SPINE Volume 42, Number 15, pp E876–E882

ß 2017 Wolters Kluwer Health, Inc. All rights reserved.

RANDOMIZED TRIAL

Effects of Low-Load Motor Control Exercises

and a High-Load Lifting Exercise on Lumbar
Multifidus Thickness
A Randomized Controlled Trial

Lars Berglund, PhD, RPT,,y,{ Björn Aasa, MSc, RPT,y,z

Peter Michaelson, PhD, RPT,§ and Ulrika Aasa, PhD, RPT ,{

with the large side in both intervention groups, without

Study Design. Randomized controlled trial.
influence of pain at baseline, or change in pain intensity.
Objective. The aim of this study was to compare the effects of
Conclusion. At baseline, there was a difference in thickness of
low-load motor control (LMC) exercises and a high-load lifting
the LM muscles between sides. It seems that exercises focusing
(HLL) exercise, on lumbar multifidus (LM) thickness on either
on spinal alignment may increase the thickness of the LM
side of the spine and whether the effects were affected by pain
muscles on the small side, irrespective of exercise load. The
intensity or change in pain intensity.
increase in LM thickness does not appear to be mediated by
Summary of Background Data. There is evidence that
either current pain intensity or the magnitude of change in pain
patients with low back pain (LBP) may have a decreased size of
intensity.
the LM muscles with an asymmetry between sides in the lower
Key words: asymmetry, deadlift, exercise, low back pain,
back. It has also been shown that LMC training can affect this
morphology, motor control, movement control, pain intensity,
asymmetry. It is, however, not known whether a high-load physical therapy, resistance training, stabilization training,
exercise has the same effect. ultrasonography multifidus.
Methods. Sixty-five participants diagnosed with nociceptive Level of Evidence: 2
mechanical LBP were included and randomized into LMC Spine 2017;42:E876–E882
exercises or a HLL exercise, the deadlift. The LM thickness was
measured using rehabilitative ultrasound imaging (RUSI), at

T
baseline and after a 2-month training period. he lumbar multifidus (LM) muscles are important for
Results. There were no differences between interventions controlling the magnitude of motions in the lumbar
regarding effect on LM muscle thickness. However, the analysis spine.1 It has been shown that patients with low back
showed a significant effect for asymmetry. The thickness of the pain (LBP) might have impairments related to LM muscle
LM muscle on the small side increased significantly compared size.2–4 More specifically, imaging studies have reported
smaller LM muscle cross-sectional area and muscle thick-
ness2,4–7 and more fatty infiltration6,8,9 in patients with
From the Department of Community Medicine and Rehabilitation, Umeå
University, Sweden; yDepartment of Surgical and Perioperative Sciences,
LBP. Also, it seems that patients with LBP may display
Umeå University, Sweden; zNorrlandskliniken Health Care Centre, Umeå, an asymmetry of LM size between the left and right sides4,10
Sweden; §Division of Health and Rehabilitation, Department of Health and that a greater ratio of asymmetry may be associated
Science, Luleå University of Technology, Sweden; and {Umeå School of
Sport Sciences, Umeå University, Sweden.
with higher pain intensity levels.11 Further, it seems import-
Acknowledgment date: August 17, 2015. First revision date: February 12,
ant to investigate which exercises can affect size and sym-
2016. Second revision date: June 10, 2016. Third revision date: August 19, metry between sides since it has been shown that exercises
2016. Fourth revision date: October 21, 2016. Acceptance date: November targeting the LM muscles can reduce recurrence of LBP.12
2, 2016.
However, there are also studies that indicate that asymmetry
The manuscript submitted does not contain information about medical
device(s)/drug(s).
between sides is frequent in both symptomatic and asymp-
Visare Norr and Norrbottens County Council grant funds were received in
tomatic populations.13
support of this work. Previously, only two exercise approaches have aimed to
Relevant financial activities outside the submitted work: employment. restore size and symmetry of LM muscles. First, low-load
Address correspondence and reprint requests to Lars Berglund, PhD, RPT, motor control (LMC) exercises,14 which have been shown to
Department of Community Medicine and Rehabilitation, Umeå University, reduce pain intensity with concomitant improvements
SE-901 87 Umeå, Sweden; E-mail: [email protected]
in LM size and symmetry in patients with LBP.10 Second,
DOI: 10.1097/BRS.0000000000001989 high-load resistance training of the back extensors15 have
E876 www.spinejournal.com August 2017

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

been evaluated in three studies on patients with LBP (NCT01061632).20 Here, we investigated the effects on
where in Willemink et al,16 patients showed a reduction percentage change [(follow-up  baseline/baseline)  100]
in disability, without a change in LM size, while Steele in thickness of the LM muscle at the fifth lumbar vertebra
et al17 showed reduced disability and pain intensity, and at the small and large sides. The study protocol was
Danneels et al18 showed an increase of the size of the LM approved by The Regional Ethical Review Board in Umeå
muscles. Notably, Steele et al17 did not evaluate LM size (No. 09–200M).
and Danneels et al18 did not evaluate pain intensity or
disability. In conclusion, both exercise approaches seem Participants
effective in accordance with present evidence stating that Consecutive individuals seeking care for LBP with a
most exercise interventions have positive and clinically duration of more than 3 months at two occupational health
relevant effects on pain and disability.19 However, there care centers, and classified as having nociceptive mechanical
are no studies comparing low-load exercises to high-load LBP21 were screened for eligibility (n ¼ 85).20 The eligibility
exercises regarding their effects on size and/or symmetry screening controlled for inclusion and exclusion criteria
of the LM muscles and no studies that have evaluated such as red (i.e., pathological processes) and yellow flags
the influence of perceived pain intensity on LM (i.e., pain and pain behavior secondary to a dominance of
thickness change. psycho-social factors),22 and assured that the dominant
A recent study from our research group showed that in 70 underlying pain pattern was of nociceptive mechanical
patients with nociceptive mechanical LBP, both LMC exer- character.21 Participants who agreed to participate in reha-
cises and a high-load lifting (HLL) exercise, the barbell bilitative ultrasound imaging (RUSI) of LM thickness were
deadlift, in combination with education, lead to clinically included. Detailed description of the selection process has
relevant improvements in pain intensity for a majority of the been presented earlier.20 Baseline characteristics of the
participants in the study.20 Still, neither the thickness of the participants are presented in Table 1.
LM muscles, nor the effects of LMC and HLL on LM
thickness was reported. Therefore, the primary aim of this Procedure
study was, to investigate the LM thickness and compare After the selection process, participants answered a
the effects of LMC exercises and a HLL exercise, on LM questionnaire and participated in RUSI measurements.
thickness among patients with nociceptive mechanical Thereafter, the randomization procedure was performed
LBP. Also, a secondary aim was to investigate whether by an investigator who was blinded to the characteristics
changes in LM thickness was affected by baseline or change of the participants. Participants were stratified by means
in pain intensity. of age and sex, and randomly allocated to intervention
groups by a computer generated procedure.20 A second
MATERIALS AND METHODS investigator thereafter contacted each participant giv-
ing times for first appointment. The physical therapist
Study Design (PT) performing the RUSI measurements, after the inter-
This study is part of a larger data collection evaluating vention period, was blinded to baseline data, but not
the effects of LMC exercises and an HLL exercise group allocation.

TABLE 1. Baseline Values of Age, BMI, VAS 7 Days at Baseline and Change in VAS 7 Days from
Baseline to Follow-Up, Displayed as Mean and Standard Deviation
Age BMI VAS Baseline VAS Change
LMC n ¼ 33 43.3 (10.3) 25.0 (3.0) 48.4 (27.0) 18.5 (26.7) n ¼ 31
HLL n ¼ 32 42.3 (9.8) 25.4 (3.8) 41.3 (23.8) 19.0 (25.5) n ¼ 29
P 0.67 0.57 0.27 0.95
Men
All n ¼ 28 45.6 (9.2) 25.1 (3.3) 38.8 (22.8) 14.1 (20.8) n ¼ 27
LMC n ¼ 13 48.0 (7.3) 25.0 (3.0) 39.8 (24.2) 10.5 (19.7) n ¼ 12
HLL n ¼ 15 43.6 (10.4) 25.1 (3.6) 37.9 (22.4) 17.1 (21.9) n ¼ 15
P 0.21 0.97 0.83 0.43
Women
All n ¼ 37 40.6 (10.1) 25.3 (3.6) 49.6 (26.8) 22.5 (29.2) n ¼ 33
LMC n ¼ 20 40.3 (10.9) 24.9 (3.1) 54.0 (27.8) 23.6 (29.7) n ¼ 19
HLL n ¼ 17 41.1 (9.4) 25.8 (4.1) 44.4 (25.4) 21.0 (29.6) n ¼ 14
P 0.82 0.48 0.28 0.80

P values for independent-samples t test.
BMI indicates body mass index; HLL, high-load lifting intervention group; LMC, low-load motor control intervention group; VAS, visual analogue scale.

Spine www.spinejournal.com E877

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

Interventions Rehabilitative Ultrasound Imaging

The interventions were carried out with a maximum of 12 Ultrasound imaging of the thickness of LM muscles at both
sessions over 2-months provided by senior PTs that have used sides of the fifth lumbar vertebra was conducted by a PT
motor control exercises and/or the deadlift exercise in their certified in ultrasound imaging using an EsaoteMyLab 25
clinical practice for at least 5 years. Education regarding basic Gold scanner with a 10–12 mHz linear probe. The partici-
aspects of pain and tissue mechanisms with emphasis on how pant being assessed was positioned on the examination table
non-ideal movement and alignment strategies could contrib- in a pain-free, prone and relaxed position with both hands
ute to the participants’ LBP was also provided for the partici- under the belly and both index fingers on the navel. The
pants within each intervention. A detailed description of both participants were first instructed to contract the muscles of
interventions has been provided earlier.20 the trunk and thereafter relax, to ensure a measurement of
the LM muscles in a relaxed state. The LM muscle was
High-Load Lifting Exercise imaged in a transverse section allowing visualization of the
The deadlift exercise was the only exercise in the HLL thoracolumbar fascia, muscle bulk, and transverse process.
intervention. During training, the PT emphasized instruc- The LM diameter was measured using on-screen calipers,
tions to ensure that the participants maintained a neutral from the highest point of the thoracolumbar fascia in a
alignment of the low back as well as activation of the straight line to the vertebral lamina/transverse process of the
stabilizing muscle of the lumbar spine while lifting and segment. To increase the intra-rater reliability, at the first
lowering the barbell from the floor. The initial sessions RUSI measurement the exact points of measurement along
focused on establishing a proper technique with relatively with anatomical landmarks and typical characteristics such
low loads (10–20 kg) at three to five sets of 10 repetitions. as scars and birthmarks were marked on the lower back and
Thereafter the progression started with the goal of increas- thereafter traced onto a transparent sheet. At the second
ing the total number of kilograms lifted compared with the measurement, the same transparent sheet was placed over
previous session in order to stimulate hypertrophy and the lumbar spine, thus enabling accurate relocation of the
maximal strength. This was done by increasing number imaging site. As per earlier studies4,13 the larger and smaller
of sets to five to eight per session while repetitions were sides (right or left) were determined to assess the thickness of
reduced to three to five per set with increased weight on the the LM muscle.
bar. The exact progression was controlled by the PT super-
vising the training in communication with each participant. Data Analysis
Baseline values of age, body mass index (BMI), pain inten-
Low-Load Motor Control Exercises sity, and LM thickness were compared between the two
The LMC intervention was performed using exercises with- intervention groups and between men and women using
out external resistance. The exercises were tailored to the independent samples t tests. Paired samples t test was used to
participants’ movement impairments and aberrant muscle compare baseline values of the small side versus large side
activation patterns.22 Static and dynamic exercises focused for men and women, respectively.
on encouraging movement patterns relieving stress from Linear mixed model (LMM) analysis was used to analyze
their area of pain and improving control and timing of the effect of the intervention. For missing data, the last value
the magnitude of movements in adjacent areas to the low carried forward (i.e., percentage change equals zero) was
back. Specific training of muscle activation of local stabil- applied. The dependent variable was percentage change in
izers of the trunk was employed throughout the intervention LM thickness. Fixed factors were Group (LMC, HLL),
as an integral part of all aspects of the movement behavior Asymmetry (small side, large side), and Sex (male, female)
retraining. The strategy for progressing the training was to and the interactions for GroupAsymmetry, SexAsymmetry,
increase the difficulty of exercises from a coordinative and SexGroup. Since no previous study has investigated the
standpoint where training began at a basic level where relation between LM thickness and pain intensity in regard to
the exercises targeted the participants’ ability to find and the effects of exercise interventions, separate analyses were
maintain their lumbar neutral position while performing performed including: 1) baseline pain intensity and 2) pain
movements with arms or legs. Thereafter, the training was intensity change (baseline to follow-up) as covariates in each
progressed to consist of exercises including dynamic move- LMM. BMI was also included as a covariate as suggested in
ments of the spine in activities the participants had reported previous studies.10,24
difficulty in performing. A P value <0.05 was considered significant for
the analyses.
Measurements
RESULTS
Questionnaire Participant attrition throughout the study is presented in
Information about age, sex, height, and weight was col- Figure 1. None of the participants in the LMC group
lected at baseline. Pain intensity during the last 7 days reported adverse effects, but one participant withdrew from
(100 mm visual analogue scale [VAS 7 days])23 was the study due to stomach surgery. In the HLL group, two
measured at baseline and at follow-up. participants dropped out; one due to adverse effects and one
E878 www.spinejournal.com August 2017

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

Figure 1. Flow of participants: low-load motor con-

trol exercises versus a high-load lifting exercise.

without giving an explanation. Mean attendance in the intervention groups for the baseline values of the small or
LMC group was 6.0 (2.0) sessions and 11.0 (2.7) in the large side of the LM muscles. Men had a significantly
HLL group. In the HLL group, the training intensity pro- (P ¼ 0.03) greater LM thickness on the large side and a
gressed from 20.0 (min–max 10.0–35.0) to 90.0 (55–200) near significantly (P ¼ 0.06) greater LM thickness on the
kg for men, and from 20.0 (10–20) kg to 55.0 (15–102.5) small side at baseline compared with women. There was a
kg for women. significant (P < 0.001) difference in thickness of the LM
There were no significant differences between the LMC muscle between the small and large side for both men
and HLL groups for baseline values (Table 1). The values for and women. This asymmetry ([thickness on large side -
LM thickness and percentage change are described in  thickness on small side]/thickness on large side100)10
Table 2. There were no significant differences between was 9.3% for men and 8.8% for women.

TABLE 2. Unadjusted Values (Mean and Standard Deviation) of the Lumbar Multifidus Thickness
(cm) at Baseline and at 2-Months Follow-Up, and Percentage Change from Baseline to
Follow-Up
n ¼ Baseline/ Large Side Large Side Large Side Small Side Small Side Small Side
Follow-Up Baseline Follow-Up Change (%) Baseline Follow-Up Change (%)
LMC n ¼ 33/32 2.70 (0.44) 2.68 (0.48) 0.41 (17.96) 2.46 (0.42) 2.62 (0.50) 7.95 (20.89)
HLL n ¼ 32/29 2.58 (0.47) 2.66 (0.60) 1.67 (14.13) 2.35 (0.45) 2.65 (0.51) 11.18 (18.08)
Men
All: n ¼ 28/28 2.78 (0.43) 2.80 (0.56) 0.71 (15.73) 2.52 (0.45) 2.66 (0.53) 6.38 (17.75)
LMC: n ¼ 13/13 2.83 (0.34) 2.71 (0.49) -4.14 (16.33) 2.59 (0.36) 2.62 (0.51) 1.35 (17.75)
HLL: n ¼ 15/15 2.73 (0.50) 2.87 (0.63) 4.91 (14.43) 2.46 (0.52) 2.70 (0.56) 10.73 (17.13)
Women
All: n ¼ 37/33 2.54 (0.45) 2.56 (0.49) 1.27 (16.54) 2.31 (0.41) 2.61 (0.48) 11.94 (20.60)
LMC: n ¼ 20/19 2.61 (0.48) 2.65 (0.48) 3.37 (18.75) 2.37 (0.45) 2.62 (0.50) 12.24 (22.10)
HLL: n ¼ 17/14 2.45 (0.40) 2.45 (0.50) 1.19 (13.64) 2.25 (0.37) 2.59 (0.46) 11.59 (19.39)

Number of participants out of total 65 at baseline and 61 at follow-up.
HLL indicates high-load lifting intervention group; LMC, low-load motor control intervention group.

Spine www.spinejournal.com E879

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

For the LMM with baseline pain intensity and BMI as receptive to short-term hypertrophic changes34 than the
covariates, the analysis for percentage change in LM thick- large side, as shown by Hides et al.10
ness showed a significant effect of Asymmetry (effect esti- Regarding differences between sexes, the HLL exercise
mate 9.58, 95% confidence interval [CI] 4.00–15.15, might be more effective than LMC exercises to increase LM
P ¼ 0.001) in favor of the small side. There were no thickness in men than in women. However, this observed
significant effects for Group (P ¼ 0.522), Sex (P ¼ 0.349) difference (Table 2) was not significant. We recommend
nor for the interactions (GroupAsymmetry, P ¼ 0.497, future studies to include a higher number of both men and
SexAsymmetry, P ¼ 0.144 and SexGroup, P ¼ 0.162), or women to find out whether this is due to us having a limited
the covariates baseline pain intensity (P ¼ 0.529) or BMI sample size.
(P ¼ 0.385). The LMM analyses showed that neither baseline pain
For the LMM with pain intensity change and BMI as intensity nor pain intensity change influenced LM thickness
covariates, the analysis for percentage change in LM change at the L5 level following interventions, as previously
thickness showed a significant effect of Asymmetry (effect indicated.11 We did not find these results surprising, since
estimate 9.58, 95% CI 4.00–15.15, P ¼ 0.001) in favor of the experience of LBP is complex and cannot be expected to
the small side. There were no significant effects for depend solely on a biological change of a tissue structure.35
Group (P ¼ 0.495), Sex (P ¼ 0.357) nor for the interactions
(GroupAsymmetry, P ¼ 0.497, SexAsymmetry, P ¼ 0.144 Methodological Considerations
and SexGroup, P ¼ 0.178), or the covariates pain intensity When using RUSI, the thickness of the LM muscle can only
change (P ¼ 0.411) or BMI (P ¼ 0.336). be measured in a posterior-anterior direction, in comparison
with magnetic resonance imaging (MRI), which measures
DISCUSSION area. When a recent study compared RUSI with MRI, poor
The results showed that LM thickness on the small side to moderate correlation between measurements were
increased significantly more over time compared with the found.36 However, RUSI measurements have commonly
large side following both interventions and that neither been used2,7,37,38 and the fact that earlier studies37,38 present
baseline pain intensity, nor change in pain intensity, affected similar values of LM thickness speaks in favor of the validity
change in LM thickness. For the LMC intervention, the of our measurements.
increase in LM muscle size on the small side is in accordance Second, since patients with nociceptive LBP seem to
with the study by Hides,10 who found that cricket players experience less pain and disability than patients with other
with LBP showed decreased LM asymmetry after 6 weeks of dominating pain pattern,39 the results may not be applicable
motor control training. The finding that a high-load exercise for other subgroups of LBP.39
could have the same effect on the LM muscles has not been Third, there was a significant difference in number of
previously shown, although at least two previous studies sessions attended between the intervention groups. The
have investigated this. In the study by Willemink et al,16 reason for this was mainly a compliance issue; the partici-
resistance training of the back extensors affected self-rated pants in the LMC group did not want to attend all 12
function, but not LM size. In a study by Danneels et al18 a sessions as they considered themselves adequately rehabili-
combination of strengthening and motor control exercises tated after about six sessions, which might have influenced
was superior in increasing the size of the LM muscles, when the effect on LM thickness for the LMC group.
compared to strengthening exercises or motor control exer-
cises, alone. Notably, Danneels et al18 did not regard asym- CONCLUSION
metry; instead, used the sum of the cross-sectional areas of The LM thickness in patients with nociceptive mechanical
the left and right LM muscles as outcome measure. We LBP seems to be similar to healthy individuals and patients
believe that the reason why the deadlift exercise was effec- with non-specific LBP. However, there was a difference in
tive, is that the participants were instructed to keep their thickness of the LM muscle between sides. It seems that
spine in a neutral position during the lift25,26 and to perform exercises focusing on spinal alignment, and thereby facili-
Valsalva maneuver27 in combination with abdominal brac- tating activation of stabilizing muscles, may increase the
ing28 prior to the lifts in order to increase intra-abdominal thickness of the LM muscles on the small side and reduce
pressure,27 lumbar stability,29 and coactivation of all trunk LM asymmetry between sides, irrespective of exercise load.
muscles.30 It has earlier been suggested that maintaining a Moreover, an increase in LM thickness does not appear to
neutral lumbar position during the lift would optimize be mediated by either current pain intensity or the magni-
biomechanical performance25 and activate31 the lumbar tude of change in pain intensity.
muscles. In fact, abdominal bracing has been shown to elicit
a significant activation of the LM muscles in various pos-
itions.28 For both interventions it is possible that the neuro-
muscular stimuli32 on the small LM muscle was sufficient Key Points
enough to achieve hypertrophy. This is especially so if we
regard the asymmetry between sides as a consequence of The present study is the first to investigate the
effects of exercise on LM thickness in patients
disuse of the small LM muscle,33 and therefore being more
E880 www.spinejournal.com August 2017

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

17. Steele J, Bruce-Low S, Smith D, et al. A randomized controlled trial

with nociceptive mechanical LBP as their of limited range of motion lumbar extension exercise in chronic
dominating pain pattern. low back pain. Spine 2013;38:1245–52.
Both low-load motor control exercises and a high- 18. Danneels LAVG, Cambier DC, Witvrouw EE, et al. Effects of three
load lifting exercise can reduce asymmetry of the different training modalities on the cross sectional area of the
lumbar multifidus muscle in patients with chronic low back pain.
lumbar multifidus muscles by increasing the Br J Sports Med 2001;35:186–91.
thickness of the small side. 19. Searle A, Spink M, Ho A, et al. Exercise interventions for the
No association between changes in LM thickness treatment of chronic low back pain: a systematic review and meta-
and pain intensity was found. analysis of randomised controlled trials. Clin Rehab 2015;
29:1155–67.
20. Aasa B, Berglund L, Michaelson P, et al. Individualized low-load
motor control exercises and education versus a high-load
lifting exercise and education to improve activity, pain intensity,
References and physical performance in patients with low back pain: a
1. Wilke HJ, Wolf S, Claes LE, et al. Stability increase of the lumbar randomized controlled trial. J Orthop Sports Phys Ther 2015;
spine with different muscle groups. A biomechanical in vitro study. 45:77–85.
Spine 1995;20:192–8. 21. Smart KM, Blake C, Staines A, et al. Mechanisms-based classifi-
2. Wallwork TL, Stanton WR, Freke M, et al. The effect of chronic cations of musculoskeletal pain: part 3 of 3: symptoms and signs of
low back pain on size and contraction of the lumbar multifidus nociceptive pain in patients with low back (R/S leg) pain. Man
muscle. Man Ther 2009;14:496–500. Ther 2012;17:352–7.
3. Hides JA, Stokes MJ, Saide M, et al. Evidence of lumbar multifidus 22. O’Sullivan P. Diagnosis and classification of chronic low back pain
muscle wasting ipsilateral to symptoms in patients with acute/ disorders: maladaptive movement and motor control impairments
subacute low back pain. Spine 1994;19:165–72. as underlying mechanism. Man Ther 2005;10:242–55.
4. Hides J, Gilmore C, Stanton W, et al. Multifidus size and symmetry 23. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures
among chronic LBP and healthy asymptomatic subjects. Man Ther for chronic pain clinical trials: IMMPACT recommendations. Pain
2008;13:43–9. 2005;113:9–19.
5. Danneels LA, Vanderstraeten GG, Cambier DC, et al. CT imaging 24. Sitilertpisan P, Hides J, Stanton W, et al. Multifidus muscle size
of trunk muscles in chronic low back pain patients and healthy and symmetry among elite weightlifters. Phys Ther Sport
control subjects. Eur Spine J 2000;9:266–72. 2012;13:11–5.
6. Parkkola R, Rytokoski U, Kormano M. Magnetic resonance 25. McGill SM, Hughson RL, Parks K. Changes in lumbar lordosis
imaging of the discs and trunk muscles in patients with chronic modify the role of the extensor muscles. Clin Biomech (Bristol,
low back pain and healthy control subjects. Spine 1993;18:830–6. Avon) 2000;15:777–80.
7. Dickx NCB, Parlevliet T, Lavens A, et al. The effect of unilateral 26. Dankaerts W, O’Sullivan P, Burnett A, et al. Altered patterns of
muscle pain on recruitment of the lumbar multifidus during superficial trunk muscle activation during sitting in nonspecific
automatic contraction. An experimental pain study. Man Ther chronic low back pain patients: importance of subclassification.
2010;15:364–9. Spine 2006;31:2017–23.
8. Kjaer P, Bendix T, Sorensen JS, et al. Are MRI-defined fat infiltra- 27. Hackett DA, Chow CM. The Valsalva maneuver: its effect on
tions in the multifidus muscles associated with low back pain? intra-abdominal pressure and safety issues during resistance exer-
BMC Med 2007;5:2. cise. J Strength Cond Res 2013;27:2338–45.
9. Mengiardi B, Schmid MR, Boos N, et al. Fat content of lumbar 28. Matthijs OC, Dedrick GS, James CR, et al. Co-contractive acti-
paraspinal muscles in patients with chronic low back pain and in vation of the superficial multifidus during volitional preemptive
asymptomatic volunteers: quantification with MR spectroscopy. abdominal contraction. PM R 2014;6:13–21.
Radiology 2006;240:786–92. 29. Grenier SG, McGill SM. Quantification of lumbar stability by
10. Hides JA, Stanton WR, McMahon S, et al. Effect of stabilization using 2 different abdominal activation strategies. Arch Phys Med
training on multifidus muscle cross-sectional area among young Rehab 2007;88:54–62.
elite cricketers with low back pain. J Orthop Sports Phys Ther 30. Maeo S, Takahashi T, Takai Y, et al. Trunk muscle activities
2008;38:101–8. during abdominal bracing: comparison among muscles and exer-
11. Huang Q, Zhang Y, Li D, et al. The evaluation of chronic low back cises. J Sports Sci Med 2013;12:467–74.
pain by determining the ratio of the lumbar multifidus muscle 31. Dankaerts W, O’Sullivan P, Burnett A, et al. Altered patterns of
cross-sectional areas of the unaffected and affected sides. J Phys superficial trunk muscle activation during sitting in nonspecific
Ther Sci 2014;26:1613–4. chronic low back pain patients: importance of subclassification.
12. Hides JA, Jull GA, Richardson CA. Long-term effects of specific Spine 2006;31:2017–23.
stabilizing exercises for first-episode low back pain. Spine 32. Kristensen J, Franklyn-Miller A. Resistance training in musculos-
2001;26:E243–8. keletal rehabilitation: a systematic review. Br J Sports Med
13. Niemelainen R, Briand MM, Battie MC. Substantial asymmetry in 2012;46:719–26.
paraspinal muscle cross-sectional area in healthy adults questions 33. Steele J, Bruce-Low S, Smith D. A reappraisal of the deconditioning
its value as a marker of low back pain and pathology. Spine (Phila hypothesis in low back pain: review of evidence from a triumvirate
Pa 1976) 2011;36:2152–7. of research methods on specific lumbar extensor deconditioning.
14. Richardson C, Hodges PW, Hides J, et al. Therapeutic Exercise for Curr Med Res Opin 2014;30:865–911.
Lumbopelvic Stabilization: A Motor Control Approach for the 34. Stevens JE, Pathare NC, Tillman SM, et al. Relative contributions
Treatment and Prevention of Low Back Pain, 2nd ed. Edinburgh, of muscle activation and muscle size to plantarflexor torque
New York: Churchill Livingstone; 2004. during rehabilitation after immobilization. J Orthop Res 2006;
15. Steele J, Bruce-Low S, Smith D. A review of the clinical value of 24:1729–36.
isolated lumbar extension resistance training for chronic low back 35. Parkin-Smith GF, Amorin-Woods LG, Davies SJ, et al. Spinal pain:
pain. PM R 2015;7:169–87. current understanding, trends, and the future of care. J Pain Res
16. Willemink MJ, van Es HW, Helmhout PH, et al. The effects of 2015;8:741–52.
dynamic isolated lumbar extensor training on lumbar multifidus 36. Belavy DL, Armbrecht G, Felsenberg D. Real-time ultrasound
functional cross-sectional area and functional status of patients measures of lumbar erector spinae and multifidus: reliability
with chronic nonspecific low back pain. Spine (Phila Pa 1976) and comparison to magnetic resonance imaging. Physiol Meas
2012;37:E1651–8. 2015;36:2285–99.

Spine www.spinejournal.com E881

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 RANDOMIZED TRIAL Exercise and Lumbar Multifidus Thickness  Berglund et al

37. Wallwork TL, Hides JA, Stanton WR. Intrarater and interrater 39. Smart KM, Blake C, Staines A, et al. Self-reported
reliability of assessment of lumbar multifidus muscle thickness pain severity, quality of life, disability, anxiety and
using rehabilitative ultrasound imaging. J Orthop Sports Phys depression in patients classified with ’nociceptive’,
Ther 2007;37:608–12. ’peripheral neuropathic’ and ’central sensitisation’ pain. The
38. Wong AY, Parent EC, Kawchuk GN. Reliability of 2 ultrasonic discriminant validity of mechanisms-based classifications
imaging analysis methods in quantifying lumbar multifidus thick- of low back (R/Sleg) pain. Man Ther 2012;17:
ness. J Orthop Sports Phys Ther 2013;43:251–62. 119–25.

E882 www.spinejournal.com August 2017

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.