Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p.

Journal of Back and Musculoskeletal Rehabilitation -1 (2016) 1–10 1

DOI 10.3233/BMR-160695
IOS Press

Respiratory muscle endurance training

reduces chronic neck pain: A pilot study
B. Wirtha,c,∗ , T. Duarte Ferreirab , M. Mittelholzerb , B.K. Humphreysc and U. Boutellierb
a
Motor Control and Learning, Institute for Human Movement Sciences and Sports, ETH Zurich, Zurich,
Switzerland
b
Exercise Physiology Lab, Institute for Human Movement Sciences and Sports, ETH Zurich, Zurich, Switzerland
c
Department of Chiropractic Medicine, University Hospital Balgrist, Zurich, Switzerland

o n
si
Abstract.

er
BACKGROUND: Patients with chronic neck pain show also respiratory dysfunctions.
OBJECTIVE: To investigate the effects of respiratory muscle endurance training (RMET) on chronic neck pain.
fv
METHODS: In this pilot study (single-subject design: 3 baseline measurements, 4 measurements during RMET), 15 neck pa-
tients (49.3 ± 13.7 years; 13 females) conducted 20 sessions of home-based RMET using a SpiroTiger R
(normocapnic hyperp-
oo

noea). Maximal voluntary ventilation (MVV), maximal inspiratory (Pimax ) and expiratory (Pemax ) pressure were measured be-
fore and after RMET. Neck flexor endurance, cervical and thoracic mobility, forward head posture, chest wall expansion and self-
pr

assessed neck disability [Neck Disability Index (NDI), Bournemouth questionnaire] were weekly assessed. Repeated measure
ANOVA (Bonferroni correction) compared the first and last baseline and the last measurement after RMET.
RESULTS: RMET significantly increased MVV (p = 0.025), Pimax (p = 0.001) and Pemax (p < 0.001). During RMET, neck
ed

disability significantly decreased (NDI: p = 0.001; Bournemouth questionnaire: p = 0.002), while neck flexor endurance (p <
0.001) and chest wall expansion (p < 0.001) increased. The changes in respiratory and musculoskeletal parameters did not
ct

correlate.
CONCLUSIONS: RMET emerged from this pilot study as a feasible and effective therapy for reducing disability in patients
rre

with chronic neck pain. The underlying mechanisms, including blood gas analyses, need further investigation in a randomized
controlled study.
co

Keywords: Neck pain, respiratory dysfunction, respiratory muscle endurance training

un

1 1. Introduction mechanics due to the reduced stability of the cervi- 10

cal and thoracic spine or from alterations in the force- 11

2 Neck pain is one of the leading causes of years lived length curve of the sternocleidomastoids [5]. 12

3 with disability worldwide [1]. So far it has been re- Therefore, respiratory muscle endurance training 13

4 garded as a musculoskeletal condition. Recently, there (RMET) might be an interesting therapy in neck pain 14

5 has been growing evidence that pulmonary function is patients because it improves Pimax and Pemax in pa- 15

6 also affected in these patients [2–5]. The observed re- tients with spinal cord injury, myasthenia gravis or 16

7 duction in maximal voluntary ventilation (MVV), and chronic obstructive pulmonary disease [6–8], but not 17

8 maximal inspiratory and expiratory pressure (Pimax in healthy subjects [9], while RMET increased MVV 18

in all these studies. Furthermore, RMET decreased the 19

9 and Pemax ) [3,4] might result from changes in rib cage
angles of the spinal curvature and reduced stiffness 20

of the rib cage in young healthy swimmers [10]. As 21

∗ Corresponding author: Brigitte Wirth, University Hospital Bal-

RMET requires the trainee to breath as hard and as 22

grist, Forchstr. 340, CH-8008 Zürich, Switzerland. Tel.: +41 44 386 deep as they can, it maximally mobilizes the entire 23

57 03; Fax: +41 44 386 57 09; E-mail: [Link]@[Link]. rib cage and the diaphragm. This rib cage mobiliz- 24

ISSN 1053-8127/16/$35.00 
c 2016 – IOS Press and the authors. All rights reserved
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 2

2 B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study

25 ing effect of RMET inspired some physiotherapists to [11.4 ± 3.5 of maximal 50 points in the German ver- 72

26 use the method in neck pain patients with quite some sion of the Neck Disability Index (NDI-G), range: 7.0– 73

27 success. However, there exists so far no study, which 20.0 points; 20.3 ± 8.8 of maximal 70 points in the 74

28 showed that neck pain patients are able to train as hard Bournemouth questionnaire, range: 8.7–42.0 points. 75

29 as healthy subjects and that RMET reduces neck pain. These assessments are described in detail in the sub- 76

30 The aim of this pilot study was to investigate for section “Self-assessment of neck disability”]. 77

31 the first time the feasibility and effects of 4 weeks

32 of RMET in a few patients with chronic neck pain.
2.2. Study design 78
33 The hypotheses were that also neck pain patients tol-
34 erate hard RMET and that the self-assessed neck func-
35 tion, the endurance of the neck flexor muscles, and The longitudinal study had a control and an inter- 79

36 chest mobility would improve. Finally, we expected vention phase (single-subject design). 3 baseline mea- 80

37 that these improvements would correlate with the in- surements served as a control, 4 measurements were 81

38 creases in the respiratory parameters (MVV, Pimax , performed during the intervention (intervals of 1 week 82

n
39 and Pemax ) induced by RMET. between 2 measurements), of which the last measure- 83

o
ment took place right at the end of the intervention. 84

si
Prior to the first baseline measurement, the participants 85

40 2. Methods were familiarized with all procedures on a separate 86

er
day. 2 months after the last training session, a follow- 87

2.1. Subjects up assessment was conducted by phone.

41
fv 88

Patients (> 18 years) having neck pain for longer

oo
42
2.3. Respiratory muscle endurance training (RMET) 89
43 than 6 months, who never underwent spinal surgery
44 and did not suffer from a neurological or inflamma-
RMET was performed with a SpiroTiger
pr

R
(idiag, 90
45 tory spinal pathology, were recruited from a medical
Fehraltorf, Switzerland), a hand-held device that al- 91
46 practice by an electronic newsletter and by a newspa-
lows for hyperpnoea ensuring normocapnia by partial
ed

92
47 per advertisement. Chronic whiplash-associated neck
48 pain was not an exclusion criterion [11]. Patients were CO2 rebreathing from a bag. The size of the hyperp- 93

noea can be individually set by adjusting the breathing 94

ct

49 advised not to change their medications and physio-

50 therapy, and to record all these therapies in a journal, frequency (fR ) and the bag volume of the SpiroTiger. 95
rre

51 which was regularly checked by an investigator. To ex- The participants conducted 5 sessions of RMET per 96

52 clude possible risk factors for the lung function mea- week, each session lasting for 30 min, for 4 weeks (20 97

training session in total). After a thorough familiariza-

co

98
53 surements, the participants completed a questionnaire
54 for the performance tests used in the exercise physiol- tion with the SpiroTiger R
(idiag, Fehraltorf, Switzer- 99

land) and with RMET under supervision of the exam- 100

un

55 ogy laboratory of ETH Zurich (Switzerland). Physical

56 activity was assessed by the Baecke questionnaire of iner at the occasion of the first introductory session, the 101

57 habitual physical activity [12]. All procedures were ap- participants performed RMET at home. To ensure the 102

58 proved by the ethics committee of the canton of Zurich correct use of the device and to adapt the bag volume 103

59 and performed in accordance with the Declaration of to training improvements, 1 training session per week 104

60 Helsinki. All patients gave written informed consent was conducted in the lab under the supervision of an 105

61 for their participation. investigator. 106

62 Between December 2013 and June 2014, 21 sub- RMET was performed with an fR that was individu- 107

63 jects entered the study. 6 participants terminated the ally calculated as 50% MVV divided by 50% vital ca- 108

64 study early mainly due to lack of time or motivation. pacity (VC). The initial training settings were as fol- 109

65 Thus, 15 patients (13 females; age: 49.3 ± 13.7 years; lows: if the participant could keep fR for the entire 30 110

66 height: 166.9 ± 7.2 cm; body mass: 67.8 ± 14.3 kg; min of RMET, the fR was increased by 2 breaths/min 111

67 physical activity: 9.3 ± 2.1 of maximal 15 points in for the next training session. The bag volume was in- 112

68 the Baecke questionnaire of habitual physical activity) dividually calculated from VC. The participants were 113

69 completed the study. 10 patients suffered from non- asked to record the daily training settings, and any pos- 114

70 traumatic neck pain. 11 patients suffered from neck sible adverse symptoms (e.g. headache or dizziness) in 115

71 pain for longer than 5 years. Neck disability was mild a journal. 116
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 3

B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study 3

117 2.4. Experimental procedures non-invasive electromechanical device, was used to de- 164

termine sagittal ROM of the thoracic spine [17]. The 165

118 2.4.1. Respiratory endurance test (RET) device was rolled down the spine starting with the 166

119 A RET to exhaustion with the SpiroTiger R

(idiag, 7th cervical vertebra (C7) and ending above the anal 167

120 Fehraltorf, Switzerland) was performed at the first crease. The measurements were performed in the neu- 168

121 baseline and the last intervention measurement. The in- tral, maximally flexed and maximally extended posi- 169

122 dividual fR was calculated as 60% MVV divided by tion. The angle of the thoracic spine in the neutral po- 170

123 50% VC so that a maximal test time of 2 to 10 min sition, and the ROM between maximal flexion and ex- 171

124 was achieved in the first RET. The second RET was tension (thoracic mobility) were used for further anal- 172

125 performed with settings identical to the first RET. The ysis. Chest mobility was measured at the axillary and 173

126 tests were stopped when the participant could not sus- xiphoid levels [18]. A flat measuring tape was drawn 174

127 tain either the target fR or the tidal volume or, in tight around the subject’s chest, and the difference in 175

128 case of no sign of exhaustion, after a maximum of circumference between maximal inspiration and maxi- 176

129 40 min [13]. mal expiration for each level was measured twice. The 177

n
mean of the 2 results was recorded. If the difference 178

o
130 2.4.2. Respiratory function testing between the 2 trials was greater than 1 cm, a 3rd mea- 179

si
131 Respiratory function was tested according to stan- surement was taken, and the average of the 2 largest 180

132 dard recommendations [14,15]. Spirometry was con- results was recorded. To determine the forward head 181

er
133 ducted with a MasterScope PC spirometer (Jaeger, posture (FHP; craniovertebral angle), a profile photo- 182

Hoechberg, Germany) that was calibrated prior to graph of the left side of the face was taken [19] in the
134
fv 183

135 each testing session. Testing parameters included VC, standing position [4,20]. The tragus of the ear and the 184

forced VC (FVC), forced expiratory volume in 1 s spinal process of C7 served as anatomical landmarks to
oo
136 185

137 (FEV1), peak expiratory flow (PEF), maximum expi- calculate the angle between the horizontal line running 186

138 ratory flow (MEF75%, MEF50%, MEF25%) and 12 s through C7 and the line from tragus to C7. Endurance 187
pr

139 MVV. The tests were repeated 3 to 5 times, depend- of the neck flexor muscle synergy was tested as recom- 188

140 ing on the standard between-maneuver criteria [14] and mended in a review paper [21]. The subjects were in 189
ed

141 the largest result was recorded. Pimax and Pemax were the supine position and were instructed to lift their head 190

142 tested with a digital respiratory pressure meter (Mi- approximately 2 cm while keeping their chins tucked 191
ct

143 croRPM, CareFusion, Hoechberg, Germany), a hand- in. The time until the subject was no longer able to hold 192

144 held device with a built-in small air leak to prevent this position was measured. The examiner monitored 193
rre

145 pressure generation by glottis closure. For Pimax and chin position visually and by a light finger touch [22]. 194

146 Pemax , subjects had to maximally inhale and exhale,

co

147 respectively, against a resistance for at least 1 s. A min- 2.6. Self-assessment of neck disability 195

148 imum of 5 and a maximum of 10 trials were performed

to ensure that the 3 largest measurements varied by
un

149 Neck disability was assessed using the German ver- 196

150 less than 5%. The maximum result was recorded. All sion of the NDI (NDI-G), which covers 10 areas of 197

151 tests were performed in a seated position with subjects daily living [23]. Each item is given a score in the range 198

152 wearing a nose clip. of 0 (no impairment) to 5 points (maximal impair- 199

ment), resulting in a maximal score of 50 points [24]. 200

153 2.5. Musculoskeletal assessments In line with the understanding of musculoskeletal dis- 201

orders as multidimensional problems, a German ver- 202

154 The musculoskeletal assessments resembled those sion of the Bournemouth questionnaire that covers 7 203

155 of a previous study [16]. Range of motion (ROM) dimensions of the bio-psycho-social pain model was 204

156 of the cervical spine was measured using Zebris R

applied [25,26]. The maximal score is 70 points (0 to 205

157 CMS10 (Zebris Medical, Isny, Germany). This ultra- 10 points per dimension). Both the total score and the 206

158 sound-based coordinate system measured 3 move- single items of the Bournemouth questionnaire were 207

159 ments of maximal flexion-extension, lateral flexion, analyzed. To determine the overall effect of RMET, 208

160 and rotation, and calculated the average for each di- a German version of the patient global impression of 209

161 rection. For rotation and lateral flexion, the mean of change (PGIC), was used. PGIC is a seven-point Lik- 210

162 left and right ROM was further analyzed. The Spinal ert scale with the extreme scores “much better” and 211

163 Mouse R
(idiag, Fehraltorf, Switzerland), a hand-held, “much worse” [27] and is recommended as a core out- 212
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 4

4 B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study

213 come of global improvement in chronic pain [28]. Pa- ters (Table 1). The other respiratory parameters did not 258

214 tients were asked how they felt at the end of 4 weeks change with the exception of MEF25%. 259

215 of RMET compared to before the intervention. Only After RMET, 12 participants reported clinically rel- 260

216 the 2 extreme scores (“much better” and “better”) were evant improvement in the PGIC (much better: N = 5; 261

217 defined as clinically significant improvement [29]. The better: N = 7); the remaining 3 participants felt only 262

218 Bournemouth questionnaire and the PGIC (“How do slightly better. Both the NDI-G and the Bournemouth 263

219 you feel now compared to before the study?”) were score significantly decreased during RMET (Fig. 1). 264

220 part of the follow-up assessment. After RMET, the average reduction in NDI-G was 265

−3.1 ± 2.0 points (range: −8.0 to +1.3 points), while 266

the mean reduction in the total Bournemouth score 267

221 2.7. Data analysis and statistics was −9.3 ± 5.5 points (range: −1.0 to −23.0 points). 268

Pain intensity, disability in daily activities, anxiety and 269

222 Because the differences between the parameters work related fear avoidance were the single items of 270

223 before and after training were normally distributed the Bournemouth questionnaire that significantly im- 271

(tested with Shapiro Wilk tests), their significance was proved through RMET with large ES (Table 2).

n
224 272

225 tested with paired t-tests. Changes in the scores of the RMET significantly increased endurance of the neck 273

o
226 NDI-G and the Bournemouth questionnaire were cal- flexor muscles (Fig. 2). The increase between the first 274

si
227 culated as differences between the mean of the 3 base- baseline measurement and the last intervention mea- 275

surement was significant and of large ES (r = 0.75).

er
276
228 line measurements and the last intervention measure-
229 ment. The questionnaire data and the musculoskele- Analogously, the increase between the last baseline 277
fv
230 tal assessments were tested with repeated measures measurement and the last intervention measurement 278

231 ANOVA and post-hoc paired t-tests. The first (Ses- was of the same size (r = 0.73), while no difference 279
oo
occurred between the 2 baseline measurements. 280
232 sion 1) and last baseline measurements (Session 3) as
Chest expansion significantly increased during 281
233 well as the last measurement of the intervention (Ses-
RMET (Fig. 3). The increase was more pronounced 282
pr

234 sion 7) were compared. In the post-hoc t-tests, the sig-

at the axillary level compared with the xiphoid level. 283
235 nificance level was adjusted to 0.05/3 = 0.017 (Bon- At the axillary level, the increase between the first 284
ferroni correction). The significance level α was set
ed

236
baseline measurement and the last intervention mea- 285
237 at 0.05 in all other analyses. To quantify the impor- surement was significant and of large ES (r = 0.77). 286
238 tance of the findings, the effect sizes (ES) were cal-
√ Analogously, the increase between the second base-
ct

287
239 culated for the significant results [r = t2 /(t2 + df ) line measurement and the last intervention measure- 288
for the t-tests; partial η 2 for the repeated measures
rre

240 ment was also significant (r = 0.84), while the av- 289

241 ANOVA] [30]. Correlations between changes (result erage of the 2 baseline results was identical. At the 290

242 of session 7 as percentage of the mean of the 3 base- xiphoid level, there was no change between the first
co

291

243 line results) in those respiratory and musculoskeletal and the second baseline measurement. The last inter- 292

244 parameters that significantly changed through RMET vention measurement was significantly higher than the 293
un

245 were calculated for all participants using the Pearson second baseline result (r = 0.75) but did not differ 294

246 correlation coefficient r. In all tests, r-values > 0.3 and from the first baseline result. All other musculoskeletal 295

247 < 0.5 were regarded as medium, and values > 0.5 as parameters did not differ between baseline and inter- 296

248 a large ES [30]. Data sets with missing values were vention (Table 3). 297

249 excluded from the corresponding analyses (available Correlation analyses revealed no significant correla- 298

250 case-analysis). All analyses used IBM SPSS Statistics tion between the observed changes in the respiratory 299

251 20.0 (SPSS, Chicago, IL, USA). parameters and the musculoskeletal parameters which 300

improved through RMET (Table 4). 301

After the end of this study, 8 patients continued with 302

RMET (7 bought a SpiroTiger R

and 1 borrowed an in- 303
252 3. Results strument). In the follow-up, 9 patients reported clini- 304

cally relevant improvement in the PGIC compared to 305

253 RMET increased the RET from 5.4 ± 1.8 to 38.3 ± the beginning of the study (much better: N = 3; bet- 306
254 6.4 min (p < 0.001). In the final RET, all but 1 partic- ter: N = 6), while 5 were slightly better and 1 pa- 307

255 ipant (who stopped at 15 min) reached 40 min. RMET tient felt no change. The mean reduction in the to- 308

256 significantly increased VC, MVV, Pemax and Pimax , tal Bournemouth score was −5.6 ± 6.9 points (range: 309

257 and these increases were of large ES for all parame- −16.7 to + 12.3 points). 310
Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 5

B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study 5

Table 1
Respiratory parameters before and after respiratory muscle endurance training
Before intervention mean (SD) End intervention mean (SD) p-value (ES)
VC (l) 3.8 (0.7) 4.0 (0.7) 0.004 (0.68)
FVC (l) 3.6 (0.6) 3.6 (0.6) 0.470
FEV1 (l) 3.1 (0.6) 3.1 (0.6) 0.747
PEF (l/s) 6.4 (1.9) 6.8 (1.7) 0.193
MEF 75% (l/s) 6.0 (1.6) 6.0 (1.2) 0.816
MEF 50% (l/s) 3.6 (1.0) 3.7 (1.2) 0.830
MEF 25% (l/s) 1.3 (0.7) 0.9 (0.5) 0.020 (0.57)
MVV (l/min) 111.4 (25.6) 121.0 (27.9) 0.025 (0.56)
Pemax (cm H2 O) 114.7 (30.6) 135.2 (38.3) < 0.001 (0.83)
Pimax (cm H2 O) 77.9 (26.8) 95.7 (23.6) 0.001 (0.73)
VC vital capacity; FVC forced vital capacity; FEV1 forced expiratory flow in 1 s; PEF peak expiratory flow; MEF 75% maximum expiratory
flow at 75% of FVC; MEF 50% maximum expiratory flow at 50% of FVC; MEF 25% maximum expiratory flow at 25% of FVC; MVV maximal
voluntary ventilation; Pemax maximal expiratory pressure; Pimax maximal inspiratory pressure; ES effect size: proportion of total variance that
is explained by the effect (calculated only for the significant results).

o n
Table 2
Single items of the Bournemouth questionnaire before and after respiratory muscle endurance training

si
BB EB EI ANOVA p-value Paired t-tests Paired t-tests Paired t-tests

er
mean (SD) mean (SD) mean (SD) (part η2 ) p (ES) (EB−EI) p (ES) (BB−EI) p (ES) (BB−EB)
Pain intensity 4.6 (2.5) 3.9 (1.9) 2.2 (1.5) 0.001 (0.48) < 0.001 (0.80) 0.001 (0.74) 0.159
fv
Disability in daily activities 2.1 (1.5) 2.7 (1.9) 1.1 (1.3) 0.001 (0.42) < 0.001 (0.78) 0.006 (0.65) 0.178
Disability in social activities 1.3 (1.4) 2.1 (2.2) 1.1 (1.5) 0.121 (0.16)
Anxiety 3.9 (2.5) 3.6 (1.9) 1.9 (1.7) 0.018 (0.31) < 0.001 (0.84) 0.014 (0.60) 0.724
oo

Depression 2.0 (1.8) 2.1 (1.9) 1.5 (1.7) 0.369 (0.06)

Work related fear avoidance 3.5 (2.9) 3.2 (1.8) 1.8 (1.5) 0.009 (0.32) 0.004 (0.67) 0.013 (0.60) 0.511
Pain locus of control 3.1 (2.6) 3.6 (2.6) 1.5 (1.4) 0.051 (0.23)
pr

BB begin baseline; EB end baseline; EI end intervention; part η2 measure of effect size for repeated measures ANOVA: proportion of variance
that a variable explains which is not explained by other variables (calculated only for the significant results); ES effect size for t-tests: proportion
ed

of total variance that is explained by the effect (calculated only for the significant results).

Table 3
ct

Musculoskeletal parameters before and after respiratory muscle endurance training

rre

Begin baseline mean (SD) End baseline mean (SD) End intervention mean (SD) ANOVA p-value
CS Flexion (◦ ) 52.1 (14.0) 49.5 (12.7) 55.9 (14.4) 0.060
CS Extension (◦ ) 50.3 (12.8) 50.1 (13.1) 49.9 (14.8) 0.985
co

CS Rotation (◦ ) 65.5 (7.9) 63.3 (7.5) 64.9 (7.9) 0.246

CS Latero-Flexion (◦ ) 32.2 (5.9) 31.2 (5.4) 32.9 (6.2) 0.204
TS Neutral position (◦ ) 45.9 (10.4) 46.5 (12.1) 46.0 (10.1) 0.901
un

TS Mobility (◦ ) 22.9 (12.1) 21.1 (15.6) 20.1 (12.8) 0.708

Forward head posture (◦ ) 45.9 (5.5) 45.0 (5.5) 44.5 (5.4) 0.293
CS cervical spine; TS thoracic spine.

Table 4
Correlations between changes in respiratory parameters and changes of those musculoskeletal parameters that improved through respiratory
muscle endurance training
VC r (p-value) MEF 25% r (p-value) MVV r (p-value) Pemax r (p-value) Pimax r (p-value)
NDI 0.08 (0.778) −0.03 (0.916) 0.50 (0.056) 0.14 (0.622) 0.43 (0.108)
BM total score 0.22 (0.434) −0.13 (0.635) 0.34 (0.213) 0.11 (0.685) 0.36 (0.192)
BM pain intensity −0.15 (0.595) −0.15 (0.595) 0.03 (0.906) −0.22 (0.437) −0.15 (0.587)
BM daily activities 0.35 (0.206) −0.45 (0.093) 0.12 (0.675) 0.01 (0.961) 0.22 (0.439)
BM anxiety −0.12 (0.675) −0.02 (0.932) 0.51 (0.054) −0.29 (0.292) 0.44 (0.105)
BM fear avoidance 0.08 (0.789) 0.18 (0.512) 0.44 (0.099) 0.09 (0.756) 0.23 (0.419)
Neck flexor endurance −0.02 (0.934) 0.01 (0.976) −0.11 (0.685) −0.15 (0.598) −0.10 (0.733)
Chest expansion axilla −0.12 (0.668) 0.32 (0.248) 0.19 (0.495) −0.22 (0.427) 0.11 (0.698)
NDI Neck disability index; BM Bournemouth questionnaire; VC vital capacity; MEF 25% maximum expiratory flow at 25% of FVC; MVV
maximal voluntary ventilation; Pemax maximal expiratory pressure; Pimax maximal inspiratory pressure; r Pearson correlation coefficient.
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 6

6 B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study

o n
si
er
fv
oo
pr
ed
ct
rre
co
un

Fig. 1. Changes in self-assessed neck function through respiratory muscle endurance training. The repeated measure ANOVA showed that both
the NDI-G [F(1.45,20.24) = 11.86, p = 0.001, partial η2 = 0.46] and the Bournemouth questionnaire [F(1.35,18.90) = 10.85, p = 0.002, partial
η2 = 0.44] significantly decreased during RMET. The improvement from the first baseline measurement to the last intervention measurement was
significant and of large ES for both the NDI-G (p < 0.001, r = 0.82) and the Bournemouth questionnaire (p = 0.001; r = 0.74). Analogously,
both scores significantly improved from the last baseline measurement to the last intervention measurement (NDI-G: p = 0.001, r = 0.77;
Bournemouth: p < 0.001; r = 0.86), while the baseline measurements did not differ from each other (NDI-G: p = 0.345; Bournemouth: p =
0.834). n.s.: not significant; ∗∗∗ p  0.001; O: outlier (value > third quartile + 1.5∗ interquartile range).

311 4. Discussion ever, the observed changes in the respiratory and the 318

musculoskeletal parameters did not correlate. 319

312 RMET significantly improved VC, MVV and the The respiratory function of the neck patients in the 320

313 maximal respiratory pressures. It also significantly im- present study was at the beginning as impaired as that 321

314 proved self-perceived neck function, predominantly of the chronic neck pain patients in earlier studies [3, 322

315 with respect to pain intensity, daily life activity, anxi- 4]. After RMET, Pemax improved to a level compara- 323

316 ety and work-related fear avoidance. Objectively, neck ble to healthy controls, but VC, MVV and Pimax re- 324

317 flexor endurance and chest expansion increased. How- mained reduced in comparison of another study [4]. 325
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 7

B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study 7

o n
si
er
Fig. 2. Changes in neck flexor endurance through respiratory muscle endurance training. The repeated measure ANOVA showed that the neck
flexor endurance significantly increased during RMET [F(1.63,22.86) = 13.61, p < 0.001, partial η2 = 0.49]. n.s.: not significant; ∗∗∗ p 
fv
0.001; O: outlier (value > third quartile + 1.5∗ interquartile range); ∗ extreme value (> third quartile + 3∗ interquartile range).
oo
326 Previous studies in healthy subjects found a signifi- pants in a previous study that found an effect of RMET 356

327 cant increase in VC after RMET, but unchanged Pimax on spinal curvature [10]. Thus, it appears that the pos- 357

328 and Pemax [9], while patients with incomplete cervical tural changes we observed had developed over years, 358
pr

329 spinal cord injury, myasthenia gravis, and chronic ob- and are not as reversible as those of young subjects. 359

330 structive pulmonary disease significantly improved in It is difficult to allocate the improvements in self- 360
ed

331 VC, Pimax and Pemax through RMET [6–8]. The sig- assessed neck function and in the musculoskeletal pa- 361

332 nificant reduction in MEF 25% that we observed in the rameters to a pathophysiological mechanism as no cor- 362
ct

333 present study after RMET might be explained by the relations with the changes in the respiratory parame- 363

334 increased expiratory force that most likely closed the ters were found. Previous studies on the effects of res- 364
rre

335 small bronchioles before the end of the expiration. This piratory training on neck pain and musculoskeletal pa- 365

336 could also explain why FVC was unchanged despite a rameters generally reported musculoskeletal changes 366
co

337 somewhat larger VC after RMET. in parallel to changes in respiratory parameters, rather 367

338 RMET also increased neck flexor endurance and than investigating the relation between these chan- 368
un

339 chest wall expansion, predominantly at the axillary ges [10,32]. Thus, further studies on the underlying 369

340 level. The sternocleidomastoids and the scalene mus- mechanisms of the effects of RMET are needed. 370

341 cles are involved not only in neck motion but also An alternative explanation is that anxiety and fear of 371

342 in respiration. Thus, strengthening these muscles by motion induce hyperventilation in these patients, lead- 372

343 RMET might improve the kinetic control in the cervi- ing to hypocapnia [reduction of arterial carbon dioxide 373

344 cal and thoracic spine, leading to changes in rib cage (CO2 )]. Hyperventilation and resulting hypocapnia in 374

345 mechanics [5]. Although a strong association between patients with chronic neck pain have been reported pre- 375

346 FHP and respiratory muscle strength has previously viously [2]. Accordingly, there is some evidence from a 376

347 been shown [4], RMET did not affect head posture small sample of neck patients that breathing re-training 377

348 or the thoracic spine in our patients. Our FHP results by biofeedback with a capnograph led to normalization 378

349 were comparable to those of healthy control subjects of end-tidal CO2 , to pain reduction, and to improve- 379

350 and neck pain patients in an earlier study [4], but they ment in neck function [32]. Additionally, it has re- 380

351 were lower than those of asymptomatic participants in cently been shown that disability in chronic neck pain 381

352 a large study [19]. Reduced craniovertebral angles (in- patients can be predicted by anxiety and catastrophiz- 382

353 creased FHP) correlate with greater thoracic kypho- ing [33]. Remarkably, particularly the Bournemouth 383

354 sis [31]. Indeed, the subjects in this study presented questionnaire items for pain intensity, anxiety and ki- 384

355 with larger thoracic angles than young healthy partici- nesiophobia improved in our subjects through RMET. 385
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 8

8 B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study

o n
si
er
fv
oo
pr
ed
ct
rre
co
un

Fig. 3. Changes in chest expansion at axilla and xiphoid through respiratory muscle endurance training. The repeated measure ANOVA showed
that both chest expansions at the axilla [F(1.35,18.95) = 22.42, p < 0.001, partial η2 = 0.62] as well as at the xiphoid [F(1.61,22.54) = 7.48,
p = 0.005, partial η2 = 0.35] significantly increased during RMET. n.s.: not significant; ∗∗∗ p  0.001; O: outlier (value > third quartile + 1.5∗
interquartile range or value < first quartile − 1.5∗ interquartile range).

386 Thus, the effects of RMET might result predomi- The main limitation of the present study was that 397

387 nantly from reducing hyperventilation and hypocap- the selected single-subject design cannot control for 398

388 nia. Hypocapnia was reported to raise the muscle tone motivational aspects. However, a phased approach in- 399

389 by increasing the excitability of nerves and muscles cluding an exploratory study preceding a randomized 400

390 (changing the membrane potentials towards the thresh- controlled trial is recommended for the evaluation of 401

391 old for the action-potential) [34]. Accordingly, the im- complex health interventions [35]. The single case re- 402

392 provement of headaches, which was reported by some search design, in which every participant serves as 403

393 participants, might result from reduced hypocapnia. its own control, was suggested to be particularly ap- 404

394 Thus, further studies on the effects of RMET on neck propriate for the evaluation of hypotheses that might 405

395 pain should include blood gas analyses and correlate be further investigated with other research methods 406

396 any changes in CO2 with changes in neck function. at a later date [36]. Furthermore, besides being eth- 407
 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 9

B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study 9

408 ically questionable, conducting placebo training with [3] Dimitriadis Z, Kapreli E, Strimpakos N, Oldham J. Respira- 450

409 the SpiroTiger R

is difficult because ineffective settings tory weakness in patients with chronic neck pain. Man Ther. 451
2013; 18(3): 248. 452
410 are likely to be identified by the participants. Never- [4] Kapreli E, Vourazanis E, Billis E, Oldham JA, Strimpakos N. 453
411 theless in a next step, a randomized controlled study Respiratory dysfunction in chronic neck pain patients. A pilot 454
412 that includes patients with more serious neck prob- study. Cephalalgia. 2009; 29(7): 701. 455

413 lems, who might present with a more manifest respira- [5] Kapreli E, Vourazanis E, Strimpakos N. Neck pain causes res- 456
piratory dysfunction. Med Hypotheses. 2008; 70(5): 1009. 457
414 tory pathology [37], should compare RMET to a con- [6] Berlowitz DJ, Tamplin J. Respiratory muscle training for cer- 458
415 trol group or to conventional neck rehabilitation strate- vical spinal cord injury. Cochrane Database Syst Rev. 2013; 459
416 gies. Thereby, the craniocervical flexion test using an 7: CD008507. 460

417 inflatable pressure biofeedback unit might alternatively [7] Rassler B, Hallebach G, Kalischewski P, Baumann I, Schauer 461
J, Spengler CM. The effect of respiratory muscle endurance 462
418 be conducted, but needs further investigation, as it has
training in patients with myasthenia gravis. Neuromuscul Dis- 463
419 so far not been recommended [21]. Lastly, although the ord. 2007; 17(5): 385. 464
420 compliance with RMET of those participants who fin- [8] Scherer TA, Spengler CM, Owassapian D, Imhof E, Boutel- 465

421 ished the study was excellent, there were 6 drop-outs. lier U. Respiratory muscle endurance training in chronic ob- 466
structive pulmonary disease: impact on exercise capacity, dys- 467

n
422 RMET may only be of value to highly motivated pa-
pnea, and quality of life. Am J Respir Crit Care Med. 2000; 468

o
423 tients, which would restrict its applicability. 162(5): 1709. 469

si
[9] Verges S, Boutellier U, Spengler CM. Effect of respiratory 470
muscle endurance training on respiratory sensations, respira- 471

er
424 5. Conclusion tory control and exercise performance: A 15-year experience. 472
Respir Physiol Neurobiol. 2008; 161(1): 16. 473
fv
[10] Obayashi H, Urabe Y, Yamanaka Y, Okuma R. Effects of 474
425 RMET emerged from this pilot study as a feasible respiratory-muscle exercise on spinal curvature. J Sport Re- 475

426 and effective method to rehabilitate chronic neck pain. habil. 2012; 21(1): 63. 476
oo

427 Although some biomechanical changes were observed, [11] Verhagen AP, Lewis M, Schellingerhout JM, Heymans MW, 477
Dziedzic K, de Vet HC, et al. Do whiplash patients differ from 478
428 the underlying mechanisms, including blood gas anal- other patients with non-specific neck pain regarding pain, 479
pr

429 yses, need further investigation in a randomized con- function or prognosis? Man Ther. 2011; 16(5): 456. 480
430 trolled study. [12] Baecke JA, Burema J, Frijters JE. A short questionnaire for 481
the measurement of habitual physical activity in epidemiolog- 482
ed

ical studies. Am J Clin Nutr. 1982; 36(5): 936. 483

[13] Markov G, Spengler CM, Knopfli-Lenzin C, Stuessi C, 484
431 Acknowledgements Boutellier U. Respiratory muscle training increases cycling 485
ct

endurance without affecting cardiovascular responses to exer- 486

We thank all patients for participation. Furthermore, cise. Eur J Appl Physiol. 2001; 85(3-4): 233. 487
rre

432

433 we thank Anja Trepp from idiag (Fehraltorf, Switzer- [14] Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, 488
Coates A, et al. Standardisation of spirometry. Eur Respir J. 489
434 land) for providing the SpiroTiger R
and the Spinal 2005; 26(2): 319. 490
co

435 Mouse. We also thank the Institute of Physiology (Uni- [15] ATS/ERS. ATS/ERS statement on respiratory muscle testing. 491

436 versity of Zurich, Switzerland) for providing a room Am J Respir Crit Care Med. 2002; 166: 518. 492
un

437 and additional equipment. [16] Wirth B, Amstalden M, Perk M, Boutellier U, Humphreys 493
BK. Respiratory dysfunction in patients with chronic neck 494
pain – influence of thoracic spine and chest mobility. Man 495
Ther. 2014; 19(5): 440. 496

438 Conflict of interest [17] Mannion AF, Knecht K, Balaban G, Dvorak J, Grob D. A new 497
skin-surface device for measuring the curvature and global 498
and segmental ranges of motion of the spine: Reliability of 499
439 The authors have no conflict of interest to report. measurements and comparison with data reviewed from the 500
literature. Eur Spine J. 2004; 13(2): 122. 501
[18] Lunardi AC, Marques da Silva CC, Rodrigues Mendes FA, 502

440 References Marques AP, Stelmach R, Fernandes Carvalho CR. Mus- 503
culoskeletal dysfunction and pain in adults with asthma. J 504

441 [1] Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ez- Asthma. 2011; 48(1): 105. 505

442 zati M, et al. Years lived with disability (YLDs) for 1160 se- [19] Raine S, Twomey LT. Head and shoulder posture variations in 506

443 quelae of 289 diseases and injuries 1990–2010: A systematic 160 asymptomatic women and men. Arch Phys Med Rehabil. 507

444 analysis for the Global Burden of Disease Study 2010. Lancet. 1997; 78(11): 1215. 508

445 2012; 380(9859): 2163. [20] Fernandez-de-las-Penas C, Perez-de-Heredia M, Molero- 509

446 [2] Dimitriadis Z, Kapreli E, Strimpakos N, Oldham J. Hypocap- Sanchez A, Miangolarra-Page JC. Performance of the cran- 510

447 nia in patients with chronic neck pain: Association with pain, iocervical flexion test, forward head posture, and headache 511

448 muscle function, and psychologic states. Am J Phys Med Re- clinical parameters in patients with chronic tension-type 512

449 habil. 2013; 92(9): 746.

 Galley Proof 18/03/2016; 16:53 File: bmr–[Link]; BOKCTP/wyn p. 10

10 B. Wirth et al. / Respiratory muscle endurance training reduces chronic neck pain: A pilot study

513 headache: A pilot study. J Orthop Sports Phys Ther. 2007; CS, Farrar JT, et al. Interpreting the clinical importance of 542
514 37(2): 33. treatment outcomes in chronic pain clinical trials: IMMPACT 543
515 [21] de Koning CH, van den Heuvel SP, Staal JB, Smits- recommendations. J Pain. 2008; 9(2): 105. 544
516 Engelsman BC, Hendriks EJ. Clinimetric evaluation of meth- [29] Hurst H, Bolton J. Assessing the clinical significance of 545
517 ods to measure muscle functioning in patients with non- change scores recorded on subjective outcome measures. J 546
518 specific neck pain: A systematic review. BMC Musculoskelet Manipulative Physiol Ther. 2004; 27(1): 26. 547
519 Disord. 2008; 9: 142. [30] Field A. Discovering statistics using SPSS. 3rd edition ed. 548
520 [22] Grimmer K. Measuring the endurance capacity of the cervical London: SAGE Publications; 2009. 549
521 short flexor muscle group. Australian Journal of Physiother- [31] Quek J, Pua YH, Clark RA, Bryant AL. Effects of thoracic 550
522 apy. 1994; 40: 251. kyphosis and forward head posture on cervical range of mo- 551
523 [23] Swanenburg J, Humphreys K, Langenfeld A, Brunner F, tion in older adults. Man Ther. 2013; 18(1): 65. 552
524 Wirth B. Validity and reliability of a German version of the [32] McLaughlin L, Goldsmith CH, Coleman K. Breathing evalua- 553
525 Neck Disability Index (NDI-G). Man Ther. 2014; 19(1): 52. tion and retraining as an adjunct to manual therapy. Man Ther. 554
526 [24] Vernon H, Mior S. The Neck Disability Index: A study of reli- 2011; 16(1): 51. 555
527 ability and validity. J Manipulative Physiol Ther. 1991; 14(7): [33] Dimitriadis Z, Kapreli E, Strimpakos N, Oldham J. Do psy- 556
528 409. chological states associate with pain and disability in chronic 557
529 [25] Bolton JE, Humphreys BK. The Bournemouth Questionnaire: neck pain patients? J Back Musculoskelet Rehabil. 2015. 558
530 A short-form comprehensive outcome measure. II. Psychome- [34] Folgering H, Snik A. Hyperventilation syndrome and muscle 559

n
531 tric properties in neck pain patients. J Manipulative Physiol fatigue. J Psychosom Res. 1988; 32(2): 165. 560

o
532 Ther. 2002; 25(3): 141. [35] Campbell M, Fitzpatrick R, Haines A, Kinmonth AL, Sander- 561

si
533 [26] Soklic M, Peterson C, Humphreys BK. Translation and vali- cock P, Spiegelhalter D, et al. Framework for design and 562
534 dation of the German version of the Bournemouth Question- evaluation of complex interventions to improve health. BMJ. 563

er
535 naire for Neck Pain. Chiropr Man Therap. 2012; 20(1): 2. 2000; 321(7262): 694. 564
536 [27] Salaffi F, Stancati A, Silvestri CA, Ciapetti A, Grassi W. Min- [36] Aldridge D. Single-case research designs for the clinician. J 565
537 imal clinically important changes in chronic musculoskele-
fv
R Soc Med. 1991; 84(5): 249. 566
538 tal pain intensity measured on a numerical rating scale. Eur J [37] Dimitriadis Z, Kapreli E, Strimpakos N, Oldham J. Pul- 567
539 Pain. 2004; 8(4): 283. monary function of patients with chronic neck pain: A spirom- 568
oo

540 [28] Dworkin RH, Turk DC, Wyrwich KW, Beaton D, Cleeland etry study. Respir Care. 2014; 59(4): 543. 569
541
pr
ed
ct
rre
co
un