Functional roles of abdominal and back muscles ...

Journal of Orthopaedic Research 19 (2001) 463±471

www.elsevier.nl/locate/orthres

Functional roles of abdominal and back muscles during isometric axial rotation of the trunk Joseph K.-F. Ng a,*, Mohamad Parnianpour b, Carolyn A. Richardson c, Vaughan Kippers d a

Department of Physiotherapy and Department of Anatomical Sciences, The University of Queensland, Australia b Department of Industrial, Welding and Systems Engineering, The Ohio State University, USA c Department of Physiotherapy, The University of Queensland, Australia d Department of Anatomical Sciences, The University of Queensland, Australia Received 15 May 2000; accepted 6 September 2000

Abstract Electromyographic (EMG) studies have shown that a large number of trunk muscles are recruited during axial rotation. The functional roles of these trunk muscles in axial rotation are multiple and have not been well investigated. In addition, there is no information on the coupling torque at dierent exertion levels during axial rotation. The aim of the study was to investigate the functional roles of rectus abdominis, external oblique, internal oblique, latissimus dorsi, iliocostalis lumborum and multi®dus during isometric right and left axial rotation at 100%, 70%, 50% and 30% maximum voluntary contractions (MVC) in a standing position. The coupling torques in sagittal and coronal planes were measured during axial rotation to examine the coupling nature of torque at dierent levels of exertions. Results showed that the coupled sagittal torque switches from nil to ¯exion at maximum exertion of axial rotation. Generally, higher EMG activities were shown at higher exertion levels for all the trunk muscles. Signi®cant dierences in activity between the right and left axial rotation exertions were demonstrated in external oblique, internal oblique, latissimus dorsi and iliocostalis lumborum while no dierence was shown in rectus abdominis and multi®dus. These results demonstrated the different functional roles of trunk muscles during axial rotation. This is important considering that the abdominal and back muscles not only produce torque but also maintain the spinal posture and stability during axial rotation exertions. The changing coupling torque direction in the sagittal plane when submaximal to maximal exertions were compared may indicate the complex nature of the kinetic coupling of trunk muscles. Ó 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved.

Introduction Twisting of the trunk has been commonly identi®ed as a risk factor associated with back pain [3,6]. About 18% of back injuries could be attributed to twisting [46]. It is the third most common body movement contributing to non-accidental injuries of the back [20]. Twisting while lifting is also a signi®cant contributing factor to back injuries in the work place [50] and at home [31]. It has been found that the risk of prolapsed disk increases if twisting is combined with lifting [12]. An increase in trunk twisting velocity during lifting was associated with a higher risk of low back disorders [24]. There have been fewer studies investigating the biome* Corresponding author. Present address: Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hung Hom, Hong Kong. Tel.: +852-2766-6765; fax: +852-2330-8656. E-mail address: [email protected] (J.K.-F. Ng).

chanics and muscle activity in axial rotation when compared to those examining ¯exion and extension of the trunk. One of the reasons may be the complicated nature of axial rotation in regard to the number of muscles involved in the maneuver. Electromyographic (EMG) studies have shown that a large number of trunk muscles are involved in axial rotation and the roles of these trunk muscles during axial rotation are multiple. The muscles have been classi®ed as prime movers, antagonists and stabilizers [15,33]. Through biomechanical analysis, prime movers in axial rotation have been identi®ed as the contralateral external oblique and the ipsilateral internal oblique as well as the latissimus dorsi [5,43]. Contraction of the ipsilateral iliocostalis lumborum may also contribute to the axial rotation torque production [5]. The abdominal obliques and latissimus dorsi are muscles with large areas of attachment and also with wide muscle ®ber orientation. Hence, contraction of these muscles has

0736-0266/01/$ - see front matter Ó 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 0 ) 9 0 0 2 7 - 5

464

J.K.-F. Ng et al. / Journal of Orthopaedic Research 19 (2001) 463±471

been found to produce moments in three planes [5]. For example, abdominal obliques can produce moment in axial rotation accompanied with ¯exion and lateral ¯exion. It has been shown in vivo that coupling of torque in sagittal and coronal planes is found during axial rotation [22,35,36]. To maintain the trunk position and spinal stability during the production of torque in axial rotation, cocontractions in antagonists and synergists are necessary. These co-contractions are produced at the expense of axial torque production and may indicate that stabilization is very important during axial rotation [27]. It has been demonstrated that higher cocontractions are produced during axial rotation when compared with movements in the coronal and sagittal planes [22,23,49]. Various studies have investigated the functional roles of trunk muscles in producing torque in axial rotation. The role of the prime mover has been identi®ed by the line of force and their moment arms in biomechanical models as well as by EMG investigations. The role of antagonists and stabilizers have received less attention. In the past, a number of studies have examined both the torque and EMG activity of the trunk muscles during axial rotation in standing [17,21,22,27,28,39,40,42,49] and in sitting [14,25]. Con¯icting results on the activity of abdominal and back muscles have been demonstrated in these previous studies because of dierent experimental procedures. For example, Pope et al. [39] found that there was no dierence in activity (non-normalized EMG data) between ipsilateral and contralateral internal oblique. On the other hand, Thelen et al. [49] demonstrated higher activity (normalized EMG data) in ipsilateral internal oblique than that of contralateral side. In addition, there is very limited information on the coupling of torque during dierent levels of axial rotation exertions. It has been suggested that it is necessary to measure the torque output triaxially in order to understand the mechanics of axial rotation [34]. More information could be gained by examination of the EMG activity at dierent levels of exertions and in dierent directions of axial rotation. Studies are required to further consider the functional role of the trunk muscles. This knowledge would also be useful in understanding the biomechanics of the trunk in axial rotation exertions. Therefore, the present study aimed to investigate the torque output in three planes and EMG activity of six trunk muscles bilaterally during isometric right and left axial rotation at dierent exertion levels. It was hypothesized that there would be an eect of functional role of trunk muscle on the activation of individual muscle at various exertion levels during right and left axial rotation. We also hypothesized that the coupling torques in sagittal and coronal planes during axial rotation would dier at various levels of exertion.

Methods Subjects Twenty-three healthy male subjects without any history of back pain were recruited for this study. All the subjects were right-handed. Their mean (S.D.) age, height, weight were 30:2 7:9 yr, 1:8 0:1 m, and 68:1 10:3 kg, respectively. All subjects gave their written informed consent to participate. The study was approved by the Medical Research Ethics Committee of The University of Queensland. Equipment A triaxial dynamometer, B200 Isostation (Isotechnologies, Hillsborough, USA) was used to measure the torque produced by the trunk about the three planes of the body in six directions. These included the primary torque in the transverse plane (right/left axial rotation), the coupling torque in the sagittal plane (¯exion/extension) and coronal plane (right/left lateral ¯exion). EMG signals from both abdominal and back muscles were recorded by surface electrodes and were ampli®ed, band pass ®ltered at 5±500 Hz, and sampled at 1000 Hz. The torque data and EMG signals were collected with a data acquisition system, AMLAB II workstation (Associative Measurement, Australia). Electrodes Surface electrodes were placed over three pairs of abdominal muscles and three pairs of back muscles. Prior to placement of the recording electrodes, the skin at the electrode sites was shaved, cleaned with alcohol and prepared using ®ne sandpaper. The skin resistance of less than 5 kX was considered acceptable. To have the optimal pick-up of the EMG signals, the electrodes were placed in parallel with the muscle ®ber orientation. The electrodes for rectus abdominis (RA) were placed 1 cm above the umbilicus and 2 cm lateral to the midline. For external oblique (EO), electrodes were placed just below the rib cage and along a line connecting the most inferior point of the costal margin and the contralateral pubic tubercle [32]. For internal oblique (IO), electrodes were placed 1 cm medial to the anterior superior iliac spine (ASIS) and beneath a line joining both ASISs [32]. It has been demonstrated that similar surface electrode positions used in the present study for EO and IO were found to record most of the signals from these muscles [26]. The electrodes for latissimus dorsi (LD) were placed over the muscle belly at T12 level and along a line connecting the most superior point of the posterior axillary fold and the S2 spinous process. The T12 level was selected so as to avoid the pressure of the thoracic pad on the electrode [47]. For the iliocostalis lumborum (IL), the electrodes were placed at the L2 level and aligned parallel to the line between the posterior superior iliac spine (PSIS) and the lateral border of the muscle at the 12th rib [4]. For the multi®dus (MU), the electrodes were placed at the L5 level and aligned parallel to the line between the PSIS and the L1-2 interspinous space [4]. This surface electrode position for multi®dus had been shown recently to have similar ®ndings to those measured with intramuscular electrodes [1]. Experimental procedure The subject was positioned in erect standing with the L5-S1 interspinous space aligned with the ¯exion/extension axis of the B200 Isostation. The pelvis and lower legs of the subjects were stabilized by the pelvic restraint, as well as thigh and knee straps. The torso was ®xed by the chest restraint and thoracic pad according to the operation manual. To prepare for the isometric testing, the machine was mechanically locked in three planes. The tasks included maximum isometric contractions in all three planes with additional submaximal contractions at three dierent exertion levels in right and left axial rotation. The six maximum and six submaximal exertions were each arranged in a randomized order. The subjects were asked to fold their arms across their chest, hold their arms above the chest restraint and maintain this position during all exertions. To decrease the chance of injury, subjects were instructed to avoid any jerky movements during the exertions. Maximum voluntary contractions (MVC) in ¯exion, extension, right and left lateral ¯exion, as well as right and left axial rotation were

J.K.-F. Ng et al. / Journal of Orthopaedic Research 19 (2001) 463±471 measured for 5 s with a 2-min rest between trials. Visual feedback and verbal encouragement were given to ensure that the maximal eort was produced by the subject. Subjects were given a rest of 15 min after the maximal exertion testing protocol. Submaximal contractions at 70%, 50% and 30% MVC in both right and left axial rotation were then computed and each displayed as a reference line on the visual feedback monitor. Each contraction lasted for 5 s and a 2-min rest was given between contractions. Each subject attended two sessions ± a familiarization and a testing session. The familiarization session was included to minimize the learning eect, and allow the subject to gain some knowledge of the equipment and testing procedure (without placement of surface electrodes). The familiarization session was held at least 3 days before the testing session. Data of the torque produced and the EMG activity in the trunk muscles during dierent exertions were collected in the testing session. Data analysis The duration of the data to be analyzed were identi®ed based on the stable 3-s data of measured axial rotation torque [47]. The coupling torque values and EMG data for this period were further analyzed. The means of the primary and coupling torque values were computed. Root mean square values of the EMG data were calculated to quantify the amplitude of EMG signals. The baseline torque and EMG values were subtracted from those during exertions. The baseline EMG data were the minimum EMG values observed in any exertion [16] while the baseline torque values were those obtained during relaxed standing in the B200 Isostation. To compare between subjects, the torque values and EMG data of individual muscles were normalized with respect to the maximum torque values and EMG data acquired during the maximum exertions in three planes. Paired t-test was performed to ®nd any signi®cant dierences between the right and left axial rotation maximal isometric strength. Multivariate analysis of variance (MANOVA) with repeated measures design was applied to the coupling torque values and EMG data of trunk muscles to determine the main eects of exertion levels, direction of exertion and their interaction eects. Subsequent to signi®cant MANOVA, univariate analysis of variance (ANOVA) with repeated measures design was applied to individual coupling torque values and EMG data for each individual muscle. The t-test was applied to the coupling torques in sagittal and coronal planes to determine if the mean values were signi®cantly dierent from zero at each exertion level.

Results Torque values Maximal isometric strengths in right and left axial rotation were 78:3 24:9 and 77:6 18:6 Nm, respec-

465

tively. The direction of exertion did not signi®cantly aect the axial rotation isometric strength. Fig. 1 demonstrates the coupling of torque in sagittal and coronal planes during axial rotation at dierent exertion levels. Signi®cant dierences in coupling torque at dierent exertion levels of axial rotation were shown in the sagittal and coronal planes (Table 1). No signi®cant main eect for the direction of exertion and a non-signi®cant interaction for level by direction were demonstrated for coupling torque in the sagittal plane. In contrast, signi®cant main eect for direction of exertion and a signi®cant interaction for level by direction were found for coupling torque in the coronal plane. Since there was no dierence between direction for the coupling torque in the sagittal plane, the data for right and left axial rotation were pooled and averaged. Coupling torque in the sagittal plane during axial rotation was signi®cantly dierent (P 0:003) from zero at 100% MVC but not at the other exertion levels. In coronal plane, the coupling torque was signi®cantly dierent from zero at all exertion levels (all P values 6 0:002) for both right and left axial rotation exertions. EMG activity The normalized EMG activity of the abdominal and back muscles during dierent levels of exertion in right and left axial rotation is shown in Figs. 2 and 3, respectively. Table 1 shows that the exertion level in¯uenced the EMG activity of all the trunk muscles. Signi®cant dierences in EMG activity between the right and left axial rotation and a signi®cant interaction for level by direction were found in external oblique, internal oblique, latissimus dorsi and iliocostalis lumborum. On the other hand, no signi®cant main eect for direction of exertion and a non-signi®cant interaction for level by direction were demonstrated in rectus abdominis and multi®dus.

Fig. 1. Mean (S.D.) coupling of torque in sagittal and coronal planes at dierent exertion levels during axial rotation. The positive exertion levels correspond to right axial rotation and the negative exertion levels correspond to left axial rotation.

466


Table 1 MANOVA and ANOVA results (F values) for the eects of exertion level and direction of exertion on the coupling torque and EMG activity of the trunk muscles during axial rotation Eects MANOVA ± all torques ANOVA for individual coupling torque Flexion/extension Lateral Flexion MANOVA ± all muscles ANOVA for individual muscles (R) rectus abdominis (L) rectus abdominis (R) external oblique (L) external oblique (R) internal oblique (L) internal oblique (R) latissimus dorsi (L) latissimus dorsi (R) iliocostalis lumborum (L) iliocostalis lumborum (R) multi®dus (L) multi®dus

Exertion level

Direction of exertion

Level x direction

10.9

29.5

5.5

15.7 8.3 12.6

0.0a 58.2 83.8

0.8a 11.5 10.3

23.8 59.4 97.9 179.6 63.9 81.6 142.2 89.4 43.0 72.8 26.1 18.1

3.0a 3.3a 6.0 14.7 115.8 255.5 241.0 71.4 88.1 110.6 2.3a 1.6a

1.1a 0.8a 4.4 5.3 27.0 57.6 118.1 61.0 29.8 49.1 2.6a 1.1a

a

NS: Non-signi®cant. P < 0:05. ** P < 0:01. *

Discussion Torque measurement Coupling of torque in three planes was observed in the present study. The lateral ¯exion coupling torque was more evident than coupling torque in sagittal plane during axial rotation exertions. It was also found that the coupling of torque in the coronal and transverse planes is largely ipsilateral, i.e. right lateral ¯exion coupled with right axial rotation. The coupling of torques in dierent planes may be attributed to the wide orientation and attachment of abdominal and back muscles that could produce moments in dierent planes. In a cadaveric study [5] that examined the lines of force for trunk muscles, it was found that the largest moment of a trunk muscle is highly related to the expected contribution of that muscle but the contributions to other directions were also signi®cant. It has been demonstrated in biomechanical modeling that some degree of lateral ¯exion moment must be produced in order to attain a maximum level of axial rotation torque [37]. An interesting novel ®nding was observed in the coupling of torque during axial rotation. At maximum exertion of axial rotation torque, the coupled sagittal torque switches from nil to ¯exion. This ®nding suggests that the kinetic coupling of trunk muscles is a more complex phenomenon than that previously explained by the geometrical con®guration of trunk anatomy [5,36,37]. It appears that neural strategies adapt to the required task demand aecting the coordinated trunk recruitment during axial rotation exertions. To maxi-

mize the axial rotation exertions the antagonist muscles balancing the accompanying ¯exion moments generated by the agonists (primary axial rotators) must not increase at the same rate (proportions). Hence, the spine must tolerate the cost of higher ¯exion moments being unbalanced at higher exertion levels of axial rotations. It appears that the neural control system must prioritize the satisfaction of the required task demands with the limitations imposed by the neural and anatomical structures to maintain a coordinated and controlled exertion. Any condition that deleteriously aects the coordination and control of the trunk muscles may pose additional risks to the spine due to alterations in loading patterns on the spine [36]. It seems that higher axial rotation torque demands may pose two interacting risk mechanisms: higher loads in addition to lower controllability of accompanying sagittal and coronal exertions. EMG ®ndings The main ®nding of the present study was that there was a dierence between the directions of axial rotation exertions in the activity for external oblique, internal oblique, latissimus dorsi and iliocostalis lumborum. On the other hand, no such dierence was demonstrated in rectus abdominis and multi®dus. This may be explained by the dierent functional roles of the trunk muscles during axial rotation. High activity of contralateral external oblique and ipsilateral internal oblique was demonstrated in the present study. During trunk axial rotation, it has been established in biomechanical analysis that contralateral


467

Fig. 2. Mean (S.D.) normalized EMG amplitude of abdominal muscles at dierent exertion levels during axial rotation. The positive exertion levels correspond to right axial rotation and the negative exertion levels correspond to left axial rotation.

external oblique and ipsilateral internal oblique abdominal muscles are the agonists for axial rotation [43,44]. In addition to the present study, this pattern has been con®rmed in EMG studies using surface electrodes [25,27,28,33,39,40,43,49] and also using intramuscular electrodes [11]. Of interest are the results of a recent study that demonstrated that there was a regional difference in the activation of external oblique during axial rotation [30]. This may have to be taken into account in future studies on the role of the abdominal obliques in axial rotation. Signi®cant antagonistic muscle activity of the ipsilateral external and contralateral internal oblique muscles during axial rotation exertions is also observed in the present study. This antagonistic muscle activity is produced at the expense of the axial rotation torque production and is considered as muscle co-contraction. Such co-contractions are important in stabilizing the

spine during exertion [49]. This premise is supported by the observations that co-contraction increases the spinal load [7], and that this increase in compressive load on the spine augments the torsional stiness of the lumbar segments [9]. Although the external oblique and internal oblique muscles are classi®ed as either a prime mover or an antagonist depending on the direction of axial rotation, the dierence in external oblique activity between sides as well as between directions of axial rotation was less than that in internal oblique. This ®nding was similarly demonstrated in some previous studies [27,28,49] but not in another study by Pope et al. [39]. One of the reasons may be due to the EMG data of Pope et al.'s study [39] not being normalized, so the comparison between muscles may not be the same as other studies where normalization of the EMG data occurred. The dierence in the behavior of external

468


Fig. 3. Mean (S.D.) normalized EMG amplitude of back muscles at dierent exertion levels during axial rotation. The positive exertion levels correspond to right axial rotation and the negative exertion levels correspond to left axial rotation.

oblique when compared with internal oblique may indicate that the role of external oblique is more complicated than just acting as a prime mover or an antagonist. It is interesting to note that in one previous study [22], the activity of external oblique in its antagonistic role during axial rotation was more than that in its prime mover role. No dierence in rectus abdominis activity between right and left axial rotation exertions was shown in the present study. Due to the vertical alignment of the rectus abdominis, there is minimal contribution to the production of torque in transverse plane [8,39]. The similarity in activity between sides for the rectus abdominis may suggest that the muscle acts as a synergist during axial rotation. The co-contractions of the rectus abdominis are considered to relate to its role in the stabilization of the lumbar spine [22,33,39,49].

Classically, latissimus dorsi has been commonly regarded as an arm muscle although it is attached to the thoracolumbar fascia. Early study has predicted that latissimus dorsi only contributes to about 5% of the axial rotation torque [43]. It is interesting to note that in a recent comparison between dierent biomechanical models, quite dierent predictions on the contribution of latissimus dorsi in axial rotation have been shown [37]. To date, few studies have investigated the EMG activity of latissimus dorsi during axial rotation of the trunk. High activation of ipsilateral latissimus dorsi was found while low activity was shown in contralateral latissimus dorsi [22,27,28,49]. Similar ®ndings were found in the present study. The high activity of ipsilateral latissimus dorsi during axial rotation may be due to the production of torque [22,27,28,49] as well as providing stability to the lumbar spine [22]. Marras and Granata


[22] suggested that part of the activity in the latissimus dorsi was to balance the ancillary torque produced in rectus abdominis. It should be noted that the arm position may aect the EMG activation of latissimus dorsi during axial rotation. No previous studies have been devoted to investigating the eect of arm position on the activation of latissimus dorsi during axial rotation of the trunk. However, it was demonstrated that dierent activation patterns of latissimus dorsi were found at various arm positions during shoulder exertion [38]. Of interest, functional dierentiation within latissimus dorsi has also been shown [38]. The ®ndings of the present study remain valid since the arm position was kept consistent across all exertions. In axial rotation, it has been established that ipsilateral erector spinae (including iliocostalis lumborum and longissimus thoracis) are more active than the contralateral erector spinae [17,25,28,39,40,43,49]. A similar dierence in activity between sides for iliocostalis lumborum was demonstrated in the present study. On the other hand, the present study did not demonstrate any dierence between sides for multi®dus. Opinions dier among researchers on the role of erector spinae during axial rotation. Some studies proposed that erector spinae could contribute to the production of axial rotation torque [5,15,22]. Other researches have veri®ed that the erector spinae are in a mechanically disadvantaged position for producing axial rotation [19,27,43] and hence a stabilization role has been attributed to this muscle group. It has been suggested that, during axial rotation, back muscles maintain the spinal posture and stabilize the lumbar spine [27,39,49]. In addition, back muscles are capable of counteracting the ¯exion or lateral ¯exion produced by other trunk muscles during axial rotation [27,43]. The lateral ¯exion produced during axial rotation associated with the contraction of external oblique may be counteracted by the back muscles. The higher activity of the iliocostalis lumborum (52±59%) than multi®dus (19±23%) could be attributed to iliocostalis lumborum being in a better mechanical position to produce lateral ¯exion than the lumbar multi®dus [18,43]. EMG studies by Jonsson [10] and Thelen et al. [48] have shown that iliocostalis lumborum is more active than multi®dus in lateral ¯exion exertions. One of the functions of multi®dus is to balance the unwanted ¯exion moment generated by the abdominal obliques as they rotate the trunk [18,19]. It has been suggested by McGill [29] that contraction of the anterior portion of internal oblique and external oblique produced a ¯exor moment to the trunk via their attachment to the rectus abdominis sheath (linea semilunaris). The other function of multi®dus during axial rotation is to maintain the normal lumbar curvature. Curvature of the lumbar spine has been shown to aect the internal

469

loading of the spine [45]. It has also been demonstrated that higher torque output is produced if the neutral spinal curvature is maintained during axial rotation when compared to hyperlordotic and hypolordotic postures [28]. EMG activity of multi®dus found in the present study not only showed similar levels between right and left axial rotation, but also increased to a lesser extent with exertion level when compared with the other synergist, rectus abdominis. This can be attributed to the dierence in morphology and biomechanics between rectus abdominis and multi®dus. Rectus abdominis spans across the abdomen and attaches between the thoracic cage and pelvis, it functions to maintain the global stability of the spine. While the multi®dus, with its attachment to the lumbar vertebrae directly, is important in local stabilization of the lumbar spine [2,13]. The stabilization role of multi®dus has been veri®ed in the in vitro biomechanical studies [41,51]. Conclusion Dierent functional roles of trunk muscles during axial rotation were demonstrated by the varied activity patterns in dierent directions and levels of exertion. This is important in view that abdominal and back muscles not only produce torque but also maintain the spinal posture and stability during axial rotation exertions. The changing coupling torque direction in the sagittal plane when submaximal to maximal exertions were compared may indicate the complex nature of kinetic coupling of trunk muscles. More attention should be paid to the important role of neural control system in the assessment of trunk muscle performance during asymmetric exertions. Acknowledgements We are most grateful to the sta of the Worker's Compensation Board of Queensland for their invaluable assistance during the whole data collection process. Funding for this project was provided by the Dorothy Hopkins Award for Clinical Study and the research support grant of Manipulative Therapists Special Group of Queensland, Australia. References [1] Arokoski JPA, Kankaanpaa M, Valta T, Juvonen I, Partanen J, Taimela S, Lindgren K-A, Airaksinen O. Back and hip extensor muscle function during therapeutic exercises. Arch Phys Med Rehabil 1999;80:842±50. [2] Bergmark A. Stability of the lumbar spine. A study in mechanical engineering. Acta Orthop Scand Suppl 1989;230:1±54.

470


[3] Burdorf A, Sorock G. Positive and negative evidence of risk factors for back disorders. Scand J Work Environ Health 1997;23:243±56. [4] De Foa JL, Forrest W, Biedermann H-J. Muscle ®bre direction of longissimus, iliocostalis and multi®dus: landmark-derived reference lines. J Anat 1989;163:243±7. [5] Dumas GA, Poulin MJ, Roy B, Gagnon M, Jovanovic M. Orientation and moment arms of some trunk muscles. Spine 1991;16:293±303. [6] Frymoyer JW, Pope MH, Costanza MC, Rosen JC, Goggin JE, Wilder DG. Epidemiologic studies of low-back pain. Spine 1980;5:419±23. [7] Granata KP, Marras WS. The in¯uence of trunk muscle coactivity on dynamic spinal loads. Spine 1995;20:913±9. [8] Hughes RE, Chan DB. The eect of strict muscle stress limits on abdominal muscle force predictions for combined torsion and extension loadings. J Biomech 1995;28:527±33. [9] Janevic J, Ashton-Miller JA, Schultz AB. Large compressive preloads decrease lumbar motion segment ¯exibility. J Orthop Res 1991;9:228±36. [10] Jonsson B. The functions of individual muscles in the lumbar part of the spinae muscle. Electromyography 1970;10:5±21. [11] Juker D, McGill S, Kropf P, Steen T. Quantitative intramuscular myoelectric activity of lumbar portions of psoas and the abdominal wall during a wide variety of tasks. Med Sci Sports Exerc 1998;30:301±10. [12] Kelsey JL, Githens PB, White III AA, Holford TR, Walter SD, O'Connor T, Ostfeld AM, Weil U, Southwick WO, Calogero JA. An epidemiologic study of lifting and twisting on the job and risk for acute prolapsed lumbar intervertebral disc. J Orthop Res 1984;2:61±6. [13] Kiefer A, Shirazi-Adl A, Parnianpour M. Synergy of the human spine in neutral postures. Eur Spine J 1998;7:471±9. [14] Kumar S, Narayan Y. EMG spectral characteristics of spinalmuscles during isometric axial rotation. J Electromyogr Kinesiol 1999;9:21±37. [15] Kumar S, Narayan Y, Zedka M. An electromyographic study of unresisted trunk rotation with normal velocity among healthy subjects. Spine 1996;21:1500±12. [16] Lavender S, Tra®mow J, Andersson GBJ, Mayer RS, Chen I-H. Trunk muscle activation. The eects of torso ¯exion, moment direction and moment magnitude. Spine 1994;19:771±8. [17] Lo¯and KR, Mumby PB, Cassisi JE, Palumbo NL, Camic PM. Assessment of lumbar EMG during static and dynamic activity in pain-free normals: implications for muscle scanning protocols. Biofeedback Self Regul 1995;20:3±18. [18] Macintosh JE, Bogduk N. The biomechanics of the lumbar multi®dus. Clin Biomech 1986;1:205±13. [19] Macintosh JE, Pearcy MJ, Bogduk N. The axial torque of the lumbar back muscles: torsion strength of the back muscles. Aust N Z J Surg 1993;63:205±12. [20] Manning DP, Mitchell RG, Blanch®eld LP. Body movements and events contributing to accidental and nonaccidental back injuries. Spine 1984;9:734±9. [21] Marras WS, Davis KG, Granata KP. Trunk muscle activities during asymmetric twisting motions. J Electromyogr Kinesiol 1998;8:247±56. [22] Marras WS, Granata KP. A biomechanical assessment and model of axial twisting in the thoracolumbar spine. Spine 1995;20:1440± 51. [23] Marras WS, Granata KP. Spine loading during trunk lateral bending motions. J Biomech 1997;30:697±703. [24] Marras WS, Lavender SA, Leurgans SE, Rajulu SL, Allread WG, Fathallah FA, Ferguson SA. The role of dynamic three-dimensional trunk motion in occupationally-related low back disorders. The eects of workplace factors, trunk position, and trunk motion characteristics on risk of injury. Spine 1993;18:617±28.

[25] Mayer TG, Smith SS, Kondraske G, Gatchel RJ, Carmichael TW, Mooney V. Quanti®cation of lumbar function. Part 3. Preliminary data on isokinetic torso rotation testing with myoelectric spectral analysis in normal and low-back pain subjects. Spine 1985;10:912± 20. [26] McGill S, Juker D, Kropf P. Appropriately placed surface EMG electrodes re¯ect deep muscle activity (psoas, quadratus lumborum, abdominal wall) in the lumbar spine. J Biomech 1996; 29:1503±7. [27] McGill SM. Electromyographic activity of the abdominal and low back musculature during the generation of isometric and dynamic axial trunk torque: implications for lumbar mechanics. J Orthop Res 1991;9:91±103. [28] McGill SM. The in¯uence of lordosis on axial trunk torque and trunk muscle myoelectric activity. Spine 1992;17:1187±93. [29] McGill SM. A revised anatomical model of the abdominal musculature for torso ¯exion eorts. J Biomech 1996;29:973±7. [30] Mirka G, Kelaher D, Baker A, Harrison A, Davis J. Selective activation of the external oblique musculature during axial torque production. Clin Biomech 1997;12:172±80. [31] Mundt DJ, Kelsey JL, Golden AL, Pastides H, Berg AT, Sklar J, Hosea T, Panjabi MM, Northeast collaborative group on low back pain. Anepidemiologic study of non-occupational lifting as a risk factor forherniated lumbar intervertebral disc. Spine 1993;18:595±602. [32] Ng JK-F, Kippers V, Richardson CA. Muscle ®bre orientation of abdominal muscles and suggested surface EMG electrode positions. Electromyogr Clin Neurophysiol 1998;38:51±8. [33] O'Brien PR, Potvin JR. Fatigue related EMG responses of trunk muscles to a prolonged, isometric twist exertion. Clin Biomech 1997;12:306±13. [34] Parnianpour M. Modeling of trunk muscle recruitment during isometric exertion. IEEE Eng Med Biol Magazine 1991;10:51±4. [35] Parnianpour M, Campello M, Sheikhzadeh A. The eect of posture on triaxial trunk strength in dierent directions: its biomechanical consideration with respect to incidence of low-back problem in construction industry. Int J Ind Ergonom 1991;8:279±87. [36] Parnianpour M, Nordin M, Kahanovitz N, Frankel V. The triaxial coupling of torque generation of trunk muscles during isometric exertions and the eect of fatiguing isoinertial movements on the motor output and movement patterns. Spine 1988;13:982±92. [37] Parnianpour M, Wang JL, Shirazi-Adl A, Sparto P, Wilke H-J. The eect of variations in trunk models in predicting muscle strength and spinal loading. J Musculoskeletal Res 1997;1:55±69. [38] Paton ME, Brown JMM. Functional dierentiation within latissimus dorsi. Electromyogr Clin Neurophysiol 1995;35:301±9. [39] Pope MH, Andersson GBJ, Broman H, Svensson M, Zetterberg C. Electromyographic studies of the lumbar trunk musculature during the development of axial torques. J Orthop Res 1986;4:288±97. [40] Pope MH, Svensson M, Andersson GBJ, Broman H, Zetterberg C. The role of prerotation of the trunk in axial twisting eorts. Spine 1987;12:1041±5. [41] Quint U, Wilke H-J, Shirazi-Adl A, Parnianpour M, L oer F, Claes LE. Importance of the intersegmental trunk muscles for the stability of the lumbar spine. A biomechanical study in vitro. Spine 1998;23:1937±45. [42] Schultz A, Cromwell R, Warwick D, Andersson G. Lumbar trunk muscle use in standing isometric heavy exertions. J Orthop Res 1987;5:320±9. [43] Schultz A, Haderspeck K, Warwick D, Portillo D. Use of lumbar trunk muscles in isometric performance of mechanically complex standing tasks. J Orthop Res 1983;1:77±91. [44] Schultz AB, Andersson GBJ, Haderspeck K, Ortengren R, Nordin M, Bj ork R. Analysis and measurement of lumbar trunk loads in tasks involving bends and twists. J Biomech 1982;15:669±75.

J.K.-F. Ng et al. / Journal of Orthopaedic Research 19 (2001) 463±471 [45] Shirazi-Adl A, Parnianpour M. Role of posture in mechanics of the lumbar spine in compression. J Spinal Disord 1996;9:277±86. [46] Snook SH, Campanelli RA, Hart JW. A study of three preventive approaches to low back injury. J Occup Med 1978;20:478±81. [47] Tan JC, Parnianpour M, Nordin M, Hofer H, Willems B. Isometric maximal and submaximal trunk extension at dierent ¯exed positions instanding. Triaxial torque output and EMG. Spine 1993;18:2480±90. [48] Thelen DG, Ashton-Miller JA, Schultz AB. Lumbar muscle activities in rapid three-dimensional pulling tasks. Spine 1996; 21:605±13.

471

[49] Thelen DG, Schultz AB, Ashton-Miller JA. Co-contraction of lumbar muscles during the development of time-varying triaxial moments. J Orthop Res 1995;13:390±8. [50] Troup JDG, Martin JW, Lloyd DCEF. Back pain in industry. A prospective survey. Spine 1981;6:61±9. [51] Wilke H-J, Wolf S, Claes LE, Arand M, Wiesend A. Stability increase of the lumbar spine with dierent muscle groups. A biomechanical in vitro study. Spine 1995;20:192±8.