Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
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
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
For peer review only
2
Data sharing statement: additional data are available from Karsten Kaping, 1
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
For peer review only
21
Data Sharing: Additional data is available from Karsten Kaping,
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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