Measurement of lumbar multifidus muscle contraction with rehabilitative ultrasound imaging

(1)

Manual Therapy 12 (2007) 161–166

Original article

Measurement of lumbar multiﬁdus muscle contraction with

rehabilitative ultrasound imaging

Kyle B. Kiesel

a,

, Tim L. Uhl

b

, Frank B. Underwood

c

, Donald W. Rodd

d

, Arthur J. Nitz

a,e

a_{Department of Rehabilitation Sciences, Rehabilitation Sciences Doctoral Program, University of Kentucky College of Health Sciences,}

900 South Limestone, CHS 126, Lexington, KY 40536-0200, USA

b

Department of Rehabilitation Sciences, Division of Athletic Training, University of Kentucky, USA

c

Department of Physical Therapy, University of Evansville, USA

d

Department of Human Kinetics and Sport Studies, University of Evansville, USA

e

Department of Rehabilitation Sciences, Division of Physical Therapy, University of Kentucky, USA

Received 7 March 2005; received in revised form 14 February 2006; accepted 2 June 2006

Abstract

Rehabilitative Ultrasound Imaging (RUSI) has been validated as a noninvasive method to measure activation of selected muscles. The purpose of this study was to determine the relationship between muscle thickness change, as measured by ultrasonography, and electromyography (EMG) activity of the lumbar multiﬁdus (LM) muscle in normal subjects.

Bipolar ﬁne wire electrodes were inserted into the LM at the L4 level of ﬁve subjects. Simultaneous EMG and RUSI data (muscle thickness) were collected while subjects performed increasingly demanding postural response tasks thought to activate the LM muscle. To determine the relationship between muscle thickness change and EMG activity, the normalized EMG data were correlated to normalized RUSI data. To determine if the tasks increased the demand on the LM, the mean EMG data were compared over each of the four tasks.

Muscle thickness change as measured by RUSI was highly correlated with EMG activity of LM in asymptomatic subjects (r¼:79;Po:001).

Mean EMG data showed increasing levels of activation across tasks (19–34% of maximum voluntary isometric contraction (MVIC)). The results of the repeated measures ANOVA demonstrated theses differences were signiﬁcant (F3;12¼25:39;Po:001). Measurement of muscle thickness change utilizing RUSI is a valid and potentially useful method to measure activation of the LM muscle in a narrow range (19–34% of MVIC) in an asymptomatic population.

Keywords:Rehabilitation; Electromyography; Ultrasonography

1. Introduction

Lumbar paraspinal musculature plays a key role in providing stability during dynamic tasks (Cholewicki and McGill, 1996). Of particular interest in the literature of late has been study of the lumbar multiﬁdus (LM) muscle. Several abnormal characteristics have been

identified in the LM in low back pain subjects. Structural abnormalities include histological changes, where fatty infiltrates replace LM muscle tissue (Weber et al., 1997; Zhao et al., 2000; Yoshihara et al., 2003), and atrophy (Kader et al., 2000; Hides et al., 2001). Motor control deficits including altered recruitment patterns (Hodges and Richardson, 1997;Danneels et al., 2002) as well as endurance deficits (Biedermann et al., 1991) have also been reported.

Quantiﬁcation of multiﬁdus activation in those with low back pain may be helpful in determining effective

www.elsevier.com/locate/math

_{Corresponding author. Department of Physical Therapy,} Uni-versity of Evansville, Evansville, IN 47714. Tel.: +812 479 2646; fax: +812 479 2717.

(2)

intervention. The gold standard measurement tool used to assess muscle activation is electromyography (EMG). EMG measures the electrical activity in the muscle and can be interpreted to represent muscle activation. To ensure a reliable signal is obtained from the multiﬁdus, an indwelling electrode should be used (Stokes et al.,

2003). RUSI offers a noninvasive method to measure

muscle activation (Hodges et al., 2003;McMeeken et al., 2004) and has gained popularity in various aspects of low back pain rehabilitation (Hides et al., 1992, 1996, 1998, 2001;Critchley, 2002;Bunce et al., 2004). ‘‘RUSI is a procedure used by physical therapists to evaluate muscle and related soft tissue morphology and function during exercise and physical tasks. RUSI is used to assist in the application of therapeutic interventions aimed at improving neuromuscular function. This includes provid-ing feedback to the patient and physical therapist to improve clinical outcomes. Additionally, RUSI is used in basic, applied, and clinical rehabilitative research to inform clinical practice. Currently, the international community is developing education and safety guidelines in accordance with World Federation for Ultrasound in Medicine and Biology (WFUMB)’’ (Teyhen, 2006). RUSI utilizing high-frequency sound waves to evaluate tissue properties such as thickness. Ultrasound exam-ination is considered low risk. According to the safety committee of the European Committee for

Medical Ultrasound (ECMUS), ‘‘Based on scientific

evidence of ultrasonically induced biological effects to date, there is no reason to withhold scanning for any clinical application’’.

It is known that muscle thickness changes when the muscle is activated (Hodges et al., 2003). The amount of thickness change that occurs with muscle activation can be quantiﬁed with the use of RUSI by comparing resting muscle thickness values to those obtained during muscle activation. Measurement of muscle thickness change compared to EMG activity has been performed on the gastrocnemius muscle (Maganaris et al., 1998), on the transverse abdominis (McMeeken et al., 2004) and on

other trunk and peripheral muscles (Hodges et al.,

2003). To our knowledge this type of study has not been performed on the LM. The purpose of this study is to determine the relationship between muscle thickness change, as measured by RUSI, and EMG activity of the LM muscle in normal subjects.

2. Methods 2.1. Subjects

Five healthy subjects, three of which were female

(mean age¼28.0 years SD 5.6, mean height¼

170.7 cm, SD 9.4, mean weight¼70.3 kg SD 15.9)

volunteered for this study. Subjects were excluded if

they had current or recent history (within 6 months) of LBP or hip pain, a history of lumbar/sacral surgery, congenital lumbar/sacral condition such as spondylo-lithesis, or spina biﬁda, or boney pathology such as a fracture. All volunteering subjects signed an institu-tional-review-board-approved consent form following verbal instructions of the procedure.

2.2. Procedures

Subjects were positioned in prone on a standard plinth. An inclinometer was placed longitudinally over the lumbo/sacral junction and pillows were used to ﬂatten the lumbar curve to less than 101. Subjects were

then oriented to and practiced the maximum voluntary isometric contraction (MVIC) procedure which was performed with the elbows ﬂexed to approximately 901

and shoulders abducted to approximately 1201. Subjects

then lifted their head, trunk, and upper extremities and held with maximum effort against a load applied at the elbow by one of the researchers. The contralateral upper extremity lifting movement, used to activate the multi-ﬁdus, was then practiced. It consisted of the upper extremity lift with four levels of graded resistance as described below.

To study the multifidus muscle, fine wire (California Fine Wire Company, Grover Beach, CA) electrodes were fabricated from pairs of nylon coated 50mm wires which were inserted into a 27 ga hypodermic needle. Approximately 1–2 mm of coating was removed from the tip of the wire, the tips were bent back at 2–3 and 3– 4 mm, respectively, and the needle and wires were sterilized. The L4 spinous process was identified, and the needle was inserted just lateral to the spinous process to the depth of the lamina, then withdrawn, leaving the electrode in the deepest portion of the LM muscle. A surface ground electrode was placed over the subject’s lateral malleolus.

The ultrasound images were generated at 25 Hz

utilizing computerized ultrasonography (Sonosite

180plus, Sonosite Inc, Seattle, WA). The primary investigator operated the ultrasound unit and did all the scanning for this study. A 70 mm 5-MHz curvilinear transducer was placed longitudinally along the spine with the mid-point over the L4 spinous process. It was moved laterally and angled slightly medially until the L4/5 zygapophyseal joint could be identiﬁed ( Richard-son et al., 1999) (Fig. 1). This scan point is directly over the LM multiﬁdus and a measurement from this landmark to the plane between the muscle and subcutaneous tissue was used for the linear

measure-ment of the LM (Richardson et al., 1999) at rest

(Fig. 2a) and during activation (Fig. 2b). An on-screen caliper was used to obtain the resting measurement (Fig. 2a and b), captured simultaneously with resting EMG data. The reliability of capturing LM images and

(3)

measuring them on the screen was established in a pilot study of eight asymptomatic subjects (ICC3,1¼.85).

Subsequent images taken during the arm lifting tasks were saved and printed for off-screen manual measure-ment. The reliability of the manual measurement was calculated by having a second researcher, blinded to the grade of activation, measure the muscle thickness. This

represents reliability of measuring the same image

between raters and was found to be excellent

(ICC3,1¼.95). The muscle thickness measurements

obtained during each task were normalized to the resting measurement and the percent change from rest was calculated (ActivityRest/Rest100). This percen-tage change in muscle thickness from rest to activation represented muscle activation as measured by RUSI.

The MVIC data were collected as the subject performed the maximum upper extremity and trunk lift as described above. Two trials of 5 s each were performed and the greatest root mean square (RMS) peak amlitude was used to normalize the EMG activity. Normalization provides a standard reference of elec-trical activity and all data are reported as a percentage of the MVIC (Bamman et al., 1997).

The contralateral arm lifting tasks were performed in the same plane as the MVIC’s. The subjects were instructed to lift their extremity straight up off of the table approximately 5 cm and hold for approximately 8 s which allowed enough time to capture two images of the contracted LM muscle. Two trials each of the four levels of increasingly demanding upper extremity lifting tasks were performed while EMG data and the images were obtained simultaneously. The ﬁrst level (no load) had resistance of only the limb; the next three levels (low, medium, and high load) had graded resistance

based on the subject’s body weight (Table 1). The

average of the two trials for each task was used for

Fig. 2. (a) Ultrasound image of the lumbar multiﬁdus muscle at the L4/5 level with caliper measurement of the muscle thickness at rest. (b) Ultrasound image of the lumbar multiﬁdus muscle activation during the contralateral arm lift.

Fig. 1. Position of transducer for imaging of the lumbar multiﬁdus muscle.

Table 1

Graded resistance levels for upper extremity lifting tasks in kilograms

Subject mass (kg) Low Medium High

o68.2 .45 .68 .90

68.2–79.5 .45 .68 1.14

79.5–90.9 .45 .90 1.14

490.9 .45 .90 1.36

Table 2

Mean and standard deviation values for the lumbar multiﬁdus muscle during rest and each of the lifting task conditions

Rest No load Low Medium High

Ultrasound Mean 2.48 3.28 3.50 3.60 3.68 SD .19 .35 .29 .33 .29 EMG Mean na 19.50 25.31 32.21 34.31 SD na 5.94 7.15 7.58 8.85

Ultrasound values are thickness measured in centimeters and EMG values are expressed as a percent of MVIC.

(4)

analysis. See Table 2 for mean ultrasound and EMG data.

2.3. EMG analysis

The EMG data were sampled at 2000 Hz using the Biopac II Student Lab Pro (Biopac System, Inc. Santa

Barbara, CA) ampliﬁed 1000 and ﬁltered at 30–

500 Hz. The Biopac has a signal-to-noise ratio of

490 dB and an input impedance of 1.0 MO. The data

were saved and imported to a PC for analysis with Datapac software (Run Technologies, Mission Viejo, CA). RMS peak amplitudes were calculated for each .5 s period. Data from the middle three seconds of each trial was averaged and expressed as a percent of MVIC. The average of the two trials for each task was used for analysis.

2.4. Statistical analysis

To determine if a correlation exists between the EMG and RUSI data points, a Pearson’s correlation coefﬁ-cient was calculated.

To determine if the individual tasks adequately increased muscle activation, a repeated measures analy-sis of variance with a Bonferroni post hoc analyanaly-sis (alevel .05) was performed on the EMG data.

3. Results

The results of the Pearson’s product moment correla-tion coefﬁcients revealed that LM muscle thickness change as measured by RUSI correlated highly (r¼:79;

Po:001) with LM EMG activity (Fig. 3).

The repeated measures ANOVA for the EMG data demonstrated that the tasks studied were signiﬁcantly different from each other (F3;12 ¼25:39;Po:001). Post

hoc analysis revealed signiﬁcant differences between the no-load and medium and high load tasks and between low load and high load tasks.

4. Discussion

Our main ﬁnding was that a high correlation exists between EMG activity and thickness change in the LM muscle during typical contractions. This adds to the limited body of knowledge related to the use of RUSI as a measurement tool for muscle activation. We demon-strated a linear relationship between thickness change in the LM muscle and EMG activity during the graded contralateral upper extremity lifting tasks. Previous research assessing the relationship between muscle thickness change and EMG activity in the transverses abdominis muscle utilized volitional activation matched

to percent of MVIC (Hodges et al., 2003; McMeeken

et al., 2004) through a large range of activation levels. The tasks we chose produced average activation from 19% to 34% of MVIC, a narrow range. Analysis of variance showed a signiﬁcant difference between tasks, with post hoc testing demonstrating a difference between no-load and medium and high load tasks and between low load and high load. Although not statistically signiﬁcant, the difference between the no load and low task was 5%, which is consistent with increases between levels of activation in previously cited studies of the transverse abdominis. Isolated volitional activation of the LM is discussed in the literature (Hides et al., 1996), studying subjects trained to perform this activity may be a method for future research to study a broader range of activation levels.

Direct comparison of our EMG results is not possible as we did not locate earlier studies that isolated EMG activity of the multifidus during contralateral limb movement in the prone position. Arokoski et al. in two separate papers (Arokoski et al., 1999, 2001) studied a variety of stabilization exercises and reported an average 41% MVIC for the LM during a standing, alternating shoulder flexion movement with an average load of 1.5 kg. Our average load across each task was .8 kg, which produces an average output of 28% of MVIC. Despite these methodological differences, re-search to measure multifidus activity during various lumbar stabilization exercises, involving loaded limb movements, has shown somewhat similar activation levels to our study.

Previous studies measuring thickness change and EMG activity of other muscles have reported conﬂicting results.Hodges et al. (2003)compared EMG activity to architectural change measured by RUSI in several

Fig. 3. Correlation of EMG activity and thickness change of lumbar multiﬁdus as measured by RUSI during each of the four graded activation tasks.

(5)

muscles across a broad range of activation levels. This study measured thickness change and EMG activity of the tibialis anterior, biceps brachii, brachialis, internal oblique and transverse abdominis and reported a curvilinear relationship where RUSI could detect changes at low levels of contraction (up to approxi-mately 20% of MVIC) and higher levels of contraction produce little further thickness change.McMeeken et al. (2004) measured the transverse abdominis during abdominal hollowing from 5% to 80% of MVIC and demonstrated a linear relationship between thickness change and EMG activity across all levels of activation measured.

Our methods differed somewhat from similar research in that we did not match a volitional contraction to a set level of activation; rather we chose tasks thought to activate the LM at progressively greater levels. This resulted in measurement in a narrow range of muscle activation and is a limitation of our study. We cannot assume a linear relationship exists across the entire range of muscle activation as we tested a narrow range. We demonstrated RUSI can detect changes in LM EMG activity from and average of 19% of MVIC (no load) to of 34% of MVIC (high load).

Further research is needed to determine if RUSI is a valid measure of LM activation across a greater range of activation levels, and in those with low back pain. If RUSI can be validated as a noninvasive measurement of LM muscular activity in the low back pain population, this may be useful for clinicians who use therapeutic exercise as an intervention in this population. RUSI could be used to measure potential LM activation impairment and how various interventions effect the impairment.

5. Conclusion

These results provide preliminary data on the potential use of RUSI in the measurement of LM muscle activation. The measurement of muscle thickness change utilizing RUSI is valid noninvasive method to measure activation of the LM muscle as it is highly correlated with EMG in a limited range (19–34% of MVIC) in an asymptomatic population.

Acknowledgments

The authors would like to thank the University of Evansville’s Undergraduate Research Committee for funding this study. The authors would like to thank ProRehab PC in Evansville, IN for the use of the sonography equipment used in this study.

References

Arokoski JP, Kankaanpaa M, Valta T, Juvonen I, Partanen J, Taimela S, et al. Back and hip extensor muscle function during therapeutic exercises. Archives of Physical Medicine and Rehabilitation 1999;80(7):842–50.

Arokoski JP, Valta T, Airaksinen O, Kankaanpaa M. Back and abdominal muscle function during stabilization exercises. Archives of Physical Medicine and Rehabilitation 2001;82(8): 1089–98.

Bamman MM, Ingram SG, Caruso JF, Greenisen MC. Evaluation of surface electromyography during maximal voluntary contrac-tion. Journal of Strength and Conditioning Resident 1997;11: 68–72.

Biedermann HJ, Shanks GL, Forrest W, JInglis J. Power spectrum analyses of electromyographic activity. Discriminators in the differential assessment of patients with chronic low-back pain. Spine 1991;16(10):1179–84.

Bunce SM, Hough AD, Moore AP. Measurement of abdominal muscle thickness using M-mode ultrasound imaging during functional activities. Manual Therapy 2004;9(1):41–4.

Cholewicki J, McGill SM. Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Clincal Biomechanics 1996;11(1):1–15.

Critchley D. Instructing pelvic ﬂoor contraction facilitates transversus abdominis thickness increase during low-abdominal hollowing. Physiotherapy Research International 2002;7(2):65–75.

Danneels LA, Coorevits PL, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE. Differences in electromyographic activity in the multiﬁdus muscle and the iliocostalis lumborum between healthy subjects and patients with sub-acute and chronic low back pain. European Spine Journal 2002;11(1):13–9.

Hides J, Richardson C, Stokes M. Diagnostic ultrasound imaging for measurement of lumbar multiﬁdus muscle in normal young adults. Physiotherapy Theory and Practice 1992;8:19–26.

Hides J, Richardson C, Jull G. Multiﬁdus muscle recovery is not automatic after resolution of acute, ﬁrst-episode low back pain. Spine 1996;21(23):2763–9.

Hides J, Richardson C, Jull G. Use of real-time ultrasound imaging for feedback in rehabilitation. Manual Therapy 1998;3(3):125–31. Hides J, Jull G, Richardson C. Long-term effects of speciﬁc stabilizing

exercises for ﬁrst-episode low back pain. Spine 2001;26(11): E243–8.

Hodges PW, Richardson CA. Contraction of the abdominal muscles associated with movement of the lower limb. Physical Therapy 1997;77(2):132–42.

Hodges PW, Pengel LH, Herbert RD, Gandevia SC. Measurement of muscle contraction with ultrasound imaging. Muscle and Nerve 2003;27(6):682–92.

Kader DF, Wardlaw D, Smith FW. Correlation between the MRI changes in the lumbar multiﬁdus muscles and leg pain. Clincal Radiology 2000;55(2):145–9.

Maganaris CN, Baltzopoulos V, Sargeant AJ. In vivo measurements of the triceps surae complex architecture in man: implications for muscle function. Journal of Physiology 1998;512(Pt 2): 603–14.

McMeeken JM, Beith ID, Newham DJ, Milligan P, Critchley DJ. The relationship between EMG and change in thickness of transversus abdominis. Clinical Biomechanics 2004;19(4):337–42.

Richardson C, Jull G, Hodges P, Hides JA. Therapeutic exercise for spinal segmental stabilization in low back pain. Ednburgh, London, New York, Philadelphia, Sydney and Toronto: Churchill Livingstone; 1999. p. 117 [Chapter 8].

Stokes IA, Henry SM, Single RM. Surface EMG electrodes do not accurately record from lumbar multiﬁdus muscles. Clinical Biomechanics 2003;18(1):9–13.

(6)

Teyhen DS. Rehabilitative ultrasound imaging symposium. Journal of Orthopaedic & Sports Physical Therapy 2006;36(8):A1–A17. Weber BR, Grob D, Dvorak J, Muntener M. Posterior surgical

approach to the lumbar spine and its effect on the multiﬁdus muscle. Spine 1997;22(15):1765–72.

Yoshihara K, Nakayama Y, Fujii N, Aoki TIto H. Atrophy of the multiﬁdus muscle in patients with lumbar disk herniation:

histochemical and electromyographic study. Orthopedics 2003; 26(5):493–5.

Zhao WP, Kawaguchi Y, Matsui H, Kanamori M, Kimura T. Histochemistry and morphology of the multiﬁdus muscle in lumbar disc herniation: comparative study between diseased and normal sides. Spine 2000;25(17):2191–9.