Ultrasonographic Measurements of Paraspinal Muscle Thickness in Adolescent Idiopathic Scoliosis: A Comparison and Reliability Study : Pediatric Physical Therapy

Secondary Logo

Journal Logo

Advanced Search
RESEARCH ARTICLE

Ultrasonographic Measurements of Paraspinal Muscle Thickness in Adolescent Idiopathic Scoliosis

A Comparison and Reliability Study

Zapata, Karina A. PT, DPT, PhD; Wang-Price, Sharon S. PT, PhD; Sucato, Daniel J. MD, MS; Dempsey-Robertson, Molly MD

Author Information

Physical Therapy (Dr Zapata), Orthopedics (Dr Sucato), and Radiology (Dr Dempsey-Robertson), Texas Scottish Rite Hospital for Children, Dallas, Texas; School of Physical Therapy (Dr Wang-Price), Texas Woman's University, Dallas, Texas.

Correspondence: Karina A. Zapata, PT, DPT, PhD, Texas Scottish Rite Hospital for Children, 2222 Welborn Street, Dallas, TX 75219 ([email protected]).

At the time the work reported in the article was undertaken, Karina A. Zapata was a PhD student in physical therapy at Texas Woman's University, Dallas, TX.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.pedpt.com).

The authors declare no conflicts of interest.

Pediatric Physical Therapy 27(2):p 119-125, Summer 2015. | DOI: 10.1097/PEP.0000000000000131

Purpose: 

To examine paraspinal muscle asymmetries in patients with mild adolescent idiopathic scoliosis (AIS) compared with controls using ultrasound (US) imaging and to determine reliability of the US imaging.

Methods: 

Twenty adolescents (10 with and 10 without AIS) whose muscle thickness of the deep paraspinals at T8 and the multifidus at L1 and L4 vertebral levels was determined using US imaging.

Results: 

Ultrasound imaging had good intrarater reliability (ICC3,3 = 0.83-0.99) and interrater reliability (ICC2,3 = 0.93-0.99). Significantly greater muscle thicknesses were found on the concave side at T8 and L1 compared with the corresponding convex side. Significantly greater normalized muscle thickness values were found on the concave side at T8, L1, and L4 in patients with mild AIS compared with controls.

Conclusions: 

Ultrasound imaging is reliable for measuring the deep thoracic paraspinals and lumbar multifidus in adolescents with and without idiopathic scoliosis. Significant paraspinal muscle asymmetries were found in mild AIS.

For more insights from the authors, access Supplemental Digital Content 1, at https://links.lww.com/PPT/A76.

INTRODUCTION

One possible contributing cause of adolescent idiopathic scoliosis (AIS) is continued asymmetric vertebral loading that leads to curve progression, especially during rapid growth.1 Paraspinal muscles, which stabilize the spine, are believed to be affected considerably by this asymmetric loading.2,3 Therefore, conservative treatments such as scoliosis-specific exercises and trunk rotational strengthening exercises are designed to increase paraspinal muscle symmetry with the goals of improving spinal stability and preventing curve progression in AIS.4,5 Although researchers have not yet demonstrated which specific paraspinal muscles are affected most in AIS, the lumbar multifidus and/or deep paraspinal muscles are of special interest because of their anatomical proximity to the spine and their function in spinal stability.4,6,7

Ultrasound (US) imaging technology is being used increasingly to study deep layers of muscle such as the lumbar multifidus. Ultrasound imaging is noninvasive compared with fine-wire electromyography and less expensive than magnetic resonance imaging (MRI). The reliability and validity of US imaging have been established for quantifying muscle morphology of the lumbar multifidus muscle in adults.8–11 In particular, muscle thickness measurements with parasagittal views are more reliable and easier to interpret than cross-sectional area measurements with transverse views.9,12 Paraspinal muscle symmetry is expected in a population without spinal deformity, and the lumbar multifidus thickness has not been found to be significantly different between sides in healthy adults without back pain or pathology.12 When comparing differences between sides in adolescents with idiopathic scoliosis (IS), Kennelly and Stokes13 reported asymmetries of the lumbar multifidus muscle for different curve types using US imaging. However, a wide range of skeletal maturity and curve severity was included. Therefore, these findings cannot be generalized to skeletally immature adolescents with mild curves. Knowledge of paraspinal muscle mass asymmetries may improve the understanding and efficacy of conservative treatments in AIS.

To date, the parasagittal view of US imaging has not been used to study the lumbar multifidus muscle in adolescents. In addition, we were interested in studying thoracic multifidus muscle because the thoracic spine is affected most often in AIS.14 However, US imaging has not been used to quantify the thoracic multifidus muscle because this muscle is too small to isolate from other deep thoracic paraspinals.13 Recognizing this limitation, we grouped the deepest muscle layer of the thoracic spine including the thoracic multifidus, semispinalis, and rotator muscles, and referred to them collectively as the deep thoracic paraspinals. Therefore, the primary purpose of this study was to compare the difference in muscle thickness of the deep thoracic paraspinals at T8 and lumbar multifidus at L1 and L4 between the concave and the convex sides in adolescents with mild IS as well as between the right and left sides in adolescents without IS (between-side comparisons). The second purpose was to compare normalized muscle thickness at the 3 above mentioned levels in adolescents with mild IS to that of adolescents without IS using US imaging (between-group comparisons). The third purpose was to determine the reliability of ultrasonographic measurements of muscle thickness at the 3 vertebral levels in adolescents with and without IS.

METHODS

The study comprised a comparison and reliability study. Informed assent from both participants and legal guardians was obtained. The Institutional Review Board at the affiliated hospital approved this study.

Participants

Twenty 10 to 12 year-old participants (mean age [SD], 11.6 [1.0] years), 10 adolescents with mild IS (AIS group) and 10 adolescents without IS (control group), were recruited from the affiliated hospital. The criteria for participants in the AIS group included a Cobb angle of 15° to 24° on the posteroanterior radiograph,15 no prior treatment for scoliosis, Risser grades 0 to 1,16,17 a thoracic, double (thoracic and lumbar), lumbar or thoracolumbar curve type, and no transitional vertebrae.18 Risser grades along with triradiate cartilage status and menarchal status are used clinically to estimate skeletal and sexual immaturity and have been shown to be indicative of risk of curve progression.16,19,20 Risser grades and triradiate cartilage status were collected from the radiographs of the participants in the AIS group. A Risser grade is based on the ossification of the iliac apophysis on a scale of 0 (no ossification) to 5 (fused ossified apophysis).17 Risser grades 0 to 1 indicate no to minimal ossification of the iliac apophysis, signifying skeletal immaturity.16,17 The triradiate cartilage status of the acetabulum is another radiographic index of maturity and typically closes before Risser 1 and menarche.20 The inclusion criteria for the control group consisted of participants who failed a school screening for scoliosis, but who demonstrated no scoliosis by radiograph (n = 3), and volunteers from the community deemed healthy who were evaluated by a spine deformity surgeon and demonstrated no scoliosis by clinical examination including the Adam's forward bend test (n = 7).21 Radiographs were not performed for those 7 volunteers because we did not wish to expose them to unnecessary radiation. Participants were excluded from both the AIS and control groups if they had a history of back injury or back pain, abdominal surgery, limb length discrepancy more than 2 cm, or were pregnant. Body mass index (BMI) has been associated with muscle size and was computed by dividing the participant's weight by the square of their height (kg/m2).10,22 Body mass index was used to normalize muscle thickness (muscle thickness/BMI, cm/kg/m2) for between-group comparisons.

Ultrasound Imaging

Ultrasound images were generated using a US machine (ACUSON, Mountain View, California) and a 1 to 4Hz curved array transducer. The curved transducer was used because it fit over the contour of the spinal curve and the area between the spinous process and the transverse process where the paraspinals are located. The lower frequency was chosen, because the lower frequency is ideal for imaging the deep muscle layer of paraspinals that is directly above the zygapophyseal joints of the spine.23 A sonographer (S.Z.) with over 20 years' experience, who was blinded to the groups, took the US images of all participants.

Prior to US imaging, the principal investigator (K.Z.) marked T8, L1, and L4 by palpating the spinous processes when participants were lying in the prone position. Because of potential errors during palpation of the spinous processes of participants with spinal deformity, metal markers were placed at these 3 levels on participants with AIS before taking standard standing radiographs. The radiograph was used to determine the Cobb angle, indicating the severity of scoliosis. Also, the marks displayed on the radiographs allowed for potential relabeling of each vertebral level to ensure correct landmark placement for US imaging. Participants were imaged in 2 positions: lying prone relaxed with arms by the sides, followed by standing upright relaxed with arms by the side.8 In the prone position, a pillow was placed under the pelvis to minimize thoracic kyphosis at T8 and lumbar lordosis at L1 and L4. To ensure a neutral position, an inclinometer was used to determine that the spine was within 10° of horizontal.8

Three US images were taken on 1 side of the spine for both right and left sides at 3 vertebral levels (T8, L1, and L4) for 2 positions (prone relaxed and standing relaxed). The right and left sides were later categorized as either the concave or convex side for the AIS group, or the dominant or nondominant hand for the control group. A standard US imaging protocol was adopted and used for parasagittal imaging of the deep thoracic paraspinal and the lumbar multifidus muscles.24 For each US image, the sonographer placed the transducer centrally and longitudinally over the spinous process and then moved laterally to locate the zygapophyseal joints with the transducer angled 30° to 45° toward the spinous processes. The sonographer imaged the multifidus muscle at L4, then L1, then the deep thoracic paraspinals at T8 with the participant first in the prone position, and then with the participant in the standing position. For each segment of each side, 3 images were taken. The average of 3 trials was used for data analysis. To determine intrarater reliability, the same sonographer (S.Z.) conducted 2 testing sessions, at least 10 minutes apart, for each participant. A radiologist (M.D-R.) blinded to group assignment performed postcapture ultrasonographic measurements of muscle thickness for both testing sessions to determine the intrarater reliability of the US image acquisitions. To determine the interrater reliability of ultrasonographic measurements of muscle thickness, the radiologist (M.D-R.) and the principal investigator (K.Z.), who was educated in reading US images, performed off-line analysis independently and measured muscle thickness of the lumbar multifidus muscles and the deep thoracic paraspinals for all participants. Electronic calipers were used to determine the muscle thickness measurement, the length from the most posterior portion of the zygapophyseal joints to the inner edge of the fascia between the lumbar multifidus/deep thoracic paraspinals and superficial tissue following a reported protocol (Figure 1).8,11 In participants with AIS, each side of the spine at each level was labeled as concave or convex, depending on the curve type.

F1-5
Fig. 1:
Ultrasonographic measurements of multifidus muscle thickness in the parasagittal view at L4 in the prone (left image) and standing (right image) positions. F = fascia; M = muscle tissue; Z = zygapophyseal joint.

SPSS version 18 (SPSS Inc, Chicago, Illinois) was used to analyze the collected data. Intraclass correlation coefficients were used to determine the intrarater reliability (ICC3,3) and inter-rater reliability (ICC2,3) between sides (concave and convex side for the scoliosis group, right and left sides for the control group), and at each vertebral level. When comparing the difference between sides, side-difference percentages were calculated by dividing the muscle thickness measurement of the expected larger side (the concave side in the AIS group and the dominant hand in the control group) by the muscle thickness measurement of the expected smaller side. For statistical analysis, 2-tailed paired t tests were used to compare the differences between sides for both groups, respectively. When comparing the difference between the 2 groups, normalized data were calculated by dividing muscle thickness measurement by BMI. Two-tailed independent t tests were used to examine differences between the 2 groups. Additional independent t tests were performed to compare the difference in age and BMI between the 2 groups. Significance was defined as α = .05.

RESULTS

Participant Characteristics

Characteristics of participants are listed in Table 1, including age, gender, ethnicity, BMI, and menarchal status. Radiographic characteristics such as the status of the triradiate cartilage, Risser grade, curve magnitude, and type are also reported in Table 1. Independent t tests showed no significant differences in age or BMI between the groups.

T1-5
TABLE 1:
Characteristics of Participants

Ultrasonographic Measurements of Muscle Thickness

Table 2 lists the deep thoracic paraspinal muscle thickness at T8 and multifidus muscle thickness in centimeters at L1 and L4 for all participants, regardless of side, and separately by concave and convex sides for the scoliosis group and by right and left sides for the control group. For both the scoliosis and control groups, the ultrasonographic measurements show that the multifidus muscle thickness at L4 was greater than the thickness of the multifidus muscle at L1 (P < .001) and the deep thoracic paraspinals at T8 (P ≤ .001) for both the prone and standing positions and for both sides. No muscle thickness differences were found between the deep thoracic paraspinals at T8 and the lumbar multifidus muscle thickness at L1 for the control group. However, for the scoliosis group, the muscle thickness at T8 was smaller on the concave side in the prone position (P = .04) and larger on the convex side in the standing position (P = .02). The average muscle thickness was significantly larger in the standing position than in the prone position for all 3 segments (P ≤ .04) except at T8 on the concave side for the scoliosis group where no significant difference was found.

T2-5
TABLE 2:
Ultrasonographic Measurements (Mean ± SD) of Deep Thoracic Paraspinals for T8 and Multifidus Muscle Thickness for L1 and L4 (cm) in the Relaxed Prone and Standing Positions

Between-Side Comparisons

Significant differences were found when comparing the concave to convex sides of the AIS group. The muscle thickness was significantly greater on the concave side in the relaxed prone position at T8 (P = .03) and L1 (P = .04), but not at L4. Side-difference percentages in the relaxed prone condition averaged 6.9% ± 0.7% at T8, 5.7% ± 0.8% at L1, and 3.2% ± 0.9% at L4. However, no significant side-differences were found in relaxed standing, averaging 4.0% ± 11.7% at T8, 5.3% ± 15.2% at L1, and 1.3% ± 5.3% at L4 (Figure 2). When comparing the dominant with nondominant sides in the control group, no significant differences were found. Side-difference percentages in the control group averaged 0.2 ± 1.0% in the relaxed prone condition and 0.2 ± 1.3% in standing position.

F2-5
Fig. 2:
Ultrasonographic measurements of deep thoracic paraspinals for T8 and multifidus muscle thickness for L1 and L4 for the concave and convex sides in the scoliosis group. Asterisks indicate statistical significance.

Between-Group Comparisons

Differences were found when comparing the scoliosis group with the control group. We compared the concave side of the curve in the scoliosis group with the dominant hand side in the control group, since these sides were expected to be the larger of the 2 sides. The muscle thickness was significantly greater on the concave side of the curve compared with the control's dominant side in the relaxed prone position at T8 (P = .05), L1 (P = .04), and L4 (P = .005). In relaxed standing, the muscle thickness on the concave side of the curve in the AIS group was significantly greater than the dominant side in the control group at T8 (P = .04) and approached significance at L1 (P = .055). No significant difference at L4 was found in the standing position (P = .75). Also, no significant differences were found between the convex side of the curve in the AIS group and the nondominant hand side in the control group in the prone and standing positions.

Reliability of Ultrasonographic Measurements

The results showed overall good intrarater reliability of US measurements of muscles thickness for all 3 vertebral levels on both sides and for both the relaxed prone and standing positions (ICC3,3 = 0.83-0.99) (Table 3). However, intrarater reliabilities of the AIS group were lower (ICC3,3 = 0.83-0.98) than those of the control group (ICC3,3 = 0.94-0.99).

T3-5
TABLE 3:
Intrarater Reliability (ICC3,3 ± 95% CI) of Ultrasonographic Measurements of Deep Thoracic Paraspinals for T8 and Multifidus Muscle Thickness for L1 and L4 in the Relaxed Prone and Standing Positions
T4-5
TABLE 4:
Interrater Reliability (ICC2,3 ± 95% CI) of Ultrasonographic Measurements of Deep Thoracic Paraspinals for T8 and Multifidus Muscle Thickness for L1 and L4 in the Relaxed Prone and Standing Positions

The results also showed good interrater reliability for all 3 vertebral levels on both sides and positions (ICC2,3 = 0.93-0.99) (Table 4).

DISCUSSION

The results showed good intrarater and interrater reliabilities of muscles thickness measurements from US images. However, the lower intrarater reliabilities of the AIS group compared with the control group indicate that structural changes from spinal deformity make imaging more difficult and less reliable.

The results also showed significant differences in the deep thoracic paraspinal and lumbar multifidus muscle thickness (ie, muscle asymmetries) on the concave side of the curve in adolescents with IS. These findings strengthen the hypothesis that changes in muscle morphology occur on the concave side.2,13 The muscle thickness was significantly larger on the concave side than on the convex side in the prone relaxed position at T8 and L1, but not at L4. The lack of a significant difference between sides at L4 may be because only 1 participant had a curve apex within 1 level of L4, rendering changes in muscle size less likely. In contrast, 8 participants had curve apexes within 1 level of L1. Although only 3 curves spanning the thoracic spine had apexes within 1 level of T8, 7 of the 10 participants had primary curves in the thoracic spine that may have contributed to greater muscle asymmetries at T8. Future studies should image participants at the apex of the curve to clarify whether or not muscle asymmetries are most notable at the curve apex.

Contrary to the results in the prone position in the AIS group, side differences were not found in the standing position. We speculate that adolescents with IS may be contracting the deep thoracic paraspinals and lumbar multifidus muscles differently to stabilize the spine in a gravity-resisted position. However, muscle thickness was greater in the standing position than in the prone position for all 3 segments in both groups. This finding is consistent with the results of other studies, suggesting an increased muscle demand of the multifidus in an upright position.25 Also, a large variation of muscle thickness for the standing position was likely due to individual differences in muscle contractions, depending on factors such as curve type and apex location. Participants may have used different muscle activation strategies to adjust for abnormal loading of the spine.26

Kennelly and Stokes13 also found larger muscles on the concave side of the curve in the prone position in double and thoracolumbar curves. Comparisons could not be made to their findings for thoracic curves because they only analyzed muscle size at L4. Their study also included older ages (12-19 years) and larger curves (13°-53°), whereas our study was limited to ages 10 to 12 years and curves with Cobb angles of 15° to 24°. Side-difference percentages in the Kennelly and Stokes13 study averaged 13.9% ± 6.4%, which is much greater than the percentage difference at L4 of 3.2% ± 0.9% in our study. Their higher between-side asymmetries may be due to including participants with larger curves and from measuring cross-sectional area instead of muscle thickness. When using MRI, Chan et al2 similarly found abnormality in the multifidus muscle at the concave side of the apex of the curve in AIS. Hyper-intense signals were noted at the apex of the concave side of the curve, and signal intensity increased with curve severity. Possible explanations for increased signal intensities included changes in muscle fiber type and chronic multifidus overuse. No significant side differences were found in the control group (adolescents without IS), corroborating other studies that reported symmetrical muscles in individuals deemed healthy.12 In addition, the average values of multifidus muscle thickness of the control group may provide preliminary normative data for adolescents.

In this study, we examined deep thoracic paraspinal muscle thickness, which had not been previously reported in any patient population. The deep thoracic paraspinal thickness was generally found to be similar to the thickness of the lumbar multifidus at L1, but both were much smaller than the thickness of the lumbar multifidus at L4. During the imaging analysis, we had difficulty isolating the thoracic multifidus muscle from other deep thoracic paraspinal muscles. This observation was consistent with Kennelly and Stokes,13 who indicated that the thoracic multifidus muscle was too small to quantify.

Ultrasound imaging may provide a reliable and objective method for further understanding the pathophysiology of AIS. Since muscle asymmetries were found in mild curves of adolescents with IS at high risk of progression, it can be hypothesized that muscle asymmetry may play a role in the development or the progression of these curves. One limitation of the study is that MRI was not used to validate US imaging measurements. Because of the 3-dimensional nature of scoliosis, parasagittal images may not have captured the entire muscle asymmetry. Further studies should investigate the effect of nonoperative treatments such as bracing and exercise, on muscle asymmetry, since little is understood regarding muscle mass and strength changes in scoliosis, especially in adolescents with a high likelihood of curve progression.

CONCLUSIONS

Ultrasound imaging is reliable for quantifying muscle thickness of the deep thoracic paraspinals and lumbar multifidus in adolescents with and without IS in the relaxed prone and standing positions. Significant paraspinal muscle asymmetries in mild curves of adolescents at risk of curve progression indicate the need to examine whether nonoperative treatments can change these imbalances.

ACKNOWLEDGMENT

The authors thank Shuzhen Zhang, RD, MS, for performing the US imaging.

REFERENCES

1. Stokes IAF. Mechanical effects on skeletal growth. J Musculoskel Neuron Interact. 2002;2(3):277–280.
2. Chan YL, Cheng JCY, Guo X, et al. MRI evaluation of multifidus muscles in adolescent idiopathic scoliosis. Pediatr Radiol. 1999;29(5):360–363.
3. Gonyea WJ, Moore-Woodard C, Moseley B, et al. An evaluation of muscle pathology in idiopathic scoliosis. J Pediatr Orthop. 1985;5(3):323–329.
4. Romano M, Minozzi S, Zaina F, et al. Exercises for adolescent idiopathic scoliosis. Cochrane Database Syst Rev. 2012;8:CD007837.
5. Mordecai SC, Dabke HV. Efficacy of exercise therapy for the treatment of adolescent idiopathic scoliosis: a review of the literature. Eur Spine J. 2012;21(3):382–389.
6. McIntire KL, Asher MA, Burton DC, Liu W. Treatment of adolescent idiopathic scoliosis with quantified trunk rotational strength training: a pilot study. J Spinal Disord Tech. 2008;21(5):349–358.
7. Koppenhaver SL, Hebert JJ, Fritz JM, et al. Reliability of rehabilitative ultrasound imaging of the transversus abdominis and lumbar multifidus muscles. Arch Phys Med Rehabil. 2009;90(1):87–94.
8. Hides JA, Richardson CA, Jull GA. Magnetic resonance imaging and ultrasonography of the lumbar multifidus muscle: comparison of two different modalities. Spine. 1995;20(1):54–58.
9. Hebert JJ, Koppenhaver SL, Parent EC, Fritz JM. A systematic review of the reliability of rehabilitative ultrasound imaging for the quantitative assessment of the abdominal and lumbar trunk muscles. Spine. 2009;34(23):E848–E856.
10. Stokes M, Rankin G, Newham D. Ultrasound imaging of lumbar multifidus muscle: normal reference ranges for measurements and practical guidance on the technique. Man Ther. 2005;10(2):116–126.
11. Wallwork TL, Hides JA, Stanton WR. Intrarater and interrater reliability of assessment of lumbar multifidus muscle thickness using rehabilitative ultrasound imaging. J Orthop Sports Phys Ther. 2007;37(10):608–612.
12. Stokes M, Hides J, Elliott J, et al. Rehabilitative ultrasound imaging of the posterior paraspinal muscles. J Orthop Sports Phys Ther. 2007;37(10):581–595.
13. Kennelly KP, Stokes MJ. Pattern of asymmetry of paraspinal muscle size in adolescent idiopathic scoliosis examined by real-time ultrasound imaging: a preliminary study. Spine. 1993;18(7):913–917.
14. Lenke LG, Betz RR, Clements D, et al. Curve prevalence of a new classification of operative adolescent idiopathic scoliosis: does classification correlate with treatment? Spine. 2002;27:604–611.
15. Cobb JR. Outline for the study of scoliosis. In: Blount WR, Banks SW, eds. The American Academy of Orthopedic Surgeons Instructional Course Lectures. Vol. 2. Ann Arbor, MI: Edwards; 1948:261–275.
16. Risser JC. The iliac apophysis; an invaluable sign in the management of scoliosis. Clin Orthop. 1958;11:111–119.
17. Manring MM, Calhoun J., Joseph C. Risser Sr., 1892-1992. Clin Orthop Relat Res. 2010;468:643–645.
18. Tini PG, Wieser C, Zinn WM. The transitional vertebra of the lumbosacral spine: its radiological classification, incidence, prevalence, and clinical significance. Rheumatol Rehabil. 1977;16:180–185.
19. Sanders JO, Browne RH, Cooney TE, Finegold DN, McConnell SJ, Margraf SA. Correlates of the peak height velocity in girls with idiopathic scoliosis. Spine. 2006;31(20):2289–2295.
20. Sanders JO, Little DG, Richards BS. Prediction of the crankshaft phenomenon by peak height velocity. Spine. 1997;22(12):1352–1357.
21. Adams W. Lectures on the Pathology and Treatment of Lateral and Other Forms of Curvature of the Spine. London: Churchill; 1865.
22. Hides J, Cooper DH, Stokes MJ. Diagnostic ultrasound imaging for measurement of the lumbar multifidus muscle in normal young adults. Physiother Theor Pract. 1992;8:19–26.
23. Whittaker JL, Teyhen DS, Elliott JM, et al. Rehabilitative ultrasound imaging: understanding the technology and its applications. J Orthop Sports Phys Ther. 2007;37(8):434–449.
24. Kiesel KB, Uhl TL, Underwood FB, et al. Measurement of lumbar multifidus muscle contraction with rehabilitative ultrasound imaging. Man Ther. 2007;12(2):161–166.
25. Lee S, Chan CK, Lam T, et al. Relationship between low back pain and lumbar multifidus size at different postures. Spine. 2006;31(19):2258–2262.
26. Stokes IAF, Gardner-Morse M. Muscle activation strategies and symmetry of spinal loading in the lumbar spine with scoliosis. Spine. 2004;29(19):2103–2107.
Keywords:

adolescent; female; humans; idiopathic scoliosis; lumbar vertebra/pathology; male; muscles pathology; muscles/ultrasonography; reliability; scoliosis/etiology; scoliosis/pathology; scoliosis/ultrasonography; thoracic vertebrae/pathology

Supplemental Digital Content

Copyright © 2015 Academy of Pediatric Physical Therapy of the American Physical Therapy Association