Original articles Changes of contour of the spine caused by load carrying J.J. Vacheron1, G. Poumarat1, R. Chandezon1 and G. Vanneuville2 1 Laboratoire de la Performance Motrice, Unité de Biomécanique, U.F.R. S.T.A.P.S, F-63177, Aubière Cedex, France 2 Laboratoire d'Anatomie, Faculté de Médecine, BP 38, F-63001 Clermont-Ferrand Cedex 1, France Received July 16, 1998 / Accepted in final formOctober 23, 1998 Key words: Load carriage – Trunk – Posture – Backpack Correspondence to: J.J. Vacheron

Abstract The development of new leisure activities such as walking has spread the use of the backpack as a means of carrying loads. The aim of this work was to present a way of defining the movements imposed on the trunk by this type of load carrying. A 20 kg load situated at the thoracic level (T9) of the trunk, was placed in a backpack (2.5 kg). The 12 subjects were average mountain guides of Auvergne region, intermediate level and complete beginners. External markers were glued to the projecting contours of the spinous processes of the C7, T7, T12, L3 and S1 vertebrae, the shin and the external occipital tuberosity (EOT). Using a Vicon 140 3-D system we measured the effective mobility of the different spinal segments in the sagittal plane during one step. For every subject, we noticed a significant decrease of the effective inter-segmental mobility (EISM) between S1-L3-T12 (p < .01) while backpacking a 22.5 kg load. A decrease of EISM also appeared at the next level between L3-T12-T7 (p < .05). An increase of the EISM between T7-C7-EOT was noted (p < .05). We supposed that strength loss of the back muscles and/or angular oscillations of the trunk could be a common cause of symptoms during backpacking. The subjects using this type of load carrying have to adopt an adequate position of the lumbar, dorsal and cervical vertebrae.

Mechanical loads on the low back appear to play a major role in the etiology of low back pain and clinical research reveals a high prevalence of spinal disorders in adults [17, 21]. The development of leisure or military activities involving a backpack justifies studies of their effects on the human body. Most of these studies have explored the metabolic effect of load carrying under various circumstances (type of ground, velocity, load mass, etc). However, the findings of Soule and Goldman [18], Goldman and Iampietro [4] and Pandolf et al [14] indicated that biomechanical assessment provided more information about load-carrying than metabolic data alone [8, 15]. In biomechanics, few researches have been conducted on body posture modification to study the effect on spinal loading and spinal problems. The classic study of Demeny [3] dealt with changes in trunk position and alignment of the centre of gravity in subjects carrying loads supported by the shoulders alone. Forward inclination of the trunk was observed by Kinoshita [7], Martin et al [11] and Bloom and Woodhull-McNeal [1]. Other adaptations to load include reduced pelvic rotation and gait pattern accommodation [7, 11]. The kinetic approach to spinal movement and the mechanical strength of the spinal structure have also been investigated [12], but the displacements that occur at different trunk levels during load carriage have received little attention [16]. Radiography is the method generally used for the study of postural variations of the spine [10]. But with radiologic techniques only small segments can be studied over a narrow range of movement, and the risk of overexposure to ionizing radiation should be avoided. Noninvasive methods using photography or high-speed cameras give more accurate results in large segments of the trunk [1, 7]. According to the state of knowledge the main question that arises is whether the posture used during load carrying with a backpack can induce low back pain. The aim of this work was to describe spinal

movements in expert and novice subjects, knowing that beginners are more sensitive than experts when using a backpack. The purpose of this study was to assess three-dimensionally the kinetics of the trunk associated with walking while load-carrying. This original report focusses on trunk movements at different levels under load-carrying. Material and methods Twelve subjects (age 32 years ± 10.2; height 174.5 cm ± 4.6 and weight 71.3 kg ± 6.3) took part in this experiment and gave their informed consent. Only subjects with no history of functional disorders were included in the study. 4 subjects were mountain guides in the Auvergne region (experienced hikers), 4 were occasional hikers, and 4 were novice hikers. The backpack used was a prototype designed and made in the laboratory. It had a steel frame with a tray supporting a load located at T9 level (Fig. 1). The empty backpack weighed 2.5 kg. The subjects were asked to walk twice: first with shoes and an empty backpack and then with shoes and loaded backpack (22.5 kg). Four 50 hz video cameras were used to capture the 3 coordinates (O, x, y, z) of reflective markers placed on the skin. These external markers were glued to the projecting contours of the spinous processes at C7, T7, T12, L3 and S1 vertebral levels, the shin (S) and the EOT (external occipital tuberosity). Two cameras were located behind and two in front of the subject in order to have at least two cameras capable of visualising the same markers during movement. Each camera was calibrated using a volume calibration object. Each video signal was recorded and digitalized via a data-station (Vicon 140 system). The system produced less than 0.5 mm noise in marker coordinates and the overall error of 3-D coordinates from the system tested within a 3¥2¥2 m 3-D space was less than 0.5 cm. Effective intersegmental mobility (obtained from angle between two consecutive segments, identified by projection of the vertebral levels on the effective spinal curve in the sagittal plane (Fig. 2) assigned to markers placed on S1, L3, T12, T7, C7, S and EOT characterized the effective movement of the segments S1-L3-T12, L3-T12-T7, T12-T7-C7, T7-C7-EOT and S-C7-T7. This technique was designed to record the angular displacement of each vertebral segment.

Fig. 1. Prototype backpack designed in the laboratory: steel frame with a tray, supporting a 21 kg load located at thoracic level (T9) of the trunk

Fig. 2. Effective intersegmental mobility (EISM): angular distance between two consecutive segments Subjects wore shorts and their own shoes, with markers placed on the anatomic landmarks previously noted. They were required to maintain the same posture of the upper limbs while walking (i.e. with left and right hands at chest level). The subjects were instructed to start walking at their own pace and to place the left foot on a reference surface. The kinetic data collected during one step phase were used to determine the mean displacement and mean angular oscillation of each vertebral segment in the sagittal plane. To assess angular oscillation, we used the value corresponding to 80% of the full range of EISM measurements in order to avoid extreme displacements. For the same experimental conditions, the different vertebral EISMs were compared using an unpaired t-test. The mean values and the full range of measurements were compared. The effect of load (empty backpack vs. loaded backpack) was studied at different trunk levels by a paired test. Results All the mean values, standard deviation and full range of measures (80%) of the EISM, calculated according to the total population, are given in Table 1.

Table 1. Mean, standard deviations and 80% fluctuations for each level of expertise calculated for each angle. The population comprised 4 novices, 4 semi-trained and 4 experts Comparison of the angular variation (empty backpack vs. loaded backpack) The subjects leaned forward while wearing a loaded backpack. The EISMs between S1-L3-T12 and between L3-T12-T7 were all significantly different between the loaded backpack carrying task and the unloaded control stance (10.9° ± 5.0° vs 15.6° ± 4.2° ; p < .01 and 1.2° ± 7.7° vs 5.9° ± 6.6° ; p < .05). The subjects seemed to have their maximum angular variation between the thoracic and lumbar vertebral levels and no significant difference existed related to the subjects' degree of expertise. Comparison of the angular variation between each stage The angular variation between the vertebral stage S1-L3-T12 and T12-T7-C7 (p < .001), S1-L3-T12 and T7-C7-EOT (p < .001), S1-L3-T12 and S-C7-T7(p < .01) were significantly different. Comparatively, the EISM between L3-T12-T7 evolved in the same way as the ESIM for S1-L3-T12 (p < .05) and no significant difference existed between the other stages observed. Comparison of the angular oscillations (or fluctuations: full range of measurements: empty backpack vs. full backpack) The angular oscillation, which represented 80% of the full range of measurements observed for every EISM, did not change whatever the load transported. We noted no significant difference in the analysis of the oscillations between the different spinous processes. The small number of subjects within each group precluded considering each group separately. However, we did notice a tendency for angular oscillations to

differ according to the experience level of the subject. Figure 4 indicates that the skilled subjects had smaller fluctuations of the EISM when carrying loads. Only the (T7-C7-EOT) EISM showed an increased oscillation (+2.7°). The semi-trained subjects resembled the analysis of the total population since no major change appeared in the oscillations. The novices seemed to be more sensitive to the effect of the load, especially at the level of the T7-C7-T7 EISM, with an increase of +15.6° in the oscillations in the comparison with the skilled subjects.

Fig. 3. Angular variations of the spinous processes during walking with shoes and without load in the backpack

Fig. 4. Angular fluctuations of the spinous processes during walking with shoes and without load in the backpack The average displacement of the subject in the sagittal plane, measured at S1 vertebral level, is noted in Table 2 during the different modalities. No significant differences were noted between the different experimental modalities.

Table 2. Average speed of S1 vertebra in different modalities. Discussion Low back symptoms are felt by sportsmen and women while carrying a backpack [2, 9]. Teenagers also suffer from hyperlordosis due to exaggerated curvature of the bottom of the back. It is a fact that the weight of school backpacks continues to increase despite legal measures to limit these risks (BO n° 339, 1995 in France). The few researches concerning biomechanical parameters indicate that the load, whether light or heavy (20% to 40% of the body-weight) changes the normal pattern of walking significantly [11]. These studies did not take into account the movement of the different spinal segments. Postural analysis in relation to the different loads transported with a backpack remains succinct. Kinoshita [7] showed from an analysis of 2D images that the angular variations of segments such as the body, thighs, legs and feet were significantly

different whether carrying a load or not. Bloom and Woodhull McNeal [1] also noted postural compensation in standing subjects carrying a 19 kg backpack. "The multiple functions of the spine to stabilise the body in an erect position, to co-ordinate the movements of the body as a whole or in its parts and to ensure reciprocal signal-transfer within the body and to its environment also demand from the different stages of the trunk and particularly the lumbo-sacral transition to be as stable as possible and flexible as necessary!" (Tittel [19]). This implies the inevitable question of the loadability (and its limits) of the trunk, especially when the three tasks mentioned overlap, as is the case in a decrease of stability linked to increased flexibility of the trunk. According to the few researches on the kinetic carrying of the load, the value of this study lies in an accurate and reliable measurement of postural change related to the lumbar, thoracic and cervical segments and consequently to spinal stress. The analysis described here also takes into account the skill level of the subjects. We used a medium load (22.5 kg), corresponding to the weight necessary for two days of trekking or to the weight carried during a family excursion with a rucksack or child carrier. Such users are the most endangered. The backpack chosen was an ordinary standard model like those used by unskilled persons. A set of previous experiments performed in the laboratory [23, 24] showed that the slight difference observed between external markers and spinous process movements allowed the study of spinal movement by noninvasive methods, especially for thin subjects, as was the case in this study. For each subject, we noticed a significant decrease of the EISM between S1-L3-T12 (p < .01) when carrying a load of 22.5 kg. The decrease of the EISM also appeared at the next level between L3-T12-T7 (p < .05) The explanation of this phenomenon is twofold. As specified by Martin [11], this process is initially the translation of a rocking movement of the trunk forward, in order to place the centre of gravity of both subject and load vertically over the feet on the ground. The decrease of lordosis is also a consequence of a retroversion movement of the pelvis, which leads to horizontalisation of the superior S1 level [16]. This modification of posture, based on the coxo-femoral joint, has already been mentioned for squat exercises, when the subject keeps this position to decrease the shearing and sliding effect on the vertebrae. All the subjects showed a decrease of the thoracic and lumbar EISM, but the skilled subjects maintained control of this new posture by decreasing the oscillations. The studies of static weight-lifting in weight-lifters conducted by Vanneuville et al [22] showed that the locking of the pelvis depended on the weight-training status. Some studies indicate that there are cyclic stresses of the back muscles during different types of load-carrying [5, 13]. More precise EMG studies on the abdominal wall muscles during walking with a backpack may confirm that the level of skill of the subject is of major importance in postural management of the load. We may assume that the skilled, because of their daily training sessions (> 100 a year) are more efficient than novices in maintaining good positioning of the spine, thus reducing the incidence of some injuries and causing less perceived strain. While the EISM between T12-T7-C7 remained the same with or without load (except for the skilled subjects, in whom a slight decrease of the angle appeared) the EISM between T7-C7-EOT increased according to the variation registered on the inferior segment (p < .05). This increase tended to bring the head of the subject upward. This raising of the head started as early as the dorsal segment with the unskilled and semi-trained subjects, but tend to start this decrease at the level of the neck in the skilled subjects. The first explanation of this increase comes from the subject who must manage with the movement of the pelvis previously noted. The only way to preserve his vision is to raise his head. The second explanation comes from the torque induced to hold the load backward. In fact, the head tends to align with the axis of the spine in order to compensate for the tension imposed backward on the shoulders and head by the backpack. Biomechanical studies of the forces imposed at the cervical and dorsal levels show that the torque recorded on these vertebral segments is very great compared to other segments [20]. This must be compensated by specific muscle action, which makes it possible to limit the intervertebral rocking movement noted in the unskilled subjects. The importance of the dorsal muscles is shown by Holewijn [6] who confirms that the influence of the load on the trapezius mm. is considerable (+2.7% of EMG activity).

Conclusion This study confirmed that the subjects using this type of load-carrying have to adopt an adequate position of the cervical, dorsal and lumbar vertebrae. All the subjects showed a decrease of the thoracic and lumbar EISM, but spinal oscillations differed according to the expertise level of the subject. The skilled subjects maintained better control of this new posture by decreasing the oscillations than the novices, who seemed to be more sensitive to the action of the load. This expert postural comportment can be explained by a voluntary response, enhanced by muscle development and training. This adaptive postural comportment was necessary to improve performance and protect the shoulders and spine during load carrying. References

1. Bloom D, Woodhull-McNeal AP (1987) Postural adjustments while standing with two types of loaded backpacks. Ergonomics 30: 1425-1430 2. Dalen A, Nilsson J, Thorstensson A (1978) Factors influencing a prolonged foot march. Karolinska Institute FOA, Stockholm Sweden, Report C50601-H6 3. Demeny G (1924) Mécanisme et éducation des mouvements. Bibliothèque Scientifique Internationale Revue EPS, Paris, pp 435-436 4. Goldman RF, Iampietro PF (1962) Energy cost of carriage. J Appl Physiol 17: 675-676 5. Harman E, Han KH, Frykman P, Johnson M, Russell F, Rosenstein M (1992) The effects on gait timing kinetics and muscle activity of vanous loads carried on the back. Med Sci Sports Exer 24: S129 6. Holewijn M (1990) Physiological strain due to load carrying. J Appl Physiol 61: 237-245 7. Kinoshita H (1985) Effects of different loads and carrying systems on selected biomechanical parameters describing walking gait. Ergonomics 28: 1347-1362 8. Kirk J, Schneider DA (1992) Physiological and perceptual responses to load carrying in female subjects using internal external frame backpack. Ergonomics 35: 445-455 9. Knapik J, Reynolds K, Staab J, Vogel JA, Jones B (1992) Injuries associated with strenous road marching. Military Med 157: 64-67 10. Lindhal O (1966) Determination of the sagittal mobility of the lumbar spine a critical method. Acta Orthop Scand 37: 241-254 11. Martin PE, Nelson RC (1986) The effects of carried loads on the walking patterns of men and women. Ergonomics 29: 1191-1202 12. McGill SM, Norman RW (1986) Partitioning of the L4-L5 dynamic moment into disc ligamentous and muscular components during lifting. Spine 7: 666-678 13. Norman RW (1979) The utility of combining EMG and mechanical work rate data in load carriage studies. Proc 4th Congr Int Soc Electrophysiological Kinesiology, Boston MA 14. Pandolf KB, Givonni B, Goldman RF (1977) Predicting energy expenditure with loads while standing or walking very slowly. J Appl Physiol 43: 577-581

15. Pierrynowski MR, Norman RW, Winter DA (1981) Mechanical energy analyses of the human during load carriage on a treadmill. Ergonomics 24: 1-14 16. Poumarat G, Vanneuville G, Vacheron JJ (1994) Rachis et port de charge. Archives Internationales de Physiologie, de Biochimie et de Biophysique C51-C52 17. Sanders HWA, Gijn J, Veraart BE, Tulleken CA, Luken NJ, Vermeer JP (1983) Klinishe betekenis van degeneratieve afwijkingen van de lumbale wervelkolom. Radiologishe afwijkingen Ned T Geneesk 127: 1374-1385 18. Soule RG, Goldman RF (1972) Terrain coefficients for energy cost prediction. J Appl Physiol 32: 706-708 19. Tittel K (1990) The loadability and relieveability of the lumbo-sacral transition in sports. J Sports Med Physical Fitness 30: 113-121 20. Vacheron JJ, Poumarat G, Chandezon R, Dabonneville M, Vanneuville G (1994) Torque applied by a load on trunk. J Biomechanics 27: 817 21. Van Dieen JH (1993) Functional load of the low back. IMAG-DLO -III -Amsterdam, Report 93 22. Vanneuville G, Garcier JM, Poumarat G, Guillot M, Chazal J (1992) Mechanism of orientation of the pelvifemoral base during static loading of the lumbar spine in weight-lifters. Surg Radiol Anat 14: 29-33 23. Vanneuville G, Monnet JP, Vacheron JJ, Garcier JM, Guillot M, Poumarat G (1996) Determination of the position of the pelvic girdle by the method of external markers in the sagittal plane: a preliminary feasibility study. Surg Radiol Anat 18: 245-247 24. Vanneuville G, Poumarat G, Vacheron JJ, Ferry B (1994) Measurement of spine movement from external markers. J Biomechanics 27: 818

Surg Radiol Anat 21: 109-113 © Springer-Verlag France 1999