Introduction
Introduction
Gait and posture are both controlled by motor programs that are partly genetically determined and partly learned. Locomotion may be regarded as a program which uses a combination of both. Sedentary adults have an automatic, alternate bilateral, totally symmetric, and reproducible locomotion [17]
[22]. The characteristic patterns of locomotion that are seen in the adult are not achieved until the child reaches the age of 7 to 9 years [24]. Thus, high level sportsmen usually begin their activity before completely achieving their locomotor apprenticeship. Intensive sports training involves learning automatic movements. The execution of these movements has to become stereotyped and reproducible. The motor programs required are inevitably different from one sport to another. The playing of some sports such as soccer, tennis or basketball demands a high degree of postural balance and quick movement changes. Postural stability is necessary to avoid any risk of falling during such activities [8]
[13]
[15]. The preferential use of an item such as a racket, a ball or a sword by one segment from the upper or the lower limb permits the definition of a sport activity as bilateral, preferential unilateral or strictly unilateral. The intensive training of a unilateral sport can induce the development of muscular asymmetry [1]
[16].
In a previous study, it has been shown that the locomotion of sportsmen practicing intensively preferential unilateral sports such as basketball and fencing results in the development of some asymmetrical gait variables [11]. In right-handed high level fencers and basketball players, the main difference was that the right propulsion double support duration was found to be longer than the left one.
Basketball is to some extent a unilateral sport. Most basketball players have a preferential lay up (right then left stance). These preferences depend on the handedness. A right-handed basketball player has most of the time a right-left lay up (83 %) [2]. Basketball is particularly an activity of jumps (about 30 jumps per match), of a succession of accelerations and decelerations, and a great number of changes of direction.
In soccer, the player often runs with a ball at his/her feet. Soccer players' foot laterality (dominant foot) depends on their shoot and dribble foot. Thus, soccer is also a preferential unilateral activity. This activity requires long runs with or without the ball, shots and dribbles with one preferential lower limb and few jumps. A right-handed soccer players usually uses his right foot to shoot and dribble the ball, and his left foot to support the body before a jump or a shot. This requires a predominantly left monopodous equilibrium stance to be learned [9].
Swimming is essentially a bilateral activity. Whatever the swimming style used, the two segments of the lower or upper limbs reproduce almost the same symmetric movements, simultaneously or alternatively according to the style. The aim of this study was to evaluate whether an intensive sport activity can have any effect upon the normal walking. If so, have these sportsman a sport-specific gait pattern? Gait variables of a preferentially unilateral handtool activity (basketball), of a preferentially unilateral foot tool activity (soccer) both using locomotion and of a bilateral activity not using locomotion (swimming) were compared. Any differences in spatial and temporal gait variables between right and left sides within and between each of these three groups of high level athletes were sought.
Method
Method
Studied population: Thirty-six right-handed male subjects (10 basketball players, 10 swimmers and 16 soccer players) were recruited during a medical check. They accepted their gait being recorded. Since they were all athletes, a sports physician checked that no players had any of the following problems: disorders of the musculoskeletal system, foot disorders likely to impair walking, a prior orthopedic corrective or joint replacement procedure in the lower limbs, low back pain or/and nerve root pain, congenital or acquired lumbar spinal stenosis; neurologic disorders, including stroke-related residual disabilities responsible for gait disorders, and cerebellar or vestibular syndrome. Moreover, those athletes with obesity, lower limb length discrepancy exceeding 1.5 cm, recent sprain of a knee or ankle, knee or ankle instability, and inflammatory joint disease were excluded. Handedness was checked by the “Edinburgh Handedness Inventory” which permitted the distinction between absolute, preferential or ambidextrous right or left handedness [19]. Ten swimmers aged 20.0 ± 3.6 (M ± SD), ten basketball players aged 22.9 ± 7.4, and 16 soccer players aged 23.0 ± 3.0 were included in this study. They were all competing at a national level. They had all played for at least 5 years, 8 hours a week and were regularly examined in our laboratory. All of the athletes were male and right-handed. Basketball players all preferentially jumped with their left leg, whereas soccer players all kicked with their right foot (being on a left stance). Swimmers used either a crawl or butterfly stroke. The age, stature, weight of the studied sportsmen are shown in Table [1]. Use of an ANOVA test failed to indicate any statistical difference between the three groups for these anthropometric data, although the data for basketball players were more variable.
Material: The apparatus developed by Bessou et al. [4]
[8] simultaneously records longitudinal displacements of both feet during walking (Fig. [1]). Each foot is secured to a non-elastic rope, which is wrapped around a pulley system connected to a potentiometer. The pulley system comprises two pulley blocks, each of which has four pairs of pulleys of decreasing diameters. The upper block is fixed and the lower one is mobile. After leaving the lower block, the cord wraps around a large pulley whose center is attached to the axis of a potentiometer (one for each foot). The cords are kept taut by a ten kilo counterweight. After reduction, displacements of each foot cause rotation of the axis of the potentiometer, which receives a direct current. The signal is produced by changes in voltage over time. The voltage delivered by each potentiometer is proportional to the distance travelled by the foot to which it is attached. The acquired signal is converted by a computer (200 Hz) and analyzed with a specific software allowing measurement to the nearest 5 mm and timing to the nearest 5 ms [12]. Fig. [2] shows a reference gait recording. Bessou's gait analyzer permits simultaneous recording of the longitudinal displacements of both feet during the gait. Thus spatial (stride length and step length), temporal variables (cycle, stance, swing, double support, early swing phase, and late swing phase durations), and walking speed were characterized. Stance, swing and double support durations are expressed as percentage of the cycle durations. Early and late swing phase durations are expressed as percentage of the swing durations. The right propulsion double support duration is the same as the left loading double support duration. Each subject was instructed to walk from the apparatus to an object placed at the end of the recording distance (seven meters), at their normal speed, striking out with the right foot and pulling on the wires to activate the apparatus. Walking was not hindered by the pull of the wires, because the load was only 3 Newtons per foot. Subjects wore their own shoes. No instruction concerning speed was given. Each subject was recorded twice; the first recording was used to make the athletes familiar with the device, and these initial data were not used.
Statistic analysis: The Student's t-test for paired data was used to compare quantitative variables between right and left sides of each group of subjects. ANOVA was used to compare quantitative variables between basketball players, swimmers and soccer players. When significant differences were found, a post hoc test (Sheffé test) was performed to determine differences between groups. All results are expressed as Mean ± Standard Deviation (M ± SD). Values of p under 0.05 were considered significant. Finally, two discriminant analyses were performed on 36 subjects of the three groups respectively with the anthropometric data only, and with all the gait variables. A discriminant analysis is used to determine if some groups are different according to their variables. A first equation is calculated to optimize the combination of variables in order to provide the most discriminative function (X-axis), then a second equation which provides the second and high discrimination function (Y-axis). Moreover, the two functions are independent. This means that their contributions to the discrimination between groups will not overlap.
Results
Results
Analysis by sport group: Table [2] shows mean values for spatiotemporal gait variables in each sport group. As regards the swimmers, comparing right and left sides, the Student's t-test for paired data failed to point out any statistical difference for all variables. In the analysis of the basketball players, the right cycle duration (p = 0.022), right propulsion double support percentage (p = 0.004), right late swing phase percentage (p = 0.025) were found to be significantly higher than the left side, while the left early swing phase percentage was longer than the right side (p = 0.025). In the soccer players group, the right step length (p = 0.016), right propulsion double support percentage (p = 0.014) and right late swing phase percentage (p = 0.001) were significantly higher than the left side. The left late swing phase percentage was found to be longer than the left side (p = 0.001). The right and left propulsion double support durations in basketball players, soccer players and swimmers are represented in Fig. [3].
Analysis between the three groups: ANOVA revealed that some variables were significantly different between the soccer players, the swimmers and the basketball players. Walking speeds of the three groups were comparable whereas cadences were significantly different (p < 0.001). The statistical data did not reach significance between the three groups for the stride length, nor for the step length. Cycle durations for both sides were different between the three groups (p = 0.001). The gait cycle duration is equal to the sum of the swing duration and of the stance duration. The stance and swing percentages (right and left sides) did not differ among these three groups. The swing duration is the sum of the early swing phase duration and of the late swing phase duration. The early swing phase is the period during which the swinging leg leaves the floor (toe off) to crossing the other leg (Fig. [2]). The late swing phase is the following period from the cross of the two legs to the heel stance. Percentages of these two variables were not different between the three groups for both sides. Sheffé tests were used to compare groups the two by two for the cadence. This test failed to point out any difference between the swimmers and the basketball players. The soccer players walked with a higher cadence (p = 0.001 for basketball players, and p = 0.015 for swimmers). These results are represented in Fig. [4]. The anthropometric data discriminant analysis did not reveal any differences associated with a sport (Fig. [5 a]). When comparing the gait variables by a discriminant analysis, the three athletic groups present greatly different gait patterns (Fig. [5 b]).
Discussion
Discussion
The aim of this study was to check whether the playing of a particular sport exerted any influence on gait pattern. Previous studies from this laboratory [22] or from Bessou et al. found no differences in gait between the right and left sides in healthy males and females of any age group. Murray et al. showed that walking was a highly reproducible activity in subjects younger than 60 years [18]. Moreover, the gait of athletes was not different from the gait variables of sedentary subjects obtained using the same apparatus in the same laboratory [22]. Thus, playing a sport like basketball, swimming or soccer does not induce a very different way of walking from normal nonathletic subjects. These results suggest that there is rather close inter-individual reproducibility of the human locomotion program, as has previously been reported [17]. However, despite the fact that the walking pattern of athletes is not very different from that of nonathletic subjects, some differences depending on the sport played were still found.
In those engaged in preferentially unilateral sports, the right propulsion double support duration was found to be longer than that of the left side, confirming a previous study [11]. However, this has not been observed in normal nonathletic subjects [4]
[17]
[18]
[22]
[24] nor was it seen in swimmers in this study. Most often, when right-handed sportsmen need to perform a shot, a dribble, a high or a long jump, the impulse side is the left foot [2]. In this study, all the subjects chosen were right-handed, and used their left foot as the impulse foot. During a jump or a shot, just after the right propulsion double support duration, the equilibrium of the body is based upon a monopodal left stance, while the right is swinging. Before each body movement, it is necessary to anticipate the postural equilibrium to keep a perfect balance during the jump or the shot. The anticipatory postural adjustments (APA) occur before starting a voluntary movement. They correspond to dynamic phenomena which are centrally pre-programmed [5]. These APA anticipate the perturbations of posture and equilibrium connected to the movement [14]. APA depends on the variables of the planned movement, posture and the task confronted. The final body equilibrium is the major parameter controlled. The postural equilibrium may be more difficult to establish if at the same time the upper limbs have a particular task to do (sending a ball in a precise direction). The postural anticipated equilibrium is mainly adjusted during the double support duration [7]
[13]
[15]
[23]. We suggest that our results imply that the higher right propulsion double support duration could be a consequence of a particular APA programmed to be used before the shot or the jump.
Some other variables (step length, cycle duration and early and late swing phase) were also found to be affected. Soccer and basketball are sports of movements and contact. The ball is gained or lost by a succession of sprints (soccer), accelerations/decelerations (basketball), violent efforts, jumps, and shots. These different actions not only involve some modifications of the central motor program, such as APA [15], but also a particular muscular development [20]. The muscular development of the soccer and basketball players has been found to be asymmetric since the dominant leg (the right one) was found to be stronger than the left one measured by isokinetic dynamometer [16]. An alternative explanation for differences (right vs. left) in gait variables might therefore be based upon differences of visco-elasticity properties of muscles between right and left sides, a consequence of the asymmetric muscular development. As regards the swimmers, during their activity in the water, their training does not involve any asymmetric muscular development, since swimming is totally bilateral.
In conclusion, playing sports such as basketball, soccer or fencing [11] over a period of years seems to have induced some permanent differences in the locomotion pattern between right and left sides, and from one sport to another. Longitudinal studies of the locomotion pattern of children practicing soccer or basketball might permit a better understanding of the appearance of this gait asymmetry. In this study, we have been unable to clarify which were due to asymmetric muscle development, which were a consequence of the central motor program, and which were due to the combination of these two factors. Nevertheless, the fact that the sportsmen recruited began to specialize in their sports before adult APA and locomotor programming was complete could be a factor explaining these asymmetries.
Acknowledgment
Acknowledgment
The authors thank Mr. Richard Medeiros for his advice in editing the manuscript. Excellent technical assistance was provided by Isabelle Bertoldi and Eric Dréano. Sincere appreciation is expressed to the subjects for their commitment to the study.
Fig. 1The gait analyzer developed by Bessou. a) Device to reduce displacements; b) Potentiometer; c) Motion recording system. Electric signals corresponding to the displacements are transmitted to the potentiometer. Signals are then retrieved by the motion recording system during a gait recording called locogram. This locogram is analyzed by a software which calculates gait variables.
Fig. 2Reference gait recording (locogram) obtained with the Bessou gait analyzer. Each curve represents the movement of one leg. “r” means propulsion double support duration.
Fig. 3Comparison of right and left propulsion double support percentage of the cycle (basketball players, swimmers and soccer players).
Fig. 4Comparison of cadence (basketball players, swimmers and soccer players).
Fig. 5 aDiscriminant analysis performed with the anthropometric data.
Fig. 5 bDiscriminant analysis performed with gait variables.
Table 1Anthropometric data of basketball players, swimmers and soccer players
| Age (years) | Height (cm) | Weight (kg) |
| M ± SD | M ± SD | M ± SD |
| Basketball players (n = 10) | 22.9 ± 7.4 | 185.2 ± 10.4 | 79.1 ± 15.7 |
| Swimmers (n = 10) | 20.0 ± 3.6 | 179.0 ± 8.7 | 73.1 ± 8.9 |
| Soccer players (n = 10) | 23.0 ± 3.0 | 179.1 ± 6.8 | 75.8 ± 5.1 |
Table 2Spatiotemporal gait parameters of basketball players, swimmers and soccer players
| Basketball Players (n = 10) | Swimmers (n = 10) | Soccer Players (n = 16) |
| Right Side | Left Side | Right Side | Left Side | Right Side | Left Side |
| Mean ± Standard Deviation | Mean ± Standard Deviation | Mean ± Standard Deviation |
| Stride length (m) | 1.56 ± 0.16 | 1.56 ± 0.15 | 1.57 ± 0.13 | 1.57 ± 0.13 | 1.49 ± 0.11 | 1.48 ± 0.10 |
| Step length (m) | 0.79 ± 0.08 | 0.76 ± 0.08 | 0.80 ± 0.08 | 0.77 ± 0.06 |
0.75 ± 0.05*
|
0.73 ± 0.05
|
| Cycle duration (s) |
1.26 ± 0.09*
|
1.24 ± 0.09
| 1.22 ± 0.04 | 1.22 ± 0.04 | 1.14 ± 0.06 | 1.14 ± 0.05 |
| Stance duration (% of the cycle) | 62.30 ± 3.06 | 62.23 ± 2.76 | 62.74 ± 1.03 | 61.84 ± 2.22 | 62.91 ± 1.15 | 63.16 ± 1.23 |
| swing duration (% of the cycle) | 37.70 ± 3.06 | 37.77 ± 2.76 | 37.26 ± 1.03 | 38.16 ± 2.22 | 37.09 ± 1.15 | 36.84 ± 1.23 |
| Propulsion double stance duration (% of the stance) |
12.72 ± 3.06*
|
10.90 ± 3.25
| 12.30 ± 1.71 | 12.27 ± 1.15 |
13.27 ± 1.13*
|
12.36 ± 1.38
|
| Early swing phase duration (% of the swing duration) |
57.59 ± 1.57*
|
59.95 ± 1.54
| 58.46 ± 2.01 | 58.73 ± 1.22 |
57.91 ± 0.97*
|
59.49 ± 1.18
|
| Late swing phase duration (% of the swing duration) |
42.21 ± 1.57*
|
40.05 ± 1.54
| 41.54 ± 2.01 | 41.27 ± 1.22 |
42.09 ± 0.97*
|
40.51 ± 1.18
|
| Cadence (cycles/min) | 96.2 ± 7.5 | 98.4 ± 3.5 | 105.38 ± 7.3 |
| Walking speed (m/s) | 1.25 ± 0.13 | 1.28 ± 0.11 | 1.30 ± 0.10 |
| *: p < 0.005 between the right and left sides |
