Abstract
Several hypotheses have been proposed to explain the mechanism of injury in whiplash including, pressure on nerve root ganglia, stretching of facet capsules, or damage to facet articular cartilage. These injury mechanisms have not been directly compared in the same study. A comparison could provide insight into the most likely mechanism of whiplash injury. Twenty eight volunteers underwent rear impacts with head and chest acceleration data collected. The same apparatus was used to test 11 cervico-thoracic human cadaveric spines with an instrumented headform attached. Head acceleration, individual vertebral kinematics from high speed video, local nerve root pressure, and facet joint contact pressures were collected during impacts. Each specimen was tested first at an impact acceleration similar to that of volunteers, who reported minimal or no symptoms after the test, then at double the acceleration. Head X (forward) and Z (upward) accelerations of cadaveric specimens were very similar in time sequence and magnitude to those of unprepared volunteers. Pressure around the lower cervical nerve roots ranged from 2.7kPa to 10kPa, and occurred generally after chest but before peak head acceleration. Facets at C4–5 and C5–6 had the highest probability (64% and 71% respectively) of pinching. Neither pressure rise nor pinching changed significantly with increased acceleration. Vertebral intersegmental extension rotations (4°–9.5°) and posterior translations (3.7–8.9 mm) peaked near maximum head excursion into the head restraint, at the time of peak head acceleration. Vertebral shear translations showed the largest (and only significant) increases with increased impact acceleration. This data implies that facet shearing was most sensitive to the increased acceleration in this experiment and may be a primary mechanism of cervical spine injury in rear impacts.
An average of 805,581 whiplash injuries occurred annually in the United States from 1988 to 1996, with an annual estimated cost of $5.2 billion dollars [12]. The diagnosis of whiplash is confounded by a general lack of objective symptoms, but may result in a long lasting chronic condition [1,2,4]. Chronic whiplash symptoms derive from the force applied during a rear impact, and may result from abnormal intersegmental displacements of the cervical vertebrae. In this study a comparison was made of the changes in vertebral rotations, facet displacements, facet joint pinching, and nerve root compression to increasing impact acceleration, all of which have been proposed as mechanisms of whiplash injury.
The response of an occupant to a rear-end impact has been well documented [3,5,6,8,9–11, 13–20]. Several recent studies, most performed using cadaveric cervical spine preparations, have shown that an initial transient “S” shape of the cervical spine occurs in response to a rear-end impact due to the lower cervical spine being thrust forward with the torso by the seat while the head remained initially level. This “S” shape was related, in different reports, to non-physiologic extension of the lower cervical segments [14], pinching instead of gliding of the facets [3,13], increased facet capsular tissue strains [5], and transient compression of the neural tissues [6]. Any or all of these mechanisms may account for symptoms experienced by victims of rear impact. However, the studies cited employed different experimental conditions making comparisons of the relative importance of a particular mechanism difficult. The first goal of this study was to determine if cadaveric spinal preparations behave in a similar manner, kinematically, to unprepared volunteers and therefore whether they could be used to provide meaningful information related to mechanisms of whiplash injury. If so, the second goal was to determine which mechanisms, observed in other studies, occurred in our specimens, during an impact. The third goal was to determine in which, if any, tissue deformations changed as peak acceleration doubled.
METHODS
DESIGN OF THE STUDY
Before studying the response of cadavers to impact, it was necessary to demonstrate that cadaveric spines have kinematics similar to unprepared volunteers. Therefore head and chest (sternum, or T1) X (forward) and Z (upward) accelerations were compared based on the assumption that if the magnitudes and times of occurrence of the forces acting on the head and torso were similar from volunteers to cadavers, then the internal spinal forces should be similar as well. To determine the most likely mechanism of injury in cadaver specimens, the four proposed mechanisms (spinal rotation, spinal shear translation, nerve root pressure, and facet impingement) were first measured. This first experiment, in cadavers, was run with accelerations similar to those experienced by volunteers without injury. Then the acceleration was doubled, so that it might be within the range of injury, and the measurements that increased significantly were determined. Measurements which changed significantly with increased acceleration were more likely to be probable mechanisms of injury, compared to those which were unchanged from levels tolerable by volunteers.
VOLUNTEER TESTING
One of 19 seats previously tested for impact properties [20] was selected as having midrange stiffness and energy absorption properties. It was mounted on a 6 wheeled sled running inside guide rails and had an impact absorbing bumper constructed of two aluminum channels separated by two sets of rubber doughnuts. Energy was provided by a pendulum with a total mass of about 73 kg. Its drop height produced a velocity of the pendulum at impact of about 6.4 kph (4 mph) and bumper compression resulted in an impact with an acceleration rising linearly to a peak of about 2.5g in about 66 msec, Fig 1, followed by a deceleration reaching a peak of −1.3g at 103 msec due to frictional interaction of the wheels of the sled with the frame rails. Overall the sled traveled about 40 cm after impact reaching a peak velocity of 3.9 kph. This pulse represented both the acceleration and deceleration components in a rear end impact of a stopped vehicle with the brakes applied.
Figure 1.
Mean input acceleration pulse in volunteer testing, including deceleration of sled by frictional contact with rails.
Testing was performed with the approval of the Institutional Review Board of the University of Washington. A total of 28 subjects were tested from which 26 intact data records were analyzed. The subjects were recruited from hospital employees and included 14 females (age range 22–64 yrs) and 12 males (age range 28 to 50 years). Each subject was seated in the sled, and restrained with lap and shoulder belts. A light plastic headband was secured on the subject’s head with an elastic strap under the chin. It contained 5 accelerometers (PCB Piezotronics Inc, Depew, NY), 2 uniaxial, measuring X (forward-backward) and Z (upward-downward) accelerations at approximately the level of each ear, (in the sagittal plane) and one triaxial, located at the apex of the head forming a frontal plane with the accelerometers at both ears. The orientations correspond to those described in SAE-J211, and the anatomical planes, as described by White and Panjabi [23]. The accelerometers at ear level were located approximately at the estimated center of gravity of the volunteer’s head, in the sagittal plane [12,28].
Signals were sampled using a laptop computer (Powermac G3, Apple Computer Co, Cupertino, CA) with an A/D card and software (LabVIEW, National Instruments, Austin, TX) and were filtered according to SAE J 211 using a 5th order 4 pole Butterworth digital filter implemented in LabVIEW. To describe the translation and rotation of the subject’s head in the XZ plane, it was transformed into a translation with reference to the lower part of the ear near the TMJ joint and an XZ plane rotation about this point. The actual origin for each accelerometer pair was at the intersection of the axes of the X and Z accelerometers.
CADAVERIC TESTING
Fresh frozen cadaveric preparations were thawed and dissected into a cervical and upper thoracic spine unit. Muscle and soft tissues were removed maintaining the discs, facet capsules, and ligaments intact. Each spine was examined by manual palpation and with lateral and AP radiographs. Disc degeneration at each level in the cervical spine was scored from the radiographs by two experienced observers using the following grades; Grade 1: end plate (EP) uniform thickness, vertebral body (VB) rounded margins, Grade 2, EP irregular thickness, VB pointed margins, Grade 3, EP focal defects, VB chondrophytes, Grade 4, EP fibrocartilage, VB < 2mm osteophytes, Grade 5, EP diffuse sclerosis, VB > 2mm osteophytes, [29]. A mean score for each spine, averaging scores for each level, was determined. Those with significantly restricted, degenerated, or hypermobile segments were not used. Eleven cervico-thoracic specimens were selected from a total of 16 available. Table 1 lists the characteristics of the specimens.
Table 1.
Characteristics of donors of the cervical spine specimens used in this study. 1 see text for degeneration scoring system
| Specimen number | Gender | Age at death (y) | Degeneration score1 |
|---|---|---|---|
| 1 | F | 63 | 1.00 |
| 2 | F | 77 | 1.33 |
| 3 | F | 59 | 2.50 |
| 4 | M | 82 | 1.83 |
| 5 | M | 85 | 2.67 |
| 6 | F | 66 | 1.00 |
| 7 | F | 83 | 3.33 |
| 8 | F | 85 | 1.83 |
| 9 | M | 83 | 4.00 |
| 10 | M | 80 | 1.17 |
| 11 | F | 84 | 1.83 |
The sled previously used for testing volunteer response to rear-impact [20] was modified by removing the seat and replacing it with a frame and platform to support the cadaveric specimens. The sled frame, bumper, wheels and guide rails remained unaltered from the volunteer experiment. The head restraint of the seat used in that experiment was mounted onto the platform, shown in Figure 2. The lower end of the specimen was flexibly mounted in a clamping box with rubber pads contacting the thoracic vertebrae.
Figure 2.
Cadaveric test arrangement showing (left) (A) skull replica attached to cervical spine, (B) head restraint, (C) vertebral and lateral mass video markers on C3–T1. (right) (A) pressure sensitive film in facet joint, and (B) pressure transducer in neuroforaminal space.
A load cell (WSM Industries, Inc) placed under the specimen mounting box was used to measure spinal shear force during impact. A plastic replica of a human skull (Anatomical Chart Co, Skokie, IL) was instrumented with a triaxial accelerometer (Kistler Instrument Corp, Amhurst, NY) placed at its estimated center of gravity and was filled with ultrasonic jelly to mimic the weight and mass distribution of the brain. The center of gravity of the skull was determined by balance weighing of the skull. The plastic skull was attached to a segment of bone remaining from the base of the skull of the specimen. All cervical spinal joints remained intact. Using a standardized plastic skull removed some variability from responses of the specimens due to different head weights and allowed the use of an accelerometer at the head CG. A stop was placed forward of the head to prevent excessive flexion after impact since no muscles were available to control and arrest head motion.
Local vertebral kinematics were determined from high speed video (Kodak, EktaPro, Rochester, NY) of markers placed onto each vertebra, from C3 to T1, taken at 1000 frames/sec. Orientations were consistent with SAE J-211. Pins placed anteriorly into each vertebral body were used to determine vertebral angulation and two 3.5 mm screws placed into each lateral mass allowed determination of facet orientation and relative displacement. Imaging software (WINanalyze Mikromak GmbH, Erlangen, Germany), accurate to 0.01 pixels, was used to determine marker coordinate positions in the video frames. This analysis was limited to the sagittal plane. A marker was present in each video and was used for calibration and scaling.
Neural tissue compression was determined by mounting miniature pressure transducers, Figure 2, approximately 1 mm thick and 3 mm in diameter, (model EPL-B02-5P, Entran Corp, Fairfield, NJ) into the right side foraminal spaces at C5–6, C6–7, and C7-T1. The transducers were inserted to a depth of about 2 cm, just posterior to the nerve roots and the void space was filled with ultrasonic jelly. The nerves appeared to remain intact and were easily identifiable. Facet articular contact pressures (pinching) were measured with pressure sensitive film (Prescale Super Low, Fuji Corp, Tokyo, Japan). A scalpel was used to make a cut, about 1 cm wide, in the lateral margin of the facet capsule, Figure 2. The rest of the capsule remained intact. Film packets, about 1 cm wide and about 0.3 mm thick were inserted into the joints at C3–4, C4–5, C5–6 and C6–7, however a facet space was sometimes too degenerated to allow insertion of a transducer. Several studies have shown that at least a 50% to 75% resection of the facet capsule is required to cause significant changes in displacement under load [7,22] so these small capsule incisions were not considered to change the specimens’ response to force. The transducers were not analyzed quantitatively. Instead, the facets were rated either pinched, indicated by a small high pressure print, dark in color, on the film, or not pinched, demonstrated by a diffuse light colored print. Because of the qualitative nature of this part of the analysis, two investigators separately analyzed the pressure films and their results were compared.
The weight of the pendulum which impacted the rear bumper of the sled, was set to 267N, and raised to a height of about 38 cm. Upon impact a chest acceleration, detected by a uniaxial accelerometer mounted to the body of vertebra T1 of the specimen, was generated that was similar to accelerations detected at the sternums of volunteers in a previous study [20], Figure 3. Sampling was performed at 3 KHz, as described previously for volunteers. After head rebound from the head restraint, a stop was used to prevent excessive forward excursion of the head. After the first test at about 2.5g peak chest acceleration, the pendulum weight was increased to 1068N. The impact was repeated, doubling the chest acceleration to about 4g.
Figure 3.
Forward (X) accelerations of volunteers, at the sternum, (n = 26, +/− 1 sd) and cadavers, at T1 (N = 11, +/− 1 sd). (Thicker curves represent means, thinner curves, mean +/− 1 sd)
RESULTS
VOLUNTEER VS CADAVERIC ACCELERATIONS
The comparisons of head accelerations from cadaveric tests at the approximate c.g of the head, (identical head restraint, 4cm head-to-restraint gap) and volunteers for the same conditions are shown in Figure 4 for both X and Z directions. Tests in which the volunteer did not hit the head restraint were not used because they indicate significant muscle tensing and preparedness for the impact and it would be unreasonable to compare those tests to cadaver results. They also do not simulate a crash in which the occupant is unaware and surprised by the impact, a common circumstance.
Figure 4.
Head acceleration (g) vs time (msec) (upper) mean head X acceleration and mean +/− 1 sd for volunteers and cadavers, (lower) mean Z head accelerations.
Peak acceleration occurred at approximately the same time for both volunteers and cadavers but cadaver mean peak acceleration reached about 2.9g while volunteer acceleration (peak + 1 sd) was about 2.6g. The rebound peak (negative) acceleration was higher for cadavers since the heads of the cadavers all hit the flexion stop, which did not occur in volunteer tests. However, the peak intervertebral motions, which will be described in the following section, coincided with peak accelerations at the time of head restraint contact, so the rebound phase was not significant for our analysis. Peak Z accelerations in cadavers reached a mean of −1g, while volunteer accelerations (mean −1 sd) were about −0.9g. For both head X and Z accelerations, peaks occurred at similar times for volunteers and cadavers. Although the mean cadaveric accelerations were somewhat greater than mean volunteer accelerations, the mean +/− 1 sd values overlapped so they were not statistically different. Therefore, the cadaveric tests can be considered to reasonably replicate the kinematics of unprepared volunteers.
CADAVERIC INTERSEGMENTAL MOTIONS
Figure 5 shows representative time history data along with frames from the high speed video of the cadaveric testing, and from a representative volunteer, to demonstrate the position of the head and neck at the time of peaks in the values measured. Head X acceleration increased as the impact pulse was applied by the pendulum, peaking in this case at about 75 msec. During the same time interval the head acceleration rose to a first smaller peak representing its acceleration by the impact pulse. To this point there was only small C5–6 extension (positive rotation), C5–6 foramenal pressure rise, and posterior shear translation of C5 on C6. This indicates that the initial impact is not responsible for large tissue deformations in the cervical spine.
Figure 5.
(upper) Comparison of relative intervertebral posterior shear(X) displacements (mm), from one specimen in one test (middle) positions of a specimen during test, 1 msec, 71 msec, 141 msec, 237 msec, (lower) corresponding positions of a volunteer test subject.
During the second interval, from 75 msec until the peak head acceleration was reached at 144 msec, the impact pulse ended (at about 120 msec) and the chest acceleration became negative because the sled has a friction brake which slowed its forward motion, similar to the result of an occupant of a vehicle in a rear impact having his or her foot on the brake. The peak chest negative (rearward) acceleration occurred at nearly the same time as the peak positive (forward) acceleration of the head. As shown in the frames from the video clip, this occurred as the head contacted the head restraint. It was during this interval that the peak foramenal pressure occurred. The extension rotation of C5–6 and the posterior shear translation of C5 on C6 reached their maxima as the head reached its peak rearward excursion into the head restraint.
During the third interval from 145 msec to 237 msec, the head was thrust forward by the head restraint, decreasing both intersegmental extension rotation and posterior translation. During this time, the tissues of the neck started to develop tension that slowed the head forward motion. The shear translation became positive (anterior). The peak positive shear force, which is not shown, but is given in Table 2, corresponded to peak positive chest acceleration, and the peak negative shear force to peak negative chest acceleration. The interaction of the head with the head restraint was important in defining the maximum spinal intersegmental motions.
Table 2.
Peak head acceleration (g), cervical spine shear force (N), maximum neural foramenal pressure (kPa), and frequency of focal pressure development (“pinching”) (%) in facet joint surfaces in 11 specimens subjected to 2g and 4g peak sled accelerations.
| Parameter | 2g Acceleration mean (95% CI) | 4g Acceleration mean (95%CI) | p-value |
|---|---|---|---|
| Maximum acc’n (g) or force(N) | |||
| Chest acceleration in x-axis | 2.37 (2.15, 2.58) | 4.25 (3.91, 4.59) | <.0001 |
| Head acceleration in x-axis | 3.58 (3.38, 3.78) | 5.73 (4.95, 6.51) | <.0001 |
| Head acceleration in z-axis | 0.69 (0.17, 1.20) | 1.36 (0.26, 2.46) | 0.042 |
| Shear force in the sagittal plane | 98.5 (77, 120) | 167.9 (130, 206) | 0.004 |
| Maximum pressure(kPa) | |||
| C5–6 neural foramen | 2.75 (1.04, 4.46) | 3.42 (1.48, 5.36) | N.S. |
| C6–7 neural foramen | 4.00 (1.30, 9.30) | 4.38 (0.93, 9.68) | N.S. |
| C7-T1 neural foramen | 7.41 (2.74, 12.1) | 10.1 (4.24, 15.9) | N.S |
| Frequency (%) | |||
| C3–4 facet joint | 60 (0, 100) | 80 (24, 100) | N.S. |
| C4–5 facet joint | 86 (51, 100) | 71 (26, 100) | N.S. |
| C5–6 facet joint | 71 (26, 100) | 75 (36, 100) | N.S. |
| C6–7 facet joint | 33 (0, 88) | 33 (0, 88) | N.S. |
| C7-T1 facet joint | 25 (0, 100) | 25 (0, 100) | N.S. |
A summary of all measured parameters is given in Tables 2 and 3. Some trends can be observed from this data. The peak positive head acceleration in the forward (X) direction was larger than the chest acceleration and it occurred much later in the interval (compare Figs 3 and 4) indicating that the head acceleration was occurring mainly due to its interaction with the head restraint and not from the impact pulse. The head Z (upward) acceleration was much smaller than the X acceleration indicating that the motion of the head was primarily horizontal. The peak neuroforaminal pressure was greatest at C7-T1, while the frequency of facet pinching was greatest in the mid cervical levels (C4–C6) and least at C7-T1. Intersegmental extension was greater than flexion and posterior shear of the superior vertebra on the inferior vertebra was greater than anterior translation. Both extension and posterior translation occurred as the head moved rearward. Posterior translation was greatest at C3–4 and decreased towards C7-T1. Extension was lowest at C3–4.
Table 3.
Vertebral intersegmental displacements at 2 and 4g peak accelerations
| Parameter | 2g Acceleration mean (95% CI) | 4g Acceleration mean (95%CI) | p value |
|---|---|---|---|
| Flexion(deg) | |||
| C3–4 level | −1.76 (−3.09, −0.43) | −1.33 (−2.75, −0.09) | N.S. |
| C4–5 level | −0.46 (−2.80, −0.12) | −0.89 (−1.87, −0.10) | N.S. |
| C5–6 level | −0.82 (−1.61, −0.03) | −0.51 (−0.94, −0.07) | N.S. |
| C6–7 level | −0.59 (−1.14, −0.04) | −0.26 (−0.44, − 0.09) | N.S. |
| Extension (deg) | |||
| C3–4 level | 2.90 (1.53, 4.28) | 4.00 (2.26, 5.74) | N.S. |
| C4–5 level | 5.53 (3.65, 7.40) | 6.60 (4.32, 8.88) | N.S. |
| C5–6 level | 7.17 (4.85, 9.48) | 6.84 (5.07, 8.60) | N.S. |
| C6–7 level | 5.90 (3.08, 8.72) | 9.45 (4.7, 14.15) | 0.027 |
| Anterior translation(mm) | |||
| C3–4 level | 2.48 (1.29, 3.66) | 0.29 (0.03, 0.54) | 0.006 |
| C4–5 level | 1.90 (1.02, 2.77) | 0.21 (0.02, 0.40) | 0.003 |
| C5–6 level | 0.85 (0.24, 1.45) | 0.18 (0.02, 0.35) | 0.013 |
| C6–7 level | 0.69 (0.16, 1.22) | 0.40 (−0.04, 0.80) | N.S. |
| C7-T1 level | 0.47 (−0.28, 1.22) | 0.11 (−0.02, 0.24) | N.S. |
| Posterior translation(mm) | |||
| C3–4 level | −6.30 (−7.68, −4.92) | −8.94 (−10.7, −7.15) | <0.0001 |
| C4–5 level | −5.54 (−6.43, −4.66) | −7.88 (−9.20, −6.56) | 0.0001 |
| C5–6 level | −4.52 (−5.39, −3.66) | −6.46 (−7.69, −5.23) | <0.0001 |
| C6–7 level | −3.59 (−4.85, −2.33) | −5.52 (−7.30, −3.73) | 0.001 |
| C7-T1 level | −1.95 (−2.72, −1.18) | −3.72 (−4.63, −2.80) | 0.001 |
| Distraction (mm) | |||
| C3–4 level | 1.66 (0.87, 2.45) | 2.04 (1.37, 2.72) | N.S. |
| C4–5 level | 1.60 (1.05, 2.15) | 2.15 (1.09, 3.21) | N.S. |
| C5–6 level | 0.55 (0.18, 0.92) | 1.69 (0.88, 2.49) | 0.006 |
| C6–7 level | 1.11 (0.29, 1.93) | 0.86 (0.23, 1.49) | N.S. |
| C7-T1 level | 0.29 (0.07, 0.52) | 0.79 (−0.07, 1.65) | N.S. |
| Compression (mm) | |||
| C3–4 level | −0.54 (−1.08, −0.00) | −0.47 (−0.93, −0.01) | N.S. |
| C4–5 level | −1.04 (−1.58, −0.50) | −0.86 (−1.52, −0.20) | N.S. |
| C5–6 level | −1.40 (−2.10, −0.70) | −0.36 (−0.77, −0.05) | 0.001 |
| C6–7 level | −0.66 (−1.13, −0.20) | −1.21 (−1.98, −0.45) | N.S. |
| C7-T1 level | −0.91 (−1.57, −0.24) | −1.08 (−1.96, −0.19) | N.S. |
CHANGES IN SPINAL DEFORMATIONS WITH ACCELERATION
Tables 2 and 3 summarize the changes that occurred in tissue deformations as the acceleration pulse was doubled. Figure 6 also shows the percent changes in those measures found to be significantly different with increased impact acceleration. When the chest X acceleration increased by 79%, head X acceleration increased by 60%, as did head Z acceleration and shear force. The most consistent effect was on the intersegmental posterior shear translation of the superior vertebra at all levels, which increased from 42% to 91%. There were no significant effects on neuroforaminal pressure or the frequency of facet pinching.
Figure 6.
Summary of the percent changes in parameters with significant differences (p < 0.05) for impact accelerations of 2g or 4g.
DISCUSSION
In this study several questions related to the mechanics of whiplash to the cervical spine in rear-end motor vehicle impacts were addressed. In general there are two distinct kinematic motions. The torso is first thrust forward by the seat, however the resulting tissue deformations are small during this phase. As the head and cervical spine extend into the head restraint, the greatest intersegmental rotations and posterior shear displacements occur along, generally, with peak neuroforaminal pressures. Therefore it is the manner in which the head contacts the head restraint which should be of concern when addressing the mechanism and prevention of cervical spine deformations which could result in whiplash. The most sensitive measured parameter to increasing impact acceleration was the posterior shear translation of the superior on the inferior vertebra during the extension of the head and neck.
Four different mechanisms proposed by others, related to whiplash symptoms from rear impact were compared. Panjabi, et al, [14] proposed that non-physiologic vertebral intersegmental rotations occurred during rear-impact. In our study, at 4 g acceleration, intersegmental rotations reached a maximum mean of 9.45 deg at C6–7. White and Panjabi [23] defined a normal range of combined flexion and extension angular motion between 6 and 26 deg. The limit of normal intersegmental angular range of motion could have been exceeded at some levels in some specimens. Facet capsule shearing was proposed by Deng, et al, [5] due to shear forces between vertebral segments. White and Panjabi [23] suggest about 3.5 mm for the upper limit of normal sagittal plane intersegmental translation. We measured greater facet translations at all levels except for C7-T1 at 2g acceleration. Facet pinching, demonstrated by Cusick, et al [3] was observed as small high pressure zones on pressure prints occurring around the posterior and lateral borders of facet joints, with the greatest frequency in the upper to middle cervical segments. It was less likely in degenerated facets, probably because motion in these joints was restricted naturally. Finally, repeatable measurable rises in pressure were observed around the cervical nerve roots, with maximums occurring after peak forward chest acceleration and before peak head acceleration. Although all of the proposed mechanisms of whiplash were measurable, it was significant that only posterior intersegmental shear translations of the superior to inferior vertebra changed significantly as impact acceleration increased. The initial impact acceleration was chosen to be in the range of that found tolerable by volunteers without incident. This implies that facet pinching, nerve root pressures, and vertebral displacements of the magnitudes determined in the experiment, at 2g, may be tolerable. Since they didn’t change significantly as acceleration was increased to a range above that tolerable by volunteers, it seems reasonable to deduce that facet translations, which did change significantly, exceeding estimated normal limits, may be a more likely mechanism of whiplash symptoms.
There are at least two possible explanations for the foramenal pressure rise, intersegmental motion changing the volume of the foramenal space and the motion of the fluid components in the foramen against the bony components. It was not possible to separate these two effects in this study however it can be noted that doubling the impact acceleration had a modest (but not significant) effect on foramenal pressure.
The cadaveric experiments have limitations that should be recognized. The cervical spines were from quite old donors (Table 1) and some degeneration was present in them, which caused nonuniformity in the motion response, with some segments being hypermobile and others quite restricted in motion. On the other hand, it is known that the great majority of older adults have some degree of cervical spinal degeneration. A second limitation was not having the torso and seat. The cervical spine and head restraint were purposely isolated by not using a complete cadaver and seat. This reduced the confounding potential of other variables such as how the torso interacted with the seatback. The accelerations were tuned so that at the base of the spine they were similar to those experienced by volunteers, both horizontally and vertically. A third limitation is that about 1/3rd of the circumference of the facet capsules on the right side for 3 or 4 facet joints were sectioned as available. Separate studies [7,21] have shown that sectioning less than 50% of the facet capsules had little effect on the cervical intersegmental motion. The responses of cadaver specimens were also compared only to those volunteers whose heads did hit the head restraint during testing, which was obvious both from the video and from the head acceleration pulse. Volunteers whose heads did not hit the head restraint were assumed to be guarding against the impact through tensing of neck muscles and would not be representative of the typical occupant who is surprised by the impact.
A fourth limitation relates to the use of a head form instead of maintaining the natural head. Panjabi, et al used a model of whiplash in which a cervical spine was mounted to a sled with a head form at the C0 joint [14]. Conceptually, in whiplash the head acts as a rigid mass, loading the spine. Therefore, a head form with similar mass and geometric properties to the natural head should apply forces during whiplash which are similar to that of the natural head. The fact that the accelerations of the head form in the cadaver experiment were very similar to those of the heads of volunteers demonstrates that the head form was functioning mechanically as would the natural head. It is true that the 5% female would have a lower weight head than the 50% male, and therefore under the same acceleration, lower forces would be applied to the spines with smaller heads. However females have less neck musculature to resist the acceleration of the head, therefore higher loads may result, offsetting the effect of the smaller weight head. In most biomechanics experiments, specimens are exposed to the same loads, not a load proportional to size, age, or gender of the specimen.
In summary, the overall findings of this study are that all of the mechanisms proposed by others during whiplash, including facet pinching, intersegmental vertebral rotation and translation, and foramenal pressure rises do occur simultaneously during impact. Peak intersegmental facet translation occurred simultaneously with peak head acceleration and was most sensitive to increased impact acceleration. Facet shearing may be a primary mechanism of whiplash injury.
ACKNOWLEDGEMENT
A grant from the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, Atlanta, GA, is gratefully acknowledged
REFERENCES
- 1.Barnsley L, Lord SM, Wallis BJ, Bogduk N. The prevalence of chronic cervical zygapophysial joint pain after whiplash. Spine. 1995;20:20–26. doi: 10.1097/00007632-199501000-00004. [DOI] [PubMed] [Google Scholar]
- 2.Clark CR. Cervical spine: Neck pain, radiculopathy, and myelopathy. Instructional Course Lectures, American Academy of Orthopedic Surgeons; Atlanta, GA. Feb 1996. [Google Scholar]
- 3.Cusick J, Pintar FA, Yoganandan N. Whiplash syndrome Kinematic factors influencing pain patterns. Spine. 2001;26:1252–1258. doi: 10.1097/00007632-200106010-00015. [DOI] [PubMed] [Google Scholar]
- 4.Deans GT, Magalliard JN, Kerr M, Rutherford WH. Neck sprain-A major cause of disability following car events. Injury. 1987;18:10–12. doi: 10.1016/0020-1383(87)90375-5. [DOI] [PubMed] [Google Scholar]
- 5.Deng B, Begeman PC, Yang KH, Tashman S, King AI. Kinematics of human cadaver cervical spine during low speed rear-end impacts. STAPP Car Crash Journal. 2000;44:171–188. doi: 10.4271/2000-01-SC13. [DOI] [PubMed] [Google Scholar]
- 6.Eichburger A, Darok M, Steffan H, Leinzinger PE, Bostrom O, Svensson MY. Pressure measurements in the spinal canal of post-mortem human subjects during rear-end impact and the correlation of results to the neck injury criterion (NIC). World Congress on Whiplash and Associated Disorders; Vancouver, BC. Feb, 1999; 1999. pp. 346–359. [Google Scholar]
- 7.Kumaresan S, Yoganandan N, Pintar FA, Voo LM, Cusick JF, Larson SJ. Finite element modeling of cervical laminectomy with graded facetectomy. J Spinal Disord. 1997;10:40–6. [PubMed] [Google Scholar]
- 8.Matsushita T, Sato TB, Hirabayashi K, Fujimura S, Asazuma T. Takatori, X-ray study of the human neck motion due to head inertia loading. Transactions of the Society of Automotive Engineers, 942208; 1994. [Google Scholar]
- 9.McConnell WE, Howard PR, Guzman HM, et al. Analysis of human test subject kinematic responses to low velocity rear end impacts. Transactions of the Society of Automotive Engineers, 930889; 1993. [Google Scholar]
- 10.McConnell WE, Howard RP, Van Poppel J, Krause R, Guzman HM, Bomar JB, Raddin JH, Benedict JV, Hatsell CP. Human head and neck kinematics after low velocity rear-end impacts: Understanding whiplash. Transactions of the Society of Automotive Engineers, 952724; 1995. [Google Scholar]
- 11.Mertz HJ, Patrick LM. Investigation of the kinematics and kinetics of whiplash. Transactions of the Society of Automotive Engineers, 670919; 1967. [Google Scholar]
- 12.National Highway Traffic Safety Administration. Federal Register. 2001 Jan 4;66(3):967–999. [PubMed] [Google Scholar]
- 13.Ono K, Kaneoka K, Wittek A, Kajzer J. Cervical injury mechanism based on the analysis of human cervical vertebral motion and head-neck-torso kinematics during low speed rear impacts. Transactions of the Society of Automotive Engineers, 973340; 1997. [Google Scholar]
- 14.Panjabi MM, Grauer JN, Cholewicki J, Nibu K, Babat LB, Dvorak J. Whiplash trauma: a biomechanical viewpoint. In: Gunzburg R, Szpalski M, editors. Whiplash injuries: Current concepts in prevention, diagnosis, and treatment of the cervical whiplash syndrome. Lippincott-Raven; Philadelphia: 1998. [Google Scholar]
- 15.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. Transactions of the Society of Automotive Engineers, 973341; 1997. [Google Scholar]
- 16.Siegmund GP, Myers BS, Davis MB, Bohnet HF, Winkelstein BA. Human cervical motion segment flexibility and facet capsular ligament strain under combined posterior shear, extension, and axial compression. STAPP Car Crash Journal. 2000;44:159–170. doi: 10.4271/2000-01-SC12. [DOI] [PubMed] [Google Scholar]
- 17.Szabo TJ, Welcher JB, Anderson RD, Rice MM, Ward JA, Paulo LR, Carpenter NJ. Human occupant kinematic response to low speed rear end impacts. Transactions of the Society of Automotive Engineers, 940532; 1994. [Google Scholar]
- 18.Szabo TJ, Welcher JB. Human subject kinematics and electromyographic activity during low speed rear impacts. Transactions of the Society of Automotive Engineers, 962432; 1996. [Google Scholar]
- 19.Tencer AF, Martin D, Goodwin V, Schaefer G, Mirza S. Do Initial Head Rotation or Head CG Distance Above the Head Restraint Affect Head Kinematics in Response to A Rear-End Impact ?. SAE Whiplash 98 conference; Phoenix, AZ. Nov 1988. [Google Scholar]
- 20.Tencer AF, Mirza SK, Bensel K. The response of human volunteers to rear-end impacts: The role of head restraint properties. Spine. 2001;22:2432–2442. doi: 10.1097/00007632-200111150-00007. [DOI] [PubMed] [Google Scholar]
- 21.Thompson JP, Pearce RH, Schechter MT, Adams ME, Tsang IKY, Bishop PB. Preliminary evaluation of a scheme for grading the gross morphology of the human intervertebral disc. Spine. 1990;15:411–415. doi: 10.1097/00007632-199005000-00012. [DOI] [PubMed] [Google Scholar]
- 22.Voo LM, Kumaresan S, Yoganandan N, Pintar FA, Cusick JF. Finite element analysis of cervical facetectomy. Spine. 1997;22:964–9. doi: 10.1097/00007632-199705010-00006. [DOI] [PubMed] [Google Scholar]
- 23.White AA, Panjabi, MM. Clinical biomechanics of the Spine. 2. Lippincott; New York: [Google Scholar]