- Email: [email protected]
Lumbar intradiscal pressure and whole-body vibration – first results
Clinical Biomechanics 16 Supplement No. 1 (2001) S127±S134
www.elsevier.com/locate/clinbiomech
Lumbar intradiscal pressure and whole-body vibration ± ®rst results Aõcha El-Khatib *, Francßois Guillon University Research Group on Spinal Biomechanics, University Hospital Raymond Poincar e , 104, Boulevard Raymond Poincar e, 92380 Garches, France
Abstract Objective. To contribute to the scienti®c background for the assessment of the health risk at the lumbar spine from whole-body vibration. Design. Experimental study. Background. Many workers have monitored the vibration at spinal locations on the skin or using skeleton mounted devices in order to assess the vertebral response when sitting. They have shown that resonance occurred in the 4±7 Hz range. Considering the dierent structures of the intervertebral joint, it seemed interesting to assess their behaviour separately, using intranuclear pressures at the lumbar discs monitored simultaneously with the vertebral accelerations. Methods. Seven unembalmed cadavers were submitted to 5 min. whole-body random vibration, in four seated postures (erect or as in a car, both with or without a lumbar support). Power spectral density functions were estimated at each lumbar level and each posture and the dominant frequency identi®ed. The energy of the pressure signal was also estimated in the 0±25 Hz band. Analysis of variance was then performed to study the eects of subject, disc level, and posture. Results. Energy of the intranuclear pressure variation decreased when leaning the backrest backwards. The eect of the lumbar support depended on the discal level and on posture. The shape of the power spectral density function suggested the existence of a cyclic loading of the nucleus pulposus, while more complex phenomena were observed at the vertebral body. Conclusions. The response of the lumbar spine cannot be assessed by examining only vertebral acceleration at one level. Relevance To understand how posture aects lumbar behaviour to minimize low back disorders. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Whole-body vibration; Lumbar spine; Intradiscal pressure; Posture
1. Introduction The role of whole-body vibration in the genesis of back disorders is documented by many epidemiological studies [1,2]. The biomechanical data showed that the intensity of the mechanical exposure of the spine also depends on factors like the anthropometry of the subjects and the posture [3±5]. The protection of the exposed workers thus requires a better comprehension of the biodynamic behaviour of the spine in a vibratory environment. Many authors measured the vertical (caudal-cephalic) vertebral accelerations, at the lumbar level, using skeleton mounted devices [6,7], or skin mounted devices [8,9], on volunteers. But the way in which the vibration is transmitted from one lumbar level *
Corresponding author. E-mail address: [email protected] (A. ElKhatib).
to the other remains badly known. In a previous report, El-Khatib et al. [10] showed that vertical seat-to-vertebrae vibration transmission through the lumbar spine (from L5 to L1) of cadavers was almost the same at each lumbar vertebra, as well for the frequency as for the modulus. But the behaviour of the lumbar spine depends on the stacking of two essentially very dierent structures: an osseous structure for the vertebrae, and a ligamentous and cartilaginous structure for the intervertebral discs and the vertebral endplates. The measurements taken on the vertebrae, in particular at the level of the vertebral body, then give an indication on the total behaviour of the intervertebral disc. They do not allow to distinguish easily the response of the nucleus pulposus from that of the annulus ®brosus, whereas low back disorders caused by whole-body vibration are frequently attributed to the intervertebral disc [11] and for some of them compensated as occupational diseases in France [12].
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 0 ) 0 0 1 0 4 - 2
S128
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
One way to assess separately the behaviour of the nucleus pulposus is the use of the intradiscal pressure measurement [13] which will be called more appropriately the intranuclear pressure hereafter. Indeed, the stresses measured in the annulus show that its behaviour is dierent from that of the nucleus [14] thus they are considered independently when modelling [15]. There are some experimental data concerning the variations of the intranuclear pressure measured on lumbar specimens [13,16±18] or in vivo [19±25]. The hydrostatic behaviour of the nucleus of the non-degenerated disc has been demonstrated by several authors [26±28]. This measurement is used to evaluate the compression forces of the intervertebral discs in dierent postures. The available in vivo data concern mainly static postures or daily activities. Little is known about shocks [29] or vibratory [30] exposures. In all cases, published measures concern one level, mostly the L3=L4 lumbar disc. In this work, the authors sought to investigate the variations of the intranuclear pressure measured simultaneously at all lumbar disc levels (L1=L2, L2=L3, L3=L4 and L4=L5), in cadavers submitted to whole-body vibration. The main questions we aimed to test were as follows: (1) what is the frequency content of the intranuclear pressure variation when submitted to a band-limited random vibration, compared to vertical acceleration at the vertebrae; (2) does the power spectral density function of the intranuclear pressure variation depend on the lumbar disc level; (3) does it depend on posture? 2. Methods The subjects were seven unembalmed cadavers from the Corpses Donation Department of Paris, tested as soon as possible after their death (Table 1). During this period, they were stored at 4°C. Intradiscal pressure measurements were realized in each of the four lumbar intervertebral discs L1=L2; L2=L3; L3=L4; L4=L5. The pressure transducers (EPI127*-7SC, Entran, Les Clayes-sous-Bois, France) with a
measuring sealed range up to 0.7 MPa (7 bars) and a resonant frequency of 1 MHz (the useful frequency range being linear from 0 Hz to 20% of the resonant frequency), were implanted from an anterior approach, after the removal of the abdominal viscera, into the centre of the nucleus pulposus. Their positions were checked on lateral and front plane radiographs. The sensitive tip (1.3 mm2 ) of the pressure sensor was mounted at the end of an 80 mm length tube of 1.6 external diameter. The pressure transducer was inserted into the disc using a needle guide of 1.8 mm interior diameter, and 2 mm external diameter. It was secured to the guide using plastic screws. The pressure guide was screwed to the sensor supporting device ®xed to the lower vertebra (Fig. 1). The lower part of this device was ®xed to the vertebral body. The higher part was slightly behind the aforementioned. It comprised a dovetail guide making it possible to make slide, upwards or downwards, a small moving part provided with a hole through which the pressure guide was introduced into the disc. This moving part (tenon) was positioned in order to ensure that the pressure transducer would be located opposite to the centre of the disc, before inserting it into the nucleus pulposus. It was then ®xed to the mortise by means of two plastic screws on each side. Accelerations of the lumbar vertebrae were also measured on the anterior face of the vertebral body [10] using 5g damped biaxial accelerometers (EGAS2CM*-5VC, Entran, Les Clayes-Sous-Bois, France). The inclination of each accelerometer was measured on the lateral X-ray taken before each vibratory exposure and all acceleration measurements were set in the same reference. The subjects were submitted to 5 min. of whole-body, vertical, broad-band, white random vibration in the bandwidth 0.8±25 Hz of about 1.5 m/s2 r.m.s., in different seated postures on a rigid seat: erect or leaning backwards as in a car, with or without a lumbar support which was positioned opposite the third lumbar vertebra in order to create a real hyper-lordosis. The signals were conditioned (MM35 for the pressure transducers and MC 402 for the accelerometers, Entran, Les Clayes-
Table 1 Subjects' characteristics Subject
Sex
Age (yr)
Cause of death
Post-mortem days before test (days)
Weight (kg)
Weight abd. visc. (kg)
Height (cm)
1 2 3 4
Female Male Male Male
63 66 63 49
3 3 6 10
77.5 77.4 85.4 69.5
? 9.8 9.4 8.5
169 164 178 160
5 6 7
Male Male Male
50 63 53
Asthma ± cardiac crisis Myocardial infarction Road accident ± skull trauma Septicemia ± oesophagus carcinoma Cardiac arrest Kidney carcinoma Suicide by barbiturate
11 7 10
82 53 66
7.5 7.25 6.5
179 171 174
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
S129
Fig. 1. Description of the instrumentation at one vertebral level. The lower part of the sensor supporting device was ®xed to the anterior wall of the vertebral body by two screws, one on each side of the mid-sagittal plane where the vertebral accelerometer was mounted. The upper part consisted of a dovetail guide. It was slit in the middle. It allowed to position a small piece, by moving it downward or upward, opposite to the centre of the intervertebral disc above the instrumented vertebra. This piece was ®xed to the guide by two plastic screws, before inserting the pressure guide into the nucleus pulposus, through the hole located at its centre.
sous-Bois, France) then stored on an analogue tape recorder (Teac XR-9000, Signal/Noise 40 dB). Only 51.2 s of the original signal were digitized. Analogue recordings were digitized after low-pass ®ltering at 40 Hz, at a sampling rate of 80 Hz. Spectral density functions Gx;x f of each channel were estimated with a frequency resolution of 0.3 Hz, on the basis of Welch's method. The dominant frequency at each level, and for each trial, was then calculated, as well as an approximation of the energy of the pressure signal in the 0±25 Hz band, which corresponded to the integral of the power spectral density across this band. One-way analysis of variance (A N O V A ) was then performed on the dominant frequency and the energy to study the eects of the individual, discal level, and posture. The null hypothesis H0 was that the means would be equal, and the alternative hypothesis H1 that they were in fact, dierent. 3. Results The power spectral density functions from the intranuclear pressure variation measurements at the lumbar discs, in the seated erect posture without a lumbar support on a rigid seat, are shown in Fig. 2. The mean square value of the data was distributed in the 1±15 Hz bandwidth (>90% of the energy in this frequency band) and was mainly concentrated at a single frequency ranging from 5.9 to 8.1 Hz. In spite of the interindividual variability, the shape of the power spectral density function of the intranuclear pressure variation was
the same for all tested subjects (except in some cases for subject 6) and in all tested postures (Fig. 3). In all cases, the dominant frequency was identical at all lumbar discs, for each subject and each posture hence it did not depend on the discal level. It was signi®cantly dierent from one subject to another (P 0). It did not change signi®cantly when changing posture (P 0:47). The energy in the 0±25 Hz bandwidth depended very signi®cantly not only on the subject (P 0:000037) (Fig. 4), but also on the discal level (P 0:0000076) (Fig. 5) and on posture (P 0:0134) (Figs. 3 and 6), mainly when moving from a seated erect posture to a seated posture as in a car (P 0:086) which produced a decrease of the energy at all discal levels. However, when considering each discal level individually, posture did not seem to aect signi®cantly the energy of the intranuclear pressure variation (0:1345 6 P 6 0:3831). The use of a lumbar support had a dierent eect on the energy when seated erect or leaning backwards. It also acted dierently for the discs above or below the third lumbar vertebra, although the dierences were generally not signi®cant. These eects are summarized in Table 2. When seated erect, the energy at the upper discs increased when using a lumbar support (P 0:2832), while it decreased at the lower levels (P 0:2018). When seated as in a car, the lumbar support did not change the energy at the upper discs (P 0:7001) while it provoked a signi®cant decrease at the lower discs (P 0:0288). While the dominant frequency remained unchanged from the vertical acceleration at the lumbar vertebrae and the intranuclear pressure at the lumbar discs, the
S130
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Fig. 2. Individual power spectral density function of the intranuclear pressure variation in the seated erect posture without a lumbar support on a rigid seat (1 bar 0.1 MPa): ±±± , disc L4=L5; ...., disc L3=L4; - - - -, disc L2=L3; -:-:, disc L1=L2.
shape of the power spectral density functions was very dierent (Fig. 7). The most signi®cant dierence concerned the frequency content. The mean square value of the data of both parameters showed a quite similar distribution until 10 Hz. Above this frequency, the frequency response of the intranuclear pressure tended to zero, while it remained rich in all the exposure bandwidth for the vertical vertebral acceleration, i.e. until 25 Hz.
4. Discussion Although unembalmed human cadavers remain the best ®tted physiological and anatomical model of the living volunteers, still they also present major limitations which should not be neglected when examining the experimental results. These have been largely addressed previously [31]. They mainly concern the absence of myoelectric activity, the body temperature, the conser-
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
S131
Fig. 3. Mean value (for seven subjects) of the power spectral density function of the intranuclear pressure variation at each intervertebral disc L1=L2; L2=L3; L3=L4; L4=L5 for dierent postures on a rigid seat (1 bar 0.1 MPa): ±±± , seated erect without a lumbar support; ...., seated erect with a lumbar support; - - - -, seated as in a car without a lumbar support; -:-:, seated as in a car with a lumbar support.
Fig. 4. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band in all tested postures and at all discal levels (1 bar 0.1 MPa) ± Eect of the subject.
vation state, and the eect of removing of the abdominal viscera. Quandieu [32] showed that increased muscle contraction improved the low-pass ®lter characteristics of the intervertebral disc, while provoked muscle relaxation shifted the ®lter characteristics towards higher frequencies. Guillon [30] showed similar intranuclear pressure variation in cadavers compared to in vivo
Fig. 5. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band, in all tested postures for seven subjects (1 bar 0.1 MPa) ± Eect of the discal level.
measures by Nachemson and Morris [19] and by Andersson et al. [20±22] in static seated postures. However, the pressure values were lower than those obtained in vivo, which was attributed to the absence of muscle contraction as they had the same magnitude as the in vivo measures realized in para- and tetraplegic patients by Hein-Sorensen et al. [23]. The eect of a lower body temperature in cadavers (about 10°C) was not controlled. It can be assumed that this may result in a stier
S132
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Fig. 6. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band at all discal levels for seven subjects (1 bar 0.1 MPa) ± Eect of posture: Posture 1, seated erect without a lumbar support; Posture 2, seated erect with a lumbar support; Posture 3, seated as in a car without a lumbar support; Posture 4, seated as in a car with a lumbar support.
Table 2 Number of times the energy of the power spectral density function of the intranuclear pressure variation in the 0±25 Hz band increases or decreases when changing posture by adding a lumbar support for seven subjects, at each discal level Seated erect Seated as in a car Without ! with a lumbar Without ! with a lumbar support support
L1=L2 L2=L3 L3=L4 L4=L5
Number of increase
Number of decrease
Number of increase
Number of decrease
6 4 1 1
1 3 6 6
2 3 0 1
5 4 7 6
behaviour than in vivo. The disc desiccation is another uncontrolled factor through the experimentation. But disc dehydration is also observed in daily activities and is revealed by the shortening observed between the morning and the night [33]. Also, postures were tested in a semi-randomized order, so no systematic eect could be attributed to desiccation. The absence of the abdominal viscera, even though modifying the total equilibrium of the trunk, did not have probably a major eect on the intranuclear pressure measurements, as the intra-abdominal pressure is mostly related to muscle contraction and the main eect would be due to the absence of muscle tone. In vivo studies showed that axial motion occurred in the spine when submitted to vertical shocks or wholebody vibration in seated postures [6]. But generally, measures were realized at only one level [7], in the best
Fig. 7. Mean value (for seven subjects) of the power spectral density of the vertical vertebral acceleration at each lumbar vertebra (black continuous lines) and of the intranuclear pressure variation at each lumbar disc level (grey lines): ±±± , L1=L2; ...., L2=L3; - - - -, L3=L4; -:-:, L4/L5 (1 bar 0.1 MPa).
cases at two locations of the lumbar spine [6]. Still, they suggested a great rigidity in the lumbar spine. In a previous study, the authors also reported a rigid behaviour of the lumbar spine, by examining the vertical seat-to-vertebrae transfer function [10,31]. One original aspect of this experiment was the monitoring of the intranuclear pressure simultaneously to the vertebral
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
acceleration, at all lumbar levels (®ve vertebrae and four discs) in seven subjects and four seated postures. As the vertebral column is constituted alternatively of two very dierent components, the vertebrae and the intervertebral discs, it was interesting to look at the response of each of the structures separately, which may allow a direct assessment of the risk. In the present study, the authors compared measurements at the nucleus pulposus to those at the vertebral body, in order to understand what was their contribution to vibration transmission through the lumbar spine, and how it was modi®ed by posture. The results showed behaviour dierent from the nucleus pulposus and the vertebral body. In the ®rst case, frequency response was concentrated at a single frequency ranging from 5.9 to 8.1 Hz in the 1±15 Hz bandwidth, while in the second case, frequency response was distributed in all the exposure bandwidth (1±25 Hz) with a peak at the same dominant frequency as the intranuclear pressure variation. Therefore, responses of both structures suggested the existence of a cyclical phenomenon at the same frequency as for a single-degree-of-freedom system. While accelerations measured at the vertebral body may give an indication on the total behaviour of the intervertebral disc, pressures measured in the nucleus pulposus re¯ected only the behaviour at the centre of the disc. Hence, dierences observed in the power spectral density function of the vertebral acceleration and the intranuclear pressure variation, re¯ecting a more stochastic phenomenon, may arise from the annulus ®brosus or be attributed to the facet joints [34,35]. The power spectral density function of the intranuclear pressure variation as well as of the vertebral acceleration peaked at the same dominant frequency at all levels, for each posture and each subject. On average, resonance frequency at the vertebrae (vertical seat-tovertebrae transmissibility) was at 6.3 Hz in the seated erect posture without lumbar support [10]. It was slightly higher than the published in vivo results by Panjabi et al. [6] (4.4 Hz), Pope et al. [7] (5 Hz), Broman et al. [36] (5±6.3 Hz) and Magnusson et al. [37] (5.8 Hz). These dierences were partly explained by variations between the experimental methods [31]. To our knowledge, there are no published data concerning intranuclear pressure variations when submitted to whole-body vibration in dierent seated postures. Andersson et al. [20±22] as well as Guillon [30] showed, in static, that the intranuclear pressure at the L3=L4 disc level decreased when leaning backwards and when using a lumbar support. Guillon [30] showed that the maximum intranuclear pressure variation at the L3=L4 intervertebral disc was also proportional to the kinetic energy for impacts on the head from dierent heights. In the present study, intranuclear pressure was measured simultaneously at all lumbar disc levels in four seated postures, for a random vibration. The energy in the 0±25
S133
Hz range was then estimated and this parameter was used to study the eect of measurement level or posture. As for the static measurements [20±22,30], the energy of the intranuclear pressure variation decreased when moving from a seated erect posture to a seated posture as in a car, at all lumbar disc levels. However, lumbar support did not have the same eect whether considering the upper discs (above L3) or the lower discs (below L3), when seated erect or seated as in a car. The energy tended to increase when using a lumbar support at the L1=L2 and L2=L3 intervertebral discs, while it always decreased at the L3=L4 and L4=L5 intervertebral discs. Dierent tendencies were observed on the vertical seatto-vertebrae transmissibility [10,31]: the transmissibility peak increased when moving from a seated erect posture to a seated as in a car posture. It decreased when using a lumbar support in the seated erect posture and increased when using a lumbar support in the seated as a car posture. It should be noticed that the experimental lumbar support used in this study (located opposite to L3, about 4 cm ahead of the backrest) intended to realize a real hyper-lordosis and would be unbearable in a reallife driving posture. The intranuclear pressure variation energy depended on the vertebral level, contrary to the vertical vertebral acceleration. However, these variations should be considered with some caution, as they should be related to the intervertebral disc degree of degeneration and to the surface of the intervertebral disc (with that of the nucleus pulposus specially). 5. Conclusions These results suggest that the health risk at the lumbar spine from whole-body vibration cannot be assessed only by examining the acceleration measurements, even if they were monitored at the vertebral level, and that it might be hazardous to conclude from observations at one level. Indeed, results from vertical seat-to-vertebrae transfer functions implied a sti connection between the lumbar vertebrae. Those from the intranuclear pressure measurements showed an eect of the discal level as well as of posture. Although, changing posture did not aect the same way at all discal levels. While the transmissibility peaks of vertical seatto-vertebrae were increased when leaning backwards, the energy of the intranuclear pressure variation decreased as for static postures. The lumbar support increased the energy at the upper disc levels (above L3) when seated erect, while the energy always decreased at the lower discs (below L3). The shape of the power spectral density function suggested the existence of a cyclic loading of the nucleus pulposus, while more complex phenomena were observed at the vertebral level.
S134
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Acknowledgements This work could not have been completed without the participation of the Corpses Donation to medicine. In memory of the donors. References [1] Burdorf A, Sorock G. Positive and negative evidence of risk factors for back disorders. Scand J Rehabil Med 1997;23:243±56. [2] Bovenzi M, Hulshof CTJ. An updated review of epidemiological studies on the relationship between exposure to whole-body vibration and low back pain. J Sound Vib 1998;215:595±612. [3] Messenger AJ, Grin MJ. Eects of anthropometric and postural variables on the transmission of whole-body vertical vibration from seat-to-head. Report. Southampton (UK): Institute of Sound and Vibration, Human Factors Research Unit; 1989 Sept. Report No. 172. [4] Grin MJ. Handbook of human vibration. London: Academic Press; 1990. [5] Paddan GS, Grin MJ. Individual variability in the transmission of vertical vibration from seat to head. Report. Southampton (UK): Institute of Sound and Vibration, Human Factors Research Unit; 1994 Nov. Report No. 236. [6] Panjabi MM, Andersson GBJ, Jorneus L, Hult E, Mattsson L. In vivo measurements of spinal column vibrations. J Bone Joint Surg 1986;68-A:695±702. [7] Pope MH, Broman H, Hansson T. The dynamic response of a subject seated on various cushions. Ergonomic 1989;32:1155±66. [8] Hinz B, Seidel H, Brauer D, Menzel G, Bluthner R, Erdmann U. Bidimensional accelerations of lumbar vertebrae and estimation of internal spinal load during sinusoidal vertical whole-body vibration: a pilot study. Clin Biomech 1988;3:241±8. [9] Matsumoto Y, Grin MJ. Movement of the upper-body of seated subjects exposed to vertical whole-body vibration at the principal resonance frequency. J Sound Vib 1998;215:743±62. [10] El-Khatib A, Guillon F, Domont A. Vertical vibration transmission through the lumbar spine of the seated subject ± ®rst results. J Sound Vib 1998;215:763±74. [11] Kelsey JL, White III AA. Epidemiology and impact of low-back pain. Spine 1980;5:133±42. [12] Guillon F, El-Khatib A. Deux nouveaux tableaux de maladies professionnelles concernant les aections douloureuses chroniques du rachis. Revue de Medecine Orthopedique 1999;58:17± 22. [13] Nachemson A. Lumbar intradiscal pressure. Experimental studies on post-mortem material. Acta Orthop Scand 1960;43 Suppl. [14] McNally DS, Adams MA. Internal intervertebral disc mechanics as revealed by stress pro®lometry. Spine 1992;17:66±73. [15] McNally DS, Arridge RGC. An analytical model of intervertebral disc mechanics. J Biomech 1995;28:53±68. [16] Ranu HS, King AI. Correlation of intradiscal pressure with vertebral endplate pressure. Ind Mech Eng 1980;C125/80:9. [17] Merriam WF, Quinnell RC, Stockdale HR, Willis DS. The eect of postural changes on the inferred pressures within the nucleus pulposus during lumbar discography. Spine 1984;9:405±8. [18] Ramos G, Martin W. Eects of vertebral axial decompression on intradiscal pressure. J Neurosurg 1994;81:350±3. [19] Nachemson A, Morris JM. In vivo measurements of intradiscal pressure. Discometry, a method for the determination of pressure in the lower lumbar discs. J Bone Joint Surg 1964;46-A:1077±91.
[20] Andersson GBJ, Ortengren R, Nachemson A, Elfstrom G. Lumbar disc pressure and myoelectric back muscle activity during sitting. I. Studies on an experimental chair. Scand J Rehabil Med 1974;6:104±14. [21] Andersson GBJ, Ortengren R. Lumbar disc pressure and myoelectric back muscle activity during sitting. III. Studies on a wheelchair. Scand J Rehabil Med 1974;6:122±7. [22] Andersson GBJ, Ortengren R, Nachemson A, Elfstrom G. Lumbar disc pressure and myoelectric back muscle activity during sitting. IV. Studies on a car driver's seat. Scand J Rehabil Med 1974;6:128±33. [23] Hein-Sorensen O, Elfstrom G, Nachemson A. Disc pressure measurements in para- and tetraplegic patients. Scand J Rehabil Med 1979;11:1±11. [24] Schultz A, Andersson G, Ortengren R, Haderspeck K, Nachemson A. Loads on the lumbar spine. Validation of a biomechanical analysis by measurements of intradiscal pressures and myoelectric signals. J Bone Joint Surg 1982;64-A:713±20. [25] Wilke HJ, Neef P, Caimi M, Hoogland T, Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755±62. [26] McNally DS, Adams MA, Goodship AE. Development and validation of a new transducer for intradiscal pressure measurement. J Biomed Eng 1992;14:495±8. [27] El-Khatib A, Piriou P, Guillon F, Somenzi G, Tarriere C. Study of the hydrostatic behaviour of the lumbar nucleus pulposus in dynamic by the measurement of the intranuclear pressures in cadavers and in vitro. In: Proceedings of the XIVth Congress of the International Society of Biomechanics, 1993 July 4±8; Paris, France. [28] Abe E, Nickel T, Buttermann GR, Lewis JL, Transfeldt EE. Lumbar intradiscal pressure after posterolateral fusion and pedicle screw ®xation. Tohoku J Exp Med 1998;186:243±53. [29] Guillon F, Pignolet F, Somenzi G, Tarriere C, Proteau J. Eets d'un choc mou vertical sur la pression intradiscale en position assise: premiers resultats sur 4 cadavres. J de Biophysique et de Biomecanique 1986;10(Suppl 1):7±8. [30] Guillon F. Inter^et de la mesure de la pression intranucleaire pour l'etude du comportement dynamique du disque intervertebral humain. Ph.D. Dissertation, Paris XI University, Orsay (F), 1992. [31] El-Khatib A. Contribution a l'etude de la transmission des vibrations dans le rachis lombaire du sujet assis. Ph.D. Dissertation, Paris XI University, Orsay (F), 1999. [32] Quandieu P. Etude, en regime vibratoire, des fonctions de transfert des disques intervertebraux lombaires et de la masse apparente d'un primate. Ph.D. Dissertation. ReneDescartes University, Paris (F), 1981. [33] Magnusson M. Eects of seated whole-body vibrations on the spine. An experimental study in man. Ph.D. Dissertation, G oteborg University, G oteborg (Sweden), 1991. [34] Prasad P, King AI, Ewing CL. The r^ ole of articular facets during +Gz acceleration. J Appl Mech, Trans ASME 1974;41:321±6. [35] Sharma M, Langrana NA, Rodriguez J. Modelling of facet articulation as a nonlinear moving contact problem: sensitivity study on lumbar facet response. J Biomech Eng, Trans ASME 1998;120:118±25. [36] Broman H, Pope MH, Benda M, Svensson M, Ottosson C, Hansson T. The impact response of the seated subject. J Orthop Res 1991;9:150±4. [37] Magnusson M, Pope M, Rostedt M, Hansson T. Eect of backrest inclination on the transmission of vertical vibrations through the lumbar spine. Clin Biomech 1993;8:5±12.
www.elsevier.com/locate/clinbiomech
Lumbar intradiscal pressure and whole-body vibration ± ®rst results Aõcha El-Khatib *, Francßois Guillon University Research Group on Spinal Biomechanics, University Hospital Raymond Poincar e , 104, Boulevard Raymond Poincar e, 92380 Garches, France
Abstract Objective. To contribute to the scienti®c background for the assessment of the health risk at the lumbar spine from whole-body vibration. Design. Experimental study. Background. Many workers have monitored the vibration at spinal locations on the skin or using skeleton mounted devices in order to assess the vertebral response when sitting. They have shown that resonance occurred in the 4±7 Hz range. Considering the dierent structures of the intervertebral joint, it seemed interesting to assess their behaviour separately, using intranuclear pressures at the lumbar discs monitored simultaneously with the vertebral accelerations. Methods. Seven unembalmed cadavers were submitted to 5 min. whole-body random vibration, in four seated postures (erect or as in a car, both with or without a lumbar support). Power spectral density functions were estimated at each lumbar level and each posture and the dominant frequency identi®ed. The energy of the pressure signal was also estimated in the 0±25 Hz band. Analysis of variance was then performed to study the eects of subject, disc level, and posture. Results. Energy of the intranuclear pressure variation decreased when leaning the backrest backwards. The eect of the lumbar support depended on the discal level and on posture. The shape of the power spectral density function suggested the existence of a cyclic loading of the nucleus pulposus, while more complex phenomena were observed at the vertebral body. Conclusions. The response of the lumbar spine cannot be assessed by examining only vertebral acceleration at one level. Relevance To understand how posture aects lumbar behaviour to minimize low back disorders. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Whole-body vibration; Lumbar spine; Intradiscal pressure; Posture
1. Introduction The role of whole-body vibration in the genesis of back disorders is documented by many epidemiological studies [1,2]. The biomechanical data showed that the intensity of the mechanical exposure of the spine also depends on factors like the anthropometry of the subjects and the posture [3±5]. The protection of the exposed workers thus requires a better comprehension of the biodynamic behaviour of the spine in a vibratory environment. Many authors measured the vertical (caudal-cephalic) vertebral accelerations, at the lumbar level, using skeleton mounted devices [6,7], or skin mounted devices [8,9], on volunteers. But the way in which the vibration is transmitted from one lumbar level *
Corresponding author. E-mail address: [email protected] (A. ElKhatib).
to the other remains badly known. In a previous report, El-Khatib et al. [10] showed that vertical seat-to-vertebrae vibration transmission through the lumbar spine (from L5 to L1) of cadavers was almost the same at each lumbar vertebra, as well for the frequency as for the modulus. But the behaviour of the lumbar spine depends on the stacking of two essentially very dierent structures: an osseous structure for the vertebrae, and a ligamentous and cartilaginous structure for the intervertebral discs and the vertebral endplates. The measurements taken on the vertebrae, in particular at the level of the vertebral body, then give an indication on the total behaviour of the intervertebral disc. They do not allow to distinguish easily the response of the nucleus pulposus from that of the annulus ®brosus, whereas low back disorders caused by whole-body vibration are frequently attributed to the intervertebral disc [11] and for some of them compensated as occupational diseases in France [12].
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 0 ) 0 0 1 0 4 - 2
S128
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
One way to assess separately the behaviour of the nucleus pulposus is the use of the intradiscal pressure measurement [13] which will be called more appropriately the intranuclear pressure hereafter. Indeed, the stresses measured in the annulus show that its behaviour is dierent from that of the nucleus [14] thus they are considered independently when modelling [15]. There are some experimental data concerning the variations of the intranuclear pressure measured on lumbar specimens [13,16±18] or in vivo [19±25]. The hydrostatic behaviour of the nucleus of the non-degenerated disc has been demonstrated by several authors [26±28]. This measurement is used to evaluate the compression forces of the intervertebral discs in dierent postures. The available in vivo data concern mainly static postures or daily activities. Little is known about shocks [29] or vibratory [30] exposures. In all cases, published measures concern one level, mostly the L3=L4 lumbar disc. In this work, the authors sought to investigate the variations of the intranuclear pressure measured simultaneously at all lumbar disc levels (L1=L2, L2=L3, L3=L4 and L4=L5), in cadavers submitted to whole-body vibration. The main questions we aimed to test were as follows: (1) what is the frequency content of the intranuclear pressure variation when submitted to a band-limited random vibration, compared to vertical acceleration at the vertebrae; (2) does the power spectral density function of the intranuclear pressure variation depend on the lumbar disc level; (3) does it depend on posture? 2. Methods The subjects were seven unembalmed cadavers from the Corpses Donation Department of Paris, tested as soon as possible after their death (Table 1). During this period, they were stored at 4°C. Intradiscal pressure measurements were realized in each of the four lumbar intervertebral discs L1=L2; L2=L3; L3=L4; L4=L5. The pressure transducers (EPI127*-7SC, Entran, Les Clayes-sous-Bois, France) with a
measuring sealed range up to 0.7 MPa (7 bars) and a resonant frequency of 1 MHz (the useful frequency range being linear from 0 Hz to 20% of the resonant frequency), were implanted from an anterior approach, after the removal of the abdominal viscera, into the centre of the nucleus pulposus. Their positions were checked on lateral and front plane radiographs. The sensitive tip (1.3 mm2 ) of the pressure sensor was mounted at the end of an 80 mm length tube of 1.6 external diameter. The pressure transducer was inserted into the disc using a needle guide of 1.8 mm interior diameter, and 2 mm external diameter. It was secured to the guide using plastic screws. The pressure guide was screwed to the sensor supporting device ®xed to the lower vertebra (Fig. 1). The lower part of this device was ®xed to the vertebral body. The higher part was slightly behind the aforementioned. It comprised a dovetail guide making it possible to make slide, upwards or downwards, a small moving part provided with a hole through which the pressure guide was introduced into the disc. This moving part (tenon) was positioned in order to ensure that the pressure transducer would be located opposite to the centre of the disc, before inserting it into the nucleus pulposus. It was then ®xed to the mortise by means of two plastic screws on each side. Accelerations of the lumbar vertebrae were also measured on the anterior face of the vertebral body [10] using 5g damped biaxial accelerometers (EGAS2CM*-5VC, Entran, Les Clayes-Sous-Bois, France). The inclination of each accelerometer was measured on the lateral X-ray taken before each vibratory exposure and all acceleration measurements were set in the same reference. The subjects were submitted to 5 min. of whole-body, vertical, broad-band, white random vibration in the bandwidth 0.8±25 Hz of about 1.5 m/s2 r.m.s., in different seated postures on a rigid seat: erect or leaning backwards as in a car, with or without a lumbar support which was positioned opposite the third lumbar vertebra in order to create a real hyper-lordosis. The signals were conditioned (MM35 for the pressure transducers and MC 402 for the accelerometers, Entran, Les Clayes-
Table 1 Subjects' characteristics Subject
Sex
Age (yr)
Cause of death
Post-mortem days before test (days)
Weight (kg)
Weight abd. visc. (kg)
Height (cm)
1 2 3 4
Female Male Male Male
63 66 63 49
3 3 6 10
77.5 77.4 85.4 69.5
? 9.8 9.4 8.5
169 164 178 160
5 6 7
Male Male Male
50 63 53
Asthma ± cardiac crisis Myocardial infarction Road accident ± skull trauma Septicemia ± oesophagus carcinoma Cardiac arrest Kidney carcinoma Suicide by barbiturate
11 7 10
82 53 66
7.5 7.25 6.5
179 171 174
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
S129
Fig. 1. Description of the instrumentation at one vertebral level. The lower part of the sensor supporting device was ®xed to the anterior wall of the vertebral body by two screws, one on each side of the mid-sagittal plane where the vertebral accelerometer was mounted. The upper part consisted of a dovetail guide. It was slit in the middle. It allowed to position a small piece, by moving it downward or upward, opposite to the centre of the intervertebral disc above the instrumented vertebra. This piece was ®xed to the guide by two plastic screws, before inserting the pressure guide into the nucleus pulposus, through the hole located at its centre.
sous-Bois, France) then stored on an analogue tape recorder (Teac XR-9000, Signal/Noise 40 dB). Only 51.2 s of the original signal were digitized. Analogue recordings were digitized after low-pass ®ltering at 40 Hz, at a sampling rate of 80 Hz. Spectral density functions Gx;x f of each channel were estimated with a frequency resolution of 0.3 Hz, on the basis of Welch's method. The dominant frequency at each level, and for each trial, was then calculated, as well as an approximation of the energy of the pressure signal in the 0±25 Hz band, which corresponded to the integral of the power spectral density across this band. One-way analysis of variance (A N O V A ) was then performed on the dominant frequency and the energy to study the eects of the individual, discal level, and posture. The null hypothesis H0 was that the means would be equal, and the alternative hypothesis H1 that they were in fact, dierent. 3. Results The power spectral density functions from the intranuclear pressure variation measurements at the lumbar discs, in the seated erect posture without a lumbar support on a rigid seat, are shown in Fig. 2. The mean square value of the data was distributed in the 1±15 Hz bandwidth (>90% of the energy in this frequency band) and was mainly concentrated at a single frequency ranging from 5.9 to 8.1 Hz. In spite of the interindividual variability, the shape of the power spectral density function of the intranuclear pressure variation was
the same for all tested subjects (except in some cases for subject 6) and in all tested postures (Fig. 3). In all cases, the dominant frequency was identical at all lumbar discs, for each subject and each posture hence it did not depend on the discal level. It was signi®cantly dierent from one subject to another (P 0). It did not change signi®cantly when changing posture (P 0:47). The energy in the 0±25 Hz bandwidth depended very signi®cantly not only on the subject (P 0:000037) (Fig. 4), but also on the discal level (P 0:0000076) (Fig. 5) and on posture (P 0:0134) (Figs. 3 and 6), mainly when moving from a seated erect posture to a seated posture as in a car (P 0:086) which produced a decrease of the energy at all discal levels. However, when considering each discal level individually, posture did not seem to aect signi®cantly the energy of the intranuclear pressure variation (0:1345 6 P 6 0:3831). The use of a lumbar support had a dierent eect on the energy when seated erect or leaning backwards. It also acted dierently for the discs above or below the third lumbar vertebra, although the dierences were generally not signi®cant. These eects are summarized in Table 2. When seated erect, the energy at the upper discs increased when using a lumbar support (P 0:2832), while it decreased at the lower levels (P 0:2018). When seated as in a car, the lumbar support did not change the energy at the upper discs (P 0:7001) while it provoked a signi®cant decrease at the lower discs (P 0:0288). While the dominant frequency remained unchanged from the vertical acceleration at the lumbar vertebrae and the intranuclear pressure at the lumbar discs, the
S130
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Fig. 2. Individual power spectral density function of the intranuclear pressure variation in the seated erect posture without a lumbar support on a rigid seat (1 bar 0.1 MPa): ±±± , disc L4=L5; ...., disc L3=L4; - - - -, disc L2=L3; -:-:, disc L1=L2.
shape of the power spectral density functions was very dierent (Fig. 7). The most signi®cant dierence concerned the frequency content. The mean square value of the data of both parameters showed a quite similar distribution until 10 Hz. Above this frequency, the frequency response of the intranuclear pressure tended to zero, while it remained rich in all the exposure bandwidth for the vertical vertebral acceleration, i.e. until 25 Hz.
4. Discussion Although unembalmed human cadavers remain the best ®tted physiological and anatomical model of the living volunteers, still they also present major limitations which should not be neglected when examining the experimental results. These have been largely addressed previously [31]. They mainly concern the absence of myoelectric activity, the body temperature, the conser-
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
S131
Fig. 3. Mean value (for seven subjects) of the power spectral density function of the intranuclear pressure variation at each intervertebral disc L1=L2; L2=L3; L3=L4; L4=L5 for dierent postures on a rigid seat (1 bar 0.1 MPa): ±±± , seated erect without a lumbar support; ...., seated erect with a lumbar support; - - - -, seated as in a car without a lumbar support; -:-:, seated as in a car with a lumbar support.
Fig. 4. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band in all tested postures and at all discal levels (1 bar 0.1 MPa) ± Eect of the subject.
vation state, and the eect of removing of the abdominal viscera. Quandieu [32] showed that increased muscle contraction improved the low-pass ®lter characteristics of the intervertebral disc, while provoked muscle relaxation shifted the ®lter characteristics towards higher frequencies. Guillon [30] showed similar intranuclear pressure variation in cadavers compared to in vivo
Fig. 5. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band, in all tested postures for seven subjects (1 bar 0.1 MPa) ± Eect of the discal level.
measures by Nachemson and Morris [19] and by Andersson et al. [20±22] in static seated postures. However, the pressure values were lower than those obtained in vivo, which was attributed to the absence of muscle contraction as they had the same magnitude as the in vivo measures realized in para- and tetraplegic patients by Hein-Sorensen et al. [23]. The eect of a lower body temperature in cadavers (about 10°C) was not controlled. It can be assumed that this may result in a stier
S132
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Fig. 6. Box-Whiskers plots of the energy of the intranuclear pressure variation in the 0±25 Hz band at all discal levels for seven subjects (1 bar 0.1 MPa) ± Eect of posture: Posture 1, seated erect without a lumbar support; Posture 2, seated erect with a lumbar support; Posture 3, seated as in a car without a lumbar support; Posture 4, seated as in a car with a lumbar support.
Table 2 Number of times the energy of the power spectral density function of the intranuclear pressure variation in the 0±25 Hz band increases or decreases when changing posture by adding a lumbar support for seven subjects, at each discal level Seated erect Seated as in a car Without ! with a lumbar Without ! with a lumbar support support
L1=L2 L2=L3 L3=L4 L4=L5
Number of increase
Number of decrease
Number of increase
Number of decrease
6 4 1 1
1 3 6 6
2 3 0 1
5 4 7 6
behaviour than in vivo. The disc desiccation is another uncontrolled factor through the experimentation. But disc dehydration is also observed in daily activities and is revealed by the shortening observed between the morning and the night [33]. Also, postures were tested in a semi-randomized order, so no systematic eect could be attributed to desiccation. The absence of the abdominal viscera, even though modifying the total equilibrium of the trunk, did not have probably a major eect on the intranuclear pressure measurements, as the intra-abdominal pressure is mostly related to muscle contraction and the main eect would be due to the absence of muscle tone. In vivo studies showed that axial motion occurred in the spine when submitted to vertical shocks or wholebody vibration in seated postures [6]. But generally, measures were realized at only one level [7], in the best
Fig. 7. Mean value (for seven subjects) of the power spectral density of the vertical vertebral acceleration at each lumbar vertebra (black continuous lines) and of the intranuclear pressure variation at each lumbar disc level (grey lines): ±±± , L1=L2; ...., L2=L3; - - - -, L3=L4; -:-:, L4/L5 (1 bar 0.1 MPa).
cases at two locations of the lumbar spine [6]. Still, they suggested a great rigidity in the lumbar spine. In a previous study, the authors also reported a rigid behaviour of the lumbar spine, by examining the vertical seat-to-vertebrae transfer function [10,31]. One original aspect of this experiment was the monitoring of the intranuclear pressure simultaneously to the vertebral
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
acceleration, at all lumbar levels (®ve vertebrae and four discs) in seven subjects and four seated postures. As the vertebral column is constituted alternatively of two very dierent components, the vertebrae and the intervertebral discs, it was interesting to look at the response of each of the structures separately, which may allow a direct assessment of the risk. In the present study, the authors compared measurements at the nucleus pulposus to those at the vertebral body, in order to understand what was their contribution to vibration transmission through the lumbar spine, and how it was modi®ed by posture. The results showed behaviour dierent from the nucleus pulposus and the vertebral body. In the ®rst case, frequency response was concentrated at a single frequency ranging from 5.9 to 8.1 Hz in the 1±15 Hz bandwidth, while in the second case, frequency response was distributed in all the exposure bandwidth (1±25 Hz) with a peak at the same dominant frequency as the intranuclear pressure variation. Therefore, responses of both structures suggested the existence of a cyclical phenomenon at the same frequency as for a single-degree-of-freedom system. While accelerations measured at the vertebral body may give an indication on the total behaviour of the intervertebral disc, pressures measured in the nucleus pulposus re¯ected only the behaviour at the centre of the disc. Hence, dierences observed in the power spectral density function of the vertebral acceleration and the intranuclear pressure variation, re¯ecting a more stochastic phenomenon, may arise from the annulus ®brosus or be attributed to the facet joints [34,35]. The power spectral density function of the intranuclear pressure variation as well as of the vertebral acceleration peaked at the same dominant frequency at all levels, for each posture and each subject. On average, resonance frequency at the vertebrae (vertical seat-tovertebrae transmissibility) was at 6.3 Hz in the seated erect posture without lumbar support [10]. It was slightly higher than the published in vivo results by Panjabi et al. [6] (4.4 Hz), Pope et al. [7] (5 Hz), Broman et al. [36] (5±6.3 Hz) and Magnusson et al. [37] (5.8 Hz). These dierences were partly explained by variations between the experimental methods [31]. To our knowledge, there are no published data concerning intranuclear pressure variations when submitted to whole-body vibration in dierent seated postures. Andersson et al. [20±22] as well as Guillon [30] showed, in static, that the intranuclear pressure at the L3=L4 disc level decreased when leaning backwards and when using a lumbar support. Guillon [30] showed that the maximum intranuclear pressure variation at the L3=L4 intervertebral disc was also proportional to the kinetic energy for impacts on the head from dierent heights. In the present study, intranuclear pressure was measured simultaneously at all lumbar disc levels in four seated postures, for a random vibration. The energy in the 0±25
S133
Hz range was then estimated and this parameter was used to study the eect of measurement level or posture. As for the static measurements [20±22,30], the energy of the intranuclear pressure variation decreased when moving from a seated erect posture to a seated posture as in a car, at all lumbar disc levels. However, lumbar support did not have the same eect whether considering the upper discs (above L3) or the lower discs (below L3), when seated erect or seated as in a car. The energy tended to increase when using a lumbar support at the L1=L2 and L2=L3 intervertebral discs, while it always decreased at the L3=L4 and L4=L5 intervertebral discs. Dierent tendencies were observed on the vertical seatto-vertebrae transmissibility [10,31]: the transmissibility peak increased when moving from a seated erect posture to a seated as in a car posture. It decreased when using a lumbar support in the seated erect posture and increased when using a lumbar support in the seated as a car posture. It should be noticed that the experimental lumbar support used in this study (located opposite to L3, about 4 cm ahead of the backrest) intended to realize a real hyper-lordosis and would be unbearable in a reallife driving posture. The intranuclear pressure variation energy depended on the vertebral level, contrary to the vertical vertebral acceleration. However, these variations should be considered with some caution, as they should be related to the intervertebral disc degree of degeneration and to the surface of the intervertebral disc (with that of the nucleus pulposus specially). 5. Conclusions These results suggest that the health risk at the lumbar spine from whole-body vibration cannot be assessed only by examining the acceleration measurements, even if they were monitored at the vertebral level, and that it might be hazardous to conclude from observations at one level. Indeed, results from vertical seat-to-vertebrae transfer functions implied a sti connection between the lumbar vertebrae. Those from the intranuclear pressure measurements showed an eect of the discal level as well as of posture. Although, changing posture did not aect the same way at all discal levels. While the transmissibility peaks of vertical seatto-vertebrae were increased when leaning backwards, the energy of the intranuclear pressure variation decreased as for static postures. The lumbar support increased the energy at the upper disc levels (above L3) when seated erect, while the energy always decreased at the lower discs (below L3). The shape of the power spectral density function suggested the existence of a cyclic loading of the nucleus pulposus, while more complex phenomena were observed at the vertebral level.
S134
A. El-Khatib, F. Guillon / Clinical Biomechanics 16 Suppl. No. 1 (2001) S127±S134
Acknowledgements This work could not have been completed without the participation of the Corpses Donation to medicine. In memory of the donors. References [1] Burdorf A, Sorock G. Positive and negative evidence of risk factors for back disorders. Scand J Rehabil Med 1997;23:243±56. [2] Bovenzi M, Hulshof CTJ. An updated review of epidemiological studies on the relationship between exposure to whole-body vibration and low back pain. J Sound Vib 1998;215:595±612. [3] Messenger AJ, Grin MJ. Eects of anthropometric and postural variables on the transmission of whole-body vertical vibration from seat-to-head. Report. Southampton (UK): Institute of Sound and Vibration, Human Factors Research Unit; 1989 Sept. Report No. 172. [4] Grin MJ. Handbook of human vibration. London: Academic Press; 1990. [5] Paddan GS, Grin MJ. Individual variability in the transmission of vertical vibration from seat to head. Report. Southampton (UK): Institute of Sound and Vibration, Human Factors Research Unit; 1994 Nov. Report No. 236. [6] Panjabi MM, Andersson GBJ, Jorneus L, Hult E, Mattsson L. In vivo measurements of spinal column vibrations. J Bone Joint Surg 1986;68-A:695±702. [7] Pope MH, Broman H, Hansson T. The dynamic response of a subject seated on various cushions. Ergonomic 1989;32:1155±66. [8] Hinz B, Seidel H, Brauer D, Menzel G, Bluthner R, Erdmann U. Bidimensional accelerations of lumbar vertebrae and estimation of internal spinal load during sinusoidal vertical whole-body vibration: a pilot study. Clin Biomech 1988;3:241±8. [9] Matsumoto Y, Grin MJ. Movement of the upper-body of seated subjects exposed to vertical whole-body vibration at the principal resonance frequency. J Sound Vib 1998;215:743±62. [10] El-Khatib A, Guillon F, Domont A. Vertical vibration transmission through the lumbar spine of the seated subject ± ®rst results. J Sound Vib 1998;215:763±74. [11] Kelsey JL, White III AA. Epidemiology and impact of low-back pain. Spine 1980;5:133±42. [12] Guillon F, El-Khatib A. Deux nouveaux tableaux de maladies professionnelles concernant les aections douloureuses chroniques du rachis. Revue de Medecine Orthopedique 1999;58:17± 22. [13] Nachemson A. Lumbar intradiscal pressure. Experimental studies on post-mortem material. Acta Orthop Scand 1960;43 Suppl. [14] McNally DS, Adams MA. Internal intervertebral disc mechanics as revealed by stress pro®lometry. Spine 1992;17:66±73. [15] McNally DS, Arridge RGC. An analytical model of intervertebral disc mechanics. J Biomech 1995;28:53±68. [16] Ranu HS, King AI. Correlation of intradiscal pressure with vertebral endplate pressure. Ind Mech Eng 1980;C125/80:9. [17] Merriam WF, Quinnell RC, Stockdale HR, Willis DS. The eect of postural changes on the inferred pressures within the nucleus pulposus during lumbar discography. Spine 1984;9:405±8. [18] Ramos G, Martin W. Eects of vertebral axial decompression on intradiscal pressure. J Neurosurg 1994;81:350±3. [19] Nachemson A, Morris JM. In vivo measurements of intradiscal pressure. Discometry, a method for the determination of pressure in the lower lumbar discs. J Bone Joint Surg 1964;46-A:1077±91.
[20] Andersson GBJ, Ortengren R, Nachemson A, Elfstrom G. Lumbar disc pressure and myoelectric back muscle activity during sitting. I. Studies on an experimental chair. Scand J Rehabil Med 1974;6:104±14. [21] Andersson GBJ, Ortengren R. Lumbar disc pressure and myoelectric back muscle activity during sitting. III. Studies on a wheelchair. Scand J Rehabil Med 1974;6:122±7. [22] Andersson GBJ, Ortengren R, Nachemson A, Elfstrom G. Lumbar disc pressure and myoelectric back muscle activity during sitting. IV. Studies on a car driver's seat. Scand J Rehabil Med 1974;6:128±33. [23] Hein-Sorensen O, Elfstrom G, Nachemson A. Disc pressure measurements in para- and tetraplegic patients. Scand J Rehabil Med 1979;11:1±11. [24] Schultz A, Andersson G, Ortengren R, Haderspeck K, Nachemson A. Loads on the lumbar spine. Validation of a biomechanical analysis by measurements of intradiscal pressures and myoelectric signals. J Bone Joint Surg 1982;64-A:713±20. [25] Wilke HJ, Neef P, Caimi M, Hoogland T, Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755±62. [26] McNally DS, Adams MA, Goodship AE. Development and validation of a new transducer for intradiscal pressure measurement. J Biomed Eng 1992;14:495±8. [27] El-Khatib A, Piriou P, Guillon F, Somenzi G, Tarriere C. Study of the hydrostatic behaviour of the lumbar nucleus pulposus in dynamic by the measurement of the intranuclear pressures in cadavers and in vitro. In: Proceedings of the XIVth Congress of the International Society of Biomechanics, 1993 July 4±8; Paris, France. [28] Abe E, Nickel T, Buttermann GR, Lewis JL, Transfeldt EE. Lumbar intradiscal pressure after posterolateral fusion and pedicle screw ®xation. Tohoku J Exp Med 1998;186:243±53. [29] Guillon F, Pignolet F, Somenzi G, Tarriere C, Proteau J. Eets d'un choc mou vertical sur la pression intradiscale en position assise: premiers resultats sur 4 cadavres. J de Biophysique et de Biomecanique 1986;10(Suppl 1):7±8. [30] Guillon F. Inter^et de la mesure de la pression intranucleaire pour l'etude du comportement dynamique du disque intervertebral humain. Ph.D. Dissertation, Paris XI University, Orsay (F), 1992. [31] El-Khatib A. Contribution a l'etude de la transmission des vibrations dans le rachis lombaire du sujet assis. Ph.D. Dissertation, Paris XI University, Orsay (F), 1999. [32] Quandieu P. Etude, en regime vibratoire, des fonctions de transfert des disques intervertebraux lombaires et de la masse apparente d'un primate. Ph.D. Dissertation. ReneDescartes University, Paris (F), 1981. [33] Magnusson M. Eects of seated whole-body vibrations on the spine. An experimental study in man. Ph.D. Dissertation, G oteborg University, G oteborg (Sweden), 1991. [34] Prasad P, King AI, Ewing CL. The r^ ole of articular facets during +Gz acceleration. J Appl Mech, Trans ASME 1974;41:321±6. [35] Sharma M, Langrana NA, Rodriguez J. Modelling of facet articulation as a nonlinear moving contact problem: sensitivity study on lumbar facet response. J Biomech Eng, Trans ASME 1998;120:118±25. [36] Broman H, Pope MH, Benda M, Svensson M, Ottosson C, Hansson T. The impact response of the seated subject. J Orthop Res 1991;9:150±4. [37] Magnusson M, Pope M, Rostedt M, Hansson T. Eect of backrest inclination on the transmission of vertical vibrations through the lumbar spine. Clin Biomech 1993;8:5±12.