- Email: [email protected]
Kinematics of head movement in simulated low velocity rear-end impacts
Clinical Biomechanics 20 (2005) 1011–1018 www.elsevier.com/locate/clinbiomech
Kinematics of head movement in simulated low velocity rear-end impacts Ivonne A. Herna´ndez a, Ken R. Fyfe b, Giseon Heo a, Paul W. Major a
a,*
Faculty of Medicine and Dentistry, Department of Dentistry, University of Alberta, Room 4051b, Dentistry/Pharmacy Centre, Edmonton, AB, Canada T6G 2N8 b Department of Mechanical Engineering, University of Alberta, AB, Canada Received 24 November 2004; accepted 11 July 2005
Abstract Objectives. To evaluate kinematics of head movement related to impact velocity, gender and awareness in simulated low velocity rear-end impacts. Methods. Thirty individuals were subjected in random order to three rear-end impacts: two unexpected impacts causing chair acceleration of 4.5 m/s2 (slow) and 10.0 m/s2 (fast) and one 10.0 m/s2 expected impact. Rearward head displacement, and linear and angular head accelerations were recorded. Results. Angular head displacement was almost two times higher for the fast than the slow unexpected-impacts (P = 0.04). Rearward and forward angular head accelerations increased two to three times with increased impact magnitude (P < 0.05). Rearward and forward linear head accelerations were two and a half to three and a half times higher for the fast than for the slow unexpected impacts (P < 0.05). Males presented two times higher upward linear head acceleration than females in the unexpected fast impact. No significant magnitude differences were identified for impact awareness in kinematics of head movement (P > 0.05). Rearward angular head acceleration reached the peak between 62 and 84 ms later than the rearward linear head acceleration (P < 0.05) in all impacts. No significant differences were identified for timing of kinematics of head movement (P > 0.05) with increased impact magnitude; however, statistical powers were low. Interpretation. Kinematics of head movement increases with increased impact magnitude. Gender differences exist for vertical linear head acceleration only. Temporal and amplitude awareness do not change the magnitude in kinematics of head movement. There are temporal differences between angular and linear head accelerations. 2005 Elsevier Ltd. All rights reserved. Keywords: Whiplash; Kinematics of head movement; Head angular acceleration; Linear head acceleration; Angular head displacement
1. Introduction Research on ‘‘whiplash injury’’ has been conducted through three main types of studies: retrospective and prospective analysis of patient symptoms, studies that use human volunteers in simulated low velocity rearend collisions, and those that involve use of human
*
Corresponding author. E-mail address: [email protected] (P.W. Major).
0268-0033/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2005.07.002
cadavers. In turn, the data from these studies have been used to build mathematical models that simulate the biomechanics of whiplash. The main advantage of using human volunteers over mathematical or human cadaver models is biofidelity. Although the kinematics of the head–neck complex in simulated rear-end impacts have been described, the influence of awareness and gender remains controversial. The kinematics of the head–neck complex in simulated impacts has been depicted as an initial forward and upward movement of the torso followed by a rearward
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1012
movement of the head (McConnell et al., 1993; McConnell et al., 1995; Ono et al., 1997; Severy et al., 1955; Siegmund et al., 1997; Szabo et al., 1994). However, discrepancy regarding the timing of the kinematics of the head movement has been reported. This disagreement can be attributed to differences in pre-impact muscle contraction, collision severity, seat back and head restraint (Siegmund et al., 1997; Viano, 2003). Kumar et al. (2000) reported that awareness of an impact reduced the acceleration of the head. Siegmund et al. (2003) reported that event awareness of an impact produced kinematic responses that differed according to gender. Gender differences have also been reported by Siegmund et al. (1997) and by Brault et al. (2000), but not by Kumar et al. (2002). Therefore, this study was designed to determine how awareness and gender influence kinematics of head movement in human volunteers exposed to simulated rear-end impacts at low velocities. The objectives of this study were to evaluate: • Relationship of impact awareness with head acceleration. • Gender differences for head kinematics. • Behaviour of kinematics of head movement at two impact magnitudes.
2. Methods The Human Research Ethics Board at the University of Alberta approved the protocol for this study. Thirtyone individuals were screened and deemed eligible to participate in this study. Participants were recruited by poster advertisement on the University of Alberta campus. One male subject did not complete the experimental phase due to a panic attack type reaction leaving a final sample of 30 subjects. Demographic data are presented in Table 1. Participants were between 18 and 35 years old, healthy and asymptomatic from any masticatory, cervical or temporomandibular joint pain. Exclusion criteria included; any systemic medical condition, restricted cervical or mandibular movements, history of car accident or trauma to the back or neck within the last twelve months, more than one missing tooth by quadrant with
Table 1 Descriptive statistics of demographic data
Females Males
Number
Age (SD)
Height (SD)
Weight (SD)
12 18
25 (3.14) 25 (2.9)
167 (8.90) 178 (6.6)
62 (11.95) 75 (12.02)
Age is expressed in years, height in centimeters, and weight in kilograms.
the exception of the third molar, and the wearing of any type of occlusal appliance. Participants attended two appointments. At the first appointment subjects read and signed the information sheet, completed a medical history form and gave informed consent. Clinical examination of the head and neck was performed on each subject. The second appointment was the experimental phase; each subject underwent three impacts. Kinematics of head movement was recorded. 2.1. Experimental set up Two recording systems were used in the present study. One system recorded the chair and torso acceleration and the other the kinematics of head movement. Accelerometers were used to measure acceleration of the sled, subjectÕs torso and head. 2.1.1. Sled acceleration set up The sled system consisted of a 250 cm by 125 cm raised wooden platform, with two 200 cm long parallel tracks mounted along the length of the platform. A Volvo car seat was sturdily mounted on a rectangular sliding board coupled with the tracks for frictionreduced travel upon impact. One uniaxial accelerometer (25 g) (Crossbow Technology, San Jose´, CA, USA) was located in the car seat to measure acceleration of the sled relative to the floor. The magnitude of the acceleration of each impact was achieved with an assembly of a pneumatic cylinder connected to an air supply through a pressure regulator calibrated for the delivery of known forces. 2.1.2. Torso acceleration set up A second triaxial accelerometer (5 g) (Crossbow Technology, San Jose´, CA, USA) was located on top of the sternum, to measure torso acceleration relative to the sled. 2.1.3. Head acceleration set up A custom designed accelerometer system was developed to measure head acceleration. The system included: a multipurpose circuit board SB1 (Ross Stirling, Edmonton, AB, Canada), a 16 channel 12 bit A/D converter, ±5 V input range (National Instruments Corporation, Austin, TX, USA) and two biaxial accelerometers ±10 g (item model ADXL 210, Analog Devices, Norwood, MA, USA). The accelerometer board was attached to the maxillary teeth with a custom dental tray. A clear plastic extension attached to the tray, positioned the accelerometer system outside of the mouth. The accelerometer board was aligned with the midline of the subject. A linear power supply Hewlett Packard (Mississauga, ON, Canada) set to 5 V was the power source for this accelerometer board.
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
2.1.4. High speed video cameras Three reflectors in conjunction with three ProReflex Cameras (item model MCO 240, Qualisys, Savedalen, Sweden) with a PC reflex software (Qualisys, Savedalen, Sweden) were also used to record head motion. Two reflectors were placed on the plastic extension attached to the dental tray and one reflector was placed in the anterior temporal region of the head. 2.2. Data acquisition The directions of the global axis (Fig. 1) were defined as follows: • The Y-axis was defined as parallel to the direction of the earthÕs gravity and positive up. • The X-axis was defined as parallel to the floor with positive toward the front of the sled and the subject. • The Z-axis was defined as the medial–lateral plane with positive toward the subjectÕs right. Linear acceleration of the head in the X and Y axes at the site of the anterior temporal region was measured. Angular acceleration and displacement of the head were determined from the accelerometer board system, whereas, the angular displacement was obtained from the video cameras. Onset time and peak time for each variable was determined. Onset and peak time were relative to the onset of chair movement. Onset time was defined as the time in which 5% of the magnitude value of the peak occurred. Peak time was defined as the time in which the maxi-
1013
mum value of the variable was reached. Data acquisition was restricted to the first 750 ms after impact. 2.3. Experimental phase After measurement of the weight and height of each participant, subjects were prepared for impacts. The triaxial accelerometer was placed on the sternum of each participant. The custom tray with the accelerometer board and reflective balls was loaded with polyvinyl siloxane impression material (Kerr, Romulus, MI, USA) and placed in the upper jaw of each subject. Each subject underwent three impacts. Two unexpected impacts: slow and fast, and one expected impact of the same magnitude of the unexpected fast impact. The mean chair acceleration peaks in the three axes for each acceleration level are presented in Table 2. The peak acceleration values in the Z-axis were significantly lower than those of the X (P = 0.001, power = 0.999) or Y-axis (P = 0.001, power = 0.999) for both slow and fast impacts. This confirmed the assumption that the movement occurred mostly in the sagittal plane. The order of impacts was randomized. For the unexpected impacts, each subject listened to loud music and fabric blindfold was used to cover each participantÕs eyes. Subjects were aware that there would be an impact, but were not advised of timing or impact magnitude. There was no attempt to deceive the subject with a ‘‘surprise’’ impact. In the expected impact, subjects were told the magnitude of the impact in qualitative terms, i.e. a fast impact, and when the impact would happen. The experimental sessions lasted approximately 1 h. 2.4. Data processing A mechanical engineer, who was not a member of this research team, processed the raw accelerometer and video camera data files. The video camera files were processed in such a way that the reflective balls were labelled, missing data were
Table 2 Mean peak linear chair acceleration in the X, Y and Z axes
Acceleration X-axis (m/s2) Acceleration Y-axis (m/s2) Acceleration Z-axis (m/s2) Sample
Fig. 1. Graphic representation of the global reference axes.
Slow impact (SD)
Fast impact (SD)
Slow–fast P value and power
4.54 (0.81)
10.03 (1.81)
2.16 (1.51)
5.79 (3.20)
0.36 (0.14)
1.81 (0.28)
0.001* 0.999 0.001* 0.999 0.001* 0.999
30
60
One way ANOVA. Magnitude of the impacts expressed as mean of the linear acceleration peaks of the sled. P value and power is presented for each analysis. * The mean difference is significant at the .05 level.
1014
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
identified and the data could be read in a text editor. These new files contained the positional data for the three markers to a resolution of 1 mm. The three markers had their position monitored at a sampling rate of 200 Hz for 5 s. No video camera data were recorded prior to the start of the test. The accelerometer board system incorporated a low pass filter with a frequency cut-off of 50 Hz. The raw files from the accelerometer system recorded 10 s: 5 s pre-impact and 5 s post-impact. However, a 750 ms window at a sampling rate of 4000 Hz with 5 s of pre-recorded data was analyzed. Files for the different impacts were coded for blinding. Missing data points and data points presenting a technical error were identified and eliminated. The remaining sample size varied for specific kinematic measurements. The recording systems had as time 0 the firing of the pneumatic cylinder piston, which caused the acceleration of the chair. The timing of the response variables was adjusted so the onset of acceleration of the chair represented time 0 in each impact and for each subject. 2.5. Statistical analysis The data was organized in an Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA) and analyzed using SPSS (version 13.0). Repeated measures (MANOVA) statistical method was used to analyze the response of the variable with an increase of the magnitude of the impact, and its response in the expected and unexpected impacts. Gender was entered into the analysis. The same statistical method was used to compare the timing among the kinematic events of the head. The significance level alpha equal 0.05 was used to determine the level of significance of the data. Raw data of the mean peak magnitude for the linear head acceleration was transformed to natural logarithm in order to obtain a normal distribution and homogenize the variance between impact level and gender.
Fig. 2. Comparison of video camera and accelerometer system recording.
3. Results The video camera and accelerometer data presented a good agreement (Fig. 2). A representative sequence of kinematic events is depicted in Fig. 3. 3.1. Kinematics of the torso 3.1.1. Linear torso acceleration The mean peak magnitude of the linear acceleration for the torso relative to the chair is presented in Table 3. The peak acceleration of the torso was increased approximately 2.5 times in the X-axis and 3 times in the Y-axis from the slow unexpected to fast unexpected
Fig. 3. Representation of timing of head kinematics for one representative subject.
impacts. There were no significant differences in the peak acceleration of the torso in the X and Y axes regarding awareness. Difference by gender in the peak acceleration was not observed in the X-axis (P = 0.108, power = 0.360) and Y-axis (P = 0.285, power = 0.199).
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1015
Table 3 Mean peak of the linear torso acceleration in the X and Y axes Sample
Slow unexpected impact (SD)
Fast unexpected impact (SD)
Fast expected impact (SD)
Slow–fast unexpected P value and power
Fast: unexpected–expected P value and power
Acceleration X-axis (m/s2)
16 M 8F
3.34 (0.45)
8.47 (1.39)
8.33 (1.40)
0.001* 0.999
0.999 0.999
Acceleration Y-axis (m/s2)
16 M 8F
2.29 (0.42)
7.04 (1.68)
7.26 (2.05)
0.001* 0.999
0.999 0.999
Repeated measure test (MANOVA). Linear acceleration expressed as raw data. P value and power is presented for each analysis. * The mean difference is significant at the .05 level.
impact magnitude or expectation; however, the statistical power was low.
3.2. Kinematics of the head Mean peaks of the kinematic events of the head are presented in Table 4. Onset and peak times of the kinematic events are presented in Table 5. Values for females and males are presented separately when a difference by gender was detected. 3.2.1. Angular head displacement Rearward angular head displacement almost doubled with an increase of the impact magnitude. Gender differences were not observed for timing or magnitude of head displacement for any impact. However the statistical power was low for these variables. There was no difference in the magnitude of angular displacement regarding awareness. Statistical power was adequate for this variable. 3.2.2. Angular head acceleration The peak magnitude of rearward angular acceleration approximately doubled and the peak magnitude of forward angular acceleration approximately tripled with higher impact magnitude. No significant differences in peak rearward and forward angular head acceleration regarding expectation were detected. Gender differences for timing or magnitude of angular acceleration were not detected for any impact; however, the statistical power was low. No differences in the onset or peak angular head acceleration times were observed regarding
3.2.3. Linear head acceleration X-axis Rearward and forward linear peak acceleration were approximately 2.5–3.5 greater with increased impact magnitude. No differences in the peaks of the rearward or forward linear accelerations regarding gender were identified; however, the statistical power was low. There were no statistical differences regarding expectation. 3.2.4. Linear head acceleration Y-axis Upward and downward linear peak acceleration was approximately 2–3 times greater with increased magnitude. In the fast-unexpected impact, female subjects presented approximately two times higher upward peak acceleration than male subjects. No difference in the upward or downward peak linear head acceleration regarding awareness was detected. 3.2.5. Timing of head acceleration In the expected fast impact, the rearward peak angular acceleration started approximately 40 ms later than rearward peak linear acceleration (P = 0.0438, power = 0.999). The rearward angular acceleration reached the peak acceleration approximately 76 ms, 62 ms and 84 ms later than the rearward linear acceleration in the slow (P = 0.001, power = 0.977), unexpected-fast
Table 4a Mean peaks of the angular displacement, angular acceleration of the head Peaks
Units
Rearward ang displ Rearward ang acc Forward ang acc
Degrees Rad/s2 Rad/s2
Slow unexpected impact (SD)
Gender P Fast value unexpected power impact (SD)
5.86 (3.61) 0.411 0.125 16.31 (4.40) 0.075 0.432 15.10 (6.22) 0.065 0.461
11.20 (7.71)
Gender P Fast value expected Power Impact (SD)
0.411 0.125 31.51 (11.23) 0.075 0.432 48.45 (28.00) 0.065 0.461
12.83 (5.41)
Sample Gender P Slow–fast Fast: value unexpected P unexpected– power value, Power expected P value, power
0.411 0.125 30.75 (11.88) 0.075 0.432 44.85 (23.65) 0.065 0.461
0.040* 0.996 0.000* 0.999 0.000* 0.999
0.999 0.996 0.999 0.999 0.999 0.999
10M 8F 10 M 9F 10 M 9F
Repeated measures test. Data presented as raw data. P value and power is presented for each analysis. Rearward ang displ refers to rearward angular displacement. Rearward ang acc refers to the rearward angular acceleration. Forward ang acc refers to the forward angular acceleration of the head. * The mean difference is significant at the .05 level.
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1016
Table 4b Mean peaks of the linear acceleration of the head Peaks
Slow unexpected impact (SD)
Gender P value power
Fast unexpected impact (SD)
Gender P value power
Fast expected impact (SD)
Gender P value power
Slow–fast unexpected P value, power
Fast: unexpected– expected P value, power
Sample
Rearward li acc X-axis Forward li acc X-axis Upward li acc Y-axis
1.55 (0.31)
0.960 0.384 0.412 0.126 0.584 0.210
2.45 (0.42)
0.960 0.384 0.412 0.126 0.033* 0.581
2.44 (0.34)
0.960 0.384 0.412 0.126 0.955 0.210
0.001* 0.999 0.001* 0.999 0.001* 0.999
0.999 0.999 0.112 0.999 0.999 0.999
17 8 17 8 18 10
0.803 0.057
0.001* 0.999
0.644 0.999
18 M 10 F
Downward li acc Y-axis
0.71 (0.35) 0.70 (0.77)
0.71 (0.69)
0.803 0.057
1.94 (0.54) 1.90 1.14 1.63 1.83
(0.94) M (0.66) F (0.92) (0.80)
0.803 0.057
1.70 (0.39) 1.47 (0.64)
1.58 (0.64)
M F M F M F
Linear acceleration of the head was measured in m/s2 and it is expressed in natural logarithm data. Repeated measures test for detecting gender differences in head linear acceleration (power values are included), and response differences between impacts. One way ANOVA analysis for detecting gender differences in the upward linear acceleration Y-axis. M refers to males and F refers to females. P value and power is presented for all comparisons. Rearward li acc refers to rearward linear acceleration. Li acceleration was measured in m/s2. Upward li acc refers to the upward linear acceleration. Downward li acc refers to the downward linear acceleration. * The mean difference is significant at the .05 level.
Table 5 Mean of the onset and peak times of the angular displacement, angular acceleration and linear acceleration of the head
Onset time Rearward ang displ Peak time Rearward ang displ Onset time Rearward ang acc Peak time Rearward ang acc Onset time Forward ang acc Peak time Forward ang acc Onset time Rearward li acc Peak time Rearward li acc Onset time Forward li acc Peak time Forward li acc
Sample
Slow unexpected impact (SD)
Fast unexpected impact (SD)
Fast expected impact (SD)
Slow–fast unexpected P value, power
Fast: unexpected–expected P value and power
10 8 10 8 10 9 10 9 10 9 10 9 17 8 17 8 17 8 17 8
100.7 (63.58) T
121.05 (64.56) T
116.38 (51.54) T
245.27 (79.25)
252.72 (101.38)
266.66 (65.32)
37.84 (19.12)
58.11 (65.15)
63.40 (43.10)
108.71 (21.24)
137.03 (69.39)
124.94 (30.07)
216.92 (87.55)
205.97 (71.46)
200.69 (57.47)
263.85 (87.04)
259.46 (71.60)
252.96 (58.22)
20.57 (56.49)
48.82 (94.14)
16.44 (56.96)
31.68 (60.94)
69.84 (109.09)
27.92 (68.66)
276.64 (172.40)
295.76 (181.62)
309.28 (137.81)
289.52 (163.36)
317.84 (163.34)
326.80 (127.79)
0.433 0.238 0.999 0.114 0.616 0.581 0.244 0.829 0.999 0.084 0.999 0.066 0.564 0.194 0.300 0.269 0.999 0.079 0.999 0.102
0.999 0.238 0.999 0.114 0.999 0.581 0.999 0.829 0.999 0.084 0.999 0.066 0.999 0.194 0.999 0.269 0.999 0.079 0.999 0.102
M F M F M F M F M F M F M F M F M F M F
Repeated measures test. Data presented as raw data. Time was measured in milliseconds. Combined male and female sample. P value and power is presented for each analysis. Ang displ refers to angular displacement, ang acc refers to angular acceleration and li acc refers to linear head acceleration.
(P = 0.029, power = 0.614) and in the expected (P = 0.001, power = 0.999) impacts respectively. Differences in timing of head acceleration were not identified regarding impact speed or expectation; however, most of the statistical powers were low.
4. Discussion The current study analyzed the head kinematics in simulated rear-end impacts. The larger head accelera-
tion observed with an increase in impact magnitude in the present study are in agreement with previous studies (Welcher et al., 2001; Kumar et al., 2000; Kumar et al., 2002; Siegmund et al., 1997). The magnitude of angular displacements reported in the present study are also in agreement with the findings of other studies (Castro et al., 1997; McConnell et al., 1995; McConnell et al., 1993; Siegmund et al., 1997; Szabo et al., 1994; Szabo and Welcher, 1996; Matsushita et al., 1994). The smaller acceleration values reported in the current study compared with previous studies might be due to the lesser
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
impact magnitudes (Siegmund et al., 1997; McConnell et al., 1995). The use of human subjects in simulated rear-end impacts has two limitations: ethics and apprehension. Participants can only be exposed to minimal impacts and the observation of increased head displacement and acceleration with head movement even with a small increase in impact magnitude supports the mechanical role of head movement in whiplash injury. Unfortunately, the minimum impact magnitude at which patients report symptoms cannot be established. Furthermore the threshold of head displacement and head acceleration for injury remains unknown. In the present study, the initial head movement was translation followed by combined rotation and translation and ultimately by rotation. There was an instantaneous rather than fixed axis of head rotation and, therefore, the axis of rotation was not calculated. The linear acceleration of the head is specific to the location in which it was measured. Its value varies across the head depending on distance from the rotational axis. On the other hand, the value of the angular acceleration of the head is valid for any point of the head. Linear and angular acceleration values of the current study presented a significant increase at higher levels of impacts; however, the linear acceleration presented a larger increase, an earlier onset, and reached the peak acceleration faster than the angular acceleration. Therefore, it might be more appropriate to analyze both types of acceleration rather than assume the linear acceleration as the key movement of the head. Vertical linear peak head acceleration was lower in females. The findings of Siegmund et al. (2003) also reported similar gender differences in head accelerations. Kumar et al. (2000) reported gender differences in rearward linear head acceleration. Differences in vertical acceleration may be a function of the car seat relationship to the subjectÕs body. Males are generally taller than females and the contact with the back against the car seat in males may result in more upward movement. Greater angular head acceleration in females would increase the risk of cervical injury. The present study did not identify gender differences for horizontal or angular head accelerations and therefore increased risk of mechanical injury due to head movement in females was not supported. The prevalence of pain complaints in female subjects may be a function of muscle injury, or secondary to centrally mediated pain modulation. Awareness of impact may affect kinematics of head movement in a simulated rear end impact. Awareness of an event refers to the anticipation of such an event, and implies three dimensions: temporal, event and amplitude awareness (Siegmund, 2001; Frank, 1986). Temporal awareness refers to whether the subject knows about the exact timing in which an event will occur, event awareness describes whether the subject knows
1017
an event will occur, and amplitude awareness refers to the awareness of the amplitude of an imminent event (Siegmund, 2001). Kumar et al. (2000) reported that awareness resulted in 30% reduction in peak backward linear head acceleration. The current findings did not report significant differences in kinematics of head movement regarding temporal awareness. Siegmund et al. (2003) reported that lack of event awareness significantly affected kinematic gender response; therefore, they concluded that event awareness is the component of anticipation event that plays an influential role in the kinematics of the head and neck in a simulated whiplash event. The unexpected events in the present study are similar to the ‘‘unalerted’’ group in the Siegmund et al. (2003) study. Subjects were aware that they would eventually sustain impact and therefore had event awareness. To remove event awareness requires deception such as used by Siegmund et al. (2003). The expected impact events in the present study are similar to the ‘‘alerted’’ group in the Siegmund et al. (2003) study. The Kumar et al. (2000) study tested expectancy based on the subject being advised of timing and magnitude of impact. There was no record of deception for the unexpected impact. The paper does not state how long the subject was kept blindfolded with loud music prior to the impact. It seems likely; however, that there was event awareness but not temporal or magnitude awareness. A possible explanation for the differences reported in these studies and the present study may relate to the subjects being informed of the magnitude of impact. The present study utilized only one ‘‘alerted’’ impact magnitude. It is possible that the subjects did not perceive the simulated impact sufficiently noxious to get injured. This study presents some limitations. It is possible that the mechanism by which the accelerometer board was attached to the subject created an additional sensory stimulus that might have altered the head motion. The distance between the head of the participant and the headrest was not recorded, and might account for the gender differences reported in this study. The restriction of the kinematics analysis to the first 750 ms after impact did not allow insight regarding angular head displacement in the rebound phase. The low magnitude of impacts in this study might have not been sufficient to create an actual simulation of a whiplash event; and therefore, to be perceived sufficiently noxious by the participants in the expected condition.
5. Interpretation Based on the findings of the current study, the following can be concluded: There is increased angular head acceleration, increased linear head acceleration and increased angular head displacement with an increased
1018
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
impact magnitude. Females have less upward head movement than males. Temporal and amplitude awareness of a simulated impact do not produce differences in the magnitude of head kinematic events. Linear and angular magnitude peak accelerations as well as onset and peak time accelerations are not coincident. Conclusions regarding timing in kinematics of head movement cannot be drawn due to the low statistical power in most of the kinematic events. Acknowledgements The authors would like to thank Dr. S. Kumar for use of his lab, equipment and technical support. This research was supported by the University of Alberta, Fund for Dentistry Grant # 2002-01 and the McIntyre Memorial Research Fund. References Brault, J.R., Siegmund, G.P., Wheeler, J.B., 2000. Cervical muscle response during whiplash: evidence of a lengthening muscle contraction. Clin. Biomech. 15, 426–435. Castro, W.H., Schilgen, M., Meyer, S., Weber, M., Peuker, C., Wortler, K., 1997. Do ‘‘whiplash injuries’’ occur in low-speed rear impacts? Eur. Spine J. 6, 366–375. Frank, J.S., 1986. Spinal motor preparation in humans. Electroencephalogr. Clin. Neurophysiol. 63, 361–370. Kumar, S., Narayan, Y., Amell, T., 2000. Role of awareness in head– neck acceleration in low velocity rear-end impacts. Accid. Anal. Prev. 32, 233–241. Kumar, S., Narayan, Y., Amell, T., 2002. An electromyographic study of low-velocity rear-end impacts. Spine 27, 1044–1055. Matsushita, T., Sato, T., Hirabayashi, K., Fujimora, S., Asazuma, T., Takatori, T., 1994. Paper 942208 X-ray study of the human neck motion due to head inertia loading. In: 38th Stapp Car Crash Conference. Warrendale, PA, USA: Society of Automotive Engineers, Inc., pp. 55–64. McConnell, W., Howard, R.P., Guzman, H.M., et al., 1993. Paper 930889 Analysis of human test subject kinematic responses to low
velocity rear-end impacts. Presentation to SAE International Congress and Exposition. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 21–30. McConnell, W., Howard, R.P., Poppel, J.V., et al., 1995. Paper 952724 Human head and neck kinematics after low velocity rear-end impacts—Understanding ‘‘whiplash. In: 39th Stapp Car Crash Conference. Warrendale, PA, USA: Society Automotive Engineers, Inc., pp. 3106–3129. Ono, K., Kaneoka, K., Wittek, A., Kazjer, J., 1997. Paper 973340 Cervical injury mechanism based on the analysis of human cervical vertebral motion and head–neck–torso kinematics during low speed rear impacts. In: 41st Annual Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 339–356. Severy, D.M., Mathewson, J.H., Bechtol, C.O., 1955. Controlled automobile rear-end collisions, an investigation of related engineering and medical phenomena. Can. Serv. Med. J. 11, 727–759. Siegmund, G., 2001. The reflex response of human neck muscles to whiplash-like perturbations [Doctoral Thesis]. Vancouver, British Columbia, Canada: University of British Columbia. Siegmund, G.P., King, D.J., Lawrence, J.M., Wheeler, J.B., Brault, J.R., Smith, T.A., 1997. Paper 973341 Head/neck kinematic response of human subjects in low-speed rear-end collisions. In: 41st Annual Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society Automotive Engineers, Inc., pp. 357–385. Siegmund, G.P., Sanderson, D.J., Myers, B.S., Inglis, J.T., 2003. Awareness affects the response of human subjects exposed to a single whiplash-like perturbation. Spine 28, 671–679. Szabo, T.J., Welcher, J.B., 1996. Paper 962432 Human subject kinematics and electromyographic activity during low speed rear impacts. In: 40th Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 295–314. Szabo, T., Welcher, J., Anderson, R., et al., 1994. Paper 940532 Human occupant kinematic response to low speed rear-end impacts. In: 38th Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 23–36. Viano, D., 2003. Seat influences on female neck responses in rear crashes: A reason why women have higher whiplash rates. Traffic Inj. Prev. 4, 228–239. Welcher, J., Szabo, T., Voss, D., 2001. 2001-01-0899 Human Occupant Motion in Rear-end Impacts: Effects of incremental increases in velocity change. SAE 2001 World Congress. Warrendale, Pennsylvania: Society of Automotive Engineers, Inc.
Kinematics of head movement in simulated low velocity rear-end impacts Ivonne A. Herna´ndez a, Ken R. Fyfe b, Giseon Heo a, Paul W. Major a
a,*
Faculty of Medicine and Dentistry, Department of Dentistry, University of Alberta, Room 4051b, Dentistry/Pharmacy Centre, Edmonton, AB, Canada T6G 2N8 b Department of Mechanical Engineering, University of Alberta, AB, Canada Received 24 November 2004; accepted 11 July 2005
Abstract Objectives. To evaluate kinematics of head movement related to impact velocity, gender and awareness in simulated low velocity rear-end impacts. Methods. Thirty individuals were subjected in random order to three rear-end impacts: two unexpected impacts causing chair acceleration of 4.5 m/s2 (slow) and 10.0 m/s2 (fast) and one 10.0 m/s2 expected impact. Rearward head displacement, and linear and angular head accelerations were recorded. Results. Angular head displacement was almost two times higher for the fast than the slow unexpected-impacts (P = 0.04). Rearward and forward angular head accelerations increased two to three times with increased impact magnitude (P < 0.05). Rearward and forward linear head accelerations were two and a half to three and a half times higher for the fast than for the slow unexpected impacts (P < 0.05). Males presented two times higher upward linear head acceleration than females in the unexpected fast impact. No significant magnitude differences were identified for impact awareness in kinematics of head movement (P > 0.05). Rearward angular head acceleration reached the peak between 62 and 84 ms later than the rearward linear head acceleration (P < 0.05) in all impacts. No significant differences were identified for timing of kinematics of head movement (P > 0.05) with increased impact magnitude; however, statistical powers were low. Interpretation. Kinematics of head movement increases with increased impact magnitude. Gender differences exist for vertical linear head acceleration only. Temporal and amplitude awareness do not change the magnitude in kinematics of head movement. There are temporal differences between angular and linear head accelerations. 2005 Elsevier Ltd. All rights reserved. Keywords: Whiplash; Kinematics of head movement; Head angular acceleration; Linear head acceleration; Angular head displacement
1. Introduction Research on ‘‘whiplash injury’’ has been conducted through three main types of studies: retrospective and prospective analysis of patient symptoms, studies that use human volunteers in simulated low velocity rearend collisions, and those that involve use of human
*
Corresponding author. E-mail address: [email protected] (P.W. Major).
0268-0033/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2005.07.002
cadavers. In turn, the data from these studies have been used to build mathematical models that simulate the biomechanics of whiplash. The main advantage of using human volunteers over mathematical or human cadaver models is biofidelity. Although the kinematics of the head–neck complex in simulated rear-end impacts have been described, the influence of awareness and gender remains controversial. The kinematics of the head–neck complex in simulated impacts has been depicted as an initial forward and upward movement of the torso followed by a rearward
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1012
movement of the head (McConnell et al., 1993; McConnell et al., 1995; Ono et al., 1997; Severy et al., 1955; Siegmund et al., 1997; Szabo et al., 1994). However, discrepancy regarding the timing of the kinematics of the head movement has been reported. This disagreement can be attributed to differences in pre-impact muscle contraction, collision severity, seat back and head restraint (Siegmund et al., 1997; Viano, 2003). Kumar et al. (2000) reported that awareness of an impact reduced the acceleration of the head. Siegmund et al. (2003) reported that event awareness of an impact produced kinematic responses that differed according to gender. Gender differences have also been reported by Siegmund et al. (1997) and by Brault et al. (2000), but not by Kumar et al. (2002). Therefore, this study was designed to determine how awareness and gender influence kinematics of head movement in human volunteers exposed to simulated rear-end impacts at low velocities. The objectives of this study were to evaluate: • Relationship of impact awareness with head acceleration. • Gender differences for head kinematics. • Behaviour of kinematics of head movement at two impact magnitudes.
2. Methods The Human Research Ethics Board at the University of Alberta approved the protocol for this study. Thirtyone individuals were screened and deemed eligible to participate in this study. Participants were recruited by poster advertisement on the University of Alberta campus. One male subject did not complete the experimental phase due to a panic attack type reaction leaving a final sample of 30 subjects. Demographic data are presented in Table 1. Participants were between 18 and 35 years old, healthy and asymptomatic from any masticatory, cervical or temporomandibular joint pain. Exclusion criteria included; any systemic medical condition, restricted cervical or mandibular movements, history of car accident or trauma to the back or neck within the last twelve months, more than one missing tooth by quadrant with
Table 1 Descriptive statistics of demographic data
Females Males
Number
Age (SD)
Height (SD)
Weight (SD)
12 18
25 (3.14) 25 (2.9)
167 (8.90) 178 (6.6)
62 (11.95) 75 (12.02)
Age is expressed in years, height in centimeters, and weight in kilograms.
the exception of the third molar, and the wearing of any type of occlusal appliance. Participants attended two appointments. At the first appointment subjects read and signed the information sheet, completed a medical history form and gave informed consent. Clinical examination of the head and neck was performed on each subject. The second appointment was the experimental phase; each subject underwent three impacts. Kinematics of head movement was recorded. 2.1. Experimental set up Two recording systems were used in the present study. One system recorded the chair and torso acceleration and the other the kinematics of head movement. Accelerometers were used to measure acceleration of the sled, subjectÕs torso and head. 2.1.1. Sled acceleration set up The sled system consisted of a 250 cm by 125 cm raised wooden platform, with two 200 cm long parallel tracks mounted along the length of the platform. A Volvo car seat was sturdily mounted on a rectangular sliding board coupled with the tracks for frictionreduced travel upon impact. One uniaxial accelerometer (25 g) (Crossbow Technology, San Jose´, CA, USA) was located in the car seat to measure acceleration of the sled relative to the floor. The magnitude of the acceleration of each impact was achieved with an assembly of a pneumatic cylinder connected to an air supply through a pressure regulator calibrated for the delivery of known forces. 2.1.2. Torso acceleration set up A second triaxial accelerometer (5 g) (Crossbow Technology, San Jose´, CA, USA) was located on top of the sternum, to measure torso acceleration relative to the sled. 2.1.3. Head acceleration set up A custom designed accelerometer system was developed to measure head acceleration. The system included: a multipurpose circuit board SB1 (Ross Stirling, Edmonton, AB, Canada), a 16 channel 12 bit A/D converter, ±5 V input range (National Instruments Corporation, Austin, TX, USA) and two biaxial accelerometers ±10 g (item model ADXL 210, Analog Devices, Norwood, MA, USA). The accelerometer board was attached to the maxillary teeth with a custom dental tray. A clear plastic extension attached to the tray, positioned the accelerometer system outside of the mouth. The accelerometer board was aligned with the midline of the subject. A linear power supply Hewlett Packard (Mississauga, ON, Canada) set to 5 V was the power source for this accelerometer board.
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
2.1.4. High speed video cameras Three reflectors in conjunction with three ProReflex Cameras (item model MCO 240, Qualisys, Savedalen, Sweden) with a PC reflex software (Qualisys, Savedalen, Sweden) were also used to record head motion. Two reflectors were placed on the plastic extension attached to the dental tray and one reflector was placed in the anterior temporal region of the head. 2.2. Data acquisition The directions of the global axis (Fig. 1) were defined as follows: • The Y-axis was defined as parallel to the direction of the earthÕs gravity and positive up. • The X-axis was defined as parallel to the floor with positive toward the front of the sled and the subject. • The Z-axis was defined as the medial–lateral plane with positive toward the subjectÕs right. Linear acceleration of the head in the X and Y axes at the site of the anterior temporal region was measured. Angular acceleration and displacement of the head were determined from the accelerometer board system, whereas, the angular displacement was obtained from the video cameras. Onset time and peak time for each variable was determined. Onset and peak time were relative to the onset of chair movement. Onset time was defined as the time in which 5% of the magnitude value of the peak occurred. Peak time was defined as the time in which the maxi-
1013
mum value of the variable was reached. Data acquisition was restricted to the first 750 ms after impact. 2.3. Experimental phase After measurement of the weight and height of each participant, subjects were prepared for impacts. The triaxial accelerometer was placed on the sternum of each participant. The custom tray with the accelerometer board and reflective balls was loaded with polyvinyl siloxane impression material (Kerr, Romulus, MI, USA) and placed in the upper jaw of each subject. Each subject underwent three impacts. Two unexpected impacts: slow and fast, and one expected impact of the same magnitude of the unexpected fast impact. The mean chair acceleration peaks in the three axes for each acceleration level are presented in Table 2. The peak acceleration values in the Z-axis were significantly lower than those of the X (P = 0.001, power = 0.999) or Y-axis (P = 0.001, power = 0.999) for both slow and fast impacts. This confirmed the assumption that the movement occurred mostly in the sagittal plane. The order of impacts was randomized. For the unexpected impacts, each subject listened to loud music and fabric blindfold was used to cover each participantÕs eyes. Subjects were aware that there would be an impact, but were not advised of timing or impact magnitude. There was no attempt to deceive the subject with a ‘‘surprise’’ impact. In the expected impact, subjects were told the magnitude of the impact in qualitative terms, i.e. a fast impact, and when the impact would happen. The experimental sessions lasted approximately 1 h. 2.4. Data processing A mechanical engineer, who was not a member of this research team, processed the raw accelerometer and video camera data files. The video camera files were processed in such a way that the reflective balls were labelled, missing data were
Table 2 Mean peak linear chair acceleration in the X, Y and Z axes
Acceleration X-axis (m/s2) Acceleration Y-axis (m/s2) Acceleration Z-axis (m/s2) Sample
Fig. 1. Graphic representation of the global reference axes.
Slow impact (SD)
Fast impact (SD)
Slow–fast P value and power
4.54 (0.81)
10.03 (1.81)
2.16 (1.51)
5.79 (3.20)
0.36 (0.14)
1.81 (0.28)
0.001* 0.999 0.001* 0.999 0.001* 0.999
30
60
One way ANOVA. Magnitude of the impacts expressed as mean of the linear acceleration peaks of the sled. P value and power is presented for each analysis. * The mean difference is significant at the .05 level.
1014
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
identified and the data could be read in a text editor. These new files contained the positional data for the three markers to a resolution of 1 mm. The three markers had their position monitored at a sampling rate of 200 Hz for 5 s. No video camera data were recorded prior to the start of the test. The accelerometer board system incorporated a low pass filter with a frequency cut-off of 50 Hz. The raw files from the accelerometer system recorded 10 s: 5 s pre-impact and 5 s post-impact. However, a 750 ms window at a sampling rate of 4000 Hz with 5 s of pre-recorded data was analyzed. Files for the different impacts were coded for blinding. Missing data points and data points presenting a technical error were identified and eliminated. The remaining sample size varied for specific kinematic measurements. The recording systems had as time 0 the firing of the pneumatic cylinder piston, which caused the acceleration of the chair. The timing of the response variables was adjusted so the onset of acceleration of the chair represented time 0 in each impact and for each subject. 2.5. Statistical analysis The data was organized in an Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA) and analyzed using SPSS (version 13.0). Repeated measures (MANOVA) statistical method was used to analyze the response of the variable with an increase of the magnitude of the impact, and its response in the expected and unexpected impacts. Gender was entered into the analysis. The same statistical method was used to compare the timing among the kinematic events of the head. The significance level alpha equal 0.05 was used to determine the level of significance of the data. Raw data of the mean peak magnitude for the linear head acceleration was transformed to natural logarithm in order to obtain a normal distribution and homogenize the variance between impact level and gender.
Fig. 2. Comparison of video camera and accelerometer system recording.
3. Results The video camera and accelerometer data presented a good agreement (Fig. 2). A representative sequence of kinematic events is depicted in Fig. 3. 3.1. Kinematics of the torso 3.1.1. Linear torso acceleration The mean peak magnitude of the linear acceleration for the torso relative to the chair is presented in Table 3. The peak acceleration of the torso was increased approximately 2.5 times in the X-axis and 3 times in the Y-axis from the slow unexpected to fast unexpected
Fig. 3. Representation of timing of head kinematics for one representative subject.
impacts. There were no significant differences in the peak acceleration of the torso in the X and Y axes regarding awareness. Difference by gender in the peak acceleration was not observed in the X-axis (P = 0.108, power = 0.360) and Y-axis (P = 0.285, power = 0.199).
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1015
Table 3 Mean peak of the linear torso acceleration in the X and Y axes Sample
Slow unexpected impact (SD)
Fast unexpected impact (SD)
Fast expected impact (SD)
Slow–fast unexpected P value and power
Fast: unexpected–expected P value and power
Acceleration X-axis (m/s2)
16 M 8F
3.34 (0.45)
8.47 (1.39)
8.33 (1.40)
0.001* 0.999
0.999 0.999
Acceleration Y-axis (m/s2)
16 M 8F
2.29 (0.42)
7.04 (1.68)
7.26 (2.05)
0.001* 0.999
0.999 0.999
Repeated measure test (MANOVA). Linear acceleration expressed as raw data. P value and power is presented for each analysis. * The mean difference is significant at the .05 level.
impact magnitude or expectation; however, the statistical power was low.
3.2. Kinematics of the head Mean peaks of the kinematic events of the head are presented in Table 4. Onset and peak times of the kinematic events are presented in Table 5. Values for females and males are presented separately when a difference by gender was detected. 3.2.1. Angular head displacement Rearward angular head displacement almost doubled with an increase of the impact magnitude. Gender differences were not observed for timing or magnitude of head displacement for any impact. However the statistical power was low for these variables. There was no difference in the magnitude of angular displacement regarding awareness. Statistical power was adequate for this variable. 3.2.2. Angular head acceleration The peak magnitude of rearward angular acceleration approximately doubled and the peak magnitude of forward angular acceleration approximately tripled with higher impact magnitude. No significant differences in peak rearward and forward angular head acceleration regarding expectation were detected. Gender differences for timing or magnitude of angular acceleration were not detected for any impact; however, the statistical power was low. No differences in the onset or peak angular head acceleration times were observed regarding
3.2.3. Linear head acceleration X-axis Rearward and forward linear peak acceleration were approximately 2.5–3.5 greater with increased impact magnitude. No differences in the peaks of the rearward or forward linear accelerations regarding gender were identified; however, the statistical power was low. There were no statistical differences regarding expectation. 3.2.4. Linear head acceleration Y-axis Upward and downward linear peak acceleration was approximately 2–3 times greater with increased magnitude. In the fast-unexpected impact, female subjects presented approximately two times higher upward peak acceleration than male subjects. No difference in the upward or downward peak linear head acceleration regarding awareness was detected. 3.2.5. Timing of head acceleration In the expected fast impact, the rearward peak angular acceleration started approximately 40 ms later than rearward peak linear acceleration (P = 0.0438, power = 0.999). The rearward angular acceleration reached the peak acceleration approximately 76 ms, 62 ms and 84 ms later than the rearward linear acceleration in the slow (P = 0.001, power = 0.977), unexpected-fast
Table 4a Mean peaks of the angular displacement, angular acceleration of the head Peaks
Units
Rearward ang displ Rearward ang acc Forward ang acc
Degrees Rad/s2 Rad/s2
Slow unexpected impact (SD)
Gender P Fast value unexpected power impact (SD)
5.86 (3.61) 0.411 0.125 16.31 (4.40) 0.075 0.432 15.10 (6.22) 0.065 0.461
11.20 (7.71)
Gender P Fast value expected Power Impact (SD)
0.411 0.125 31.51 (11.23) 0.075 0.432 48.45 (28.00) 0.065 0.461
12.83 (5.41)
Sample Gender P Slow–fast Fast: value unexpected P unexpected– power value, Power expected P value, power
0.411 0.125 30.75 (11.88) 0.075 0.432 44.85 (23.65) 0.065 0.461
0.040* 0.996 0.000* 0.999 0.000* 0.999
0.999 0.996 0.999 0.999 0.999 0.999
10M 8F 10 M 9F 10 M 9F
Repeated measures test. Data presented as raw data. P value and power is presented for each analysis. Rearward ang displ refers to rearward angular displacement. Rearward ang acc refers to the rearward angular acceleration. Forward ang acc refers to the forward angular acceleration of the head. * The mean difference is significant at the .05 level.
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
1016
Table 4b Mean peaks of the linear acceleration of the head Peaks
Slow unexpected impact (SD)
Gender P value power
Fast unexpected impact (SD)
Gender P value power
Fast expected impact (SD)
Gender P value power
Slow–fast unexpected P value, power
Fast: unexpected– expected P value, power
Sample
Rearward li acc X-axis Forward li acc X-axis Upward li acc Y-axis
1.55 (0.31)
0.960 0.384 0.412 0.126 0.584 0.210
2.45 (0.42)
0.960 0.384 0.412 0.126 0.033* 0.581
2.44 (0.34)
0.960 0.384 0.412 0.126 0.955 0.210
0.001* 0.999 0.001* 0.999 0.001* 0.999
0.999 0.999 0.112 0.999 0.999 0.999
17 8 17 8 18 10
0.803 0.057
0.001* 0.999
0.644 0.999
18 M 10 F
Downward li acc Y-axis
0.71 (0.35) 0.70 (0.77)
0.71 (0.69)
0.803 0.057
1.94 (0.54) 1.90 1.14 1.63 1.83
(0.94) M (0.66) F (0.92) (0.80)
0.803 0.057
1.70 (0.39) 1.47 (0.64)
1.58 (0.64)
M F M F M F
Linear acceleration of the head was measured in m/s2 and it is expressed in natural logarithm data. Repeated measures test for detecting gender differences in head linear acceleration (power values are included), and response differences between impacts. One way ANOVA analysis for detecting gender differences in the upward linear acceleration Y-axis. M refers to males and F refers to females. P value and power is presented for all comparisons. Rearward li acc refers to rearward linear acceleration. Li acceleration was measured in m/s2. Upward li acc refers to the upward linear acceleration. Downward li acc refers to the downward linear acceleration. * The mean difference is significant at the .05 level.
Table 5 Mean of the onset and peak times of the angular displacement, angular acceleration and linear acceleration of the head
Onset time Rearward ang displ Peak time Rearward ang displ Onset time Rearward ang acc Peak time Rearward ang acc Onset time Forward ang acc Peak time Forward ang acc Onset time Rearward li acc Peak time Rearward li acc Onset time Forward li acc Peak time Forward li acc
Sample
Slow unexpected impact (SD)
Fast unexpected impact (SD)
Fast expected impact (SD)
Slow–fast unexpected P value, power
Fast: unexpected–expected P value and power
10 8 10 8 10 9 10 9 10 9 10 9 17 8 17 8 17 8 17 8
100.7 (63.58) T
121.05 (64.56) T
116.38 (51.54) T
245.27 (79.25)
252.72 (101.38)
266.66 (65.32)
37.84 (19.12)
58.11 (65.15)
63.40 (43.10)
108.71 (21.24)
137.03 (69.39)
124.94 (30.07)
216.92 (87.55)
205.97 (71.46)
200.69 (57.47)
263.85 (87.04)
259.46 (71.60)
252.96 (58.22)
20.57 (56.49)
48.82 (94.14)
16.44 (56.96)
31.68 (60.94)
69.84 (109.09)
27.92 (68.66)
276.64 (172.40)
295.76 (181.62)
309.28 (137.81)
289.52 (163.36)
317.84 (163.34)
326.80 (127.79)
0.433 0.238 0.999 0.114 0.616 0.581 0.244 0.829 0.999 0.084 0.999 0.066 0.564 0.194 0.300 0.269 0.999 0.079 0.999 0.102
0.999 0.238 0.999 0.114 0.999 0.581 0.999 0.829 0.999 0.084 0.999 0.066 0.999 0.194 0.999 0.269 0.999 0.079 0.999 0.102
M F M F M F M F M F M F M F M F M F M F
Repeated measures test. Data presented as raw data. Time was measured in milliseconds. Combined male and female sample. P value and power is presented for each analysis. Ang displ refers to angular displacement, ang acc refers to angular acceleration and li acc refers to linear head acceleration.
(P = 0.029, power = 0.614) and in the expected (P = 0.001, power = 0.999) impacts respectively. Differences in timing of head acceleration were not identified regarding impact speed or expectation; however, most of the statistical powers were low.
4. Discussion The current study analyzed the head kinematics in simulated rear-end impacts. The larger head accelera-
tion observed with an increase in impact magnitude in the present study are in agreement with previous studies (Welcher et al., 2001; Kumar et al., 2000; Kumar et al., 2002; Siegmund et al., 1997). The magnitude of angular displacements reported in the present study are also in agreement with the findings of other studies (Castro et al., 1997; McConnell et al., 1995; McConnell et al., 1993; Siegmund et al., 1997; Szabo et al., 1994; Szabo and Welcher, 1996; Matsushita et al., 1994). The smaller acceleration values reported in the current study compared with previous studies might be due to the lesser
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
impact magnitudes (Siegmund et al., 1997; McConnell et al., 1995). The use of human subjects in simulated rear-end impacts has two limitations: ethics and apprehension. Participants can only be exposed to minimal impacts and the observation of increased head displacement and acceleration with head movement even with a small increase in impact magnitude supports the mechanical role of head movement in whiplash injury. Unfortunately, the minimum impact magnitude at which patients report symptoms cannot be established. Furthermore the threshold of head displacement and head acceleration for injury remains unknown. In the present study, the initial head movement was translation followed by combined rotation and translation and ultimately by rotation. There was an instantaneous rather than fixed axis of head rotation and, therefore, the axis of rotation was not calculated. The linear acceleration of the head is specific to the location in which it was measured. Its value varies across the head depending on distance from the rotational axis. On the other hand, the value of the angular acceleration of the head is valid for any point of the head. Linear and angular acceleration values of the current study presented a significant increase at higher levels of impacts; however, the linear acceleration presented a larger increase, an earlier onset, and reached the peak acceleration faster than the angular acceleration. Therefore, it might be more appropriate to analyze both types of acceleration rather than assume the linear acceleration as the key movement of the head. Vertical linear peak head acceleration was lower in females. The findings of Siegmund et al. (2003) also reported similar gender differences in head accelerations. Kumar et al. (2000) reported gender differences in rearward linear head acceleration. Differences in vertical acceleration may be a function of the car seat relationship to the subjectÕs body. Males are generally taller than females and the contact with the back against the car seat in males may result in more upward movement. Greater angular head acceleration in females would increase the risk of cervical injury. The present study did not identify gender differences for horizontal or angular head accelerations and therefore increased risk of mechanical injury due to head movement in females was not supported. The prevalence of pain complaints in female subjects may be a function of muscle injury, or secondary to centrally mediated pain modulation. Awareness of impact may affect kinematics of head movement in a simulated rear end impact. Awareness of an event refers to the anticipation of such an event, and implies three dimensions: temporal, event and amplitude awareness (Siegmund, 2001; Frank, 1986). Temporal awareness refers to whether the subject knows about the exact timing in which an event will occur, event awareness describes whether the subject knows
1017
an event will occur, and amplitude awareness refers to the awareness of the amplitude of an imminent event (Siegmund, 2001). Kumar et al. (2000) reported that awareness resulted in 30% reduction in peak backward linear head acceleration. The current findings did not report significant differences in kinematics of head movement regarding temporal awareness. Siegmund et al. (2003) reported that lack of event awareness significantly affected kinematic gender response; therefore, they concluded that event awareness is the component of anticipation event that plays an influential role in the kinematics of the head and neck in a simulated whiplash event. The unexpected events in the present study are similar to the ‘‘unalerted’’ group in the Siegmund et al. (2003) study. Subjects were aware that they would eventually sustain impact and therefore had event awareness. To remove event awareness requires deception such as used by Siegmund et al. (2003). The expected impact events in the present study are similar to the ‘‘alerted’’ group in the Siegmund et al. (2003) study. The Kumar et al. (2000) study tested expectancy based on the subject being advised of timing and magnitude of impact. There was no record of deception for the unexpected impact. The paper does not state how long the subject was kept blindfolded with loud music prior to the impact. It seems likely; however, that there was event awareness but not temporal or magnitude awareness. A possible explanation for the differences reported in these studies and the present study may relate to the subjects being informed of the magnitude of impact. The present study utilized only one ‘‘alerted’’ impact magnitude. It is possible that the subjects did not perceive the simulated impact sufficiently noxious to get injured. This study presents some limitations. It is possible that the mechanism by which the accelerometer board was attached to the subject created an additional sensory stimulus that might have altered the head motion. The distance between the head of the participant and the headrest was not recorded, and might account for the gender differences reported in this study. The restriction of the kinematics analysis to the first 750 ms after impact did not allow insight regarding angular head displacement in the rebound phase. The low magnitude of impacts in this study might have not been sufficient to create an actual simulation of a whiplash event; and therefore, to be perceived sufficiently noxious by the participants in the expected condition.
5. Interpretation Based on the findings of the current study, the following can be concluded: There is increased angular head acceleration, increased linear head acceleration and increased angular head displacement with an increased
1018
I.A. Herna´ndez et al. / Clinical Biomechanics 20 (2005) 1011–1018
impact magnitude. Females have less upward head movement than males. Temporal and amplitude awareness of a simulated impact do not produce differences in the magnitude of head kinematic events. Linear and angular magnitude peak accelerations as well as onset and peak time accelerations are not coincident. Conclusions regarding timing in kinematics of head movement cannot be drawn due to the low statistical power in most of the kinematic events. Acknowledgements The authors would like to thank Dr. S. Kumar for use of his lab, equipment and technical support. This research was supported by the University of Alberta, Fund for Dentistry Grant # 2002-01 and the McIntyre Memorial Research Fund. References Brault, J.R., Siegmund, G.P., Wheeler, J.B., 2000. Cervical muscle response during whiplash: evidence of a lengthening muscle contraction. Clin. Biomech. 15, 426–435. Castro, W.H., Schilgen, M., Meyer, S., Weber, M., Peuker, C., Wortler, K., 1997. Do ‘‘whiplash injuries’’ occur in low-speed rear impacts? Eur. Spine J. 6, 366–375. Frank, J.S., 1986. Spinal motor preparation in humans. Electroencephalogr. Clin. Neurophysiol. 63, 361–370. Kumar, S., Narayan, Y., Amell, T., 2000. Role of awareness in head– neck acceleration in low velocity rear-end impacts. Accid. Anal. Prev. 32, 233–241. Kumar, S., Narayan, Y., Amell, T., 2002. An electromyographic study of low-velocity rear-end impacts. Spine 27, 1044–1055. Matsushita, T., Sato, T., Hirabayashi, K., Fujimora, S., Asazuma, T., Takatori, T., 1994. Paper 942208 X-ray study of the human neck motion due to head inertia loading. In: 38th Stapp Car Crash Conference. Warrendale, PA, USA: Society of Automotive Engineers, Inc., pp. 55–64. McConnell, W., Howard, R.P., Guzman, H.M., et al., 1993. Paper 930889 Analysis of human test subject kinematic responses to low
velocity rear-end impacts. Presentation to SAE International Congress and Exposition. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 21–30. McConnell, W., Howard, R.P., Poppel, J.V., et al., 1995. Paper 952724 Human head and neck kinematics after low velocity rear-end impacts—Understanding ‘‘whiplash. In: 39th Stapp Car Crash Conference. Warrendale, PA, USA: Society Automotive Engineers, Inc., pp. 3106–3129. Ono, K., Kaneoka, K., Wittek, A., Kazjer, J., 1997. Paper 973340 Cervical injury mechanism based on the analysis of human cervical vertebral motion and head–neck–torso kinematics during low speed rear impacts. In: 41st Annual Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 339–356. Severy, D.M., Mathewson, J.H., Bechtol, C.O., 1955. Controlled automobile rear-end collisions, an investigation of related engineering and medical phenomena. Can. Serv. Med. J. 11, 727–759. Siegmund, G., 2001. The reflex response of human neck muscles to whiplash-like perturbations [Doctoral Thesis]. Vancouver, British Columbia, Canada: University of British Columbia. Siegmund, G.P., King, D.J., Lawrence, J.M., Wheeler, J.B., Brault, J.R., Smith, T.A., 1997. Paper 973341 Head/neck kinematic response of human subjects in low-speed rear-end collisions. In: 41st Annual Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society Automotive Engineers, Inc., pp. 357–385. Siegmund, G.P., Sanderson, D.J., Myers, B.S., Inglis, J.T., 2003. Awareness affects the response of human subjects exposed to a single whiplash-like perturbation. Spine 28, 671–679. Szabo, T.J., Welcher, J.B., 1996. Paper 962432 Human subject kinematics and electromyographic activity during low speed rear impacts. In: 40th Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 295–314. Szabo, T., Welcher, J., Anderson, R., et al., 1994. Paper 940532 Human occupant kinematic response to low speed rear-end impacts. In: 38th Stapp Car Crash Conference. Warrendale, Pennsylvania, USA: Society of Automotive Engineers, Inc., pp. 23–36. Viano, D., 2003. Seat influences on female neck responses in rear crashes: A reason why women have higher whiplash rates. Traffic Inj. Prev. 4, 228–239. Welcher, J., Szabo, T., Voss, D., 2001. 2001-01-0899 Human Occupant Motion in Rear-end Impacts: Effects of incremental increases in velocity change. SAE 2001 World Congress. Warrendale, Pennsylvania: Society of Automotive Engineers, Inc.