94
For peer review only The abdominal drawing-in maneuver for detecting activity in the deep abdominal muscles: is this clinical tool reliable and valid? Journal: BMJ Open Manuscript ID: bmjopen-2015-008711 Article Type: Research Date Submitted by the Author: 07-May-2015 Complete List of Authors: Kaping, Karsten; NVS-Sjukgymnastik, Karolinska Institutet; Rehabtjanst Äng, Björn; NVS-Sjukgymnastik, Karolinska Institutet Rasmussen-Barr, Eva; NVS-Sjukgymnastik, Karolinska Institutet <b>Primary Subject Heading</b>: Rehabilitation medicine Secondary Subject Heading: Sports and exercise medicine, Radiology and imaging, Diagnostics, Rehabilitation medicine Keywords: Back pain < ORTHOPAEDIC & TRAUMA SURGERY, Musculoskeletal disorders < ORTHOPAEDIC & TRAUMA SURGERY, Orthopaedic sports trauma < ORTHOPAEDIC & TRAUMA SURGERY, Spine < ORTHOPAEDIC & TRAUMA SURGERY, REHABILITATION MEDICINE, ULTRASONOGRAPHY For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open on June 23, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2015-008711 on 9 December 2015. Downloaded from

BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* [email protected];

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

The abdominal drawing-in maneuver for detecting activity in the deep abdominal muscles: is this clinical tool reliable

and valid?

Journal: BMJ Open

Manuscript ID: bmjopen-2015-008711

Article Type: Research

Date Submitted by the Author: 07-May-2015

Complete List of Authors: Kaping, Karsten; NVS-Sjukgymnastik, Karolinska Institutet; Rehabtjanst Äng, Björn; NVS-Sjukgymnastik, Karolinska Institutet Rasmussen-Barr, Eva; NVS-Sjukgymnastik, Karolinska Institutet

<b>Primary Subject Heading</b>:

Rehabilitation medicine

Secondary Subject Heading: Sports and exercise medicine, Radiology and imaging, Diagnostics, Rehabilitation medicine

Keywords:

Back pain < ORTHOPAEDIC & TRAUMA SURGERY, Musculoskeletal disorders < ORTHOPAEDIC & TRAUMA SURGERY, Orthopaedic sports trauma < ORTHOPAEDIC & TRAUMA SURGERY, Spine < ORTHOPAEDIC & TRAUMA SURGERY, REHABILITATION MEDICINE, ULTRASONOGRAPHY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open on June 23, 2020 by guest. P

rotected by copyright.http://bm

jopen.bmj.com

/B

MJ O

pen: first published as 10.1136/bmjopen-2015-008711 on 9 D

ecember 2015. D

ownloaded from

Page 2: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

1

The abdominal drawing-in maneuver for detecting activity in the 1

deep abdominal muscles: is this clinical tool reliable and valid? 2

3

Statements 4

Financial support from the Swedish Confederation of Physiotherapists-Memorial Fund is 5

gratefully acknowledged. 6

7

Authors’ contributions 8

All authors participated in the conception and design of the study, and all made substantial 9

contributions to the acquisition of data. BOÄ and ERB performed the statistical analysis. All 10

authors participated in the interpretation of data. KK drafted the manuscript and BOÄ and 11

ERB critically revised the manuscript for important intellectual content. All the authors read 12

and approved the final manuscript 13

14

Funding 15

KK works as a salaried private practitioner, BOÄ and ERB receive their salaries from the 16

Karolinska Institutet 17

18

Competing interests 19

None. 20

21

22

Ethics approval 23

The regional ethical review board in Stockholm, Sweden, approved the study (Dnr. 24

2011/1859–31). 25

26

Provenance and peer review: not commissioned; externally peer reviewed. 27

28

Page 1 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 3: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

2

Data sharing statement: additional data are available from Karsten Kaping, 1

[email protected] 2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

Page 2 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 4: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

3

The abdominal drawing-in maneuver for detecting activity in the 1

deep abdominal muscles: is this clinical tool reliable and valid? 2

3

4

5

Karsten Kaping, RPT, MT, MSc *,1,2 Björn O. Äng, RPT, Associate Professor, 1 6

Eva Rasmussen-Barr. RPT, PhD 1,2 7

8

9

Address: 1 Department of Neurobiology, Care Sciences and Society, Division of 10

Physiotherapy, Karolinska Institutet, Alfred Nobels Allé 12, 141 83 Huddinge 11

Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 12

Telephone: +46-8-441 16 60 13

Email: Karsten Kaping* [email protected]; 14

[email protected]; [email protected] 15

*Corresponding author 16

17

Keywords: 18

Abdominal muscles, concurrent validity, lumbar spine, transversus abdominis, ultrasound 19

imaging 20

Word count: 3561 21

22

23

24

Page 3 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 5: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

4

ABSTRACT (word count 288) 1

Objective: The abdominal drawing-in maneuver (ADIM) is a common clinical tool for 2

manually assessing as “correct” or not a preferential activation of the deep abdominal muscles 3

in patients with low-back pain (LBP). The validity and reliability of manual assessment of the 4

ADIM are, however, as yet unknown. This study evaluated the concurrent and discriminative 5

validity and the reliability of the manually assessed ADIM. 6

Design: Single-blinded cross-sectional study 7

Settings: General population in Stockholm County, Sweden 8

Participants: The study sample comprised 38 subjects seeking care for low-back pain and 15 9

healthy subjects. 10

Measures: The manual ADIM was assessed as correct or not following a standard procedure. 11

Ultrasound imaging (USI) was used as the concurrent reference (gold standard) for the 12

manually assessed ADIM by calculating a ratio of the change in muscle thickness between the 13

resting and the contracted states: the correlation between manual test and the USI was 14

calculated. Discriminative validity was analysed by calculating sensitivity and specificity. A 15

sample of 24 subjects was analysed with kappa coefficients for inter-observer reliability 16

between two raters. 17

Results: The concurrent validity between the manual ADIM and the ADIM-USI ratios 18

showed poor correlations (r=.13-.40). The discriminative validity of the manually assessed 19

ADIM to predict LBP showed a sensitivity/specificity of 0.30/0.73, while the ADIM-USI 20

ratio to predict LBP showed 0.19/0.87. The inter-observer reliability for the manually-21

assessed ADIM revealed substantial agreement: (K= 0.71, CI (95%) 0.41-1.00. 22

Conclusion: Although the inter-observer reliability of the manually-assessed ADIM was 23

high, both the concurrent and discriminative validity was low for examining the preferential 24

activity of the deep abdominal muscles. Neither the manually-assessed ADIM nor the 25

ultrasound testing discriminated between subjects with LBP and healthy subjects regarding 26

preferential activity of the transversus muscle as this ability/inability was also present in 27

healthy subjects. 28

29

30

31

Page 4 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 6: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

5

1

2

3

4

5

6

STRENGTHS AND LIMITATIONS 7

Strengths and limitations of this study 8

Strengths and limitations of this study 9

•Strengths of this study are that the abdominal-drawing-in-manoeuvre was analysed for both 10

concurrent and discriminative validity and that the investigators were blinded to all subjects’ 11

health status. 12

•Strength is also that the study used a gold-standard tool (ultrasound) to measure the quantity 13

of muscle activity for comparison with the manual estimation. 14

•Possible limitations are that our study design resulted in a disproportionate sample between 15

healthy subjects and those with low-back pain. 16

•In addition, the patients estimated their low-back pain only as low-to-moderate 17

18

19

20

21

22

23

24

25

26

27

Page 5 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 7: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

6

INTRODUCTION (word count 3388) 1

The lumbar spine depends on complex interaction between passive and active structures such 2

as contracting muscles, ligaments and fascia to maintain optimal spinal control during bodily 3

movement and loading [1-3]. Contraction of the lateral abdominal muscles, and especially the 4

anticipatory activity of the transversus abdominus muscle (TrA), is suggested to be a key 5

motor-control element for transferring spinal loads to the pelvis and upper trunk, thus 6

avoiding injury to spinal tissue or pain incidents following sub-failure loading. [2-5] 7

Studies have reported delayed anticipatory activity of TrA in low-back-pain sufferers (LBP) 8

in comparison to healthy subjects [5-7], although such differences in temporal activity have 9

also been questioned. [8] Specific exercises have been proposed for re-training motor control 10

of the abdominal muscles; a popular and common treatment for people with recurrent and 11

chronic LBP.[9-10] Following such interventions, recent systematic reviews have reported 12

improvement of symptoms such as pain and functional deficits compared to other exercises or 13

treatment.[11-12] 14

To assess the activity of the deep abdominal muscles a specific clinical test is available: the 15

abdominal drawing-in maneuver (ADIM). [13] Through manual palpation and visual 16

inspection, the ADIM is used to assess the success or not in performing a voluntary and 17

preferential contraction of the deep abdominal muscles, especially the TrA [13-14], in order to 18

target interventions with specific exercises and follow-up progression. 19

In addition, ultrasound imaging (USI) has for several years been used in clinical settings to 20

quantify the activity of the deep abdominal muscles. Specifically, it has also been used in 21

research to quantify change in muscle thickness or muscle onset of TrA before and after 22

interventions using motor control exercises.[15] The USI has been validated for measuring 23

Page 6 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 8: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

7

change in muscle thickness against electromyography (EMG) [16] and magnetic resonance 1

imaging ( MRI) [17] and demonstrates high inter-observer reliability in both healthy people 2

and subjects with LBP. [14, 18-21] However, not all clinicians have access to USI and the 3

manually assessed ADIM therefore remains the most common clinical tool for addressing and 4

examining a preferential contraction of the deep abdominal muscles. 5

A clinically convenient test should show both good reliability and high validity. Specifically, 6

it should be able to demonstrate concurrent validity against a reliable gold-standard 7

assessment (e.g., USI verifying TrA activity); proved to measure what it purports to 8

measure.[22] If the clinical test is to reflect impairment, not merely normal variation in 9

healthy subjects, the test should also exhibit empirical, discriminative validity to justify its use 10

among patients. [23] For the present purpose discriminative validity was defined as the extent 11

to which manually-assessed and USI-assessed ADIM differentiated between subjects with 12

LBP and those without.[24] 13

To date, the concurrent and discriminative validity of the manually assessed ADIM is 14

unknown, and only one study addresses inter-observer reliability.[25] The present aims were 15

therefore to assess the concurrent validity of the manually assessed ADIM in relation to the 16

USI during guided abdominal contractions, to assess its discriminative ability, and to 17

determine the inter-observer reliability. 18

METHODS 19

Study design 20

A cross-sectional, single-blinded design was used to evaluate the concurrent validity, 21

discriminating validity and inter-observer reliability of the ADIM in subjects with LBP and 22

in healthy subjects. A sample size of 50 subjects was considered adequate to power the 23

study to obtain proper correlation data as used in concurrent analysis, [26], while a sample 24

Page 7 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 9: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

8

of 27 subjects was calculated for kappa agreement data (confidence interval 0.3; expected 1

percentage agreement 0.8 and chance agreement 0.5) used in reliability statistics.[27] 2

Planned enrolment was undertaken to meet these criteria. 3

4

Study sample 5

Subjects with LBP were recruited through information about the study distributed by e-6

mail and telephone to six private physiotherapy settings in Stockholm, Sweden. Healthy 7

subjects were found through written and verbal information distributed at two large 8

companies in Stockholm, each with approximately 700 employees. Fifty-three subjects, 38 9

with LBP and 15 healthy participants aged between 18 and 65 were included (Table 1). 10

Subjects with LBP were considered eligible and included if they had had LBP for at least 11

two weeks and estimated their pain level to > 20 on a visual analogue scale (VAS 0-100). 12

[23, 28] They were excluded if pain was severe and thought likely to prevent the planned 13

test procedure; if they were pregnant; had undergone back or abdominal surgery within the 14

last six months; had a neurological disease, neurological signs of nerve root compression, 15

inflammatory autoimmune disease, or a malignancy. Healthy subjects were excluded if 16

they reported any pain on the VAS or had undergone back rehabilitation within the 17

previous five years. Both patients with LBP and healthy subjects signed written informed-18

consent forms. Subjects were enrolled 2012-02-10 to 2012-06-27. 19

20

Investigators 21

The manual investigators (n = 3) were physiotherapists specialized in manual therapy for 22

13–32 years and with several years of experience assessing ADIM in patients with LBP. 23

Another investigator skilled in musculoskeletal USI performed all the USI assessments. 24

Page 8 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 10: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

9

The intra-rater reliability for the USI examination of TrA at rest and during ADIM (ICC 1

2.1) was very good (ICC 0.94) as recorded for this study. Both the manual and the USI 2

investigator were blinded to the subject´s health status at the time of testing. At inclusion, 3

prior to the clinical intervention, all subjects completed demographic and medical-history, 4

and self-assessment questionnaires on pain and function. 5

6

Insert table 1 about here 7

8

Manual assessment of ADIM 9

ADIM was explained, performed and assessed as a slow and gentle “abdominal hollowing” 10

as originally described by Richardson et al. [13] The subjects were in a supine hook-lying 11

with hips at 40–60° and knees flexed between 90-100°, and arms placed along the torso 12

(Figure 1). They were instructed to gently draw the navel and in towards the spine, to hold 13

the contraction and breathe normally. The manual investigator palpated for muscle 14

contraction medially to the spina iliaca anterior superior (SIAS). The success of a correctly 15

performed contraction of the TrA was considered to be a slowly-developed, deep tension in 16

the abdominal wall continuously for up to 10 seconds. [13] Compensatory patterns were 17

observed during the test, such as: holding the breath; backward movement of the pelvis; 18

visual or palpable contraction of the obliquus externus muscle; pressing the heels towards 19

the floor. If any of these patterns appeared, the ADIM was assessed as incorrect. [13, 29-20

31] Before the investigation, the manual investigators underwent a training session to reach 21

a consensus on how to judge whether the manually-assessed ADIM was correct. 22

23

24

Page 9 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 11: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

10

Insert figure 1 about here 1

2

Ultrasonography assessment of ADIM and data processing 3

Ultrasound images were recorded with the Prosonix 600 DCPD Swissmed (HP 4

Medical GMBH, Wollerau, Switzerland), and motions were registered using a linear-5

array transducer (5-10 MHz). Using B (brightness) mode ultrasound, the transducer 6

was positioned across the abdominal wall at a point between the inferior angle of the 7

rib cage and the iliac crest, approximately 100 mm from the umbilicus. [32-33] On-8

screen callipers measured the distance between the inner edge of the fascial lines of 9

the TrA, the obliquus internus (OI), and the obliquus externus (OE) muscles. The 10

subject could not see the computer screen. Three pictures during rest and ADIM were 11

recorded for each body side. The images of muscles at rest were collected at the end of 12

exhalation, and images during ADIM were collected when the subject had reached the 13

expected peak of activity (Figure 2). 14

15

Insert figure 2 about here 16

17

Mean values of the measurements from the ADIM and resting position were 18

calculated for TrA, OI, and OE (Table 2). From these values, three ratios were 19

calculated: (1) The TrA contraction ratio = TrA thickness contracted/TrA thickness at 20

rest, (2) The TrA preferential activation ratio= relative change in proportion of TrA 21

relative to total lateral abdominal muscle thickness= (TrA contracted/ TrA+OE+OI 22

contracted)-(TrA at rest/TrA+OE?+OI at rest) and (3) OI + OE contraction ratio = 23

OE+OI thickness contracted/OE+OI thickness at rest . [14] 24

25

Page 10 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 12: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

11

The contraction ratio represents the relative change in muscle thickness as a result of 1

muscle contraction, and the preferential activation ratio, the change in the proportion 2

of the TrA relative to the total lateral abdominal muscle thickness. [14] 3

4

Insert table 2 about here 5

6

The mean ratio value minus one standard deviation (-1 SD) was selected as the cut-off 7

to define the preferential activation of TrA.[34] In assessing whether the ADIM, i.e. 8

the preferential activation of TrA, was correctly performed all individual ratios had to 9

exceed this defined cut-off to be accepted as passed (table 3). [34-35] 10

11

12

Insert table 3 about here 13

14

15

Experimental setup 16

All included subjects were assigned by lot to start with either the manual or the USI 17

assessment of the ADIM. They were briefed by the first investigator (either manual or 18

USI) about the purpose of the ADIM and were instructed how to perform a correct 19

ADIM. Following this, each subject was asked to perform an ADIM lasting up to 10 20

seconds without feedback; this was repeated three times. The manual investigator filled 21

out a paper chart with yes or no; yes indicating a correct contraction. If two of the three 22

performed ADIMs were assessed as a correct contraction; the test was noted and filled 23

out as correct. The subject rested for 10 minutes between the manual assessments and 24

the USI assessment. To investigate the inter-observer reliability of the manually-25

Page 11 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 13: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

12

assessed ADIM, a second investigator (M2) observed and assessed the ADIM ten 1

minutes following the first assessment (M1). Both investigators were blinded to each 2

other’s results. 3

4

Statistical analysis 5

For the concurrent validity analyses, Spearman’s rank correlation coefficient (r) was 6

used to calculate associations between the results of the manual test and the USI 7

investigation of the preferential activation of the TrA. The coefficient (r) was defined 8

as follows: 0.00–0.25 = little or no correlation; 0.26–0.49 = low correlation; 0.50–9

0.69 = moderate correlation; 0.70–0.89 = high correlation; and 0.9–1.0 = very high 10

correlation.[22, 36] 11

12

Discriminative validity was analysed by calculating sensitivity and specificity, with a 13

95% confidence interval for both patients with LBP and healthy subjects. We 14

considered at least 70% sensitivity and 50% specificity to be valid.[36] 15

16

The inter-observer reliability of the manual test was calculated using kappa (κ), and 17

was presented with 95% CI, percentage agreement (PA), and standard error (SE). 18

We defined the strength of the kappa coefficient according to Landis and Koch: 0 = 19

poor, 0.01–0.2 = slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 20

substantial and 0.81–1 = almost perfect.[37] 21

22

23

Page 12 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 14: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

13

RESULTS 1

Fifty-three subjects were allocated to the study. Their characteristics are presented in 2

Table 1. All the subjects invited to participate agreed and were enrolled. One subject 3

with LBP was excluded owing to un-interpretable images. Fifty-two subjects (LBP n = 4

37, healthy participants n = 15) were thus included in the analyses. The subjects with 5

LBP had moderate pain (Visual analogue Scale; median 34, range 21-50) and low-to-6

moderate disability (Roland Morris Disability Score; median 7, range 3-14). There was 7

no difference between subjects with LBP and healthy subjects regarding age, height and 8

BMI. Sixty-eight percent of the total sample was women. 9

10

Concurrent validity 11

Table 4 presents the associations between the manually-assessed ADIM and the USI-12

measured TrA contraction ratio (TrA-CR) and the TrA preferential ratio (TrA-Pref. Ratio). 13

None of the values for association were significant. 14

15

The associations between the manually-assessed ADIM and the USI-measured TrA 16

contraction ratio (TrA-CR) varied and were overall low (r=-0.38-0.1) as were the results 17

for the manually-assessed ADIM and the TrA preferential activation ratio (TrA-Pref. 18

Ratio) ( r = -0.19-0.53). 19

20

Insert table 4 about here 21

22

A subgroup analysis of the subjects presenting with LBP (n=37) was also performed. 23

Here, the agreement between the manually-assessed ADIM and the USI-measured TrA-24

CR and TrA-Pref. ratio also varied (r= 0.09-0.17). 25

Page 13 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 15: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

14

1

Discriminative validity 2

The discriminative validity of the manually assessed ADIM—that is, the ability of the 3

test to discriminate between subjects with LBP and healthy ones—showed a sensitivity 4

of 0.30 and a specificity of 0.73. Regarding USI measurement of the ADIM the 5

discriminative validity for TrA-CR showed a sensitivity of 0.35 and a specificity of 0.80 6

while the TrA preferential ratio showed a sensitivity of 0.41 and a specificity of 0.8 7

(Table 5). 8

9

Reliability 10

The inter-observer reliability of the ADIM manually assessed by two of the assessors 11

(M1 and M2), revealed substantial agreement of 87.5%, (κ = 0.71, CI (95%) 0.41-12

1.00) while the standard error was 0.15. 13

14

Insert table 5 about here 15

16

DISCUSSION 17

The present aim was to establish the concurrent and discriminative validity, and the inter-18

observer reliability, of the manually-assessed abdominal-draw-in maneuver (ADIM). The 19

ADIM is a clinical tool for assessing preferential activity of the deep abdominal muscles in 20

subjects with LBP. 21

22

Page 14 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 16: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

15

Our results showed poor concurrent validity, meaning weak associations between the 1

manually-assessed ADIM and the two calculated ratios of muscle thickness from the USI 2

measurements. In addition to the low agreement, the results varied widely (TrA-CR r = -3

0.38-0.1; TrA-Pref. Ratio r = -0.19-0.53, respectively). Further, no concurrent correlation 4

between the manually-assessed test and the USI measurements emerged when computed 5

only for the group of subjects with LBP. The validity of the manually-assessed ADIM to 6

discriminate those with LBP from healthy participants only reached low sensitivity 7

regarding both contraction ratio and preferential ratio. The inter-observer reliability of the 8

manual assessment of ADIM in the present study was, however, substantial (κ = 0.71). 9

10

Clearly, the manually assessed ADIM failed to agree with the USI measurements, either 11

against TrA contraction ratio or against the TrA preferential activation ratio. Reasons for 12

the poor concurrent correlation may be the complexity involved in determining an 13

isolated preferential contraction of a specific muscle when the activity seemingly appears 14

in several muscles. Urquhart et al [30] suggested that there is often a co-activation between 15

TrA and OI as both muscles have to stabilize the lumbar segments. This is also supported 16

by that the fact that the OI has both a deeper and a superficial layer.[30-32] Further, the 17

anatomical position where the lateral abdominal muscles overlie one another contributes to 18

the difficulty in palpating for an isolated TrA contraction.[34-35] This raises the question 19

of what this test actually adds, or what it actually measures since the inter-observer 20

reliability was substantial 21

22

Our results revealed good specificity to detect individuals presenting with a correct test 23

among all subjects. However, the results showed lower sensitivity, i.e. the preferential 24

Page 15 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 17: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

16

activation, assessed both manually and with USI, did not meet the fundamental criterion of 1

discriminating between subjects with LBP and healthy subjects. Again, the question arises 2

of whether there is a clinical use for the manually assessed ADIM to detect impaired 3

preferential activation of TrA in subjects with LBP. It must be taken into consideration if 4

the test is to be used to identifying functional impairments in patients under consideration 5

for rehabilitation. Our results are supported by Pulkovsky et al [34], who reported 6

inconsistent outcomes in USI-measured preferential TrA activity in LBP and healthy 7

subjects, concluding that the test was unable to differentiate between the two groups. The 8

same results have been presented by other authors [34, 37-38] who also conclude that the 9

values of associations between healthy participants and subjects with LBP measured with 10

USI are highly variable and no association can be reached between groups during ADIM-11

USI. Even so, manual assessment of ADIM is still commonly used in the clinic to 12

differentiate between patients for guiding specific exercises to re-train motor control. On 13

the present results, this can be questioned since the ability to perform seems simply to be a 14

normal variation independent of LBP. 15

16

Ferreira et al [38] reported improved TrA muscle activity measured with ultrasound 17

following an intervention with specific motor-control exercises to activate the deeper 18

abdominals in people with a poorer ability to preferentially contract TrA. [41] However, 19

only low associations between improved functional level and the activity of the deep 20

abdominals were reported. Moreover, recent studies report inconsistent relationships 21

between alleviation of pain and disability and changes in activity or muscle thickness in 22

TrA as measured with USI B or M mode. Here the subjects were people with LBP 23

following interventions with specific motor control exercises.[15, 39] 24

25

Page 16 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 18: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

17

The values obtained from the muscles measured with USI itself were consistent with 1

previous findings. [14, 40-41] Small differences in mean values for anatomical distances 2

appeared at rest and during measured muscle contraction. [18, 33, 35, 42] Our ratio values 3

were similar to or slightly higher than those previously reported, a phenomenon that 4

probably relates to basic data, especially the proportionally smaller value from the TrA at 5

rest.[14] 6

7

8

Strengths and limitations 9

The strength of the present study is the aim to investigate both concurrent and 10

discriminative validity, and also the reliability, of a clinical tool commonly used in the 11

rehabilitation of patients with LBP when prescribing specific exercises. A further strength 12

is the study design with investigators blinded to all subjects’ health status. Pulkovsky et al 13

[34] studied the concurrent validity of the ADIM using ultrasound and a blinded design. 14

They did not, however, investigate the manually-assessed ADIM. To our knowledge no 15

other study has performed a single blinded study of this maneuver. 16

17

While the use of USI as the criterion was believed to strengthen the results on validity, 18

the gold-standard method itself showed low discriminative validity. Indeed, a possible 19

limitation is that USI B mode was used instead of M mode in the assessment, suggesting 20

the possibility that the peak of TrA activity during the ADIM was missed. Hodges and 21

colleagues [18] proposed that a change in muscle dimension does not always correlate to 22

increased muscle activity, depending on muscle length, muscle extensibility, pennation 23

pattern and competing forces from adjacent muscles. The use of M mode with tissue 24

Page 17 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 19: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

18

Doppler imaging (TDI) may therefore be more advantageous than the use of B mode as 1

the images are sampled over time.[34] This was however not feasible in the present 2

study but should be tried in a future one. 3

Whittaker et al [38] suggest that USI B-mode may be used solely as a source of evaluation of 4

muscle activity and understanding changes in muscle thickness and neuromuscular control. 5

However USI-B mode still plays an important role in detecting altered abdominal muscle 6

function in subjects with LBP and affords helpful biofeedback for abdominal muscle function 7

when evaluating treatment programs. 8

For the selected cut-off value of the ultrasound measurements of TrA contraction and 9

preferential ratio used in the calculations of concurrent validity, we used cut-offs 10

previously defined for investigating the ability of the TrA contraction ratio to discriminate 11

between subjects with chronic LBP and healthy participants. [14, 34] We are aware that a 12

cut-off value will affect the outcome, and different cut-offs for these ratios may therefore 13

be discussed and further investigated. 14

15

For strengthening internal validity, we presented the USI investigators’ intra-rater 16

reliability (ICC 0.94), which is consistent with previously reported results of intra-rater 17

reliability when measuring TrA muscle thickness with USI (ICC = 0.92-0.99). [16] The 18

present inter-observer reliability of the manual assessment of ADIM was substantial (κ = 19

0.71), a finding that aligns with a previous study (κ = 0.52). [25] 20

21

Our main aim was to investigate whether the manually-assessed ADIM is a valid method 22

for assessing a preferential activation of the deep abdominal muscles. We did this by 23

comparing the manually-assessed test (ADIM) results to the USI findings for the same 24

action. Through blinding of the investigators by including both healthy subjects and 25

Page 18 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 20: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

19

subjects with LBP, we aimed to minimize bias and thus strengthen the results. However, 1

this led to a disproportionately large sample of subjects with LBP and this may be 2

considered a limitation to the study. 3

4

It is imperative that clinical tests are studied for both validity and reliability. Therefore the 5

manually-assessed ADIM should be further studied in order to afford clinicians better 6

directions as to whether this clinical tool really is valid for detecting correct preferential 7

activity of TrA in persons presenting with LBP. 8

9

CONCLUSION 10

Our study is the first to consider the concurrent and discriminative validity of the 11

manually-assessed abdominal draw-in maneuver (ADIM), a commonly used clinical tool. 12

Although it´s inter-observer reliability was high, both the concurrent and discriminative 13

validity was low for examining preferential activity of the deep abdominal muscles. 14

Neither the manually-assessed ADIM nor the ultrasound testing discriminated between 15

subjects with low-back pain and healthy subjects regarding preferential activity of the 16

transversus muscle. This was because this ability is excited also in healthy subjects. From 17

the present results, it thus appears that the “preferential” activity cannot be linked to 18

subjects with low-back pain, and this leads us to question the test’s utility in the clinic for 19

addressing impaired activation of the deep abdominals. 20

21

Acknowledgements 22

Financial support from the Swedish Confederation of Physiotherapists-Memorial Fund is 23

gratefully acknowledged. 24

Page 19 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 21: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

20

1

Authors’ contributions 2

All authors participated in the conception and design of the study, and all made substantial 3

contributions to the acquisition of data. BOÄ and ERB performed the statistical analysis. 4

All authors participated in the interpretation of data. KK drafted the manuscript and BOÄ 5

and ERB critically revised the manuscript for important intellectual content. All the 6

authors read and approved the final manuscript 7

8

Funding 9

KK works as a salaried private practitioner, BOÄ and ERB receive their salaries from the 10

Karolinska Institutet 11

12

Competing interests 13

None. 14

15

Ethics approval 16

The regional ethical review board in Stockholm, Sweden, approved the study (Dnr. 17

2011/1859–31). 18

License for Publication 19

The corresponding author is entitled to grant on behalf of all the authors, and does grant on 20

behalf of all the authors, an exclusive license (or non-exclusive for government) on a 21

worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be 22

published in any BMJPGL products and sublicenses such use and exploitation of all 23

Page 20 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 22: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

21

subsidiary rights, as set out in our license. 1

(http://group.bmj.com/products/journals/instructions-for-authors/licence-forms) 2

Competing Interest: None declared. 3

4

Provenance and peer review: not commissioned; externally peer reviewed. 5

6

Data sharing statement: additional data are available from Karsten Kaping, 7

[email protected] 8

9

10

11

12

13

14

REFERENCES 15

1. Vleeming, A., et al., The functional coupling of the deep abdominal and paraspinal muscles: 16

the effects of simulated paraspinal muscle contraction on force transfer to the middle and 17

posterior layer of the thoracolumbar fascia. J Anat, 2014. 225(4): p. 447-62. 18

2. Panjabi, M.M., The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, 19

and enhancement. J Spinal Disord, 1992. 5(4): p. 383-9; discussion 397. 20

3. Panjabi, M.M., The stabilizing system of the spine. Part II. Neutral zone and instability 21

hypothesis. J Spinal Disord, 1992. 5(4): p. 390-6; discussion 397. 22

4. Panjabi, M.M., A hypothesis of chronic back pain: ligament subfailure injuries lead to muscle 23

control dysfunction. Eur Spine J, 2006. 15(5): p. 668-76. 24

5. Hodges, P.W., Changes in motor planning of feedforward postural responses of the trunk 25

muscles in low back pain. Exp Brain Res, 2001. 141(2): p. 261-6. 26

6. Hodges, P.W. and G.L. Moseley, Pain and motor control of the lumbopelvic region: effect and 27

possible mechanisms. J Electromyogr Kinesiol, 2003. 13(4): p. 361-70. 28

7. Hodges, P. and G. Jull, Does strengthening the abdominal muscles prevent low back pain? J 29

Rheumatol, 2000. 27(9): p. 2286-8. 30

Page 21 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 23: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

22

8. Allison, G.T., S.L. Morris, and B. Lay, Feedforward responses of transversus abdominis are 1

directionally specific and act asymmetrically: implications for core stability theories. J Orthop 2

Sports Phys Ther, 2008. 38(5): p. 228-37. 3

9. Macedo, L.G., et al., Effect of motor control exercises versus graded activity in patients with 4

chronic nonspecific low back pain: a randomized controlled trial. Phys Ther, 2012. 92(3): p. 5

363-77. 6

10. Rasmussen-Barr, E., et al., Graded exercise for recurrent low-back pain: a randomized, 7

controlled trial with 6-, 12-, and 36-month follow-ups. Spine (Phila Pa 1976), 2009. 34(3): p. 8

221-8. 9

11. Bystrom, M.G., E. Rasmussen-Barr, and W.J. Grooten, Motor control exercises reduces pain 10

and disability in chronic and recurrent low back pain: a meta-analysis. Spine (Phila Pa 1976), 11

2013. 38(6): p. E350-8. 12

12. Macedo, L.G., et al., Motor control exercise for persistent, nonspecific low back pain: a 13

systematic review. Phys Ther, 2009. 89(1): p. 9-25. 14

13. Richardson, C., Jull, G., Hodges, P. , Therapeutic Exercise for spinal segmental stabilisation in 15

low back pain. Scientific basis and clinical approach. 1st ed. 1999, United Kingdom: Churchill 16

Livingstone 17

14. Teyhen, D.S., et al., The use of ultrasound imaging of the abdominal drawing-in maneuver in 18

subjects with low back pain. J Orthop Sports Phys Ther, 2005. 35(6): p. 346-55. 19

15. Wong, A.Y., et al., Do changes in transversus abdominis and lumbar multifidus during 20

conservative treatment explain changes in clinical outcomes related to nonspecific low back 21

pain? A systematic review. J Pain, 2014. 15(4): p. 377 e1-35. 22

16. Ferreira, P.H., et al., Discriminative and reliability analyses of ultrasound measurement of 23

abdominal muscles recruitment. Man Ther, 2011. 16(5): p. 463-9. 24

17. Hides, J., et al., An MRI investigation into the function of the transversus abdominis muscle 25

during "drawing-in" of the abdominal wall. Spine (Phila Pa 1976), 2006. 31(6): p. E175-8. 26

18. Hodges, P.W., et al., Measurement of muscle contraction with ultrasound imaging. Muscle 27

Nerve, 2003. 27(6): p. 682-92. 28

19. Koppenhaver, S.L., et al., Reliability of rehabilitative ultrasound imaging of the transversus 29

abdominis and lumbar multifidus muscles. Arch Phys Med Rehabil, 2009. 90(1): p. 87-94. 30

20. Park, B.H., et al., Reliability and Validity of the Korean Version VISA-P Questionnaire for 31

Patellar Tendinopathy in Adolescent Elite Volleyball Athletes. Ann Rehabil Med, 2013. 37(5): 32

p. 698-705. 33

21. Ghamkhar, L., et al., Application of rehabilitative ultrasound in the assessment of low back 34

pain: a literature review. J Bodyw Mov Ther, 2011. 15(4): p. 465-77. 35

22. Carter RE, L.J., Domholdt E. , ed. Rehabilitation research:prinicipals and applications. 4th ed. 36

ed. 2011, Elsevier Saunders: St.Louis Missouri. 37

23. Ford, J.D., et al., Development and validation of a brief mental health screening instrument 38

for newly incarcerated adults. Assessment, 2007. 14(3): p. 279-99. 39

24. Streiner DL, N.G., ed. Health Measurement Scales. A practical guide to their development and 40

use. 3rd ed. 2003, Oxford University Press: Oxford. 41

25. Costa, L.O., et al., Short report: intra-tester reliability of two clinical tests of transversus 42

abdominis muscle recruitment. Physiother Res Int, 2006. 11(1): p. 48-50. 43

26. Altman, D.G., ed. Practical statistics for medical research. 1991, Chapman and Hall: London. 44

27. Norman GR, S.D., ed. The bare essentials. 2008, People's medical publishing house: Shelton, 45

CT. 46

28. Stanton, T.R., et al., After an episode of acute low back pain, recurrence is unpredictable and 47

not as common as previously thought. Spine (Phila Pa 1976), 2008. 33(26): p. 2923-8. 48

Page 22 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 24: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

23

29. Henry, S.M. and K.C. Westervelt, The use of real-time ultrasound feedback in teaching 1

abdominal hollowing exercises to healthy subjects. J Orthop Sports Phys Ther, 2005. 35(6): p. 2

338-45. 3

30. Urquhart, D.M., et al., Abdominal muscle recruitment during a range of voluntary exercises. 4

Man Ther, 2005. 10(2): p. 144-53. 5

31. Urquhart, D.M., et al., Regional morphology of the transversus abdominis and obliquus 6

internus and externus abdominis muscles. Clin Biomech (Bristol, Avon), 2005. 20(3): p. 233-7

41. 8

32. Costa, L.O., et al., An investigation of the reproducibility of ultrasound measures of abdominal 9

muscle activation in patients with chronic non-specific low back pain. Eur Spine J, 2009. 18(7): 10

p. 1059-65. 11

33. Westad, C., P.J. Mork, and O. Vasseljen, Location and sequence of muscle onset in deep 12

abdominal muscles measured by different modes of ultrasound imaging. J Electromyogr 13

Kinesiol, 2010. 20(5): p. 994-9. 14

34. Pulkovski, N., et al., Ultrasound assessment of transversus abdominis muscle contraction ratio 15

during abdominal hollowing: a useful tool to distinguish between patients with chronic low 16

back pain and healthy controls? Eur Spine J, 2012. 21 Suppl 6: p. S750-9. 17

35. Mannion, A.F., et al., Abdominal muscle size and symmetry at rest and during abdominal 18

hollowing exercises in healthy control subjects. J Anat, 2008. 213(2): p. 173-82. 19

36. Turner, D., et al., Using the entire cohort in the receiver operating characteristic analysis 20

maximizes precision of the minimal important difference. J Clin Epidemiol, 2009. 62(4): p. 21

374-9. 22

37. Landis, J.R. and G.G. Koch, An application of hierarchical kappa-type statistics in the 23

assessment of majority agreement among multiple observers. Biometrics, 1977. 33(2): p. 24

363-74. 25

38. Ferreira, P.H., et al., Changes in recruitment of transversus abdominis correlate with disability 26

in people with chronic low back pain. Br J Sports Med, 2010. 44(16): p. 1166-72. 27

39. Wong, A.Y., et al., Do various baseline characteristics of transversus abdominis and lumbar 28

multifidus predict clinical outcomes in nonspecific low back pain? A systematic review. Pain, 29

2013. 154(12): p. 2589-602. 30

40. Whittaker, J.L., et al., Association between changes in electromyographic signal amplitude 31

and abdominal muscle thickness in individuals with and without lumbopelvic pain. J Orthop 32

Sports Phys Ther, 2013. 43(7): p. 466-77. 33

41. Beazell, J.R., et al., Changes in lateral abdominal muscle thickness during an abdominal 34

drawing-in maneuver in individuals with and without low back pain. Res Sports Med, 2011. 35

19(4): p. 271-82. 36

42. Rankin, G., M. Stokes, and D.J. Newham, Abdominal muscle size and symmetry in normal 37

subjects. Muscle Nerve, 2006. 34(3): p. 320-6. 38

39

40

41

42

43

Page 23 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 25: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

24

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Page 24 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 26: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

25

Table 1 Demographic and self-assessment questionnaire data for all subjects included (n=52)

Total

(n=52)

Healthy

(n=15)

LBP

(n= 37)

Female

(n=36)

Male

(n=16)

Gender (female/male) 36/16 12/3 24/13

Age (yrs.) (SD) 42 (5.82) 41.5 (10.71) 42.4 (6.97) 42 (7.01) 42 (10.71)

Height (m) (SD) 1.72 (0.24) 1.69 (0.44) 1.73 (0.28) 1.68 (0.18) 1.83 (0.32)

Weight (kg) (SD) 72.2 (10.01) 68.6 (17.71) 73.6 (19.00) 65.5 (10.9) 87.1 (22.4)

BMI (SD) 24 (3.33) 23.5 (6.06) 24.3 (3.99) 23.2 (3.84) 26.1 (6.60)

VAS (0-100)* 34 (21-50)

Roland Morris (0-24)* 7 (3-14)

Oswestry (0-100)* 22 (14-28)

Previous seeking

treatment for low back

pain 5 years back (?*) (% of group)

65 0 65

Subjects’ work status

(% of group)

92 100 89

Pain duration

∟1-4 weeks (n) 3

∟1-3 months (n) 3

∟3 > months (n) 31

First-time episode LBP

n (% of group) 5 (14)

Recurrent LBP

n (% of groups) 32 (86)

SD=standard deviation, BMI=body mass index, VAS=visual analogue scale, *data

presented as median (25th /75th percentiles).

Page 25 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 27: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

26

1

2

3

4

5

6

Figure 1. Test position for abdominal-drawing-in-maneuver (ADIM) 7

8

9

10

11

12

A B 13

Figure 2. Ultrasound image of abdominal muscles, A= rest B=activity during 14

abdominal-draw-in-maneuver (ADIM). Distance 1 represents M. transversus 15

abdominis (TrA) distance 2 M. obliquus internus (OI) and distance 3 M. obliquus 16

externus (OE). Note the increase in TrA, OI and the decrease in OE during the 17

ADIM, picture B. 18

Page 26 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 28: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

27

1

2

3

Table 2. Data from the USI measurements of muscle thickness (mm) at rest and

during preferential activation of TrA, presented with mean and SD (n=52).

At rest

(n=52)

Preferential

contraction (n=52)

Change

(%)

TrA Left 2.28 (0.96) 4.41 (0.80) 93.4

TrA Right 2.88 (0.93) 4.49 (0.79) 55.9

TrA total abdomen 2.78 (0.93) 4.45 (0.79) 60.0

OI Left 8.02 (0.49) 9.01 (0.76) 12.3

OI Right 8.01 (0.49) 8.50 (0.63) 6.1

OI total abdomen 8.02 (0.49) 8.75 (0.69) 9.2

OE Left 5.91 (0.58) 5.53 (0.65) -6.4

OE Right 5.68 (0.62) 5.71 (0.62) 0.5

OE total abdomen 5.80 (0.60) 5.62 (0.63) -3.1

USI=ultrasound imaging, TrA= m. transversus abdominis, OI= m. obliquus internus,

OE= m. obliquus externus, SD= Standard deviation, % change= percentage change

from resting to activated state during preferential activation.

4

5

6

7

8

9

10

11

12

Page 27 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 29: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

28

1

2

3

Table 3. Calculated ratios from ultrasound imaging of lateral abdominal muscles

during preferential activation of TrA

Contraction ratios All

n=52

Healthy

n=15

LBP

n=37

TrA Contraction ratio,

mean (SD)

1.69 (0.18)

∟Left 1.71 (0.24) 1.75 (0.45) 1.69 (0.28)

∟Right 1.67 (0.23) 1.76 (0.45) 1.63 (0.27)

TrA Preferential ratio, mean (SD)

0.07 (0.01)

∟Left 0.07 (0.04) 0.072 (0.019) 0.076 (0.019)

∟Right 0.07 (0.04) 0.066 (0.003) 0.076 (0.019)

OE+OI Contraction ratio, mean (SD)

1.05 (0.14)

∟Left 1.047 (0.15) 1.064 (0.27) 1.04 (0.17)

∟Right 1.048 (0.15) 1.057 (0.27) 1.04 (0.17)

TrA=m. transversus abdominis, OI= m. obliquus internus, OE= m. obliquus externus Ratio (SD)=calculated ratio with standard deviation

4

5

6

7

8

9

Page 28 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 30: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

29

1

2

3

4

Table 4. Associations between manually-assessed ADIM and USI-assessed, expressed

in TrA-CR ratio and TrA preferential ratio in all subject (n=52) and in subjects by

investigator.

TrA-CR

(r)

TrA Pref. Ratio

(r)

All investigators (n= 52) .09 .12

Investigator 1 (n=30) 0.1 .06

Investigator 2 (n=13) .03 .53

Investigator 3 (n=9) -.38 -.19

TrA-CR Ratio = transversus abdominis contraction ratio, TrA-Pref. Ratio = transversus abdominis preferential activation ratio. Bold represents good agreement.

5

Table 5 Sensitivity and specificity for the USI-measured TrA contractions ratio

(TrA-CR) and TrA preferential ratios (TrA-Pref. Ratio) and manually-assessed

ADIM; presented with 95% CI.

Sensitivity

(95% CI)

Specificity

(95%CI)

TrA Contraction Ratio 0.35 (0.20-0.53) 0.80 (0.51-0.95)

TrA Preferential Ratio 0.41 (0.25-0.58) 0.80 (0.51-0.95)

ADIM-Manual 0.30 (0.16-0.47) 0.73 (0.44-0.91)

ADIM= Abdominal drawing-in maneuver , CI = 95% confidence interval , TrA

= M. transversus abdominis, USI = Ultrasound imaging

6

7

Page 29 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 31: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

30

1

Page 30 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 32: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

STARD checklist for reporting of studies of diagnostic accuracy

(version January 2003)

Section and Topic Item

#

On page #

TITLE/ABSTRACT/

KEYWORDS

1 Identify the article as a study of diagnostic accuracy (recommend MeSH

heading 'sensitivity and specificity').

2

INTRODUCTION 2 State the research questions or study aims, such as estimating diagnostic

accuracy or comparing accuracy between tests or across participant

groups.

5

METHODS

Participants 3 The study population: The inclusion and exclusion criteria, setting and

locations where data were collected.

6

4 Participant recruitment: Was recruitment based on presenting symptoms,

results from previous tests, or the fact that the participants had received

the index tests or the reference standard?

6

5 Participant sampling: Was the study population a consecutive series of

participants defined by the selection criteria in item 3 and 4? If not,

specify how participants were further selected.

6

6 Data collection: Was data collection planned before the index test and

reference standard were performed (prospective study) or after

(retrospective study)?

7-9

Test methods 7 The reference standard and its rationale. 7-9

8 Technical specifications of material and methods involved including how

and when measurements were taken, and/or cite references for index

tests and reference standard.

7-9

9 Definition of and rationale for the units, cut-offs and/or categories of the

results of the index tests and the reference standard.

9

10 The number, training and expertise of the persons executing and reading

the index tests and the reference standard.

6-7

11 Whether or not the readers of the index tests and reference standard

were blind (masked) to the results of the other test and describe any

other clinical information available to the readers.

7

Statistical methods 12 Methods for calculating or comparing measures of diagnostic accuracy,

and the statistical methods used to quantify uncertainty (e.g. 95%

confidence intervals).

10

13 Methods for calculating test reproducibility, if done. 7

RESULTS

Participants 14 When study was performed, including beginning and end dates of

recruitment.

15 Clinical and demographic characteristics of the study population (at least

information on age, gender, spectrum of presenting symptoms).

6

16 The number of participants satisfying the criteria for inclusion who did or

did not undergo the index tests and/or the reference standard; describe

why participants failed to undergo either test (a flow diagram is strongly

recommended).

11

Test results 17 Time-interval between the index tests and the reference standard, and

any treatment administered in between.

9

18 Distribution of severity of disease (define criteria) in those with the target

condition; other diagnoses in participants without the target condition.

6

19 A cross tabulation of the results of the index tests (including

indeterminate and missing results) by the results of the reference

standard; for continuous results, the distribution of the test results by the

results of the reference standard.

11-12

20 Any adverse events from performing the index tests or the reference

standard.

11

Estimates 21 Estimates of diagnostic accuracy and measures of statistical uncertainty

(e.g. 95% confidence intervals).

12

22 How indeterminate results, missing data and outliers of the index tests

were handled.

11

23 Estimates of variability of diagnostic accuracy between subgroups of

participants, readers or centers, if done.

11

24 Estimates of test reproducibility, if done. 11-12

DISCUSSION 25 Discuss the clinical applicability of the study findings. 13-14

Page 31 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 33: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

The abdominal drawing-in maneuver for detecting activity in the deep abdominal muscles: is this clinical tool reliable

and valid?

Journal: BMJ Open

Manuscript ID bmjopen-2015-008711.R1

Article Type: Research

Date Submitted by the Author: 17-Sep-2015

Complete List of Authors: Kaping, Karsten; NVS-Sjukgymnastik, Karolinska Institutet; Rehabtjanst Äng, Björn; NVS-Sjukgymnastik, Karolinska Institutet Rasmussen-Barr, Eva; NVS-Sjukgymnastik, Karolinska Institutet

<b>Primary Subject Heading</b>:

Rehabilitation medicine

Secondary Subject Heading: Sports and exercise medicine, Radiology and imaging, Diagnostics, Rehabilitation medicine

Keywords:

Back pain < ORTHOPAEDIC & TRAUMA SURGERY, Musculoskeletal disorders < ORTHOPAEDIC & TRAUMA SURGERY, Orthopaedic sports trauma < ORTHOPAEDIC & TRAUMA SURGERY, Spine < ORTHOPAEDIC & TRAUMA SURGERY, REHABILITATION MEDICINE, ULTRASONOGRAPHY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open on June 23, 2020 by guest. P

rotected by copyright.http://bm

jopen.bmj.com

/B

MJ O

pen: first published as 10.1136/bmjopen-2015-008711 on 9 D

ecember 2015. D

ownloaded from

Page 34: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

1

The abdominal drawing-in maneuver for detecting activity in the

deep abdominal muscles: is this clinical tool reliable and valid?

Page 1 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 35: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

2

The abdominal drawing-in maneuver for detecting activity in the

deep abdominal muscles: is this clinical tool reliable and valid?

Karsten Kaping, RPT, MT, MSc *,1,2 Björn O. Äng, RPT, Associate Professor, 1

Eva Rasmussen-Barr. RPT, PhD 1,2

Address: 1 Department of Neurobiology, Care Sciences and Society, Division of

Physiotherapy, Karolinska Institutet, Alfred Nobels Allé 12, 141 83 Huddinge

Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden

Telephone: +46-8-441 16 60

Email: Karsten Kaping* [email protected];

[email protected]; [email protected]

*Corresponding author

Keywords:

Abdominal muscles, concurrent validity, lumbar spine, transversus abdominis, ultrasound

imaging

Word count: 3561

Page 2 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 36: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

3

ABSTRACT (word count 288)

Objective: The abdominal drawing-in maneuver (ADIM) is a common clinical tool for

manually assessing as “correct” or not a preferential activation of the deep abdominal muscles

in patients with low-back pain (LBP). The validity and reliability of manual assessment of the

ADIM are, however, as yet unknown. This study evaluated the concurrent and discriminative

validity and the reliability of the manually assessed ADIM.

Design: Single-blinded cross-sectional study

Settings: General population in Stockholm County, Sweden

Participants: The study sample comprised 38 subjects seeking care for low-back pain and 15

healthy subjects.

Measures: The manual ADIM was assessed as correct or not following a standard procedure.

Ultrasound imaging (USI) was used as the concurrent reference (gold standard) for the

manually assessed ADIM by calculating a ratio of the change in muscle thickness between the

resting and the contracted states: the correlation between manual test and the USI was

calculated. Discriminative validity was analysed by calculating sensitivity and specificity. A

sample of 24 subjects was analysed with kappa coefficients for inter-observer reliability

between two raters.

Results: The concurrent validity between the manual ADIM and the ADIM-USI ratios

showed poor correlations (r=.13-.40). The discriminative validity of the manually assessed

ADIM to predict LBP showed a sensitivity/specificity of 0.30/0.73, while the ADIM-USI

ratio to predict LBP showed 0.19/0.87. The inter-observer reliability for the manually-

assessed ADIM revealed substantial agreement: (K= 0.71, CI (95%) 0.41-1.00.

Conclusion: Although the inter-observer reliability of the manually-assessed ADIM was

high, both the concurrent and discriminative validity was low for examining the preferential

activity of the deep abdominal muscles. Neither the manually-assessed ADIM nor the

ultrasound testing discriminated between subjects with LBP and healthy subjects regarding

preferential activity of the transversus muscle as this ability/inability was also present in

healthy subjects.

Page 3 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 37: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

4

STRENGTHS AND LIMITATIONS

Strengths and limitations of this study

Strengths and limitations of this study

•Strength of this study is that the abdominal-drawing-in-manoeuvre was analysed for both

concurrent and discriminative validity and that the investigators were blinded to all subjects’

health status.

•Strength is also that the study used a gold-standard tool (ultrasound) to measure the quantity

of muscle activity for comparison with the manual estimation.

•Possible limitation is that our study design resulted in a disproportionate sample between

healthy subjects and those with low-back pain.

•In addition, the patients estimated their low-back pain only as low-to-moderate

Page 4 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 38: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

5

INTRODUCTION (word count 3388)

The lumbar spine depends on complex interaction between passive and active structures such

as contracting muscles, ligaments and fascia to maintain optimal spinal control during bodily

movement and loading [1-3]. Contraction of the lateral abdominal muscles, and especially the

anticipatory activity of the transversus abdominus muscle (TrA), is suggested to be a key

motor-control element for transferring spinal loads to the pelvis and upper trunk, thus

avoiding injury to spinal tissue or pain incidents following sub-failure loading. [2-5]

Studies have reported delayed anticipatory activity of TrA in low-back-pain sufferers (LBP)

in comparison to healthy subjects [5-7], although such differences in temporal activity have

also been questioned. [8] Specific exercises have been proposed for re-training motor control

of the abdominal muscles; a popular and common treatment for people with recurrent and

chronic LBP.[9-10] Following such interventions, recent systematic reviews have reported

improvement of symptoms such as pain and functional deficits compared to other exercises or

treatment.[11-12]

To assess the activity of the deep abdominal muscles a specific clinical test is available: the

abdominal drawing-in maneuver (ADIM). [13] Through manual palpation and visual

inspection, the ADIM is used to assess the success or not in performing a voluntary and

preferential contraction of the deep abdominal muscles, especially the TrA [13-14], in order to

target interventions with specific exercises and follow-up progression.

In addition, ultrasound imaging (USI) has for several years been used in clinical settings to

quantify the activity of the deep abdominal muscles. Specifically, it has also been used in

research to quantify change in muscle thickness or muscle onset of TrA before and after

interventions using motor control exercises.[15] The USI has been validated for measuring

Page 5 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 39: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

6

change in muscle thickness against electromyography (EMG) [16] and magnetic resonance

imaging ( MRI) [17] and demonstrates high inter-observer reliability in both healthy people

and subjects with LBP. [14, 18-21] However, not all clinicians have access to USI and the

manually assessed ADIM therefore remains the most common clinical tool for addressing and

examining a preferential contraction of the deep abdominal muscles.

A clinically convenient test should show both good reliability and high validity. Specifically,

it should be able to demonstrate concurrent validity against a reliable gold-standard

assessment (e.g., USI verifying TrA activity); proved to measure what it purports to

measure.[22] If the clinical test is to reflect impairment, not merely normal variation in

healthy subjects, the test should also exhibit empirical, discriminative validity to justify its use

among patients. [23] For the present purpose discriminative validity was defined as the extent

to which manually-assessed and USI-assessed ADIM differentiated between subjects with

LBP and those without.[24]

To date, the concurrent and discriminative validity of the manually assessed ADIM is

unknown, and only one study addresses inter-observer reliability.[25] The present aims were

therefore to assess the concurrent validity of the manually assessed ADIM in relation to the

USI during guided abdominal contractions, to assess its discriminative ability, and to

determine the inter-observer reliability.

METHODS

Study design

A cross-sectional, single-blinded design was used to evaluate the concurrent validity,

discriminating validity and inter-observer reliability of the ADIM in subjects with LBP and in

healthy subjects. A sample size of 50 subjects was considered adequate to power the study to

obtain proper correlation data as used in concurrent analysis, [26], while a sample of 27

Page 6 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 40: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

7

subjects was calculated for kappa agreement data (confidence interval 0.3; expected

percentage agreement 0.8 and chance agreement 0.5) used in reliability statistics.[27] Planned

enrolment was undertaken to meet these criteria. The regional ethical review board approved

the study (Dnr 2011/1859-31).

Study sample

Subjects with LBP were recruited through information about the study distributed by e-mail

and telephone to six private physiotherapy settings in Stockholm, Sweden. Healthy subjects

were found through written and verbal information distributed at two large companies in

Stockholm, each with approximately 700 employees. Fifty-three subjects, 38 with LBP and

15 healthy participants aged between 18 and 65 were included (Table 1). Subjects seeking

care for a primary complaint of LBP were considered eligible and included if their LBP had

lasted for at least two weeks and if they estimated their pain level to > 20 on a visual analogue

scale (VAS 0-100). [23, 28] They were excluded if pain was severe and thought likely to

prevent the planned test procedure; if they were pregnant; had undergone back or abdominal

surgery within the last six months; had a neurological disease, neurological signs of nerve root

compression, inflammatory autoimmune disease, or a malignancy. Subjects with spondylosis,

or spondylolisthesis were included if no signs of nerve root compression existed. Healthy

subjects were excluded if they reported any pain on the VAS or had undergone back

rehabilitation within the previous five years. Both patients with LBP and healthy subjects

signed written informed-consent forms. Subjects were enrolled 2012-02-10 to 2012-06-27.

Page 7 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 41: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

8

Investigators

The manual investigators (n = 3) were physiotherapists specialized in manual therapy for 13–

32 years and with several years of experience assessing ADIM in patients with LBP. Another

investigator skilled in musculoskeletal USI performed all the USI assessments. A same day

intra-rater reliability was performed and calculated from the obtained USI measurements on

35 subjects and considered very good (ICC 2.1, 0.94).

Both the manual and the USI investigator were blinded to the subject´s health status at the

time of testing. At inclusion, prior to the clinical intervention, all subjects completed

demographic and medical-history, and self-assessment questionnaires on pain and

function.

Insert table 1 about here

Manual assessment of ADIM

ADIM was explained, performed and assessed as a slow and gentle “abdominal hollowing”

as originally described by Richardson et al. [13] The subjects were in a supine hook-lying

with hips at 40–60° and knees flexed between 90-100°, and arms placed along the torso

(Figure 1). They were instructed to gently draw the navel and in towards the spine, to hold

the contraction and breathe normally. The manual investigator palpated for muscle

contraction medially to the spina iliaca anterior superior (SIAS). The success of a correctly

performed contraction of the TrA was considered to be a slowly-developed, deep tension in

the abdominal wall continuously for up to 10 seconds. [13] Compensatory patterns were

observed during the test, such as: holding the breath; backward movement of the pelvis;

visual or palpable contraction of the obliquus externus muscle; pressing the heels towards

Page 8 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 42: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

9

the floor. If any of these patterns appeared, the ADIM was assessed as incorrect. [13, 29-

31] Before the investigation, the manual investigators underwent a training session to reach

a consensus on how to judge whether the manually-assessed ADIM was correct.

Insert figure 1 about here

Ultrasonography assessment of ADIM and data processing

Ultrasound images were recorded with the Prosonix 600 DCPD Swissmed (HP

Medical GMBH, Wollerau, Switzerland), and motions were registered using a linear-

array transducer (5-10 MHz). Using B (brightness) mode ultrasound, the transducer

was positioned across the abdominal wall at a point between the inferior angle of the

rib cage and the iliac crest, approximately 100 mm from the umbilicus. [32-33] On-

screen callipers measured the distance between the inner edge of the fascial lines of

the TrA, the obliquus internus (OI), and the obliquus externus (OE) muscles. The

subject could not see the computer screen. Three pictures during rest and ADIM were

recorded for each body side. The images of muscles at rest were collected at the end of

exhalation, and images during ADIM were collected when the subject had reached the

expected peak of activity (Figure 2).

Insert figure 2 about here

Mean values of the measurements from the ADIM and resting position were

calculated for TrA, OI, and OE (Table 2). From these values, three ratios were

calculated: (1) The TrA contraction ratio = TrA thickness contracted/TrA thickness at

Page 9 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 43: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

10

rest, (2) The TrA preferential activation ratio= relative change in proportion of TrA

relative to total lateral abdominal muscle thickness= (TrA contracted/ TrA+OE+OI

contracted)-(TrA at rest/TrA+OE?+OI at rest) and (3) OI + OE contraction ratio =

OE+OI thickness contracted/OE+OI thickness at rest . [14]

The contraction ratio represents the relative change in muscle thickness as a result of

muscle contraction, and the preferential activation ratio, the change in the proportion

of the TrA relative to the total lateral abdominal muscle thickness. [14]

Insert table 2 about here

The mean ratio value minus one standard deviation (-1 SD) was selected as the cut-off

to define the preferential activation of TrA.[34] In assessing whether the ADIM, i.e.

the preferential activation of TrA, was correctly performed all individual ratios had to

exceed this defined cut-off to be accepted as passed (table 3). [34-35]

Insert table 3 about here

Experimental setup

All included subjects were assigned by lot to start with either the manual or the USI

assessment of the ADIM. They were briefed by the first investigator (either manual or USI)

about the purpose of the ADIM and were instructed how to perform a correct ADIM.

Following this, each subject was asked to perform an ADIM lasting up to 10 seconds without

Page 10 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 44: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

11

feedback; this was repeated three times. The manual investigator filled out a paper chart with

yes or no; yes indicating a correctly performed ADIM.

If two of the three performed ADIMs were assessed as a correct contraction; the test

was noted and filled out as correct. The subject rested for 10 minutes between the

manual assessments and the USI assessment. To investigate the inter-observer reliability

of the manually-assessed ADIM, a second investigator (M2) observed and assessed the

ADIM ten minutes following the first assessment (M1). Both investigators were blinded

to each other’s results.

Statistical analysis

For the concurrent validity analyses, Spearman’s rank correlation coefficient (r) was

used to calculate associations between the results of the manual test and the USI

investigation of the preferential activation of the TrA. The coefficient (r) was defined

as follows: 0.00–0.25 = little or no correlation; 0.26–0.49 = low correlation; 0.50–

0.69 = moderate correlation; 0.70–0.89 = high correlation; and 0.9–1.0 = very high

correlation.[22, 36]

Discriminative validity was analysed by calculating sensitivity and specificity, with a 95%

confidence interval for both patients with LBP and healthy subjects. We considered at least

70% sensitivity and 50% specificity to be valid.[36] A correctly or not correctly performed

ADIM was scored as “yes” or “no” both for subjects with LBP and for healthy.

Page 11 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 45: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

12

The inter-observer reliability of the manual test was calculated using kappa (κ), and was

presented with 95% CI, percentage agreement (PA), and standard error (SE). We defined

the strength of the kappa coefficient according to Landis and Koch: 0 = poor, 0.01–0.2 =

slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 substantial and 0.81–1 = almost

perfect.[37]

RESULTS

Fifty-three subjects were allocated to the study. Their characteristics are presented in

Table 1. All the subjects invited to participate agreed and were enrolled. One subject

with LBP was excluded owing to un-interpretable images. Fifty-two subjects (LBP n =

37, healthy participants n = 15) were thus included in the analyses. The subjects with

LBP had moderate pain (Visual analogue Scale; median 34, range 21-50) and low-to-

moderate disability (Roland Morris Disability Score; median 7, range 3-14). There was

no difference between subjects with LBP and healthy subjects regarding age, height and

BMI. Sixty-eight percent of the total sample was women.

Concurrent validity

Table 4 presents the associations between the manually-assessed ADIM and the USI-

measured TrA contraction ratio (TrA-CR) and the TrA preferential ratio (TrA-Pref. Ratio).

None of the values for association were significant.

Page 12 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 46: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

13

The associations between the manually-assessed ADIM and the USI-measured TrA

contraction ratio (TrA-CR) varied and were overall low (r=-0.38-0.1) as were the results

for the manually-assessed ADIM and the TrA preferential activation ratio (TrA-Pref.

Ratio) ( r = -0.19-0.53).

Insert table 4 about here

A subgroup analysis of the subjects presenting with LBP (n=37) was also performed.

Here, the agreement between the manually-assessed ADIM and the USI-measured TrA-

CR and TrA-Pref. ratio also varied (r= 0.09-0.17).

Discriminative validity

The discriminative validity of the manually assessed ADIM—that is, the ability of the

test to discriminate between subjects with LBP and healthy ones—showed a sensitivity

of 0.30 and a specificity of 0.73. Regarding USI measurement of the ADIM the

discriminative validity for TrA-CR showed a sensitivity of 0.35 and a specificity of 0.80

while the TrA preferential ratio showed a sensitivity of 0.41 and a specificity of 0.8

(Table 5).

Reliability

The inter-observer reliability of the ADIM manually assessed by two of the assessors

(M1 and M2), revealed substantial agreement of 87.5%, (κ = 0.71, CI (95%) 0.41-

1.00) while the standard error was 0.15.

Page 13 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 47: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

14

Insert table 5 about here

DISCUSSION

The present aim was to establish the concurrent and discriminative validity, and the inter-

observer reliability, of the manually-assessed ADIM. The ADIM is a clinical tool for

assessing preferential activity of the deep abdominal muscles in subjects with LBP.

Our results showed poor concurrent validity, meaning weak associations between the

manually-assessed ADIM and the two calculated ratios of muscle thickness from the USI

measurements. In addition to the low agreement, the results varied widely (TrA-CR r = -

0.38-0.1; TrA-Pref. Ratio r = -0.19-0.53, respectively). Further, no concurrent correlation

between the manually-assessed test and the USI measurements emerged when computed

only for the group of subjects with LBP. The validity of the manually-assessed ADIM to

discriminate those with LBP from healthy participants only reached low sensitivity

regarding both contraction ratio and preferential ratio. The inter-observer reliability of the

manual assessment of ADIM in the present study was, however, substantial (κ = 0.71).

Clearly, the manually assessed ADIM failed to agree with the USI measurements, either

against TrA contraction ratio or against the TrA preferential activation ratio. Reasons for

the poor concurrent correlation may be the complexity involved in determining an

isolated preferential contraction of a specific muscle when the activity seemingly appears

in several muscles. Urquhart et al [30] suggested that there is often a co-activation between

Page 14 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 48: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

15

TrA and OI as both muscles have to stabilize the lumbar segments. This is also supported

by that the fact that the OI has both a deeper and a superficial layer.[30-32] Further, the

anatomical position where the lateral abdominal muscles overlie one another contributes to

the difficulty in palpating for an isolated TrA contraction.[34-35] This raises the question

of what this test actually adds, or what it actually measures since the inter-observer

reliability was substantial

Our results revealed good specificity to detect individuals presenting with a correct test

among all subjects. However, the results showed lower sensitivity, i.e. the preferential

activation, assessed both manually and with USI, did not meet the fundamental criterion of

discriminating between subjects with LBP and healthy subjects. Again, the question arises of

whether there is a clinical use for the manually assessed ADIM to detect impaired preferential

activation of TrA in subjects with LBP. It must be taken into consideration if the test is to be

used to identifying functional impairments in patients under consideration for rehabilitation.

Our results are supported by Pulkovsky et al [34], who reported inconsistent outcomes in USI-

measured preferential TrA activity in LBP and healthy subjects, concluding that the test was

unable to differentiate between the two groups. The same results have been presented by other

authors [34, 37-38] who also conclude that the values of associations between healthy

participants and subjects with LBP measured with USI are highly variable and no association

can be reached between groups during ADIM-USI. In contrast to these results, Teyhen et

al.[38] showed significant differences in activity of the deep abdominal muscles assessed with

USI in persons with LBP compared to healthy, performing an active straight leg raise test

(ASLR). It should however be noted that their study design differed from ours as no

instruction of ADIM was given to the subjects prior performing the ASLR.

Page 15 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 49: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

16

.

Even so, manual assessment of ADIM is still commonly used in the clinic to differentiate

between patients for guiding specific exercises to re-train motor control. On the present

results, this can be questioned since the ability to perform seems simply to be a normal

variation independent of LBP.

Ferreira et al [38] reported improved TrA muscle activity measured with ultrasound

following an intervention with specific motor-control exercises to activate the deeper

abdominals in people with a poorer ability to preferentially contract TrA. [41] However,

only low associations between improved functional level and the activity of the deep

abdominals were reported. Moreover, recent studies report inconsistent relationships

between alleviation of pain and disability and changes in activity or muscle thickness in

TrA as measured with USI B or M mode. Here the subjects were people with LBP

following interventions with specific motor control exercises.[15, 39]

The values obtained from the muscles measured with USI itself were consistent with

previous findings. [14, 39-40] Small differences in mean values for anatomical distances

appeared at rest and during measured muscle contraction. [18, 33, 35, 41] Our ratio values

were similar to or slightly higher than those previously reported, a phenomenon that

probably relates to basic data, especially the proportionally smaller value from the TrA at

rest.[14]

Strengths and limitations

Page 16 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 50: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

17

The strength of the present study is the aim to investigate both concurrent and

discriminative validity, and also the reliability, of a clinical tool commonly used in the

rehabilitation of patients with LBP when prescribing specific exercises. A further strength

is the study design with investigators blinded to all subjects’ health status. Pulkovsky et al

[34] studied the concurrent validity of the ADIM using ultrasound and a blinded design.

They did not, however, investigate the manually-assessed ADIM. To our knowledge no

other study has performed a single blinded study of this maneuver.

While the use of USI as the criterion was believed to strengthen the results on validity,

the gold-standard method itself showed low discriminative validity. Indeed, a possible

limitation is that USI B mode was used instead of M mode in the assessment, suggesting

the possibility that the peak of TrA activity during the ADIM was missed. Hodges and

colleagues [18] proposed that a change in muscle dimension does not always correlate to

increased muscle activity, depending on muscle length, muscle extensibility, pennation

pattern and competing forces from adjacent muscles. The use of M mode with tissue

Doppler imaging (TDI) may therefore be more advantageous than the use of B mode as

the images are sampled over time.[34] This was however not feasible in the present

study but should be tried in a future one.

Whittaker et al [38] suggest that USI B-mode may be used solely as a source of evaluation of

muscle activity and understanding changes in muscle thickness and neuromuscular control.

However USI-B mode still plays an important role in detecting altered abdominal muscle

function in subjects with LBP and affords helpful biofeedback for abdominal muscle function

when evaluating treatment programs.

Page 17 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 51: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

18

For the selected cut-off value of the ultrasound measurements of TrA contraction and

preferential ratio used in the calculations of concurrent validity, we used cut-offs

previously defined for investigating the ability of the TrA contraction ratio to discriminate

between subjects with chronic LBP and healthy participants. [14, 34] We are aware that a

cut-off value will affect the outcome, and different cut-offs for these ratios may therefore

be discussed and further investigated.

For strengthening internal validity, we presented the USI investigators’ intra-rater

reliability (ICC 0.94), which is consistent with previously reported results of intra-rater

reliability when measuring TrA muscle thickness with USI (ICC = 0.92-0.99). [16] The

present inter-observer reliability of the manual assessment of ADIM was substantial (κ =

0.71), a finding that aligns with a previous study (κ = 0.52). [25]

Our main aim was to investigate whether the manually-assessed ADIM is a valid method

for assessing a preferential activation of the deep abdominal muscles. We did this by

comparing the manually-assessed test (ADIM) results to the USI findings for the same

action. Through blinding of the investigators by including both healthy subjects and

subjects with LBP, we aimed to minimize bias and thus strengthen the results. However,

this led to a disproportionately large sample of subjects with LBP and this may be

considered a limitation to the study.

It is imperative that clinical tests are studied for both validity and reliability. Therefore the

manually-assessed ADIM should be further studied in order to afford clinicians better

directions as to whether this clinical tool really is valid for detecting correct preferential

activity of TrA in persons presenting with LBP.

Page 18 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 52: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

19

CONCLUSION

Our study is the first to consider the concurrent and discriminative validity of the

manually-assessed ADIM, a commonly used clinical tool. Although it´s inter-observer

reliability was high, both the concurrent and discriminative validity was low for examining

preferential activity of the deep abdominal muscles. Neither the manually-assessed ADIM

nor the ultrasound testing discriminated between subjects with low-back pain and healthy

subjects regarding preferential activity of the transversus muscle. This was because this

ability is excited also in healthy subjects. From the present results, it thus appears that the

“preferential” activity cannot be linked to subjects with low-back pain, and this leads us to

question the test’s utility in the clinic for addressing impaired activation of the deep

abdominals.

Acknowledgements

Financial support from the Swedish Confederation of Physiotherapists-Memorial Fund is

gratefully acknowledged.

Authors’ contributions

All authors participated in the conception and design of the study, and all made substantial

contributions to the acquisition of data. BOÄ and ERB performed the statistical analysis.

All authors participated in the interpretation of data. KK drafted the manuscript and BOÄ

Page 19 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 53: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

20

and ERB critically revised the manuscript for important intellectual content. All the

authors read and approved the final manuscript

Funding

KK works as a salaried private practitioner, BOÄ and ERB receive their salaries from the

Karolinska Institutet

Competing interests

No, there are no competing interests.

Ethics approval

The regional ethical review board in Stockholm, Sweden, approved the study (Dnr.

2011/1859–31).

License for Publication

The corresponding author is entitled to grant on behalf of all the authors, and does grant on

behalf of all the authors, an exclusive license (or non-exclusive for government) on a

worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be

published in any BMJPGL products and sublicenses such use and exploitation of all

subsidiary rights, as set out in our license.

(http://group.bmj.com/products/journals/instructions-for-authors/licence-forms)

Competing Interest: None declared.

Provenance and peer review: not commissioned; externally peer reviewed.

Page 20 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 54: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

21

Data Sharing: Additional data is available from Karsten Kaping,

[email protected]

REFERENCES

1. Vleeming, A., et al., The functional coupling of the deep abdominal and paraspinal muscles:

the effects of simulated paraspinal muscle contraction on force transfer to the middle and

posterior layer of the thoracolumbar fascia. J Anat, 2014. 225(4): p. 447-62.

2. Panjabi, M.M., The stabilizing system of the spine. Part I. Function, dysfunction, adaptation,

and enhancement. J Spinal Disord, 1992. 5(4): p. 383-9; discussion 397.

3. Panjabi, M.M., The stabilizing system of the spine. Part II. Neutral zone and instability

hypothesis. J Spinal Disord, 1992. 5(4): p. 390-6; discussion 397.

4. Panjabi, M.M., A hypothesis of chronic back pain: ligament subfailure injuries lead to muscle

control dysfunction. Eur Spine J, 2006. 15(5): p. 668-76.

5. Hodges, P.W., Changes in motor planning of feedforward postural responses of the trunk

muscles in low back pain. Exp Brain Res, 2001. 141(2): p. 261-6.

6. Hodges, P.W. and G.L. Moseley, Pain and motor control of the lumbopelvic region: effect and

possible mechanisms. J Electromyogr Kinesiol, 2003. 13(4): p. 361-70.

7. Hodges, P. and G. Jull, Does strengthening the abdominal muscles prevent low back pain? J

Rheumatol, 2000. 27(9): p. 2286-8.

8. Allison, G.T., S.L. Morris, and B. Lay, Feedforward responses of transversus abdominis are

directionally specific and act asymmetrically: implications for core stability theories. J Orthop

Sports Phys Ther, 2008. 38(5): p. 228-37.

9. Macedo, L.G., et al., Effect of motor control exercises versus graded activity in patients with

chronic nonspecific low back pain: a randomized controlled trial. Phys Ther, 2012. 92(3): p.

363-77.

10. Rasmussen-Barr, E., et al., Graded exercise for recurrent low-back pain: a randomized,

controlled trial with 6-, 12-, and 36-month follow-ups. Spine (Phila Pa 1976), 2009. 34(3): p.

221-8.

11. Bystrom, M.G., E. Rasmussen-Barr, and W.J. Grooten, Motor control exercises reduces pain

and disability in chronic and recurrent low back pain: a meta-analysis. Spine (Phila Pa 1976),

2013. 38(6): p. E350-8.

Page 21 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 55: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

22

12. Macedo, L.G., et al., Motor control exercise for persistent, nonspecific low back pain: a

systematic review. Phys Ther, 2009. 89(1): p. 9-25.

13. Richardson, C., Jull, G., Hodges, P. , Therapeutic Exercise for spinal segmental stabilisation in

low back pain. Scientific basis and clinical approach. 1st ed. 1999, United Kingdom: Churchill

Livingstone

14. Teyhen, D.S., et al., The use of ultrasound imaging of the abdominal drawing-in maneuver in

subjects with low back pain. J Orthop Sports Phys Ther, 2005. 35(6): p. 346-55.

15. Wong, A.Y., et al., Do changes in transversus abdominis and lumbar multifidus during

conservative treatment explain changes in clinical outcomes related to nonspecific low back

pain? A systematic review. J Pain, 2014. 15(4): p. 377 e1-35.

16. Ferreira, P.H., et al., Discriminative and reliability analyses of ultrasound measurement of

abdominal muscles recruitment. Man Ther, 2011. 16(5): p. 463-9.

17. Hides, J., et al., An MRI investigation into the function of the transversus abdominis muscle

during "drawing-in" of the abdominal wall. Spine (Phila Pa 1976), 2006. 31(6): p. E175-8.

18. Hodges, P.W., et al., Measurement of muscle contraction with ultrasound imaging. Muscle

Nerve, 2003. 27(6): p. 682-92.

19. Koppenhaver, S.L., et al., Reliability of rehabilitative ultrasound imaging of the transversus

abdominis and lumbar multifidus muscles. Arch Phys Med Rehabil, 2009. 90(1): p. 87-94.

20. Park, B.H., et al., Reliability and Validity of the Korean Version VISA-P Questionnaire for

Patellar Tendinopathy in Adolescent Elite Volleyball Athletes. Ann Rehabil Med, 2013. 37(5):

p. 698-705.

21. Ghamkhar, L., et al., Application of rehabilitative ultrasound in the assessment of low back

pain: a literature review. J Bodyw Mov Ther, 2011. 15(4): p. 465-77.

22. Carter RE, L.J., Domholdt E. , ed. Rehabilitation research:prinicipals and applications. 4th ed.

ed. 2011, Elsevier Saunders: St.Louis Missouri.

23. Ford, J.D., et al., Development and validation of a brief mental health screening instrument

for newly incarcerated adults. Assessment, 2007. 14(3): p. 279-99.

24. Streiner DL, N.G., ed. Health Measurement Scales. A practical guide to their development and

use. 3rd ed. 2003, Oxford University Press: Oxford.

25. Costa, L.O., et al., Short report: intra-tester reliability of two clinical tests of transversus

abdominis muscle recruitment. Physiother Res Int, 2006. 11(1): p. 48-50.

26. Altman, D.G., ed. Practical statistics for medical research. 1991, Chapman and Hall: London.

27. Norman GR, S.D., ed. The bare essentials. 2008, People's medical publishing house: Shelton,

CT.

28. Stanton, T.R., et al., After an episode of acute low back pain, recurrence is unpredictable and

not as common as previously thought. Spine (Phila Pa 1976), 2008. 33(26): p. 2923-8.

29. Henry, S.M. and K.C. Westervelt, The use of real-time ultrasound feedback in teaching

abdominal hollowing exercises to healthy subjects. J Orthop Sports Phys Ther, 2005. 35(6): p.

338-45.

30. Urquhart, D.M., et al., Abdominal muscle recruitment during a range of voluntary exercises.

Man Ther, 2005. 10(2): p. 144-53.

31. Urquhart, D.M., et al., Regional morphology of the transversus abdominis and obliquus

internus and externus abdominis muscles. Clin Biomech (Bristol, Avon), 2005. 20(3): p. 233-

41.

32. Costa, L.O., et al., An investigation of the reproducibility of ultrasound measures of abdominal

muscle activation in patients with chronic non-specific low back pain. Eur Spine J, 2009. 18(7):

p. 1059-65.

33. Westad, C., P.J. Mork, and O. Vasseljen, Location and sequence of muscle onset in deep

abdominal muscles measured by different modes of ultrasound imaging. J Electromyogr

Kinesiol, 2010. 20(5): p. 994-9.

Page 22 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 56: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

23

34. Pulkovski, N., et al., Ultrasound assessment of transversus abdominis muscle contraction ratio

during abdominal hollowing: a useful tool to distinguish between patients with chronic low

back pain and healthy controls? Eur Spine J, 2012. 21 Suppl 6: p. S750-9.

35. Mannion, A.F., et al., Abdominal muscle size and symmetry at rest and during abdominal

hollowing exercises in healthy control subjects. J Anat, 2008. 213(2): p. 173-82.

36. Turner, D., et al., Using the entire cohort in the receiver operating characteristic analysis

maximizes precision of the minimal important difference. J Clin Epidemiol, 2009. 62(4): p.

374-9.

37. Landis, J.R. and G.G. Koch, An application of hierarchical kappa-type statistics in the

assessment of majority agreement among multiple observers. Biometrics, 1977. 33(2): p.

363-74.

38. Teyhen, D.S., et al., Ultrasound characteristics of the deep abdominal muscles during the

active straight leg raise test. Arch Phys Med Rehabil, 2009. 90(5): p. 761-7.

39. Whittaker, J.L., et al., Association between changes in electromyographic signal amplitude

and abdominal muscle thickness in individuals with and without lumbopelvic pain. J Orthop

Sports Phys Ther, 2013. 43(7): p. 466-77.

40. Beazell, J.R., et al., Changes in lateral abdominal muscle thickness during an abdominal

drawing-in maneuver in individuals with and without low back pain. Res Sports Med, 2011.

19(4): p. 271-82.

41. Rankin, G., M. Stokes, and D.J. Newham, Abdominal muscle size and symmetry in normal

subjects. Muscle Nerve, 2006. 34(3): p. 320-6.

Page 23 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 57: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

24

Table 1 Demographic and self-assessment questionnaire data for all subjects included (n=52)

Total (n=52)

Healthy (n=15)

LBP (n= 37)

Female (n=36)

Male (n=16)

Gender (female/male) 36/16 12/3 24/13

Age (yrs.) (SD) 42 (5.82) 41.5 (10.71) 42.4 (6.97) 42 (7.01) 42 (10.71)

Height (m) (SD) 1.72 (0.24) 1.69 (0.44) 1.73 (0.28) 1.68 (0.18) 1.83 (0.32)

Weight (kg) (SD) 72.2 (10.01) 68.6 (17.71) 73.6 (19.00) 65.5 (10.9) 87.1 (22.4)

BMI (SD) 24 (3.33) 23.5 (6.06) 24.3 (3.99) 23.2 (3.84) 26.1 (6.60)

VAS (0-100)* 34 (21-50)

Roland Morris (0-24)* 7 (3-14)

Oswestry (0-100)* 22 (14-28)

Seeking care for LBP during the previous 5

years

(% of group)

65 0 65

Subjects’ work status

(% of group)

92 100 89

Pain duration

∟2-4 weeks (n) 3

∟1-3 months (n) 3

∟>3 months (n) 31

Recurrent LBP n (% of group)

32 (86)

Page 24 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 58: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

25

LBP=Low-back pain, SD=standard deviation, BMI=body mass index, VAS=visual

analogue scale, *data presented as median (25th /75th percentiles).

Table 2. Data from the USI measurements of muscle thickness (mm) at rest and

during ADIM, presented with mean and SD (n=52).

At rest

(n=52)

ADIM

(n=52)

Change

(%)

TrA Left 2.28 (0.96) 4.41 (0.80) 93.4

TrA Right 2.88 (0.93) 4.49 (0.79) 55.9

TrA total abdomen 2.78 (0.93) 4.45 (0.79) 60.0

OI Left 8.02 (0.49) 9.01 (0.76) 12.3

OI Right 8.01 (0.49) 8.50 (0.63) 6.1

OI total abdomen 8.02 (0.49) 8.75 (0.69) 9.2

OE Left 5.91 (0.58) 5.53 (0.65) -6.4

OE Right 5.68 (0.62) 5.71 (0.62) 0.5

OE total abdomen 5.80 (0.60) 5.62 (0.63) -3.1

USI=ultrasound imaging, TrA= m. transversus abdominis, OI= m. obliquus internus, OE= m. obliquus externus, SD= Standard deviation, % change= percentage change from rested state to activation of the deep abdominal muscles during ADIM.

Page 25 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 59: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

26

Table 3. Calculated ratios from ultrasound imaging of lateral abdominal muscles

during preferential activation of TrA

Contraction ratios All

n=52

Healthy

n=15

LBP

n=37

TrA Contraction ratio,

mean (SD)

1.69 (0.18)

∟Left 1.71 (0.24) 1.75 (0.45) 1.69 (0.28)

∟Right 1.67 (0.23) 1.76 (0.45) 1.63 (0.27)

TrA Preferential ratio,

mean (SD)

0.07 (0.01)

∟Left 0.07 (0.04) 0.072 (0.019) 0.076 (0.019)

∟Right 0.07 (0.04) 0.066 (0.003) 0.076 (0.019)

OE+OI Contraction ratio,

mean (SD)

1.05 (0.14)

∟Left 1.047 (0.15) 1.064 (0.27) 1.04 (0.17)

∟Right 1.048 (0.15) 1.057 (0.27) 1.04 (0.17)

TrA=m. transversus abdominis, OI= m. obliquus internus, OE= m. obliquus

externus Ratio (SD)=calculated ratio with standard deviation

Page 26 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 60: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

27

Table 4. Associations between manually-assessed ADIM and USI-assessed, expressed

in TrA-CR ratio and TrA preferential ratio in all subject (n=52) and in subjects by

investigator.

TrA-CR

(r)

TrA Pref. Ratio

(r)

All investigators (n= 52) .09 .12

Investigator 1 (n=30) 0.1 .06

Investigator 2 (n=13) .03 .53

Investigator 3 (n=9) -.38 -.19

TrA-CR Ratio = transversus abdominis contraction ratio, TrA-Pref. Ratio = transversus abdominis preferential activation ratio. Bold represents good agreement.

Table 5 Sensitivity and specificity for the USI-measured TrA contractions ratio

(TrA-CR) and TrA preferential ratios (TrA-Pref. Ratio) and manually-assessed

ADIM; presented with 95% CI.

Sensitivity

(95% CI)

Specificity

(95%CI)

TrA Contraction Ratio 0.35 (0.20-0.53) 0.80 (0.51-0.95)

TrA Preferential Ratio 0.41 (0.25-0.58) 0.80 (0.51-0.95)

ADIM-Manual 0.30 (0.16-0.47) 0.73 (0.44-0.91)

ADIM= Abdominal drawing-in maneuver , CI = 95% confidence interval , TrA

= M. transversus abdominis, USI = Ultrasound imaging

Page 27 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 61: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

28

Page 28 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 62: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

Figure 1. Test position for abdominal-drawing-in-maneuver (ADIM)

Page 29 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 63: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

A B

Figure 2. Ultrasound image of abdominal muscles, A= rest B=activity during

abdominal-draw-in-maneuver (ADIM). Distance 1 represents M. transversus abdominis (TrA) distance 2 M. obliquus internus (OI) and distance 3 M. obliquus

externus (OE). Note the increase in TrA, OI and the decrease in OE during the

ADIM, picture B.

Page 30 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 64: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

STARD checklist for reporting of studies of diagnostic accuracy

(version January 2003)

Section and Topic Item

#

On page #

TITLE/ABSTRACT/

KEYWORDS

1 Identify the article as a study of diagnostic accuracy (recommend MeSH

heading 'sensitivity and specificity').

2

INTRODUCTION 2 State the research questions or study aims, such as estimating diagnostic

accuracy or comparing accuracy between tests or across participant

groups.

5

METHODS

Participants 3 The study population: The inclusion and exclusion criteria, setting and

locations where data were collected.

6

4 Participant recruitment: Was recruitment based on presenting symptoms,

results from previous tests, or the fact that the participants had received

the index tests or the reference standard?

6

5 Participant sampling: Was the study population a consecutive series of

participants defined by the selection criteria in item 3 and 4? If not,

specify how participants were further selected.

6

6 Data collection: Was data collection planned before the index test and

reference standard were performed (prospective study) or after

(retrospective study)?

7-9

Test methods 7 The reference standard and its rationale. 7-9

8 Technical specifications of material and methods involved including how

and when measurements were taken, and/or cite references for index

tests and reference standard.

7-9

9 Definition of and rationale for the units, cut-offs and/or categories of the

results of the index tests and the reference standard.

9

10 The number, training and expertise of the persons executing and reading

the index tests and the reference standard.

6-7

11 Whether or not the readers of the index tests and reference standard

were blind (masked) to the results of the other test and describe any

other clinical information available to the readers.

7

Statistical methods 12 Methods for calculating or comparing measures of diagnostic accuracy,

and the statistical methods used to quantify uncertainty (e.g. 95%

confidence intervals).

10

13 Methods for calculating test reproducibility, if done. 7

RESULTS

Participants 14 When study was performed, including beginning and end dates of

recruitment.

15 Clinical and demographic characteristics of the study population (at least

information on age, gender, spectrum of presenting symptoms).

6

16 The number of participants satisfying the criteria for inclusion who did or

did not undergo the index tests and/or the reference standard; describe

why participants failed to undergo either test (a flow diagram is strongly

recommended).

11

Test results 17 Time-interval between the index tests and the reference standard, and

any treatment administered in between.

9

18 Distribution of severity of disease (define criteria) in those with the target

condition; other diagnoses in participants without the target condition.

6

19 A cross tabulation of the results of the index tests (including

indeterminate and missing results) by the results of the reference

standard; for continuous results, the distribution of the test results by the

results of the reference standard.

11-12

20 Any adverse events from performing the index tests or the reference

standard.

11

Estimates 21 Estimates of diagnostic accuracy and measures of statistical uncertainty

(e.g. 95% confidence intervals).

12

22 How indeterminate results, missing data and outliers of the index tests

were handled.

11

23 Estimates of variability of diagnostic accuracy between subgroups of

participants, readers or centers, if done.

11

24 Estimates of test reproducibility, if done. 11-12

DISCUSSION 25 Discuss the clinical applicability of the study findings. 13-14

Page 31 of 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 65: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

The abdominal drawing-in maneuver for detecting activity in the deep abdominal muscles: is this clinical tool reliable

and valid?

Journal: BMJ Open

Manuscript ID bmjopen-2015-008711.R2

Article Type: Research

Date Submitted by the Author: 21-Oct-2015

Complete List of Authors: Kaping, Karsten; NVS-Sjukgymnastik, Karolinska Institutet; Rehabtjanst Äng, Björn; NVS-Sjukgymnastik, Karolinska Institutet Rasmussen-Barr, Eva; NVS-Sjukgymnastik, Karolinska Institutet

<b>Primary Subject Heading</b>:

Rehabilitation medicine

Secondary Subject Heading: Sports and exercise medicine, Radiology and imaging, Diagnostics, Rehabilitation medicine

Keywords:

Back pain < ORTHOPAEDIC & TRAUMA SURGERY, Musculoskeletal disorders < ORTHOPAEDIC & TRAUMA SURGERY, Orthopaedic sports trauma < ORTHOPAEDIC & TRAUMA SURGERY, Spine < ORTHOPAEDIC & TRAUMA SURGERY, REHABILITATION MEDICINE, ULTRASONOGRAPHY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open on June 23, 2020 by guest. P

rotected by copyright.http://bm

jopen.bmj.com

/B

MJ O

pen: first published as 10.1136/bmjopen-2015-008711 on 9 D

ecember 2015. D

ownloaded from

Page 66: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

1

The abdominal drawing-in maneuver for detecting activity in the

deep abdominal muscles: is this clinical tool reliable and valid?

Karsten Kaping, RPT, MT, MSc *,1,2 Björn O. Äng, RPT, Associate Professor, 1

Eva Rasmussen-Barr. RPT, PhD 1,2

Address: 1 Department of Neurobiology, Care Sciences and Society, Division of

Physiotherapy, Karolinska Institutet, Alfred Nobels Allé 12, 141 83 Huddinge

Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden

Telephone: +46-8-441 16 60

Email: Karsten Kaping* [email protected];

[email protected]; [email protected]

*Corresponding author

Keywords:

Abdominal muscles, concurrent validity, lumbar spine, transversus abdominis, ultrasound

imaging

Word count: 3561

Page 1 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 67: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

2

ABSTRACT (word count 288)

Objective: The abdominal drawing-in maneuver (ADIM) is a common clinical tool for

manually assessing as “correct” or not a preferential activation of the deep abdominal muscles

in patients with low-back pain (LBP). The validity and reliability of manual assessment of the

ADIM are, however, as yet unknown. This study evaluated the concurrent and discriminative

validity and the reliability of the manually assessed ADIM.

Design: Single-blinded cross-sectional study

Settings: General population in Stockholm County, Sweden

Participants: The study sample comprised 38 subjects seeking care for low-back pain and 15

healthy subjects.

Measures: The manual ADIM was assessed as correct or not following a standard procedure.

Ultrasound imaging (USI) was used as the concurrent reference (gold standard) for the

manually assessed ADIM by calculating a ratio of the change in muscle thickness between the

resting and the contracted states: the correlation between manual test and the USI was

calculated. Discriminative validity was analysed by calculating sensitivity and specificity. A

sample of 24 subjects was analysed with kappa coefficients for inter-observer reliability

between two raters.

Results: The concurrent validity between the manual ADIM and the ADIM-USI ratios

showed poor correlations (r=.13-.40). The discriminative validity of the manually assessed

ADIM to predict LBP showed a sensitivity/specificity of 0.30/0.73, while the ADIM-USI

ratio to predict LBP showed 0.19/0.87. The inter-observer reliability for the manually-

assessed ADIM revealed substantial agreement: (K= 0.71, CI (95%) 0.41-1.00.

Conclusion: Although the inter-observer reliability of the manually-assessed ADIM was

high, both the concurrent and discriminative validity was low for examining the preferential

activity of the deep abdominal muscles. Neither the manually-assessed ADIM nor the

ultrasound testing discriminated between subjects with LBP and healthy subjects regarding

preferential activity of the transversus muscle as this ability/inability was also present in

healthy subjects.

Page 2 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 68: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

3

STRENGTHS AND LIMITATIONS

Strengths and limitations of this study

Strengths and limitations of this study

•Strength of this study is that the abdominal-drawing-in-manoeuvre was analysed for both

concurrent and discriminative validity and that the investigators were blinded to all subjects’

health status.

•Strength is also that the study used a gold-standard tool (ultrasound) to measure the quantity

of muscle activity for comparison with the manual estimation.

•Possible limitation is that our study design resulted in a disproportionate sample between

healthy subjects and those with low-back pain.

•In addition, the patients estimated their low-back pain only as low-to-moderate

Page 3 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 69: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

4

INTRODUCTION (word count 3388)

The lumbar spine depends on complex interaction between passive and active structures such

as contracting muscles, ligaments and fascia to maintain optimal spinal control during bodily

movement and loading [1-3]. Contraction of the lateral abdominal muscles, and especially the

anticipatory activity of the transversus abdominus muscle (TrA), is suggested to be a key

motor-control element for transferring spinal loads to the pelvis and upper trunk, thus

avoiding injury to spinal tissue or pain incidents following sub-failure loading. [2-5]

Studies have reported delayed anticipatory activity of TrA in low-back-pain sufferers (LBP)

in comparison to healthy subjects [5-7], although such differences in temporal activity have

also been questioned. [8] Specific exercises have been proposed for re-training motor control

of the abdominal muscles; a popular and common treatment for people with recurrent and

chronic LBP.[9-10] Following such interventions, recent systematic reviews have reported

improvement of symptoms such as pain and functional deficits compared to other exercises or

treatment.[11-12]

To assess the activity of the deep abdominal muscles a specific clinical test is available: the

abdominal drawing-in maneuver (ADIM). [13] Through manual palpation and visual

inspection, the ADIM is used to assess the success or not in performing a voluntary and

preferential contraction of the deep abdominal muscles, especially the TrA [13-14], in order to

target interventions with specific exercises and follow-up progression.

In addition, ultrasound imaging (USI) has for several years been used in clinical settings to

quantify the activity of the deep abdominal muscles. Specifically, it has also been used in

research to quantify change in muscle thickness or muscle onset of TrA before and after

interventions using motor control exercises.[15] The USI has been validated for measuring

Page 4 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 70: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

5

change in muscle thickness against electromyography (EMG) [16] and magnetic resonance

imaging ( MRI) [17] and demonstrates high inter-observer reliability in both healthy people

and subjects with LBP. [14, 18-21] However, not all clinicians have access to USI and the

manually assessed ADIM therefore remains the most common clinical tool for addressing and

examining a preferential contraction of the deep abdominal muscles.

A clinically convenient test should show both good reliability and high validity. Specifically,

it should be able to demonstrate concurrent validity against a reliable gold-standard

assessment (e.g., USI verifying TrA activity); proved to measure what it purports to

measure.[22] If the clinical test is to reflect impairment, not merely normal variation in

healthy subjects, the test should also exhibit empirical, discriminative validity to justify its use

among patients. [23] For the present purpose discriminative validity was defined as the extent

to which manually-assessed and USI-assessed ADIM differentiated between subjects with

LBP and those without.[24]

To date, the concurrent and discriminative validity of the manually assessed ADIM is

unknown, and only one study addresses inter-observer reliability.[25] The present aims were

therefore to assess the concurrent validity of the manually assessed ADIM in relation to the

USI during guided abdominal contractions, to assess its discriminative ability, and to

determine the inter-observer reliability.

METHODS

Study design

A cross-sectional, single-blinded design was used to evaluate the concurrent validity,

discriminating validity and inter-observer reliability of the ADIM in subjects with LBP and in

healthy subjects. A sample size of 50 subjects was considered adequate to power the study to

obtain proper correlation data as used in concurrent analysis, [26], while a sample of 27

Page 5 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 71: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

6

subjects was calculated for kappa agreement data (confidence interval 0.3; expected

percentage agreement 0.8 and chance agreement 0.5) used in reliability statistics.[27] Planned

enrolment was undertaken to meet these criteria. The regional ethical review board approved

the study (Dnr 2011/1859-31).

Study sample

Subjects with LBP were recruited through information about the study distributed by e-mail

and telephone to six private physiotherapy settings in Stockholm, Sweden. Healthy subjects

were found through written and verbal information distributed at two large companies in

Stockholm, each with approximately 700 employees. Fifty-three subjects, 38 with LBP and

15 healthy participants aged between 18 and 65 were included (Table 1). Subjects seeking

care for a primary complaint of LBP were considered eligible and included if their LBP had

lasted for at least two weeks and if they estimated their pain level to > 20 on a visual analogue

scale (VAS 0-100). [23, 28] They were excluded if pain was severe and thought likely to

prevent the planned test procedure; if they were pregnant; had undergone back or abdominal

surgery within the last six months; had a neurological disease, neurological signs of nerve root

compression, inflammatory autoimmune disease, or a malignancy. Subjects with spondylosis,

or spondylolisthesis were included if no signs of nerve root compression existed. Healthy

subjects were excluded if they reported any pain on the VAS or had undergone back

rehabilitation within the previous five years. Both patients with LBP and healthy subjects

signed written informed-consent forms. Subjects were enrolled 2012-02-10 to 2012-06-27.

Page 6 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 72: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

7

Investigators

The manual investigators (n = 3) were physiotherapists specialized in manual therapy for 13–

32 years and with several years of experience assessing ADIM in patients with LBP. Another

investigator skilled in musculoskeletal USI performed all the USI assessments. A same day

intra-rater reliability was performed and calculated from the obtained USI measurements on

35 subjects and considered very good (ICC 2.1, 0.84).

Both the manual and the USI investigator were blinded to the subject´s health status at the

time of testing. At inclusion, prior to the clinical intervention, all subjects completed

demographic and medical-history, and self-assessment questionnaires on pain and

function.

Insert table 1 about here

Manual assessment of ADIM

ADIM was explained, performed and assessed as a slow and gentle “abdominal hollowing”

as originally described by Richardson et al. [13] The subjects were in a supine hook-lying

with hips at 40–60° and knees flexed between 90-100°, and arms placed along the torso

(Figure 1). They were instructed to gently draw the navel and in towards the spine, to hold

the contraction and breathe normally. The manual investigator palpated for muscle

contraction medially to the spina iliaca anterior superior (SIAS). The success of a correctly

performed contraction of the TrA was considered to be a slowly-developed, deep tension in

the abdominal wall continuously for up to 10 seconds. [13] Compensatory patterns were

observed during the test, such as: holding the breath; backward movement of the pelvis;

visual or palpable contraction of the obliquus externus muscle; pressing the heels towards

Page 7 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 73: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

8

the floor. If any of these patterns appeared, the ADIM was assessed as incorrect. [13, 29-

31] Before the investigation, the manual investigators underwent a training session to reach

a consensus on how to judge whether the manually-assessed ADIM was correct.

Insert figure 1 about here

Ultrasonography assessment of ADIM and data processing

Ultrasound images were recorded with the Prosonix 600 DCPD Swissmed (HP

Medical GMBH, Wollerau, Switzerland), and motions were registered using a linear-

array transducer (5-10 MHz). Using B (brightness) mode ultrasound, the transducer

was positioned across the abdominal wall at a point between the inferior angle of the

rib cage and the iliac crest, approximately 100 mm from the umbilicus. [32-33] On-

screen callipers measured the distance between the inner edge of the fascial lines of

the TrA, the obliquus internus (OI), and the obliquus externus (OE) muscles. The

subject could not see the computer screen. Three pictures during rest and ADIM were

recorded for each body side. The images of muscles at rest were collected at the end of

exhalation, and images during ADIM were collected when the subject had reached the

expected peak of activity (Figure 2).

Insert figure 2 about here

Mean values of the measurements from the ADIM and resting position were

calculated for TrA, OI, and OE (Table 2). From these values, three ratios were

calculated: (1) The TrA contraction ratio = TrA thickness contracted/TrA thickness at

rest, (2) The TrA preferential activation ratio= relative change in proportion of TrA

Page 8 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 74: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

9

relative to total lateral abdominal muscle thickness= (TrA contracted/ TrA+OE+OI

contracted)-(TrA at rest/TrA+OE?+OI at rest) and (3) OI + OE contraction ratio =

OE+OI thickness contracted/OE+OI thickness at rest . [14]

The contraction ratio represents the relative change in muscle thickness as a result of

muscle contraction, and the preferential activation ratio, the change in the proportion

of the TrA relative to the total lateral abdominal muscle thickness. [14]

Insert table 2 about here

The mean ratio value minus one standard deviation (-1 SD) was selected as the cut-off

to define the preferential activation of TrA.[34] In assessing whether the ADIM, i.e.

the preferential activation of TrA, was correctly performed all individual ratios had to

exceed this defined cut-off to be accepted as passed (table 3). [34-35]

Insert table 3 about here

Experimental setup

All included subjects were assigned by lot to start with either the manual or the USI

assessment of the ADIM. They were briefed by the first investigator (either manual or USI)

about the purpose of the ADIM and were instructed how to perform a correct ADIM.

Following this, each subject was asked to perform an ADIM lasting up to 10 seconds without

Page 9 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 75: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

10

feedback; this was repeated three times. The manual investigator filled out a paper chart with

yes or no; yes indicating a correctly performed ADIM.

If two of the three performed ADIMs were assessed as a correct contraction; the test

was noted and filled out as correct. The subject rested for 10 minutes between the

manual assessments and the USI assessment. To investigate the inter-observer reliability

of the manually-assessed ADIM, a second investigator (M2) observed and assessed the

ADIM ten minutes following the first assessment (M1). Both investigators were blinded

to each other’s results.

Statistical analysis

For the concurrent validity analyses, Spearman’s rank correlation coefficient (r) was

used to calculate associations between the results of the manual test and the USI

investigation of the preferential activation of the TrA. The coefficient (r) was defined

as follows: 0.00–0.25 = little or no correlation; 0.26–0.49 = low correlation; 0.50–

0.69 = moderate correlation; 0.70–0.89 = high correlation; and 0.9–1.0 = very high

correlation.[22, 36]

Discriminative validity was analysed by calculating sensitivity and specificity, with a 95%

confidence interval for both patients with LBP and healthy subjects. We considered at least

70% sensitivity and 50% specificity to be valid.[36] A correctly or not correctly performed

ADIM was scored as “yes” or “no” both for subjects with LBP and for healthy.

Page 10 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 76: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

11

The inter-observer reliability of the manual test was calculated using kappa (κ), and was

presented with 95% CI, percentage agreement (PA), and standard error (SE). We defined

the strength of the kappa coefficient according to Landis and Koch: 0 = poor, 0.01–0.2 =

slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 substantial and 0.81–1 = almost

perfect.[37]

RESULTS

Fifty-three subjects were allocated to the study. Their characteristics are presented in

Table 1. All the subjects invited to participate agreed and were enrolled. One subject

with LBP was excluded owing to un-interpretable images. Fifty-two subjects (LBP n =

37, healthy participants n = 15) were thus included in the analyses. The subjects with

LBP had moderate pain (Visual analogue Scale; median 34, range 21-50) and low-to-

moderate disability (Roland Morris Disability Score; median 7, range 3-14). There was

no difference between subjects with LBP and healthy subjects regarding age, height and

BMI. Sixty-eight percent of the total sample was women.

Concurrent validity

Table 4 presents the associations between the manually-assessed ADIM and the USI-

measured TrA contraction ratio (TrA-CR) and the TrA preferential ratio (TrA-Pref. Ratio).

None of the values for association were significant.

Page 11 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 77: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

12

The associations between the manually-assessed ADIM and the USI-measured TrA

contraction ratio (TrA-CR) varied and were overall low (r=-0.38-0.1) as were the results

for the manually-assessed ADIM and the TrA preferential activation ratio (TrA-Pref.

Ratio) ( r = -0.19-0.53).

Insert table 4 about here

A subgroup analysis of the subjects presenting with LBP (n=37) was also performed.

Here, the agreement between the manually-assessed ADIM and the USI-measured TrA-

CR and TrA-Pref. ratio also varied (r= 0.09-0.17).

Discriminative validity

The discriminative validity of the manually assessed ADIM—that is, the ability of the

test to discriminate between subjects with LBP and healthy ones—showed a sensitivity

of 0.30 and a specificity of 0.73. Regarding USI measurement of the ADIM the

discriminative validity for TrA-CR showed a sensitivity of 0.35 and a specificity of 0.80

while the TrA preferential ratio showed a sensitivity of 0.41 and a specificity of 0.8

(Table 5).

Reliability

The inter-observer reliability of the ADIM manually assessed by two of the assessors

(M1 and M2), revealed substantial agreement of 87.5%, (κ = 0.71, CI (95%) 0.41-

1.00) while the standard error was 0.15.

Page 12 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 78: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

13

Insert table 5 about here

DISCUSSION

The present aim was to establish the concurrent and discriminative validity, and the inter-

observer reliability, of the manually-assessed ADIM. The ADIM is a clinical tool for

assessing preferential activity of the deep abdominal muscles in subjects with LBP.

Our results showed poor concurrent validity, meaning weak associations between the

manually-assessed ADIM and the two calculated ratios of muscle thickness from the USI

measurements. In addition to the low agreement, the results varied widely (TrA-CR r = -

0.38-0.1; TrA-Pref. Ratio r = -0.19-0.53, respectively). Further, no concurrent correlation

between the manually-assessed test and the USI measurements emerged when computed

only for the group of subjects with LBP. The validity of the manually-assessed ADIM to

discriminate those with LBP from healthy participants only reached low sensitivity

regarding both contraction ratio and preferential ratio. The inter-observer reliability of the

manual assessment of ADIM in the present study was, however, substantial (κ = 0.71).

Clearly, the manually assessed ADIM failed to agree with the USI measurements, either

against TrA contraction ratio or against the TrA preferential activation ratio. Reasons for

the poor concurrent correlation may be the complexity involved in determining an

isolated preferential contraction of a specific muscle when the activity seemingly appears

in several muscles. Urquhart et al [30] suggested that there is often a co-activation between

Page 13 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 79: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

14

TrA and OI as both muscles have to stabilize the lumbar segments. This is also supported

by that the fact that the OI has both a deeper and a superficial layer.[30-32] Further, the

anatomical position where the lateral abdominal muscles overlie one another contributes to

the difficulty in palpating for an isolated TrA contraction.[34-35] This raises the question

of what this test actually adds, or what it actually measures since the inter-observer

reliability was substantial

Our results revealed good specificity to detect individuals presenting with a correct test

among all subjects. However, the results showed lower sensitivity, i.e. the preferential

activation, assessed both manually and with USI, did not meet the fundamental criterion of

discriminating between subjects with LBP and healthy subjects. Again, the question arises of

whether there is a clinical use for the manually assessed ADIM to detect impaired preferential

activation of TrA in subjects with LBP. It must be taken into consideration if the test is to be

used to identifying functional impairments in patients under consideration for rehabilitation.

Our results are supported by Pulkovsky et al [34], who reported inconsistent outcomes in USI-

measured preferential TrA activity in LBP and healthy subjects, concluding that the test was

unable to differentiate between the two groups. The same results have been presented by other

authors [34, 37-38] who also conclude that the values of associations between healthy

participants and subjects with LBP measured with USI are highly variable and no association

can be reached between groups during ADIM-USI. In contrast to these results, Teyhen et

al.[38] showed significant differences in activity of the deep abdominal muscles assessed with

USI in persons with LBP compared to healthy, performing an active straight leg raise test

(ASLR). It should however be noted that their study design differed from ours as no

instruction of ADIM was given to the subjects prior performing the ASLR.

Page 14 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 80: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

15

.

Even so, manual assessment of ADIM is still commonly used in the clinic to differentiate

between patients for guiding specific exercises to re-train motor control. On the present

results, this can be questioned since the ability to perform seems simply to be a normal

variation independent of LBP.

Ferreira et al [38] reported improved TrA muscle activity measured with ultrasound

following an intervention with specific motor-control exercises to activate the deeper

abdominals in people with a poorer ability to preferentially contract TrA. However, only

low associations between improved functional level and the activity of the deep

abdominals were reported. Moreover, recent studies report inconsistent relationships

between alleviation of pain and disability and changes in activity or muscle thickness in

TrA as measured with USI B or M mode. Here the subjects were people with LBP

following interventions with specific motor control exercises.[15, 39]

The values obtained from the muscles measured with USI itself were consistent with

previous findings. [14, 39-40] Small differences in mean values for anatomical distances

appeared at rest and during measured muscle contraction. [18, 33, 35, 41] Our ratio values

were similar to or slightly higher than those previously reported, a phenomenon that

probably relates to basic data, especially the proportionally smaller value from the TrA at

rest.[14]

Page 15 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 81: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

16

Strengths and limitations

The strength of the present study is the aim to investigate both concurrent and

discriminative validity, and also the reliability, of a clinical tool commonly used in the

rehabilitation of patients with LBP when prescribing specific exercises. A further strength

is the study design with investigators blinded to all subjects’ health status. Pulkovsky et al

[34] studied the concurrent validity of the ADIM using ultrasound and a blinded design.

They did not, however, investigate the manually-assessed ADIM. To our knowledge no

other study has performed a single blinded study of this maneuver.

While the use of USI as the criterion was believed to strengthen the results on validity,

the gold-standard method itself showed low discriminative validity. Indeed, a possible

limitation is that USI B mode was used instead of M mode in the assessment, suggesting

the possibility that the peak of TrA activity during the ADIM was missed. Hodges and

colleagues [18] proposed that a change in muscle dimension does not always correlate to

increased muscle activity, depending on muscle length, muscle extensibility, pennation

pattern and competing forces from adjacent muscles. The use of M mode with tissue

Doppler imaging (TDI) may therefore be more advantageous than the use of B mode as

the images are sampled over time.[34] This was however not feasible in the present

study but should be tried in a future one.

Whittaker et al [38] suggest that USI B-mode may be used solely as a source of evaluation of

muscle activity and understanding changes in muscle thickness and neuromuscular control.

However USI-B mode still plays an important role in detecting altered abdominal muscle

function in subjects with LBP and affords helpful biofeedback for abdominal muscle function

when evaluating treatment programs.

Page 16 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 82: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

17

For the selected cut-off value of the ultrasound measurements of TrA contraction and

preferential ratio used in the calculations of concurrent validity, we used cut-offs

previously defined for investigating the ability of the TrA contraction ratio to discriminate

between subjects with chronic LBP and healthy participants. [14, 34] We are aware that a

cut-off value will affect the outcome, and different cut-offs for these ratios may therefore

be discussed and further investigated.

For strengthening internal validity, we presented the USI investigators’ intra-rater

reliability (ICC 0.84), which is consistent with previously reported results of intra-rater

reliability when measuring TrA muscle thickness with USI (ICC = 0.92-0.99). [16] The

present inter-observer reliability of the manual assessment of ADIM was substantial (κ =

0.71), a finding that aligns with a previous study (κ = 0.52). [25]

Our main aim was to investigate whether the manually-assessed ADIM is a valid method

for assessing a preferential activation of the deep abdominal muscles. We did this by

comparing the manually-assessed test (ADIM) results to the USI findings for the same

action. Through blinding of the investigators by including both healthy subjects and

subjects with LBP, we aimed to minimize bias and thus strengthen the results. However,

this led to a disproportionately large sample of subjects with LBP and this may be

considered a limitation to the study.

It is imperative that clinical tests are studied for both validity and reliability. Therefore the

manually-assessed ADIM should be further studied in order to afford clinicians better

directions as to whether this clinical tool really is valid for detecting correct preferential

activity of TrA in persons presenting with LBP.

Page 17 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 83: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

18

CONCLUSION

Our study is the first to consider the concurrent and discriminative validity of the

manually-assessed ADIM, a commonly used clinical tool. Although it´s inter-observer

reliability was high, both the concurrent and discriminative validity was low for examining

preferential activity of the deep abdominal muscles. Neither the manually-assessed ADIM

nor the ultrasound testing discriminated between subjects with low-back pain and healthy

subjects regarding preferential activity of the transversus muscle. This was because this

ability was excited also in healthy subjects. From the present results, it thus appears that

the “preferential” activity cannot be linked to subjects with low-back pain, and this leads us

to question the test’s utility in the clinic for addressing impaired activation of the deep

abdominals.

Page 18 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 84: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

19

Acknowledgements

Financial support from the Swedish Confederation of Physiotherapists-Memorial Fund is

gratefully acknowledged.

Authors’ contributions

All authors participated in the conception and design of the study, and all made substantial

contributions to the acquisition of data. BOÄ and ERB performed the statistical analysis.

All authors participated in the interpretation of data. KK drafted the manuscript and BOÄ

and ERB critically revised the manuscript for important intellectual content. All the

authors read and approved the final manuscript

Funding

KK works as a salaried private practitioner, BOÄ and ERB receive their salaries from the

Karolinska Institutet

Competing interests

None.

Ethics approval

The regional ethical review board in Stockholm, Sweden, approved the study (Dnr.

2011/1859–31).

License for Publication

Page 19 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 85: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

20

The corresponding author is entitled to grant on behalf of all the authors, and does grant on

behalf of all the authors, an exclusive license (or non-exclusive for government) on a

worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be

published in any BMJPGL products and sublicenses such use and exploitation of all

subsidiary rights, as set out in our license.

(http://group.bmj.com/products/journals/instructions-for-authors/licence-forms)

Competing Interest: None declared.

Provenance and peer review: not commissioned; externally peer reviewed.

Data sharing statement: additional data are not available.

Page 20 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 86: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

21

REFERENCES

1. Vleeming, A., et al., The functional coupling of the deep abdominal and paraspinal muscles:

the effects of simulated paraspinal muscle contraction on force transfer to the middle and

posterior layer of the thoracolumbar fascia. J Anat, 2014. 225(4): p. 447-62.

2. Panjabi, M.M., The stabilizing system of the spine. Part I. Function, dysfunction, adaptation,

and enhancement. J Spinal Disord, 1992. 5(4): p. 383-9; discussion 397.

3. Panjabi, M.M., The stabilizing system of the spine. Part II. Neutral zone and instability

hypothesis. J Spinal Disord, 1992. 5(4): p. 390-6; discussion 397.

4. Panjabi, M.M., A hypothesis of chronic back pain: ligament subfailure injuries lead to muscle

control dysfunction. Eur Spine J, 2006. 15(5): p. 668-76.

5. Hodges, P.W., Changes in motor planning of feedforward postural responses of the trunk

muscles in low back pain. Exp Brain Res, 2001. 141(2): p. 261-6.

6. Hodges, P.W. and G.L. Moseley, Pain and motor control of the lumbopelvic region: effect and

possible mechanisms. J Electromyogr Kinesiol, 2003. 13(4): p. 361-70.

7. Hodges, P. and G. Jull, Does strengthening the abdominal muscles prevent low back pain? J

Rheumatol, 2000. 27(9): p. 2286-8.

8. Allison, G.T., S.L. Morris, and B. Lay, Feedforward responses of transversus abdominis are

directionally specific and act asymmetrically: implications for core stability theories. J Orthop

Sports Phys Ther, 2008. 38(5): p. 228-37.

9. Macedo, L.G., et al., Effect of motor control exercises versus graded activity in patients with

chronic nonspecific low back pain: a randomized controlled trial. Phys Ther, 2012. 92(3): p.

363-77.

10. Rasmussen-Barr, E., et al., Graded exercise for recurrent low-back pain: a randomized,

controlled trial with 6-, 12-, and 36-month follow-ups. Spine (Phila Pa 1976), 2009. 34(3): p.

221-8.

11. Bystrom, M.G., E. Rasmussen-Barr, and W.J. Grooten, Motor control exercises reduces pain

and disability in chronic and recurrent low back pain: a meta-analysis. Spine (Phila Pa 1976),

2013. 38(6): p. E350-8.

12. Macedo, L.G., et al., Motor control exercise for persistent, nonspecific low back pain: a

systematic review. Phys Ther, 2009. 89(1): p. 9-25.

13. Richardson, C., Jull, G., Hodges, P. , Therapeutic Exercise for spinal segmental stabilisation in

low back pain. Scientific basis and clinical approach. 1st ed. 1999, United Kingdom: Churchill

Livingstone

14. Teyhen, D.S., et al., The use of ultrasound imaging of the abdominal drawing-in maneuver in

subjects with low back pain. J Orthop Sports Phys Ther, 2005. 35(6): p. 346-55.

15. Wong, A.Y., et al., Do changes in transversus abdominis and lumbar multifidus during

conservative treatment explain changes in clinical outcomes related to nonspecific low back

pain? A systematic review. J Pain, 2014. 15(4): p. 377 e1-35.

16. Ferreira, P.H., et al., Discriminative and reliability analyses of ultrasound measurement of

abdominal muscles recruitment. Man Ther, 2011. 16(5): p. 463-9.

17. Hides, J., et al., An MRI investigation into the function of the transversus abdominis muscle

during "drawing-in" of the abdominal wall. Spine (Phila Pa 1976), 2006. 31(6): p. E175-8.

18. Hodges, P.W., et al., Measurement of muscle contraction with ultrasound imaging. Muscle

Nerve, 2003. 27(6): p. 682-92.

19. Koppenhaver, S.L., et al., Reliability of rehabilitative ultrasound imaging of the transversus

abdominis and lumbar multifidus muscles. Arch Phys Med Rehabil, 2009. 90(1): p. 87-94.

20. Park, B.H., et al., Reliability and Validity of the Korean Version VISA-P Questionnaire for

Patellar Tendinopathy in Adolescent Elite Volleyball Athletes. Ann Rehabil Med, 2013. 37(5):

p. 698-705.

Page 21 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 87: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

22

21. Ghamkhar, L., et al., Application of rehabilitative ultrasound in the assessment of low back

pain: a literature review. J Bodyw Mov Ther, 2011. 15(4): p. 465-77.

22. Carter RE, L.J., Domholdt E. , ed. Rehabilitation research:prinicipals and applications. 4th ed.

ed. 2011, Elsevier Saunders: St.Louis Missouri.

23. Ford, J.D., et al., Development and validation of a brief mental health screening instrument

for newly incarcerated adults. Assessment, 2007. 14(3): p. 279-99.

24. Streiner DL, N.G., ed. Health Measurement Scales. A practical guide to their development and

use. 3rd ed. 2003, Oxford University Press: Oxford.

25. Costa, L.O., et al., Short report: intra-tester reliability of two clinical tests of transversus

abdominis muscle recruitment. Physiother Res Int, 2006. 11(1): p. 48-50.

26. Altman, D.G., ed. Practical statistics for medical research. 1991, Chapman and Hall: London.

27. Norman GR, S.D., ed. The bare essentials. 2008, People's medical publishing house: Shelton,

CT.

28. Stanton, T.R., et al., After an episode of acute low back pain, recurrence is unpredictable and

not as common as previously thought. Spine (Phila Pa 1976), 2008. 33(26): p. 2923-8.

29. Henry, S.M. and K.C. Westervelt, The use of real-time ultrasound feedback in teaching

abdominal hollowing exercises to healthy subjects. J Orthop Sports Phys Ther, 2005. 35(6): p.

338-45.

30. Urquhart, D.M., et al., Abdominal muscle recruitment during a range of voluntary exercises.

Man Ther, 2005. 10(2): p. 144-53.

31. Urquhart, D.M., et al., Regional morphology of the transversus abdominis and obliquus

internus and externus abdominis muscles. Clin Biomech (Bristol, Avon), 2005. 20(3): p. 233-

41.

32. Costa, L.O., et al., An investigation of the reproducibility of ultrasound measures of abdominal

muscle activation in patients with chronic non-specific low back pain. Eur Spine J, 2009. 18(7):

p. 1059-65.

33. Westad, C., P.J. Mork, and O. Vasseljen, Location and sequence of muscle onset in deep

abdominal muscles measured by different modes of ultrasound imaging. J Electromyogr

Kinesiol, 2010. 20(5): p. 994-9.

34. Pulkovski, N., et al., Ultrasound assessment of transversus abdominis muscle contraction ratio

during abdominal hollowing: a useful tool to distinguish between patients with chronic low

back pain and healthy controls? Eur Spine J, 2012. 21 Suppl 6: p. S750-9.

35. Mannion, A.F., et al., Abdominal muscle size and symmetry at rest and during abdominal

hollowing exercises in healthy control subjects. J Anat, 2008. 213(2): p. 173-82.

36. Turner, D., et al., Using the entire cohort in the receiver operating characteristic analysis

maximizes precision of the minimal important difference. J Clin Epidemiol, 2009. 62(4): p.

374-9.

37. Landis, J.R. and G.G. Koch, An application of hierarchical kappa-type statistics in the

assessment of majority agreement among multiple observers. Biometrics, 1977. 33(2): p.

363-74.

38. Teyhen, D.S., et al., Ultrasound characteristics of the deep abdominal muscles during the

active straight leg raise test. Arch Phys Med Rehabil, 2009. 90(5): p. 761-7.

39. Whittaker, J.L., et al., Association between changes in electromyographic signal amplitude

and abdominal muscle thickness in individuals with and without lumbopelvic pain. J Orthop

Sports Phys Ther, 2013. 43(7): p. 466-77.

40. Beazell, J.R., et al., Changes in lateral abdominal muscle thickness during an abdominal

drawing-in maneuver in individuals with and without low back pain. Res Sports Med, 2011.

19(4): p. 271-82.

41. Rankin, G., M. Stokes, and D.J. Newham, Abdominal muscle size and symmetry in normal

subjects. Muscle Nerve, 2006. 34(3): p. 320-6.

Page 22 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 88: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

23

Table 1 Demographic and self-assessment questionnaire data for all subjects included (n=52)

Total

(n=52)

Healthy

(n=15)

LBP

(n= 37)

Female

(n=36)

Male

(n=16)

Gender (female/male) 36/16 12/3 24/13

Age (yrs.) (SD) 42 (5.82) 41.5 (10.71) 42.4 (6.97) 42 (7.01) 42 (10.71)

Height (m) (SD) 1.72 (0.24) 1.69 (0.44) 1.73 (0.28) 1.68 (0.18) 1.83 (0.32)

Weight (kg) (SD) 72.2 (10.01) 68.6 (17.71) 73.6 (19.00) 65.5 (10.9) 87.1 (22.4)

BMI (SD) 24 (3.33) 23.5 (6.06) 24.3 (3.99) 23.2 (3.84) 26.1 (6.60)

VAS (0-100)* 34 (21-50)

Roland Morris (0-24)* 7 (3-14)

Oswestry (0-100)* 22 (14-28)

Seeking care for LBP

during the previous 5 years

(% of group)

65 0 65

Subjects’ work status

(% of group)

92 100 89

Pain duration

∟2-4 weeks (n) 3

∟1-3 months (n) 3

∟>3 months (n) 31

Recurrent LBP

n (% of group) 32 (86)

Page 23 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 89: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

24

LBP=Low-back pain, SD=standard deviation, BMI=body mass index, VAS=visual

analogue scale, *data presented as median (25th /75th percentiles).

Table 2. Data from the USI measurements of muscle thickness (mm) at rest and

during ADIM, presented with mean and SD (n=52).

At rest

(n=52)

ADIM

(n=52)

Change

(%)

TrA Left 2.28 (0.96) 4.41 (0.80) 93.4

TrA Right 2.88 (0.93) 4.49 (0.79) 55.9

TrA total abdomen 2.78 (0.93) 4.45 (0.79) 60.0

OI Left 8.02 (0.49) 9.01 (0.76) 12.3

OI Right 8.01 (0.49) 8.50 (0.63) 6.1

OI total abdomen 8.02 (0.49) 8.75 (0.69) 9.2

OE Left 5.91 (0.58) 5.53 (0.65) -6.4

OE Right 5.68 (0.62) 5.71 (0.62) 0.5

OE total abdomen 5.80 (0.60) 5.62 (0.63) -3.1

USI=ultrasound imaging, TrA= m. transversus abdominis, OI= m. obliquus internus, OE= m. obliquus externus, SD= Standard deviation, % change= percentage change from rested state to activation of the deep abdominal muscles during ADIM.

Page 24 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 90: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

25

Table 3. Calculated ratios from ultrasound imaging of lateral abdominal muscles

during preferential activation of TrA

Contraction ratios All n=52

Healthy n=15

LBP n=37

TrA Contraction ratio,

mean (SD)

1.69 (0.18)

∟Left 1.71 (0.24) 1.75 (0.45) 1.69 (0.28)

∟Right 1.67 (0.23) 1.76 (0.45) 1.63 (0.27)

TrA Preferential ratio,

mean (SD)

0.07 (0.01)

∟Left 0.07 (0.04) 0.072 (0.019) 0.076 (0.019)

∟Right 0.07 (0.04) 0.066 (0.003) 0.076 (0.019)

OE+OI Contraction ratio,

mean (SD)

1.05 (0.14)

∟Left 1.047 (0.15) 1.064 (0.27) 1.04 (0.17)

∟Right 1.048 (0.15) 1.057 (0.27) 1.04 (0.17)

TrA=m. transversus abdominis, OI= m. obliquus internus, OE= m. obliquus

externus Ratio (SD)=calculated ratio with standard deviation

Page 25 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 91: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

26

Table 4. Associations between manually-assessed ADIM and USI-assessed, expressed

in TrA-CR ratio and TrA preferential ratio in all subject (n=52) and in subjects by

investigator.

TrA-CR

(r)

TrA Pref. Ratio

(r)

All investigators (n= 52) .09 .12

Investigator 1 (n=30) 0.1 .06

Investigator 2 (n=13) .03 .53

Investigator 3 (n=9) -.38 -.19

TrA-CR Ratio = transversus abdominis contraction ratio, TrA-Pref. Ratio = transversus abdominis preferential activation ratio. Bold represents good agreement.

Table 5 Sensitivity and specificity for the USI-measured TrA contractions ratio

(TrA-CR) and TrA preferential ratios (TrA-Pref. Ratio) and manually-assessed

ADIM; presented with 95% CI.

Sensitivity

(95% CI)

Specificity

(95%CI)

TrA Contraction Ratio 0.35 (0.20-0.53) 0.80 (0.51-0.95)

TrA Preferential Ratio 0.41 (0.25-0.58) 0.80 (0.51-0.95)

ADIM-Manual 0.30 (0.16-0.47) 0.73 (0.44-0.91)

ADIM= Abdominal drawing-in maneuver , CI = 95% confidence interval , TrA

= M. transversus abdominis, USI = Ultrasound imaging

Page 26 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 92: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

Figure 1. Test position for abdominal-drawing-in-maneuver (ADIM) 119x90mm (300 x 300 DPI)

Page 27 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 93: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

Figure 2. Ultrasound image of abdominal muscles, Left= rest Right=activity during abdominal-draw-in-maneuver (ADIM). Distance 1 represents M. transversus abdominis (TrA) distance 2 M. obliquus internus (OI) and distance 3 M. obliquus externus (OE). Note the increase in TrA, OI and the decrease in OE during

the ADIM, picture B. 119x90mm (300 x 300 DPI)

Page 28 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from

Page 94: BMJ Open€¦ · 12 Stockholm, Sweden. 2 Rehabtjänst, S:t Eriksgatan 48, 112 32 Stockholm, Sweden 13 Telephone: +46-8-441 16 60 14 Email: Karsten Kaping* karsten.kaping@rehabtjanst.se;

For peer review only

STARD checklist for reporting of studies of diagnostic accuracy

(version January 2003)

Section and Topic Item

#

On page #

TITLE/ABSTRACT/

KEYWORDS

1 Identify the article as a study of diagnostic accuracy (recommend MeSH

heading 'sensitivity and specificity').

2

INTRODUCTION 2 State the research questions or study aims, such as estimating diagnostic

accuracy or comparing accuracy between tests or across participant

groups.

5

METHODS

Participants 3 The study population: The inclusion and exclusion criteria, setting and

locations where data were collected.

6

4 Participant recruitment: Was recruitment based on presenting symptoms,

results from previous tests, or the fact that the participants had received

the index tests or the reference standard?

6

5 Participant sampling: Was the study population a consecutive series of

participants defined by the selection criteria in item 3 and 4? If not,

specify how participants were further selected.

6

6 Data collection: Was data collection planned before the index test and

reference standard were performed (prospective study) or after

(retrospective study)?

7-9

Test methods 7 The reference standard and its rationale. 7-9

8 Technical specifications of material and methods involved including how

and when measurements were taken, and/or cite references for index

tests and reference standard.

7-9

9 Definition of and rationale for the units, cut-offs and/or categories of the

results of the index tests and the reference standard.

9

10 The number, training and expertise of the persons executing and reading

the index tests and the reference standard.

6-7

11 Whether or not the readers of the index tests and reference standard

were blind (masked) to the results of the other test and describe any

other clinical information available to the readers.

7

Statistical methods 12 Methods for calculating or comparing measures of diagnostic accuracy,

and the statistical methods used to quantify uncertainty (e.g. 95%

confidence intervals).

10

13 Methods for calculating test reproducibility, if done. 7

RESULTS

Participants 14 When study was performed, including beginning and end dates of

recruitment.

15 Clinical and demographic characteristics of the study population (at least

information on age, gender, spectrum of presenting symptoms).

6

16 The number of participants satisfying the criteria for inclusion who did or

did not undergo the index tests and/or the reference standard; describe

why participants failed to undergo either test (a flow diagram is strongly

recommended).

11

Test results 17 Time-interval between the index tests and the reference standard, and

any treatment administered in between.

9

18 Distribution of severity of disease (define criteria) in those with the target

condition; other diagnoses in participants without the target condition.

6

19 A cross tabulation of the results of the index tests (including

indeterminate and missing results) by the results of the reference

standard; for continuous results, the distribution of the test results by the

results of the reference standard.

11-12

20 Any adverse events from performing the index tests or the reference

standard.

11

Estimates 21 Estimates of diagnostic accuracy and measures of statistical uncertainty

(e.g. 95% confidence intervals).

12

22 How indeterminate results, missing data and outliers of the index tests

were handled.

11

23 Estimates of variability of diagnostic accuracy between subgroups of

participants, readers or centers, if done.

11

24 Estimates of test reproducibility, if done. 11-12

DISCUSSION 25 Discuss the clinical applicability of the study findings. 13-14

Page 29 of 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on June 23, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2015-008711 on 9 Decem

ber 2015. Dow

nloaded from