- Email: [email protected]
The effects of real and artificial Leg Length Discrepancy on mechanical work and energy cost during the gait
Gait & Posture 59 (2018) 147–151
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
The effects of real and artificial Leg Length Discrepancy on mechanical work and energy cost during the gait
MARK
⁎
T.F. Assogbaa, , S. Bouletb, C. Detrembleura, P. Mahaudensa,b,c a Université Catholique de Louvain, Secteur des Sciences de la Santé, Institut de Recherche Expérimentale et Clinique, Neuro Musculo Skeletal Lab (NMSK), Avenue Mounier 53, B-1200 Brussels, Belgium b Cliniques Universitaires St Luc, Service de médecine physique et réadaptation, Avenue Hippocrate 10, B-1200 Brussels, Belgium c Cliniques Universitaires St Luc, Service d’orthopédie et de traumatologie de l’appareil locomoteur, Avenue Hippocrate 10, B-1200 Brussels, Belgium
A R T I C L E I N F O
A B S T R A C T
Keywords: Leg length discrepancy Mechanical work Energy cost Gait
Background: The impacts of Leg Length Discrepancy (LLD) on the kinematic and dynamic parameters of walking have been widely discussed. But little is known on total mechanical work and energy cost. These two variables are more representative of the functional impairment undergone by the LLD patients. Aim: To assess the changes of the mechanical work and energy cost of walking in subjects with real LLD and to compare their results with healthy subjects in whom the LLD has been simulated. Method: The mechanical work and energy cost data of 60 healthy subjects (speed: 4 km/h) with artificial LLD induced by soles (2 and 4 cm), 20 patients (speed: 3.75 ± 0.5 km/h) with real LLD and 20 matched subjects (speed: 3.75 ± 0.5 km/h) were collected. Statistical comparisons between the groups were performed using a tpaired test and ANOVA. Results: Patients with a real LLD showed a significant decrease in mechanical work and energy cost when compared to norms. Patients with real LLD provide a better recovery when compared to subjects with artificial LLD of 2 cm, and a decrease of energy cost and higher muscular efficiency (mechanical work/energy cost) when compared to subjects with artificial LLD of 4 cm. Conclusions: Our results showed that patients with a real LLD develop compensatory strategies during gait, probably to minimize the displacement of the body center of mass and consequently reduce the amount of energy expenditure useful for their displacement. Moreover, they adopt a better gait strategy compared to the subjects in whom LLD was simulated.
1. Introduction Walking is the usual mode of human locomotion [1]. However, musculoskeletal pathology may alter our ability to move, disturbing gait, increasing energy costs and, consequently, reducing our autonomy. For example, the use of walking cane in patients with knee osteoarthritis has been observed to cause an increase in energy expenditure during walking [2]. Leg Length Discrepancy (LLD), whether or not linked to a pathological process, is one of the most frequent musculoskeletal conditions encountered in clinical practice. LLD, defined as a condition in which the lower limbs are noticeably unequal in length [3], LLD is identified in as many as 40% [4] to 70% [5] of the population. Two etiological groups can be characterized, based on structural versus functional alterations: structural LLD is defined as real LLD associated with shortening of bony structures; functional LLD is defined as LLD that is the
⁎
result of altered mechanics of the lower extremities [6]. Several authors have shown that LLD (greater than 20–30 mm) can produce changes in gait. These include increased ground reaction forces [7–9], increased lower extremity kinetic energy [10] and increased mechanical work [11,12], defined as the total positive mechanical work (Wtot) done by the muscles during walking, in subjects with real LLD, and increased oxygen consumption (VO2) in subjects with artificial LLD induced by wearing a shoe with a sole [13]. Song et al. measured mechanical work during gait in children with real LLD. They observed that the children used several techniques to compensate for the LLD, including toe walking, vaulting, circumduction, and increased flexion of the longer limb. In addition, they noted that the longer limb performed more mechanical work than did the shorter. The average LLD for subjects with no observable compensatory strategy was 16.4 mm [12]. To our knowledge, energy expenditure has not been studied in real LLD. Gurney et al. created an artificial LLD
Corresponding author. E-mail addresses: [email protected], [email protected] (T.F. Assogba).
http://dx.doi.org/10.1016/j.gaitpost.2017.10.004 Received 15 May 2017; Received in revised form 24 August 2017; Accepted 3 October 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
trajectories of reflective markers positioned on specific anatomical landmarks. The subject was fitted with a mask to measure VO2 and carbon dioxide production (VCO2) throughout the treadmill test using an ergospirometer (Medisoft, Belgium). The sessions began with a rest period, in which the subjects stood barefoot for the static calibration of kinematic and energetic variables. Thereafter, the subjects were asked to walk on a treadmill for a few minutes. For group 1, all subjects walked at a constant speed of 4 km h−1 for each trial (with and without soles). For group 2, the subjects walked at a spontaneous speed that had been measured earlier using the 10 m walk test [15]. Group 3 subjects walked on the treadmill at different speeds of 1, 2, 3, 4 and 5 km h−1 (in order to establish norms). Each subject walked for a few minutes until a steady state was reached and maintained for at least 2 min to record energetic variables. Kinematics and kinetics variables were then recorded simultaneously and averaged over 10 successive strides. The mean for each data was used for statistical analysis. In group 1, LLD was induced using 2 cm and 4 cm soles (Airlit) attached to a sports shoe. The order in which the soles were used was random and determined by drawing lots. Before gait analysis, the subjects of all groups walked on the treadmill for at least 10 min in a practice session to get used to the treadmill, and also to the different soles for group 1.
using shoe soles in healthy older persons (ages 55–86) during gait. The subjects had significantly greater energy consumption at 20, 30 and 40 mm of LLD compared to no LLD [13]. Normal human walking is governed by a multitude of interrelated parameters: kinematics, kinetics and energetics. Understanding these parameters that describe gait biomechanics is essential to improve the functional activity of persons suffering from pathology that affects walking. The role played by energetics, which may be a good indicator of the difficulty a subject has to walk, is often neglected; its evaluation could contribute to better therapeutic decisions for improving gait in individuals with LLD. We hypothesized that people with real LLD have an altered walking mechanism and adapt their walking strategy to limit excessive energy expenditure better than do people with experimentally artificial LLD. The current study was therefore designed first to assess the effects of real LLD on mechanical work and energy cost during walking and second to compare the energetic changes in subjects with real and artificial LLD. 2. Method 2.1. Participants The subjects were divided into three groups. The first was composed of 60 healthy adult subjects aged 18–29 years (22 ± 1 years, group 1) who participated in the experiment, and walked with and without soles of 2 and 4 cm. The second group was composed of 20 young subjects with real LLD aged 5–19 years (12 ± 3 years, group 2). The third group was composed of 20 young subjects (12 ± 3 years, group 3, norm) without LLD matched in age and speed with each subject of group 2. The sex ratio and demographic data of the three groups are shown in Table 1. Group 1 subjects were recruited from university personnel between January and March 2016. Group 2 subjects were recruited after consultation in the orthopedic service of our hospital between January and December 2016. Subjects with an LLD of less than 1 cm or who had undergone any surgery, fracture or immobilization of the lower limbs in the six months prior to the evaluation were excluded. Group 3 subjects, with no lower limb orthopedic history were recruited from our entourage (professional and social). Subjects less than 5 years of age or with more than 1 cm of inequality or who had trauma to the pelvis or limb in the six months prior to the study were not included. Every subject or parent gave signed consent and participated freely in the study, which was approved by the local ethics board.
2.3. Parameters Mechanical work was computed as follows: the total positive mechanical work (Wtot) done by the muscles during walking was divided into the external work (Wext) performed to move the body Center of Mass (COM) relative to surroundings and the internal work (Wint) performed to move the body segments relative to the COM [16]. Wext was computed from four force transducers located at the four corners of the treadmill. These transducers measured the 3D-ground reaction forces according to Cavagna [17]. Wext represents the positive work and is divided into the forward work (Wekf), vertical work (Wekv) and lateral work (Wekl) necessary to accelerate the COM in the three directions and lift the COM during a stride. The Recovery, quantifying the percentage of mechanical energy saved by a pendulum-like exchange between the gravitational potential energy and the kinetic energy of the COM (i.e., an index reflecting the effectiveness of the pendulum-like mechanical mode of walking), was measured [17,18]. Wint, the work required to move the limbs relative to the COM was computed from kinematic data following the methods described by Willems et al. [16] and by Detrembleur et al. [19]. The metabolic cost was determined by the subject’s VO2 and VCO₂. The mass specific gross energy consumption rate (W kg−1) was obtained from the VO2 using the energy equivalent of oxygen, taking into account the measured respiratory exchange ratio (RER). Each energy measurement started with a rest period in which the subject was standing on the treadmill. Thereafter, they walked until a steady state was reached and maintained for at least 2 min. The Joules of energy expended per liter of oxygen consumed was computed depending on the RER according to the Lusk equation [20]. The energy expended above the resting value (standing subtracted from walking consumption) was divided by the walking speed to obtain the net energy cost of walking (C, J/kg m−1) [21]. The efficiency (ɳ) of the positive work production by the muscles was calculated as the ratio between Wtot and C [21].
2.2. Procedures Gait was assessed using a three-dimensional analysis, which included synchronous kinematics, kinetics, mechanical and energetic measurements [14] (Fig. 1). All subjects walked on a motor-driven treadmill (Mercury LTmed, HP Cosmos®, Germany) equipped with force captors and surrounded by eight infrared cameras at 200 Hz (BTS, Milan, Italy) to measure the
Table 1 Sex ratioand demographic data. Group 1 n = 60 Mean ± SD
Group 2 n = 20 Mean ± SD
Group 3 n = 20 Mean ± SD
22 ± 1 173.0 ± 8.0 67.6 ± 9.1
12 ± 3 153.3 ± 23.5 51.5 ± 25.8
12 ± 3 157 ± 18.7 39.9 ± 13.0
2.4. Statistical analysis Age (Years) Height (cm) Weight (kg)
All variables that had a normal distribution and equality of variance are presented as means ( ± SD). Other variables are expressed as medians and quartiles [25–75%]. Statistical analysis was performed
SD: Standard Deviation.
148
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
Fig. 1. Energetic and mechanical measurements. Illustration of energy expenditure (oxygen consumption and energy cost) measured by the classical indirect calorimetric method and mechanical work assessed by the three-dimensional ground reaction force. The infrared cameras (A) are positioned so that at least two visualize each reflective marker at any given time. From the reflective markers movements we can calculate the 3D trajectories of the body segments (B). A force platform located under the treadmill (B) records the patient’s ground reaction forces. Energy expenditure is measured indirectly based on the rate of oxygen consumption by the patient using an ergo spirometer (C). Finally the mechanical work (D) is calculated as the work performed by muscles to raise and accelerate the center of body mass (external mechanical work) and to move the body segments relative to the center of body mass (internal mechanical work). Adapted from Lobet et al. [14]. Adapted from Lobet et al. [14]
3.2. Real LLD versus artificial LLD (Table 3)
using SigmaPlot version 13.0 (SPSS 2014). The significance level was set at P ≤ 0.05. A paired t-test was used to determine whether the subjects with real LLD had different mechanical work and energy cost compared to group 3. Subjects with real LLD were thus matched in age and speed with subjects without LLD (group 3). To compare real and artificial LLD, age was normalized using a Zscore transformation, and we then we used a One-Way Analysis of Variance (ANOVA) or Kruskal-Wallis One Way Analysis of Variance on Ranks (if normality and equality of variance tests not passed) with one factor (factor = groups i.e. real LLD, artificial 2 cm and 4 cm). A posthoc test with Dunn’s method was used to identify significantly different variables.
There were no statistically significant differences in the Wext, Wint and Wtot Z-scores of subjects with real LLD and artificial LLD of 2 and 4 cm. There were significant differences in the Z-scores of recovery, net energy cost and efficiency, with p values of 0.005, ˂0.001 and 0.004, respectively, for real LLD and artificial LLD of 2 and 4 cm. Post-hoc
Table 2 Results of Paired t-test on mechanical and energetic variables in subjects with real LLD versus matched subjects without LLD. Group 2 n = 20 Mean ± SD
Group 3 n = 20 Mean ± SD
P-value
Mechanics Wext (J/kg m) Wint (J/kg m) Wtot (J/kg m) Wekf (J/kg m) Wekv (J/kg m) Wekl (J/kg m) Recovery (%)
0.25 ± 0.06 0.25 ± 0.03 0.50 ± 0.07 0.26 ± 0.05 0.37 ± 0.07 0.009 ± 0.007 58.6 ± 12.9
0.29 ± 0.05 0.28 ± 0.06 0.57 ± 0.06 0.33 ± 0.04 0.44 ± 0.03 0.009 ± 0.005 62.4 ± 5.8
0.019 0.013 0.002 < 0.001 0.002 0.657 0.182
Energetics Net energy cost (J/kg m) Muscle efficiency (%)
2.05 ± 0.45 22.10 ± 5.34
2.34 ± 0.28 25.78 ± 5.12
0.032 0.152
3. Results 3.1. Real LLD versus matched subjects without LLD (Table 2) The Wtot was 12% lower in subjects with real LLD than in matched subjects without LLD (P = 0.002) and the different components of Wtot, i.e., external work and internal work, were 13% (P = 0.019) and 10% (P = 0.013) lower, respectively. For the three components of the Wext, only Wekf and Wekv were statistically significantly lower in the subjects with real LLD than in matched subjects without LLD, both by 0.07 J/kg. (P = ˂0.001 and 0.002, respectively). The net energy cost was 12% lower (P = 0.032). There were no significant differences in recovery and efficiency between real LLD subjects without LLD.
Significant differences are typed in bold and are accepted for P-value ≤ 0.05. SD: Standard Deviation.
149
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
increasing pelvic obliquity, circumduction, increased hip and/or knee flexion (steppage gait), more ankle dorsiflexion, or any combination of these actions [8]. Further studies should compare kinematic and mechanical parameters, for example the 3D displacement of COM in LLD subjects during gait, to corroborate this hypothesis. The decrease in Wint could also be explained by an alteration in some walking determinants, such as reduced angular speed of lower limbs joints, which were not evaluated in this study. This assumption is reinforced by a study by Lobet et al. [23]. These authors reported, in walking in hemophiliacs, that a reduction in Wint could be provided by a decrease in the range of motion at the level of the affected ankle and hip, and a subsequent decrease in joint angular speed. The decrease in Wext and Wint induced a decrease in the Wtot. Our result suggests that subjects with real LLD economize their positive muscular work, which may facilitate the recovery of the pendular mechanism necessary to walk close to normal and consequently may help guide the choice of appropriate therapies for these subjects. In fact, our subjects with real LLD had a recovery that was close to that of subjects without LLD (58% vs 62%), conserving the energy exchange between the kinetic and the potential energy of COM. Our results are in accordance with other studies showing that this pendulum-like mechanism is also intact in some orthopedic pathologies [23,24]. These studies demonstrated that subjects with lowest-level gait disorder were able to move their bodies in a way that allowed a normal amount of external muscular work to be performed per unit distance. Our results showed that subjects with real LLD (group 2) had a lower energy cost of walking compared to subjects without LLD (group 3). Studies by Cavagna [21] and Willems [16] have highlighted that energy cost depends on muscle work requirements and on the efficiency of muscle work production. In addition, Kerrigan et al. [25], demonstrated that vertical displacement of COM reliably predicted VO2 during gait. Based on these observations, a low vertical displacement of the COM may have induced a low Wekv and thus a low Wext, with a reduction in the VO2 in our subjects with real LLD. This observation is strengthened by a study by Massaad et al. who investigated the energetics of human walking by decreasing (flat) and increasing (bouncy) the vertical COM displacement in healthy subjects walking at different speeds and showed that the vertical displacement of COM appeared to be a determinant of the energy cost of walking [26]. According to Kaufman, it is possible that individuals use compensatory mechanisms to dynamically shorten the long limb, probably to minimize the displacement of the COM and consequently reduce energy expenditure [8]. Inman showed that compensatory strategies dampen oscillations of the COM and decrease over-all energy expenditure during gait [27]. Our results are in agreement with these results from other studies.
Table 3 Results of ANOVA on mechanical and energetic variable in subjects with real LLD versus artificial LLD (2 cm and 4 cm). Median
P-value
[25%–75]
Mechanics Z-score Wext* Z-score Wint Z-score Wtot. Z-score Recovery Energetics Z-score net energy cost Z-score efficiency
Real LLD n = 20
aLLD 2 cm n = 60
aLLD 4 cm n = 60
−0.48 ± 0.99
−0.29 ± 1.33
−0.58 ± 1.36
0.497
−0.70 [−1.06/ −0.23] −0.75 [−1.22/ −0.24] −0.58 [−2.78/ 0.79]
−0.41 [−0.69/ −0.13] −0.59 [−1.12/ 0.05] −1.74 [−2.11/ −1.12]
−0.45 [−0.73/ −0.10] −0.80 [−1.36/ 0.03] −1.05 [−1.68/ −0.45]
0.224
−0.55 [−0.90/ 0.53]
0.26 [−0.29/ 0.89]
0.59 [0.09/ 1.24]
< 0.001b
0.07 [−0.81/ 1.20]
−0.64 [−1.11/ −0.12]
−0.87 [−1.27/ −0.40]
0.004c
0.479 0.005a
Significant differences are typed in bold and are accepted for P-value ≤ 0.05. a Real LLD ≠ aLLD 2 cm (P = 0.029). b Real LLD ≠ aLLD 4 cm (P ˂ 0.001). c Real LLD ≠ aLLD 4 cm (P = 0.003). * Mean aLLD = artificial LLD.
analysis using Dunn’s method also revealed significant differences. Thus, for recovery, real LLD ≠ artificial 2 cm LLD (Z-score real LLD closer to normal than Z-score 2 cm; P-value = 0.029), for net energy cost and efficiency, real LLD ≠ artificial 4 cm LLD (Z-score real LLD closer to normal than Z-score 4 cm LLD, P-value respectively ˂0.001 and 0.003). 4. Discussion 4.1. Real LLD versus matched subjects without LLD We assessed the effects of real LLD on mechanical work and energy cost during walking. These two parameters were significantly reduced in subjects with real LLD (group 2) compared to those without LLD (group 3). Recovery and efficiency were closer to the values of the subjects without LLD. The reduced mechanical work in subjects with real LLD shown in this study has not been observed in other studies. However, our subjects with real LLD had a length difference of 2.02 cm, by contrast with previous studies that have assessed inequalities of more than 2–3 cm, generating increased mechanical work [11,12]. The decrease in Wext observed in the real LLD subjects, may be due to the decrease in both forward (Wekf) and vertical (Wekv) components, which were 21% and 16% lower, respectively. We hypothesized that the decrease in these two components would smooth the COM displacement in vertical and forward directions. Various kinematics strategies can explain this COM smoothing, including excessive rotation of the pelvis, increased pelvic obliquity or knee flexion during the single leg stance or significant lateral movements of the pelvis during gait [22]. According to Saunders, smoothing of COM displacement can in turn explain the low energy cost of walking [22]. These various elements lead these subjects to develop various strategies of compensation. Kaufman suggested that subjects with real LLD compensate for their shorter limb by increasing downward pelvic obliquity, performing knee extension in midstance, vaulting, toe walking, or any combination of these. In addition, the longer leg can be artificially shortened by
4.2. Real LLD versus artificial LLD When we compare our results obtained in subjects with real LLD to those with artificial LLD, the recovery of real LLD subjects was better than that of subjects with 2 cm artificial LLD. We hypothesized that subjects with real LLD would adopt a better walking strategy than subjects with an artificial LLD, improving pendular mechanism of gait. Our subjects with real LLD patients also had lower energy cost and greater walking efficiency compared to subjects with a 4 cm artificial LLD. These differences would indicate a long-term positive effect and likely sustained motor learning that is not possible when someone is asked to walk with an artificial LLD. According to Brand and Yack [28], adults did not to show compensatory strategies when walking with experimentally induced uneven leg lengths over short periods. The compensatory strategies developed by subjects with real LLD were acquired by experience and long-term learning effects, unlike the simulated subjects who did not have time to adapt. This suggestion is reinforced by Maassad et al. [25] who demonstrated that after only 6 weeks of a gait training program with COM biofeedback, patient with hemiparesis decreased their energy consumption by 30% by practicing 150
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
syndrome), J. Orthop. Sports Phys. Ther. 3 (1) (1981) 11–16. [5] A.L. Woerman, S.A. Binder-MacLeod, Leg length discrepancy assessment: accuracy and precision in five clinical methods of evaluation, J. Orthop. Sports Phys. Ther. 5 (5) (1984) 230–239. [6] R.L. Blake, B. Ferguson, Limb length discrepancies, J. Am. Podiatr. Med. Assoc. 82 (1) (1992) 33–38. [7] A. Bhave, D. Paley, J.E. Herzenberg, Improvement in gait parameters after lengthening for the treatment of limb-length discrepancy, J. Bone Joint Surg. Am. 81 (4) (1999) 529–534. [8] K.R. Kaufman, L.S. Miller, D.H. Sutherland, Gait asymmetry in patients with limblength inequality, J. Pediatr. Orthop. 16 (2) (1996) 144–150. [9] D. Schuit, M. Adrian, P. Pidcoe, Effect of heel lifts on ground reaction force patterns in subjects with structural leg-length discrepancies, Phys. Ther. 69 (8) (1989) 663–670. [10] F.G. Delacerda, O.D. Wikoff, Effect of lower extremity asymmetry on the kinematics of gait, J. Orthop. Sports Phys. Ther. 3 (3) (1982) 105–107. [11] X.C. Liu, G. Fabry, G. Molenaers, J. Lammens, P. Moens, Kinematic and kinetic asymmetry in patients with leg-length discrepancy, J. Pediatr. Orthop. 18 (2) (1998) 187–189. [12] K.M. Song, S.E. Halliday, D.G. Little, The effect of limb-length discrepancy on gait, J. Bone Joint Surg. Am. 79 (11) (1997) 1690–1698. [13] B. Gurney, C. Mermier, R.A. Robergs, A.L. Gibson, D. Rivero, Effects of leg length discrepancy on gait economy and lower extremity muscle activity in older adults, J. Bone Joint Surg. Am. 83-A (6) (2001) 907–915. [14] S. Lobet, C. Detrembleur, F. Massaad, C. Hermans, Three-dimensional gait analysis can shed new light on walking in patients with haemophilia, Sci. World J. 13 (2013) 284358 7 pages. [15] D.T. Wade, Measurement in neurological rehabilitation, Curr. Opin. Neurol. Neurosurg. 5 (1992) 682–686. [16] P.A. Willems, G.A. Cavagna, N.C. Heglund, External, internal and total work in human locomotion, J. Exp. Biol. 198 (Pt. 2) (1995) 379–393. [17] G.A. Cavagna, Force platforms as ergometers, J. Appl. Physiol. 39 (1) (1975) 174–179. [18] G.A. Cavagna, H. Thys, A. Zamboni, The sources of external work in level walking and running, J. Physiol. 262 (3) (1976) 639–657. [19] C. Detrembleur, J.M. Vanmarsenille, F. De Cuyper, F. Dierick, Relationship between energy cost, gait speed, vertical displacement of centre of body mass and efficiency of pendulum-like mechanism in unilateral amputee gait, Gait Posture 21 (3) (2005) 333–340. [20] W.D. McArdle, F.I. Katch, V.L. Katch, Exercise Physiology: Energy, Nutrition, and Human Performance, 4th ed., William & Wilkins, Baltimore, 1996. [21] G.A. Cavagna, M. Kaneko, Mechanical work and efficiency in level walking and running, J. Physiol. 268 (2) (1977) 467–481. [22] J.B. Saunders, V.T. Inman, H.D. Eberhart, The major determinants in normal and pathological gait, J. Bone Joint Surg. Am. 35-A (3) (1953) 543–558. [23] S. Lobet, C. Hermans, G.J. Bastien, F. Massaad, C. Detrembleur, Impact of ankle osteoarthritis on the energetics and mechanics of gait: the case of hemophilic arthropathy, Clinl. Biomech. (Bristol. Avon) 27 (6) (2012) 625–631. [24] C. Detrembleur, A. Van den Hecke, F. Dierick, Motion of the body centre of gravity as a summary indicator of the mechanics of human pathological gait, Gait Posture 12 (3) (2000) 243–250. [25] D.C. Kerrigan, B.E. Viramontes, P.J. Corcoran, P.J. LaRaia, Measured versus predicted vertical displacement of the sacrum during gait as tool to measure biomechanical gait performance, Am. J. Phys. Med. Rehabil. 74 (1) (1995) 3–8. [26] F. Massaad, T.M. Lejeune, C. Detrembleur, Reducing the energy cost of hemiparetic gait using center of mass feedback: a pilot study, Neurorehabil. Neural Repair 4 (4) (2010) 338–347. [27] V.T. Inman, H.J. Ralston, F. Todd, Human Walking, Baltimore, Williams & Wilkins, London, 1981 Pp.154. [28] R.A. Brand, H.J. Yack, Effects of leg length discrepancies on the forces at the hip joint, Clinl. Orthop. Relat. Res. 17 (333) (1996) 2–80.
a gait pattern with less vertical COM displacement, as in normal walking. A follow-up of these patients 6 months after the training ended revealed that a 15% decrease in the energy cost was sustained. Thus, these compensatory strategies provide real LLD subjects walking similar to that of normal subjects. Several authors [8,26,27] have shown that a better energy cost may be explained by a significant decrease in Wtot. However, this observation was not corroborated by our results. Indeed, despite the significant difference in the energy cost between real LLD and a 4-cm artificial LLD, there was no significant difference in Wtot (P = 0.497). Increased co-activation of the antagonist muscles in the lower limb (possibly an adaptation to ensure adequate joint stability) may partly explain the observed high energy cost and muscle efficiency of our subjects with a 4 cm artificial LLD. In summary, our subjects with real LLD had adopted compensatory strategies that allowed them to minimize energy consumption. Moreover, the compensatory strategies were probably acquired by experience, unlike the simulated subjects who did not have time to adapt. Nevertheless, our results underline the importance of assessing kinematics and kinetics data in future studies in order to establish the relationship between these parameters and the alterations in gait and to measure the impact of LLD on the joints and muscles. 5. Conclusion We showed a decrease in total mechanical work and the energy cost of walking in subjects with real LLD compared to subjects without LLD and a lower energy cost compared to subjects with artificial LLD. The probable use of compensatory strategies, acquired by experience in subjects with real LLD patients, allowed them to reduce energy expenditure thus facilitating their ability to walk. Quantitative analysis of the energy variables is fundamental in the management of subjects with LLD because it can help guide therapeutic choices. Conflict of interest The authors of this manuscript have no financial and personal relationships with other people or organizations that could inappropriately influence/biased their work. References [1] PA. Willems, B. Schepens, C. Detrembleur, Marche normale. EMC (Elsevier Masson SAS, Paris) Kinésithérapie-Médecine physique-Réadaptation 2012; 26-007-B-75. [2] A. Jones, P.G. Silva, M. Colucci, A. Tuffanin, J.R. Jardim, J. Natour, Evaluation of immediate impact of cane use on energy expenditure during gait in patients with knee osteoarthritis, Gait Posture 35 (3) (2012) 435–439. [3] B. Gurney, Leg length discrepancy, Gait Posture 15 (2) (2002) 195–206. [4] S.I. Subotnick, Limb length discrepancies of the lower extremity (the short leg
151
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
The effects of real and artificial Leg Length Discrepancy on mechanical work and energy cost during the gait
MARK
⁎
T.F. Assogbaa, , S. Bouletb, C. Detrembleura, P. Mahaudensa,b,c a Université Catholique de Louvain, Secteur des Sciences de la Santé, Institut de Recherche Expérimentale et Clinique, Neuro Musculo Skeletal Lab (NMSK), Avenue Mounier 53, B-1200 Brussels, Belgium b Cliniques Universitaires St Luc, Service de médecine physique et réadaptation, Avenue Hippocrate 10, B-1200 Brussels, Belgium c Cliniques Universitaires St Luc, Service d’orthopédie et de traumatologie de l’appareil locomoteur, Avenue Hippocrate 10, B-1200 Brussels, Belgium
A R T I C L E I N F O
A B S T R A C T
Keywords: Leg length discrepancy Mechanical work Energy cost Gait
Background: The impacts of Leg Length Discrepancy (LLD) on the kinematic and dynamic parameters of walking have been widely discussed. But little is known on total mechanical work and energy cost. These two variables are more representative of the functional impairment undergone by the LLD patients. Aim: To assess the changes of the mechanical work and energy cost of walking in subjects with real LLD and to compare their results with healthy subjects in whom the LLD has been simulated. Method: The mechanical work and energy cost data of 60 healthy subjects (speed: 4 km/h) with artificial LLD induced by soles (2 and 4 cm), 20 patients (speed: 3.75 ± 0.5 km/h) with real LLD and 20 matched subjects (speed: 3.75 ± 0.5 km/h) were collected. Statistical comparisons between the groups were performed using a tpaired test and ANOVA. Results: Patients with a real LLD showed a significant decrease in mechanical work and energy cost when compared to norms. Patients with real LLD provide a better recovery when compared to subjects with artificial LLD of 2 cm, and a decrease of energy cost and higher muscular efficiency (mechanical work/energy cost) when compared to subjects with artificial LLD of 4 cm. Conclusions: Our results showed that patients with a real LLD develop compensatory strategies during gait, probably to minimize the displacement of the body center of mass and consequently reduce the amount of energy expenditure useful for their displacement. Moreover, they adopt a better gait strategy compared to the subjects in whom LLD was simulated.
1. Introduction Walking is the usual mode of human locomotion [1]. However, musculoskeletal pathology may alter our ability to move, disturbing gait, increasing energy costs and, consequently, reducing our autonomy. For example, the use of walking cane in patients with knee osteoarthritis has been observed to cause an increase in energy expenditure during walking [2]. Leg Length Discrepancy (LLD), whether or not linked to a pathological process, is one of the most frequent musculoskeletal conditions encountered in clinical practice. LLD, defined as a condition in which the lower limbs are noticeably unequal in length [3], LLD is identified in as many as 40% [4] to 70% [5] of the population. Two etiological groups can be characterized, based on structural versus functional alterations: structural LLD is defined as real LLD associated with shortening of bony structures; functional LLD is defined as LLD that is the
⁎
result of altered mechanics of the lower extremities [6]. Several authors have shown that LLD (greater than 20–30 mm) can produce changes in gait. These include increased ground reaction forces [7–9], increased lower extremity kinetic energy [10] and increased mechanical work [11,12], defined as the total positive mechanical work (Wtot) done by the muscles during walking, in subjects with real LLD, and increased oxygen consumption (VO2) in subjects with artificial LLD induced by wearing a shoe with a sole [13]. Song et al. measured mechanical work during gait in children with real LLD. They observed that the children used several techniques to compensate for the LLD, including toe walking, vaulting, circumduction, and increased flexion of the longer limb. In addition, they noted that the longer limb performed more mechanical work than did the shorter. The average LLD for subjects with no observable compensatory strategy was 16.4 mm [12]. To our knowledge, energy expenditure has not been studied in real LLD. Gurney et al. created an artificial LLD
Corresponding author. E-mail addresses: [email protected], [email protected] (T.F. Assogba).
http://dx.doi.org/10.1016/j.gaitpost.2017.10.004 Received 15 May 2017; Received in revised form 24 August 2017; Accepted 3 October 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
trajectories of reflective markers positioned on specific anatomical landmarks. The subject was fitted with a mask to measure VO2 and carbon dioxide production (VCO2) throughout the treadmill test using an ergospirometer (Medisoft, Belgium). The sessions began with a rest period, in which the subjects stood barefoot for the static calibration of kinematic and energetic variables. Thereafter, the subjects were asked to walk on a treadmill for a few minutes. For group 1, all subjects walked at a constant speed of 4 km h−1 for each trial (with and without soles). For group 2, the subjects walked at a spontaneous speed that had been measured earlier using the 10 m walk test [15]. Group 3 subjects walked on the treadmill at different speeds of 1, 2, 3, 4 and 5 km h−1 (in order to establish norms). Each subject walked for a few minutes until a steady state was reached and maintained for at least 2 min to record energetic variables. Kinematics and kinetics variables were then recorded simultaneously and averaged over 10 successive strides. The mean for each data was used for statistical analysis. In group 1, LLD was induced using 2 cm and 4 cm soles (Airlit) attached to a sports shoe. The order in which the soles were used was random and determined by drawing lots. Before gait analysis, the subjects of all groups walked on the treadmill for at least 10 min in a practice session to get used to the treadmill, and also to the different soles for group 1.
using shoe soles in healthy older persons (ages 55–86) during gait. The subjects had significantly greater energy consumption at 20, 30 and 40 mm of LLD compared to no LLD [13]. Normal human walking is governed by a multitude of interrelated parameters: kinematics, kinetics and energetics. Understanding these parameters that describe gait biomechanics is essential to improve the functional activity of persons suffering from pathology that affects walking. The role played by energetics, which may be a good indicator of the difficulty a subject has to walk, is often neglected; its evaluation could contribute to better therapeutic decisions for improving gait in individuals with LLD. We hypothesized that people with real LLD have an altered walking mechanism and adapt their walking strategy to limit excessive energy expenditure better than do people with experimentally artificial LLD. The current study was therefore designed first to assess the effects of real LLD on mechanical work and energy cost during walking and second to compare the energetic changes in subjects with real and artificial LLD. 2. Method 2.1. Participants The subjects were divided into three groups. The first was composed of 60 healthy adult subjects aged 18–29 years (22 ± 1 years, group 1) who participated in the experiment, and walked with and without soles of 2 and 4 cm. The second group was composed of 20 young subjects with real LLD aged 5–19 years (12 ± 3 years, group 2). The third group was composed of 20 young subjects (12 ± 3 years, group 3, norm) without LLD matched in age and speed with each subject of group 2. The sex ratio and demographic data of the three groups are shown in Table 1. Group 1 subjects were recruited from university personnel between January and March 2016. Group 2 subjects were recruited after consultation in the orthopedic service of our hospital between January and December 2016. Subjects with an LLD of less than 1 cm or who had undergone any surgery, fracture or immobilization of the lower limbs in the six months prior to the evaluation were excluded. Group 3 subjects, with no lower limb orthopedic history were recruited from our entourage (professional and social). Subjects less than 5 years of age or with more than 1 cm of inequality or who had trauma to the pelvis or limb in the six months prior to the study were not included. Every subject or parent gave signed consent and participated freely in the study, which was approved by the local ethics board.
2.3. Parameters Mechanical work was computed as follows: the total positive mechanical work (Wtot) done by the muscles during walking was divided into the external work (Wext) performed to move the body Center of Mass (COM) relative to surroundings and the internal work (Wint) performed to move the body segments relative to the COM [16]. Wext was computed from four force transducers located at the four corners of the treadmill. These transducers measured the 3D-ground reaction forces according to Cavagna [17]. Wext represents the positive work and is divided into the forward work (Wekf), vertical work (Wekv) and lateral work (Wekl) necessary to accelerate the COM in the three directions and lift the COM during a stride. The Recovery, quantifying the percentage of mechanical energy saved by a pendulum-like exchange between the gravitational potential energy and the kinetic energy of the COM (i.e., an index reflecting the effectiveness of the pendulum-like mechanical mode of walking), was measured [17,18]. Wint, the work required to move the limbs relative to the COM was computed from kinematic data following the methods described by Willems et al. [16] and by Detrembleur et al. [19]. The metabolic cost was determined by the subject’s VO2 and VCO₂. The mass specific gross energy consumption rate (W kg−1) was obtained from the VO2 using the energy equivalent of oxygen, taking into account the measured respiratory exchange ratio (RER). Each energy measurement started with a rest period in which the subject was standing on the treadmill. Thereafter, they walked until a steady state was reached and maintained for at least 2 min. The Joules of energy expended per liter of oxygen consumed was computed depending on the RER according to the Lusk equation [20]. The energy expended above the resting value (standing subtracted from walking consumption) was divided by the walking speed to obtain the net energy cost of walking (C, J/kg m−1) [21]. The efficiency (ɳ) of the positive work production by the muscles was calculated as the ratio between Wtot and C [21].
2.2. Procedures Gait was assessed using a three-dimensional analysis, which included synchronous kinematics, kinetics, mechanical and energetic measurements [14] (Fig. 1). All subjects walked on a motor-driven treadmill (Mercury LTmed, HP Cosmos®, Germany) equipped with force captors and surrounded by eight infrared cameras at 200 Hz (BTS, Milan, Italy) to measure the
Table 1 Sex ratioand demographic data. Group 1 n = 60 Mean ± SD
Group 2 n = 20 Mean ± SD
Group 3 n = 20 Mean ± SD
22 ± 1 173.0 ± 8.0 67.6 ± 9.1
12 ± 3 153.3 ± 23.5 51.5 ± 25.8
12 ± 3 157 ± 18.7 39.9 ± 13.0
2.4. Statistical analysis Age (Years) Height (cm) Weight (kg)
All variables that had a normal distribution and equality of variance are presented as means ( ± SD). Other variables are expressed as medians and quartiles [25–75%]. Statistical analysis was performed
SD: Standard Deviation.
148
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
Fig. 1. Energetic and mechanical measurements. Illustration of energy expenditure (oxygen consumption and energy cost) measured by the classical indirect calorimetric method and mechanical work assessed by the three-dimensional ground reaction force. The infrared cameras (A) are positioned so that at least two visualize each reflective marker at any given time. From the reflective markers movements we can calculate the 3D trajectories of the body segments (B). A force platform located under the treadmill (B) records the patient’s ground reaction forces. Energy expenditure is measured indirectly based on the rate of oxygen consumption by the patient using an ergo spirometer (C). Finally the mechanical work (D) is calculated as the work performed by muscles to raise and accelerate the center of body mass (external mechanical work) and to move the body segments relative to the center of body mass (internal mechanical work). Adapted from Lobet et al. [14]. Adapted from Lobet et al. [14]
3.2. Real LLD versus artificial LLD (Table 3)
using SigmaPlot version 13.0 (SPSS 2014). The significance level was set at P ≤ 0.05. A paired t-test was used to determine whether the subjects with real LLD had different mechanical work and energy cost compared to group 3. Subjects with real LLD were thus matched in age and speed with subjects without LLD (group 3). To compare real and artificial LLD, age was normalized using a Zscore transformation, and we then we used a One-Way Analysis of Variance (ANOVA) or Kruskal-Wallis One Way Analysis of Variance on Ranks (if normality and equality of variance tests not passed) with one factor (factor = groups i.e. real LLD, artificial 2 cm and 4 cm). A posthoc test with Dunn’s method was used to identify significantly different variables.
There were no statistically significant differences in the Wext, Wint and Wtot Z-scores of subjects with real LLD and artificial LLD of 2 and 4 cm. There were significant differences in the Z-scores of recovery, net energy cost and efficiency, with p values of 0.005, ˂0.001 and 0.004, respectively, for real LLD and artificial LLD of 2 and 4 cm. Post-hoc
Table 2 Results of Paired t-test on mechanical and energetic variables in subjects with real LLD versus matched subjects without LLD. Group 2 n = 20 Mean ± SD
Group 3 n = 20 Mean ± SD
P-value
Mechanics Wext (J/kg m) Wint (J/kg m) Wtot (J/kg m) Wekf (J/kg m) Wekv (J/kg m) Wekl (J/kg m) Recovery (%)
0.25 ± 0.06 0.25 ± 0.03 0.50 ± 0.07 0.26 ± 0.05 0.37 ± 0.07 0.009 ± 0.007 58.6 ± 12.9
0.29 ± 0.05 0.28 ± 0.06 0.57 ± 0.06 0.33 ± 0.04 0.44 ± 0.03 0.009 ± 0.005 62.4 ± 5.8
0.019 0.013 0.002 < 0.001 0.002 0.657 0.182
Energetics Net energy cost (J/kg m) Muscle efficiency (%)
2.05 ± 0.45 22.10 ± 5.34
2.34 ± 0.28 25.78 ± 5.12
0.032 0.152
3. Results 3.1. Real LLD versus matched subjects without LLD (Table 2) The Wtot was 12% lower in subjects with real LLD than in matched subjects without LLD (P = 0.002) and the different components of Wtot, i.e., external work and internal work, were 13% (P = 0.019) and 10% (P = 0.013) lower, respectively. For the three components of the Wext, only Wekf and Wekv were statistically significantly lower in the subjects with real LLD than in matched subjects without LLD, both by 0.07 J/kg. (P = ˂0.001 and 0.002, respectively). The net energy cost was 12% lower (P = 0.032). There were no significant differences in recovery and efficiency between real LLD subjects without LLD.
Significant differences are typed in bold and are accepted for P-value ≤ 0.05. SD: Standard Deviation.
149
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
increasing pelvic obliquity, circumduction, increased hip and/or knee flexion (steppage gait), more ankle dorsiflexion, or any combination of these actions [8]. Further studies should compare kinematic and mechanical parameters, for example the 3D displacement of COM in LLD subjects during gait, to corroborate this hypothesis. The decrease in Wint could also be explained by an alteration in some walking determinants, such as reduced angular speed of lower limbs joints, which were not evaluated in this study. This assumption is reinforced by a study by Lobet et al. [23]. These authors reported, in walking in hemophiliacs, that a reduction in Wint could be provided by a decrease in the range of motion at the level of the affected ankle and hip, and a subsequent decrease in joint angular speed. The decrease in Wext and Wint induced a decrease in the Wtot. Our result suggests that subjects with real LLD economize their positive muscular work, which may facilitate the recovery of the pendular mechanism necessary to walk close to normal and consequently may help guide the choice of appropriate therapies for these subjects. In fact, our subjects with real LLD had a recovery that was close to that of subjects without LLD (58% vs 62%), conserving the energy exchange between the kinetic and the potential energy of COM. Our results are in accordance with other studies showing that this pendulum-like mechanism is also intact in some orthopedic pathologies [23,24]. These studies demonstrated that subjects with lowest-level gait disorder were able to move their bodies in a way that allowed a normal amount of external muscular work to be performed per unit distance. Our results showed that subjects with real LLD (group 2) had a lower energy cost of walking compared to subjects without LLD (group 3). Studies by Cavagna [21] and Willems [16] have highlighted that energy cost depends on muscle work requirements and on the efficiency of muscle work production. In addition, Kerrigan et al. [25], demonstrated that vertical displacement of COM reliably predicted VO2 during gait. Based on these observations, a low vertical displacement of the COM may have induced a low Wekv and thus a low Wext, with a reduction in the VO2 in our subjects with real LLD. This observation is strengthened by a study by Massaad et al. who investigated the energetics of human walking by decreasing (flat) and increasing (bouncy) the vertical COM displacement in healthy subjects walking at different speeds and showed that the vertical displacement of COM appeared to be a determinant of the energy cost of walking [26]. According to Kaufman, it is possible that individuals use compensatory mechanisms to dynamically shorten the long limb, probably to minimize the displacement of the COM and consequently reduce energy expenditure [8]. Inman showed that compensatory strategies dampen oscillations of the COM and decrease over-all energy expenditure during gait [27]. Our results are in agreement with these results from other studies.
Table 3 Results of ANOVA on mechanical and energetic variable in subjects with real LLD versus artificial LLD (2 cm and 4 cm). Median
P-value
[25%–75]
Mechanics Z-score Wext* Z-score Wint Z-score Wtot. Z-score Recovery Energetics Z-score net energy cost Z-score efficiency
Real LLD n = 20
aLLD 2 cm n = 60
aLLD 4 cm n = 60
−0.48 ± 0.99
−0.29 ± 1.33
−0.58 ± 1.36
0.497
−0.70 [−1.06/ −0.23] −0.75 [−1.22/ −0.24] −0.58 [−2.78/ 0.79]
−0.41 [−0.69/ −0.13] −0.59 [−1.12/ 0.05] −1.74 [−2.11/ −1.12]
−0.45 [−0.73/ −0.10] −0.80 [−1.36/ 0.03] −1.05 [−1.68/ −0.45]
0.224
−0.55 [−0.90/ 0.53]
0.26 [−0.29/ 0.89]
0.59 [0.09/ 1.24]
< 0.001b
0.07 [−0.81/ 1.20]
−0.64 [−1.11/ −0.12]
−0.87 [−1.27/ −0.40]
0.004c
0.479 0.005a
Significant differences are typed in bold and are accepted for P-value ≤ 0.05. a Real LLD ≠ aLLD 2 cm (P = 0.029). b Real LLD ≠ aLLD 4 cm (P ˂ 0.001). c Real LLD ≠ aLLD 4 cm (P = 0.003). * Mean aLLD = artificial LLD.
analysis using Dunn’s method also revealed significant differences. Thus, for recovery, real LLD ≠ artificial 2 cm LLD (Z-score real LLD closer to normal than Z-score 2 cm; P-value = 0.029), for net energy cost and efficiency, real LLD ≠ artificial 4 cm LLD (Z-score real LLD closer to normal than Z-score 4 cm LLD, P-value respectively ˂0.001 and 0.003). 4. Discussion 4.1. Real LLD versus matched subjects without LLD We assessed the effects of real LLD on mechanical work and energy cost during walking. These two parameters were significantly reduced in subjects with real LLD (group 2) compared to those without LLD (group 3). Recovery and efficiency were closer to the values of the subjects without LLD. The reduced mechanical work in subjects with real LLD shown in this study has not been observed in other studies. However, our subjects with real LLD had a length difference of 2.02 cm, by contrast with previous studies that have assessed inequalities of more than 2–3 cm, generating increased mechanical work [11,12]. The decrease in Wext observed in the real LLD subjects, may be due to the decrease in both forward (Wekf) and vertical (Wekv) components, which were 21% and 16% lower, respectively. We hypothesized that the decrease in these two components would smooth the COM displacement in vertical and forward directions. Various kinematics strategies can explain this COM smoothing, including excessive rotation of the pelvis, increased pelvic obliquity or knee flexion during the single leg stance or significant lateral movements of the pelvis during gait [22]. According to Saunders, smoothing of COM displacement can in turn explain the low energy cost of walking [22]. These various elements lead these subjects to develop various strategies of compensation. Kaufman suggested that subjects with real LLD compensate for their shorter limb by increasing downward pelvic obliquity, performing knee extension in midstance, vaulting, toe walking, or any combination of these. In addition, the longer leg can be artificially shortened by
4.2. Real LLD versus artificial LLD When we compare our results obtained in subjects with real LLD to those with artificial LLD, the recovery of real LLD subjects was better than that of subjects with 2 cm artificial LLD. We hypothesized that subjects with real LLD would adopt a better walking strategy than subjects with an artificial LLD, improving pendular mechanism of gait. Our subjects with real LLD patients also had lower energy cost and greater walking efficiency compared to subjects with a 4 cm artificial LLD. These differences would indicate a long-term positive effect and likely sustained motor learning that is not possible when someone is asked to walk with an artificial LLD. According to Brand and Yack [28], adults did not to show compensatory strategies when walking with experimentally induced uneven leg lengths over short periods. The compensatory strategies developed by subjects with real LLD were acquired by experience and long-term learning effects, unlike the simulated subjects who did not have time to adapt. This suggestion is reinforced by Maassad et al. [25] who demonstrated that after only 6 weeks of a gait training program with COM biofeedback, patient with hemiparesis decreased their energy consumption by 30% by practicing 150
Gait & Posture 59 (2018) 147–151
T.F. Assogba et al.
syndrome), J. Orthop. Sports Phys. Ther. 3 (1) (1981) 11–16. [5] A.L. Woerman, S.A. Binder-MacLeod, Leg length discrepancy assessment: accuracy and precision in five clinical methods of evaluation, J. Orthop. Sports Phys. Ther. 5 (5) (1984) 230–239. [6] R.L. Blake, B. Ferguson, Limb length discrepancies, J. Am. Podiatr. Med. Assoc. 82 (1) (1992) 33–38. [7] A. Bhave, D. Paley, J.E. Herzenberg, Improvement in gait parameters after lengthening for the treatment of limb-length discrepancy, J. Bone Joint Surg. Am. 81 (4) (1999) 529–534. [8] K.R. Kaufman, L.S. Miller, D.H. Sutherland, Gait asymmetry in patients with limblength inequality, J. Pediatr. Orthop. 16 (2) (1996) 144–150. [9] D. Schuit, M. Adrian, P. Pidcoe, Effect of heel lifts on ground reaction force patterns in subjects with structural leg-length discrepancies, Phys. Ther. 69 (8) (1989) 663–670. [10] F.G. Delacerda, O.D. Wikoff, Effect of lower extremity asymmetry on the kinematics of gait, J. Orthop. Sports Phys. Ther. 3 (3) (1982) 105–107. [11] X.C. Liu, G. Fabry, G. Molenaers, J. Lammens, P. Moens, Kinematic and kinetic asymmetry in patients with leg-length discrepancy, J. Pediatr. Orthop. 18 (2) (1998) 187–189. [12] K.M. Song, S.E. Halliday, D.G. Little, The effect of limb-length discrepancy on gait, J. Bone Joint Surg. Am. 79 (11) (1997) 1690–1698. [13] B. Gurney, C. Mermier, R.A. Robergs, A.L. Gibson, D. Rivero, Effects of leg length discrepancy on gait economy and lower extremity muscle activity in older adults, J. Bone Joint Surg. Am. 83-A (6) (2001) 907–915. [14] S. Lobet, C. Detrembleur, F. Massaad, C. Hermans, Three-dimensional gait analysis can shed new light on walking in patients with haemophilia, Sci. World J. 13 (2013) 284358 7 pages. [15] D.T. Wade, Measurement in neurological rehabilitation, Curr. Opin. Neurol. Neurosurg. 5 (1992) 682–686. [16] P.A. Willems, G.A. Cavagna, N.C. Heglund, External, internal and total work in human locomotion, J. Exp. Biol. 198 (Pt. 2) (1995) 379–393. [17] G.A. Cavagna, Force platforms as ergometers, J. Appl. Physiol. 39 (1) (1975) 174–179. [18] G.A. Cavagna, H. Thys, A. Zamboni, The sources of external work in level walking and running, J. Physiol. 262 (3) (1976) 639–657. [19] C. Detrembleur, J.M. Vanmarsenille, F. De Cuyper, F. Dierick, Relationship between energy cost, gait speed, vertical displacement of centre of body mass and efficiency of pendulum-like mechanism in unilateral amputee gait, Gait Posture 21 (3) (2005) 333–340. [20] W.D. McArdle, F.I. Katch, V.L. Katch, Exercise Physiology: Energy, Nutrition, and Human Performance, 4th ed., William & Wilkins, Baltimore, 1996. [21] G.A. Cavagna, M. Kaneko, Mechanical work and efficiency in level walking and running, J. Physiol. 268 (2) (1977) 467–481. [22] J.B. Saunders, V.T. Inman, H.D. Eberhart, The major determinants in normal and pathological gait, J. Bone Joint Surg. Am. 35-A (3) (1953) 543–558. [23] S. Lobet, C. Hermans, G.J. Bastien, F. Massaad, C. Detrembleur, Impact of ankle osteoarthritis on the energetics and mechanics of gait: the case of hemophilic arthropathy, Clinl. Biomech. (Bristol. Avon) 27 (6) (2012) 625–631. [24] C. Detrembleur, A. Van den Hecke, F. Dierick, Motion of the body centre of gravity as a summary indicator of the mechanics of human pathological gait, Gait Posture 12 (3) (2000) 243–250. [25] D.C. Kerrigan, B.E. Viramontes, P.J. Corcoran, P.J. LaRaia, Measured versus predicted vertical displacement of the sacrum during gait as tool to measure biomechanical gait performance, Am. J. Phys. Med. Rehabil. 74 (1) (1995) 3–8. [26] F. Massaad, T.M. Lejeune, C. Detrembleur, Reducing the energy cost of hemiparetic gait using center of mass feedback: a pilot study, Neurorehabil. Neural Repair 4 (4) (2010) 338–347. [27] V.T. Inman, H.J. Ralston, F. Todd, Human Walking, Baltimore, Williams & Wilkins, London, 1981 Pp.154. [28] R.A. Brand, H.J. Yack, Effects of leg length discrepancies on the forces at the hip joint, Clinl. Orthop. Relat. Res. 17 (333) (1996) 2–80.
a gait pattern with less vertical COM displacement, as in normal walking. A follow-up of these patients 6 months after the training ended revealed that a 15% decrease in the energy cost was sustained. Thus, these compensatory strategies provide real LLD subjects walking similar to that of normal subjects. Several authors [8,26,27] have shown that a better energy cost may be explained by a significant decrease in Wtot. However, this observation was not corroborated by our results. Indeed, despite the significant difference in the energy cost between real LLD and a 4-cm artificial LLD, there was no significant difference in Wtot (P = 0.497). Increased co-activation of the antagonist muscles in the lower limb (possibly an adaptation to ensure adequate joint stability) may partly explain the observed high energy cost and muscle efficiency of our subjects with a 4 cm artificial LLD. In summary, our subjects with real LLD had adopted compensatory strategies that allowed them to minimize energy consumption. Moreover, the compensatory strategies were probably acquired by experience, unlike the simulated subjects who did not have time to adapt. Nevertheless, our results underline the importance of assessing kinematics and kinetics data in future studies in order to establish the relationship between these parameters and the alterations in gait and to measure the impact of LLD on the joints and muscles. 5. Conclusion We showed a decrease in total mechanical work and the energy cost of walking in subjects with real LLD compared to subjects without LLD and a lower energy cost compared to subjects with artificial LLD. The probable use of compensatory strategies, acquired by experience in subjects with real LLD patients, allowed them to reduce energy expenditure thus facilitating their ability to walk. Quantitative analysis of the energy variables is fundamental in the management of subjects with LLD because it can help guide therapeutic choices. Conflict of interest The authors of this manuscript have no financial and personal relationships with other people or organizations that could inappropriately influence/biased their work. References [1] PA. Willems, B. Schepens, C. Detrembleur, Marche normale. EMC (Elsevier Masson SAS, Paris) Kinésithérapie-Médecine physique-Réadaptation 2012; 26-007-B-75. [2] A. Jones, P.G. Silva, M. Colucci, A. Tuffanin, J.R. Jardim, J. Natour, Evaluation of immediate impact of cane use on energy expenditure during gait in patients with knee osteoarthritis, Gait Posture 35 (3) (2012) 435–439. [3] B. Gurney, Leg length discrepancy, Gait Posture 15 (2) (2002) 195–206. [4] S.I. Subotnick, Limb length discrepancies of the lower extremity (the short leg
151