The influence of head restraint and occupant factors on peak head/neck kinematics in low-speed rear-end collisions

https://doi.org/10.1016/S0001-4575(98)00077-3Get rights and content

Abstract

Prior two-way analyses of variance showed that the peak kinematic response of the head and neck of subjects exposed to low-speed rear-end collisions was related to speed change and gender, however potential reasons for this gender dependence were not determined. Using multiple linear regression, this study further examined these response data to determine the relative influence of specific factors, including subject anthropometry, neck strength, cervical range of motion, seated posture and head restraint position, which may have been responsible for the previously-observed gender dependence. The results of this analysis showed that vehicle speed change and relative head restraint position explained the largest proportion of the observed variation in peak occupant kinematic response. Seated posture measures also explained some of the variation in kinematic response. The current analysis prioritizes which variables to explore more thoroughly in future research and which variables should be carefully controlled in future studies.

Introduction

After a low-speed rear-end automobile impact, occupants sometimes complain of symptoms commonly referred to as whiplash-associated disorders (WAD) (Spitzer et al., 1995). The exact mechanisms producing WAD remain unclear, although numerous factors that may influence the incidence or duration of WAD have been reported. These factors are generally grouped into three categories: seat factors, occupant factors and external factors (Viano and Gargan, 1995). Seat factors include seat and head restraint geometry, stiffness, strength and inclination; occupant factors include gender, anthropometry, seated posture and preparedness; and external factors include vehicle mass, vehicle stiffness, bumper design and collision severity (States and Balcerak, 1973, Foret-Bruno et al., 1991, Viano and Gargan, 1995).

Many studies have attempted to draw a direct link between the production or duration of WAD and specific seat, occupant, or external factors (see A in Fig. 1). With seat factors, for instance, a higher head restraint has been shown to reduce the incidence of neck injury (Nygren et al., 1985), and increased head restraint backset, defined as the horizontal gap between the back of the head and the front surface of the head restraint, has correlated significantly with increased neck symptom duration (Olsson et al., 1990). With occupant factors, a higher incidence of WAD has been reported in females than males (O’Neill et al., 1972, Balla, 1980, Kahane, 1982, Lövsund et al., 1988, Otremski et al., 1989), and other occupant factors such as height (Carlsson et al., 1985), age (Otremski et al., 1989), preparedness and pre-impact posture (Sturzenegger et al., 1994), and seat belt use (Otremski et al., 1989, Maag et al., 1990) have also been observed to affect the incidence or duration of WAD. External factors such as collision speed change have been associated with initial measures of neck strain severity (Ryan et al., 1993).

Fewer studies have addressed either the relationship between the seat, occupant, or external factors and the physical responses of occupants (see B in Fig. 1) or the relationship between the physical and clinical responses (see C in Fig. 1). Seat factors have been shown to have a larger effect on occupant kinematics than vehicle construction (an external factor) at a given collision speed change (Haland et al., 1996), and specific seat back and head restraint modifications have been shown to affect the kinematic and kinetic response of anthropomorphic test devices (ATD’s) (Svensson et al., 1996). Occupant kinematic response has also been shown to increase with collision speed change (Svensson et al., 1996, Siegmund et al., 1997). There is, however, limited information regarding the influence of occupant anthropometry, physiology and seated posture on the physical response of vehicle occupants.

In this paper the relationship of some occupant and seat factors to the peak kinematic response data of 42 human subjects exposed to low-speed rear-end impacts was examined using multiple linear regression. A previous two-way analysis of variance (ANOVA) of these kinematic data found that peak amplitude and time to peak amplitude of some sagittal-plane kinematic response parameters varied significantly with gender (Siegmund et al., 1997), however, specific gender-based factors responsible for this correlation were not identified. This study examined inter-subject differences in anthropometry, neck strength and cervical range of motion to identify occupant factors that may be responsible for the previously-observed gender differences in peak kinematic response. Potentially confounding variables such as head restraint position (backset and height) and seated posture of the occupant immediately before impact were also incorporated into the analysis.

Section snippets

Methods

Predictors used in the regression analyses are designated in capital letters where they are defined.

Results

Data from 21 male and 21 female subjects were analyzed. Three subjects (1M, 2F) withdrew between their 4 and 8 km/h tests, yielding data from 81 collisions. High speed video data (position and angle data) were successfully acquired from all tests, whereas occasional sensor problems yielded 75 complete data sets of acceleration and velocity data.

A total of 31 response peaks common to all subjects were observed (Fig. 3). This produced 31 responses for peak amplitude (PA) and 31 responses for time

Discussion

The relative influence of 22 selected occupant, seat and external factors on 31 kinematic response peaks obtained from 42 human subjects exposed to 81 low-speed rear-end automobile collisions has been assessed using multiple linear regression. The purpose of this analysis was to potentially identify the occupant-related factors responsible for the previously observed gender-dependency of peak kinematic responses using a two-way ANOVA (Siegmund et al., 1997). Overall, the previous correlation

Acknowledgements

Partial funding for this project was received under the Technology BC program administered by the Science Council of British Columbia.

References (34)

  • G.M. Furnival et al.

    Regression by leaps and bounds

    Technometrics

    (1974)
  • Haland, Y., Lindh, F., Fredriksson, R., Svensson, M., 1996. The influence of the car body and the seat on the loading...
  • Kahane, C.J., 1982. An evaluation of head restraints-Federal motor vehicle safety standard 202 (DOT HS-806 108)....
  • J.M. Lawrence et al.

    Measuring head restraint force and point of application during low-speed rear-end automobile collisions (970397)

  • Lövsund, P., Nygren, A., Salen, B., Tingvall, C., 1988. Neck injuries in rear end collisions among front and rear seat...
  • Maag, U., Desjardins, D., Bourbeau, R., Laberge-Nadeau, C., 1990. Seat belts and neck injuries. Proceedings of 1990...
  • Najjar, M.F., Rowland, M., 1987. Anthropometric Reference Data and Prevalence of Overweight, United States, 1976-80,...
  • Cited by (28)

    • Whiplash injury prevention with active head restraint

      2009, Clinical Biomechanics
      Citation Excerpt :

      To reduce injurious head and neck loads and motions, the active head restraint rotates forward based on occupant momentum pressing into the seatback during whiplash (Viano and Olsen, 2001). Previous biomechanical studies have evaluated the effects of head restraint position on head and neck loads and motions during simulated whiplash of volunteers (Siegmund et al., 1999), whole cadavers (Geigl et al., 1994), anthropometric test dummies (Svensson et al., 1996; Szabo et al., 2003), and computer models (Eriksson, 2004; Szabo and Welcher, 1996; Tencer et al., 2002; Tencer et al., 2001). Volunteer and cadaver experiments have shown that standard, non-active automobile seats and head restraints were insufficient to limit excessive head motions (Geigl et al., 1994).

    • Risk of injury of the cervical spine in sled tests in female volunteers

      2007, Clinical Biomechanics
      Citation Excerpt :

      Further in vivo data, especially with regard to chronological occurrence, are not available for comparison at this time. In the literature also the head-to-head restraint contact occurs between 100–120 ms (Matsushita et al., 1994; Mühlbauer et al., 1999; Siegmund et al., 1999). The analysis of the acceleration and motion data shows that even with optimum vertical positioning of the head restraint the extension motion is not adequately reduced.

    • The effect of collision pulse properties on seven proposed whiplash injury criteria

      2005, Accident Analysis and Prevention
      Citation Excerpt :

      These findings suggest that there may be a measure based on vehicle speed change that predicts whiplash injury, though further work is required. The general applicability of the current results is limited by the use of a single seat, a single seated posture and a single head restraint position – all factors known to affect occupant response (Håland et al., 1996; Svensson et al., 1996; Siegmund et al., 1999; Boström et al., 2000; Welcher and Szabo, 2001). Since many automobile seats are alike in construction and have monotonic force/deflection curves (up to failure, at least), the general pattern of results observed here are anticipated to apply to seats other than the one used in these tests.

    View all citing articles on Scopus
    View full text