- Email: [email protected]
Effects of trunk exertion force and direction on postural control of the trunk during unstable sitting
Available online at www.sciencedirect.com
Clinical Biomechanics 23 (2008) 505–509 www.elsevier.com/locate/clinbiomech
Effects of trunk exertion force and direction on postural control of the trunk during unstable sitting HyunWook Lee a,*, Kevin P. Granata a,b, , Michael L. Madigan a,b b
a School of Biomedical Engineering and Sciences, Virginia Polytechnic Institute and State University, Mail Code 0298, Blacksburg, VA 24061, USA Department of Engineering Science and Mechanics, Virginia Polytechnic Institute and State University, Mail Code 0298, Blacksburg, VA 24061, USA
Received 3 September 2007; accepted 4 January 2008
Abstract Background. Pushing and pulling exertions have been implicated as risk factors of low-back disorders. In an attempt to investigate the mechanisms by which pushing and pulling influence risk for low-back disorders, the goal of this study was to investigate the effects of trunk exertion force and exertion direction on postural control of the trunk during unstable sitting. Methods. Seat movements were recorded while subjects maintained a seated posture on a wobbly chair against different exertion forces (0N, 40N, and 80N) and exertion directions (trunk flexion and extension). Postural control of the trunk was assessed from kinematic variability (root-mean-squared amplitude and 95% ellipse area) and non-linear stability analyses (stability diffusion exponent and maximum finite-time Lyapunov exponent). Findings. Kinematic variability and non-linear stability estimates increased as exertion force increased including root-mean-squared amplitude (P < 0.001), 95% ellipse area (P < 0.001), stability diffusion exponent (P = 0.042), and maximum finite-time Lyapunov exponent (P < 0.001). A subset of measures indicated postural control of the trunk was poorer during flexion exertions compared to extension exertions including root-mean-squared amplitude (P < 0.001), 95% ellipse area (P = 0.046), and maximum finite-time Lyapunov exponent (P = 0.002). Interpretation. Trunk exertion force and exertion direction affect postural control of the trunk. This study may aid in understanding how pushing and pulling exertions can potentially contribute to low-back disorders. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Trunk; Flexion; Extension; Low-back; Co-activation
1. Introduction Repetitive lifting has been reported as a significant risk factor of low-back disorders in the occupational setting (Andersson, 1981; Marras et al., 1995). As such, industry is rapidly modifying the workspace to reduce lifting, and frequently replacing it with pushing and pulling exertions. However, pushing and pulling exertions are also associated with low-back disorders. In the United States, Canada, and the United Kingdom, 20% of industrial low-back inju*
Corresponding author. E-mail address: [email protected] (H. Lee). Deceased.
0268-0033/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2008.01.003
ries have been attributed to pushing or pulling activities (Damkot et al., 1984; Hoozemans et al., 1998). This injury rate is expected to increase with the trend toward a growing number of push and pull-related exertions in the workplace. Despite the fact that 50% of industrial manual materials handling tasks include pushing and pulling exertions (Baril-Gingras and Lortie, 1995), the biomechanics and neuromuscular control of pushing and pulling exertions remain poorly understood (Schibye et al., 2001; van der Beek et al., 1999). Pushing and pulling exertions are associated with trunk flexion and extension exertions. These trunk exertions may influence postural control of the trunk, and any changes in postural control of the trunk could potentially have a role
506
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
in low-back injuries incurred while pushing or pulling. In support of this hypothesis, Radebold et al. (2001) reported poorer postural control of the trunk in patients with lowback pain compared to healthy controls. However, to our knowledge no studies have investigated the effects of trunk exertions on trunk postural control. Trunk exertion force and direction has the potential to influence trunk postural control. Increasing the force level of trunk flexion or extension exertions increases agonistantagonist co-activation of the trunk flexors and extensors (Lee et al., 2007; Granata et al., 2005). While increased coactivation increases trunk stiffness (Lee et al., 2006) and may be beneficial for injury prevention during some tasks, it may have a deleterious effect on postural control of the trunk. Reeves et al. (2006), for example, reported a degradation of trunk postural control when participants purposefully increased trunk co-activation during unstable sitting. Trunk exertion direction also as the potential to influence trunk postural control. Greater trunk co-activation occurs during trunk flexion exertions compared to extension exertions (Lee et al., 2007; Granata et al., 2005). As noted above, increases in trunk co-activation may impair trunk postural control. Based on our lack of understanding of the effects of trunk exertions on postural control of the trunk, the goal of this study was to investigate the effects of trunk exertion force and direction on postural control of the trunk during unstable sitting. This information could be useful in understanding the etiology of low-back disorders, and potentially aid in the development of manual material handling tasks to help minimize the risk for low-back disorders. 2. Methods Twelve healthy adults participated including 7 males (age 25.7 (6.9) years, height 178.7 (6.9) cm, mass 79.9 (9.0) kg), and 5 females (age 21.4 (1.7) years, height 161.3 (8.2) cm, mass 59.6 (8.3) kg). Participants reported no history of chronic low-back pain, and provided informed consent approved by the Institutional Review Board at Virginia Tech prior to participation. Postural control of the trunk was assessed while participants sat on a wobbly chair (Cholewicki et al., 2000). The seat pan of the chair pivoted freely in the sagittal and frontal planes (Fig. 1). Destabilizing moments arise from gravitational effects associated with small angle deviations from the neutral position (i.e. where the chair-subject system center of mass was positioned directly above the wobbly chair pivot). Adjustable springs with an effective rotational stiffness, kh, applied an elastic restorative moment to the seat to return the seat pan to the neutral position. This stiffness was set at kh = 50% rM Grav where rM Grav was the linearized gradient of gravitational moment recorded in static calibration measurements. Hence, the postural control assessment accounted for differences in subject anthropometry that were intrinsically embedded in rM Grav . A stiffness value of 100% rM Grav would mean the destabilizing
Fig. 1. Wobbly chair design adapted from Cholewicki et al. (2000). An isotonic anteriorly-directed force was applied to the trunk during extension exertion trials, and an isotonic posteriorly-directed force was applied to the trunk during flexion exertion trials. h1 = angle of seat relative to horizontal; h2 = angle of trunk relative to vertical.
gravitational moment in wobbly chair positions other than neutral would be completely offset by spring forces. Since the stabilizing restorative stiffness was less than the destabilizing gravitational moment, the biomechanical system was inherently unstable. Therefore, trunk movement was necessary to correct for disturbances in seat angle from neutral, and to maintain an upright seated posture. A foot support attached to the seat pan supported the lower limbs so that the upright seated posture was maintained predominantly by movement at the lumbar spine. External horizontal isotonic forces were applied in the median plane at the T8 level of the torso. These forces required voluntary trunk extension and flexion moments to maintain an upright posture while seated on the wobbly chair. Forces of 0, 40 and 80N were generated by tensioning elastic bands between a chest harness worn by the participants and a distant wall of the laboratory. The length of the elastic bands was 15 m, thereby assuring that change in length of the band from torso movements was small relative to the band length, i.e. forces were nearly isotonic. The externally-applied isotonic forces caused a tipping moment about the pivot point under the seat pan. To offset this tipping moment, the seat pan was adjusted anteriorly or posteriorly to achieve steady-state equilibrium under the applied load, i.e. the static gravitational moment balanced the externally-applied forces. Note, however, that the system remained unstable, as described above, due to destabilizing moments associated with gravity. Six experimental conditions were tested using a full-factorial design with three exertion forces (0, 40, and 80N) and two exertion directions (trunk flexion and trunk extension). Each subject was instructed to maintain seated balance on a wobble chair with arms-crossed over their chest and to look straight with eyes opened. The presentation order of exertion direction was counter-balanced so that half of the participants performed all flexion exertions first, and the other half performed all extension exertions first. Within each exertion direction, the presentation order of the 40N and 80N conditions was also counter-balanced.
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
A single 0N force condition was always tested first. Subjects completed a block of five trials of 60 s duration in each experimental condition with one minute of rest between trials. Prior to the test protocol, subjects performed five practice trials in each experimental condition to become accustomed to the procedure. Seat angle in both the anterior-posterior and medial-lateral planes was sampled at 100 Hz during all trials using an electromagnetic motion sensor (Ascension Technology, Burlington, VT). Elastic band force was also sampled at 100 Hz using a uniaxial load cell (Interface Inc., Scottsdale, AZ). Seat angle and elastic band force data were then lowpass filtered using a 6 Hz (4th-order Butterworth filter). Seat angle data were de-meaned so that the mean value of each trial was zero. Postural control of the trunk was quantified using summary statistics of seat angle variance, and by non-linear stability analyses. Summary statistics of seat angle included: root-mean-squared (RMS) amplitude of the seat angle, and 95% ellipse area (EA). The RMS amplitude was calculated using the total recti-linear distance from the mean position because this includes the dynamic interaction between anterior-posterior and medial-lateral movements. The 95% EA represents the region of typical movement and was computed using methods reported by Prieto et al. (1996). Non-linear stability analyses included the stabilogram diffusion analysis (SDA) and the maximum finite-time Lyapunov exponent, kmax. SDA analyses compute variance from the mean squared distances traveled hDd 2i iDt by the signal in time interval Dt (Collins and DeLuca, 1993). Cholewicki et al. (2000) showed that the stability diffusion exponent, H, can be computed from the log–log slope of hDd 2i iDt versus Dt = 0.01–2.0 s, and that this exponent demonstrated bi-linear behavior, i.e. short-term and long-term stability exponents. Only the short-term exponent has been shown to discriminate between subject groups with and without low-back pain (Radebold et al., 2001). Therefore, only the short-term exponent was used in the present analyses. Values of H < 0.5 indicate that perturbations are likely to be controlled and attracted toward the equilibrium state, whereas values of H > 0.5 indicate a disturbance will continue to grow toward infinity. H was calculated by using radial direction angle data because of same reason as RMS amplitude. Lyapunov exponents quantify the mean rate of divergence with respect to an equilibrium state. The dynamic state at each instant in time was described by a fourdimensional vector including angle and velocity in the anterior–posterior and medial–lateral planes, ~ qðtÞ ¼ ½hSx ; hSy ; h_ Sx ; h_ Sy . For every data point ~ qðtÞ in the time series we recorded the mean distance, dðtÞ, to three data points that were nearest to it (Kantz and Schreiber, 2004). In an asymptotically stable system, the distance between these points must grow smaller with time. Preliminary analyses demonstrated that averaging the three nearest neighbors reduced the effects of stochastic noise, whereas more than
507
three failed to improve signal quality. Stability was quanti changes over a finitefied from the rate at this which dðtÞ time period Dt = 0.01–0.2 s, i.e. from d(t) to d(t + Dt) averaged over all time samples hdðt þ DtÞ=dðtÞit ¼ ekMax Dt
ð3Þ
An increase in H and kmax imply impaired postural control of the trunk. A detailed description of these methods is available elsewhere (Wolf et al., 1985; Rosenstein et al., 1993). Prior to the statistical analyses, postural control measures calculated from all five trials in each experimental condition were averaged. A repeated-measures ANOVA was used to investigate the effects of exertion force (0, 40, and 80N), and exertion direction (trunk flexion and extension) on RMS, EA, H and kmax. Following significant main effects or interactions, pairwise comparisons were performed using the Tukey HSD. Significance was determined at the level of P < 0.05. 3. Results All dependent variables increased as exertion force increased (Fig. 2). Compared to 0N exertions, the RMS and EA were 14.3% and 17.7% higher during 40N exertions, and 33.0% and 46.0% higher during 80N exertions, respectively (P < 0.001 for RMS and P < 0.001 for EA). Similarly, H and kmax were 0.2% and 7.7% higher during 40N exertions, and 0.6% and 11.2% higher during 80N exertions, respectively, compared to 0N exertions (P = 0.042 for H and P < 0.001 for kmax). Flexion exertions generally exhibited higher values of the dependent variables compared to extension exertions (Fig. 2). RMS, EA, and kmax were 29.1% (P < 0.001), 7.9% (P = 0.046) and 5.8% (P = 0.002) higher, respectively, during flexion exertions, while H (P = 0.56) was not different. Both RMS and kmax exhibited a significant force direction interaction. Post-hoc analyses revealed RMS was higher during the flexion exertions than extension exertions at both the 40N and 80N force levels, and kmax was higher during the flexion exertions and extension exertions at the 40N exertion force level only. 4. Discussion The goal of this study was to investigate the effects of trunk exertion force and direction on trunk postural control during unstable sitting. The dependent variables used were based on variability and non-linear stability analyses of seat angle. Increases in these dependent variables implied a degradation of trunk postural control. Results, in general, indicated a degradation of trunk postural control as exertion force increased, and that trunk postural control was poorer during trunk flexion exertions compared to trunk extension exertions. A quantitative comparison between the results here and other studies investigating trunk postural control on a during unstable sitting (e.g. Radebold et al., 2001; Cholewicki
508
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
Fig. 2. Summary statistics of seat angle variance (RMS, EA) and non-linear stability measures were affected by exertion force and direction. *: a significant difference from 0N force. : a significant difference from 40N force. à: a significant difference between flexion and extension. #: main effect of exertion force. €: main effect of exertion direction.
et al., 2000) was not possible to due subtle differences in wobbly chair design and the fact that the postural control measures used here were based on angular displacement of the chair while these other studies were based on center of pressure. However, the mean values of H reported here (0.76–0.77) were comparable to these other studies (0.56– 0.83). The degradation of trunk postural control with increasing exertion force found here appears to be consistent with other studies. For example, Reeves et al. (2006) reported a degradation of postural control during unstable sitting with elevated trunk co-activation. While the current study did not manipulate trunk co-activation per se, the increase in exertion force investigated here has been shown to increase trunk co-activation (Lee et al., 2007; Granata et al., 2005). This increase in trunk co-activation can degrade trunk postural control because increased activation levels of muscles are associated with increased variability in force output (Christou et al., 2002; Hamilton et al., 2004). Results also indicated trunk postural control was poorer during flexion exertions compared to extension exertions. Co-activation during trunk flexion exertions has been reported to be approximately twice the level of co-activation during equivalent extension exertions (Granata et al., 2005). Knowing that increased activation levels are associated with increased variability in force output (Christou et al., 2002; Hamilton et al., 2004), these results appear to be consistent with other findings. Three limitations of this study warrant discussion. First, actual pushing and pulling tasks were not investigated even
though the main motivation for this study was the association of pushing and pulling exertions with low-back disorders. Using controlled trunk flexion and extension exertions allowed for easier experimental control of trunk exertions levels, and avoided potential variability between subjects in upper extremity performance. Second, although differences in trunk postural control during unstable sitting have been identified between individual with and without low-back pain (Radebold et al., 2001), a clear relation between trunk postural control and actual spinal pain/ injury has not been established. Third, EMG data were not included in our analysis, so it was only possible to hypothesize changes in co-activation when attempting to explain the results. In conclusion, increasing trunk exertion force degraded trunk postural control during unstable sitting, and trunk flexion exertions exhibited poorer trunk postural control compared to trunk extension exertions. This study may aid in understanding how pushing and pulling exertions can potentially contribute to low-back disorders. Further analyses are necessary to investigate trunk postural control during occupational tasks involving pushing and pulling. Acknowledgements This work is dedicated to Dr. Kevin P. Granata, who was tragically killed on April 16, 2007 at Virginia Tech. This study was supported by a grant R01 AR49923-01 from CDC of the National Institute for Occupational
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
Safety and Health. We wish to thank Steve Hanson and Emmanuel Ekwueme for their assistance in collecting data. References Andersson, G.B.J., 1981. Epidemiologic aspects on low back pain in industry. Spine 6, 53–60. Baril-Gingras, G., Lortie, M., 1995. The handling of objects other than boxes: univariate analysis of handling techniques in a large transport company. Ergonomics 38, 905–925. Cholewicki, J., Polzhofer, G.A., Radebold, A., 2000. Postural control of trunk during unstable sitting. J. Biomechanics 22, 1733–1737. Christou, E.A., Grossman, M., Carlton, L.G., 2002. Modeling variability of force during isometric contractions of the quadriceps femoris. J. Mot. Behav. 34, 67–81. Collins, J.J., DeLuca, C.J., 1993. Open-loop and closed-loop control of posture: a random-walk analysis of center of pressure. Exp. Brain Res. 96, 308–318. Damkot, D.K., Pope, M.H., Lord, J., Frymoyer, J.W., 1984. The relationship between work history, work environment and low-back pain in men. Spine 9, 395–399. Granata, K.P., Lee, P.E., Franklin, T.C., 2005. Co-contraction recruitment and spinal load during isometric trunk flexion and extension. Clin. Biomech. 20, 1029–1037. Hamilton, A.F., Jones, K.E., Wolpert, D.M., 2004. The scaling of motor noise with muscle strength and motor unit number in humans. Exp. Brain Res. 157, 417–430. Hoozemans, M.J., van der Beek, A.J., Frings-Dresen, M.H., van Dijk, F.J., van der Woude, L.H., 1998. Pushing and pulling in relation to musculoskeletal disorders: a review of risk factors. Ergonomics 41, 757–781. Kantz, H., Schreiber, T., 2004. Nonlinear Time Series Analyses, Second ed. Cambridge University Press, NY.
509
Lee, P.J., Rogers, E.L., Granata, K.P., 2006. Active trunk stiffness increases with co-contraction. J Electromyogr. Kines. 16, 51–57. Lee, P.J., Granata, K.P., Moorhouse, K.M., 2007. Active trunk stiffness during voluntary isometric flexion and extension exertions. Hum. Factors 49, 100–109. Marras, W.S., Lavender, S.A., Leurgans, S., Fathallah, F.A., Allread, W.G., Ferguson, S.A., Ferguson, S.A., Allread, W.G., Rajulu, S.L., 1995. Biomechanical risk factors for occupationally related low back disorder risk. Ergonomics 38, 377–410. Prieto, T.E., Myklebust, J.B., Hoffmann, R.G., Lovett, E.G., Myklebust, B.M., 1996. Measures of postural steadiness: differences between healthy young and elderly adults. IEEE Trans. Biomed. Eng 43, 956– 966. Radebold, A., Cholewicki, J., Polzhofer, G.A., Green, T.P., 2001. Impaired postural control of the lumbar spine is associated with delayed muscle response times in patients with chronic idiopathic low back pain. Spine 26, 724–730. Reeves, N.P., Everding, V.Q., Cholewicki, J., 2006. The effect of trunk stiffness on postural control during unstable seated balance. Exp. Brain Res. 174, 694–700. Rosenstein, M.T., Collins, J.J., DeLuca, C.J., 1993. A practical method for calculating largest Lyapunov exponents from small data sets. Physica D 65, 134. Schibye, B., Sogaard, K., Martinsen, D., Klausen, K., 2001. Mechanical load on the low back and shoulders during pushing and pulling of twowheeled waste containers compared with lifting and carrying of bags and bins. Clin. Biomech. 16, 549–559. Van der Beek, A.J., Hoozemans, M.J.M., Frings-Dresen, M.H., Burdorf, A., 1999. Assessment of exposure to pushing and pulling in epidemiological field studies: an overview of methods, exposure measures, and measurement strategies. Int. J. Ind. Ergon. 24, 417–429. Wolf, A., Swift, J.B., Swinney, H.L., Vastano, J.A., 1985. Determining Lyapunov exponents from a time-series. Physica D 16, 285–317.
Clinical Biomechanics 23 (2008) 505–509 www.elsevier.com/locate/clinbiomech
Effects of trunk exertion force and direction on postural control of the trunk during unstable sitting HyunWook Lee a,*, Kevin P. Granata a,b, , Michael L. Madigan a,b b
a School of Biomedical Engineering and Sciences, Virginia Polytechnic Institute and State University, Mail Code 0298, Blacksburg, VA 24061, USA Department of Engineering Science and Mechanics, Virginia Polytechnic Institute and State University, Mail Code 0298, Blacksburg, VA 24061, USA
Received 3 September 2007; accepted 4 January 2008
Abstract Background. Pushing and pulling exertions have been implicated as risk factors of low-back disorders. In an attempt to investigate the mechanisms by which pushing and pulling influence risk for low-back disorders, the goal of this study was to investigate the effects of trunk exertion force and exertion direction on postural control of the trunk during unstable sitting. Methods. Seat movements were recorded while subjects maintained a seated posture on a wobbly chair against different exertion forces (0N, 40N, and 80N) and exertion directions (trunk flexion and extension). Postural control of the trunk was assessed from kinematic variability (root-mean-squared amplitude and 95% ellipse area) and non-linear stability analyses (stability diffusion exponent and maximum finite-time Lyapunov exponent). Findings. Kinematic variability and non-linear stability estimates increased as exertion force increased including root-mean-squared amplitude (P < 0.001), 95% ellipse area (P < 0.001), stability diffusion exponent (P = 0.042), and maximum finite-time Lyapunov exponent (P < 0.001). A subset of measures indicated postural control of the trunk was poorer during flexion exertions compared to extension exertions including root-mean-squared amplitude (P < 0.001), 95% ellipse area (P = 0.046), and maximum finite-time Lyapunov exponent (P = 0.002). Interpretation. Trunk exertion force and exertion direction affect postural control of the trunk. This study may aid in understanding how pushing and pulling exertions can potentially contribute to low-back disorders. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Trunk; Flexion; Extension; Low-back; Co-activation
1. Introduction Repetitive lifting has been reported as a significant risk factor of low-back disorders in the occupational setting (Andersson, 1981; Marras et al., 1995). As such, industry is rapidly modifying the workspace to reduce lifting, and frequently replacing it with pushing and pulling exertions. However, pushing and pulling exertions are also associated with low-back disorders. In the United States, Canada, and the United Kingdom, 20% of industrial low-back inju*
Corresponding author. E-mail address: [email protected] (H. Lee). Deceased.
0268-0033/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2008.01.003
ries have been attributed to pushing or pulling activities (Damkot et al., 1984; Hoozemans et al., 1998). This injury rate is expected to increase with the trend toward a growing number of push and pull-related exertions in the workplace. Despite the fact that 50% of industrial manual materials handling tasks include pushing and pulling exertions (Baril-Gingras and Lortie, 1995), the biomechanics and neuromuscular control of pushing and pulling exertions remain poorly understood (Schibye et al., 2001; van der Beek et al., 1999). Pushing and pulling exertions are associated with trunk flexion and extension exertions. These trunk exertions may influence postural control of the trunk, and any changes in postural control of the trunk could potentially have a role
506
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
in low-back injuries incurred while pushing or pulling. In support of this hypothesis, Radebold et al. (2001) reported poorer postural control of the trunk in patients with lowback pain compared to healthy controls. However, to our knowledge no studies have investigated the effects of trunk exertions on trunk postural control. Trunk exertion force and direction has the potential to influence trunk postural control. Increasing the force level of trunk flexion or extension exertions increases agonistantagonist co-activation of the trunk flexors and extensors (Lee et al., 2007; Granata et al., 2005). While increased coactivation increases trunk stiffness (Lee et al., 2006) and may be beneficial for injury prevention during some tasks, it may have a deleterious effect on postural control of the trunk. Reeves et al. (2006), for example, reported a degradation of trunk postural control when participants purposefully increased trunk co-activation during unstable sitting. Trunk exertion direction also as the potential to influence trunk postural control. Greater trunk co-activation occurs during trunk flexion exertions compared to extension exertions (Lee et al., 2007; Granata et al., 2005). As noted above, increases in trunk co-activation may impair trunk postural control. Based on our lack of understanding of the effects of trunk exertions on postural control of the trunk, the goal of this study was to investigate the effects of trunk exertion force and direction on postural control of the trunk during unstable sitting. This information could be useful in understanding the etiology of low-back disorders, and potentially aid in the development of manual material handling tasks to help minimize the risk for low-back disorders. 2. Methods Twelve healthy adults participated including 7 males (age 25.7 (6.9) years, height 178.7 (6.9) cm, mass 79.9 (9.0) kg), and 5 females (age 21.4 (1.7) years, height 161.3 (8.2) cm, mass 59.6 (8.3) kg). Participants reported no history of chronic low-back pain, and provided informed consent approved by the Institutional Review Board at Virginia Tech prior to participation. Postural control of the trunk was assessed while participants sat on a wobbly chair (Cholewicki et al., 2000). The seat pan of the chair pivoted freely in the sagittal and frontal planes (Fig. 1). Destabilizing moments arise from gravitational effects associated with small angle deviations from the neutral position (i.e. where the chair-subject system center of mass was positioned directly above the wobbly chair pivot). Adjustable springs with an effective rotational stiffness, kh, applied an elastic restorative moment to the seat to return the seat pan to the neutral position. This stiffness was set at kh = 50% rM Grav where rM Grav was the linearized gradient of gravitational moment recorded in static calibration measurements. Hence, the postural control assessment accounted for differences in subject anthropometry that were intrinsically embedded in rM Grav . A stiffness value of 100% rM Grav would mean the destabilizing
Fig. 1. Wobbly chair design adapted from Cholewicki et al. (2000). An isotonic anteriorly-directed force was applied to the trunk during extension exertion trials, and an isotonic posteriorly-directed force was applied to the trunk during flexion exertion trials. h1 = angle of seat relative to horizontal; h2 = angle of trunk relative to vertical.
gravitational moment in wobbly chair positions other than neutral would be completely offset by spring forces. Since the stabilizing restorative stiffness was less than the destabilizing gravitational moment, the biomechanical system was inherently unstable. Therefore, trunk movement was necessary to correct for disturbances in seat angle from neutral, and to maintain an upright seated posture. A foot support attached to the seat pan supported the lower limbs so that the upright seated posture was maintained predominantly by movement at the lumbar spine. External horizontal isotonic forces were applied in the median plane at the T8 level of the torso. These forces required voluntary trunk extension and flexion moments to maintain an upright posture while seated on the wobbly chair. Forces of 0, 40 and 80N were generated by tensioning elastic bands between a chest harness worn by the participants and a distant wall of the laboratory. The length of the elastic bands was 15 m, thereby assuring that change in length of the band from torso movements was small relative to the band length, i.e. forces were nearly isotonic. The externally-applied isotonic forces caused a tipping moment about the pivot point under the seat pan. To offset this tipping moment, the seat pan was adjusted anteriorly or posteriorly to achieve steady-state equilibrium under the applied load, i.e. the static gravitational moment balanced the externally-applied forces. Note, however, that the system remained unstable, as described above, due to destabilizing moments associated with gravity. Six experimental conditions were tested using a full-factorial design with three exertion forces (0, 40, and 80N) and two exertion directions (trunk flexion and trunk extension). Each subject was instructed to maintain seated balance on a wobble chair with arms-crossed over their chest and to look straight with eyes opened. The presentation order of exertion direction was counter-balanced so that half of the participants performed all flexion exertions first, and the other half performed all extension exertions first. Within each exertion direction, the presentation order of the 40N and 80N conditions was also counter-balanced.
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
A single 0N force condition was always tested first. Subjects completed a block of five trials of 60 s duration in each experimental condition with one minute of rest between trials. Prior to the test protocol, subjects performed five practice trials in each experimental condition to become accustomed to the procedure. Seat angle in both the anterior-posterior and medial-lateral planes was sampled at 100 Hz during all trials using an electromagnetic motion sensor (Ascension Technology, Burlington, VT). Elastic band force was also sampled at 100 Hz using a uniaxial load cell (Interface Inc., Scottsdale, AZ). Seat angle and elastic band force data were then lowpass filtered using a 6 Hz (4th-order Butterworth filter). Seat angle data were de-meaned so that the mean value of each trial was zero. Postural control of the trunk was quantified using summary statistics of seat angle variance, and by non-linear stability analyses. Summary statistics of seat angle included: root-mean-squared (RMS) amplitude of the seat angle, and 95% ellipse area (EA). The RMS amplitude was calculated using the total recti-linear distance from the mean position because this includes the dynamic interaction between anterior-posterior and medial-lateral movements. The 95% EA represents the region of typical movement and was computed using methods reported by Prieto et al. (1996). Non-linear stability analyses included the stabilogram diffusion analysis (SDA) and the maximum finite-time Lyapunov exponent, kmax. SDA analyses compute variance from the mean squared distances traveled hDd 2i iDt by the signal in time interval Dt (Collins and DeLuca, 1993). Cholewicki et al. (2000) showed that the stability diffusion exponent, H, can be computed from the log–log slope of hDd 2i iDt versus Dt = 0.01–2.0 s, and that this exponent demonstrated bi-linear behavior, i.e. short-term and long-term stability exponents. Only the short-term exponent has been shown to discriminate between subject groups with and without low-back pain (Radebold et al., 2001). Therefore, only the short-term exponent was used in the present analyses. Values of H < 0.5 indicate that perturbations are likely to be controlled and attracted toward the equilibrium state, whereas values of H > 0.5 indicate a disturbance will continue to grow toward infinity. H was calculated by using radial direction angle data because of same reason as RMS amplitude. Lyapunov exponents quantify the mean rate of divergence with respect to an equilibrium state. The dynamic state at each instant in time was described by a fourdimensional vector including angle and velocity in the anterior–posterior and medial–lateral planes, ~ qðtÞ ¼ ½hSx ; hSy ; h_ Sx ; h_ Sy . For every data point ~ qðtÞ in the time series we recorded the mean distance, dðtÞ, to three data points that were nearest to it (Kantz and Schreiber, 2004). In an asymptotically stable system, the distance between these points must grow smaller with time. Preliminary analyses demonstrated that averaging the three nearest neighbors reduced the effects of stochastic noise, whereas more than
507
three failed to improve signal quality. Stability was quanti changes over a finitefied from the rate at this which dðtÞ time period Dt = 0.01–0.2 s, i.e. from d(t) to d(t + Dt) averaged over all time samples hdðt þ DtÞ=dðtÞit ¼ ekMax Dt
ð3Þ
An increase in H and kmax imply impaired postural control of the trunk. A detailed description of these methods is available elsewhere (Wolf et al., 1985; Rosenstein et al., 1993). Prior to the statistical analyses, postural control measures calculated from all five trials in each experimental condition were averaged. A repeated-measures ANOVA was used to investigate the effects of exertion force (0, 40, and 80N), and exertion direction (trunk flexion and extension) on RMS, EA, H and kmax. Following significant main effects or interactions, pairwise comparisons were performed using the Tukey HSD. Significance was determined at the level of P < 0.05. 3. Results All dependent variables increased as exertion force increased (Fig. 2). Compared to 0N exertions, the RMS and EA were 14.3% and 17.7% higher during 40N exertions, and 33.0% and 46.0% higher during 80N exertions, respectively (P < 0.001 for RMS and P < 0.001 for EA). Similarly, H and kmax were 0.2% and 7.7% higher during 40N exertions, and 0.6% and 11.2% higher during 80N exertions, respectively, compared to 0N exertions (P = 0.042 for H and P < 0.001 for kmax). Flexion exertions generally exhibited higher values of the dependent variables compared to extension exertions (Fig. 2). RMS, EA, and kmax were 29.1% (P < 0.001), 7.9% (P = 0.046) and 5.8% (P = 0.002) higher, respectively, during flexion exertions, while H (P = 0.56) was not different. Both RMS and kmax exhibited a significant force direction interaction. Post-hoc analyses revealed RMS was higher during the flexion exertions than extension exertions at both the 40N and 80N force levels, and kmax was higher during the flexion exertions and extension exertions at the 40N exertion force level only. 4. Discussion The goal of this study was to investigate the effects of trunk exertion force and direction on trunk postural control during unstable sitting. The dependent variables used were based on variability and non-linear stability analyses of seat angle. Increases in these dependent variables implied a degradation of trunk postural control. Results, in general, indicated a degradation of trunk postural control as exertion force increased, and that trunk postural control was poorer during trunk flexion exertions compared to trunk extension exertions. A quantitative comparison between the results here and other studies investigating trunk postural control on a during unstable sitting (e.g. Radebold et al., 2001; Cholewicki
508
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
Fig. 2. Summary statistics of seat angle variance (RMS, EA) and non-linear stability measures were affected by exertion force and direction. *: a significant difference from 0N force. : a significant difference from 40N force. à: a significant difference between flexion and extension. #: main effect of exertion force. €: main effect of exertion direction.
et al., 2000) was not possible to due subtle differences in wobbly chair design and the fact that the postural control measures used here were based on angular displacement of the chair while these other studies were based on center of pressure. However, the mean values of H reported here (0.76–0.77) were comparable to these other studies (0.56– 0.83). The degradation of trunk postural control with increasing exertion force found here appears to be consistent with other studies. For example, Reeves et al. (2006) reported a degradation of postural control during unstable sitting with elevated trunk co-activation. While the current study did not manipulate trunk co-activation per se, the increase in exertion force investigated here has been shown to increase trunk co-activation (Lee et al., 2007; Granata et al., 2005). This increase in trunk co-activation can degrade trunk postural control because increased activation levels of muscles are associated with increased variability in force output (Christou et al., 2002; Hamilton et al., 2004). Results also indicated trunk postural control was poorer during flexion exertions compared to extension exertions. Co-activation during trunk flexion exertions has been reported to be approximately twice the level of co-activation during equivalent extension exertions (Granata et al., 2005). Knowing that increased activation levels are associated with increased variability in force output (Christou et al., 2002; Hamilton et al., 2004), these results appear to be consistent with other findings. Three limitations of this study warrant discussion. First, actual pushing and pulling tasks were not investigated even
though the main motivation for this study was the association of pushing and pulling exertions with low-back disorders. Using controlled trunk flexion and extension exertions allowed for easier experimental control of trunk exertions levels, and avoided potential variability between subjects in upper extremity performance. Second, although differences in trunk postural control during unstable sitting have been identified between individual with and without low-back pain (Radebold et al., 2001), a clear relation between trunk postural control and actual spinal pain/ injury has not been established. Third, EMG data were not included in our analysis, so it was only possible to hypothesize changes in co-activation when attempting to explain the results. In conclusion, increasing trunk exertion force degraded trunk postural control during unstable sitting, and trunk flexion exertions exhibited poorer trunk postural control compared to trunk extension exertions. This study may aid in understanding how pushing and pulling exertions can potentially contribute to low-back disorders. Further analyses are necessary to investigate trunk postural control during occupational tasks involving pushing and pulling. Acknowledgements This work is dedicated to Dr. Kevin P. Granata, who was tragically killed on April 16, 2007 at Virginia Tech. This study was supported by a grant R01 AR49923-01 from CDC of the National Institute for Occupational
H. Lee et al. / Clinical Biomechanics 23 (2008) 505–509
Safety and Health. We wish to thank Steve Hanson and Emmanuel Ekwueme for their assistance in collecting data. References Andersson, G.B.J., 1981. Epidemiologic aspects on low back pain in industry. Spine 6, 53–60. Baril-Gingras, G., Lortie, M., 1995. The handling of objects other than boxes: univariate analysis of handling techniques in a large transport company. Ergonomics 38, 905–925. Cholewicki, J., Polzhofer, G.A., Radebold, A., 2000. Postural control of trunk during unstable sitting. J. Biomechanics 22, 1733–1737. Christou, E.A., Grossman, M., Carlton, L.G., 2002. Modeling variability of force during isometric contractions of the quadriceps femoris. J. Mot. Behav. 34, 67–81. Collins, J.J., DeLuca, C.J., 1993. Open-loop and closed-loop control of posture: a random-walk analysis of center of pressure. Exp. Brain Res. 96, 308–318. Damkot, D.K., Pope, M.H., Lord, J., Frymoyer, J.W., 1984. The relationship between work history, work environment and low-back pain in men. Spine 9, 395–399. Granata, K.P., Lee, P.E., Franklin, T.C., 2005. Co-contraction recruitment and spinal load during isometric trunk flexion and extension. Clin. Biomech. 20, 1029–1037. Hamilton, A.F., Jones, K.E., Wolpert, D.M., 2004. The scaling of motor noise with muscle strength and motor unit number in humans. Exp. Brain Res. 157, 417–430. Hoozemans, M.J., van der Beek, A.J., Frings-Dresen, M.H., van Dijk, F.J., van der Woude, L.H., 1998. Pushing and pulling in relation to musculoskeletal disorders: a review of risk factors. Ergonomics 41, 757–781. Kantz, H., Schreiber, T., 2004. Nonlinear Time Series Analyses, Second ed. Cambridge University Press, NY.
509
Lee, P.J., Rogers, E.L., Granata, K.P., 2006. Active trunk stiffness increases with co-contraction. J Electromyogr. Kines. 16, 51–57. Lee, P.J., Granata, K.P., Moorhouse, K.M., 2007. Active trunk stiffness during voluntary isometric flexion and extension exertions. Hum. Factors 49, 100–109. Marras, W.S., Lavender, S.A., Leurgans, S., Fathallah, F.A., Allread, W.G., Ferguson, S.A., Ferguson, S.A., Allread, W.G., Rajulu, S.L., 1995. Biomechanical risk factors for occupationally related low back disorder risk. Ergonomics 38, 377–410. Prieto, T.E., Myklebust, J.B., Hoffmann, R.G., Lovett, E.G., Myklebust, B.M., 1996. Measures of postural steadiness: differences between healthy young and elderly adults. IEEE Trans. Biomed. Eng 43, 956– 966. Radebold, A., Cholewicki, J., Polzhofer, G.A., Green, T.P., 2001. Impaired postural control of the lumbar spine is associated with delayed muscle response times in patients with chronic idiopathic low back pain. Spine 26, 724–730. Reeves, N.P., Everding, V.Q., Cholewicki, J., 2006. The effect of trunk stiffness on postural control during unstable seated balance. Exp. Brain Res. 174, 694–700. Rosenstein, M.T., Collins, J.J., DeLuca, C.J., 1993. A practical method for calculating largest Lyapunov exponents from small data sets. Physica D 65, 134. Schibye, B., Sogaard, K., Martinsen, D., Klausen, K., 2001. Mechanical load on the low back and shoulders during pushing and pulling of twowheeled waste containers compared with lifting and carrying of bags and bins. Clin. Biomech. 16, 549–559. Van der Beek, A.J., Hoozemans, M.J.M., Frings-Dresen, M.H., Burdorf, A., 1999. Assessment of exposure to pushing and pulling in epidemiological field studies: an overview of methods, exposure measures, and measurement strategies. Int. J. Ind. Ergon. 24, 417–429. Wolf, A., Swift, J.B., Swinney, H.L., Vastano, J.A., 1985. Determining Lyapunov exponents from a time-series. Physica D 16, 285–317.