🎾 Influence Of Restricted Knee Motion Duri¶
Giới Thiệu¶
Influence Of Restricted Knee Motion Duri — tài liệu 8 trang từ thư viện sách tennis.
Chủ đề chính: Strength, Giao bóng, Thể lực
Tóm tắt nội dung (trích từ tài liệu gốc): Journal of Strength and Conditioning Research, 2007, 21(3), 950�957 2007 National Strength & Conditioning Association INFLUENCE OF RESTRICTED KNEE MOTION DURING THE FLAT FIRST SERVE IN TENNIS OLIVIER GIRARD,1 JEAN-PAUL MICALLEF,1,2 AND GRE� GOIRE P. MILLET3 1UPRES-EA 2991, Faculty of Sport Sciences, Montpellier, France; 2INSERM ADR 08, Montpellier, France; 3ASPIRE, Academy for Sports Excellence, Doha, Qatar. ABSTRACT. Girard, O., J.P. Micallef, and G.P. Millet. Influence curring earlier in the serve sequence, are probably not of restricted knee motion during the flat first serve in tennis. J.
Lưu ý: Nội dung dưới đây được trích xuất tự động từ PDF gốc tiếng Anh, giữ nguyên ngôn ngữ để bảo toàn độ chính xác kỹ thuật.
Nội Dung Gốc (Tiếng Anh)¶
Journal of Strength and Conditioning Research, 2007, 21(3), 950�957
2007 National Strength & Conditioning Association
INFLUENCE OF RESTRICTED KNEE MOTION DURING
THE FLAT FIRST SERVE IN TENNIS
OLIVIER GIRARD,1 JEAN-PAUL MICALLEF,1,2 AND GRE� GOIRE P. MILLET3
1UPRES-EA 2991, Faculty of Sport Sciences, Montpellier, France; 2INSERM ADR 08, Montpellier, France;
3ASPIRE, Academy for Sports Excellence, Doha, Qatar.
ABSTRACT. Girard, O., J.P. Micallef, and G.P. Millet. Influence curring earlier in the serve sequence, are probably not
of restricted knee motion during the flat first serve in tennis. J. important (21).
Strength Cond. Res. 21(3):950�957. 2007.--The aim of this study
was to examine the influence of restricted knee motion during However, it is well known that the development of
the serve in tennis players of different performance levels. Thir- both linear and angular momentum starts with the
ty subjects distributed in 3 groups (beginner, B; intermediate, I; ground reaction forces (GRF) generated by the players
elite, E) performed 15 flat first serves with normal (normal through their footwork (3). Recently, a technical model of
serve, SN) and restricted (restricted serve, SR) knee motion. In the serve has been proposed to complement the subjective
SR, the legs were kept outstretched by splints with a knee joint analysis of the shot by coaches (14). Several mechanisms
angle fixed at 10 (0 fully extended) to prevent any knee flexion/ underlying SE have been investigated by kinematics
extension. Vertical maximum ground reaction forces (Fzmax), ball methods (17, 18, 39, 40) and discussed elsewhere (12, 13,
impact location (Limpact), and ball speed (Sball) were measured 25). Proximal-to-distal sequencing has been clearly iden-
with force platform, video analysis, and radar, respectively. tified in the tennis serve (17, 37, 40). An essential aspect
Fzmax, Limpact, and Sball were higher (p 0.001) in SN than in SR. of this proximal-to-distal speed summation is that each
Sball was significantly (p 0.001) dependent on performance lev- segment movement is performed so that maximal speed
el, with higher values recorded in E than in B or I. From SR to is generated from the kinetic chain (23). The use of elastic
SN, increase in Limpact was greater (p 0.01) in E than in other energy and muscle preload both in the dominant arm/
groups and increases in Fzmax and Sball were correlated (r 0.69, shoulder (4, 11) and in the leg muscles (20) was shown to
p 0.01) in E only. Knee motion is a significant contributor to be paramount for SE. An increase in the trajectory of the
serving effectiveness whatever the performance level. Skilled racquet during the forward swing is known to allocate
players perform faster serves than their less skilled counter- time for the development of speed prior to impact (10, 11).
parts, and this is partly related to a more forceful lower limb Lesser is known about the lower extremity function, al-
drive. though GRF data (2, 16, 20, 30, 36, 41), kinematics of
lower extremity (27), and relationships between lower
KEY WORDS. ball speed, biomechanics, ground reaction forces, body strength and ball speed (32, 35) have been reported.
performance, racquet sport, stretch-shortening cycle
The contributions of the different segments to the
INTRODUCTION overarm throwing motion have been examined in various
ways (28). One specific method approaching the problem
M odern tactics dictate that tennis players hit of body segment contribution to sport performance is the
the ball both with maximal speed and with joint immobilization or restraint paradigm. The theory
an acceptable level of control. In a tennis justifying joints immobilization is based on a relatively
serve, an optimal racquet position, trajec- simple process: the subject executes a given skill and the
tory, height, and speed are necessary at the criterion of performance (i.e., height jumped or ball re-
time of impact with the ball and are required to coordi- lease) is recorded. The individual is then restrained in
nate lower and upper body segments (6, 16). Ball speed some way in an effort to eliminate or isolate the influence
and impact height have been presented as key variables of movement in several or at a particular joint. Under
underlying serving effectiveness (SE) in tennis (20, 25, these constraint conditions, the subject attempts to per-
35). form the original skill. Decrement in the value of the per-
formance criterion is considered as a rough index of the
By using 3-dimensional kinematic analysis to better role of the immobilized segments. Although this method
understand the source of the joint torques exerted on in- has been largely employed in the field of sport biome-
dividual body segments during the serve motion, previous chanics (i.e., baseball pitching [31] or overarm throwing
studies have shown that the greatest contribution to for- [38]), it seems, however, not relevant to quantify or esti-
ward velocity of the racquet head at impact was produced mate numerically the contribution of individual joints to
by the combined actions of internal rotation of the upper the final end point speed (28). Considering that the link-
arm and wrist flexion (18, 37, 40). However, these results ing of forward (linear) and rotational (angular) motions
are debatable because Gordon and Dapena (21) recently is critical for the generation of racquet speed during the
questioned methods used in these studies to measure the serve, one may argue that the removal of the use of lower
contribution of the body segments motions to racquet extremities that constitute the starting point of the ki-
speed at impact. They have shown that skin-attached netic chain will necessarily affect service effectiveness
markers could not be used to calculate accurately the up- through probable altered actions of the trunk or upper
per arm twist, due to skin movement. For these authors, arm. In the present study, the use of the joint immobili-
considering that the racquet speed at impact reflects only zation method could then be used as a paradigm for in-
the conditions at this final instant of the process, the mo-
tions of some body parts, such as lower extremities oc-
950
terrupting the normal mechanisms used by players to RESTRICTED KNEE MOTION DURING THE TENNIS SERVE 951
perform serves. Because most GRF studies have used
skilled players, the role of the lower extremities in cre- FIGURE 1. Example of a subject performing tennis flat first
ating GRF across skill levels in the flat tennis serve re- serves with normal (A) and restricted (B) knee motion. During
mains unclear. Although Girard et al. (20) recently tennis serves with restricted knee motion, the legs were kept
showed that elite players had higher vertical forces and outstretched by splints with a knee joint angle fixed at 10 to
a different neuromuscular temporal pattern in leg drive prevent any knee flexion/extension.
than their lower counterparts during serves, it is un-
known whether the restriction of knee motion would dif- (i.e., variable resistance, isokinetics, free weights, or rub-
ferently affect players of various performance levels. ber tubing) in order to build a good strength base. In I
and E, the conditioning training was periodized with cy-
The aim of this study was therefore to examine the cles of 4 weeks. This method has been shown to be effi-
influence of restricted knee motion during the serve in cient to enhance serve velocity in advanced players (24).
tennis players of different performance levels. It was hy- Informed consent was given by all the subjects, and a
pothesized that the restricted knee motion induces a local ethics committee for the protection of individuals
greater decrease in service effectiveness in skilled players gave their approval to the project before its initiation.
than in beginners.
Procedures
METHODS Vertical Jump Testing. During CMJ subjects started from
an erect standing position and made a downward move-
Experimental Approach to the Problem ment before starting to push off vertically in one contin-
uous movement (no pause). Subjects were asked to keep
Subjects were tested in early April during a precompeti- their hands on their hips. The force platform measure-
tive period (1 month prior to the major individual cham- ments were used to calculate peak lower extremity mus-
pionship in May). After a standardized warm-up lasting cular power (W�kg1) of the center of mass as the product
10 minutes (i.e., submaximal run, knee extensions), sub- of vertical component of GRF (Fz, N) and vertical velocity
jects who where distributed into 3 experimental groups (Vv, m�s1) (9). Based on the force-time principle (New-
(beginner, intermediate, elite) according to their tennis ton's second law of motion), Vv was calculated by inte-
performance level completed a set of 3 countermovement grating the force-time curve of Fz (vertical acceleration)
jumps (CMJs) from a force platform to evaluate peak low- from the beginning of the concentric (propulsive) phase.
er extremity muscular power. Then they performed Subjects were asked to jump as high as they could 3
serves for an additional 10 minutes with increasing times, and the best performance was reported. Coefficient
speed. After this, the subjects were asked to perform ran- of variation was calculated for peak lower extremity mus-
domly sets of 15 flat first serves with normal (normal cular power (CMJ) as the ratio of standard deviation by
serve [SN]) or restricted (restricted serve [SR]) knee mo- the mean and averaged 5.5 3.8%.
tion. To test the influence of restricted knee motion, the
legs were kept outstretched in SR by splints with a knee Serving Tests. The experiment consisted of flat first
joint angle fixed at 10 (0 fully extended) to prevent any serves performed both with SN and SR knee motion (Fig-
knee flexion/extension. All experiments were conducted ure 1). All serve trials were completed from the deuce or
on an outdoor Greenset tennis court. right service court with a 30-second rest between trials
until 15 acceptable serves were accomplished. An accept-
Subjects able serve required the ball to be hit with maximum
speed relative to the ability of the player (from the judge-
A group of thirty right-handed men (age: 21.3 3.8 years; ment of a professional coach) and land in the ad-side ser-
height: 179.7 7.0 cm; body mass: 74.1 9.8 kg) com- vice area. During SN, subjects were asked to perform flat
pleted the study. Based on the international tennis num- first serves as used in official competition. In SR, the legs
ber (ITN) equivalents established by the International were kept outstretched by splints (Macrimed, Medical
Tennis Federation, participants were distributed into 3 supplies, Paese, Italia) with a knee joint angle fixed at
experimental groups: (a) beginner (B; ITN 9, N 7): ten- 10 (0 fully extended) to prevent any knee flexion/exten-
nis players with irregular practice in competitions but sion. Thus, only trunk and upper-limb segments were
regularly physically active (sport recreation); (b) inter- used to perform serves.
mediate (I; ITN 5, N 10): good club players who have
played competitive tennis for many years; and (c) elite (E; Force Platform Recordings. A force platform (Captels
ITN 2, N 15): players of national level who regularly
practice with high intensity. No significant differences in
age, height, and weight were observed between the 3
groups. Mean training characteristics of subjects were
years of practice, 9.1 6.6 years (0.9 0.1, 6.9 3.8,
and 14.9 3.0 years in B, I, and E, respectively); tech-
nical/tactical training, 5.2 4.6 h�wk1 (1.2 2.2, 5.7
3.1, and 7.4 4.8 h�wk1 in B, I, and E, respectively);
physical training, 3.6 3.1 h�wk1 (3.0 4.3, 4.3 2.9,
and 3.5 2.7 h�wk1 in B, I, and E, respectively). The
conditioning program of the 3 groups focused mainly on
aerobic and anaerobic capabilities enhancement and in-
cluded general cardiovascular exercises (i.e., long dis-
tance running [30�40 minutes], short [10, 20, 40 m], and
long [100, 200, and 400 m] interval training). Addition-
ally, subjects performed a variety of plyometric (i.e., med-
icine balls, hopping) and resistance training modalities
952 GIRARD, MICALLEF, AND MILLET
TABLE 1. Performance parameters during the flat first serve performed with restricted (restricted serve [SR]) and normal (normal
serve [SN]) knee motion for the 3 performance level groups.*
SR SN
E (n 13)
B (n 7) I (n 10) E (n 13) B (n 7) I (n 10)
Sball (km/h) 89.1 4.9 126.6 6.8 144.6 14.1 107.2 6.1 148.8 16.3 169.4 11.3
Limpact (cm) 78.8 3.5 82.1 5.7 81.4 3.8 81.7 3.7 86.9 5.3 87.9 4.0�
Himpact (%) 144.4 5.7
146.3 5.4 145.3 3.9 145.9 7.2 148.4 5.4 147.5 3.8
* Values are mean SD. B beginner group; I intermediate group; E elite group; Sball postimpact ball speed; Himpact
impact height; Limpact ball impact location. Himpact was normalized to the standing height of the subject. Limpact was determined as
the difference between height of the racquet at impact and subject's standing height.
p 0.001; significantly different from beginner group.
p 0.05; significantly different from intermediate group.
� p 0.05; significantly different from beginner group in SN.
SA, Saint Mathieu de Treviers, France) was used to mon- able was tested by the Kolmogorov�Smirnov test. When
itor Fx, Fy, and Fz orthogonal components of the GRF. the normality condition was accepted, a one-way analysis
The force platform used was 50 50 cm and surrounded of variance (ANOVA) was used to test the effect of per-
by a raised wooden surface area. For each trial, 3 com- formance level on peak lower extremity muscular power
ponents of forces were sampled at 500 Hz simultaneously and percentage of variation from SR to SN in kinetic var-
by using an analog-to-digital convertor (MP 100A-CE; iables and performance parameters. The effect of perfor-
Biopac, Santa Barbara, CA). In the present study, all sub- mance level on type of serve was verified by a 2-way AN-
jects replicated their own specific stance position. OVA with repeated measures (3 groups [B, I, E] 2 con-
ditions [SR vs. SN]) on kinetic variables (Fzmax, Fx, Fy,
Video and Radar Recordings. Postimpact ball speed and Fz) and performance parameters (Sball, Himpact, and
(Sball, km�h1) was measured for each trial by the means Limpact). This analysis showed the global effect of the type
of a radar (Stalker ATS, Plano, TX) fixed on a 2.5-meter of serve, the global effect of performance level and, the
height tripod, 2 meters behind the players in the direction effect of interactions between type of serve and perfor-
of the serve. The racquet-ball impact height (Himpact, cm) mance level conditions. When significant main effects
was measured by a video camera (JVC, Ottawa, Ontario, were observed with the 2-way ANOVA, Bonferroni post-
Canada), operating at 50 Hz, located laterally on a rigid hoc analyses were used to identify differences among
tripod behind a 3-meter guide mark made of 2 metallic means. Pearson correlation coefficient was used to assess
poles connected with colored yarns vertically every 10 cm. the relationships between changes in selected GRF data
From the tapes, a researcher visually estimated Himpact and performance parameters between the 2 serve condi-
with a precision of 1 cm. The reliability of this way of tions in each performance level group. The level of sig-
measuring Himpact was assessed by the same experimenter nificance was established at p 0.05 for all procedures
digitizing 10 successful randomized trials on 3 different (SigmaStat 2.3, Jandel Corporation, San Rafael, CA).
days and was satisfying (coefficient of variation 0.7
0.3%). RESULTS
Analysis and Treatment of Data Vertical Jump Ability
Both for SN and SR, the 10 trials with the highest Sball No statistically significant difference (F1 1.1; p 0.36)
were used for subsequent analysis. Vertical maximum in peak lower extremity muscular power (59.5 10.1,
GRF (Fzmax) was normalized to body weight and Himpact to 61.2 5.9, and 63.8 5.5 W�kg1 in B, I, and E, respec-
the standing height of the subject (Himpact, %). Ball impact tively) was observed between groups.
location (Limpact, cm) was determined as the difference be-
tween height of the racquet at impact and subject's stand- Performance Parameters
ing height. For each trial, minimal and maximal values
for the 3 GRF components during the serve were identi- Performance parameters (Sball, Himpact, and L ) impact data in
fied using Acqknowledge software (3.7.2, Biopac, Santa the 2 serve conditions for each performance level group
Barbara, CA). For analysis purposes, GRF were ex- are displayed in Table 1. There was a significant effect of
pressed as difference () between maximal and minimal type of serve on Vball (F1,6 101.9; p 0.001). No signif-
values. Serving effectiveness was evaluated through per- icant interaction effect between type of serve and perfor-
formance parameters including Sball, Himpact, and L . impact mance level (F2,12 0.8; p 0.468) was found in Sball, but
Changes between the 2 serve conditions were also calcu- there is a significant performance level effect (F2,12 69.8;
lated for performance parameters (Sball, Himpact, and L ) impact p 0.001): posthoc analysis showed a higher Vball in E
and kinetic variables (Fzmax, Fx, Fy, and Fz). Inter- than in B (p 0.001) or I (p 0.05) but also in I than in
trial variability (coefficient of variation) of dependent var- B (p 0.001). There was a significant increase (F1,6
iables, i.e. kinetic variables (5.1 vs. 4.6, 12.0 vs. 10.1, 11.7 47.8; p 0.001) in Himpact from SR to SN, but there was no
vs. 12.4, and 9.6 vs. 11.1% for Fzmax, Fx, Fy, and Fz) significant difference in Himpact between groups (F2,12
and performance parameters (7.0 vs. 7.6, 2.2 vs. 2.4 and 0.4; p 0.68), nor any significant interaction (F2,12 1.5;
0.6 vs. 0.7% for Sball, Himpact, and Limpact), was not different p 0.26). There was a significant increase (F1,6 249.1;
in SN and SR, respectively. p 0.001) in Limpact from SR to SN, and this increase tended
(F2,12 2.5; p 0.06) to be dependent on performance
Statistical Analyses level. Limpact displayed a significant interaction effect (F2,12
8.4; p 0.01) between type of serve and performance
Means and standard deviations were calculated for all level. This interaction effect was due to a significantly
variables. The normality of the distribution of each vari-
RESTRICTED KNEE MOTION DURING THE TENNIS SERVE 953
TABLE 2. Percentage of variation in kinetic and performance
parameters variables from the restricted (restricted serve [SR])
to the normal serve (normal serve [SN]) conditions for the 3 per-
formance level groups.*
B (n 7) I (n 10) E (n 13)
Sball 16.9 6.9 15.1 8.8 14.8 4.2
Limpact 3.5 2.6 5.6 2.0 7.4 1.7
Himpact 1.0 1.3 1.4 0.7 1.5 1.0
Fx
Fy 19.5 20.6 22.9 37.0 24.8 20.0
Fz 50.2 10.9 48.9 33.3 63.4 17.3
48.2 27.4 55.1 18.8 59.3 10.6
Fzmax 25.5 14.6 28.0 11.9 34.0 11.2
* Values are mean SD. B beginner group; I interme-
diate group; E elite group; Sball postimpact ball speed; Himpact
impact height; Limpact ball impact location. Himpact was nor-
malized to the standing height of the subject. Limpact was deter-
mined as the difference between height of the racquet at impact
and subject's standing height. All components of forces are ex-
pressed as difference between maximal and minimal values ()
during the serve. GRF ground reaction forces in the medio-
lateral (Fx), anteroposterior (Fy), and vertical (Fz) components.
p 0.001; significantly different from beginner group.
p 0.05; significantly different from intermediate group.
higher (p 0.05) Limpact in E than in B in SN only. Table FIGURE 2. Typical curves of ground reaction forces signals in
2 summarizes the percentage variations in performance the mediolateral (Fx), anteroposterior (Fy), and vertical (Fz)
parameters and kinetic variables for the 3 groups. components during a flat first serve performed with restricted
(restricted serve, shaded line) and normal (normal serve, solid
Kinetic Variables line) knee motion in beginner and elite players. Broken lines
correspond to the time of racquet ball impact.
Typical curves of GRF signals in the Fx, Fy, and Fz com-
ponents of successful performance of a beginner and elite no significant correlations existed between changes in ki-
player during the 2 serve conditions are shown in Figure netic variables and in performance parameters between
2. Changes in kinetic variables, that is Fx (F1,6 10.7; the 2 serve conditions in B and I.
p 0.05), Fy (F1,6 44.2; p 0.001), and Fz (F1,6
62.6; p 0.001), were higher in SN than in SR (Table 3). DISCUSSION
There was a significant effect of performance level on Fx
(F2,12 5.2; p 0.05): posthoc analysis showed a smaller This study examined the influence of restricted knee mo-
Fx in B than in E (p 0.05) but only a tendency when tion during the serve in tennis players of different per-
compared with I (p 0.06). The Fy (F2,12 1.6; p formance levels. First of all, it is of the highest interest
0.24) and Fz (F2,12 2.1; p 0.17) were not significantly for the purpose of the present study to note that peak
influenced by the performance level. The interaction of lower extremity muscular power was similar in the 3
type of serve and performance level conditions was sig- groups. This suggests that the differences in kinetic var-
nificant (F2,12 3.8; p 0.05) in Fy. This interaction iables observed during the serve were primarily due to
effect was due to a significantly higher (p 0.05) Fy in technical/coordination aspects that characterize the dif-
E than in B in SN only. Fzmax was significantly higher (F1,6 ferent levels of expertise of the subjects.
54.3; p 0.001) in SN than in SR, independently of
performance level (Figure 3). There was no significant in- To evaluate the influence of restricted knee motion
teraction between these 2 factors. during the tennis serve, the joint immobilization ap-
proach has been employed (28). Subjects attempted to
Relationships Between Kinetic and Performance perform flat first serves with normal and restricted knee
motion, and decrement in the value of performance pa-
Parameters rameters and kinetic variables from the original (SN) to
the restraint condition (SR) was recorded. This method
The relationship between increase in Fzmax and increase was previously used in overarm throwing (28) and/or
in Sball from SR to SN in E is presented in Figure 4. Finally,
TABLE 3. Peak ground reaction forces (BW) during the flat first serve performed with restricted (restricted serve [SR]) and normal
(normal serve [SN]) knee motion for the three performance level groups.*
SR SN
GRF (BW) B (n 7) I (n 10) E (n 13) B (n 7) I (n 10) E (n 13)
Fx 0.19 0.04 0.23 0.06 0.28 0.08 0.25 0.04 0.36 0.15 0.39 0.10
Fy 0.11 0.02 0.10 0.04 0.09 0.03 0.22 0.05 0.24 0.07 0.28 0.00
Fz 0.47 0.17 0.54 0.16 0.58 0.11 1.00 0.34 1.27 0.30 1.46 0.30
* Values are mean SD. All components of forces are expressed as difference between maximal and minimal values () during
the serve. B beginner group; I intermediate group; E elite group; GRF ground reaction forces in the mediolateral (Fx),
anteroposterior (Fy), and vertical (Fz) components.
p 0.05; significantly different from beginner group.
p 0.05; significantly different from beginner group in SN.
954 GIRARD, MICALLEF, AND MILLET
FIGURE 3. Vertical maximum component of ground reaction FIGURE 4. Relationship (r 0.69, p 0.01) between changes
forces (Fzmax) during the flat first serve performed with re- in vertical maximum component of ground reaction forces
stricted (restricted serve, white bar) and normal (normal serve,
shaded bar) knee motion for the 3 performance level groups. (Fzmax changes, N) and in postimpact ball speed (Sball changes,
Values are mean SD. %) during the flat first serve performed with restricted (re-
stricted serve, SR) and normal (normal serve, SN) knee motion
in elite players (n 13). Fzmax is expressed without weight of
the subject on the force plate.
jumping (26) to evaluate the influence of a particular joint the different studies. The development of linear momen-
or joints sequence to performance. Although physical im- tum in the vertical and horizontal directions depends on
mobilization of joints may provide some general insights the type of stance adopted by the player (2, 19). However,
into segmental contributions to performance, it is impor- no difference in ball speed has been reported between the
tant to note that the restriction of one or more joints can 2 stances (17) . In the present study, all beginner and
also deteriorate the coordinated action of the other body intermediate and most of the elite players used a stance
segments (28). This may be particularly true for the ten- close to ``foot-back,'' which is known to produce smaller
nis serve in which a number of body segments are coor- vertical GRF but greater peak forward propulsive force,
dinated in a sequence referred to as the kinetic chain (17, with the back leg favoring rapid displacement to the net
23). Therefore, it seems clear that this approach, although (2, 19).
used in the past, is inadequate to evaluate the contribu-
tion of a particular joint to the final outcome of a tennis As expected, from the restricted to the traditional
serve. A valuable method of approaching the problem of serve condition, significant increases in performance pa-
the contributions of the different body segments and rameters and kinetic variables were recorded (Table 2),
joints to the final velocity of the racquet head would be irrespective of the subject's expertise level. This indicates
to focus on the resultant muscle torque patterns, which that, as required by the experimental design, the influ-
involves detailed computations of the internal forces re- ence of knee motion was effectively minimized by the use
sponsible for ball speed (28). The influence of restricted of splints. This result confirms that knee motion is a sig-
knee motion, in the present study, was then used as a nificant contributor to serving effectiveness, whatever the
paradigm for interrupting the normal mechanisms used performance level.
by players to perform the serve, rather than to focus on
the contribution of lower limbs. Elliott (10) has shown that a rhythmical action is the
key to an effective serve. Several body segments have to
In the present study, the mean Sball values measured be coordinated for producing a high-speed serve with an
in the traditional serve condition in the 3 performance acceptable level of control, in a proximal-to-distal time
level groups (107, 148, and 169 km�h1 in B, I, and E, sequence (13, 16, 17). In this sequence, the acceleration
respectively) are in line with previous results in unskilled of the racquet through the ball is built up through the
(87�108 km�h1 [1, 29]) and national level (145�180 summation of the individual segments speeds, transfer-
km�h1 [5, 17, 19]) players. The range of Himpact values ring linear and angular momentum generated from the
(144�149% of standing height) compares with previous GRF to the racquet (10, 23). In the present study, the
findings (141�152%) in players of similar standard (5, 17), larger GRF values (i.e. Fx, Fy, and Fz) in the normal
the greatest values being recorded in the more skilled condition are the result of a forceful leg drive. This sug-
players as a result of a forceful leg drive (20). gests that lower extremities require some degree of knee
flexion during the backswing to generate large amounts
Although the reduced dimensions of the force plate of linear and angular momentum during the knee exten-
might have limited force production, vertical GRF mea- sion, transferring the GRF to the trunk (3, 18). This con-
sured during the traditional serve condition in the pres- sideration is also supported by previous studies hypoth-
ent study (1.68�2.12 BW) are in accordance with those esizing that the largest portions ( 50%) of kinetic energy
previously described (2, 19, 30, 41). Negligible mediolat- or force generated during the serve in world-class players
eral, low anteroposterior, and peak vertical force of one- are developed in the legs and trunk (22, 34). In the pres-
third body weight were recorded in a study by Van Ghe- ent study, only 1 force platform was used to determine
luwe and Hebbelinck (41). These peak vertical forces were GRF, which did not allow the accurate appreciation of the
lower than those reported elsewhere, in which players role of individual leg segments. However, it has been pre-
were able to generate considerable vertical forces (twice viously reported that the back leg provides most of the
their body weight) with both foot-up (the rear foot is upward and forward push, whereas the front leg provides
moved forward next to the front one during the ``push-off'' the stable post for the rotational momentum (3). It should
phase) and foot-back (the feet stay at the same relative be therefore assumed that the combined action of the low-
level) stances (2, 19). The discrepancies are largely the er extremities enhances the ability to generate trunk and
result of the performance level of the players tested in upper-arm rotations later in the action, which in turn
may contribute to enhanced SE (Sball, Himpact, and L ) impact RESTRICTED KNEE MOTION DURING THE TENNIS SERVE 955
from SR to SN.
Although the present results confirm that the knee
Given that the changes in the total angular momen- flexion before extension is a prerequisite for an efficient
tum of the body around the center of mass are primarily execution of the serve, it is important to note that the
the result of the magnitude and direction of the reaction influence of restricted knee motion on performance pa-
forces from the court (3, 17), one may argue that limiting rameters and kinetic variables was in part dependent on
leg movement has some effects not only on force produc- the performance level of the players. Although it is well
tion capabilities from GRF but also on ball toss, trunk, documented that (a) weight distribution (center of pres-
and hitting arm motions. In the serve, the trunk move- sure of GRFs) in the starting position is an individual
ment is a fundamental link in the kinetic chain that characteristic (36) and (b) different stances produce dif-
transfers energy from the extension of the lower limbs to ferent patterns of GRF curves (2, 19), one may argue that
the arm during the forward swing (3). As a consequence, changes in magnitude of kinetic variables from SR to SN
one may speculate that the expected decrease in trunk between players of various abilities primarily result in a
angular momentum in the restricted knee flexion condi- more or less efficient leg drive, leading to different levels
tion may have in turn decreased the ability of the shoul- of SE.
der to rotate rapidly internally, an action known to be a
key factor in SE (18, 37). An effective serve is characterized by vertical forces
that induce the body to be driven off the ground for im-
Efficient kinetic chain force production for the serve pact (2, 19, 41). This point is clearly supported by the
requires commonalities in the sequence, including the use vertical force curve in E (Figure 2). Again, Payne (30)
of elastic energy and muscle preload (11, 13, 20). As reported that the angular momentum developed during
shown recently (20), lower-limb activity during the serve the serve is the result of the vertical forces generating an
is characterized by a stretch-shortening cycle action, that off-center impulse behind the center of mass of the player,
is, an eccentric contraction (knee flexion) followed by a which helps to rotate the trunk forward (flexion, shoul-
concentric one (knee extension). Without knee bend dur- der-over-shoulder, and rotation) in preparation for im-
ing SR, quadriceps muscles were not stretched and there- pact. This is confirmed by the fact that skilled players
fore elastic energy was not stored in elastic components. increased to a greater extent ball impact location from the
As a consequence, it should be assumed that speed of leg restricted to the normal knee flexion condition than their
extension was certainly near 0 at the beginning of the lower counterparts.
kinetic chain force production and that the role of the
trunk was then limited. Surprisingly, the decrease in Sball from SN to SR was to
the same extent in all groups (Table 2), although skilled
It is well known that the legs require some degree of players displayed higher Sball values than their less
knee flexion during the preparation phase not only to de- skilled counterparts (Table 1). A possible explanation
crease the loading in upper limbs segments (10) but also about the difference in the restricted serve condition
to assist players in driving the racquet down, behind, and could be the development of higher muscular forces in the
away from the back (putting shoulder muscles on stretch dominant arm in skilled players, because measures of up-
[4]) and increasing the trajectory of the racquet prior to per extremity's flexibility and muscular strength were
impact (17). It should therefore be assumed that the ac- found to be linked to postimpact ball speed during serves
tion of the lower extremities enhances the trunk and up- performed by elite performers (8).
per-arm rotations and facilitates the downward racquet
motion. Although the type of backswing was shown to It is interesting to note that from the restricted to the
have minimal influence on service performance or on normal knee flexion condition, increase in Fzmax account-
loading of the shoulder and elbow joints (10), it is of in- ed for 48% of the variance of increase in Sball in the high-
terest to note that when leg participation was allowed, est skilled players only. This finding emphasizes that a
greater values in kinetic variables (i.e., Fy, Fz, Fzmax) forceful lower-limb drive is used to improve SE in skilled
were recorded in the 3 groups of various expertise levels. players (20, 27). Bartlett et al. (5) added further support
For the flat and slice serves, Bahamonde (4) has shown to the relationship between leg drive and SE in skilled
that leg drive and trunk rotations produce a forced exter- players by demonstrating that the difference in ball speed
nal (away from the direction of the serve) rotation of the at impact between British national and county players
upper arm, resulting in the stretch of the internal rota- was to a great extent the result of the timing of the move-
tors muscles. On movement reversal, these stretched ment of the back-foot forward during the preparation
muscles are creating a higher speed of rotation of the hit- phase. Another interesting result is the greater increase
ting arm and consequently a higher postimpact ball speed in Fy from SR to SN in skilled players than in their less
(13). This phenomenon was certainly present during the skilled counterparts. So the highest skilled players, pro-
traditional serve condition to accelerate the upper-arm ducing a greater shift of the center of mass forward (36),
segments and as a consequence the ball. Research has are able to generate greater somersault (forward) angular
shown that 10 to 20% additional speed is achieved after momentum, which is a well-known factor contributing to
a stretch-shortening cycle (11, 15). However, the ability the development of racquet and ball linear velocities at
to store elastic energy is affected by numerous factors, impact (3). Inversely, incorrect timing in the leg muscles
such as the level of preactivation, the muscle stiffness and activation and smaller magnitudes of GRF have been
compliance, the velocity and magnitude of stretch, and identified in unskilled players. It was suggested that ki-
the coupling time between eccentric and concentric phas- netic chain breakage occurs in beginners who use com-
es (42). In this context, one may argue that the storage pensative mechanisms to stabilize body segments in an
of elastic energy and muscle preload were inevitably re- attempt to hit the ball (20, 23). In the present study, these
duced during SR as a result of the limited motions of the findings are supported by the lack of correlation between
lower extremities and trunk. Restricting leg drive may kinetic variables and performance parameters in un-
have also altered the ideal positioning of trunk and up- skilled individuals.
per-arm segments and therefore the rotation amplitude.
CONCLUSIONS
This study examined the influence of restricted knee mo-
tion during the serve in tennis players of different per-
956 GIRARD, MICALLEF, AND MILLET 10. ELLIOTT, B., G. FLEISIG, R. NICHOLLS, AND R. ESCAMILIA. Technique ef-
fects on upper limb loading in the tennis serve. J. Sci. Med. Sport. 6:76�
formance levels. The present results confirm that the 87. 2003.
knee flexion before extension is a prerequisite for an ef-
ficient execution of the serve, whatever the performance 11. ELLIOTT, B.C. Biomechanics of tennis. In: Tennis. A. Renstro�m, ed. Ox-
level. However, several differences in performance pa- ford: Blackwell, 2002. pp. 1�28.
rameters variables were identified between players of
various performance levels. From the restricted to the 12. ELLIOTT, B.C. Biomechanics of the serve in tennis: A biomedical per-
normal serve condition, skilled players displayed larger spective. Sports Med. 6:285�294. 1988.
increases in anteroposterior GRF and in ball impact lo-
cations than in other groups. Skilled players also per- 13. ELLIOTT, B.C. The development of racquet speed. In: Biomechanics of
formed faster serves than their less skilled counterparts, Advanced Tennis. B.C. Elliott, M. Reid, and M. Crespo, eds. London: ITF
which are related to a forceful lower limb drive. Taken Ltd, 2003. pp. 33�47.
together, these results reinforce the role of lower extrem-
ities to produce an effective high-speed serve through pos- 14. ELLIOTT, B.C., AND J. ALDERSON. Biomechanical performance models:
sible mechanisms, including the use of coordinated move- The basis for stroke analysis. In: Biomechanics of Advanced Tennis. M.
ment and/or the use of elastic energy and muscle preload. Crespo, ed. London: ITF Ltd, 2004. pp. 157�175.
However, further investigation is needed to better under-
stand the relationships between lower extremity function 15. ELLIOTT, B.C., K.G. BAXTER, AND T.F. BESIER. Internal rotation of the
and racquet kinematics among different players. upper-arm segment during a stretch-shorten cycle movement. J. Appl.
Biomech. 15:381�395. 1999.
PRACTICAL APPLICATIONS
16. ELLIOTT, B.C., AND R. KILDERRY. The Art and Science of Tennis. Phila-
The daily and experienced observations of player move- delphia: Saunders College Publishing, 1983. p. 41.
ments by the coaches can be completed by numerical in-
formation to establish the optimal range of flexion exten- 17. ELLIOTT, B.C., T. MARSH, AND B. BLANKSBY. A three-dimensional cine-
sion in lower extremities for generating high-speed matographic analysis of the tennis serve. Int. J. Sports Biomech. 2:260�
serves. The large involvement of the lower extremities in 271. 1986.
the tennis serve reinforces the importance of their
strength and flexibility training to improve explosive 18. ELLIOTT, B.C., R.N. MARSHALL, AND G. NOFFAL. Contributions of upper
power, speed, and endurance (33). Explosive strength or limb segment rotations during the power serve in tennis. J. Appl. Bio-
plyometric training is known to be useful for improving mech. 11:433�442. 1995.
lower body strength. Plyometric training, such as bound-
ing, jumping, and hopping, enhances the muscle's ability 19. ELLIOTT, B.C., AND G.A. WOOD. The biomechanics of the foot-up and foot-
to generate power by optimizing stretch-shortening cycle. back tennis service techniques. Aust. J. Sport Sci. 3:3�6. 1983.
Plyometric training in addition to flexibility, cardiorespi-
ratory endurance, general strength, and muscular endur- 20. GIRARD, O., J.P. MICALLEF, AND G.P. MILLET. Lower-limb activity during
ance was shown to be efficient for improving general fit- the power serve in tennis: Effects of performance level. Med. Sci. Sports
ness and preventing injuries (7). Appropriate leg exercis- Exerc. 37:1021�1029. 2005.
es in tennis also include isokinetics, weight machines,
and rubber tubing systems. Considering that improve- 21. GORDON, B.J., AND J. DAPENA. Contributions of joint rotations to racquet
ment in speed can also be caused by a better reaction speed in the tennis serve. J. Sports Sci. 24:31�49. 2006.
time, coaches should prescribe exercises that invoke spe-
cific patterns (direction, amplitude, speed) of neuromus- 22. KIBLER, W.B. Biomechanical analysis of the shoulder during tennis ac-
cular recruitment and activation. Also, plyometric medi- tivities. Clin. Sports Med. 14:79�85. 1995.
cine ball throws that activate trunk and upper arm mus-
cles are efficient. 23. KIBLER, W.B., AND D. VAN DER MEER. Mastering the kinetic chain. In:
World-Class Tennis Technique. P. Roetert and J. Groppel, eds. Cham-
REFERENCES paign, IL: Human Kinetics, 2001. pp. 99�113.
1. ANDERSON, M.B. Comparison of muscle patterning in the overarm throw 24. KRAEMER, W.J., K. HAKKINEN, N.T. TRIPLETT-MCBRIDE, A.C. FRY, L.P.
and tennis serve. Res. Q. 50:541�553. 1979. KOZIRIS, N.A. RATAMESS, J.E. BAUER, J.S. VOLEK, T. MCCONNELL, R.U.
NEWTON, S.E. GORDON, D. CUMMINGS, J. HAUTH, F. PULLO, J.M. LYNCH,
2. BAHAMONDE, R., AND D. KNUDSON. Ground reaction forces of two types S.J. FLECK, S.A. MAZZETTI, AND H.G. KNUTTGEN. Physiological changes
of stances and tennis serves. Med. Sci. Sports Exerc. 33:S102, 2001. with periodized resistance training in women tennis players. Med. Sci.
Sports Exerc. 35:157�168, 2003.
3. BAHAMONDE, R.E. Changes in angular momentum during the tennis
serve. J. Sports Sci. 18:579�592. 2000. 25. LEES, A. Science and the major racket sports: A review. J. Sports Sci.
21:707�732. 2003.
4. BAHAMONDE, R.E. Joint power production during flat and slice tennis
serves. In: 15th International Symposium on Biomechanics in Sports. 26. LEES, A., J. VANRENTERGHEM, AND D. DE CLERCQ. Understanding how
Denton, TX, 1997. pp. 489�494. an arm swing enhances performance in the vertical jump. J. Biomech.
37:1929�1940. 2004.
5. BARTLETT, R., J. PILLER, AND S. MILLER. A three-dimensional analysis
of the tennis serves of National (British) and county standard players. 27. LO, K.C., L.H. WANG, C.C. WU, AND F.C. SU. Kinematics of lower ex-
In: Science and Racket Sports. T. Reilly, M. Hughes, and A. Lees, eds. tremity in tennis flat and spine serve. J. Med. Biol. Eng. 24:209�212.
London: E & FN Spon, 1995. pp. 98�102. 2004.
6. BRODY, H. Tennis Science for Tennis Players. Philadelphia: University of 28. MILLER, D.I. Body segment contributions to sport skill performance: Two
Pennsylvania Press, 1987. contrasting approaches. Res. Q. Exerc. Sport. 51:219�233. 1980.
7. CHANDLER, T.J. Exercise training for tennis. Clin. Sports Med. 14:33�46. 29. MIYASHITA, M., T. TSUNODA, S. SAKURAI, H. NISHIZONA, AND T. MIZUNO.
1995. Muscular activities in the tennis serve and overhand throwing. Scand.
J. Sports Sci. 2:52�58. 1980.
8. COHEN, D.B., M.A. MONT, K.R. CAMPBELL, B.N. VOGELSTEIN, AND J.W.
LOEWY. Upper extremity physical factors affecting tennis serve velocity. 30. PAYNE, A.H. Comparison of the ground reaction forces in golf drive and
Am. J. Sports Med. 22:746�750. 1994. tennis service. Aggressologie 19:53�54. 1978.
9. DRISS, T., H. VANDEWALLE, J. QUIEVRE, C. MILLER, AND H. MONOD. Ef- 31. PETERSON, M.W. The segmental components in skilled baseball throw-
fects of external loading on power output in a squat jump on a force ing. Master's thesis, University of Illinois, Urbana-Champaign, 1973.
platform: A comparison between strength and power athletes and sed-
entary individuals. J. Sports Sci. 19:99�105. 2001. 32. PUGH, S.F., J.E. KOVALESKI, R.J. HEITMAN, AND W.F. GILLEY. Upper and
lower body strength in relation to ball speed during a serve by male
collegiate tennis players. Percept. Mot. Skills 3:867�872. 2003.
33. REID, M., A. QUINN, AND M. CRESPO. Strength and Conditioning for Ten-
nis. London: ITF Ltd, 2003.
34. SCHONBORN, R. Advanced Techniques for Competitive Tennis. Aachen,
Germany: Meyer and Meyer, 1999. pp. 1�280.
35. SIGNORILE, J.F., D.J. SANDLER, W.N. SMITH, M. STOUTENBERG, AND A.C.
PERRY. Correlation analyses and regression modeling between isokinetic
testing and on-court performance in competitive adolescent tennis play-
ers. J. Strength Cond. Res. 19:519�526. 2005.
36. SMITH, S.L. Comparison of selected kinematic and kinetic parameters
associated with the flat and slice serves of male intercollegiate tennis
players. Doctoral dissertation, Indiana University, 1979.
37. SPRIGINGS, E., R. MARSHALL, B. ELLIOTT, AND L. JENNINGS. A three-
dimensional kinematic method for determining the effectiveness of arm
segment rotations in producing racquet-head speed. J. Biomech. 27:245�
254. 1994.
38. TOYOSHIMA, S., T. HOSHIKAWA, M. MIYASHITA, AND T. OGURI. Contribu-
tion of the body parts to throwing performance. In: Biomechanics IV. R.C.
Nelson and C.A. Morehouse, eds. Baltimore: University Park Press,
1974.
39. VAN GHELUWE, B., I. DE RUYSSCHER, AND J. CRAENHALS. Pronation and RESTRICTED KNEE MOTION DURING THE TENNIS SERVE 957
endorotation of the racket arm in a tennis serve. In: Biomechanics X-B.
B. Jonsson, ed. Champaign, IL: Human Kinetics, 1987. pp. 667�672. 42. WILSON, G.J., B.C. ELLIOTT, AND G.A. WOOD. The effect on performance
of imposing a delay during a stretch-shorten cycle movement. Med. Sci.
40. VAN GHELUWE, B., AND M. HEBBELINCK. The kinematics of the service Sports Exerc. 23:364�370. 1991.
movement in tennis: A three dimensional cinematographical approach.
In: Biomechanics IX-B. D.A. Winter, R.W. Norman, R.P. Wells, and A.E. Acknowledgments
Patla, eds. Champaign, IL: Human Kinetics, 1985. pp. 521�526. Thanks to Sebastien Racinais for help in statistical analysis.
41. VAN GHELUWE, B., AND M. HEBBELINCK. Muscle action and ground re- Address correspondence to Olivier Girard, oliv.girard@
action forces in tennis. Int. J. Sport Biomech. 2:88�99. 1986. gmail.com.