literature review. Neuroepidemiology. 1974;54(5):494-496.
2014;43(3/4):259-268. doi:10.1159/000369343. 24. Hoberman M. Physical medicine and rehabilitation; its value and
3. Venugopal V, Pavlakis S. Duchenne Muscular Dystrophy. StatsPearls limitations in progressive muscular dystrophy. Am J Phys Med.
[Internet] 2018. https://www-ncbi-nlm-nih-gov.proxy.missouristate. 1955;34(1):109-115.
edu/books/NBK482346. 25. Vilozni D, Bar-Yishay E, Gur I, Shapira Y, Meyer S, Godfrey S. Comput-
4. Gawor M, Proszynski TJ. The molecular cross talk of the dystrophin- erized respiratory muscle training in children with Duchenne muscular
glycoprotein. Ann N Y Acad Sci. 2017;1412:62-72. doi:10.1111/ dystrophy. Neuromuscular Disord. 1994;4(3):249-255.
nyas.13500. 26. Topin N, Matecki S, Le Bris S, et al. Dose-dependent effect of individu-
5. Glanzman AM, Kusler A, Stuberg K. Muscular dystrophies and spinal alized respiratory muscle training in children with Duchenne muscular
muscular atrophy In: Palisano RJ, Orlin MN, Schreiber J, eds. Campbell’s dystrophy. Neuromuscular Disord. 2002:12(6);576-583.
Physical Therapy for Children. 5th ed. St Louis, MO: Elsevier; 2017. 27. Rodillo E, Nomble-Jamieson CM, Aber V, Heckmatt JZ, Muntoni F,
6. Serrano AL, Muñoz-Ćanoves P. Fibrosis development in early- Dubowitz V. Respiratory muscle training in Duchenne muscular dys-
onset muscular dystrophies: mechanisms and translational impli- trophy. Arch Dis Child. 1994:64;736-738. doi:10.1136/adc.64.5.736.
cations. Semin Cell Dev Biol. 2017;64:181-190. doi:10.1016/ 28. Wanke T, Toifl K, Merkle M, Formanek D, Lahrmann H, Swick H. Inspi-
j.semcdb.2016.013. ratory muscle training in patients with Duchenne muscular dystrophy.
7. Barnabei MS, Martindale JM, Townsend D, Metzger JM. Exercise and Chest. 1994:105(2):475-482.
muscular dystrophy: implications and analysis of effects on muscu- 29. Stern LM, Martin AJ, Jones N, Garrett R, Yeates J. Training inspiratory
loskeletal and cardiovascular systems. Compr Physiol. 2011;1(3):1353- resistance in Duchenne dystrophy using adapted computer games. Dev
1363. doi:10.1oos/cphy.c100062. Med Child Neurol. 1989:31(4);494-500.
8. Ansved T. Muscle training in muscular dystrophies. Acta Physiol Scand. 30. Martin AJ, Stern L, Yeates J, Lepp D, Little J. Respiratory muscle training
2001;171:359-366. in Duchenne muscular dystrophy. Dev Med Child Neuro. 1986;28:314-
9. Rochester DF. The diaphragm: contractile properties and fatigue. J Clin 318.
Invest. 1985;75(5):1397-1402. 31. Field AP, Gillett R. How to do a meta-analysis. Br J Math Stat Psychol.
10. Webster C, Silberstein L, Hays AP, Blau HM. Fast muscle fibers 2010;63(pt 3):665-694.
are preferentially affected in Duchenne muscular dystrophy. Cell. 32. Haas CF, Loik PS, Gay SE. Airway clearance applications in the elderly
1988;52(4):503-513. and in patients with neurologic or neuromuscular compromise. Respir
11. Illi SK, Held U, Frank I, Spengler CM. Effect of respiratory muscle Care. 2007;52:1362-1381.
training on exercise performance in healthy individuals a systematic 33. Syabbalo N. Assessment of respiratory muscle function and strength.
review and meta-analysis. Sports Med. 2012;42(8):707-724. Postgrad Med J. 1998;74:208-215.
12. Pollock RD, Rafferty GF, Moxham J, Kalra L. Respiratory muscle 34. LoMauro A, D’Angelo MG, Aliverti A. Assessment and management of
strength and training in stroke and neurology: a systematic review. Int respiratory function in patients with Duchenne muscular dystrophy:
J Stroke. 2013;8(2):124-130. doi:10.1111/j.1747-4949.2012.00811.x. current and emerging options. Ther Clin Risk Manage. 2015;11:1475-
13. Smith PE, Coakley JH, Edwards RH. Respiratory muscle training in 1488.
Duchenne muscular dystrophy. Muscle Nerve. 1988;11(7):784-785. 35. Glass GV. Primary, secondary and meta-analysis of research. Educational
14. Stedman HH, Sweeney HL, Shrager JB, et al. The mdx mouse diaphragm Res. 1976;5(10):3-8.
reproduces the degenerative changes of Duchene muscular dystrophy. 36. Howell DC. Confidence Intervals of Effect Size—University of Ver-
Nature. 1991;353(6335):536-539. mont 2010. https://www.uvm.edu/∼dhowell/methods/Supplements/
15. Selby JT, Acosta P, Sleeper MM, Barton ER, Sweeney HL. Long-term Confidence%20Intervals%20on%20Effect%20Size.pdf. Accessed Sep-
wheel running compromises diaphragm function but improves car- tember 3, 2018.
diac and plantarflexor function in the mdx mouse. Appl Physio (1985). 37. Matecki S, Topin N, Hayot M, et al. A standardized method for the eval-
2013;115(5):660-666. doi:10.1152/japplphysiol.00252.2013. uation of respiratory muscle endurance in patients with Duchenne mus-
16. Barbin IC, Pereira JA, Bersan Rovere M, et al. Diaphragm degenera- cular dystrophy. Neuromuscular Disord. 2001;11(2):171-177.
tion and cardiac structure in mdx mouse: potential clinical implica- 38. Winkler G, Zifko U, Nader A, et al. Dose-dependent effects of inspi-
tions for Duchenne muscular dystrophy. J Anat. 2016;228(5):784-791. ratory muscle training in neuromuscular disorders. Muscle Nerve.
doi:10.1111/joa.12443. 2000;23(8):1257-1260.

Pediatric Physical Therapy The Effect of Inspiratory Muscle Training on Duchenne Muscular Dystrophy 329

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.
39. Koessler W, Wanke T, Winkler G, et al. 2 years’ experience with inspi- 42. Percival JM, Whitehead NP, Adams ME, Adamo CM, Beavo JA, Froehner
ratory muscle training in patients with neuromuscular disorders. Chest. SC. Sildenafil reduces respiratory muscle weakness and fibrosis in
2001;120(3):765-769. the mdx mouse model of Duchenne muscular dystrophy. J Pathol.
40. Gloss D, Moxley RT III, Ashwal S, Oskoui M. Practice guideline update 2012;228(1):77-87. doi:101002/path.4054.
summary: corticosteroid treatment of Duchenne muscular dystrophy: 43. Kostek MC, Gordon B. Exercise is an adjuvant to contempo-
Report of the Guideline Development Subcommittee of the American rary dystrophy treatments. Exerc Sport Sci Rev. 2018;46:34-41.
Academy of Neurology. Neurol. 2016;86(5):465-472. doi:10.1249/JES.0000000000000131.
41. Butse GM, Voit T, Schara U, et al. Efficacy of idebenone on respi- 44. Portney LG, Watkins MP. Foundations of Clinical Research Applications to
ratory function in patients with Duchenne muscular dystrophy not Practice. 3rd ed. Upper Saddle River, NJ: Pearson Education; 2009.
using glucocorticoids (DELOS): a double-blind randomized placebo- 45. Moeyaert M, Ugille M, Ferron JM, et al. Estimating intervention effects
controlled phase 3 trial. Lancet. 2015;385(9979):1748-1757. doi:10. across different types of single-subject experimental designs: empirical
1016/S0140-6736(15)60025-3. illustration. Sch Psychol Q. 2015;30(1):50-63. doi:10.1037/spq000068.

CLINICAL BOTTOM LINE

Commentary on “The Effect of Inspiratory Muscle Training on Duchenne Muscular

Dystrophy: A Meta-analysis”

“How could I apply this information?”

Children with muscular dystrophy have progressive muscle weakness of striated and respiratory muscles. The appli-
cation of the inspiratory muscle training (IMT) methods and the results of this meta-analysis are limited. Little infor-
mation is provided to the reader about the specific method of IMT, which is important for a clinician interested in
providing IMT for a person with muscular dystrophy. For example, 3 of the included studies use IMT methods with
more laboratory-based equipment (despite 2 of the studies being home-based) and the fourth study used an incen-
tive spirometer (with less potential for IMT). One of the studies provided both inspiratory and expiratory muscle
training. These important clinical factors need further discussion to facilitate the application of the information in this
meta-analysis to clinical practice, as well as, with interpreting the results.
“What should I be mindful about when applying this information?”
Generalization of the results of this meta-analysis should be applied cautiously due to the above comments as well
as (1) substantial heterogeneity in the 4 included studies, which, although not reported, can be seen in Figures 3 and
4, (2) no reporting on the quality of the included studies using the PEDro or other scale, (3) use of less common meta-
analytic methods, and (4) a modest number of relevant studies identified via PubMed of “IMT and breathing exercises
in patients with muscular dystrophy” (n = 23 and 47 publications, respectively) some of which could have been
included in this article as a “systemic review and meta-analysis” providing more useful information for the interested
clinician. Although briefly discussed in this meta-analysis, the possibility of respiratory muscle fatigue and subsequent
respiratory failure is concerning in this population. Further discussion on clinical signs and symptoms of respiratory
muscle fatigue and failure, as well as IMT methods to prevent these problems, is needed.

James G. Moore, PT, PhD

Pediatric Residency Program, Nicklaus Children’s Hospital, University of Miami
Miami, Florida
Lawrence P. Cahalin, PT, PhD
Department of Physical Therapy, University of Miami
Coral Gables, Florida
The authors declare no conflicts of interest.
DOI: 10.1097/PEP.0000000000000641

330 Williamson et al Pediatric Physical Therapy

Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association.
Unauthorized reproduction of this article is prohibited.