Link to Figures Show Figures
	Article Text Figures References
	Send Link Send HTML Article

T12 and S2 Vertebrae Oscillation During the Stance Phase of Gait in Asymptomatic Young Adults Fitted With Accommodative Foot Orthoses

George Marinakis, MEng, MPhil, PhD, CEng, MISPO
Paola Catalfamo, MBEng

ABSTRACT

The authors investigated the effect on the oscillation of the twelfth thoracic and the second sacral vertebrae of healthy individuals during level walking when accommodative foot orthoses are used. Three-dimensional motion analysis was performed on eight healthy young male adults, fitted bilaterally with separated-arms foot orthoses while performing gait trials. Six spherical retroreflective skin-mounted markers were placed at specific anatomical points. Each participant performed six consecutive gait cycles at self-selected speed under three experimental conditions: barefoot, shod, and shod with insoles. Kinematic variables were then calculated, and repeated measures one-way analysis of variance and Tukey-Kramer multiple comparison tests were performed to characterize the differences observed among the test conditions. Statistically significant differences were found for the range of motion, the range of velocity, the maximum position and the maximum velocity instants of the thoracic vertebra, and for the range of motion and the maximum velocity instant of the sacrum. From these differences, the maximum position and the maximum velocity instants of the thoracic vertebra and the maximum velocity instant of the sacrum are attributed to the use of the orthoses. In addition, the increment of the damping ratio of both vertebrae, associated with the dissipation effect of the orthoses, was clear. These results indicate that there are some immediate effects of the tested foot orthoses on the oscillation pattern of the lower spine of the examined group. These insoles might contribute to motion-related low back pain prevention or relief; however, the actual mechanism remains unclear, so additional investigation is needed. ( J Prosthet Orthot. 2005;17: 57–65.)

Foot orthoses are extensively used by healthy and disabled individuals to support and/or correct foot function.

According to a 2000 survey of "attitudes toward foot care," conducted by the American Podiatric Medical Association, with a sample of 39.6 million Americans 18 years and older, 53% were currently wearing prescribed (8.3 million) or nonprescribed (12.7 million) foot orthoses or foot supports.^1a In the United Kingdom, almost 30% of the ฃ38 million 1994 budget for orthoses in the National Health Service clinics alone was spent on foot orthoses.^1b

The clinical effectiveness of foot orthoses in the treatment of various pathological states and syndromes, including low back pain, has been proven by several studies.^2–5 However, research on the biomechanical effects to reveal the mechanisms through which certain benefits can be derived is very limited. Studies on the biomechanical behavior of foot orthoses were mainly focused on the effect of insoles upon the kinematics of the lower limbs during running^6,7 and the plantar pressure redistribution.^8,9

Walking is the most common form of human work that introduces impulsive stresses into the skeleton. The shock wave generated during heel strike is normally dissipated and attenuated by the body's natural shock absorbers (ie, the foot, ankle, knee, hip, pelvis, and intervertebral disks). Furthermore, the shock absorption capacity of the human locomotor system has been correlated with a variety of degenerative joint diseases.¹⁰ Foot orthoses aim to act as artificial shock absorbers, helping the human body prevent overload conditions. Biomechanical studies on the effect of foot orthoses on the lower limbs and pelvis kinematics during normal walking are scarce.^11,12

Work has been performed previously on the dissipation effect of insoles during walking using accelerometers mounted at the ankle,¹³ tibial tuberosity,¹⁴ sacrum and forehead, 5 and between the teeth.¹⁵ The results of all these studies showed a significant reduction of skeletal shock when the insoles were used.

Studies of the trunk, spine, and pelvis during locomotor activities using three-dimensional motion analysis systems indicated that the upper-body kinematics are important to locomotor control,^16,17 and that the observed patterns within and among the thoracic and lumbar spine segments and the pelvis are consistent and comparable among different conditions.^18,19 However, foot orthoses were not included in these studies.

In a previous study,²⁰ we investigated how normal subjects responded in terms of the motion of their lower limbs, pelvis, and trunk during level walking when fitted with separated medial and lateral arms foot orthoses. There were some immediate effects of these orthoses to the walking pattern of the examined group. The way the orthoses contribute to relief of pain and fatigue, especially in the lower spine, could not be revealed.

In the current study, we investigated how the same subjects responded in terms of the motion of their spine during level walking when fitted with the same foot orthoses. The Cartesian coordinates of T12 and S2 collected during gait analysis were used to derive (apart from the corresponding accelerations) various other kinematic variables associated with the lower spine motion pattern and the shock absorption effect of the tested insoles.

MATERIALS AND METHODS

DESIGN

The mass-produced (off-the-shelf) foot orthosis used in this study was a three-quarters length, U-shaped insole with separated medial and lateral arms (Backease, Seton Scholl Healthcare, Oldham, UK). These insoles use a three-layer construction. The top layer consists of a soft nonslip cloth; the second layer is made of cellular polyethylene foam, and the heel area has an additional shaped plug of the same material. The insole is commercially available in two different sizes (large and small). In this study, eight large insoles (fitting shoes with sizes from 7 to 11) were used, one pair for each subject. More information about this orthosis can be found in a previous study.20

All subjects wore the insoles in conjunction with their own personal shoes. All shoes were leisure shoes, similar in style, and had been routinely used by the subjects before the beginning of the study; all shoes were in good condition. The subjects were asked to wear their own shoes because these were considered a familiar technical aid to comfortable walking. The alternative could have been to use the same "standard" shoes for all participants; however, this requires time for the participants to acclimate with the new footwear and it could affect the characteristic gait pattern of each individual. All shoes had a full length, flat sole, made of rubber, with a closed heel and a heel height difference between 13.0 and 15.0 mm. All shoes had lace enclosures and square toes.

To ensure correct and consistent shoe fit with and without orthosis, the shoes were tested for adequate space to accommodate and complement the added material of the orthoses. To avoid too-narrow or "snug" fit, the width of the shoes at the widest portion of the foot was tested to ensure that a small piece of shoe could be grasped (i.e., the pinch test). This way the natural expansion of the foot during each weightbearing step was not prevented. The heel area of the shoes was tested to be narrow enough to keep the heels from moving up and down during walking. All participants were fitted with foot orthoses by the same investigator (the first author), and detailed instructions were given on the proper wear and maintenance.

Three conditions were tested: barefoot, shod, and shod with insoles. The dependent variables were lower spine kinematic variables.

Gait analysis was conducted in two separate sessions. During the first session, subjects were tested while barefoot and while wearing shoes. Insoles were worn during the second session, held 3 weeks after the first to achieve a minimum degree of acclimatization to the insoles.

PARTICIPANTS

Eight healthy male young adults participated in this study. Their mean age was 24.4 ฑ 4.3 years, mean weight 64.5 ฑ 8.3 kg, and mean height 1.77 ฑ 0.08 m. Their activity level was characterized as "medium" because they participated in modest physical activity on a regular basis. None of the subjects was a foot orthosis user before the current study. Inclusion criteria were: 1) no current musculoskeletal system complications; 2) right-handed and right-footed; 3) absence of any cognitive problems; 4) absence of any other conditions that could affect and/or limit walking ability, and 5) continuously wearing the insoles for 3 weeks preceding the second gait analysis session.

The foot structure of the participants was characterized as "neutrally" aligned, with no deformities that could justify a pes planus or pes cavus type. None had experienced any neurological or muscular disorder or injury to the tibialis posterior tendon or any condition in which one or more of the arches of the foot had fallen. Furthermore, no foot deformities that might lead to abnormal positions of the calcaneus were revealed under static and dynamic observation. Finally, there were no other signs, such as muscle atrophy, bony prominences, or toe and metatarsal head deformities, that could be related to the presence of foot abnormalities.

All participants reported they were feeling comfortable in all three test conditions and that they had not experienced any low back problems during the last 2 months.

The measuring procedure was carried out according to the ethical guidelines for teaching and research edited by the Advisory Committee on Ethics of the University of Surrey, and informed consent was obtained from all participants.

MEASURING PROCEDURE

Three-dimensional (3-D) gait analysis, using seven CCD cameras (ProReflex MCU, Qualisys Medical AB, Gothenburg, Sweden) and passive retroreflective skin-mounted markers, was performed on all participants. Before any image capture, the system was calibrated using a reference structure for defining the calibration coordinate system and a wand to provide the camera system with measurement points. The distance between two 19.3-mm markers affixed to a rigid rod (wand) such that their centers were 750 mm apart was recorded as the wand was moving in all three directions and in the entire measurement area within a space of 4000 X 2000 X 2000 mm. During all calibration measurements, the standard deviation was less than 0.5% and the range of difference less than 1%.

The participants were asked to wear close-fitting, darkcolored swimsuits made of elastic nonreflective material. Six spherical retroreflective markers with a diameter of 19.3 mm were placed at specific anatomical points: between the second and third metatarsal bones of both feet (15 mm proximally of the metatarsal heads); posterior of the calcaneus of both feet (at the same horizontal plane as the previous marker); on the sacrum (the spinous process of S2 was located as the midpoint between bilateral posterior superior iliac spine), and on T12 (the interspace between the third and fourth lumbar vertebrae was found as lying at the same level as the tops of the iliac crests, the spinous processes of L3-T12 were then palpated and the marker was applied) (Figure 1 ).

The skin at the sites of marker placement was cleaned with an alcohol wipe to remove any oils or lotions. The markers were then placed with hypoallergenic tape. The placement of the skin-mounted markers was performed in a consistent manner by the same investigator.

During each session, and before actual image capture, an orientation period allowed the participants to practice walking under testing conditions. Subjects were asked to walk on a 10-m walkway four times at their comfortable speed. The participants' performance during six consecutive gait cycles was then recorded at a capture rate of 240 Hz. All subjects started from a position 4 m away from the measurement area to reach a natural continuous walking pattern as they entered the calibrated area.

The captured images were simultaneously digitized and all 3-D segments were processed and edited using the QTrac software (Qualisys Medical AB, Gothenburg, Sweden). The procedure included tracking motion data (creating 3-D marker trajectories); sorting the 3-D data according to the markers used in the measurement; and selecting appropriate data to export for analysis.

DATA REDUCTION

To remove any additive noise (i.e., any components of the final signal that were not attributable to the process itself), the collected raw data for the Cartesian coordinates of the six markers were smoothed by means of a digital second order low-pass Butterworth filter, applied both backward and forward, resulting in a fourth order filter, with a cut-off frequency of 10 Hz implemented using Matlabฎ (The Math-Works, Natick, MA).

With the use of the filtered data, the first heel strike was defined as the instant when the velocity of the heel marker along the progression line dropped below a predefined threshold (480 mm/second). In the same way, the toe-off was defined as the instant when the velocity of the toe marker exceeded the same threshold.

Each vertebra moving along the vertical axis was considered to be a linear system with viscous damping. To calculate the dimensionless damping ratio (r), the following formula was used:

where d is the logarithmic decrement (per unit cycle) , expressed by:

with A_i and A_i+1 representing two successive peak position points, being one period apart in the time history.

For the purposes of this study, the oscillation of each vertebra was characterized by kinematic variables associated with its vertical position, velocity, and acceleration. In the case of position, the damping ratio, the range of motion (in mm), and the instant when the highest position was reached (in percentage stance) were calculated. In the case of velocity and acceleration, the maximum values (in mm/second and mm/second²), the range of values (in mm/second and mm/second²), and the instant when the maximum values were observed (in percentage stance) were calculated. Thus, 18 oscillation variables were calculated for each of the three test conditions.

STATISTICAL ANALYSIS

Repeated measures one-way analysis of variance (ANOVA) tests were performed to characterize the differences of the measured parameters in the three experimental conditions. The Tukey-Kramer multiple comparison tests were used when significant differences were found ( p < 0.05).

RESULTS

Typical diagrams of the position, Z, the velocity, U_z, and the acceleration, A_z, of T12 and S2 during the stance phase of walking at each test condition are presented in Figures 2 and 3 . The calculated variables for all participants together with the mean values and their standard deviation are presented in Table 1 , Table 2 , and Table 3 .

TWELFTH THORACIC VERTEBRA

When the insoles were used, the damping ratio was noticeably increased by 27.8% and 65.4% when compared with the barefoot and shod conditions, respectively. All subjects except subject 3 exhibited higher damping ratio when changing from the shod to the insole condition. However, in most of the subjects the damping ratio was decreased from barefoot to shod. None of the observed differences was statistically significant ( p < 0.3794). Range of motion increased by 11.2% and decreased by 5.2%, when compared with the barefoot and shod conditions. ANOVA test showed that these differences were statistically significant ( p = 0.0053), and Tukey-Kramer post-tests revealed that the observed difference was significant ( p < 0.010) when comparing the barefoot to the shod condition only. Most subjects showed their highest range of motion when shod, which was then reduced when fitted with the insoles. The maximum position instant was significantly increased ( p = 0.0027) from barefoot to shod and insole. Post-tests showed that the observed increment was not significant when the shod condition was compared with the insole.

When the velocity variables were considered, the highest maximum velocity was observed under the shod condition. However, the average value of the maximum velocity was decreased when changing from shod to insole, although this was not true for all subjects. Range of velocity exhibited a similar behavior, with the maximum velocity reaching its highest value when the subjects were shod and being reduced when the orthoses were used. ANOVA test showed that these differences were statistically significant ( p = 0.0469), and Tukey-Kramer post-tests revealed that the observed difference was significant ( p < 0.050) when comparing the barefoot to the shod condition only. The maximum velocity instant was significantly increased ( p = 0.0231) from the barefoot to the insole condition, with all subjects showing the highest value when fitted with the orthoses. It must be noted that the differences observed between the barefoot and shod condition were quite small and not statistically significant.

When the acceleration variables were considered, the average value of the maximum acceleration was decreased by 9.9% and slightly increased by 1.9% when the insole condition was compared with the barefoot and shod conditions, respectively. Although subject 2 showed high values of maximum acceleration influencing the average value, all subjects except subject 5 showed consistently higher maximum acceleration when changing from the shod to the insole condition. Range of acceleration showed a similar behavior, with the maximum acceleration exhibiting its highest value when the subjects were barefoot and being reduced when subjects were shod. However, for most of the subjects the range of acceleration was additionally decreased when the insoles were used. Finally, the mean value of the maximum acceleration instant was increased by 6.3% from barefoot to shod, and by 3.7% from shod to insole. The differences observed in the acceleration variables of T12 were not statistically significant.

SECOND SACRAL VERTEBRA

When the insoles were used, the damping ratio was noticeably increased by 57.9% and 35.6% when compared with the barefoot and shod conditions, respectively. However, the observed differences were not statistically significant. Range of motion was increased by 8.4% and decreased by 4.6% when compared with the barefoot and shod conditions. Most subjects exhibited their highest range of motion when shod, which was reduced when fitted with the insoles. The most prominent exception was subject 7, who exhibited the highest range of motion when the insoles were used. The difference observed between the barefoot and shod condition was statistically significant (ANOVA test showed that p = 0.0277 and Tukey-Kramer post-test showed p < 0.05). The maximum position instant was increased by 4.2% and 0.2%, in comparison with the barefoot and shod conditions.

When the velocity variables were considered, the highest maximum velocity was observed under the shod condition. However, most of the subjects exhibited a decreased maximum velocity from shod to insole. Range of velocity was increased when changing from the barefoot condition to both the shod condition and insole condition. The observed decrement from shod to insole was very small. The maximum velocity instant was increased when comparing the insole condition to both the barefoot and shod conditions, with most subjects exhibiting the highest value when fitted with the orthoses. The difference observed between the barefoot and the insole conditions was statistically significant ( p = 0.0231).

When the acceleration variables were considered, the highest maximum acceleration was observed under the insole condition; it was increased by 2.6% and 4.9% when compared with the barefoot and shod conditions, respectively. With regard to the range of acceleration, the use of insoles for most of the subjects resulted in the highest value. Finally, when the subjects were fitted with the orthoses, the maximum acceleration instant was increased, with the higher difference of 13.2% observed between the insole and barefoot conditions. None of the differences observed in the acceleration variables of S2 was statistically significant.

DISCUSSION

The damping ratio for both tested vertebrae was small (typically below 0.05), indicating that T12 and S2 experienced a low damping oscillation. In addition, the formula used to calculate the damping ratio could be reduced to:

When the subjects were fitted with the insoles, the damping ratio of both tested vertebrae was noticeably increased when compared with both the barefoot and shod conditions. This could be associated with the dissipation effect of the orthoses. This seemed to be more important when the thoracic vertebra was considered because the damping ratio was actually decreased from barefoot to shod, indicating that the shock absorption capacity of the shoes alone was not sufficient to result in the desirable increased damping.

The effect of the insoles on the range of motion of both vertebrae was statistically significant. This can be attributed to the shoes alone. The range of motion of both vertebrae was increased from barefoot to shod. On the other hand, most of the subjects exhibited a decreased range of motion of both vertebrae from shod to insole. This may lead to a decrement of the displacement of the body's center of gravity from the line of progression, which is associated with the energy consumption and thus the efficiency of walking. It must be noted that for most subjects, the lowest range of motion of both vertebrae was observed when barefoot.

The maximum position instant for the thoracic vertebra was significantly increased from the barefoot to both the shod and insole conditions. When the behavior of the sacral vertebra was considered, the increments observed from barefoot to both shod and insole conditions were not statistically significant. In general, when the subjects were shod, both tested vertebrae needed more time to reach their highest positions during the stance phase, with the insoles appearing not to have an effect.

When the velocity variables were considered, for both tested vertebrae, the highest average maximum velocity was found under the shod condition, decreasing when the insoles were used. Because the kinetic energy of the body segments is directly proportional to their velocity and walking activity is characterized by a continuous exchange between kinetic and potential energy, the observed trend of the maximum velocity might be related to the way the energy is transferred between one segment and another. Thus, there could be a relation between the observed decrement of the velocity of the vertebrae with a smoother translation of the upper body, which affects favorably the efficiency of walking. However, not all of the subjects exhibited this behavior.

Regarding the range of velocity, both vertebrae followed a similar pattern, with the maximum velocity reaching its highest value when the subjects were shod and decreasing when fitted with the insoles.

The maximum velocity instant of the thoracic vertebra was significantly increased when comparing the insole condition to the barefoot one. In most of the subjects, the shoes did not seem to have a significant effect on this variable, so the use of insoles appeared to be the most important factor. In addition, most subjects exhibited the highest value of the maximum velocity instant of both vertebrae when fitted with the orthoses. A similar behavior was followed by the sacral vertebra, with the differences observed between the barefoot and insole conditions being statistically significant. It must be noted that for most subjects the lowest maximum velocity, range of velocity, and maximum velocity instant of both vertebrae were found when the subjects were barefoot.

A specific pattern in the rate of change of position (i.e., velocity) does not necessarily prejudge the pattern in the rate of change of velocity (i.e., acceleration). For example, the highest maximum velocity of both tested vertebrae was observed under the shod condition. This was not the case for the maximum acceleration, for which the highest value was reached by the thoracic vertebra when the subjects were barefoot, and by the sacral vertebra when the subjects were fitted with the insoles.

The high values of the maximum acceleration of the thoracic vertebra observed under the barefoot condition were decreased when changing to the shod condition. However, for all subjects except one the maximum acceleration of the thoracic vertebra was increased when the orthoses were used.

Similar to the maximum acceleration, the range of acceleration of the thoracic vertebra exhibited the highest value under the barefoot condition, which decreased when subjects were shod, and further decreased when subjects wore the insoles. This was not the case for the sacrum, for which the use of insoles resulted in the highest range of acceleration for most subjects.

In general, statistically significant differences were found for the range of motion, the range of velocity, the maximum position, and the maximum velocity instants of the thoracic vertebra and for the range of motion and the maximum velocity instant of the sacrum. From these differences, the maximum position and the maximum velocity instants of the thoracic vertebra and the maximum velocity instant of the sacrum can be attributed to the use of the orthoses. In addition, the increment of the damping ratio of both vertebrae, which is associated with the dissipation effect of the orthoses, was apparent.

This study presents three important limitations. First, the small sample size does not allow us to generalize the results. Second, averaging procedures were applied to determine the mean values and standard deviations of the derived variables at each test condition. Thus, the reported percentage differences are referred to the mean values of a single data set representing the entire sample. Reporting average values alone eliminates individual variations of subjects on the basic motion pattern and ignores the possibility of significant changes that may have occurred within subjects. For this, the corresponding tables include the data of each individual for all the derived variables, and the related comments are referred to the trend observed within and across subjects. Third, the acclimatization time allowed was relatively short. Although there were some effects to the oscillation pattern of the tested vertebrae 3 weeks after the fitting with the orthoses, the maximal effects may be apparent only after several weeks.

The main aim of the current study was to investigate whether there is any noticeable effect of the separated-arms foot orthoses on the oscillation of T12 and S2. Our results showed that there are some immediate changes of the tested vertebrae kinematics attributed to the use of these insoles. How these changes might contribute to motion-related low back pain prevention or relief remains unclear.

CONCLUSION

Because of the variability among subjects and the small number of subjects involved in this pilot study, the conclusions derived must be investigated more fully, involving a larger population.

Our study showed that there are some systematic changes in the lower spine oscillation pattern when the group of the analyzed individuals was fitted with separated-arms foot orthoses. Statistically significant differences attributed to the use of the insoles were found for the maximum position and the maximum velocity instants of T12 and the maximum velocity instant of S2. In addition, the increment of the damping ratio of both vertebrae, which is associated with the dissipation effect of the orthoses, was apparent. However, none of the differences observed in the tested acceleration variables of both vertebrae were statistically significant.

These findings provide more information to further explore the positive effects of foot orthoses, with regard to the efficiency of gait and the prevention or relief of low back pain and fatigue.

Correspondence to: Paola Catalfamo, Centre for Biomedical Engineering/ School of Engineering, Duke of Kent Building, 12DK03, University of Surrey, Guildford, Surrey, GU2 7TE, United Kingdom; e-mail: p.catalfamo@surrey.ac.uk .

NOTE: The first reference in the list below is actually "1a". Reference 1b is as follows: Fox H, Winson I. Foot orthoses: an audit of expenditure and efficacy. The Foot 1994;4:79–84.

GEORGE MARINAKIS, MEng, MPhil, PhD, CEng, MISPO, is affiliated with the Centre for Biomedical Engineering/School of Engineering and European Institute of Health and Medical Sciences, University of Surrey, Guildford, Surrey, United Kingdom.

PAOLA CATALFAMO, MBEng, is affiliated with the Centre for Biomedical Engineering/School of Engineering, University of Surrey, Guildford, Surrey, United Kingdom. The YPF-Repsol Foundation supports the PhD studies of Paola Catalfamo.

References:

1. The American Podiatric Medical Association, Attitudes Toward Foot Care 2000, Coleman and Associates, Inc., April 2000.
2. Birke J, Foto J, Pfiefer L. Effect of orthosis material hardness on walking pressure in high-risk diabetes patients. J Prosthet Orthot 1999;11:43–47.
3. Eng J, Pierrynowski M. Evaluation of soft foot orthotics in the treatment of patellofemoral pain syndrome. Phys Ther 1993;73:62–68.
4. Hodge M, Bach T, Carter G. Orthotic management of plantar pressure and pain in rheumatoid arthritis. Clin Biomech 1999;14:567–575.
5. Wosk J, Voloshin A. Low back pain. Conservative treatment with artificial shock absorbers. Arch Phys Med Rehabil 1985;66:145–148.
6. Bates B, Osternig L, Mason M, James L. Foot orthotic devices to modify selected aspects of lower extremity mechanics. Am J Sports Med 1979;7:338–342.
7. McCulloch M, Brunt D, Linder D. The effect of foot orthotics and gait velocity on lower limb kinematics and temporal events of stance. J Orthop Sports Phys Ther 1993;17:2–10.
8. Quesada P, Sawyer F. Assessment of efficacy of silicone gel-filled shoe insoles for plantar pressure relief. Gait Posture 1995;3:98–101.
9. Stacoff A, Reinschmidt C, Nigg B, et al. Effects of foot orthoses on skeletal motion during running. Clin Biomech 2000;15:54–64.
10. Voloshin A, Wosk J. Shock absorbing capacity of the human knee. Proceedings of the Special Conference of the Canadian Society of Biomechanics on "Human Locomotion I." Ontario, Canada, October 27–29, 1980.
11. Nester C, van der Linden M, Bowker P. Effect of foot orthoses on the kinematics and kinetics of normal walking gait. Gait Posture 2003;17:180–187.
12. Marinakis G, Flaieh I, Drake C. Biomechanical effects of polymer gel-filled foot orthoses on healthy individuals during normal walking. Abstracts of the 12th Annual Meeting of the European Society for Movement Analysis in Adults and Children. Gait Posture 2003;18(2):86–87.
13. Johnson G. The effectiveness of shock-absorbing insoles during normal walking. Prosthet Orthot Int 1988;12:91–95.
14. Voloshin A, Wosk J. Influence of artificial shock absorbers on human gait. Clin Orthop Rel Res 1981;160:52–56.
15. Pratt D, Rees P, Rodgers C. Assessment of some shock absorbing insoles. Prosthet Orthot Int 1986;10:43–45.
16. Krebs D, Wong D, Jevsevar D, et al. Trunk kinematics during locomotor activities. Phys Ther 1992;72:35–44.
17. Thorstensson A, Nilsson J, Carlson H, Zomlefer M. Trunk movements in human locomotion. Acta Physiol Scand 1984;121:9–22.
18. Thurston A, Harris J. Normal kinematics of the lumbar spine and pelvis. Spine 1983;2:199–205.
19. Vogt L, Pfeifer K, Banzer W. Comparison of angular lumbar spine and pelvis kinematics during treadmill and overground locomotion. Clin Biomech 2002;17:162–165.
20. Marinakis G, Catalfamo P. The effect of separated-arms foot orthoses on the lower body and trunk kinematics, during level walking. J Prosthet Orthot 2004;16:87–93.