During gait, the position of the ground reaction force (GRF) vector relative to lower-limb joint centers is an important aspect of control. In normally developing children, the vector can be seen to pass close to the joint centers in the sagittal plane, minimizing muscular demand.1 Such alignment is not always seen in those with altered neurological control, resulting in abnormal, energy-consuming gait.2 Fixed (rigid and nonarticulating) ankle-foot orthoses can be prescribed to manipulate the alignment of the GRF vector throughout stance. This is achieved by controlling the angle of the anterior border of the tibia (shank) relative to the vertical (SAV) through, for example, adding wedges and altering the sole profile. Orthotic tuning is the process of optimizing the alignment of the GRF to lower-limb joints, using video vector to visualize the vector.3
In the published work available, it is inferred that biomechanical correction is unlikely to be achieved from orthoses “as supplied.” It is suggested that orthotic tuning with video vector is necessary to optimize alignment, because accurate estimation of kinetics cannot be assumed from the observation of kinematics.3,4 Some work proposes that a tuned orthosis will improve gait kinematics,2,5,6 prevent the development of deformities,7 increase range at the ankle (Owen E, London, UK, personal communication, 2005), and provide beneficial sensorimotor learning because of walking in a corrected biomechanical environment.3,5,7,8 However, observations are from small populations, with only three articles reporting a quantitative outcome measure, “moment arm” at the knee, and to determine or illustrate the success of tuning.5,6,8
Owen9 measured the SAV retrospectively from a cohort of children who had been optimally tuned, finding SAV of 10° to 12° at mid-midstance, the event where the knee of the swing leg just progresses past the stance leg. However, no information regarding the SAV of normally developing children is available. “Moment arm” at the knee gives a true indication of abnormal control during walking.8 However, only limited normal data have been published, from up to five subjects.5,6
This study investigated the SAV and “moment arm” at the knee throughout stance in a group of normally developing children. The aim of this study was to provide preliminary evidence for tuning techniques and outcome measures that could be used both clinically and in future research. Data were recorded both barefoot and shod to give an indication of how normally developing children adapt their shank kinematics when footwear with a heel-sole differential (HSD) is worn.
The study was reviewed and given favorable ethical opinion by the University Ethics Committee. Eleven healthy children were recruited as a convenience sample. The group consisted of seven males and four females aged between 5 and 16 years, with mean (standard deviation [SD]) age 10 (4) years. Leg lengths ranged between 495 to 980 mm and had a mean of 735 (140) mm. Footwear worn had a mean HSD of 11 mm (range 0–17 mm), with the majority of subjects wearing footwear subjectively assessed as having reasonably stiff soles, positive back heels, and slightly rounded profiles at the metatarsal heads.
Kinematic data were collected using seven ProReflex cameras (Qualisys AB, Sweden) sampling at 120 Hz. Data regarding the SAV were collected from retroreflective markers placed on the tibial tuberosity and the anterior distal tibia. Further markers were placed on the lateral ankle malleolus and between the second and third metatarsals on the forefoot to allow identification of gait events. Another marker was placed laterally on the knee joint line, midway between the front (excluding the patella) and back of the knee. This provided a reference point for the knee joint center and measurements of the “moment arm”. Figure 1 shows marker placement and the GRF vector on a subject at mid-midstance. With footwear, the forefoot marker was repositioned to the top of the fifth metatarsal head. Data were processed in Qualisys Track Manager (Qualisys AB, Sweden) and exported to Microsoft Excel (Microsoft, Redmond, Washington) for analysis.
Kinetic data were collected from two AMTI force plates (model BP400600HF, Summit Medical, UK). This was integrated with sagittal video footage (collected at 25 Hz) using an in-house developed program in LabVIEW version 5.1 (National Instruments, UK), allowing the GRF vector to be visualized in real time.
All subjects practiced walking with markers until their guardians felt gait was representative. Data were then collected at self-selected normal speed until three traverses with clean force plate contacts were recorded, both barefoot and shod. SAV values were determined from both legs, and “moment arm” data were obtained only from the right leg. Sagittal plane coordinates of the tibial markers were used to calculate the SAV in 2D, because this is the measure used practically when tuning clinically. For six subjects, SAV in 3D was also calculated to understand the errors associated with using 2D data. “Moment arm” measures were made using video analysis software ProTrainer version 6.1 (Sports Motion, Cardiff, CA) and normalized to leg length.
Because of the preliminary nature of the study and the limited sample size, only descriptive statistics are presented in the results. Mean (SD) speed barefoot was 1.39 m/s (0.12 m/s) and shod 1.47 m/s (0.17 m/s). Mean (SD) step length barefoot was 0.63 m (0.15 m) and shod 0.90 m (0.21 m).
Figure 2 shows the calculated SAV throughout stance, both barefoot and shod. The mean (SD) SAV at mid-midstance for the population was 11.4° (3.4°) barefoot and 10.5° (3.6°) shod. The mean (SD) timing of the event of mid-midstance as a percentage of stance was 45% (2%) barefoot and 44% (2%) shod. The mean (SD) difference in SAV calculated in 2D and 3D in six subjects was 1.0° (0.1°) between 40% and 50% of stance both barefoot and shod. “Moment arm” data throughout stance are shown in Figure 3. The mean (SD) normalized “moment arm” at the knee at mid-midstance was 2.5% (1.7%) barefoot and 1.3% (2.4%) shod.
For both SAV and “moment arm” measures, the SDs seen are greater than the differences between the mean values for barefoot and shod data, preventing definite conclusions regarding the effect of footwear.
The SAV shows a pattern of increasing incline throughout stance as the contralateral leg swings forward, allowing the body to progress past the stance foot. Mean SAV at mid-midstance was within 10° to 12° both barefoot and shod, as found by Owen9 in children wearing ankle-foot orthosis, thereby supporting use of this angle range as a starting point when tuning. With footwear, a general trend of increased recline at initial contact and throughout early stance was seen, perhaps because of an increase in step length associated with wearing shoes. Interestingly, despite a HSD, footwear did not lead to increased SAV in mid-midstance.
The pattern seen in Figure 3 for the “moment arm” represents changes between stabilizing extension moments and shock absorbing/body progressing flexion moments. With footwear, a trend of increased extension followed by increased flexion moments was seen. One hypothesis is that positive back heels on footwear offset the GRF vector posteriorly throughout stance. When combined with sole stiffness, this may delay and impede the natural forefoot rocker.1
The magnitude of the SDs seen for both SAV and “moment arm” may be due to natural variation and sources of error introduced including normalization of the gait cycle; identification of initial contact, toe off, and mid-midstance; placement of the knee marker; alignment of the vector on the video image; image resolution; and use of 2D data.10 Error because of the latter in six healthy subjects was just 1° at 40% to 50% of stance, when comparing 2D and 3D estimations of SAV. Further investigation is needed into the differences seen in children with neurological conditions.
The GRF vector approach only provides an estimation of the “moment arm” and moments, as inertial and gravitational effects are ignored. However, measurement at mid-midstance seems acceptable as errors because of segment inclination and acceleration are minimal around 40% to 50% stance.11 The mean values and pattern of the “moment arm” at the knee throughout stance for the group were similar to the previously published work,5,6 indicating that its use as an outcome measure may be appropriate. However, the SDs seen (both group and within-subject) suggest that in a normal population, a mean measure from a number of walks should be used. Further work is required to establish the appropriateness of determining “moment arm” from just one gait cycle in children requiring tuning or alternative reliable and practical clinical methods.
This preliminary study identifies patterns of the SAV and “moment arm” at the knee, throughout stance, in normally developing children. Measures of “moment arm” at the knee were similar to the previously published work from smaller cohorts, indicating that its use as an outcome measure for tuning may be appropriate, especially if measures are taken over more than one gait cycle. Measures of SAV were taken from the front border of the shank as per tuning clinic protocols where measurements are commonly made from this position off-screen. Normal measurements of SAV measured in this way were previously unpublished. Differences seen between barefoot and shod data are likely to be due to the sole profile of the footwear and an increase in step length when worn. This information can be used to increase the understanding of orthotic tuning protocols, aid their development, and provide a point of reference to which measures from patients can be compared, in an area where there is an increasing awareness by clinicians but limited published work.
The study was reviewed and given favorable ethical opinion by the University of Surrey Ethics Committee.
1. Perry J. Gait Analysis: Normal and Pathological Function
. New Jersey: SLACK Inc; 1992.
2. Meadows CB. The scientific basis of treatment III: to improve the dynamic efficiency of gait. In: Report of a Consensus Conference on the Lower Limb Orthotic Management of Cerebral Palsy
. Denmark: International Society for Prosthetics and Orthotics (ISPO); 1995.
3. Butler PB, Nene AV. The biomechanics of fixed ankle foot orthoses and their potential in the management of cerebral palsied children. Physiotherapy
4. Stallard J. Assessment of the mechanical function of orthoses by force vector visualisation. Physiotherapy
5. Butler PB, Thompson N, Major RE. Improvement in walking performance of children with cerebral palsy: preliminary results. Dev Med Child Neurol
6. Butler PB, Farmer SE, Stewart C, et al. The effect of fixed ankle foot orthoses in children with cerebral palsy. Disabil Rehabil Assist Technol
7. Major RE. Current orthotic practice in relation to the improvement of gait efficiency. In: Report of a Consensus Conference on the Lower Limb Orthotic Management of Cerebral Palsy
. Denmark: International Society for Prosthetics and Orthotics (ISPO); 1995.
8. Butler PB, Farmer SE, Major RE. Improvement in gait parameters following late intervention in traumatic brain injury: a long-term follow-up report of a single case. Clin Rehabil
9. Owen E. Shank angle to floor measures of tuned “ankle-foot orthosis footwear combinations” used with children with cerebral palsy, spina bifida, and other conditions. Gait Posture
10. Pratt E. A review of orthotic tuning techniques and the development of a normals database [dissertation]. Surrey: School of Engineering, University of Surrey; 2006.
11. Wells RP. The projection of the ground reaction force as a predictor of internal joint moments. Bull Prosthet Res