- Email: [email protected]
Upper extremity motion during gait in adolescents with structural leg length discrepancy—An exploratory study
Accepted Manuscript Title: Upper Extremity Motion during Gait in Adolescents with Structural Leg Length Discrepancy—an Exploratory Study Author: Fabiola Angelico Marie Freslier Jacqueline Romkes Reinald Brunner Stefan Schmid PII: DOI: Reference:
S0966-6362(17)30004-8 http://dx.doi.org/doi:10.1016/j.gaitpost.2017.01.003 GAIPOS 5282
To appear in:
Gait & Posture
Received date: Revised date: Accepted date:
3-10-2016 23-11-2016 3-1-2017
Please cite this article as: Angelico Fabiola, Freslier Marie, Romkes Jacqueline, Brunner Reinald, Schmid Stefan.Upper Extremity Motion during Gait in Adolescents with Structural Leg Length Discrepancy—an Exploratory Study.Gait and Posture http://dx.doi.org/10.1016/j.gaitpost.2017.01.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Upper Extremity Motion during Gait in Adolescents with Structural Leg Length Discrepancy – an Exploratory Study
Fabiola ANGELICO1,2, Marie FRESLIER1, Jacqueline ROMKES1, Reinald BRUNNER1,3, Stefan SCHMID4,*
1
University of Basel Children’s Hospital, Laboratory for Movement Analysis, Basel,
Switzerland 2
Zurich University of Applied Sciences, School of Health Professions, Winterthur,
Switzerland 3
University of Basel Children’s Hospital, Orthopaedic Department, Basel, Switzerland
4
Bern University of Applied Sciences, Health Division, Bern, Switzerland
Corresponding author: *
Stefan Schmid, PT, PhD, Bern University of Applied Sciences, Health Division,
Murtenstrasse 10, 3008 Bern, Switzerland, +41 79 936 74 79, [email protected]
Word count: Abstract: 248, Text (without acknowledgements and references): 2905
1
RESEARCH HIGHLIGHTS
First study investigating upper limb motion in LLD patients during gait
LLD patients indicated secondary deviations in upper limb motion
Increased shoulder extension as passive physical effect of trunk position
Active strategies might be involved in controlling total body angular momentum
Basis for further investigations using complex musculoskeletal models
ABSTRACT Background and Purpose: Depending on the extent of a structural leg length discrepancy (LLD), several compensatory mechanisms take place in order to maintain function and to optimize energy consumption during gait. However, studies describing the influence of a structural LLD on upper limb motion are lacking. The current study therefore aimed at the evaluation of upper limb motion during gait in LLD patients compared to healthy controls. Methods: Motion capture data from 14 patients with structural LLD and 15 healthy controls that were collected during barefoot walking at a self-selected speed were retrospectively analyzed. Specifically, kinematic parameters of the shoulder and elbow joints as well as the trunk segment were investigated and considered in relation to a minimal clinically important difference of 5°. Results: The shoulders in LLD patients were kept constantly in a more extended and at initial contact in a more adducted position as compared to healthy controls. In addition, the patients’ elbow joints showed increased flexion motion and the trunk segment indicated a constant trunk lateral-flexion and axial rotation tendency towards the affected side. Conclusions: Patients with structural LLD indicated clinically relevant secondary deviations in shoulder and elbow motion. While some of these deviations were most likely passive physical effects, others might have occurred as active strategies to maintain balance or to regulate total body angular momentum. These findings contribute to the understanding of 2
secondary gait deviations induced by a structural LLD and might serve as a basis for further investigations using complex musculoskeletal models.
Keywords: LLD; Upper body; Arm; Secondary deviations
INTRODUCTION Leg length discrepancy (LLD) is a relatively common problem found in 40-70% of the population, whereby about one in every 1000 persons presents with a LLD of more than 20 mm [1]. Complications of LLD include functional limitations such as gait and balance problems and associated musculoskeletal disorders such as low back pain or stress fractures [1]. Depending on the extent of the LLD, pelvic obliquity was observed as a consequence during gait, whereas increased ankle plantarflexion (equinus and vaulting) on the affected side (shorter leg), persistent shortening (knee flexion and ankle dorsiflexion) and circumduction on the unaffected side (longer leg) and lateral bending of the lumbar and thoracic spine were described to occur as compensations for the unequal leg length [2-5]. No study, however, could be identified describing the influence of a structural LLD on upper limb motion in this population. During normal gait, Perry [6] defines the movement pattern of the upper limbs as a reaction to the thorax movement, with the arms swinging contrariwise to the rest of the body, whereas the range of motion of the upper limb joints is symmetric between left and right, but varies from person to person and is influenced by gait speed. For a group of healthy adolescents, Romkes et al. [7] reported shoulder movements ranging from 17.4° of extension to 6.5° of flexion and elbow movements from 29.9° to 45.9° of flexion. Based on experiments investigating the functional contribution of upper limb muscles during normal gait, it was postulated that arm swing has both passive and active components and occurs most likely to reduce energy expenditure [8-10]. When comparing walking with constrained arms to normal gait, energy 3
consumption was increased by 8% [11]. Therefore, it was suggested that not only a pathology itself, but also an altered arm swing could lead to a higher energy consumption [12]. Focusing on the effects of walking with a restricted arm swing, Bruijn et al. [13] showed that arm swing did not contribute to gait stability in normal gait, but helped to regain balance and hence contributed to an overall stability following a perturbation. In children with cerebral palsy, an altered movement pattern of the upper limbs was suggested to occur as compensation and to play a role in gaining stability and balance [7, 14-18]. Given the role of arm swing during gait as well as the fact that oxygen consumption was reported to be greater in individuals with an artificially induced LLD [19], the current study aimed at exploring upper limb motion in adolescents with a structural LLD compared to a group of matched healthy controls. Such knowledge might contribute to a better understanding of gait adaptations throughout the whole body and might therefore influence the planning of therapeutic interventions.
METHODS Participants: Data from fourteen adolescents with a structural LLD that were measured in the movement analysis laboratory of a university children’s hospital between 2006 and 2014 were included in this retrospective analysis (Table 1). Inclusion criteria for the patients were: aged between 10 and 18 years, diagnosis of structural LLD of greater than 1% of body height and at least 20 mm, no neurological basic pathology, no obesity (>95th BMI per-age percentile), no diagnosed structural deformities of the spine and no injuries to the locomotor system that led to persistent functional deficits. Hence, LLD patients were only included when demonstrating no deficits in joint range of motion, muscle length and muscle strength as well as no signs of spasticity in the lower extremities. This information was extracted from the report of the standard clinical examination that was carried out prior to the data collection in LLD patients. 4
For comparative purposes, data from fifteen matched healthy adolescents that were measured in the same laboratory were used. The study was approved by the local ethics committee and written informed consent was obtained from all participants and their legal guardians.
Data collection: The LLD was measured from the anterior superior iliac spine via the patella to the medial malleolus using a tape measure [20]. In order to evaluate upper body kinematics, retroreflective markers were placed on the skin according to the Plug-in Gait full body model as described by Gutierrez et al. [21]. Data was collected using a Vicon motion capture system with 6-12 infrared cameras (Oxford Metrics, London, UK). All participants were asked to walk barefoot at a self-selected normal speed along a 10 meter walkway. Measurements were repeated until data from at least five trials were obtained.
Data reduction: The data was analyzed using the Vicon Workstation, Nexus and the Polygon software (Oxford Metrics, London, UK). For every trial, visual setting of gait events was performed by an experienced assessor. The kinematics of the shoulder joint (flexion/extension and abduction/adduction), elbow joint (flexion/extension) and trunk segment (anterior/posterior tilt, lateral-flexion and axial rotation) were calculated according to the Plug-in Gait full body model and expressed in degrees [°]. For each participant, average profiles over five trials were generated and parameterized accordingly. The parameters of interest for the upper extremity joint and trunk angles were: value at initial contact (IC) as well as average (AVG) and range of motion (ROM) values.
Statistical analyses: 5
All statistical analyses were performed using SPSS version 22 (SPSS Inc., Chicago, IL, USA) and G*Power 3.1.9.2 [22]. Normal distribution of the majority of the subject demographic and outcome parameters was confirmed with the Shapiro-Wilk test (p>0.05). Subject demographics were compared using independent samples T-tests (statistically significant difference: p≤0.05). Comparisons of the shoulder and elbow kinematic parameters (primary outcomes) between the patients’ affected and unaffected and the healthy controls’ left side were carried out using one-way analyses of variance (ANOVA) with Tukey HSD post hoc tests, whereas comparisons of the trunk kinematic parameters (secondary outcomes) between the patients and healthy controls were carried out using independent samples T-tests. In order to account for the explorative nature of the study, the interpretation of the results was thereby mainly based on effect sizes (Cohen’s f and d) rather than p-values. Post hoc tests for the primary outcomes were thus only conducted in case of considerable effects of the ANOVA (f≥0.1) [23]. As previously suggested and extensively described [24, 25], mean differences were subsequently evaluated for their clinical relevance using a minimal clinically important difference (MCID) of 5°. Differences showing a large effect (d≥0.8) were thereby considered clinically relevant with the absolute mean difference being above the MCID and a tendency with the mean difference being below the MCID. Shoulder parameters that showed clinically relevant differences were further evaluated for a possible relationship with trunk kinematics using linear regression analyses.
RESULTS No significant differences were found for age, weight, height and gait speed between the LLD patients and the healthy controls (Table 1).
Primary outcomes
6
The statistical analysis showed considerable overall effects for several upper extremity parameters (Table 2). On the patients’ affected side, the shoulder showed clinically relevant greater extension angles at IC (mean difference (95% confidence interval): -13.4° (-19.4°,7.4°), d=-1.96; p<0.001) and on average (-7.1° (-12.5°,-1.8°), d=-1.28; p=0.007) as well as smaller abduction angles at IC (-5.4° (-10.6°,-0.2°), d=-0.92; p=0.040) as compared to the controls (Figure 1). Furthermore, patients showed clinically relevant increases for the average elbow flexion angle (8.3° (2.8°,13.9°), d=1.52; p=0.002) as well as an increased elbow ROM (19.5° (10.8°,28.3°), d=2.22; p<0.001) on the affected side compared to the control group. The shoulder on the patients’ unaffected side showed an increased extension angle at IC (8.9° (-15.0°,-2.9°), d=-1.38; p=0.002) and on average (-9.3° (-14.7°,-3.9°), d=-1.58; p<0.001) compared to the shoulder in the control group. In addition, the patients showed a clinically relevant decrease in the abduction angle at IC (-6.4° (-11.5°,-1.2°), d=-1.21; p=0.013) between the unaffected side and the controls. Similar to the patients’ affected side, the elbow joint on the unaffected side revealed clinically relevant increases for the flexion angle on average (10.0° (4.5°,15.5°), d=1.63; p<0.001) and ROM (18.0° (9.3°,26.7°), d=2.11; p<0.001) as compared to controls. The patients’ affected side showed clinically relevant greater shoulder ROM (12.3° (2.5°,22.1°), d=1.19; p=0.011) compared to the unaffected side. Regression analyses revealed a statistically significant coefficient of determination for the shoulder flexion/extension and trunk tilt angles at IC on the affected side (R2=0.306, p=0.040) (Table 3). The qualitative examination of the upper extremity joint angles indicated that the shoulder on the patients’ unaffected side remained constantly in extension (Figure A.1 in the electronic supplementary material). In addition, the frontal plane shoulder angles indicated a time shift of the instant of peak abduction angle between the two groups. While the healthy controls
7
reached their peak abduction angle during loading response, LLD patients reached it in late midstance.
Secondary outcomes The statistical analysis of the trunk angles yielded large effects for differences in tilt ROM (d=1.63, p<0.001), lateral-flexion at IC on the affected side (d=-0.99, p=0.017), lateral-flexion average and ROM (d=-1.07, p=0.008 and d=0.90, p=0.023, respectively) as well as axial rotation at IC on the unaffected side (d=-1.35, p=0.001) (Table 4). However, due to mean differences of less than 5°, they were only considered tendencies rather than clinically relevant. In addition, the qualitative examination of the trunk segment angles indicated that the trunk in LLD patients tended to be more upright during loading response (less anterior trunk tilt) as well as in constant lateral-flexion and axial rotation towards the affected side compared to the healthy controls (Figure A.1).
DISCUSSION The aim of the current study was to explore upper extremity kinematics during gait in adolescents with a structural LLD in comparison to a group of matched healthy controls. The analysis indicated that, although the trunk showed no clinically relevant differences, the LLD patients adopted clinically relevant secondary deviations in sagittal shoulder and elbow as well as frontal shoulder kinematics. The question whether these secondary deviations occurred as passive physical effects of the primary deviations or as active compensatory mechanisms to offset primary deviations and the physical effects [26], however, will be discussed in the following. The deviations found in the LLD patients’ sagittal plane shoulder angles at IC on the affected side might have resulted from a slightly more upright trunk position during loading response 8
and could therefore be considered a passive physical effect. In addition, it could be assumed that the observed tendencies of trunk lateral-flexion and axial rotation towards the affected side would have resulted in increased shoulder extension as well as decreased shoulder abduction angles on the unaffected side. However, since no relevant correlations were found between these parameters, this assumption could not be supported by the current results. Furthermore, the decreased shoulder abduction angle at IC on the affected side might have occurred as a compensatory strategy for maintaining lateral balance after “stepping down” on the shorter leg. This strategy has been previously described in healthy young and middle-aged individuals following induced lateral perturbations [27]. However, it remains unclear whether patients with a structural LLD would react in the same way as healthy individuals and since dynamic balance/gait stability was not evaluated in the current investigation, this assumption should be treated with caution. The increased elbow flexion motion might be interpreted as a passive reaction to a slightly increased sagittal shoulder motion using the double-pendulum model [28]. Another plausible explanation for the observed deviations in the upper extremities in LLD patients could be the regulation of total body angular momentum. As reported by Thielemans et al. [29], changes in angular momentum of one limb can be compensated through changes in angular momentum of the other limbs. It might therefore be possible that the LLD patients adjusted their arm swing in order to maintain a normal total body angular momentum in spite of the unequal leg length. The fact that only tendencies and no clinically relevant deviations were found for the trunk can be explained by previously described compensatory mechanisms occurring in the lower extremities, the pelvis and the spine in order to equilibrate leg length, minimize the displacement of the body’s center of mass and reduce energy expenditure [2-5]. In the late 90’s, Song et al. [4] investigated lower extremity and pelvis kinematics in a group of adolescents with structural LLDs of 0.6-11.1 cm and documented compensatory strategies 9
such as pelvis obliquity as well as toe-walking and, for the ones that walked plantigrade, vaulting, circumduction and persistent flexion of the longer limb. Almost 20 years later, Aiona et al. [2] investigated a similar cohort of LLD patients and reported pelvis obliquity, increased ankle plantarflexion on the shorter side (equinus or vaulting) as well as knee flexion on the longer side as compensations for the unequal leg length. Studies conducted on healthy individuals with artificially induced LLDs of 1-5 cm found pelvis obliquity and changes in the sagittal plane kinematics of the ankle, knee and hip joints (i.e. prolonging the shorter leg and shortening the longer leg) as well as lateral bending of the lumbar and thoracic spine [3, 5]. However, Song et al. [4] reported that patients with LLDs of less than 3% of the longer limb length (approximately 2.2 cm) showed no compensatory strategies. Similarly, Walsh et al. [5] found that several LLDs up to 2.2 cm did either not require any compensatory strategies or were compensated by pelvis obliquity and did not require any functional adaptations in the lower extremities. Since the minimal LLD in the current study was 3 cm, it can be expected that all patients displayed compensations in the lower extremities, the pelvis and the trunk/spine. It has to be considered though that these authors did not distinguish between active and passive mechanisms and interpreted several secondary deviations as compensations. According to Schmid et al. [26], however, such a distinction might be of high importance for the planning of a physical therapy treatment, since compensations are active neuromuscular processes and might therefore be controlled by voluntary actions, whereas physical effects are given based on the laws of physics and might therefore not be corrected. When considering this distinction, pelvic obliquity, such as found by Walsh et al. [5] in patients with LLDs up to 2.2 cm, would most likely be classified as a passive physical effect, occurring as a logical physical consequence of an unequal leg length that is not or not sufficiently compensated for by active neuromuscular processes in the lower extremity joints.
10
A possible explanation for an absence of compensatory strategies in patients with rather small LLDs (< 5 cm) could be that the “cost” of achieving a level pelvis is simply too great (increased work on the long side) and that these patients are therefore willing to tolerate some amount of pelvic obliquity [2]. This might have a direct impact on upper body movement since an uncompensated pelvic obliquity would most likely cause the spine and the upper extremities to actively react, whereas a fully or partially compensated pelvic obliquity would cause less deviations in the upper body. In addition, Aiona et al. [2] showed that LLD patients applied different lower limb compensations depending on the amount of LLD as well as the location of the predominant difference (i.e. femur or tibia). Therefore, it cannot be excluded that the compensatory strategies in the upper extremities also depended on these factors. An important factor that has to be always considered when investigating gait kinematics is walking speed. Previous research showed that gait characteristics such as spatio-temporal parameters, ground reaction forces as well as joint angles and moments were significantly affected by speed in healthy children [30]. Taking gait speed into account, a possible speeddependency of the observed kinematic group differences in the current study can be excluded. Furthermore, the current kinematic findings seem to be highly comparable to those of Song et al. [4] and Aiona et al. [2], since the gait speeds reported in their studies (1.23 m/s and 1.2 m/s, respectively) appear to be in agreement with the current ones. On the other hand, when comparing the upper extremity joint angles of the control subjects in the current study with the ones of Romkes et al. [7], it appears that there are slight differences (especially in frontal plane shoulder motion) that might limit the interpretation of the current findings. However, it has to be considered that the sample size of the control group in the above mentioned study was significantly smaller resulting in less representative values of the control parameters. The fact that the length of the upper limb segments was not taken into account might be considered another limitation. It can therefore not be excluded that some of
11
the observed deviations in upper limb motion were merely the result of a structural arm length difference. Further insights into the neuromuscular control mechanisms of the upper limbs in LLD patients might be gained using complex musculoskeletal models that allow the simulation of muscular force contributions and joint loading. Such computations, however, do not derive from conventional gait analysis. In conclusion, adolescents with a structural LLD of greater than 1% of body height indicated clinically relevant secondary deviations in shoulder and elbow motion. The increased shoulder extension angle on the affected side was considered to have occurred as a passive physical effect of the slightly more upright trunk position during loading response. The other changes in shoulder and elbow motion might possibly have taken place as an active strategy in order to maintain balance or to regulate total body angular momentum. These findings contribute to the understanding of secondary gait deviations induced by a structural LLD and might serve as a basis for further investigations using complex musculoskeletal models.
CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest.
Acknowledgments: The authors acknowledge Dr. Andreas Krieg for assistance in the patient recruitment process and Dr. Katrin Bracht-Schweizer for assistance in the data collection.
REFERENCES [1] Gurney B. Leg length discrepancy. Gait Posture. 2002;15:195-206.
12
[2] Aiona M, Do KP, Emara K, Dorociak R, Pierce R. Gait patterns in children with limb length discrepancy. J Pediatr Orthop. 2015;35:280-4. [3] Kakushima M, Miyamoto K, Shimizu K. The Effect of Leg Length Discrepancy on Spinal Motion during Gait: Three-Dimensional Analysis in Healthy Volunteers. Spine (Phila Pa 1976). 2003;28:2472-6. [4] Song KM, Halliday SE, Little DG. The effect of limb-length discrepancy on gait. J Bone Joint Surg Am. 1997;79:1690-8. [5] Walsh M, Connolly P, Jenkinson A, O'Brien T. Leg length discrepancy - an experimental study of compensatory changes in three dimensions using gait analysis. Gait Posture. 2000;12:156-61. [6] Perry J, Burnfield JM. Gait Analysis: Normal and Pathological Function. 2nd ed. Thorofare, NJ: SLACK Incorporated; 2010. [7] Romkes J, Peeters W, Oosterom AM, Molenaar S, Bakels I, Brunner R. Evaluating upper body movements during gait in healthy children and children with diplegic cerebral palsy. J Pediatr Orthop B. 2007;16:175-80. [8] Goudriaan M, Jonkers I, van Dieen JH, Bruijn SM. Arm swing in human walking: what is their drive? Gait Posture. 2014;40:321-6. [9] Kuhtz-Buschbeck JP, Jing B. Activity of upper limb muscles during human walking. J Electromyogr Kinesiol. 2012;22:199-206. [10] Meyns P, Bruijn SM, Duysens J. The how and why of arm swing during human walking. Gait Posture. 2013;38:555-62. [11] Umberger BR. Effects of suppressing arm swing on kinematics, kinetics, and energetics of human walking. J Biomech. 2008;41:2575-80. [12] Collins SH, Adamczyk PG, Kuo AD. Dynamic arm swinging in human walking. Proc Biol Sci. 2009;276:3679-88.
13
[13] Bruijn SM, Meijer OG, Beek PJ, van Dieen JH. The effects of arm swing on human gait stability. J Exp Biol. 2010;213:3945-52. [14] Delabastita T, Desloovere K, Meyns P. Restricted Arm Swing Affects Gait Stability and Increased Walking Speed Alters Trunk Movements in Children with Cerebral Palsy. Front Hum Neurosci. 2016;10:354. [15] Galli M, Cimolin V, Albertini G, Piccinini L, Turconi AC, Romkes J, et al. Kinematic analysis of upper limb during walking in diplegic children with Cerebral Palsy. Eur J Paediatr Neurol. 2013. [16] Meyns P, Desloovere K, Van Gestel L, Massaad F, Smits-Engelsman B, Duysens J. Altered arm posture in children with cerebral palsy is related to instability during walking. Eur J Paediatr Neurol. 2012;16:528-35. [17] Meyns P, Duysens J, Desloovere K. The arm posture in children with unilateral Cerebral Palsy is mainly related to antero-posterior gait instability. Gait Posture. 2016;49:132-5. [18] Meyns P, Van Gestel L, Massaad F, Desloovere K, Molenaers G, Duysens J. Arm swing during walking at different speeds in children with Cerebral Palsy and typically developing children. Res Dev Disabil. 2011;32:1957-64. [19] Gurney B, Mermier C, Robergs R, Gibson A, Rivero D. Effects of limb-length discrepancy on gait economy and lower-extremity muscle activity in older adults. J Bone Joint Surg Am. 2001;83-A:907-15. [20] Jamaluddin S, Sulaiman AR, Imran MK, Juhara H, Ezane MA, Nordin S. Reliability and accuracy of the tape measurement method with a nearest reading of 5 mm in the assessment of leg length discrepancy. Singapore Med J. 2011;52:681-4. [21] Gutierrez EM, Bartonek A, Haglund-Akerlind Y, Saraste H. Centre of mass motion during gait in persons with myelomeningocele. Gait Posture. 2003;18:37-46.
14
[22] Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39:175-91. [23] Cohen J. Statistical Power Analysis for the Behavioral Sciences. Revised ed. New York, NY: Academic Press, Inc.; 1977. [24] Schmid S, Romkes J, Taylor WR, Lorenzetti S, Brunner R. Orthotic correction of lower limb function during gait does not immediately influence spinal kinematics in spastic hemiplegic cerebral palsy. Gait Posture. 2016;49:457-62. [25] Schmid S, Studer D, Hasler C-C, Romkes J, Taylor WR, Lorenzetti S, et al. Quantifying Spinal Gait Kinematics using an Enhanced Optical Motion Capture Approach in Adolescent Idiopathic Scoliosis. Gait Posture. 2016;44:231–7. [26] Schmid S, Schweizer K, Romkes J, Lorenzetti S, Brunner R. Secondary gait deviations in patients with and without neurological involvement: A systematic review. Gait Posture. 2013;37:480-93. [27] Allum JH, Carpenter MG, Honegger F, Adkin AL, Bloem BR. Age-dependent variations in the directional sensitivity of balance corrections and compensatory arm movements in man. J Physiol. 2002;542:643-63. [28] Kubo M, Wagenaar RC, Saltzman E, Holt KG. Biomechanical mechanism for transitions in phase and frequency of arm and leg swing during walking. Biol Cybern. 2004;91:91-8. [29] Thielemans V, Meyns P, Bruijn SM. Is angular momentum in the horizontal plane during gait a controlled variable? Hum Mov Sci. 2014;34:205-16. [30] van der Linden ML, Kerr AM, Hazlewood ME, Hillman SJ, Robb JE. Kinematic and kinetic gait characteristics of normal children walking at a range of clinically relevant speeds. J Pediatr Orthop. 2002;22:800-6.
15
FIGURE CAPTIONS Figure 1: Primary outcomes: Results of the pairwise group comparisons (mean differences with 95% confidence intervals) for the kinematic parameters of the shoulder and elbow joints (expressed in degrees) between leg length discrepancy (LLD) patients (affected and unaffected sides) and healthy controls. The dashed vertical lines represent the minimal clinically important difference of 5°. Differences were categorized as clinically relevant (large black squares) and clinically not relevant (small white squares). IC: initial contact, AVG: average; ROM: range of motion.
16
Table 1: Demographics of the leg length discrepancy (LLD) patients and healthy controls expressed as means, standard deviations (SD) and ranges (in brackets). Group comparisons were conducted using independent samples T-tests (statistical significance p0.05). LLD Patients
Controls
(n = 14)
(n = 15)
Age [years]
14.3 SD 1.6 (11-17)
14.1 SD 1.7 (12-17)
0.804
Height [m]
1.66 SD 0.08 (1.52-1.84)
1.62 SD 0.10 (1.46-1.84)
0.245
Mass [kg]
60.5 SD 12.7 (36-79)
54.2 SD 10.0 (40-70)
0.153
Gender [male/female]
6/8
8/7
-
Affected (shorter) leg [right/left]
7/7
-
-
LLD [mm]
47.5 SD 16.1 (30-85)
3.0 SD 4.1 (0-10)
<0.001
LLD [% of height]
2.9 SD 1.0 (1.8-5.2)
0.2 SD 0.2 (0.0-0.6)
<0.001
Gait speed [m/s]
1.2 SD 0.2 (0.9-1.6)
1.2 SD 0.1 (1.1-1.5)
0.961
p-value
17
Table 2: Primary outcomes: Reported are mean, standard deviation (SD) and 95% confidence intervals (within the brackets) for the kinematic parameters of the shoulder and elbow joints (expressed in degrees) in patients with structural leg length discrepancy (LLD) and matched healthy controls. In addition, effect sizes (Cohen’s f and d, respectively) as well as results for the group comparisons (one-way analyses of variance (ANOVA) and Tukey’s post hoc tests) are presented. 1)
2) 3)
LLD patients
LLD patients
Post hoc (Tukey) ANOVA 1-3
2-3
1-2
f
d
d
d
Controls (affected side)
(unaffected side)
-23.8 SD 7.0 (-27.8,-19.7)
-19.3 SD 6.4 (-23.0,-15.7)
-10.4 SD 6.7 (-14.1,-6.7)
0.84**
-1.96++
-1.38++
-0.66
AVG
-9.7 SD 5.9 (-13.1,-6.3)
-11.9 SD 6.5 (-15.6,-8.1)
-2.5 SD 5.4 (-5.6,0.5)
0.68**
-1.28++
-1.58++
0.35
ROM
29.8 SD 12.7 (22.4,37.1)
17.5 SD 7.1 (13.4,21.6)
20.6 SD 11.3 (14.3,26.9)
0.48**
0.76
-0.33
1.19++
IC
7.2 SD 6.6 (3.4,11.0)
6.2 SD 5.5 (3.0,9.4)
12.6 SD 5.1 (9.7,15.4)
0.49**
-0.92++
-1.21++
0.16
AVG
9.7 SD 5.6 (6.5,12.9)
8.7 SD 5.1 (5.7,11.6)
11.3 SD 3.9 (9.2,13.5)
0.22*
-0.34
-0.58
0.19
ROM
8.1 SD 5.2 (5.1,11.1)
7.4 SD 3.3 (5.5,9.3)
6.1 SD 3.5 (4.1,8.0)
0.21*
0.47
0.39
0.17
IC
28.6 SD 4.9 (25.8,31.4)
30.5 SD 7.5 (26.2,34.9)
28.9 SD 4.9 (26.2,31.6)
0.14*
-0.06
0.26
-0.31
AVG
37.6 SD 6.0 (34.1,41.1)
39.3 SD 7.2 (35.1,43.5)
29.3 SD 4.9 (26.6,32.0)
0.72**
1.52++
1.63++
-0.26
ROM
25.9 SD 11.7 (19.2,32.7)
24.4 SD 11.2 (17.9,30.9)
6.4 SD 4.8 (3.7,9.1)
0.93**
2.22++
2.11++
0.13
IC Flex./ Ext. Shoulder Abd./ Add.
Elbow
Flex./ Ext.
Flex.: flexion, Ext.: extension, Abd.: abduction, Add.: adduction, IC: initial contact, ROM: range of motion, AVG: average Positive/negative values: flexion/extension (shoulder and elbow), abduction/adduction (shoulder) *
Considerable effect size for ANOVA (f≥0.1) without statistical significance (p>0.05), post hoc tests indicated Considerable effect size for ANOVA (f≥0.1) with statistical significance (p≤0.05), post hoc tests indicated
** +
Large effect size for post hoc (Tukey) without statistical significance (p>0.05) Large effect size for post hoc (Tukey) with statistical significance (p≤0.05)
++
18
Table 3: Results of the regression analyses (expressed as coefficients of determination, R2) for the shoulder parameters that showed clinically relevant differences in adolescents with a structural leg length discrepancy. Trunk (affected side) Tilt
Shoulder (affected side)
Flex./ Ext. Abd./ Add.
Shoulder (unaffected side)
Flex./ Ext. Abd./ Add.
Lat.-Flex.
Axial Rot.
IC
AVG
IC
AVG
IC
AVG
0.306*
-
0.228
-
0.002
-
-
0.190
-
0.018
-
0.030
IC
0.000
-
0.055
-
0.047
-
IC
0.002
-
0.000
-
0.004
-
-
0.069
-
0.020
-
0.004
0.000
-
0.003
-
0.091
-
IC AVG
AVG IC
Flex.: flexion, Ext.: extension, Abd.: abduction, Add.: adduction, Lat.-Flex.: lateral-flexion, Axial Rot.: axial rotation, IC: initial contact, AVG: average * Statistically significant coefficient of determination (p<0.05) Table 4: Secondary outcomes: Reported are mean, standard deviation (SD) and 95% confidence intervals (within the brackets) for the kinematic parameters of the trunk segment (expressed in degrees) in patients with structural leg length discrepancy (LLD) and matched healthy controls. In addition, effect sizes (Cohen’s d) and results for the group comparisons (independent samples T-tests) are presented.
19
1) LLD patients
2) Controls
Mean-difference (1 - 2)
T-test (d)
IC (affected)
3.3 SD 5.3 (0.3,6.3)
-2.7 (-7.3,1.8)
-0.58
0.2 (-3.5,4.0)
0.05
6.0 SD 4.2 (3.7,8.3) IC (unaffected)
6.2 SD 5.6 (3.0,9.5)
AVG
4.8 SD 5.5 (1.6,7.9)
5.4 SD 4.1 (3.1,7.6)
-0.6 (-5.2,4.0)
-0.12
ROM
5.0 SD 1.5 (4.1,5.9)
2.9 SD 0.9 (2.4,3.5)
2.1 (0.8,3.4)
1.63++
-1.9 (-4.0,0.3)
-0.99++
-1.6 (-3.4,0.3)
-0.62
Tilt
IC (affected)
-1.4 SD 2.4 (-2.7,0.0)
IC (unaffected)
-1.0 SD 3.2 (-2.9,0.8)
AVG
-1.2 SD 2.5 (-2.6,0,3)
1.0 SD 1.3 (0.3,1.7)
-2.2 (-4.1,-0.2)
-1.07++
ROM
5.1 SD 2.6 (3.6,6.7)
3.3 SD 1.3 (2.6,4.0)
1.8 (-0.2,3.9)
0.90++
IC (affected)
1.8 SD 4.2 (-0.6,4.3)
2.6 (-0.2,5.4)
-0.70
4.1 (-1.8,6.5)
-1.35++
0.5 SD 1.3 (-0.2,1.2) Trunk
Lat.-Flex.
-0.8 SD 3.2 (-2.5,1.0) IC (unaffected)
3.4 SD 3.0 (1.7,5.1)
AVG
2.6 SD 3.5 (0.6,4.6)
0.6 SD 2.2 (-0.6,1.8)
2.0 (-0.8,4.7)
0.68
ROM
6.8 SD 2.7 (5.3,8.3)
5.7 SD 2.0 (4.6,6.8)
1.1 (-1.1,3.3)
0.47
Axial Rot.
Lat.-Flex.: lateral-flexion, Axial Rot.: axial rotation, IC: initial contact, ROM: range of motion, AVG: average Positive/negative values: anterior/posterior tilt, contra (unaffected side)-/ipsilateral (affected side) flexion, external (affected side)/internal (unaffected side) axial rotation +
Large effect size for T-test without statistical significance (p>0.05) Large effect size for T-test with statistical significance (p≤0.05)
++
20
S0966-6362(17)30004-8 http://dx.doi.org/doi:10.1016/j.gaitpost.2017.01.003 GAIPOS 5282
To appear in:
Gait & Posture
Received date: Revised date: Accepted date:
3-10-2016 23-11-2016 3-1-2017
Please cite this article as: Angelico Fabiola, Freslier Marie, Romkes Jacqueline, Brunner Reinald, Schmid Stefan.Upper Extremity Motion during Gait in Adolescents with Structural Leg Length Discrepancy—an Exploratory Study.Gait and Posture http://dx.doi.org/10.1016/j.gaitpost.2017.01.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Upper Extremity Motion during Gait in Adolescents with Structural Leg Length Discrepancy – an Exploratory Study
Fabiola ANGELICO1,2, Marie FRESLIER1, Jacqueline ROMKES1, Reinald BRUNNER1,3, Stefan SCHMID4,*
1
University of Basel Children’s Hospital, Laboratory for Movement Analysis, Basel,
Switzerland 2
Zurich University of Applied Sciences, School of Health Professions, Winterthur,
Switzerland 3
University of Basel Children’s Hospital, Orthopaedic Department, Basel, Switzerland
4
Bern University of Applied Sciences, Health Division, Bern, Switzerland
Corresponding author: *
Stefan Schmid, PT, PhD, Bern University of Applied Sciences, Health Division,
Murtenstrasse 10, 3008 Bern, Switzerland, +41 79 936 74 79, [email protected]
Word count: Abstract: 248, Text (without acknowledgements and references): 2905
1
RESEARCH HIGHLIGHTS
First study investigating upper limb motion in LLD patients during gait
LLD patients indicated secondary deviations in upper limb motion
Increased shoulder extension as passive physical effect of trunk position
Active strategies might be involved in controlling total body angular momentum
Basis for further investigations using complex musculoskeletal models
ABSTRACT Background and Purpose: Depending on the extent of a structural leg length discrepancy (LLD), several compensatory mechanisms take place in order to maintain function and to optimize energy consumption during gait. However, studies describing the influence of a structural LLD on upper limb motion are lacking. The current study therefore aimed at the evaluation of upper limb motion during gait in LLD patients compared to healthy controls. Methods: Motion capture data from 14 patients with structural LLD and 15 healthy controls that were collected during barefoot walking at a self-selected speed were retrospectively analyzed. Specifically, kinematic parameters of the shoulder and elbow joints as well as the trunk segment were investigated and considered in relation to a minimal clinically important difference of 5°. Results: The shoulders in LLD patients were kept constantly in a more extended and at initial contact in a more adducted position as compared to healthy controls. In addition, the patients’ elbow joints showed increased flexion motion and the trunk segment indicated a constant trunk lateral-flexion and axial rotation tendency towards the affected side. Conclusions: Patients with structural LLD indicated clinically relevant secondary deviations in shoulder and elbow motion. While some of these deviations were most likely passive physical effects, others might have occurred as active strategies to maintain balance or to regulate total body angular momentum. These findings contribute to the understanding of 2
secondary gait deviations induced by a structural LLD and might serve as a basis for further investigations using complex musculoskeletal models.
Keywords: LLD; Upper body; Arm; Secondary deviations
INTRODUCTION Leg length discrepancy (LLD) is a relatively common problem found in 40-70% of the population, whereby about one in every 1000 persons presents with a LLD of more than 20 mm [1]. Complications of LLD include functional limitations such as gait and balance problems and associated musculoskeletal disorders such as low back pain or stress fractures [1]. Depending on the extent of the LLD, pelvic obliquity was observed as a consequence during gait, whereas increased ankle plantarflexion (equinus and vaulting) on the affected side (shorter leg), persistent shortening (knee flexion and ankle dorsiflexion) and circumduction on the unaffected side (longer leg) and lateral bending of the lumbar and thoracic spine were described to occur as compensations for the unequal leg length [2-5]. No study, however, could be identified describing the influence of a structural LLD on upper limb motion in this population. During normal gait, Perry [6] defines the movement pattern of the upper limbs as a reaction to the thorax movement, with the arms swinging contrariwise to the rest of the body, whereas the range of motion of the upper limb joints is symmetric between left and right, but varies from person to person and is influenced by gait speed. For a group of healthy adolescents, Romkes et al. [7] reported shoulder movements ranging from 17.4° of extension to 6.5° of flexion and elbow movements from 29.9° to 45.9° of flexion. Based on experiments investigating the functional contribution of upper limb muscles during normal gait, it was postulated that arm swing has both passive and active components and occurs most likely to reduce energy expenditure [8-10]. When comparing walking with constrained arms to normal gait, energy 3
consumption was increased by 8% [11]. Therefore, it was suggested that not only a pathology itself, but also an altered arm swing could lead to a higher energy consumption [12]. Focusing on the effects of walking with a restricted arm swing, Bruijn et al. [13] showed that arm swing did not contribute to gait stability in normal gait, but helped to regain balance and hence contributed to an overall stability following a perturbation. In children with cerebral palsy, an altered movement pattern of the upper limbs was suggested to occur as compensation and to play a role in gaining stability and balance [7, 14-18]. Given the role of arm swing during gait as well as the fact that oxygen consumption was reported to be greater in individuals with an artificially induced LLD [19], the current study aimed at exploring upper limb motion in adolescents with a structural LLD compared to a group of matched healthy controls. Such knowledge might contribute to a better understanding of gait adaptations throughout the whole body and might therefore influence the planning of therapeutic interventions.
METHODS Participants: Data from fourteen adolescents with a structural LLD that were measured in the movement analysis laboratory of a university children’s hospital between 2006 and 2014 were included in this retrospective analysis (Table 1). Inclusion criteria for the patients were: aged between 10 and 18 years, diagnosis of structural LLD of greater than 1% of body height and at least 20 mm, no neurological basic pathology, no obesity (>95th BMI per-age percentile), no diagnosed structural deformities of the spine and no injuries to the locomotor system that led to persistent functional deficits. Hence, LLD patients were only included when demonstrating no deficits in joint range of motion, muscle length and muscle strength as well as no signs of spasticity in the lower extremities. This information was extracted from the report of the standard clinical examination that was carried out prior to the data collection in LLD patients. 4
For comparative purposes, data from fifteen matched healthy adolescents that were measured in the same laboratory were used. The study was approved by the local ethics committee and written informed consent was obtained from all participants and their legal guardians.
Data collection: The LLD was measured from the anterior superior iliac spine via the patella to the medial malleolus using a tape measure [20]. In order to evaluate upper body kinematics, retroreflective markers were placed on the skin according to the Plug-in Gait full body model as described by Gutierrez et al. [21]. Data was collected using a Vicon motion capture system with 6-12 infrared cameras (Oxford Metrics, London, UK). All participants were asked to walk barefoot at a self-selected normal speed along a 10 meter walkway. Measurements were repeated until data from at least five trials were obtained.
Data reduction: The data was analyzed using the Vicon Workstation, Nexus and the Polygon software (Oxford Metrics, London, UK). For every trial, visual setting of gait events was performed by an experienced assessor. The kinematics of the shoulder joint (flexion/extension and abduction/adduction), elbow joint (flexion/extension) and trunk segment (anterior/posterior tilt, lateral-flexion and axial rotation) were calculated according to the Plug-in Gait full body model and expressed in degrees [°]. For each participant, average profiles over five trials were generated and parameterized accordingly. The parameters of interest for the upper extremity joint and trunk angles were: value at initial contact (IC) as well as average (AVG) and range of motion (ROM) values.
Statistical analyses: 5
All statistical analyses were performed using SPSS version 22 (SPSS Inc., Chicago, IL, USA) and G*Power 3.1.9.2 [22]. Normal distribution of the majority of the subject demographic and outcome parameters was confirmed with the Shapiro-Wilk test (p>0.05). Subject demographics were compared using independent samples T-tests (statistically significant difference: p≤0.05). Comparisons of the shoulder and elbow kinematic parameters (primary outcomes) between the patients’ affected and unaffected and the healthy controls’ left side were carried out using one-way analyses of variance (ANOVA) with Tukey HSD post hoc tests, whereas comparisons of the trunk kinematic parameters (secondary outcomes) between the patients and healthy controls were carried out using independent samples T-tests. In order to account for the explorative nature of the study, the interpretation of the results was thereby mainly based on effect sizes (Cohen’s f and d) rather than p-values. Post hoc tests for the primary outcomes were thus only conducted in case of considerable effects of the ANOVA (f≥0.1) [23]. As previously suggested and extensively described [24, 25], mean differences were subsequently evaluated for their clinical relevance using a minimal clinically important difference (MCID) of 5°. Differences showing a large effect (d≥0.8) were thereby considered clinically relevant with the absolute mean difference being above the MCID and a tendency with the mean difference being below the MCID. Shoulder parameters that showed clinically relevant differences were further evaluated for a possible relationship with trunk kinematics using linear regression analyses.
RESULTS No significant differences were found for age, weight, height and gait speed between the LLD patients and the healthy controls (Table 1).
Primary outcomes
6
The statistical analysis showed considerable overall effects for several upper extremity parameters (Table 2). On the patients’ affected side, the shoulder showed clinically relevant greater extension angles at IC (mean difference (95% confidence interval): -13.4° (-19.4°,7.4°), d=-1.96; p<0.001) and on average (-7.1° (-12.5°,-1.8°), d=-1.28; p=0.007) as well as smaller abduction angles at IC (-5.4° (-10.6°,-0.2°), d=-0.92; p=0.040) as compared to the controls (Figure 1). Furthermore, patients showed clinically relevant increases for the average elbow flexion angle (8.3° (2.8°,13.9°), d=1.52; p=0.002) as well as an increased elbow ROM (19.5° (10.8°,28.3°), d=2.22; p<0.001) on the affected side compared to the control group. The shoulder on the patients’ unaffected side showed an increased extension angle at IC (8.9° (-15.0°,-2.9°), d=-1.38; p=0.002) and on average (-9.3° (-14.7°,-3.9°), d=-1.58; p<0.001) compared to the shoulder in the control group. In addition, the patients showed a clinically relevant decrease in the abduction angle at IC (-6.4° (-11.5°,-1.2°), d=-1.21; p=0.013) between the unaffected side and the controls. Similar to the patients’ affected side, the elbow joint on the unaffected side revealed clinically relevant increases for the flexion angle on average (10.0° (4.5°,15.5°), d=1.63; p<0.001) and ROM (18.0° (9.3°,26.7°), d=2.11; p<0.001) as compared to controls. The patients’ affected side showed clinically relevant greater shoulder ROM (12.3° (2.5°,22.1°), d=1.19; p=0.011) compared to the unaffected side. Regression analyses revealed a statistically significant coefficient of determination for the shoulder flexion/extension and trunk tilt angles at IC on the affected side (R2=0.306, p=0.040) (Table 3). The qualitative examination of the upper extremity joint angles indicated that the shoulder on the patients’ unaffected side remained constantly in extension (Figure A.1 in the electronic supplementary material). In addition, the frontal plane shoulder angles indicated a time shift of the instant of peak abduction angle between the two groups. While the healthy controls
7
reached their peak abduction angle during loading response, LLD patients reached it in late midstance.
Secondary outcomes The statistical analysis of the trunk angles yielded large effects for differences in tilt ROM (d=1.63, p<0.001), lateral-flexion at IC on the affected side (d=-0.99, p=0.017), lateral-flexion average and ROM (d=-1.07, p=0.008 and d=0.90, p=0.023, respectively) as well as axial rotation at IC on the unaffected side (d=-1.35, p=0.001) (Table 4). However, due to mean differences of less than 5°, they were only considered tendencies rather than clinically relevant. In addition, the qualitative examination of the trunk segment angles indicated that the trunk in LLD patients tended to be more upright during loading response (less anterior trunk tilt) as well as in constant lateral-flexion and axial rotation towards the affected side compared to the healthy controls (Figure A.1).
DISCUSSION The aim of the current study was to explore upper extremity kinematics during gait in adolescents with a structural LLD in comparison to a group of matched healthy controls. The analysis indicated that, although the trunk showed no clinically relevant differences, the LLD patients adopted clinically relevant secondary deviations in sagittal shoulder and elbow as well as frontal shoulder kinematics. The question whether these secondary deviations occurred as passive physical effects of the primary deviations or as active compensatory mechanisms to offset primary deviations and the physical effects [26], however, will be discussed in the following. The deviations found in the LLD patients’ sagittal plane shoulder angles at IC on the affected side might have resulted from a slightly more upright trunk position during loading response 8
and could therefore be considered a passive physical effect. In addition, it could be assumed that the observed tendencies of trunk lateral-flexion and axial rotation towards the affected side would have resulted in increased shoulder extension as well as decreased shoulder abduction angles on the unaffected side. However, since no relevant correlations were found between these parameters, this assumption could not be supported by the current results. Furthermore, the decreased shoulder abduction angle at IC on the affected side might have occurred as a compensatory strategy for maintaining lateral balance after “stepping down” on the shorter leg. This strategy has been previously described in healthy young and middle-aged individuals following induced lateral perturbations [27]. However, it remains unclear whether patients with a structural LLD would react in the same way as healthy individuals and since dynamic balance/gait stability was not evaluated in the current investigation, this assumption should be treated with caution. The increased elbow flexion motion might be interpreted as a passive reaction to a slightly increased sagittal shoulder motion using the double-pendulum model [28]. Another plausible explanation for the observed deviations in the upper extremities in LLD patients could be the regulation of total body angular momentum. As reported by Thielemans et al. [29], changes in angular momentum of one limb can be compensated through changes in angular momentum of the other limbs. It might therefore be possible that the LLD patients adjusted their arm swing in order to maintain a normal total body angular momentum in spite of the unequal leg length. The fact that only tendencies and no clinically relevant deviations were found for the trunk can be explained by previously described compensatory mechanisms occurring in the lower extremities, the pelvis and the spine in order to equilibrate leg length, minimize the displacement of the body’s center of mass and reduce energy expenditure [2-5]. In the late 90’s, Song et al. [4] investigated lower extremity and pelvis kinematics in a group of adolescents with structural LLDs of 0.6-11.1 cm and documented compensatory strategies 9
such as pelvis obliquity as well as toe-walking and, for the ones that walked plantigrade, vaulting, circumduction and persistent flexion of the longer limb. Almost 20 years later, Aiona et al. [2] investigated a similar cohort of LLD patients and reported pelvis obliquity, increased ankle plantarflexion on the shorter side (equinus or vaulting) as well as knee flexion on the longer side as compensations for the unequal leg length. Studies conducted on healthy individuals with artificially induced LLDs of 1-5 cm found pelvis obliquity and changes in the sagittal plane kinematics of the ankle, knee and hip joints (i.e. prolonging the shorter leg and shortening the longer leg) as well as lateral bending of the lumbar and thoracic spine [3, 5]. However, Song et al. [4] reported that patients with LLDs of less than 3% of the longer limb length (approximately 2.2 cm) showed no compensatory strategies. Similarly, Walsh et al. [5] found that several LLDs up to 2.2 cm did either not require any compensatory strategies or were compensated by pelvis obliquity and did not require any functional adaptations in the lower extremities. Since the minimal LLD in the current study was 3 cm, it can be expected that all patients displayed compensations in the lower extremities, the pelvis and the trunk/spine. It has to be considered though that these authors did not distinguish between active and passive mechanisms and interpreted several secondary deviations as compensations. According to Schmid et al. [26], however, such a distinction might be of high importance for the planning of a physical therapy treatment, since compensations are active neuromuscular processes and might therefore be controlled by voluntary actions, whereas physical effects are given based on the laws of physics and might therefore not be corrected. When considering this distinction, pelvic obliquity, such as found by Walsh et al. [5] in patients with LLDs up to 2.2 cm, would most likely be classified as a passive physical effect, occurring as a logical physical consequence of an unequal leg length that is not or not sufficiently compensated for by active neuromuscular processes in the lower extremity joints.
10
A possible explanation for an absence of compensatory strategies in patients with rather small LLDs (< 5 cm) could be that the “cost” of achieving a level pelvis is simply too great (increased work on the long side) and that these patients are therefore willing to tolerate some amount of pelvic obliquity [2]. This might have a direct impact on upper body movement since an uncompensated pelvic obliquity would most likely cause the spine and the upper extremities to actively react, whereas a fully or partially compensated pelvic obliquity would cause less deviations in the upper body. In addition, Aiona et al. [2] showed that LLD patients applied different lower limb compensations depending on the amount of LLD as well as the location of the predominant difference (i.e. femur or tibia). Therefore, it cannot be excluded that the compensatory strategies in the upper extremities also depended on these factors. An important factor that has to be always considered when investigating gait kinematics is walking speed. Previous research showed that gait characteristics such as spatio-temporal parameters, ground reaction forces as well as joint angles and moments were significantly affected by speed in healthy children [30]. Taking gait speed into account, a possible speeddependency of the observed kinematic group differences in the current study can be excluded. Furthermore, the current kinematic findings seem to be highly comparable to those of Song et al. [4] and Aiona et al. [2], since the gait speeds reported in their studies (1.23 m/s and 1.2 m/s, respectively) appear to be in agreement with the current ones. On the other hand, when comparing the upper extremity joint angles of the control subjects in the current study with the ones of Romkes et al. [7], it appears that there are slight differences (especially in frontal plane shoulder motion) that might limit the interpretation of the current findings. However, it has to be considered that the sample size of the control group in the above mentioned study was significantly smaller resulting in less representative values of the control parameters. The fact that the length of the upper limb segments was not taken into account might be considered another limitation. It can therefore not be excluded that some of
11
the observed deviations in upper limb motion were merely the result of a structural arm length difference. Further insights into the neuromuscular control mechanisms of the upper limbs in LLD patients might be gained using complex musculoskeletal models that allow the simulation of muscular force contributions and joint loading. Such computations, however, do not derive from conventional gait analysis. In conclusion, adolescents with a structural LLD of greater than 1% of body height indicated clinically relevant secondary deviations in shoulder and elbow motion. The increased shoulder extension angle on the affected side was considered to have occurred as a passive physical effect of the slightly more upright trunk position during loading response. The other changes in shoulder and elbow motion might possibly have taken place as an active strategy in order to maintain balance or to regulate total body angular momentum. These findings contribute to the understanding of secondary gait deviations induced by a structural LLD and might serve as a basis for further investigations using complex musculoskeletal models.
CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest.
Acknowledgments: The authors acknowledge Dr. Andreas Krieg for assistance in the patient recruitment process and Dr. Katrin Bracht-Schweizer for assistance in the data collection.
REFERENCES [1] Gurney B. Leg length discrepancy. Gait Posture. 2002;15:195-206.
12
[2] Aiona M, Do KP, Emara K, Dorociak R, Pierce R. Gait patterns in children with limb length discrepancy. J Pediatr Orthop. 2015;35:280-4. [3] Kakushima M, Miyamoto K, Shimizu K. The Effect of Leg Length Discrepancy on Spinal Motion during Gait: Three-Dimensional Analysis in Healthy Volunteers. Spine (Phila Pa 1976). 2003;28:2472-6. [4] Song KM, Halliday SE, Little DG. The effect of limb-length discrepancy on gait. J Bone Joint Surg Am. 1997;79:1690-8. [5] Walsh M, Connolly P, Jenkinson A, O'Brien T. Leg length discrepancy - an experimental study of compensatory changes in three dimensions using gait analysis. Gait Posture. 2000;12:156-61. [6] Perry J, Burnfield JM. Gait Analysis: Normal and Pathological Function. 2nd ed. Thorofare, NJ: SLACK Incorporated; 2010. [7] Romkes J, Peeters W, Oosterom AM, Molenaar S, Bakels I, Brunner R. Evaluating upper body movements during gait in healthy children and children with diplegic cerebral palsy. J Pediatr Orthop B. 2007;16:175-80. [8] Goudriaan M, Jonkers I, van Dieen JH, Bruijn SM. Arm swing in human walking: what is their drive? Gait Posture. 2014;40:321-6. [9] Kuhtz-Buschbeck JP, Jing B. Activity of upper limb muscles during human walking. J Electromyogr Kinesiol. 2012;22:199-206. [10] Meyns P, Bruijn SM, Duysens J. The how and why of arm swing during human walking. Gait Posture. 2013;38:555-62. [11] Umberger BR. Effects of suppressing arm swing on kinematics, kinetics, and energetics of human walking. J Biomech. 2008;41:2575-80. [12] Collins SH, Adamczyk PG, Kuo AD. Dynamic arm swinging in human walking. Proc Biol Sci. 2009;276:3679-88.
13
[13] Bruijn SM, Meijer OG, Beek PJ, van Dieen JH. The effects of arm swing on human gait stability. J Exp Biol. 2010;213:3945-52. [14] Delabastita T, Desloovere K, Meyns P. Restricted Arm Swing Affects Gait Stability and Increased Walking Speed Alters Trunk Movements in Children with Cerebral Palsy. Front Hum Neurosci. 2016;10:354. [15] Galli M, Cimolin V, Albertini G, Piccinini L, Turconi AC, Romkes J, et al. Kinematic analysis of upper limb during walking in diplegic children with Cerebral Palsy. Eur J Paediatr Neurol. 2013. [16] Meyns P, Desloovere K, Van Gestel L, Massaad F, Smits-Engelsman B, Duysens J. Altered arm posture in children with cerebral palsy is related to instability during walking. Eur J Paediatr Neurol. 2012;16:528-35. [17] Meyns P, Duysens J, Desloovere K. The arm posture in children with unilateral Cerebral Palsy is mainly related to antero-posterior gait instability. Gait Posture. 2016;49:132-5. [18] Meyns P, Van Gestel L, Massaad F, Desloovere K, Molenaers G, Duysens J. Arm swing during walking at different speeds in children with Cerebral Palsy and typically developing children. Res Dev Disabil. 2011;32:1957-64. [19] Gurney B, Mermier C, Robergs R, Gibson A, Rivero D. Effects of limb-length discrepancy on gait economy and lower-extremity muscle activity in older adults. J Bone Joint Surg Am. 2001;83-A:907-15. [20] Jamaluddin S, Sulaiman AR, Imran MK, Juhara H, Ezane MA, Nordin S. Reliability and accuracy of the tape measurement method with a nearest reading of 5 mm in the assessment of leg length discrepancy. Singapore Med J. 2011;52:681-4. [21] Gutierrez EM, Bartonek A, Haglund-Akerlind Y, Saraste H. Centre of mass motion during gait in persons with myelomeningocele. Gait Posture. 2003;18:37-46.
14
[22] Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39:175-91. [23] Cohen J. Statistical Power Analysis for the Behavioral Sciences. Revised ed. New York, NY: Academic Press, Inc.; 1977. [24] Schmid S, Romkes J, Taylor WR, Lorenzetti S, Brunner R. Orthotic correction of lower limb function during gait does not immediately influence spinal kinematics in spastic hemiplegic cerebral palsy. Gait Posture. 2016;49:457-62. [25] Schmid S, Studer D, Hasler C-C, Romkes J, Taylor WR, Lorenzetti S, et al. Quantifying Spinal Gait Kinematics using an Enhanced Optical Motion Capture Approach in Adolescent Idiopathic Scoliosis. Gait Posture. 2016;44:231–7. [26] Schmid S, Schweizer K, Romkes J, Lorenzetti S, Brunner R. Secondary gait deviations in patients with and without neurological involvement: A systematic review. Gait Posture. 2013;37:480-93. [27] Allum JH, Carpenter MG, Honegger F, Adkin AL, Bloem BR. Age-dependent variations in the directional sensitivity of balance corrections and compensatory arm movements in man. J Physiol. 2002;542:643-63. [28] Kubo M, Wagenaar RC, Saltzman E, Holt KG. Biomechanical mechanism for transitions in phase and frequency of arm and leg swing during walking. Biol Cybern. 2004;91:91-8. [29] Thielemans V, Meyns P, Bruijn SM. Is angular momentum in the horizontal plane during gait a controlled variable? Hum Mov Sci. 2014;34:205-16. [30] van der Linden ML, Kerr AM, Hazlewood ME, Hillman SJ, Robb JE. Kinematic and kinetic gait characteristics of normal children walking at a range of clinically relevant speeds. J Pediatr Orthop. 2002;22:800-6.
15
FIGURE CAPTIONS Figure 1: Primary outcomes: Results of the pairwise group comparisons (mean differences with 95% confidence intervals) for the kinematic parameters of the shoulder and elbow joints (expressed in degrees) between leg length discrepancy (LLD) patients (affected and unaffected sides) and healthy controls. The dashed vertical lines represent the minimal clinically important difference of 5°. Differences were categorized as clinically relevant (large black squares) and clinically not relevant (small white squares). IC: initial contact, AVG: average; ROM: range of motion.
16
Table 1: Demographics of the leg length discrepancy (LLD) patients and healthy controls expressed as means, standard deviations (SD) and ranges (in brackets). Group comparisons were conducted using independent samples T-tests (statistical significance p0.05). LLD Patients
Controls
(n = 14)
(n = 15)
Age [years]
14.3 SD 1.6 (11-17)
14.1 SD 1.7 (12-17)
0.804
Height [m]
1.66 SD 0.08 (1.52-1.84)
1.62 SD 0.10 (1.46-1.84)
0.245
Mass [kg]
60.5 SD 12.7 (36-79)
54.2 SD 10.0 (40-70)
0.153
Gender [male/female]
6/8
8/7
-
Affected (shorter) leg [right/left]
7/7
-
-
LLD [mm]
47.5 SD 16.1 (30-85)
3.0 SD 4.1 (0-10)
<0.001
LLD [% of height]
2.9 SD 1.0 (1.8-5.2)
0.2 SD 0.2 (0.0-0.6)
<0.001
Gait speed [m/s]
1.2 SD 0.2 (0.9-1.6)
1.2 SD 0.1 (1.1-1.5)
0.961
p-value
17
Table 2: Primary outcomes: Reported are mean, standard deviation (SD) and 95% confidence intervals (within the brackets) for the kinematic parameters of the shoulder and elbow joints (expressed in degrees) in patients with structural leg length discrepancy (LLD) and matched healthy controls. In addition, effect sizes (Cohen’s f and d, respectively) as well as results for the group comparisons (one-way analyses of variance (ANOVA) and Tukey’s post hoc tests) are presented. 1)
2) 3)
LLD patients
LLD patients
Post hoc (Tukey) ANOVA 1-3
2-3
1-2
f
d
d
d
Controls (affected side)
(unaffected side)
-23.8 SD 7.0 (-27.8,-19.7)
-19.3 SD 6.4 (-23.0,-15.7)
-10.4 SD 6.7 (-14.1,-6.7)
0.84**
-1.96++
-1.38++
-0.66
AVG
-9.7 SD 5.9 (-13.1,-6.3)
-11.9 SD 6.5 (-15.6,-8.1)
-2.5 SD 5.4 (-5.6,0.5)
0.68**
-1.28++
-1.58++
0.35
ROM
29.8 SD 12.7 (22.4,37.1)
17.5 SD 7.1 (13.4,21.6)
20.6 SD 11.3 (14.3,26.9)
0.48**
0.76
-0.33
1.19++
IC
7.2 SD 6.6 (3.4,11.0)
6.2 SD 5.5 (3.0,9.4)
12.6 SD 5.1 (9.7,15.4)
0.49**
-0.92++
-1.21++
0.16
AVG
9.7 SD 5.6 (6.5,12.9)
8.7 SD 5.1 (5.7,11.6)
11.3 SD 3.9 (9.2,13.5)
0.22*
-0.34
-0.58
0.19
ROM
8.1 SD 5.2 (5.1,11.1)
7.4 SD 3.3 (5.5,9.3)
6.1 SD 3.5 (4.1,8.0)
0.21*
0.47
0.39
0.17
IC
28.6 SD 4.9 (25.8,31.4)
30.5 SD 7.5 (26.2,34.9)
28.9 SD 4.9 (26.2,31.6)
0.14*
-0.06
0.26
-0.31
AVG
37.6 SD 6.0 (34.1,41.1)
39.3 SD 7.2 (35.1,43.5)
29.3 SD 4.9 (26.6,32.0)
0.72**
1.52++
1.63++
-0.26
ROM
25.9 SD 11.7 (19.2,32.7)
24.4 SD 11.2 (17.9,30.9)
6.4 SD 4.8 (3.7,9.1)
0.93**
2.22++
2.11++
0.13
IC Flex./ Ext. Shoulder Abd./ Add.
Elbow
Flex./ Ext.
Flex.: flexion, Ext.: extension, Abd.: abduction, Add.: adduction, IC: initial contact, ROM: range of motion, AVG: average Positive/negative values: flexion/extension (shoulder and elbow), abduction/adduction (shoulder) *
Considerable effect size for ANOVA (f≥0.1) without statistical significance (p>0.05), post hoc tests indicated Considerable effect size for ANOVA (f≥0.1) with statistical significance (p≤0.05), post hoc tests indicated
** +
Large effect size for post hoc (Tukey) without statistical significance (p>0.05) Large effect size for post hoc (Tukey) with statistical significance (p≤0.05)
++
18
Table 3: Results of the regression analyses (expressed as coefficients of determination, R2) for the shoulder parameters that showed clinically relevant differences in adolescents with a structural leg length discrepancy. Trunk (affected side) Tilt
Shoulder (affected side)
Flex./ Ext. Abd./ Add.
Shoulder (unaffected side)
Flex./ Ext. Abd./ Add.
Lat.-Flex.
Axial Rot.
IC
AVG
IC
AVG
IC
AVG
0.306*
-
0.228
-
0.002
-
-
0.190
-
0.018
-
0.030
IC
0.000
-
0.055
-
0.047
-
IC
0.002
-
0.000
-
0.004
-
-
0.069
-
0.020
-
0.004
0.000
-
0.003
-
0.091
-
IC AVG
AVG IC
Flex.: flexion, Ext.: extension, Abd.: abduction, Add.: adduction, Lat.-Flex.: lateral-flexion, Axial Rot.: axial rotation, IC: initial contact, AVG: average * Statistically significant coefficient of determination (p<0.05) Table 4: Secondary outcomes: Reported are mean, standard deviation (SD) and 95% confidence intervals (within the brackets) for the kinematic parameters of the trunk segment (expressed in degrees) in patients with structural leg length discrepancy (LLD) and matched healthy controls. In addition, effect sizes (Cohen’s d) and results for the group comparisons (independent samples T-tests) are presented.
19
1) LLD patients
2) Controls
Mean-difference (1 - 2)
T-test (d)
IC (affected)
3.3 SD 5.3 (0.3,6.3)
-2.7 (-7.3,1.8)
-0.58
0.2 (-3.5,4.0)
0.05
6.0 SD 4.2 (3.7,8.3) IC (unaffected)
6.2 SD 5.6 (3.0,9.5)
AVG
4.8 SD 5.5 (1.6,7.9)
5.4 SD 4.1 (3.1,7.6)
-0.6 (-5.2,4.0)
-0.12
ROM
5.0 SD 1.5 (4.1,5.9)
2.9 SD 0.9 (2.4,3.5)
2.1 (0.8,3.4)
1.63++
-1.9 (-4.0,0.3)
-0.99++
-1.6 (-3.4,0.3)
-0.62
Tilt
IC (affected)
-1.4 SD 2.4 (-2.7,0.0)
IC (unaffected)
-1.0 SD 3.2 (-2.9,0.8)
AVG
-1.2 SD 2.5 (-2.6,0,3)
1.0 SD 1.3 (0.3,1.7)
-2.2 (-4.1,-0.2)
-1.07++
ROM
5.1 SD 2.6 (3.6,6.7)
3.3 SD 1.3 (2.6,4.0)
1.8 (-0.2,3.9)
0.90++
IC (affected)
1.8 SD 4.2 (-0.6,4.3)
2.6 (-0.2,5.4)
-0.70
4.1 (-1.8,6.5)
-1.35++
0.5 SD 1.3 (-0.2,1.2) Trunk
Lat.-Flex.
-0.8 SD 3.2 (-2.5,1.0) IC (unaffected)
3.4 SD 3.0 (1.7,5.1)
AVG
2.6 SD 3.5 (0.6,4.6)
0.6 SD 2.2 (-0.6,1.8)
2.0 (-0.8,4.7)
0.68
ROM
6.8 SD 2.7 (5.3,8.3)
5.7 SD 2.0 (4.6,6.8)
1.1 (-1.1,3.3)
0.47
Axial Rot.
Lat.-Flex.: lateral-flexion, Axial Rot.: axial rotation, IC: initial contact, ROM: range of motion, AVG: average Positive/negative values: anterior/posterior tilt, contra (unaffected side)-/ipsilateral (affected side) flexion, external (affected side)/internal (unaffected side) axial rotation +
Large effect size for T-test without statistical significance (p>0.05) Large effect size for T-test with statistical significance (p≤0.05)
++
20