UBC Faculty Research and Publications

Are cervical multifidus muscles active during whiplash and startle? An initial experimental study Siegmund, Gunter P; Blouin, Jean-Sébastien; Carpenter, Mark G; Brault, John R; Inglis, JTimothy Jun 5, 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12891_2007_Article_486.pdf [ 1.93MB ]
JSON: 52383-1.0223414.json
JSON-LD: 52383-1.0223414-ld.json
RDF/XML (Pretty): 52383-1.0223414-rdf.xml
RDF/JSON: 52383-1.0223414-rdf.json
Turtle: 52383-1.0223414-turtle.txt
N-Triples: 52383-1.0223414-rdf-ntriples.txt
Original Record: 52383-1.0223414-source.json
Full Text

Full Text

ralssBioMed CentBMC Musculoskeletal DisordersOpen AcceResearch articleAre cervical multifidus muscles active during whiplash and startle? An initial experimental studyGunter P Siegmund*1,2, Jean-Sébastien Blouin2,3, Mark G Carpenter2,3, John R Brault4 and J Timothy Inglis2,3Address: 1MEA Forensic Engineers & Scientists, Richmond, BC, Canada, 2School of Human Kinetics, University of British Columbia, Vancouver, BC, Canada, 3ICORD, University of British Columbia, Vancouver, BC, Canada and 4MEA Forensic Engineers & Scientists, Lake Forest, CA, USAEmail: Gunter P Siegmund* - gunter.siegmund@meaforensic.com; Jean-Sébastien Blouin - jsblouin@interchange.ubc.ca; Mark G Carpenter - mark.carpenter@ubc.ca; John R Brault - john.brault@meaforensic.com; J Timothy Inglis - tinglis@unixg.ubc.ca* Corresponding author    AbstractBackground: The cervical multifidus muscles insert onto the lower cervical facet capsularligaments and the cervical facet joints are the source of pain in some chronic whiplash patients.Reflex activation of the multifidus muscle during a whiplash exposure could potentially contributeto injuring the facet capsular ligament. Our goal was to determine the onset latency and activationamplitude of the cervical multifidus muscles to a simulated rear-end collision and a loud acousticstimuli.Methods: Wire electromyographic (EMG) electrodes were inserted unilaterally into the cervicalmultifidus muscles of 9 subjects (6M, 3F) at the C4 and C6 levels. Seated subjects were thenexposed to a forward acceleration (peak acceleration 1.55 g, speed change 1.8 km/h) and a loudacoustic tone (124 dB, 40 ms, 1 kHz).Results: Aside from one female, all subjects exhibited multifidus activity after both stimuli (8subjects at C4, 6 subjects at C6). Neither onset latencies nor EMG amplitude varied with stimulustype or spine level (p > 0.13). Onset latencies and amplitudes varied widely, with EMG activityappearing within 160 ms of stimulus onset (for at least one of the two stimuli) in 7 subjects.Conclusion: These data indicate that the multifidus muscles of some individuals are active earlyenough to potentially increase the collision-induced loading of the facet capsular ligaments.BackgroundThe cervical facet joints are a source of neck pain in abouthalf of chronic whiplash patients [1]. In addition to guid-ing better diagnostic and treatment techniques [2,3], thisfinding provides an anatomical focus to biomechanicalstudies aimed at understanding the aetiology of whiplashanism [4], but more attention has been devoted towardexcess strain of the capsular ligament itself [5-8]. Injuriouslevels of strain have been observed in some capsular liga-ments when loads simulating a rear-end collision wereapplied in-vitro [7,8]. More recently, allodynia – meas-ured as paw withdrawals in a rat model – has been corre-Published: 5 June 2008BMC Musculoskeletal Disorders 2008, 9:80 doi:10.1186/1471-2474-9-80Received: 28 November 2007Accepted: 5 June 2008This article is available from: http://www.biomedcentral.com/1471-2474/9/80© 2008 Siegmund et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 9(page number not for citation purposes)injuries. Pinching of the posterior synovial fold of the cer-vical facet capsular ligament is one possible injury mech-lated to levels of capsular ligament strain relevant towhiplash injury [9], and Group III and IV afferents fromBMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80the facet joint capsule have demonstrated a gradedresponse to mechanical loading in an in-vivo goat model[10].Anatomically, the cervical facet capsule contains fine,unmyelinated nerves that likely have nociceptive function[11]. Distending these ligaments by injection of contrastmedia has produced whiplash-like pain patterns in nor-mal individuals [12]. Tendons of the cervical multifidusmuscles insert directly onto the capsular ligaments[13,14] and it has been postulated that multifidus activa-tion during the neuromuscular response to a rear-endautomobile impact could increase the strain in the capsu-lar ligaments above that imposed passively by the impact-induced head and neck dynamics [7,13]. Prior work hasshown early multifidus activation during a whiplashresponse in one of three subjects [15], however, it remainsunclear whether this reflex response will be present in alarger group of subjects.The neuromuscular response to a whiplash exposure con-tains both a postural and a startle response [16,17]. Thiscombined postural/startle response was observed in sur-face electromyograms of the sternocleidomastoid and cer-vical paraspinal muscles with and without the loud soundof a vehicle crash [16], although muscle activity was largerwhen the acceleration was accompanied by a loud sound[18]. These prior findings suggest that a startle response –presumably evoked by the crash motion, the noise, orsome combination of the two – amplifies the superficialneck muscle response during an unexpected rear-end col-lision. If either the crash-induced motion or a loud noisealso evokes a reflex response in the deep multifidus mus-cles, then this muscle contraction could add to the loadsborne by the facet capsule during a rear-end automobilecollision and possibly affect the capsular ligament's injurypotential.Based on this line of reasoning, our broad goal is to estab-lish whether reflex activation of the cervical multifidusmuscle contributes to straining the facet capsular ligamentand thereby contributes to the genesis of whiplash injury.In this study, we examine one step in this broader goaland specifically address whether a reflex contraction of themultifidus muscle is evoked by either the posturalresponse (a seated horizontal acceleration without thenoise of impact) or a startle response (a loud noise with-out a postural perturbation). If either of these stimuli gen-erates a reflex response in multifidus, our broader goalwarrants continued investigation; if neither stimulus gen-erates a reflex response, then we can conclude that multi-fidus likely does not play a role in facet capsule injury inwhiplash.MethodsSubjects and consentNine subjects (6M, 3F) participated in the experiment.Male subjects were 30 ± 6 years old, 177 ± 6 cm tall andweighed 79 ± 5 kg; female subjects were 30 ± 1 years old,166 ± 5 cm tall and weighed 68 ± 6 kg. None of the sub-jects had a history of whiplash injury, medical conditionsthat impaired sensory or motor function, or prolongedneck or back pain during the preceding 2 years. Subjectsdid not ingest caffeine or nicotine for two hours before theexperiment. All subjects gave written inform consent andthe experiment was approved by the UBC ClinicalResearch Ethics Board and conformed to the Declarationof Helsinki.InstrumentationEMG activity of the left multifidus muscles was measuredusing twisted pairs of insulated 0.05 mm wire (Stablohm800A, California Fine Wire, Grover Beach, CA) with 1–2mm of wire exposed at each recording tip. The recordingtip of each wire was hooked to anchor it in the muscle tis-sue. After first identifying multifidus and any major ves-sels on each subject's magnetic resonance (MR) scan(Phillips Gyroscan Intera 3.0T), wires were inserted intothe multifidus muscles at the C4 and C6 levels using 25gauge needles under ultrasound guidance (Sonos 5500,Agilent Technologies, Andover, MA) (Figure 1). Ultra-sound was again used during wire extraction, andalthough the wires were harder to visualize during extrac-tion (without the needle present), tissue displacement atthe hooked end was readily apparent during wire extrac-tion. The recording ends of all wires were in the same loca-tion before and after the experiment, which meant thatthe wire has not moved substantively relative to the sur-rounding muscle tissue during the exposures. EMG signalswere amplified and band-pass filtered (30–1000 Hz)using a Neurolog system (Digitimer, Welwyn GardenCity, Hertfordshire, England). Head acceleration wasmeasured with a nine accelerometer array (Kistler8302B20S1, Amherst, NY) arranged in a 3-2-2-2 configu-ration [19] and sled acceleration was measured with auniaxial accelerometer (Sensotec JTF3629-05, Columbus,OH). Head and torso displacements were measured witha motion analysis system (Pheonix VZ4000, Burnaby,BC). Transducer signals were low-pass filtered (1000 Hz)and, together with the EMG signals, simultaneously sam-pled at 2 kHz. Displacement data were acquired at 100 Hzper marker.Test proceduresSubjects first performed 5-second isometric maximal vol-untary contractions (MVCs) from the neutral position ineight directions (flexion, extension, left and right lateralPage 2 of 9(page number not for citation purposes)flexion, and the 45° points between these 4 primary direc-tions) to provide normalizing data for the dynamic EMGBMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80recordings. For the MVCs, a subject's head was firmlyclamped to an inverted force plate (Bertec 4060H, Wor-thington, OH) and their torso firmly strapped to a rigidseat back (not the same seat used for the exposuresdescribed below) (Figure 2a). For their exposure to theacceleration and acoustic stimuli, subjects were seated inan automobile seat (1991 Honda Accord front passengertwo linear induction motors (IC55-100A7; Kollmorgen,Kommack, NY) mounted through linear bearings to 6 mhorizontal rails rigidly fastened to a concrete base. Sledmotion was programmed through a 500 Hz position-feedback controller and was thus insensitive to subjectmass. Subjects were instructed to adopt a comfortableseated posture, face forward, rest their forearms on theirlap, and relax their face and neck muscles. Seated subjectswere first exposed to a single unexpected loud acousticstimulus (124dB, 1000 Hz, 40 ms duration) capable ofevoking a startle reflex [20]. After a rest period of at least 3minutes, subjects experienced a single forward horizontalacceleration pulse (apeak = 1.55 ± 0.02 g; tpeak = 16 ± 1 ms;Δt = 59 ± 1 ms; Δv = 0.50 ± 0.01 m/s, noise<82 db, Figure2c). After the pulse, the sled traveled at 0.50 m/s for 500ms (beyond our period of interest) before deceleratinglinearly to rest at 0.05 g. A single exposure to each stimu-lus was used because multiple exposures could be con-founded by the rapid habituation observed to loudacoustic stimuli [21,22] and seated perturbations [23,24].Data reductionAll EMG data were high-pass filtered (50 Hz) to removemotion artifact present in some tests. The onset of EMGactivity was determined using a log-likelihood-ratio algo-rithm [25,26] and then confirmed visually. For both theacceleration and acoustic stimuli, the root-mean-squared(RMS) amplitude of the EMG signals was computed usinga moving 20 ms window. These RMS values were thennormalized by the maximum RMS EMG (also using a 20ms window) observed during the MVC contractions.For the acceleration and acoustic stimuli, three kinematicparameters were calculated: i) the horizontal accelerationof the head's centre of mass in the lab frame (ax), ii) thehorizontal displacement (retraction, rx) of the atlanto-occipital (AO) joint with respect to the centre of the T1vertebral body, and iii) the extension angle (θy) of thehead in the lab frame. The center of the T1 vertebral bodyrelative to the manubrium and C7 spinous process wasdetermined from each subject's pre-test MR scan. All time-varying kinematic signals were set to zero at the start of thestimulus.Statistical analysisAll dependent variables were first tested for normalityusing a Shapiro-Wilks test. Differences between the stim-uli (acceleration/acoustic) and spine level (C4/C6) weretested using a two-way repeated-measures ANOVA andTukey post-hoc tests for normally distributed variables.For variables not normally distributed, a non-parametricFriedman ANOVA and post-hoc Wilcoxon matched pairstests were used. All tests were performed using StatisticaMRI and ultrasound images showing wire insertionFigure 1MRI and ultrasound images showing wire insertion. a) axial slice of a magnetic resonance scan at C4 level showing desired insertion path to the multifidus muscle and b) ultra-sound image showing actual needle path and tip during inser-tion into the multifidus muscle.Multifidus insertionPage 3 of 9(page number not for citation purposes)with the head restraint removed) mounted to a feedback-controlled linear sled (Figure 2b). The sled consisted of(v.6.1, Statsoft, Tulsa, OK) with a significance level set atp < 0.05.BMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80ResultsEight of the nine subjects exhibited multifidus activityafter both stimuli, as demonstrated by the exemplar dataof two subjects in Figure 3. The remaining subject exhib-ited no multifidus activity to either stimulus – despiteclear multifidus activity during her MVC – and wasexcluded from the statistical analysis of the EMG data.Multifidus activity was present in all eight subjects at theC4 level and in six subjects at the C6 level. Supramaximalmuscle activity was observed either at one level or to oneof the two stimuli in six subjects.There were no differences in the muscle response ampli-tude between the acceleration and acoustic stimuli orbetween the two spine levels (p = 0.32) (Table 1). Therewere also no differences in the onset latencies betweeneither stimulus types or recording levels (p = 0.41). Therewas however considerable variation in the onset latenciesbetween subjects, with multifidus onset occurring within160 ms of stimulus onset in five subjects during the sledacceleration and six subjects (four of which were com-mon) during the acoustic stimulus.All subjects exhibited a well-defined forward head acceler-ation, head extension and retraction of the AO joint rela-tive to T1 when exposed to the sled acceleration (Figure3). The peak kinematic responses for the acoustic stimuluswere smaller (p < 0.02; Table 1) and also varied in direc-tion, with the initial head acceleration forward in foursubjects and rearward in the other subjects. Kinematicpeaks occurred earlier and were less variable for the sledacceleration compared to the acoustic stimulus (p < 0.03;Table 1).DiscussionBased on the data presented here, the cervical multifidusmuscles of some individuals are active during either pos-tural or startle responses. Of our nine subjects, eight sub-jects exhibited multifidus activity following the stimuliand seven subjects responded within 160 ms to at leastone of the two stimuli. As a result, this study establishesthat multifidus could play a role in straining the capsularligament of a large segment of the population during arear-end collision. Of course the single subject in the cur-rent study who responded to neither stimulus may repre-sent a segment of the population whose multifidusmuscles do not react to these stimuli, and this subjectshould not be dismissed outright as atypical given the lownumber of subjects we tested. Despite this caveat, thesefindings lend support to the proposition that multifidusactivation may play a role in the genesis of some whiplashinjuries, although more work is needed before we canconclude whether it actually contributes to injuring theExperimental setupFigu e 2Experimental setup. Schematic and photograph showing a) the head clamp and force plate used for the isometric con-tractions, b) the sled configuration, and c) the sled accelera-tion pulses. Three superimposed pulses are shown to illustrate the repeatability of the pulse. The dashed line shows a vehicle-to-vehicle collision pulse with a speed change of 8 km/h recorded during earlier experiments [19].Head accelerometryPage 4 of 9(page number not for citation purposes)facet capsule ligament.BMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80To be relevant to whiplash injury, the multifidus musclemust activate both early and forcefully during its responseto a collision. Peak capsular ligament strain occurs about200 ms after the onset of T1 acceleration [8], and T1 accel-eration begins about 25 to 35 ms after vehicle impact [19].Assuming meaningful levels of muscle force are generated75 to 100 ms following activation [27,28], multifidus acti-vation must occur within 125 to 160 ms after impact (i.e.,sled acceleration onset) to coincide with the collision-induced peak in capsular ligament strain. Of our nine sub-jects, multifidus was active within 125 ms in 4 subjectsfollowing the sled acceleration (see example in upperpanel of Figure 3a) and 6 subjects following the startlingtone (see example in lower panel of Figure 3b). In oneadditional subject, multifidus was active within 160 ms ofthe sled acceleration onset. Thus, the multifidus musclesresponded to at least one stimulus sufficiently early topotentially contribute to a facet capsular injury in slightlymore than half our subjects. For individuals startled by anunexpected rear-end impact, it may not matter whethermultifidus activation was mediated by a startle or posturalresponse: activation by either stimulus before peak retrac-tion could contribute to increasing facet capsule strain.For individuals whose multifidus muscles do not respondto either stimulus, or for individuals who have bothdelayed multifidus activation to a sudden accelerationand no startle response, the multifidus muscles may notaffect capsule strain and therefore play no role in generat-ing a whiplash injury.The force generated by the multifidus muscle in our sub-jects was not measured, but about half our subjectsexceeded their MVC level during either the acceleration oracoustic stimuli. Supramaximal muscle activity has beenobserved by others studying whiplash [29], but it remainsunclear why or how this occurs. Subjects may not haveexerted maximal efforts during their MVCs; however, theneck moments measured in our male subjects (flexion 12± 6 Nm; extension 20 ± 8 Nm) are similar to maximalneck moments (flexion 13 ± 3 Nm; extension 24 ± 7 Nm)measured by others [30]. It is also possible that multifidusmay not recruit fully during contractions designed to max-imize the horizontal force at the forehead or may berecruited differently in the seats used for the MVC and theacceleration/acoustic exposures. Alternatively, short,reflex muscle activations may generate more synchronousbursts of action potentials than do 5 second maximal vol-untary efforts, and summation of these synchronousaction potentials yields a greater EMG signal during tran-sient reflex activations. Further work is needed to deter-mine why this supramaximal activity occurs and how toestimate multifidus muscle force from these data.mated from data in the literature. The area of multifidusinsertion onto the facet capsule varies between 9 and 96mm2 (average of 48 ± 22 mm2) [6], which for an isometrictetanic stress of 0.44 MPa [31], corresponds to an appliedforce of 9 to 42 N (average 21 ± 10 N). During elongation,force increases by a factor of 1.2 to 1.6 depending onlengthening velocity [32] and therefore the maximumforce that could be applied to the capsule is between 11and 67 N (average 29 ± 14 N; a factor of 1.4 is assumed forthis average and SD). Prior estimates of the peak multi-fidus force (42 to 55 N) using mathematical models of themale neck are greater than this average [33,34], butslightly less than the maximum value computed here.The loads applied by a maximally-active multifidus mus-cle to the facet capsule are potentially a large proportion(about 64% on average) of the quasi-static loads (21 to 93N; 45 ± 21 N) required to cause sub-catastrophic failuresin the capsular ligament [7]. Sub-catastrophic failureloads under dynamic conditions have not been reported,but increasing catastrophic failure loads have beenobserved with increasing elongation rates (Table 2). Peakligament elongation rates of about 50 mm/s have beenestimated for whiplash loading with an average T1 accel-Table 1: Mean ± SD and median (range) of muscle onset times, normalized muscle amplitudes, kinematic amplitudes and time of peak kinematics.Sled StartleEMG onset time (ms)C4 level 129 ± 60 123 ± 64140 (34 – 191) 105 (58 – 222)C6 level 232 ± 175 182 ± 120194 (66 – 571) 126 (60 – 417)EMG amplitude (%MVC)C4 level 137 ± 168 a 136 ± 123 a107 (12 – 532) 80 (39 – 337)C6 level 40 ± 25 57 ± 53 a36 (9 – 77) 28 (15 – 150)KinematicsHead accel (m/s2) 9.7 ± 3.3 1.9 ± 1.010.1 (5.6 – 15.2) 1.6 (0.5 – 3.5)Head extension (°) 15.6 ± 5.6 1.9 ± 3.1 a12.6 (10.0 – 25.9) 0.5 (0.1 – 8.8)Retraction (mm) 18.4 ± 6.9 4.3 ± 2.4 a19.4 (8.7 – 31.3) 5.2 (1.0 – 6.6)Time to peak kinematics (ms)Head accel 123 ± 11 272 ± 168 a122 (106 – 145) 215 (124 – 581)Head extension 208 ± 51 a 557 ± 200190 (170 – 330) 510 (290 – 770)Retraction 173 ± 48 373 ± 187160 (110 – 240) 330 (130 – 710)a non-normally distributed, Shapiro-Wilks p < 0.05Page 5 of 9(page number not for citation purposes)Although we did not measure muscle force, the maximumforce that the multifidus muscle can generated can be esti-eration of 2.3 g over 100 ms [see Figure 5 in reference[35]]. This level of T1 acceleration is consistent with aBMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80Page 6 of 9(page number not for citation purposes)Exemplar dataFigure 3Exemplar data. Exemplar data for two subjects showing the raw electromyographic activity and kinematics parameters observed during a) the sled perturbation and b) the loud acoustic tone. The vertical scale bars are aligned with time t = 0 ms. EMG are presented in arbitrary units that are the same for both stimuli.enot citsuocA )b)a Sled accelerationC4 MultC6 Mult10 m/s² Head accel10º Head angle20 mm Retraction1 g enoTlecca delS-100 0 100 200 300 400 500 -100 0 100 200 300 400 500C4 MultC6 Mult10 m/s² Head accel10º Head angle20 mm Retraction1 g enoTlecca delS)sm( emiT         )sm( emiT-100 0 100 200 300 400 500 -100 0 100 200 300 400 500BMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80vehicle speed change of about 8 km/h [19], and higherstrain rates likely occur in more severe collisions. If weassume that catastrophic and sub-catastrophic failureloads scale similarly, then the data in Table 2 suggest thatsub-catastrophic failure loads will be up to 1.5 timeshigher under dynamic conditions with elongation rates ofup to 100 mm/s than under quasi-static conditions. Thisassumption suggests the dynamic loads required to causesub-catastrophic failures in the capsular ligament wouldbe in the region of 68 ± 32 N at elongation rates of 100mm/s. Thus even under dynamic conditions, the loadsapplied by a maximally-active multifidus muscle to thefacet capsular ligament are potentially large (43% on aver-age) compared to the capsule's estimated sub-catastrophicfailure loads. During severe collisions inducing capsuleelongation rates greater than 100 mm/s, the sub-cata-strophic failure loads will be higher and the proportionalcontribution of multifidus activation to capsule failurewill be lower. Joints with large multifidus/capsular inser-tion areas, low partial-rupture loads and high levels ofmultifidus activation during the collision might be partic-ularly prone to injury. Unfortunately, the large within-individual variations in insertion area [6] and partial-rup-ture load [7], combined with the large between-subjectsvariability in multifidus activation timing and amplitudeobserved here, make it difficult to predict who is a risk forthis type of injury.The relatively large force that multifidus can applythrough the capsular ligament combined with the largeactivation levels occasionally observed during a pure star-tle raises the question of why facet capsule injuries do notoccur – at least in some individuals – when startled. Wedo not know the answer to this question, but it may bethat the capsular ligaments and multifidus muscles of aparticular individual co-develop and are well matched fornormal neck movements and startle responses in normalor near-normal postures. Whiplash-induced neck motionmay pre-strain the capsular ligament and make it morevulnerable to multifidus-induced loading – or vice versa.Further work is needed to explore this question.The high percentage of Type I muscle fibres [36], the smallnumber of vertebra spanned [14] and the low momentgenerating capacity [14] of the cervical multifidus musclesmay suggest this muscle has a primarily postural role.Recent evidence, however, suggests the neck multifidusmuscles exhibit phasic activity during isometric contrac-tions and voluntary head movements in addition toreceiving neural signals common to the superficial neckmuscles [15,16]. Thus posture-related activity – and per-haps more importantly movement-related activity – in themultifidus muscle could conceivably provoke or exacer-bate pain in an injured capsular ligament. Moreover, thehigh number of muscle spindles in multifidus [37] sug-gests an important role for multifidus in directly sensingspine joint position and perhaps contributing to headposition sense. If pain from an injured capsular ligamentinhibits or alters tonic or phasic multifidus activity, thenaltered somatosensory information from muscle spindlescould also contribute to the dizziness experienced bymany whiplash patients [38]. Indeed, increased jointposition error in whiplash patients has been previouslyattributed to mechanoreceptor dysfunction in the cervicalspine [39,40], and a dysfunctional interaction of the cap-sular ligament and the multifidus muscle could explainthese symptoms.For this experiment, we used a sled acceleration that wasless severe and an acoustic stimulus that was more intensethan those experienced by many whiplash-injuredpatients. Although our acceleration pulse yielded a lowoverall speed change, its leading edge matched the onsetof a collision with a speed change of 8 km/h (Figure 2c)[41], and as a result, the onset times we observed likelyapply to more severe collisions. The amplitude of surfaceEMG from superficial neck muscles increases with bothincreasing sled acceleration and velocity change [42], andbased on this pattern, we would normally expect largermultifidus EMG activity in higher severity collisions.However, since supra-maximal EMG activity was alreadyevoked in some subjects by our low-severity sled perturba-tion, it is not clear whether there is scope for increasedmultifidus EMG with increased perturbation severity. Fur-Table 2: Catastrophic failure loads of isolated facet capsular ligaments as a function of elongation rate.Study N Gender Elongation rate (mm/s) Failure load (N)Winkelstein et al. (1999) [43] 12M 0.0083 84 ± 37Siegmund et al. (2001) [7] 13F 0.0083 94 ± 31Myklebust et al. (1988) [44] 2–5* 10 108 ± 40**Winkelstein et al. (1999) [43] 12M 100 136 ± 60Ivancic et al. (2007) [45] 32 723 220 ± 84Shim et al. (2006) [46] 15M 9600 – 13,600 260 ± 112Page 7 of 9(page number not for citation purposes)* N at each cervical level from C2/3 to C7/T1 between 2 and 5** average of the means and standard deviations at cervical level from C2/3 to C7/T1; also values reported here are half those reported by Myklebust et al (1988) because left and right ligaments were tested simultaneously.BMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/80ther work is needed to explore this issue. The acousticstimulus (124dB) was louder than generated by mostvehicle-to-vehicle collisions (92 to 110 dB, unpublishedmeasurements). In this experiment, however, we did notattempt to replicate the noise of a collision; rather we wereinterested in whether a pure startle response involves mul-tifidus muscle activation. Our loud tone thus served toevoke a startle response without the confounding effect ofa postural response, and thereby allowed us to seewhether startle – by itself – activated the multifidus mus-cle. We also did not randomize the presentation order ofthe startle and perturbation stimuli, and as a result theperturbation response may have been attenuated by thepreceding startle stimulus. In an actual rear-end collision,the startle component of the response [16] would likelybe evoked by a combination of tactile (interaction withthe seat) and acoustic (crash noise) sensory stimulation.Moreover, vehicle occupants who are unexpectedly struckfrom behind may experience a superposition of the twoneuromuscular responses we observed here.Our subjects were seated in an unmodified automobileseat similar in construction but perhaps different inmechanical properties from more recent automobileseats. Compared to our seat, some modern seats, particu-larly those with integrated seat belts, likely have a higherseatback stiffness that could produce higher occupantaccelerations that in turn could increase the neck muscleresponse amplitude. Other modern seats, such as someanti-whiplash seats, may have lower seatback restitutionvalues than the seat we used, and thus yield lower occu-pant speed changes that in turn could decrease the neckmuscle response. Further work is needed to explore theeffect of the different design trends in modern seats on thetiming and amplitude of the neck muscle response.The ability to generalize our results to the broader popu-lation is limited by the small number of subjects wetested. The goal of this experiment, however, was to estab-lish the plausibility of a potential role for multifidus in thegenesis of whiplash injury, and given the invasive natureof our experiment, nine subjects was deemed sufficient toestablish this plausibility. Based on our results, additionalwork is now justified using a larger number of subjectsand a wider range of impact severities to further explorethe role of multifidus in whiplash injury.ConclusionIn summary, cervical multifidus muscle activity is evokedby both horizontal sled accelerations and acousticallystartling tones. This finding represents one step towardsunderstanding whether reflex activation of the multifidusmuscle can exacerbate a whiplash injury involving the cer-Competing interestsAuthors GPS and JRB hold shares in and are employed bya forensic consulting company that may benefit from pub-lication of these data. Authors JSB, MGC and JTI declareno conflicts.Authors' contributionsAll authors participated in conceiving and designing thestudy. Pilot work was carried out by GPS, J–SB, JRB andJTI. GPS and J–SB conducted the experiments with assist-ance from MGC and JTI. GPS and J–SB analyzed the datawith input from the other authors. GPS drafted the man-uscript with feedback from all authors. All authors readand approved the final manuscript.AcknowledgementsExternal funding was provided by the Canadian Institute for Health Research (CIHR), Michael Smith Health Research Foundation (MSHRF), Workers' Compensation Board (WCB) of British Columbia, BC Neuro-trauma Fund, Natural Sciences and Engineering Research Council (NSERC), and University of British Columbia Startup Funding. Thanks to Dr. Bill Sheel for the use of his ultrasound equipment (funded by CFI), and Mr. Jeff Nickel and Mr. Mircea Oala-Florescu for their assistance with equipment. Thanks also to Trudy Harris and Dr. Alex MacKay of the UBC MRI Research Cen-tre.References1. Lord SM, Barnsley L, Wallis BJ, Bogduk N: Chronic cervical zyga-pophysial joint pain after whiplash. A placebo-controlledprevalence study.  Spine 1996, 21:1737-1744.2. Lord SM, Barnsley L, Wallis BJ, McDonald GJ, Bogduk N: Percutane-ous radio-frequency neurotomy for chronic cervical zygapo-physeal-joint pain.  N Engl J Med 1996, 335:1721-1726.3. McDonald GJ, Lord SM, Bogduk N: Long-term follow-up ofpatients treated with cervical radiofrequency neurotomy forchronic neck pain.  Neurosurgery 1999, 45:61-67.4. Kaneoka K, Ono K, Inami S, Hayashi K: Motion analysis of cervicalvertebrae during whiplash loading.  Spine 1999, 24:763-769.5. Yoganandan N, Pintar FA, Klienberger M: Cervical spine vertebraland facet joint kinematics under whiplash.  J Biomech Eng 1998,120:305-307.6. Winkelstein BA, Nightingale RW, Richardson WJ, Myers BS: Thecervical facet capsule and its role in whiplash injury: a biome-chanical investigation.  Spine 2000, 25:1238-1246.7. Siegmund GP, Myers BS, Davis MB, Bohnet HF, Winkelstein BA:Human cervical motion segment flexibility and facet capsu-lar ligament strain under combined posterior shear, exten-sion and axial compression.  Stapp Car Crash Journal 2000,44:159-170.8. Pearson AM, Ivancic PC, Ito S, Panjabi MM: Facet joint kinematicsand injury mechanisms during simulated whiplash.  Spine2004, 29:390-397.9. Lee KE, Thinnes JH, Gokhin DS, Winkelstein BA: A novel rodentneck pain model of facet-mediated behavioral hypersensitiv-ity: implications for persistent pain and whiplash injury.  JNeurosci Methods 2004, 137:151-159.10. Lu Y, Chen C, Kallakuri S, Patwardhan A, Cavanaugh JM: Neuro-physiological and biomechanical characterization of goatcervical facet joint capsules.  J Orthop Res 2005, 23:779-787.11. McLain RF: Mechanoreceptor endings in human cervical facetjoints.  Spine 1994, 19:495-501.12. Dwyer A, Aprill C, Bogduk N: Cervical zygapophyseal joint painpatterns. I: A study in normal volunteers.  Spine 1990,15:453-457.13. Winkelstein BA, McLendon RE, Barbir A, Myers BS: An anatomicalPage 8 of 9(page number not for citation purposes)vical facet capsule ligament. investigation of the human cervical facet capsule, quantifyingmuscle insertion area.  J Anat 2001, 198(Pt 4):455-461.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Musculoskeletal Disorders 2008, 9:80 http://www.biomedcentral.com/1471-2474/9/8014. Anderson JS, Hsu AW, Vasavada AN: Morphology, architecture,and biomechanics of human cervical multifidus.  Spine 2005,30:E86-91.15. Siegmund GP, Blouin JS, Brault JR, Hedenstierna S, Inglis JT: Electro-myography of superficial and deep neck muscles during iso-metric, voluntary and reflex contractions.  J Biomech Eng 2007,129:66-77.16. Blouin JS, Inglis JT, Siegmund GP: Startle responses elicited bywhiplash perturbations.  J Physiol 2006, 573(Pt 3):857-867.17. Blouin JS, Inglis JT, Carpenter MG, Siegmund GP: Neural control ofsuperficial and deep neck muscles in humans.  J Neurophysiol2007, 98:920-928.18. Blouin JS, Inglis JT, Siegmund GP: Auditory startle alters theresponse of human subjects exposed to a single whiplash-likeperturbation.  Spine 2006, 31:146-154.19. Siegmund GP, King DJ, Lawrence JM, Wheeler JB, Brault JR, Smith TA:Head/neck kinematic response of human subjects in low-speed rear-end collisions.  In Proceedings of the 41st Stapp CarCrash Conference: 13–14 November 1997; Lake Buena Vista, FL Warren-dale, PA: Society of Automotive Engineers; 1997:357-385. 20. Siegmund GP, Inglis JT, Sanderson DJ: Startle response of humanneck muscles sculpted by readiness to perform ballistic headmovements.  J Physiol 2001, 535:289-300.21. Brown P, Rothwell JC, Thompson PD, Britton TC, Day BL, MarsdenCD: New observations on the normal auditory startle reflexin man.  Brain 1991, 114(Pt 4):1891-1902.22. Bisdorff AR, Bronstein AM, Gresty MA: Responses in neck andfacial muscles to sudden free fall and a startling auditorystimulus.  Electroencephalogr Clin Neurophysiol 1994, 93:409-416.23. Blouin JS, Descarreaux M, Belanger-Gravel A, Simoneau M, TeasdaleN: Attenuation of human neck muscle activity followingrepeated imposed trunk-forward linear acceleration.  ExpBrain Res 2003, 150:458-464.24. Siegmund GP, Sanderson DJ, Myers BS, Inglis JT: Rapid neck muscleadaptation alters the head kinematics of aware and unawaresubjects undergoing multiple whiplash-like perturbations.  JBiomech 2003, 36:473-482.25. Staude G, Wolf W: Objective motor response onset detectionin surface myoelectric signals.  Med Eng Phys 1999, 21:449-467.26. Staude GH: Precise onset detection of human motorresponses using a whitening filter and the log-likelihood-ratio test.  IEEE Trans Biomed Eng 2001, 48:1292-1305.27. Tennyson SA, Mital NK, King AI: Electromyographic signals ofthe spinal musculature during +Gz impact acceleration.Orthop Clin North Am 1977, 8:97-119.28. Andersen LL, Aagaard P: Influence of maximal muscle strengthand intrinsic muscle contractile properties on contractilerate of force development.  Eur J Appl Physiol 2006, 96:46-52.29. Kumar S, Narayan Y, Amell T: An electromyographic study oflow-velocity rear-end impacts.  Spine 2002, 27:1044-1055.30. Vasavada AN, Li S, Delp SL: Three-dimensional isometricstrength of neck muscles in humans.  Spine 2001, 26:1904-1909.31. Myers BS, Woolley CT, Slotter TL, Garrett WE, Best TM: The influ-ence of strain rate on the passive and stimulated engineeringstress – large strain behavior of the rabbit tibialis anteriormuscle.  J Biomech Eng 1998, 120:126-132.32. Scott SH, Brown IE, Loeb GE: Mechanics of feline soleus: I. Effectof fascicle length and velocity on force output.  J Muscle Res CellMotil 1996, 17:207-219.33. Moroney SP, Schultz AB, Miller JA: Analysis and measurement ofneck loads.  J Orthop Res 1988, 6:713-720.34. Choi H, Vanderby R: Muscle forces and spinal loads at C4/5level during isometric voluntary efforts.  Med Sci Sports Exerc2000, 32:830-838.35. Stemper BD, Yoganandan N, Pintar FA: Effects of abnormal pos-ture on capsular ligament elongations in a computationalmodel subjected to whiplash loading.  J Biomech 2005,38:1313-1323.36. Boyd-Clark LC, Briggs CA, Galea MP: Comparative histochemi-cal composition of muscle fibres in a pre- and a postvertebralmuscle of the cervical spine.  J Anat 2001, 199(Pt 6):709-716.37. Boyd-Clark LC, Briggs CA, Galea MP: Muscle spindle distribution,morphology, and density in longus colli and multifidus mus-cles of the cervical spine.  Spine 2002, 27:694-701.39. Treleaven J, Jull G, Sterling M: Dizziness and unsteadiness follow-ing whiplash injury: characteristic features and relationshipwith cervical joint position error.  J Rehabil Med 2003, 35:36-43.40. Heikkila H, Wenngren B: Cervicocephalic kinesthetic sensibil-ity, active range of cervical motion, and ocular function inpatients with whiplash injury.  Arch Phys Med Rehabil 1998,79:1089-1094.41. Siegmund GP, Heinrichs BE, Lawrence JM, Phillipens MMGM: Kineticand kinematic responses of the RID2a, Hybrid III and humanvolunteers in low-speed rear-end collisions.  In Stapp Car CrashJ Volume 45. Ann Arbor, MI: Stapp Association; 2001:239-256. 42. Siegmund GP, Sanderson DJ, Inglis JT: Gradation of neck muscleresponses and head/neck kinematics to acceleration andspeed change in rear-end collisions.  Stapp Car Crash J 2004,48:419-430.43. Winkelstein BA, Nightingale RW, Richardson WJ, Myers BS: Cervi-cal facet joint mechanics: Its application to whiplash injury(99SC15).  Proc Stapp Car Crash Conference 1999, 43:243-252.44. Myklebust JB, Pintar F, Yoganadan N, Cusick JF, Maiman D, Myers TJ,Sances A: Tensile strength of spinal ligaments.  Spine 1988,13:526-531.45. Ivancic PC, Coe MP, Ndu AB, Tominga Y, Carlson EJ, Rubin W, Pan-jabi MM: Dynamic mechanical properties of intact human cer-vical spine ligaments.  The Spine Journal 2007, 7:659-665.46. Shim VPW, Liu JF, Lee VS: A technique for dynamic tensile test-ing of human cervical spine ligaments.  Experimental Mechanics2006, 46:77-89.Pre-publication historyThe pre-publication history for this paper can be accessedhere:http://www.biomedcentral.com/1471-2474/9/80/prepubyours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralPage 9 of 9(page number not for citation purposes)38. Evans RW: Some observations on whiplash injuries.  Neurol Clin1992, 10:975-997.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items