Pathophysiology 12 (2005) 243–247
Trunk muscle co-ordination during gait: Relationship between muscle function and acute low back pain Christoph Anders a,∗ , Hans-Christoph Scholle a , Heiko Wagner b , Christian Puta c , Roland Grassme a,d , Alexander Petrovitch e a
Institute for Pathophysiology and Pathobiochemistry, Motor Research Group, Friedrich-Schiller-University, D-07740 Jena, Germany b Institute for Sports Sciences, Chair of Biomechanics, Friedrich-Schiller-University Jena, Germany c Institute for Sports Sciences, Chair of Sports Medicine, Friedrich-Schiller-University Jena, Germany d Berufsgenossenschaft Nahrungsmittel & Gastst¨ atten Prevention Department Erfurt, Erfurt, Germany e Institute for Diagnostic and Interventional Radiology, Friedrich-Schiller-University Jena, Germany Received 4 May 2005; accepted 5 September 2005
Abstract Low back pain costs billions of Euros annually in all industrialized countries. Often radiological diagnosis fails to give evidence of the pathogenesis of low back pain. Although psychophysiological characteristics have an influence, it seems that insufficient muscular spinal stabilization may play the major role in the development of low back pain. Assessment of trunk muscle stabilization activity during everyday activities is rare. Therefore, in this study healthy persons were investigated during walking on a treadmill at a speed of 4 km/h. Women (n = 16) with no history of back pain were investigated before and after a static loading situation of the spine, i.e. while wearing a waistcoat. After this loading situation four women developed pain (pain subjects). Surface EMG (SEMG) was taken from five trunk muscles of both sides. Grand averaged amplitude curves over stride, amplitude normalized curves and variation between all included strides were calculated for all muscles and subjects, respectively. The normal range of all calculated parameters was defined within the span between the 5th and the 95th percentiles of all pain free subjects. Data were evaluated according to deviations from the normal range. Already before the load situation, pain subjects showed considerable deviations from the normal range, mainly of their abdominal muscles. There was no relationship between magnitude of deviation and pain intensity, but perceived exertion was highest in those subjects who showed the most symptoms in terms of number of muscles being identified as considerably deviating from the normal range. No specific “dysfunction pattern” could be identified, which argues for highly individual mechanisms instead of a single target muscle. The results suggest cumulative effects of different disturbance levels resulting in acute back pain. Since deviations could be identified already before the pain occurred, disturbed muscle function seems to be a risk factor for developing back pain. Further investigations aimed at clear identification of and, as a second step, correction of muscle function are necessary. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Gait analysis; Surface EMG; Acute low back pain; Trunk muscle co-ordination; Treadmill walking
1. Introduction Low back pain is an enormous socioeconomic problem in all industrialized countries. In Germany alone, the total costs of low back pain (LBP) are estimated at about D 17 billion per year [1]. Most of these expenses are caused by chronic LBP. There is no doubt that very high peak and also permanent ∗
Corresponding author. Tel.: +49 3641 937313; fax: +49 3641 937377. E-mail address:
[email protected] (C. Anders).
0928-4680/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pathophys.2005.09.001
loads on the spine may cause pain [2]. On the other hand a lot of people complaining about low back pain rather have to be rated as “underusing” their musculo-skeletal system. In these cases, radiological diagnosis mostly fails to give hints about morphological reasons for the pain [3]. Psychological factors like job satisfaction [4] or workplace social environment [2] are rated as important, but these rather correlate with acute back pain than with the chronic disease course. Anyhow, this “unspecific” LBP is often reduced to psychophysiological characteristics of the patients.
C. Anders et al. / Pathophysiology 12 (2005) 243–247
244
The so-called “muscle systems” [5] may give a clue for understanding the basic mechanisms of low back pain. Since our spine has to fulfill quite different demands, both stability and mobility have to be permanently ensured by the musculoskeletal system. To do this, highly co-ordinated activations are necessary. The stabilizing muscles stiffen the spine and simultaneously limit its range of motion [6]. The mobilizing muscles initiate movements and produce high forces [6]. Deficits of the stabilizing function can be measured only indirectly. Incorrect co-ordination of local muscles could be measured in low back pain patients by Hodges and Richardson [7] in terms of delayed pre-activation towards voluntary movements. The concept of the neutral zone, primarily introduced by Panjabi [8,9] is able to characterize the amount of passive control of spinal stability. Recently, this could be quantified by investigations of disturbed proprioceptive perception in LBP patients [10]. References increase, arguing for a lack of “the essential” muscle [11,12] rather than highly co-ordinated intermuscular activation for ensuring the necessary spinal stabilization. Furthermore, provocations of the whole system seem to be a promising tool to get more insight in the LBP pathogenesis. Since functional data of everyday activities are rare there is still a great gap of information. Therefore the aim of this study was to investigate trunk muscle activation and coordination during treadmill walking in healthy persons and subjects with acute low back pain.
2. Subjects and methods Altogether, 16 female subjects were recruited (Table 1). They were all free of LBP at the start of the investigation. Four of these persons complained LBP after static spinal loading (pain subjects, see below). Twelve healthy women (22–43 years, median: 24.5 years) who even after the static Table 1 Investigated subjects: age, pain level after static load situation and perceived exertion at the end of static load situation [20]
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 P1 P2 P3 P4
Age (years)
VAS (100−1 )
Borg [6–20]
27 23 24 23 22 23 24 25 40 43 28 31 24 26 25 23
0 0 0 0 0 0 0 0 0 0 0 0 14 36 26 48
10 13 13 10 9 7 12 13 13 6 14 13 12 15 15 13
H: healthy subjects; P: pain subjects.
loading remained LBP symptom free served as the reference group. The study was approved by the local ethics committee (0558-11/00) and, therefore, fulfilled the conditions of the Declaration of Helsinki. Informed written consent was obtained from every volunteer. After an adequate habituation phase of at least 5 min every person walked on treadmill for approximately 1 min at a speed of 4 km/h (1.11 m/s, speed for relaxed walking). The investigation was performed two times during the investigation day: the first one was carried out in the morning, the second in the afternoon after all subjects completed a static loading situation of their spine for 2 h. The load was applied by wearing a waistcoat, loaded with a weight of approximately 25% of each subject’s upper body mass ([13], accuracy: 1 kg). Bipolar surface EMG (SEMG, 5–700 Hz, Biovision, Wehrheim, Germany) was taken from five trunk muscles [14,15], muscles and electrode positions see Table 2. For this investigation disposable Ag–AgCl electrodes (Arbo® , Germany) with a circular uptake area of 1 cm diameter and an inter-electrode distance of 2.5 cm were used. Data were stored on computer for further offline analysis (AD-conversion at 2000 s−1 , DAQCard-AI-16E-4: 12 bit, National Instruments, USA, resolution: 1.0 V/bit). Force signals from both heels were used to identify stride cycles using a semi-automatic software algorithm (Matlab® ) including visual control. Cadence time was analyzed and only strides within 25% deviation from the calculated median time of all respective strides were used for analysis. The number of strides used for calculation varied from 40 to 70, depending on exact recording time and number of eliminated strides due to technical problems. Raw SEMG was centered and high-pass filtered (fourth-order Butterworth filter, 20 Hz) to avoid influences from movement artifacts. Root mean square (rms) was calculated subsequently using a window of 50 ms. Initially, strides were time-normalized to avoid variances originating from remaining inconsistencies in stride length. Time Table 2 Investigated trunk muscles and respective electrode positions (according to [14,15]) Muscle
Electrode orientation and position
M. rectus abdominis (upper part, RA l/r) M. obliquus internus abdominis (OI l/r) M. obliquus externus abdominis (OE l/r)
Vertical, 4 cm lateral umbilicus, caudal electrode at level of umbilicus Along horizontal line between both ASIS’s, medial from inguinal ligament Along line from most inferior point of costal margin to opposite pubic tubercle, cranial electrode directly below most inferior point of costal margin 1 cm medial and parallel to line between PSIS’s and 1st palpable spinous process, caudal electrode at L4 level Vertical, over palpable bulge of muscle (approximately 3 cm lateral from midline) caudal electrode at L1 level
M. multifidus (lumbalis, MF l/r) M. erector spinae (longissimus, ES l/r)
Muscles from both sides were investigated simultaneously. ASIS: anterior superior iliac spine; PSIS: posterior superior iliac spine.
C. Anders et al. / Pathophysiology 12 (2005) 243–247
245
Table 3 Percentage of stride time located outside the range between the 5th and 95th percentiles of investigated healthy subjects: absolute SEMG amplitudes, before/after load application
P1 P2 P3 P4
ral
rar
oil
oir
oel
oer
mfl
mfr
esl
esr
−/− 44/69 44/− 34/50
−/− 62/52 −/− −/−
92/57 −/− −/− −/−
77/46 −/73 100/51 −/−
−/− 49/43 69/63 −/−
−/− −/50 −/− 72/−
−/− −/− −/− −/−
−/− −/− −/− −/35
−/− −/− 57/76 −/−
−/− −/− −/− −/−
Table 4 Percentage of normalized stride time located outside the range between the 5th and 95th percentiles of investigated healthy subjects: relative SEMG amplitudes, before/after load application
P1 P2 P3 P4
ral
rar
oil
oir
oel
oer
mfl
mfr
esl
esr
51/− −/− 80/51 −/−
43/− −/− 43/38 −/−
−/− 54/39 −/− −/−
59/− −/− −/48 47/62
41/− 53/50 53/− −/−
−/− −/− −/− −/−
−/− 70/60 −/− −/−
−/− 48/− −/− −/−
−/− −/− −/− 53/−
−/− −/− −/− 88/−
Table 5 Percentage of stride time located outside the range between the 5th and 95th percentiles of investigated healthy subjects: variation coefficient, before/after load application
P1 P2 P3 P4
ral
rar
oil
oir
oel
oer
mfl
mfr
esl
esr
−/− −/− 41/64 −/−
−/− −/− −/69 −/−
−/− 56/69 −/36 −/−
39/42 48/50 −/64 −/−
−/− −/− −/− −/−
−/− −/− −/− −/−
−/− 46/− 44/− −/−
−/− −/− −/− −/42
−/− 60/− −/− −/−
−/− 61/44 −/− −/−
normalization was calculated with an accuracy of 0.5% (201 time points). Subsequently, grand averaged SEMG curves were calculated separately for both investigation times, all muscles and subjects, respectively. To estimate the repeatability of muscle activation patterns variation coefficients (VC) were calculated for all relative time points. Furthermore, the ensembled SEMG curves were amplitude-normalized according to the maximum level within stride to balance different rms values between subjects. By doing this inter-individual amplitude differences of the SEMG curves were reduced drastically [16,17] and activation patterns could be analyzed. For every parameter the span between the 5th and 95th percentiles of all healthy subjects was defined as the normal range and, therefore, covered 90% of all values. According to the calculation and the number of subjects that were investigated, pointwise highest and lowest values both were located outside the normal range. This, in turn, resulted in an expected range of 16.7% of stride time laying outside the defined range for every healthy volunteer and muscle, respectively. Therefore, data of the pain subjects were rated as considerably deviating if their curves remained at least 34% of the normalized stride time outside the defined normal range.
in two cases back muscles showed deviations, whereas in nine cases abdominal muscles were noticeable. Except for one subject (P3), if relative amplitudes were assessed, i.e. coordination pattern, the number of deviating muscles increased (Table 4). Four muscles identified as considerably deviating according to absolute amplitude levels also showed demonstrative curves of their relative amplitudes. Considerably increased variations between strides occurred in only eight muscles (Table 5). One pain subject (P4) did not show any demonstrative variations, whereas in pain subject P2 five muscles were detected. After load application, absolute amplitude levels of 12 muscles deviated from the normal range. These were again mainly abdominal muscles: in only two cases amplitude levels of back muscles deviated considerably from the defined normal range (Table 3). Pattern analysis revealed no distinctive feature for P1 and only one deviating muscle for P4, whereas the other pain subjects (P2, P3) showed three demonstrative muscles (Table 4). In only one case back muscles could be identified as considerably deviating. VC identification differed between pain subjects: both, P1 and P4 only showed one significant muscle, whereas P3 was characterized by four muscles with deviating VC levels (Table 5).
3. Results
4. Discussion
Before the load situation every pain subject was characterized by at least two muscles considerably deviating from the defined normal range of the absolute SEMG levels (Table 3):
The results of this study highlighted the relationship between the acute LBP and disturbed muscular coordination. This could be demonstrated for absolute SEMG
246
C. Anders et al. / Pathophysiology 12 (2005) 243–247
amplitudes, SEMG patterns, and the repeatability of the measured SEMG curves during gait. Interestingly, pain subjects already showed considerable deviating parameter levels before the load situation, i.e. at investigation times when they still reported no pain. Furthermore, the number of cases, i.e. muscles that were noticeable were slightly higher before the pain occurred in comparison to the situation after load application, when the pain was present. Therefore, to identify the primarily underlying process the question arises: do coordination disturbances lead to low back pain or does low back pain corrupt muscle coordination? In a recently developed concept about the pathogenesis of LBP [18] the vicious circle of back pain starts with sub-failure, i.e. soft tissue damage, resulting in faulty proprioception and therefore corrupted muscle coordination. This concept should therefore require preceding load exceeding the individual failure tolerance [19]. In our investigation, subjects had to wear a waistcoat with considerable load for 2 h without any break. This load was applied after they already had completed a test in the morning. Therefore, it can be assumed that this load significantly exceeded the individual tolerance in the four pain subjects. This was in accordance with perceived exertion just at the end of the loading situation which developed considerable levels (maximum level: 15 (hard: P2, P3)) according to Borg-scale [20]. Median Borg-scale level in the reference group was 12.5 (between fairly light and moderately hard). P2 and P3 pain subjects were characterized by the largest number of distinctive features: before and after load. In the alternative, load application or load levels are not by themselves sufficient prerequisites for acute pain. Individual properties, here quantified by SEMG parameters, have to be considered. The identification of functional characteristics prior to pain may help to quantify risk levels. If aimed at prevention rather than diagnosis of back pain, SEMG investigation may be a promising tool in future. However, diagnosis still seems to be fuzzy: deviations from the normal range were randomly distributed among pain subjects and investigated muscles. No specific alteration pattern could be identified. Furthermore, normal range was defined as laying between the 5th and 95th percentiles of all pain free subjects and, therefore, included 90% of all measured values. Unavoidably, 10% of the measured values of the healthy volunteers are also located outside the defined range, which were amounted to about 17% of normalized stride time for every muscle. Therefore, the data basis had to be expanded to allow more detailed diagnosis. In a recent investigation by Lamoth et al. [21] stride to stride variability was used to discriminate between pain and pain free situations for experimentally induced pain. In our investigation, SEMG variability not necessarily correlated with pain. Only if perceived exertion was rated as “hard” variance increased. This was observed predominantly for abdominal muscles. Since deviant muscle coordination plays an important role in the development of low back pain [12] establishment of normal intra- and intermuscular activation pattern, via SEMG
analysis, would be helpful in diagnostics and treatment. The concept of equilibrium point hypothesis [22] although originally applied for fast movements, may give hints for this, but was not analyzed in this study. In conclusion, a deviant trunk muscle activation pattern may occur already without low back pain. With the application of an additional load it is possible to gain significantly more information. The combination of high level of perceived exertion and the occurrence of pain was connected with the highest number of demonstrative parameters. SEMG investigations seem to be a promising tool for preventive diagnostics of low back pain.
Acknowledgements This study was supported by the Center for Interdisciplinary Prevention of Diseases related to Professional Activities, founded by FSU Jena and BGN. The authors wish to thank Ms. Marcie Matthews for language correction.
References [1] W. Bolten, A. Kempel-Waibel, W. Pforringer, Analysis of the cost of illness in backache, Med. Klin. 93 (1998) 388–393. [2] R. Norman, R. Wells, P. Neumann, J. Frank, H. Shannon, M. Kerr, A comparison of peak vs. cumulative physical work exposure risk factors for the reporting of low back pain in the automotive industry, Clin. Biomech. (Bristol, Avon) 13 (1998) 561–573. [3] M.W. van Tulder, W.J. Assendelft, B.W. Koes, L.M. Bouter, Spinal radiographic findings and nonspecific low back pain. A systematic review of observational studies, Spine 22 (1997) 427–434. [4] S.J. Bigos, M.C. Battie, D.M. Spengler, L.D. Fisher, W.E. Fordyce, T.H. Hansson, A.L. Nachemson, M.D. Wortley, A prospective study of work perceptions and psychosocial factors affecting the report of back injury, Spine 16 (1991) 1–6. [5] A. Bergmark, Stability of the lumbar spine. A study in mechanical engineering, Acta Orthop. Scand. Suppl. 60 (1989) 1–54. [6] M.J. Comerford, S.L. Mottram, Movement and stability dysfunction—contemporary developments, Man Ther. 6 (2001) 15–26. [7] P.W. Hodges, C.A. Richardson, Delayed postural contraction of transversus abdominis in low back pain associated with movement of the lower limb, J. Spinal Disord. 11 (1998) 46–56. [8] M.M. Panjabi, The stabilizing system of the spine. Part II. Neutral zone and instability hypothesis, J. Spinal Disord. 5 (1992) 390– 396. [9] M.M. Panjabi, The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, and enhancement, J. Spinal Disord. 5 (1992) 383–389. [10] V. Leinonen, M. Kankaanpaa, M. Luukkonen, M. Kansanen, O. Hanninen, O. Airaksinen, S. Taimela, Lumbar paraspinal muscle function, perception of lumbar position, and postural control in disc herniation-related back pain, Spine 28 (2003) 842–848. [11] S.M. McGill, S. Grenier, N. Kavcic, J. Cholewicki, Coordination of muscle activity to assure stability of the lumbar spine, J. Electromyogr. Kinesiol. 13 (2003) 353–359. [12] J.H. van Dieen, J. Cholewicki, A. Radebold, Trunk muscle recruitment patterns in patients with low back pain enhance the stability of the lumbar spine, Spine 28 (2003) 834–841. [13] M. Kankaanp¨aa¨ , S. Taimela, C.L. Webber, O. Airaksinen, O. H¨anninen, Lumbar paraspinal muscle fatigability in repetitive isoin-
C. Anders et al. / Pathophysiology 12 (2005) 243–247
[14]
[15]
[16]
[17]
ertial loading: EMG spectral indices, Borg scale and endurance time, Eur. J. Appl. Physiol. 76 (1997) 236–242. H.J. Hermens, B. Freriks, R. Merletti, D.F. Stegeman, J. Blok, G. Rau, C. Disselhorst-Klug, G. H¨agg, European recommendations for surface electromyography, results of the SENIAM Project, Roessingh Research and Development b.v., Roessingh, 1999. J.K. Ng, V. Kippers, C.A. Richardson, Muscle fibre orientation of abdominal muscles and suggested surface EMG electrode positions, Electromyogr. Clin. Neurophysiol. 38 (1998) 51–58. J.F. Yang, D.A. Winter, Electromyographic amplitude normalization methods: improving their sensitivity as diagnostic tools in gait analysis, Arch. Phys. Med. Rehabil. 65 (1984) 517–521. D.A. Winter, H.J. Yack, EMG profiles during normal human walking: stride-to-stride and inter-subject variability, Electroencephalogr. Clin. Neurophysiol. 67 (1987) 402–411.
247
[18] M.M. Panjabi, Consequences of a subfailure injury. A hypothesis of chronic spine pain, in: IV World Congress of Biomechanics, Calgary, 2002. [19] S. McGill, Low back disorders: evidence-based prevention and rehabilitation, Human Kinetics, Champaign, 2002. [20] G. Borg, Perceived exertion as an indicator of somatic stress, Scand. J. Rehabil. Med. 2 (1970) 92–98. [21] C.J. Lamoth, A. Daffertshofer, O.G. Meijer, G. Lorimer Moseley, P.I. Wuisman, P.J. Beek, Effects of experimentally induced pain and fear of pain on trunk coordination and back muscle activity during walking, Clin. Biomech. (Bristol, Avon) 19 (2004) 551– 563. [22] M.L. Latash, G.L. Gottlieb, An equilibrium-point model for fast, single-joint movement. I. Emergence of strategy-dependent EMG patterns, J. Mot. Behav. 23 (1991) 163–177.
