Journal of Electromyography and Kinesiology 24 (2014) 253–257
Contents lists available at ScienceDirect
Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin
Differential activation of parts of the latissimus dorsi with various isometric shoulder exercises Se-yeon Park a, Won-gyu Yoo b,⇑ a b
Department of Physical Therapy, The Graduate School, Inje University, Republic of Korea Department of Physical Therapy, College of Biomedical Science and Engineering, Inje University, 607 Obangdong, Gimhae, Gyeongsangnam-do 621-749, Republic of Korea
a r t i c l e
i n f o
Article history: Received 6 August 2013 Received in revised form 16 October 2013 Accepted 10 December 2013
Keywords: Electromyography Isometric exercise Selective activation Shoulder movement
a b s t r a c t As no study has examined whether the branches of the latissimus dorsi are activated differently in different exercises, we investigated intramuscular differences of components of the latissimus dorsi during various shoulder isometric exercises. Seventeen male subjects performed four isometric exercises: shoulder extension, adduction, internal rotation, and shoulder depression. Surface electromyography (sEMG) was used to collect data from the medial and lateral components of the latissimus dorsi during the isometric exercises. Two-way repeated analysis of variance with two within-subject factors (exercise condition and muscle branch) was used to determine the significance of differences between the branches, and which branch was activated more with the exercise variation. The root mean squared sEMG values for the muscles were normalized using the modified isolation equation (%Isolation) and maximum voluntary isometric contraction (%MVIC). Neither the %MVIC nor %Isolation data differed significantly between muscle branches, while there was a significant difference with exercise. %MVIC was significantly higher with shoulder extension, compared to the other isometric exercises. There was a significant correlation between exercise condition and muscle branch in the %Isolation data. Shoulder extension and adduction and internal rotation increased %Isolation of the medial latissimus dorsi more than shoulder depression. Shoulder depression had the highest value of %Isolation of the lateral latissimus dorsi compared to the other isometric exercises. Comparing the medial and lateral latissimus dorsi, the medial component was predominantly activated with shoulder extension, adduction, and internal rotation, and the lateral component with shoulder depression. Shoulder extension is effective for activating the latissimus dorsi regardless of the intramuscular branch. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction In the musculoskeletal system, certain broad muscles with many insertions and origins are thought to have different kinematic functions according to the fiber arrangement (Paton and Brown, 1995; Brown et al., 2007). The latissimus dorsi originates from the spinous processes of the last six thoracic vertebrae, thoracolumbar fascia, and iliac crest, and inserts into the intertubercular groove of the humerus (Kendall et al., 2005). Since the latissimus dorsi has many neuromuscular end plates and vascular and nerve branches (Watanabe et al., 2010), it used as a donor sites for reconstructive surgery (de Oliveira et al., 2010; Kwon et al., 2011). Surgical methods for preventing morbidity of the latissimus dorsi, such as splitting the thoracodorsal nerve or sparing an area through vascular anatomy, have been suggested (Kwon et al., 2011). However, the contributions of the intramuscular branches of the latissimus dorsi to shoulder exercises have not been clearly ⇑ Corresponding author. Tel.: +82 55 320 3994; fax: +82 55 329 1678. E-mail address: [email protected] (W.-g. Yoo). 1050-6411/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jelekin.2013.12.004
established, although these differences are important for the prognosis after surgery and rehabilitation. Movement at a joint related to muscular activation can be investigated using surface electromyography (sEMG). When using a sEMG, the anthropometric position of the electrode sites is important for reducing spatial variability and measuring an accurate sEMG signal (Hug, 2011). The surface electrode sites for the latissimus dorsi have differed in previous studies: some used a site lateral to T9 over the muscle belly, while others used sites lateral and inferior to the inferior angle of the scapulae (Signorile et al., 2002; Snyder and Leech, 2009; Vera-Garcia et al., 2010). These sites might be differentiated as the medial and lateral branches of the latissimus dorsi, but a few studies have investigated both medial and lateral electrode sites of the latissimus dorsi simultaneously. To our knowledge, two studies have examined the functional differences of the fibers of this broad muscle with changes in the direction of exercises and shoulder position (Paton and Brown, 1995; Brown et al., 2007). Zhao et al. (2003) suggested that the latissimus dorsi can be divided into medial and lateral branches, and that the lateral branch is activated more with shoulder motion.
254
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Although the muscle induces many shoulder actions, including internal rotation, adduction, and extension of the humerus, and depression of the shoulder girdle, activation depending on branch differentiation has not been specifically reported. This preliminary study investigated the functional differences in the intramuscular branches of the latissimus dorsi during various isometric shoulder exercises. Based on a report that the lateral branch was activated more with shoulder motion (Zhao et al., 2003), we expected that shoulder exercises would activate the lateral branch of the latissimus dorsi more than the medial branch. 2. Method 2.1. Study population Twenty asymptomatic males were recruited from a local university using convenience sampling. They had a mean ± SD age of 23.38 ± 1.02 (range 22–26) years and weight-trained twice per week. Subjects with a history of upper extremity pain or discomfort in the past 6 months were excluded. Tightness of the latissimus dorsi was examined before the experiment. Subject lay on a table in the supine position, and was asked to flex his arm fully while remaining in contact with the table (Sahmann, 2002). Three subjects whose lumbar region rose off the table, indicating a tight latissimus dorsi, were excluded from this study. The final study sample comprised 17 males, with a mean height of 175.38 ± 4.88 cm and weighing 66.28 ± 4.75 kg. All subjects were right-hand dominant, with dominance defined as the hand used for writing. Ethics approval for this study was obtained from the Inje University Faculty of Health Sciences Human Ethics Committee. The participants provided informed consent. 2.2. Instrumentation Surface electromyography data were collected using a Trigno wireless system (Delsys, Boston, MA, USA); the Trigno electrodes (Delsys) were set at a band pass of 20–450 Hz and a common mode rejection ratio of 80 dB. The 27 37 15-mm sensor has four fixed 5 10-mm contact areas, which are half the area of a Bagnoli sensor (10 10 mm), and are made of pure silver (99.9%). The sEMG data were corrected using EMG-Works-Acquisition (Delsys) at 2000 Hz. To exclude any influence of electrical noise, the electrode site was prepared by cleaning the skin with alcohol and shaving it lightly. One surface electrode was placed over the right medial latissimus dorsi (MLD), lateral to T9 over the muscle belly (Vera-Garcia et al., 2010; Frost et al., 2009). A second electrode was placed over the lateral latissimus dorsi (LLD) 4 cm below the inferior tip of the scapula, half the distance between the spine and lateral edge of the torso (Cram et al., 1998). Since the MLD is adjacent to the lower trapezius, the muscle belly was identified with manual muscle testing to prevent the electrode from overlapping both the latissimus dorsi and lower trapezius (Fig. 1) (Kendall et al., 2005). A Power-track II digital hand held dynamometer (JTECH Medical, Salt Lake City, UT, USA) was used to measure the strength of each subject. The device expresses the peak load as a numerical value (pounds) during isometric exercises and simultaneously displays the data and provides auditory feedback when exceeding a threshold. 2.3. Exercise procedure The maximum voluntary isometric contraction (MVIC) was measured to normalize the sEMG amplitude during the shoulder exercsies. Following a previous study that investigated normalization of the latissimus dorsi, three trials of MVIC were performed
while applying manual resistance (1) during shoulder extension with adduction in the prone position and (2) while pulling down against a fixed bar in the sitting position (Park and Yoo, 2013). Among the two MVICs for MLD and LLD, the highest mean value was used for normalizing the procedure. After measuring the MVIC, each subject performed three trials of each isometric exercise in shoulder extension, abduction, internal rotation, and depression. For each isomeric shoulder exercise, 50% of the maximum load was determined using a hand-held dynamometer. The participants performed two trials of maximum efforts for each of shoulder extension, adduction, internal rotation, and shoulder depression within standing position, and the 50% averaged maximum load was determined. After a 5-min rest and 5-min practice, the participants performed each exercise at 50% of the maximum load. While watching the display on the dynamometer, each participant tried to maintain the 50% maximum load for 5 s. The allowable error range of the load was 2.27 kg (5 lbs), and a trial exceeding this error range was regarded as a failure and repeated. To prevent compensatory movement, the subjects were fixed to a pillar by grasping handle bar from the pillar, and one researcher watched to exclude any compensatory movement (Fig. 1). The dynamometer was positioned at the posterior part of the elbow joint in shoulder extension, at the medial part of the elbow joint in shoulder adduction, at the medial forearm in shoulder internal rotation with elbow flexion, and at the inferior surface of the elbow in shoulder depression. Each subject conducted the exercise within the 5-s time frame using a 60-Hz metronome, and the middle 3 s were analyzed. The sequence of exercises was determined randomly. Each participant was allowed a 5-min rest between exercise sessions and a 60-s rest between trials. 2.4. Data analysis Three seconds of sEMG data during the exercises were fullwave rectified and averaged, and the data were expressed as the %MVIC and %Isolation values. The %MVIC was normalized relative to the MVIC procedure. To analyze the intramuscular response of the latissimus dorsi, the mean rectified sEMG value recorded from each branch of the latissimus dorsi was calculated using the isolation equation (Fig. 2). Previous studies used this equation to compare intramuscular branches of the trapezius and latissimus dorsi (Paton and Brown, 1995; Arlotta et al., 2011). PASW Statistics (ver. 18.0; SPSS, Chicago, IL, USA) was used to identify significant differences in the %Isolation and %MVIC between the MLD and LLD. Two-way repeated measures analysis of variance (ANOVA) was used to determine the effects of the muscle branches (MLD and LLD) and isometric exercises (isometric shoulder extension, adduction, internal rotation, and shoulder depression). When significant interactions were observed between factors, one-way repeated ANOVA was performed to evaluate the difference among the four isometric exercises for each muscle. If necessary, post hoc Bonferroni corrections were performed to identify differences among the four exercises. All significance levels were set at P < 0.05. 3. Results Descriptive statistics pertaining to the %MVIC values of the medial and lateral latissimus dorsi are shown in Table 1. The ‘‘isometric exercises’’ factor was associated with a significant difference in the %MVIC value (F3.14 = 11.913, P < 0.001, effect size = 0.719) but no significant interaction was evident between factors (F3.14 = 2.953, P > 0.05, effect size = 0.388). Shoulder
255
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Fig. 1. Illustration of electrode placements and four isometric exercise conditions. (A) Electrode placements, (B) shoulder extension, (C) shoulder adduction, (D) shoulder internal rotation, (E) shoulder depression.
than when the other isometric exercises were performed (P < 0.05). In contrast, the percentage isolation of the lateral latissimus dorsi was significantly higher upon shoulder depression compared to other conditions (P < 0.05). The mean difference in percentage isolation upon shoulder depression and extension was 18% (95% CI: 8–27%). The percentages of isolation evident upon shoulder depression, and adduction or internal rotation, differed by 10% (95% CI: 1–20%) and 13% (95% CI: 4–22%), respectively.
Fig. 2. Isolation equation used in this study.
extension was associated with significantly higher %MVIC values than were other isometric exercises. The mean difference between shoulder extension and adduction was 22% of the MVIC (95% CI: 11–33%). Shoulder extension also varied with the extent of internal rotation (by 31% of the MVIC [95% CI: 15–46%]) and depression (by 22% of the MVIC [95% CI: 7–36%]). The percentage isolation values of the medial and lateral latissimus dorsi during the four isometric exercises performed are shown in Table 2. No significant muscle effect was noted (F1.16 = 3.165, P = 0.09, effect size = 0.165) but a significant interaction between factors of muscle and isometric exercise was indeed observed (F3.14 = 10.326, P = 0.001, effect size = 0.689). Performance of repeated one-way ANOVA with pairwise comparisons revealed significant differences in the percentage isolation values of the two muscle branches as isometric exercise conditions varied (F3.14 = 10.326, P = 0.001, effect size = 0.689). Shoulder extension was associated with a significantly greater extent of medial latissimus dorsi activation than was shoulder internal rotation or depression (P < 0.05). The percentage isolation of the medial latissimus dorsi upon shoulder depression was significantly less
4. Discussion We explored whether functional differences were evident between the medial and lateral latissimus dorsi branches, and measured position-associated activation of these branches during performance of isometric exercises. We examined only two branches (the medial and lateral) of the latissimus dorsi, unlike previous studies that investigated several intramuscular branches (Brown et al., 2007). A recent cadaveric and surgical study showed that the thoracodorsal nerve (which innervates the latissimus dorsi) has both transverse and descending branches (Colohan et al., 2012). Therefore, we examined two branches of the latissimus dorsi using sEMG to calculate both isolation percentages and %MVIC values. To represent the effect of functional exercises on the medial and lateral latissimus dorsi, present study used the %MVIC. All of the isometric exercises induced %MVIC > 20%, which is considered moderate clinically (DiGiovine et al., 1992). Shoulder extension
Table 1 Descriptive statics of % MVIC values during four isometric exercise conditions. Exercise condition
P-value
Muscles
SE
SA
SI
SD
Muscle
Isometric exercises
MLD LLD
55.72 ± 19.87 52.71 ± 19.82
28.92 ± 12.91 35.63 ± 12.59
21.99 ± 13.46 25.18 ± 14.70
25.32 ± 12.79 39.05 ± 15.72
0.091
0.000*
SE: shoulder extension, SA: shoulder adduction, SI: shoulder internal rotation, SD: shoulder depression. MLD: medial latissimus dorsi, LLD: lateral latissimus dorsi. * Significant difference between conditions.
256
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Table 2 Descriptive statics of % Isolated EMG values during four isometric exercise conditions. Exercise condition
P-value
Muscles
SE
SA
SI
SD
Muscle
Isometric exercise
MLD LLD
61.58 ± 10.98 38.42 ± 10.98
54.37 ± 12.69 45.63 ± 12.69
56.84 ± 11.23 43.16 ± 11.23
43.90 ± 11.44 56.10 ± 11.44
0.094
0.001*
SE: shoulder extension, SA: shoulder adduction, SI: shoulder internal rotation, SD: shoulder depression. MLD: medial latissimus dorsi, LLD: lateral latissimus dorsi. * Significant difference between conditions.
showed a high level of muscle activation over 50% of MVIC, which was significantly higher in both the medial and lateral latissimus dorsi compared to shoulder adduction and internal rotation. This indicates that exercise including shoulder extension is effective at activating both the medial and lateral latissimus dorsi. Although the subject’s position was different to present study, a recent finding also demonstrated that shoulder extension in the prone position is the best method of inducing the maximum voluntary isometric contraction for latissimus dorsi (Park and Yoo, 2013). The %Isolation showed that the contributions of the medial and lateral latissimus dorsi branches differed in different isometric exercises. Two previous studies investigated the latissimus dorsi as six intramuscular branches (Paton and Brown, 1997; Brown et al., 2007). It was suggested that individual muscles segments could be classified as prime movers, synergists, or antagonists and that the functional classification was generally determined by the muscle segments moment arm or mechanical line of action (Brown et al., 2007). Although previous study excluded shoulder internal rotation and depression so that direct comparison with present result was impossible, Paton and Brown (1997) reported that relative %EMG value (similar to %Isolation in present study) was determined by the shoulder position and isometric exercise direction. The present result of %Isolation demonstrated that dominant activation of the medial branch was related to shoulder extension, adduction, and internal rotation, and dominant activation of the lateral branch was related to shoulder depression. We found that the lateral branch of the muscle was optimally used for shoulder depression, being less suited to shoulder extension, adduction, or internal rotation. However, the %Isolation data should not be understood as an actual activation level. The %Isolation data represent only the partial dominance of muscle activation when comparing limited intramuscular branches (Hug, 2011). Our method of calculation may be applied when it is necessary to compare the relative activation levels of various intramuscular branches, but the %MVIC data may differ from the isolation percentage values. Regarding the lateral latissimus dorsi branches, %MVIC in shoulder depression was not consistent with the %Isolation. Although shoulder depression showed a significantly higher %Isolation in the lateral latissimus dorsi compared to the other exercises, there were no major differences in the %MVIC between shoulder depression and other conditions. This inconsistent result could be interpreted as the lateral latissimus dorsi not being activated sufficiently with shoulder depression, but the relative activation was higher than that of the medial latissimus dorsi. Muscles acting synergistically to induce shoulder depression include the latissimus dorsi and the inferior fibers of the pectoralis, long head of the triceps, and subscapularis (Kendall et al., 2005). These muscles might be activated more with shoulder depression and contribute more to the exercises than the latissimus dorsi. Both the %MVIC and %Isolation results implied that a surgical procedure such as a partial latissimus dorsi flap would not cause a severe deficit of the latissimus dorsi. Ishida et al. (1999) demonstrated that sparing the medial branch in a partial latissimus dorsi flap resulted in a 25% increase in patient satisfaction, as well as less reduction in muscle power.
A recent study also demonstrated that a partial latissimus dorsi flap is a good option for leaving a minimal functional deficit in shoulder movement, compared to an extended latissimus dorsi flap that used both the medial and lateral latissimus dorsi (Kim et al., 2013). Our present results and those of previous studies suggest that attention should be paid to strengthening the medial branch of the latissimus dorsi to ensure good recovery of function after partial latissimus dorsi flap surgery. Although ours was a basic study of the activation of the middle and lateral latissimus dorsi using sEMG, it has several limitations. The most problematic is the possibility of cross-talk while measuring the sEMG of the medial latissimus dorsi. We attempted to avoid the lower trapezius and erector spine muscles when placing the electrodes, but activation of adjacent muscles could have affected our sEMG signal. Second, we could not record kinematic data throughout the experiment, so the accuracy of the neutral shoulder position relied on manual correction. Although every subject that participated in our study attempted to maintain 50% of the maximum load in all exercises, the 2.27 kg (5 lb) error range could have had a marked effect on our %MVIC data. 5. Conclusion Our data suggest that the medial and lateral branches of the latissimus dorsi are activated differently with changes in isometic exercises. The medial branch is activated more than the lateral branch with isometric exercises of shoulder extension, adduction, and internal rotation. The lateral branch is activated more than the medial branch with shoulder depression. However, this difference was not represented functionally. The actual activation level with %MVICs was higher in both of medial lateral branches with shoulder extension than with the other isometric exercises. Future work should evaluate each branch of the latissimus dorsi in detail, with considering limitations of present study. Conflict of Interest None. Acknowledgments This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. 2012R1A1B4001058). References Arlotta M, Lovasco G, McLean L. Selective recruitment of the lower fibers of trapezius muscle. J Electromyogr Kinesiol 2011;21(3):403–10. Brown JM, Wickham JB, McAndrew DJ, Huang XF. Muscles within muscles: coordination of 19 muscle segments within three shoulder muscles during isometric motor tasks. J Electromyogr Kinesiol 2007;17(1):57–73. Colohan S, Wong C, Lakhiani C, Cheng A, Maia M, Arbique G, et al. The free descending branch muscle-sparing latissimus dorsi flap: vascular anatomy and clinical applications. Plast Reconstr Surg 2012;130(6):776–87.
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257 Cram JR, Kasman GS, Holtz J. Introduction to surface electromyography. 1st ed., Aspen; 1998. de Oliveira RR, Pinto e Silva MP, Gurgel MS, Pastori-Filho L, Sarian LO. Immediate breast reconstruction with transverse latissimus dorsi flap does not affect the short-term recovery of shoulder range of motion after mastectomy. Ann Plast Surg 2010;64(4):402–8. DiGiovine NM, Jobe FW, Pink M, Perry J. An electromyographic analysis of the upper extremity in pitching. J Shoulder Elbow Surg 1992;1(1):15–25. Frost DM, Abdoli-E M, Stevenson JM. PLAD (personal lift assistive device) stiffness affects the lumbar flexion/extension moment and the posterior chain EMG during symmetrical lifting tasks. J Electromyogr Kinesiol 2009;19(6):403–12. Hug F. Can muscle coordination be precisely studied by surface electromyography? J Electromyogr Kinesiol 2011;21(1):1–12. Ishida A, Schmidt A, Fromberg G, Mo9 ckley I. Donor site morbidity: standard versus partial Latissimus Dorsi free flap. Eur J Plast Surg 1999;22(8):362–5. Kendall FP, McCreary EK, Provance PG, Rodgers MM, Romani WA. Muscles: testing and function, with posture and pain. 5th ed. Baltimore (MD): Lippincott Williams & Wilkins; 2005. Kim H, Wiraatmadja ES, Lim SY, Pyon JK, Bang SI, Oh KS, et al. Comparison of morbidity of donor site following pedicled muscle-sparing latissimus dorsi flap versus extended latissimus dorsi flap breast reconstruction. J Plast Reconstr Aesthet Surg 2013;66(5):640–6. Kwon ST, Chang H, Oh M. Anatomic basis of interfascicular nerve splitting of innervated partial latissimus dorsi muscle flap. J Plast Reconstr Aesthet Surg 2011;64(5):109–14. Park SY, Yoo WG. Comparison of exercises inducing maximum voluntary isometric contraction for the latissimus dorsi using surface electromyography. J Electromyogr Kinesiol 2013;23(5):1106–10. Paton ME, Brown JM. Functional differentiation within latissimus dorsi. Electromyogr Clin Neurophysiol 1995;35(5):301–9. Sahmann S. Diagnosis and treatment of movement impairment syndromes. St Louis: Mosby; 2002. p. 212–13. Signorile JF, Zink AJ, Szwed SP. A comparative electromyographical investigation of muscle utilization patterns using various hand positions during the lat pulldown. J Strength Cond Res 2002;16(4):539–46. Snyder BJ, Leech JR. Voluntary increase in latissimus dorsi muscle activity during the lat pull-down following expert instruction. J Strength Cond Res 2009;23(8):2204–9. Vera-Garcia FJ, Moreside JM, McGill SM. MVC techniques to normalize trunk muscle EMG in healthy women. J Electromyogr Kinesiol 2010;20(1):10–6.
257
Watanabe K, Kiyokawa K, Rikimaru H, Koga N, Yamaki K, Saga T. Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution. J Plast Reconstr Aesthet Surg 2010;63(7):1091–8. Zhao R, Qiao Q, Liu C. Anatomy of latissimus dorsi and its clinical application. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi 2003;17(1):38–40.
Se-yeon Park received his master of science in physical therapy at the Inje university in 2012 and now is a doctor of science student in physical therapy and rehabilitation sciences at Inje university. He is currently working as a researcher of National Research Foundation of Korea and a member of the Korean Physical Therapy Association. His research interests include musculoskeletal response in physical therapy interventions, and scapular dyskinesis.
Won-gyu Yoo received the Ph.D. in Physical Therapy Treatments for Musculoskeletal Disorders from the Yonsei University, the Republic of Korea, in 2008. He was the acting head of the Department of Physical Therapy at Inje University in Gimhea, Gyeongsangnamdo, Republic of Korea. He is working as the main researcher of National Research Foundation of Korea for posture correction research of computer users. His research interests include biological signal processing, chronic muscle pain and dysfunction due to overuse, and medical device development for physical therapy interventions.
Contents lists available at ScienceDirect
Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin
Differential activation of parts of the latissimus dorsi with various isometric shoulder exercises Se-yeon Park a, Won-gyu Yoo b,⇑ a b
Department of Physical Therapy, The Graduate School, Inje University, Republic of Korea Department of Physical Therapy, College of Biomedical Science and Engineering, Inje University, 607 Obangdong, Gimhae, Gyeongsangnam-do 621-749, Republic of Korea
a r t i c l e
i n f o
Article history: Received 6 August 2013 Received in revised form 16 October 2013 Accepted 10 December 2013
Keywords: Electromyography Isometric exercise Selective activation Shoulder movement
a b s t r a c t As no study has examined whether the branches of the latissimus dorsi are activated differently in different exercises, we investigated intramuscular differences of components of the latissimus dorsi during various shoulder isometric exercises. Seventeen male subjects performed four isometric exercises: shoulder extension, adduction, internal rotation, and shoulder depression. Surface electromyography (sEMG) was used to collect data from the medial and lateral components of the latissimus dorsi during the isometric exercises. Two-way repeated analysis of variance with two within-subject factors (exercise condition and muscle branch) was used to determine the significance of differences between the branches, and which branch was activated more with the exercise variation. The root mean squared sEMG values for the muscles were normalized using the modified isolation equation (%Isolation) and maximum voluntary isometric contraction (%MVIC). Neither the %MVIC nor %Isolation data differed significantly between muscle branches, while there was a significant difference with exercise. %MVIC was significantly higher with shoulder extension, compared to the other isometric exercises. There was a significant correlation between exercise condition and muscle branch in the %Isolation data. Shoulder extension and adduction and internal rotation increased %Isolation of the medial latissimus dorsi more than shoulder depression. Shoulder depression had the highest value of %Isolation of the lateral latissimus dorsi compared to the other isometric exercises. Comparing the medial and lateral latissimus dorsi, the medial component was predominantly activated with shoulder extension, adduction, and internal rotation, and the lateral component with shoulder depression. Shoulder extension is effective for activating the latissimus dorsi regardless of the intramuscular branch. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction In the musculoskeletal system, certain broad muscles with many insertions and origins are thought to have different kinematic functions according to the fiber arrangement (Paton and Brown, 1995; Brown et al., 2007). The latissimus dorsi originates from the spinous processes of the last six thoracic vertebrae, thoracolumbar fascia, and iliac crest, and inserts into the intertubercular groove of the humerus (Kendall et al., 2005). Since the latissimus dorsi has many neuromuscular end plates and vascular and nerve branches (Watanabe et al., 2010), it used as a donor sites for reconstructive surgery (de Oliveira et al., 2010; Kwon et al., 2011). Surgical methods for preventing morbidity of the latissimus dorsi, such as splitting the thoracodorsal nerve or sparing an area through vascular anatomy, have been suggested (Kwon et al., 2011). However, the contributions of the intramuscular branches of the latissimus dorsi to shoulder exercises have not been clearly ⇑ Corresponding author. Tel.: +82 55 320 3994; fax: +82 55 329 1678. E-mail address: [email protected] (W.-g. Yoo). 1050-6411/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jelekin.2013.12.004
established, although these differences are important for the prognosis after surgery and rehabilitation. Movement at a joint related to muscular activation can be investigated using surface electromyography (sEMG). When using a sEMG, the anthropometric position of the electrode sites is important for reducing spatial variability and measuring an accurate sEMG signal (Hug, 2011). The surface electrode sites for the latissimus dorsi have differed in previous studies: some used a site lateral to T9 over the muscle belly, while others used sites lateral and inferior to the inferior angle of the scapulae (Signorile et al., 2002; Snyder and Leech, 2009; Vera-Garcia et al., 2010). These sites might be differentiated as the medial and lateral branches of the latissimus dorsi, but a few studies have investigated both medial and lateral electrode sites of the latissimus dorsi simultaneously. To our knowledge, two studies have examined the functional differences of the fibers of this broad muscle with changes in the direction of exercises and shoulder position (Paton and Brown, 1995; Brown et al., 2007). Zhao et al. (2003) suggested that the latissimus dorsi can be divided into medial and lateral branches, and that the lateral branch is activated more with shoulder motion.
254
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Although the muscle induces many shoulder actions, including internal rotation, adduction, and extension of the humerus, and depression of the shoulder girdle, activation depending on branch differentiation has not been specifically reported. This preliminary study investigated the functional differences in the intramuscular branches of the latissimus dorsi during various isometric shoulder exercises. Based on a report that the lateral branch was activated more with shoulder motion (Zhao et al., 2003), we expected that shoulder exercises would activate the lateral branch of the latissimus dorsi more than the medial branch. 2. Method 2.1. Study population Twenty asymptomatic males were recruited from a local university using convenience sampling. They had a mean ± SD age of 23.38 ± 1.02 (range 22–26) years and weight-trained twice per week. Subjects with a history of upper extremity pain or discomfort in the past 6 months were excluded. Tightness of the latissimus dorsi was examined before the experiment. Subject lay on a table in the supine position, and was asked to flex his arm fully while remaining in contact with the table (Sahmann, 2002). Three subjects whose lumbar region rose off the table, indicating a tight latissimus dorsi, were excluded from this study. The final study sample comprised 17 males, with a mean height of 175.38 ± 4.88 cm and weighing 66.28 ± 4.75 kg. All subjects were right-hand dominant, with dominance defined as the hand used for writing. Ethics approval for this study was obtained from the Inje University Faculty of Health Sciences Human Ethics Committee. The participants provided informed consent. 2.2. Instrumentation Surface electromyography data were collected using a Trigno wireless system (Delsys, Boston, MA, USA); the Trigno electrodes (Delsys) were set at a band pass of 20–450 Hz and a common mode rejection ratio of 80 dB. The 27 37 15-mm sensor has four fixed 5 10-mm contact areas, which are half the area of a Bagnoli sensor (10 10 mm), and are made of pure silver (99.9%). The sEMG data were corrected using EMG-Works-Acquisition (Delsys) at 2000 Hz. To exclude any influence of electrical noise, the electrode site was prepared by cleaning the skin with alcohol and shaving it lightly. One surface electrode was placed over the right medial latissimus dorsi (MLD), lateral to T9 over the muscle belly (Vera-Garcia et al., 2010; Frost et al., 2009). A second electrode was placed over the lateral latissimus dorsi (LLD) 4 cm below the inferior tip of the scapula, half the distance between the spine and lateral edge of the torso (Cram et al., 1998). Since the MLD is adjacent to the lower trapezius, the muscle belly was identified with manual muscle testing to prevent the electrode from overlapping both the latissimus dorsi and lower trapezius (Fig. 1) (Kendall et al., 2005). A Power-track II digital hand held dynamometer (JTECH Medical, Salt Lake City, UT, USA) was used to measure the strength of each subject. The device expresses the peak load as a numerical value (pounds) during isometric exercises and simultaneously displays the data and provides auditory feedback when exceeding a threshold. 2.3. Exercise procedure The maximum voluntary isometric contraction (MVIC) was measured to normalize the sEMG amplitude during the shoulder exercsies. Following a previous study that investigated normalization of the latissimus dorsi, three trials of MVIC were performed
while applying manual resistance (1) during shoulder extension with adduction in the prone position and (2) while pulling down against a fixed bar in the sitting position (Park and Yoo, 2013). Among the two MVICs for MLD and LLD, the highest mean value was used for normalizing the procedure. After measuring the MVIC, each subject performed three trials of each isometric exercise in shoulder extension, abduction, internal rotation, and depression. For each isomeric shoulder exercise, 50% of the maximum load was determined using a hand-held dynamometer. The participants performed two trials of maximum efforts for each of shoulder extension, adduction, internal rotation, and shoulder depression within standing position, and the 50% averaged maximum load was determined. After a 5-min rest and 5-min practice, the participants performed each exercise at 50% of the maximum load. While watching the display on the dynamometer, each participant tried to maintain the 50% maximum load for 5 s. The allowable error range of the load was 2.27 kg (5 lbs), and a trial exceeding this error range was regarded as a failure and repeated. To prevent compensatory movement, the subjects were fixed to a pillar by grasping handle bar from the pillar, and one researcher watched to exclude any compensatory movement (Fig. 1). The dynamometer was positioned at the posterior part of the elbow joint in shoulder extension, at the medial part of the elbow joint in shoulder adduction, at the medial forearm in shoulder internal rotation with elbow flexion, and at the inferior surface of the elbow in shoulder depression. Each subject conducted the exercise within the 5-s time frame using a 60-Hz metronome, and the middle 3 s were analyzed. The sequence of exercises was determined randomly. Each participant was allowed a 5-min rest between exercise sessions and a 60-s rest between trials. 2.4. Data analysis Three seconds of sEMG data during the exercises were fullwave rectified and averaged, and the data were expressed as the %MVIC and %Isolation values. The %MVIC was normalized relative to the MVIC procedure. To analyze the intramuscular response of the latissimus dorsi, the mean rectified sEMG value recorded from each branch of the latissimus dorsi was calculated using the isolation equation (Fig. 2). Previous studies used this equation to compare intramuscular branches of the trapezius and latissimus dorsi (Paton and Brown, 1995; Arlotta et al., 2011). PASW Statistics (ver. 18.0; SPSS, Chicago, IL, USA) was used to identify significant differences in the %Isolation and %MVIC between the MLD and LLD. Two-way repeated measures analysis of variance (ANOVA) was used to determine the effects of the muscle branches (MLD and LLD) and isometric exercises (isometric shoulder extension, adduction, internal rotation, and shoulder depression). When significant interactions were observed between factors, one-way repeated ANOVA was performed to evaluate the difference among the four isometric exercises for each muscle. If necessary, post hoc Bonferroni corrections were performed to identify differences among the four exercises. All significance levels were set at P < 0.05. 3. Results Descriptive statistics pertaining to the %MVIC values of the medial and lateral latissimus dorsi are shown in Table 1. The ‘‘isometric exercises’’ factor was associated with a significant difference in the %MVIC value (F3.14 = 11.913, P < 0.001, effect size = 0.719) but no significant interaction was evident between factors (F3.14 = 2.953, P > 0.05, effect size = 0.388). Shoulder
255
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Fig. 1. Illustration of electrode placements and four isometric exercise conditions. (A) Electrode placements, (B) shoulder extension, (C) shoulder adduction, (D) shoulder internal rotation, (E) shoulder depression.
than when the other isometric exercises were performed (P < 0.05). In contrast, the percentage isolation of the lateral latissimus dorsi was significantly higher upon shoulder depression compared to other conditions (P < 0.05). The mean difference in percentage isolation upon shoulder depression and extension was 18% (95% CI: 8–27%). The percentages of isolation evident upon shoulder depression, and adduction or internal rotation, differed by 10% (95% CI: 1–20%) and 13% (95% CI: 4–22%), respectively.
Fig. 2. Isolation equation used in this study.
extension was associated with significantly higher %MVIC values than were other isometric exercises. The mean difference between shoulder extension and adduction was 22% of the MVIC (95% CI: 11–33%). Shoulder extension also varied with the extent of internal rotation (by 31% of the MVIC [95% CI: 15–46%]) and depression (by 22% of the MVIC [95% CI: 7–36%]). The percentage isolation values of the medial and lateral latissimus dorsi during the four isometric exercises performed are shown in Table 2. No significant muscle effect was noted (F1.16 = 3.165, P = 0.09, effect size = 0.165) but a significant interaction between factors of muscle and isometric exercise was indeed observed (F3.14 = 10.326, P = 0.001, effect size = 0.689). Performance of repeated one-way ANOVA with pairwise comparisons revealed significant differences in the percentage isolation values of the two muscle branches as isometric exercise conditions varied (F3.14 = 10.326, P = 0.001, effect size = 0.689). Shoulder extension was associated with a significantly greater extent of medial latissimus dorsi activation than was shoulder internal rotation or depression (P < 0.05). The percentage isolation of the medial latissimus dorsi upon shoulder depression was significantly less
4. Discussion We explored whether functional differences were evident between the medial and lateral latissimus dorsi branches, and measured position-associated activation of these branches during performance of isometric exercises. We examined only two branches (the medial and lateral) of the latissimus dorsi, unlike previous studies that investigated several intramuscular branches (Brown et al., 2007). A recent cadaveric and surgical study showed that the thoracodorsal nerve (which innervates the latissimus dorsi) has both transverse and descending branches (Colohan et al., 2012). Therefore, we examined two branches of the latissimus dorsi using sEMG to calculate both isolation percentages and %MVIC values. To represent the effect of functional exercises on the medial and lateral latissimus dorsi, present study used the %MVIC. All of the isometric exercises induced %MVIC > 20%, which is considered moderate clinically (DiGiovine et al., 1992). Shoulder extension
Table 1 Descriptive statics of % MVIC values during four isometric exercise conditions. Exercise condition
P-value
Muscles
SE
SA
SI
SD
Muscle
Isometric exercises
MLD LLD
55.72 ± 19.87 52.71 ± 19.82
28.92 ± 12.91 35.63 ± 12.59
21.99 ± 13.46 25.18 ± 14.70
25.32 ± 12.79 39.05 ± 15.72
0.091
0.000*
SE: shoulder extension, SA: shoulder adduction, SI: shoulder internal rotation, SD: shoulder depression. MLD: medial latissimus dorsi, LLD: lateral latissimus dorsi. * Significant difference between conditions.
256
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257
Table 2 Descriptive statics of % Isolated EMG values during four isometric exercise conditions. Exercise condition
P-value
Muscles
SE
SA
SI
SD
Muscle
Isometric exercise
MLD LLD
61.58 ± 10.98 38.42 ± 10.98
54.37 ± 12.69 45.63 ± 12.69
56.84 ± 11.23 43.16 ± 11.23
43.90 ± 11.44 56.10 ± 11.44
0.094
0.001*
SE: shoulder extension, SA: shoulder adduction, SI: shoulder internal rotation, SD: shoulder depression. MLD: medial latissimus dorsi, LLD: lateral latissimus dorsi. * Significant difference between conditions.
showed a high level of muscle activation over 50% of MVIC, which was significantly higher in both the medial and lateral latissimus dorsi compared to shoulder adduction and internal rotation. This indicates that exercise including shoulder extension is effective at activating both the medial and lateral latissimus dorsi. Although the subject’s position was different to present study, a recent finding also demonstrated that shoulder extension in the prone position is the best method of inducing the maximum voluntary isometric contraction for latissimus dorsi (Park and Yoo, 2013). The %Isolation showed that the contributions of the medial and lateral latissimus dorsi branches differed in different isometric exercises. Two previous studies investigated the latissimus dorsi as six intramuscular branches (Paton and Brown, 1997; Brown et al., 2007). It was suggested that individual muscles segments could be classified as prime movers, synergists, or antagonists and that the functional classification was generally determined by the muscle segments moment arm or mechanical line of action (Brown et al., 2007). Although previous study excluded shoulder internal rotation and depression so that direct comparison with present result was impossible, Paton and Brown (1997) reported that relative %EMG value (similar to %Isolation in present study) was determined by the shoulder position and isometric exercise direction. The present result of %Isolation demonstrated that dominant activation of the medial branch was related to shoulder extension, adduction, and internal rotation, and dominant activation of the lateral branch was related to shoulder depression. We found that the lateral branch of the muscle was optimally used for shoulder depression, being less suited to shoulder extension, adduction, or internal rotation. However, the %Isolation data should not be understood as an actual activation level. The %Isolation data represent only the partial dominance of muscle activation when comparing limited intramuscular branches (Hug, 2011). Our method of calculation may be applied when it is necessary to compare the relative activation levels of various intramuscular branches, but the %MVIC data may differ from the isolation percentage values. Regarding the lateral latissimus dorsi branches, %MVIC in shoulder depression was not consistent with the %Isolation. Although shoulder depression showed a significantly higher %Isolation in the lateral latissimus dorsi compared to the other exercises, there were no major differences in the %MVIC between shoulder depression and other conditions. This inconsistent result could be interpreted as the lateral latissimus dorsi not being activated sufficiently with shoulder depression, but the relative activation was higher than that of the medial latissimus dorsi. Muscles acting synergistically to induce shoulder depression include the latissimus dorsi and the inferior fibers of the pectoralis, long head of the triceps, and subscapularis (Kendall et al., 2005). These muscles might be activated more with shoulder depression and contribute more to the exercises than the latissimus dorsi. Both the %MVIC and %Isolation results implied that a surgical procedure such as a partial latissimus dorsi flap would not cause a severe deficit of the latissimus dorsi. Ishida et al. (1999) demonstrated that sparing the medial branch in a partial latissimus dorsi flap resulted in a 25% increase in patient satisfaction, as well as less reduction in muscle power.
A recent study also demonstrated that a partial latissimus dorsi flap is a good option for leaving a minimal functional deficit in shoulder movement, compared to an extended latissimus dorsi flap that used both the medial and lateral latissimus dorsi (Kim et al., 2013). Our present results and those of previous studies suggest that attention should be paid to strengthening the medial branch of the latissimus dorsi to ensure good recovery of function after partial latissimus dorsi flap surgery. Although ours was a basic study of the activation of the middle and lateral latissimus dorsi using sEMG, it has several limitations. The most problematic is the possibility of cross-talk while measuring the sEMG of the medial latissimus dorsi. We attempted to avoid the lower trapezius and erector spine muscles when placing the electrodes, but activation of adjacent muscles could have affected our sEMG signal. Second, we could not record kinematic data throughout the experiment, so the accuracy of the neutral shoulder position relied on manual correction. Although every subject that participated in our study attempted to maintain 50% of the maximum load in all exercises, the 2.27 kg (5 lb) error range could have had a marked effect on our %MVIC data. 5. Conclusion Our data suggest that the medial and lateral branches of the latissimus dorsi are activated differently with changes in isometic exercises. The medial branch is activated more than the lateral branch with isometric exercises of shoulder extension, adduction, and internal rotation. The lateral branch is activated more than the medial branch with shoulder depression. However, this difference was not represented functionally. The actual activation level with %MVICs was higher in both of medial lateral branches with shoulder extension than with the other isometric exercises. Future work should evaluate each branch of the latissimus dorsi in detail, with considering limitations of present study. Conflict of Interest None. Acknowledgments This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. 2012R1A1B4001058). References Arlotta M, Lovasco G, McLean L. Selective recruitment of the lower fibers of trapezius muscle. J Electromyogr Kinesiol 2011;21(3):403–10. Brown JM, Wickham JB, McAndrew DJ, Huang XF. Muscles within muscles: coordination of 19 muscle segments within three shoulder muscles during isometric motor tasks. J Electromyogr Kinesiol 2007;17(1):57–73. Colohan S, Wong C, Lakhiani C, Cheng A, Maia M, Arbique G, et al. The free descending branch muscle-sparing latissimus dorsi flap: vascular anatomy and clinical applications. Plast Reconstr Surg 2012;130(6):776–87.
S.-y. Park, W.-g. Yoo / Journal of Electromyography and Kinesiology 24 (2014) 253–257 Cram JR, Kasman GS, Holtz J. Introduction to surface electromyography. 1st ed., Aspen; 1998. de Oliveira RR, Pinto e Silva MP, Gurgel MS, Pastori-Filho L, Sarian LO. Immediate breast reconstruction with transverse latissimus dorsi flap does not affect the short-term recovery of shoulder range of motion after mastectomy. Ann Plast Surg 2010;64(4):402–8. DiGiovine NM, Jobe FW, Pink M, Perry J. An electromyographic analysis of the upper extremity in pitching. J Shoulder Elbow Surg 1992;1(1):15–25. Frost DM, Abdoli-E M, Stevenson JM. PLAD (personal lift assistive device) stiffness affects the lumbar flexion/extension moment and the posterior chain EMG during symmetrical lifting tasks. J Electromyogr Kinesiol 2009;19(6):403–12. Hug F. Can muscle coordination be precisely studied by surface electromyography? J Electromyogr Kinesiol 2011;21(1):1–12. Ishida A, Schmidt A, Fromberg G, Mo9 ckley I. Donor site morbidity: standard versus partial Latissimus Dorsi free flap. Eur J Plast Surg 1999;22(8):362–5. Kendall FP, McCreary EK, Provance PG, Rodgers MM, Romani WA. Muscles: testing and function, with posture and pain. 5th ed. Baltimore (MD): Lippincott Williams & Wilkins; 2005. Kim H, Wiraatmadja ES, Lim SY, Pyon JK, Bang SI, Oh KS, et al. Comparison of morbidity of donor site following pedicled muscle-sparing latissimus dorsi flap versus extended latissimus dorsi flap breast reconstruction. J Plast Reconstr Aesthet Surg 2013;66(5):640–6. Kwon ST, Chang H, Oh M. Anatomic basis of interfascicular nerve splitting of innervated partial latissimus dorsi muscle flap. J Plast Reconstr Aesthet Surg 2011;64(5):109–14. Park SY, Yoo WG. Comparison of exercises inducing maximum voluntary isometric contraction for the latissimus dorsi using surface electromyography. J Electromyogr Kinesiol 2013;23(5):1106–10. Paton ME, Brown JM. Functional differentiation within latissimus dorsi. Electromyogr Clin Neurophysiol 1995;35(5):301–9. Sahmann S. Diagnosis and treatment of movement impairment syndromes. St Louis: Mosby; 2002. p. 212–13. Signorile JF, Zink AJ, Szwed SP. A comparative electromyographical investigation of muscle utilization patterns using various hand positions during the lat pulldown. J Strength Cond Res 2002;16(4):539–46. Snyder BJ, Leech JR. Voluntary increase in latissimus dorsi muscle activity during the lat pull-down following expert instruction. J Strength Cond Res 2009;23(8):2204–9. Vera-Garcia FJ, Moreside JM, McGill SM. MVC techniques to normalize trunk muscle EMG in healthy women. J Electromyogr Kinesiol 2010;20(1):10–6.
257
Watanabe K, Kiyokawa K, Rikimaru H, Koga N, Yamaki K, Saga T. Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution. J Plast Reconstr Aesthet Surg 2010;63(7):1091–8. Zhao R, Qiao Q, Liu C. Anatomy of latissimus dorsi and its clinical application. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi 2003;17(1):38–40.
Se-yeon Park received his master of science in physical therapy at the Inje university in 2012 and now is a doctor of science student in physical therapy and rehabilitation sciences at Inje university. He is currently working as a researcher of National Research Foundation of Korea and a member of the Korean Physical Therapy Association. His research interests include musculoskeletal response in physical therapy interventions, and scapular dyskinesis.
Won-gyu Yoo received the Ph.D. in Physical Therapy Treatments for Musculoskeletal Disorders from the Yonsei University, the Republic of Korea, in 2008. He was the acting head of the Department of Physical Therapy at Inje University in Gimhea, Gyeongsangnamdo, Republic of Korea. He is working as the main researcher of National Research Foundation of Korea for posture correction research of computer users. His research interests include biological signal processing, chronic muscle pain and dysfunction due to overuse, and medical device development for physical therapy interventions.
