Individual profiles of spatio-temporal coordination in high intensity swimming

Human Movement Science 31 (2012) 1200–1212

Contents lists available at SciVerse ScienceDirect

Human Movement Science

journal homepage: www.elsevier .com/locate/humov

Individual profiles of spatio-temporal coordination in highintensity swimming

Pedro Figueiredo a,b, Ludovic Seifert c,⇑, João Paulo Vilas-Boas a,b,Ricardo J. Fernandes a,b

a CIFI2D, Faculty of Sport, University of Porto, Porto, Portugalb LABIOMEP, University of Porto, Porto, Portugalc CETAPS, Faculty of Sports Sciences, University of Rouen, France

a r t i c l e i n f o

Article history:Available online 24 August 2012

PsycINFO classification:23303720

Keywords:Motor controlCluster analysisFatigueSwimming

0167-9457/$ - see front matter � 2012 Elsevier B.Vhttp://dx.doi.org/10.1016/j.humov.2012.01.006

⇑ Corresponding author. Tel.: +33 232107784; faE-mail address: [email protected] (L

a b s t r a c t

The aim of the present study was to examine the inter-subject var-iability in inter-arm coordination during front crawl swimming athigh intensity. Ten male competitive swimmers swam 200 m frontcrawl at race pace. Two above water and four underwater camerasvideotaped the test and APAS was used to assess the 3D anatomicalpoints position and to calculate, afterwards, the angular position,velocity and the continuous relative phase, which was used to ana-lyze upper limbs coupling during two arm stroke cycles for each50 m lap of the 200 m front crawl event. The kinematics of thetwo arm stroke cycles was filtered and normalized to the totalcycle duration. A significant decrease, caused by fatigue, wasobserved in the biomechanical and coordinative (temporal) param-eters. The topography of the mean continuous relative phase curveshowed a higher inter-subject variability between two groups(63.1� vs. 76.0�; p < .05), suggesting that several profiles of inter-limb coordination exist. Two clusters arose from the coordinationprofiling, which mainly stemmed from organismic constraints(i.e., the swimmer’s technique).

� 2012 Elsevier B.V. All rights reserved.

1. Introduction

The dynamical systems approach to the study of coordinated movement focuses on the stability ofspatio-temporal patterns of body parts that are coordinated with each other and the environment

. All rights reserved.

x: +33 232107793.. Seifert).

http://dx.doi.org/10.1016/j.humov.2012.01.006

mailto:[email protected]

http://dx.doi.org/10.1016/j.humov.2012.01.006

http://www.sciencedirect.com/science/journal/01679457

http://www.elsevier.com/locate/humov

P. Figueiredo et al. / Human Movement Science 31 (2012) 1200–1212 1201

(e.g., Beek, Peper, & Stegeman, 1995; Kelso, 1995; Schöner & Kelso, 1988). A key feature of the approachis that the variability in motor behavior provides a window into the stable and functional states of coor-dination (Davids, Button, & Bennett, 2008; Glazier & Davids, 2009). The approach emphasizes theimportance of inter- and intra-limb coordination, which are viewed as crucial aspects of human (andanimal) motor behavior (Donker, Beek, Wagenaar, & Mulder, 2001; Lee, Swinnen, & Verschueren,1995; Seifert, Leblanc, Chollet, & Delignières, 2010; van Emmerik & Wagenaar, 1996; Swinnen et al.,1997). In human gait research, inter-limb coordination has been mostly investigated during walkingor running (e.g., Diedrich & Warren, 1995; Donker et al., 2001; van Emmerik & Wagenaar, 1996;Wagenaar & Beek, 1992; Wagenaar & van Emmerik, 2000), that is, in situations where the arms donot contribute to propulsion, but to trunk stabilization (Wannier, Bastiaanse, Colombo, & Dietz,2001). However, in aquatic locomotion (e.g., front crawl), 85% of the propulsion is due to arm actions(Bucher, 1975; Deschodt, Arsac, & Rouard, 1999; Hollander, de Groot, van Ingen Schenau, Kahman, &Toussaint, 1988), which elevates the importance of inter-arm coordination in this particular form oflocomotion; in this context, inter-arm coordination is linked to the swimming velocity, and it changesdue to the breathing actions, floating (external forces) and balance of the body segments (Chollet,Chalies, & Chatard, 2000; Seifert, Toussaint, Alberty, Schnitzler, & Chollet, 2010; Yanai, 2003).

Two elementary modes of coordination have been identified and investigated in depth in the studyof rhythmic movements (Kelso, 1984; Kelso & Jeka, 1992): (i) in-phase coordination designates thesimultaneous contraction of homologous muscles (e.g., flexing or extending the arms simultaneously)and (ii) anti-phase coordination designates the simultaneous activation of non-homologous musclegroups (e.g., flexing one arm and extending the other one). These stable patterns of coordinationcan be examined by the relative phase between the limbs and its variability (Beek et al., 1995). Thisline of research was initiated by the seminal work of Kelso on phase transitions in rhythmic fingerand hand movements (Kelso, 1984; Kelso, Holt, Rubin, & Kugler, 1981) and the modeling of this phe-nomenon in terms of the dynamics of coupled oscillators (Haken, Kelso, & Bunz, 1985). It was foundthat coordinated finger movements may manifest in-phase and anti-phase coordination modes, andhave a differential stability that co-depends on the frequency of movement.

Inter-limb coordination emerges due to changes in the constraints imposed on action, rather thanprescriptions for action (Kugler, Kelso, & Turvey, 1980; Newell, 1986). In this context, three types ofconstraints have been distinguished (Newell, 1986): organismic, environmental, and task. Organismicconstraints are structural or functional, associated with the actor, such as his or her body anthropom-etry. Environmental constraints are external to the organism and pertain to environmental factors.Task constraints refer to the goal of the activity and have been classified into three categories: taskgoal(s), rules or instructions specifying the response dynamics, and implements or machines specify-ing the response dynamics. In aquatic locomotion, several inherent environmental constraints must beconsidered, such as the density or temperature of the fluid, the direction of water flow (causing dragforce), underwater visibility, and waves on the surface of the water. For studying swimming perfor-mance, two environmental constraints are continuously applied to the swimmer’s body while displac-ing him- or herself through water: drag force and the energy wasted in accelerating masses of waterused to generate propulsion (Toussaint & Beek, 1992). If these environmental constraints are inherentto aquatic locomotion, fatigue develops in a manner that depends on the task and the internal pacingof each swimmer. Fatigue in human performance can be defined as any exercise-induced decrease inthe power produced (Bigland-Ritchie & Woods, 1984; Taylor & Gandevia, 2008), which, in a maximalswimming exercise, manifests itself in a decrease in speed. Adaptive movement patterns emerge as afunction of the organism to conserve metabolic energy, for instance as a function of fatigue (Sparrow &Newell, 1998). Individuals thus naturally adopt a preferred coordination mode (Sparrow, Hughes,Russel, & Le Rossignol, 2000; Sparrow & Newell, 1998), especially during the third stage of learningwhich concerns coordination efficiency (Newell, 1985), in interaction with the key (task) constraints,in order to exploit them to the fullest to achieve performance aims (Sparrow, 1983; Sparrow et al.,2000; Sparrow & Newell, 1998). In the context of swimming, Seifert et al. (2010) found that this adap-tation process is highly individual in nature and dependent on his or her specialty.

As each individual acts differently (in consequence of his or her organismic constraints) in achiev-ing the same goal, there is no single optimal pattern of coordination. Instead there are several solu-tions to each motor problem, culminating in considerable inter-individual variability (Newell &

1202 P. Figueiredo et al. / Human Movement Science 31 (2012) 1200–1212

Corcos, 1993). Inter-individual variability reflects the patterns of coordination adopted by individualsin exploring different possible motor solutions in relation to the constraints encountered, and permitscoordination profiling (Button, Davids, & Schöllhorn, 2006). Swimmers can react differently to fatiguedepending on how they accommodate relevant constraints (environment, task, organism) (Newell,1986); notably, breathing laterality, specialty (sprint/distance). The constraints provide the directionand restrain the movement possibilities of each swimmer, as emphasized in the constraint-led ap-proach (Davids et al., 2008). In this manner, movement variability could well have a functional role.

In addition, most of the front crawl swimmers present asymmetric arm coordination. This coordi-nation asymmetry is related to breathing laterality (preferential breathing side for a unilateral breath-ing pattern) and motor laterality (arm dominance), resulting in different profiles of coordination, withdifferences in the relative duration of stroke phases (Seifert, Chollet, & Allard, 2005).

Cluster analysis is an increasingly common technique to detect and group similar patterns withinhigh dimensional datasets. One significant advantage of coordination pattern clustering is that no apriori assumptions about the structure of the dataset are required to identify similar patterns (Rein,Button, Davids, & Summers, 2010). In the study of swimming, cluster analysis has been used to clas-sify: (i) the characteristics of subject’ physical fitness, as critical velocity or anaerobic working capacityin breaststroke (Abe et al., 2006); (ii) the backstroke start, in particular the body segment vector of theswimmers (Wilson & Howard, 1983); (iii) the aerial style of the diving start (Seifert et al., 2010); (iv)the diving start profiles, covering the entire start duration, i.e., from block phase to swimming phase to15 m (Vantorre, Seifert, Fernandes, Boas, & Chollet, 2010); and (v) groups of swimmers according totheir responses to swim training (Avalos, Hellard, & Chatard, 2003).

In the last decade, coordination in aquatic locomotion, particularly front crawl swimming, has beenstudied using a coordinative measure called the Index of Coordination, which was originally proposed byChollet et al. (2000) and which is sensitive to temporal changes (Seifert, Chollet, & Rouard, 2007; Seifert &Toussaint et al., 2010). Using this parameter, the changes in motor coordination that occur with swim-ming speed have been a topic of interest (Schnitzler, Seifert, Ernwein, & Chollet, 2008; Seifert et al., 2005;Seifert & Toussaint et al., 2010). However, its use should be complemented with spatial data to enablespatial–temporal analysis; furthermore, if calculated in a continuous way, it will provide informationabout instantaneous changes (Glazier, Wheat, Pease, & Bartlett, 2006; Hamill, Haddad, & Mc Dermott,2000). Thus, the calculation of the continuous relative phase (CRP) (Hamill, van Emmerik, Heiderscheit,& Li, 1999; Kelso, 1995; Kurz & Stergiou, 2002) would yield information on the type of relationship (in-phase or anti-phase) between a pair of joints and the relative amount of coupling (from the in-phase tothe anti-phase mode of coordination), and, even more relevant, it will enable an intra-cycle analysisthrough the swimming cycle for a better understanding of the inter-limb coordination dynamics.

The aim of this study was to examine individual coordination profiling during a 200 m maximalfront crawl effort. We hypothesized that a common behavior regarding the influence of fatigue occurs,but accompanied by a high inter-individual difference in the coordination parameters.

2. Methods

2.1. Participants

Ten male competitive swimmers volunteered to participate in the study. Their main characteristicswere 21.6 ± 2.4 (M ± SD) years old, 185.2 ± 6.8 cm of height, 188.7 ± 8.4 cm of arm span, 76.4 ± 6.1 kgof body mass, 10.1 ± 1.8% of adipose tissue, 11.0 ± 3.5 years of competitive swimming background anda performance over the 200 m freestyle event of 91.6 ± 2.1% of the short course world record. Beforetesting, participants read and signed an informed consent form according to the local ethics commit-tee and Declaration of Helsinki (2000).

2.2. Experimental procedure

Before the experimental trial, swimmers performed an individual warm-up, for a total of 1000 m.Afterwards, 21 markers were placed on anatomical landmarks defining the three-dimensional position


and orientation of the head, torso, arms, forearms, hands, thighs, shanks and feet (Zatsiorsky modeladapted by de Leva, 1996). The experimental trial consisted of performing a 200-m maximum frontcrawl in a 25-m swimming pool. Two arm movements, taken in the central part of the pool, were eval-uated to avoid start and turn effects. To eliminate the effects of breathing on the variables studied,swimmers were asked to execute a non-breathing cycle while swimming through the calibrated space.The 200-m maximum front crawl proved to be a fatiguing exercise, which induced physiological andbiomechanical changes in the course of the effort (Figueiredo, Zamparo, Sousa, Vilas-Boas, & Fernan-des, 2011; Pelayo, Mujika, Sidney, & Chatard, 1996). As an indicator of exercise intensity, post-exerciseblood lactate peak value was used. Capillary blood samples (5 ll) were collected from the ear lobe toassess rest and post-exercise blood lactate by means of a portable lactate analyzer (Lactate Pro, ArkrayInc.). Blood lactate was measured before and at 1, 3, 5, and 7 min after the test.

2.3. Video analysis and stroking parameters (velocity, stroke rate, stroke length)

Six digital video cameras, four under (optical axes of adjacent cameras varied from 75� to 110�) andtwo above-water (optical axes of adjacent cameras �100�) were used to record the trial. Underwaterand above-water views were reconstructed in 3D using a calibration volume (3 � 2 � 3 m and 30 cal-ibration points), DLT algorithm (Abdel-Aziz & Karara, 1971) and a low-pass digital filter of 6 Hz. Aflashing light was used to synchronize all the cameras. Landmarks were digitized manually andframe-by-frame at a frequency of 50 Hz, using Ariel Performance Analysis System (Ariel DynamicsInc., USA). The above-described procedures enabled to calculate the average swimming speed(m s�1, as the horizontal displacement of the center of mass in one stroke cycle over its total duration),the stroke length (m, as the horizontal displacement of the center of mass in one stroke cycle) and thestroke frequency (Hz, as the time needed to complete a stroke cycle).

2.4. Inter-arm coordination

Three-dimensional segment angle was calculated between the arm’s center of mass and trunk, andangular velocities from positional data, for the right and left side. The continuous relative phase (CRP)between inter-arm (arm’s center of mass-shoulder-trunk) angles was analyzed from two arm move-ments for each lap, taken from the central part of the swimming pool to avoid start and turn effects.The stroke cycle was considered to begin in the entry of the right hand and to finish in the re-entry ofthe right hand, and was normalized to its duration, in percentage. In accordance with Hamill et al.(2000), the data on angular displacements and angular velocities were normalized (hnorm and xnorm,respectively) in the interval [�1, +1] as follows (Fig. 1):

hnorm ¼2h

hmax � hmin� hmax þ hmin

hmax � hminð1Þ

where hmax and hmin are the maximum and minimum angular position within one complete cycle,respectively.

xnorm ¼2x

xmax �xmin�xmax þxmin

xmax �xminð2Þ

where xmax and xmin is the maximum and minimum angular velocity within one complete cycle,respectively.

Phase angles (ø) were calculated using the following formula and graphical representation (Fig. 2):

/ ¼ tan�1ðxnorm=hnormÞ ð3Þ

where hnorm and xnorm are the normalized angular displacements and angular velocities, respectively.The coupling was quantified through the CRP. CRP was defined as the difference between the nor-

malized phase angles of two segment motions throughout the arm swimming cycle. In each coupling,the left shoulder phase angle was subtracted from the right one. A CRP of 0� indicated that the respec-tive segments were in-phase, and, as the CRP increased, the segments would be more out-of-phaseuntil a CRP of 180� would indicate an anti-phase coupling. A positive CRP indicated that the proximal

Normalised angle

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Nor

mal

ised

vel

ocity

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Fig. 1. Phase portrait.

Time (% of the stroke cycle)

0 20 40 60 80 100 120

Pha

se A

ngle

(º)

-400

-300

-200

-100

0

100

200

300

400

Rigth armLeft arm

Fig. 2. Phase angle for right and left side.


segment had a greater phase angle while a negative CRP indicated that the distal segment had a great-er phase angle. CRP was computed as:

CRP ¼ right shoulder phase angle� left shoulder phase angle ð4Þ

Nevertheless, two coordination modes are theoretically possible (in-phase, 0�, and anti- phase,180�). However, we used the range of 0� < CRP < 360�, because when information about the coordina-tion between segments is required, this approach is considered more suitable (Wheat & Glazier, 2006).

Following Bardy, Oullier, Bootsma, and Stoffregen (2002), Diedrich and Warren (1995), Seifert,Delignières, Boulesteix, and Chollet (2007) and Seifert et al. (2010), a lag of ±30� was accepted in thisstudy for the determination of a coordination mode. Therefore, an in-phase mode was assumed to oc-cur for 210� < CRP < 30�, while the anti-phase mode was taken to be between 150� < CRP < 180�.

2.5. Inter-subject coordination profiling

Coordination profiling between the swimmers was established by a cluster hierarchical analysis, asproposed before (Ball & Best, 2007; Button et al., 2006; Rein et al., 2010) to classify different partici-pant profiles.


2.6. Statistical analysis

2.6.1. Differences between lapsNormality of the distribution and the variance homogeneity were checked. The compound symme-

try, or sphericity, was checked using the Mauchley test. When the assumption of sphericity was notmet, the significance of F-ratios was adjusted according to the Greenhouse–Geisser procedure whenthe epsilon correction factor was <.75 or according to the Huynh–Feld procedure when the epsiloncorrection factor was >.75 (Vincent, 1999). Comparisons between laps were performed using a one-way repeated measures ANOVA. The comparison was conducted on: (i) the stroking parameters(speed, stroke frequency, stroke length, and arm angular velocity); (ii) the mean continuous relativephase (CRPmean) (iii) the standard deviation of the continuous relative phase (CRPSD); (iv) the timespent in the in-phase mode; (v) the continuous relative phase at the catch (when the hand startsthe backward movement) and at the exit (when the hand leaves the water) of the underwater handpath (CRPcatch and CRPexit, respectively) for the right (R) and left (L) hand; (vi) the maximum peakof the continuous relative phase values (CRPmax); (vii) the minimum peak of the continuous relativephase values (CRPmin); (viii) the difference between maximum and minimum peaks of CRP values(CRPdif); (ix) the time of the glide and propulsive phases; and (x) the time at which the catch and exitpoints of the hand underwater path occurred for right and left hand (Rcatch, Rexit, Lcatch, Lexit).

Since an n = 10 sampled was used, effect size was computed with Cohen’s d and Cohen’s f for t-testand repeated measures ANOVA, respectively. It was considered (i) a small effect size if 0 6 |d| 6 .2 or0 6 |f| 6 .1; (ii) a medium effect size if .2 < |d| 6 .5 or .1 < |f| 6 .25, and (iii) a large effect size if |d| > .5or |f| > .25 (Cohen, 1988).

2.6.2. Inter-subject coordination profiling2.6.2.1. Cluster analysis to classify swimming styles. Given that a high inter-subject variability of CRPwas expected, cluster hierarchical analysis using the squared Euclidean distance dissimilarity measureand the Ward linkage method was applied to determine several profiles within the studied swimmers,using all the variables assessed.

2.6.2.2. Cluster validation. According to Breiman (1996) and Rein et al. (2010), the number of clustersand classifications of the subject within the cluster were validated by a bootstrapping procedure(called ‘‘bagging’’, as applied to dendrogram and machine learning; Breiman, 1996). One bagging pro-cedure corresponds to constructing the dendrogram, then this operation was repeated N times for allsubjects minus one in order to examine whether the obtained classifications were stable or not. Forexample, if we removed Subject 1, we observed if each subject remained in his initial cluster orswitched from one cluster to another. In addition, the variables that differed significantly betweenclusters were determined using the Fisher information. The Fisher information corresponds to the ra-tio between inter-cluster (Jb) and intra-cluster (Jw) distances:

Fisher information ¼ Jb=Jw ð5Þ

where Jb and Jw corresponds to:

Jb ¼X

gNgd2ðlg ;

�XÞ ð6Þ

where Ng is the number of element in the cluster g, d is the chosen distance, lg is the center of cluster g(i.e., the mean of all points in g), and X is the center of all the points (i.e., the mean of all the points).

Jw ¼X

g

Xi�Cg

d2ðXi;lgÞ ð7Þ

where Cg are the points in cluster g, xi is the value for each point.The higher the Fisher information, the more discriminative are the variables. Based on the Fisher

information, the cluster analysis was repeated several times, removing each variable one by one. Avariable was considered significantly discriminative when the composition of the dendrogram (num-ber of clusters and classifications of the subject in the cluster) did not change in comparison to the


initial result. All tests were conducted with Minitab 15.1.0.0� software (Minitab Inc., Paris, France,2006) with a conventional significance level of p < .05.

3. Results

3.1. Fatigue effect

Swimmers showed a significant decrease in speed (13.8%), stroke frequency (5.1%), stroke length(6.9%) and angular velocity (13.3%) along the 200 m concomitant with an increase of blood lactate(11.12 ± 1.65 mmol l�1), suggesting the appearance of fatigue (Table 1). However, no differences werefound between laps in the spatial coordinative parameters, with the exception of the Rglide

(F(3,27) = 5.57, p = .004, f = .23), Rcatch (F(3,27) = 5.57, p = .004, f = .23), Rexit (F(3,27) = 11.84, p < .001,f = .33), and Lexit (F(3,27) = 11.31, p < .001, f = .41) temporal parameters that increased from the begin-ning to the end of the 200 m event (Fig. 3).


The dendrogram enabled us to classify the swimmers in two coordination profiles: six subjectscomposed cluster #1 and four subjects composed cluster #2. Eleven variables significantly explainedthe difference between the two clusters. The number of turnovers in the bagging procedure and theFisher information was used to classify these variables from the most discriminative to the less dis-criminative variables: Lglide, Lcatch, CRPmin, Lprop, CRPcatchL, CRPSD, Rglide, Rcatch, SF, Rexit, CRPcatchR, v,SL, Lexit, Rprop, CRPmax, CRPdif, CRPmean, and CRPexitL (Table 2). The most discriminative variables werethe ones with the highest Fisher information; all variables with a Fisher information value of or above.72 presented differences between the clusters; these variables were: Lglide, Lcatch, CRPmin, Lprop,CRPcatchL, CRPSD, Rglide, Rcatch, SF, Rexit, CRPcatchR (Table 2).

Cluster #1 was characterized by a flatter CRP curve (Fig. 4, left panel) and a more symmetric coor-dination profile (Fig. 5, right panel), whereas cluster #2 was characterized by the highest CRP variabil-ity (Fig. 4, left panel), the longest time of the cycle spent in glide and the most asymmetriccoordination profile (Fig. 5, right panel).

4. Discussion

4.1. Effect of fatigue

The decrease observed in the v, SF, SL, and arm angular v along the 200 m evidenced, concomitantwith the post exercise lactate, that this high intensity effort indeed induced fatigue, as previouslyshown for the same swimming distance (Bonifazi, Martelli, Marugo, Sardella, & Carli, 1993; Figueiredoet al., 2011; Pelayo et al., 1996). Coordinative spatial parameters showed no differences across the fourlaps of the 200 m event, but differences were observed in the temporal parameters. As suggested byAujouannet, Bonifazi, Hintzy, Vuillerme, and Rouard (2006), this could be due to the very high perfor-mance level and homogeneity of the group of the swimmers tested, suggesting a stable spatial pattern

Table 1Effect of fatigue on the stroking parameters (Mean (SD)).

Parameters Lap 1 Lap 2 Lap 3 Lap 4 F(3,27) p f

v (m s�1) 1.57 (0.08) 1.39a (0.06) 1.34a (0.07) 1.35a (0.08) 24.58 <0.001 1.26SL (m) 2.29 (0.23) 2.21 (0.17) 2.19 (0.13) 2.12a,b,c (0.13) 4.55 0.01 0.32SF (Hz) 0.68 (0.09) 0.63 (0.06) 0.61a (0.05) 0.64 (0.05) 5.08 0.006 0.39Rx (m s�1) 2.73 (0.42) 2.50 (0.26) 2.43a (0.26) 2.38a (0.25) 5.18 0.006 0.40Lx(m s�1) 2.83 (0.36) 2.61a (0.31) 2.52a (0.17) 2.44a (0.16) 9.42 <0.001 0.53

a,b,cStatistical different from first, second and third lap, respectively. p < .05.

Fig. 3. Timing of the occurrence of the catch and exit points for the right and left arm, and the relative duration of the glide andpropulsion phases for the right and left arm (left panel) over the 200 m front crawl. Continuous relative phase angle at the catchand exit points for the right and left arm, and the mean, minimum, maximum, SD, and difference (maximum–minimum) (rightpanel) over the 200 m front crawl. a and b Statistical different from first, and second lap, respectively. p < .05.

Table 2Difference between cluster’s parameters (Average (SD)), p < .05.

Parameters Cluster 1 Cluster 2 t p d Inter-cluster distance Intra-cluster distance Fisher

v (m s�1) 1.43 (0.04) 1.38 (0.03) 1.89 .10 1.53 0.005 0.01 .45SF (Hz) 0.67 (0.04) 0.60 (0.05) 2.47 .04 1.78 0.05 0.14 .76SL (m) 2.15 (0.08) 2.29 (0.19) �1.67 .13 1.18 36.2 47.4 .35CRPcatchR (�) 135.6 (16.3) 100.2 (35.3) 2.39 .04 1.57 3647.9 5066.6 .72CRPcatchL (�) 257.4 (28.8) 301.7 (16.1) �2.76 .03 2.00 4693.3 4916.0 .95CRPmin (�) 87.8 (11.3) 58.4 (14.5) 3.63 .007 2.61 2078.5 1259.2 1.65CRPSD (�) 63.1 (8.5) 76.0 (5.0) 2.70 .03 1.95 398.1 436.8 .91Rglide (%) 42.2 (2.9) 48.7 (4.9) �2.66 .03 1.92 101.2 114.6 .88Lglide (%) 38.9 (3.3) 47.6 (3.5) �4.00 .004 2.88 181.2 90.7 2.00Lprop (%) 36.5 (2.7) 31.2 (1.2) 3.63 .007 2.62 68.1 41.4 1.64Rcatch (%) 42.2 (2.9) 48.7 (4.9) �2.66 .03 1.92 101.2 114.6 .88Rexit (%) 78.4 (2.6) 82.9 (3.4) �2.39 .04 1.72 48.5 67.8 .72Lcatch (%) 93.2 (4.4) 101.8 (1.7) �8.59 .006 2.65 177.0 105.9 1.67Lexit (%) 29.7 (3.9) 33.0 (1.7) �1.56 .16 1.13 25.5 84.4 .30


that is not easily changed even by the impairments imposed by fatigue. The increase of temporalparameters under fatigue was marked by an increase of the Rglide time and by the time of occurrenceof Rcatch, Rexit, and Lexit. This occurred as the swimmers increased the relative duration of the propul-sive phases, which suggests a compensation for the decline of the force generating capacity (Alberty,Sidney, Huot-Marchand, Hespel, & Pelayo, 2005; Alberty, Sidney, Pelayo, & Toussaint, 2009). Concom-itant with the increase of the phases’ relative time duration, we found a decrease in the angular veloc-ity of the arm, which also reflects a fatigue effect and contributes to the reduction of the propulsiveforce generation responsible for the decrease in speed. In previous studies, this reduced propulsiveforce production under fatigue conditions was referred in fatigue stages (Figueiredo et al., 2011; Suitoet al., 2008; Toussaint, Carol, Kranenborg, & Truijens, 2006). In fact, the 200 m event, when performedat maximum intensity, is considered to be exhausting (Aujouannet et al., 2006; Bonifazi et al., 1993;Pelayo et al., 1996), and impairments of motor control resulting from fatigue may be expected (Albertyet al., 2005, 2009). This constraint may explain the observed asymmetries, as reflected in the increaseof the Rglide and Rcatch and the absence of these effects at the contralateral side, as observed previously(Barden, Kell, & Kobsar, 2011; Potts, Chralton, & Smith, 2002; Seifert et al., 2005).

The results of the continuous relative phase along the effort suggest that the intrinsic anti-phasecoordination pattern was preserved throughout the effort, in spite of the temporal changes that oc-curred in the stroke mechanics, and the fact that the propulsion efficiency declined.

Time (% of stroke cycle)

Cou

ntin

uous

Rel

ativ

e P

hase

(º)

0

50

100

150

200

250

300

350

400

Rcatch

Rexit

Lcatch

Lexit

Rentry

Lentry


0 20 40 60 80 100 120 0 20 40 60 80 100 120

Cou

ntin

uous

Rel

ativ

e P

hase

(º)

0

50

100

150

200

250

300

350

400

Rcatch

Rexit

Lcatch

Lexit

Rentry

Lentry

Fig. 4. Continuous relative phase between right (R) shoulder phase angle (arm’s center of mass - shoulder - trunk) and left (L)shoulder phase angle through a complete cycle for the cluster 1 (right panel) and cluster 2 (left panel). Entry, catch, and exitkeys points of the stroke cycle are identified.


Ang

le (

º)

0

20

40

60

80

100

120

140

160

180

200Rigth armLeft arm

Rcatch

Rexit

Lcatch

Lexit

Rentry Lentry


0 20 40 60 80 100 120 0 20 40 60 80 100 120

Ang

le (

º)

0

20

40

60

80

100

120

140

160

180

200Rigth armLeft arm

Rcatch

Rexit

Lcatch

Lexit

Rentry Lentry

Fig. 5. Variations of the right and left shoulder angles (arm’s centre of mass - shoulder - trunk) through a complete cycle for thecluster 1 (right panel) and cluster 2 (left panel). Entry, catch, and exit keys points of the stroke cycle are identified.


It may be that the propulsion efficiency was declining somewhat because at the end of the effort SFwas increasing; however v was stable, and so, propelling efficiency was lower per arm stroke. Similarstrong coupling relationships have also been observed between arm and leg in various rhythmic activ-ities, including walking, creeping and swimming (Wannier et al., 2001), but also between the arms insprint swimming (Nikodelis, Kollias, & Hatzitaki, 2005). However, in this situation fatigue would beexpected to be sufficient to disrupt the intrinsic coordination pattern, given the high drag forces ex-erted by the water to the swimmer. Contrary to our expectations, and perhaps due to the strong influ-ence of a central pattern generator (Kuo, 2002; Nikodelis et al., 2005), however, the coordinationremained stable.

The present findings are limited to the analysis of two arm stroke cycles for each 50 m lap andtherefore critical notions related to the coordination dynamics of interest might be influenced. How-ever, the use of one cycle to represent performance at some stage of the effort is common practice inperformance-oriented sports research (Glazier, Davids, & Bartlett, 2003; Nikodelis et al., 2005), partic-ularly when dealing with a whole-body 3D approach (Figueiredo, Vilas Boas, Maia, Goncalves, &Fernandes, 2009; Psycharakis, Naemi, Connaboy, McCabe, & Sanders, 2010).


In the front crawl swimming technique the arms are temporally coupled in an intrinsic anti-phasesynchronization mode, which seems to be preserved in spite of the prevailing environmental


constraints and skill level (Nikodelis et al., 2005). Unsurprisingly therefore, all swimmers had an anti-phase coordination mode. Whereas CRPmean showed no significant differences between the two clus-ters, CRPSD, mostly because of CRPmin, proved to be different between clusters, implying that cluster#1 had a flatter CRP curve, reflecting lower variability. Differences noted in the CRPcatch between clus-ters also revealed different tendencies to spend more or less time in the glide phase. This highlightsthe individual dependence on different constraints like breathing laterality, specialty (sprint/distance)predicted by the constraint-led approach, and reflects the individual optimization of swimmers’ ownconstraints to satisfy the key performance constraint, to swim faster. Bilateral asymmetry in elite frontcrawl swimming has been reported before (Barden et al., 2011; Nikodelis et al., 2005; Potts et al.,2002; Psycharakis & Sanders, 2008; Seifert et al., 2005). Seifert et al. (2005) found that arm coordina-tion asymmetry varied greatly among swimmers and was associated with the handedness and theunilateral breathing pattern adopted by the swimmers, which suggests that the asymmetry was pre-dominantly propulsion-based. It has also been reported that swimmers might apply larger propulsiveforces when pulling with the dominant arm (Barden et al., 2011; Potts et al., 2002). We observed thatthe underwater phases of the right arm were longer, probably because the participants were right-handed; this can be related to differences in the magnitude, duration, timing, or direction of propul-sive forces between the underwater phases of the dominant and non-dominant arm (Psycharakis &Sanders, 2008). Sadeghi, Allard, Prince, and Labelle (2000) reviewed the symmetry and limb domi-nance in gait analysis and suggested that the dominant limb might be used mainly for propulsionand the non-dominant for control and support. Barden et al. (2011) found that the degree of bilateralasymmetry does not increase in relation to augmented exercise intensity; in fact, their results showedthat arm asymmetry decreased with increasing exercise intensity, and that the decreases in asymme-try were proportional for both the power (underwater) and recovery (above water) phases of thestroke cycle, confirming that higher degrees of anti-phase inter-limb coupling (i.e., less bilateral asym-metry) occur at faster speeds in front crawl swimming (Nikodelis et al., 2005). However, the presenceof a higher external power imbalance when nearing exhaustion (about 8%) between the left and rightarm was also observed before (Potts et al., 2002). In our study asymmetries seemed to be stable orslightly increased, as changes occurred, Rglide increased, while Lprop had a tendency to increase(p = .07). Whereas the results of Potts et al. (2002) were found at exhaustion using an incremental pro-tocol, the present results indicated the occurrence of fatigue, and a decrease in velocity. The presentresults also underscored the need of profiling, as arm coordination asymmetry varied greatly amongswimmers, which could augment the imbalance between propulsive and non-propulsive forces,increasing intra-cycle velocity variation and energy cost for the same average speed.

Swimmers of cluster #2 had higher values of Rglide and Lglide indicating a prevalence of swimmingwith more glide time and a tendency for higher SL, as compared to the counterparts of cluster #1. Thiscould be explained by the swimmer’s specialty, i.e., an organismic constraint, since the swimmers inthis cluster also performed well at the 400 and 800 m competitive events. These competitive distancesare associated with higher SL (Craig, Skehan, Pawelczyk, & Boomer, 1985) and a preference for a longerhigher glide phase than in shorter distances, implying higher propelling efficiency values compared tosprinters (Cappaert, Bone, & Troup, 1992; Seifert & Komer et al., 2010). Also a longer relative durationof the left arm propulsion was observed, since swimmers presented a left preference breathing side,resulting in asymmetric coordination (Barden et al., 2011). In spite of being carried out only in non-breathing cycles, Seifert, Chehensse, Tourny-Chollet, Lemaitre, and Chollet (2008) showed that evenin this condition, swimmers use a coordination mode similar to front crawl breathing cycles, that ispart of their usual repertoire, emphasizing the powerful role of learned breathing laterality. These con-clusions were attained even when the swimmer did not need to turn his head to breathe, indicatingthat unilateral breathing disturbs arm coordination, and that all the stroke key points studied occurredat a later time within the stroke cycle.

In contrast, the swimmers belonging to cluster #1 were more symmetric and more balanced, con-tributing to this, the higher SF presented in relation to cluster #2, as well as the above-discussed lat-erality (Seifert et al., 2005). Yanai (2003) and Psycharakis and Sanders (2008) showed that swimmersreduce the amplitude of the body roll as they increase their stroke frequency, maintaining both armspatterns more constant and similar. This is expected from sprinters, as was the case for the swimmersin cluster #1, as they use higher SF, a more continuous inter-limb coordination (Seifert et al., 2010),


and also present higher relative durations in the pull phase for maximal velocity bouts (McCabe,Psycharakis, & Sanders, 2011). As Seifert et al. (2010) hypothesized, the sprint swimmers modifiedtheir stroke mechanics because they usually focus on propulsion to overcome high active drag, whichis linked to the increase of overall propulsion impulse (Alberty et al., 2009) and the increase in SF. Inspite of the differences between clusters, which probably occurred as a result of the swimmers’ spe-cialty, no differences were found between clusters for the hand speed, contrary to precious reports(McCabe et al., 2011), presumably because all swimmers were 200 m specialist.

5. Conclusion

This study highlighted the changes caused by the fatigue induced by swimming 200 m front crawlat maximal effort. Significant changes were evident in physiologic, kinematic and coordinative param-eters. Although these changes occurred across all swimmers, individual adaptations were found tocluster in two groups. Differences in the general biomechanical parameters to achieve the same veloc-ity, presenting different combinations of SL and SF, promoting the occurrence of asymmetries, werealso caused by the inherent breathing pattern of the swimmers. It was possible however that the ob-served differences were mainly due to their specialty, i.e., as sprinters or distance swimmers.

Acknowledgments

This investigation was supported by grants of the Portuguese Science and Technology Foundation(SFRH/BD/38462/2007 and PTDC/DES/101224/2008).

References

Abdel-Aziz, Y., & Karara, H. (1971). Direct linear transformation: From comparator coordinates into object coordinates in closerange photogrammetry. In Proceedings of the symposium on close-range photogrammetry (pp. 1–18). Illinois: Church Falls.

Abe, D., Tokumaru, H., Nihata, S., Fukuoka, Y., Usui, S., & Yoshida, T. (2006). Assessment of short-distance breaststrokeswimming performance with critical velocity. Journal of Sports Sciences and Medicine, 5, 340–348.

Alberty, M., Sidney, M., Huot-Marchand, F., Hespel, J. M., & Pelayo, P. (2005). Intracyclic velocity variations and arm coordinationduring exhaustive exercise in front crawl stroke. International Journal of Sports Medicine, 26, 471–475.

Alberty, M., Sidney, M., Pelayo, P., & Toussaint, H. M. (2009). Stroking characteristics during time to exhaustion tests. Medicineand Science in Sports and Exercise, 41, 637–644.

Aujouannet, Y. A., Bonifazi, M., Hintzy, F., Vuillerme, N., & Rouard, A. H. (2006). Effects of a high-intensity swim test on kinematicparameters in high-level athletes. Applied Physiology, Nutrition, and Metabolism, 31, 150–158.

Avalos, M., Hellard, P., & Chatard, J. C. (2003). Modeling the training-performance relationship using a mixed model in eliteswimmers. Medicine and Science in Sports and Exercise, 35, 838–846.

Ball, K. A., & Best, R. J. (2007). Different centre of pressure patterns within the golf stroke I: Cluster analysis. Journal of SportsSciences, 25, 757–770.

Barden, J. M., Kell, R. T., & Kobsar, D. (2011). The effect of critical speed and exercise intensity on stroke phase duration andbilateral asymmetry in 200-m front crawl swimming. Journal of Sports Sciences, 29, 517–526.

Bardy, B. G., Oullier, O., Bootsma, R. J., & Stoffregen, T. A. (2002). Dynamics of human postural transitions. Journal of ExperimentalPsychology: Human Perception and Performance, 28, 499–514.

Beek, P. J., Peper, C. E., & Stegeman, D. F. (1995). Dynamical models of movement coordination. Human Movement Science, 14,573–608.

Bigland-Ritchie, B., & Woods, J. J. (1984). Changes in muscle contractile properties and neural control during human muscularfatigue. Muscle Nerve, 7, 691–699.

Bonifazi, M., Martelli, G., Marugo, L., Sardella, F., & Carli, G. (1993). Blood lactate accumulation in top level swimmers followingcompetition. The Journal of Sports Medicine and Physical Fitness, 33, 13–18.

Breiman, L. (1996). Bagging predictors. Machine Learning, 24, 123–140.Bucher, W. (1975). The influence of the leg kick and the arm stroke on the total speed during the crawl stroke. In L. Lewillie & J.

P. Clarys (Eds.), Swimming II (pp. 180–187). Baltimore: University Park Press.Button, C., Davids, K., & Schöllhorn, W. (2006). Coordination profiling of movement systems. In K. Davids, S. Bennett, & K. Newell

(Eds.), Movement system variability (pp. 133–152). Champaign, IL: Human Kinetics Publishers.Cappaert, J. M., Bone, M., & Troup, J. P. (1992). Intensity and performance related differences in propelling and mechanical

efficiencies. In D. MacLaren, T. Reilly, & A. Lees (Eds.), Swimming Science VI (pp. 49–52). London: E & FN SPON.Chollet, D., Chalies, S., & Chatard, J. C. (2000). A new index of coordination for the crawl: Description and usefulness.

International Journal of Sports Medicine, 21, 54–59.Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.Craig, A. B., Jr., Skehan, P. L., Pawelczyk, J. A., & Boomer, W. L. (1985). Velocity, stroke rate, and distance per stroke during elite

swimming competition. Medicine and Science in Sports and Exercise, 17, 625–634.


Davids, K., Button, C., & Bennett, S. (2008). Dynamics of skill acquisition: A constraints-led approach. Champaign, IL: HumanKinetics.

de Leva, P. (1996). Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters. Journal of Biomechanics, 29, 1223–1230.Deschodt, V. J., Arsac, L. M., & Rouard, A. H. (1999). Relative contribution of arms and legs in humans to propulsion in 25-m

sprint front-crawl swimming. European Journal of Applied Physiology and Occupational Physiology, 80, 192–199.Diedrich, F. J., & Warren, W. H. Jr., (1995). Why change gaits? Dynamics of the walk-run transition. Journal of Experimental

Psychology: Human Perception and Performance, 21, 183–202.Donker, S. F., Beek, P. J., Wagenaar, R. C., & Mulder, T. (2001). Coordination between arm and leg movements during locomotion.

Journal of Motor Behavior, 33, 86–102.Figueiredo, P., Vilas Boas, J. P., Maia, J., Goncalves, P., & Fernandes, R. J. (2009). Does the hip reflect the centre of mass swimming

kinematics? International Journal of Sports Medicine, 30, 779–781.Figueiredo, P., Zamparo, P., Sousa, A., Vilas-Boas, J. P., & Fernandes, R. J. (2011). An energy balance of the 200 m front crawl race.

European Journal of Applied Physiology, 111, 767–777.Glazier, P. S., & Davids, K. (2009). On analysing and interpreting variability in motor output. [Comment Letter]. Journal of Science

and Medicine in Sport, 12, e2–e3. author reply e4–e5.Glazier, P. S., Wheat, J. S., Pease, D. L., & Bartlett, R. M. (2006). Dynamic systems theory and the functional role of movement

variability. In K. Davids, S. Bennett, & K. Newell (Eds.), Movement system variability (pp. 49–72). Champaign, IL: HumanKinetics Publishers.

Glazier, S. P., Davids, K., & Bartlett, R. M. (2003). Dynamical systems theory: A relevant framework for performance-orientedsports biomechanics research. Sport Science, 7. <http://www.sportsci.org/jour/03/psg.htm>.

Haken, H., Kelso, J. A. S., & Bunz, H. (1985). A theoretical model of phase transitions in human hand movements. BiologicalCybernetics, 51, 347–356.

Hamill, J., Haddad, J. M., & Mc Dermott, W. (2000). Issues in quantifying variability from a dynamical systems perspective.Journal of Applied Biomechanics, 16, 407–418.

Hamill, J., van Emmerik, R. E. A., Heiderscheit, B. C., & Li, L. (1999). A dynamical systems approach to lower extremity runninginjuries. Clinical Biomechanics, 14, 297–308.

Hollander, A. P., de Groot, G., van Ingen Schenau, G. J., Kahman, R., & Toussaint, H. M. (1988). Contribution of the legs topropulsion in swimming. In B. E. Ungerechts, K. Wilkie, & K. Reischle (Eds.), Swimming science V (pp. 39–43). Champaign, IL:Human Kinetics Publishers.

Kelso, J. A. S. (1984). Phase transitions and critical behavior in human bimanual coordination. American Journal of Physiology,246, 1000–1004.

Kelso, J. A. S., Holt, K. G., Rubin, P., & Kugler, P. N. (1981). Patterns of human interlimb coordination emerge from the propertiesof non-linear, limit cycle oscillatory processes: Theory and data. Journal of Motor Behavior, 13, 226–261.

Kelso, J. A. S., & Jeka, J. J. (1992). Symmetry breaking dynamics of human multilimb coordination. Journal of ExperimentalPsychology: Human Perception and Performance, 18, 645–668.

Kelso, J. A. S. (1995). Dynamic patterns: The self-organization of brain and behavior. Cambridge, Mass: MIT Press.Kugler, P. J., Kelso, J. A. S., & Turvey, M. T. (1980). On the concept of coordinative structures as dissipative structures: 1.

Theoretical lines of convergence. In G. E. Stelmach & J. Requin (Eds.), Tutorials in motor behavior (pp. 3–37). Amsterdam:North-Holland.

Kuo, A. D. (2002). The relative roles of feedforward and feedback in the control of rhythmic movements. Motor Control, 6,129–145.

Kurz, M. J., & Stergiou, N. (2002). Effect of normalization and phase angle calculations on continuous relative phase. Journal ofBiomechanics, 35, 369–374.

Lee, T. D., Swinnen, S. P., & Verschueren, S. (1995). Relative phase alterations during bimanual skill acquisition. Journal of MotorBehavior, 27, 263–274.

McCabe, C. B., Psycharakis, S., & Sanders, R. (2011). Kinematic differences between front crawl sprint and distance swimmers atsprint pace. Journal of Sports Sciences, 29, 115–123.

Newell, K. M. (1985). Coordination, control and skill. In D. Goodman, R. B. Wilberg, & I. M. Franks (Eds.), Differing perspectives inmotor learning, memory and control (pp. 295–317). Amsterdam: North-Holland.

Newell, K. M. (1986). Constraints on the development of coordination. In M. G. Wade & H. T. A. Whiting (Eds.), Motordevelopment in children: Aspect of coordination and control (pp. 341–360). Dordrecht: Nijhoff.

Newell, K. M., & Corcos, D. M. (1993). Variability and motor control. Champaign, Ill: Human Kinetics.Nikodelis, T., Kollias, I., & Hatzitaki, V. (2005). Bilateral inter-arm coordination in freestyle swimming: Effect of skill level and

swimming speed. Journal of Sports Sciences, 23, 737–745.Pelayo, P., Mujika, I., Sidney, M., & Chatard, J. (1996). Blood lacate recovery measurements, training, and performance during a

23-week period of competitive swimming. European Journal of Applied Physiology, 12, 197–206.Potts, A. D., Chralton, J. E., & Smith, H. M. (2002). Bilateral arm power imbalance in swim bench exercise to exhaustion. Journal of

Sports Sciences, 20, 975–979.Psycharakis, S. G., Naemi, R., Connaboy, C., McCabe, C., & Sanders, R. H. (2010). Three- dimensional analysis of intracycle velocity

fluctuations in frontcrawl swimming. Scandinavian Journal of Medicine & Science in Sports, 20, 128–135.Psycharakis, S. G., & Sanders, R. H. (2008). Shoulder and hip roll changes during 200-m front crawl swimming. Medicine and

Science in Sports and Exercise, 40, 2129–2136.Rein, R., Button, C., Davids, K., & Summers, J. (2010). Cluster analysis of movement patterns in multiarticular actions: A tutorial.

Motor Control, 14, 211–239.Sadeghi, H., Allard, P., Prince, F., & Labelle, H. (2000). Symmetry and limb dominance in able-bodied gait: A review. Gait Posture,

12, 34–45.Schnitzler, C., Seifert, L., Ernwein, V., & Chollet, D. (2008). Arm coordination adaptations assessment in swimming. International

Journal of Sports Medicine, 29, 480–486.Schöner, G., & Kelso, J. A. S. (1988). Dynamic pattern generation in behavioral and neural systems. Science, 239(4847),

1513–1520.

http://www.sportsci.org/jour/03/psg.htm


Seifert, L., Chehensse, A., Tourny-Chollet, C., Lemaitre, F., & Chollet, D. (2008). Effect of breathing pattern on arm coordinationsymmetry in front crawl. Journal of Strength and Conditioning Research, 22, 1670–1676.

Seifert, L., Chollet, D., & Allard, P. (2005). Arm coordination symmetry and breathing effect in front crawl. Human MovementScience, 24, 234–256.

Seifert, L., Chollet, D., & Rouard, A. (2007). Swimming constraints and arm coordination. Human Movement Science, 26, 68–86.Seifert, L., Delignières, D., Boulesteix, L., & Chollet, D. (2007). Effect of expertise on butterfly stroke coordination. Journal of Sports

Sciences, 25, 131–141.Seifert, L., Komar, J., Lepretre, P. M., Lemaitre, F., Chavallard, F., Alberty, M., et al (2010). Swim specialty affects energy cost and

motor organization. International Journal of Sports Medicine, 31, 624–630.Seifert, L., Leblanc, H., Chollet, D., & Delignières, D. (2010). Inter-limb coordination in swimming: Effect of speed and skill level.

Human Movement Science, 29, 103–113.Seifert, L., Toussaint, H. M., Alberty, M., Schnitzler, C., & Chollet, D. (2010). Arm coordination, power, and swim efficiency in

national and regional front crawl swimmers. Human Movement Science, 29, 426–439.Seifert, L., Vantorre, J., Lemaitre, F., Chollet, D., Toussaint, H. M., & Vilas-Boas, J. P. (2010). Different profiles of the aerial start

phase in front crawl. Journal of Strength and Conditioning Research, 24, 507–516.Sparrow, W. A. (1983). The efficiency of skilled performance. Journal of Motor Behavior, 15, 237–261.Sparrow, W. A., Hughes, K. M., Russel, A. P., & Le Rossignol, P. F. (2000). Movement economy, preferred modes, and pacing. In W.

A. Sparrow (Ed.), Energetics of human activity. Champaign, IL: Human Kinetics.Sparrow, W. A., & Newell, K. M. (1998). Metabolic energy expenditure and the regulation of movement economy. Psychonomic

Bulletin and Review, 5, 173–196.Suito, H., Ikegami, Y., Nunome, H., Sano, S., Shinkai, H., & Tsujimoto, N. (2008). The effect of fatigue on the underwater arm

stroke motion in the 100-m front crawl. Journal of Applied Biomechanics, 24, 316–324.Swinnen, S. P., van Langendonk, L., Verschueren, S., Peeters, G., Dom, R., & de Weerdt, W. (1997). Interlimb coordination deficits

in patients with Parkinson’s disease during the production of two-joint oscillations in the sagittal plane. MovementDisorders, 12, 958–968.

Taylor, J. L., & Gandevia, S. C. (2008). A comparison of central aspects of fatigue in submaximal and maximal voluntarycontractions. Journal of Applied Physiology, 104, 542–550.

Toussaint, H. M., & Beek, P. J. (1992). Biomechanics of competitive front crawl swimming. Sports Medicine, 13, 8–24.Toussaint, H. M., Carol, A., Kranenborg, H., & Truijens, M. J. (2006). Effect of fatigue on stroking characteristics in an arms-only

100-m front-crawl race. Medicine and Science in Sports and Exercise, 38, 1635–1642.van Emmerik, R. E., & Wagenaar, R. C. (1996). Effects of walking velocity on relative phase dynamics in the trunk in human

walking. Journal of Biomechanics, 29, 1175–1184.Vantorre, J., Seifert, L., Fernandes, R. J., Boas, J. P., & Chollet, D. (2010). Kinematical profiling of the front crawl start. International

Journal of Sports Medicine, 31, 16–21.Vincent, J. W. (1999). Statistics in kinesiology (2nd ed.). Champaign, IL: Human Kinetics.Wagenaar, R. C., & Beek, W. J. (1992). Hemiplegic gait: A kinematic analysis using walking speed as a basis. Journal of

Biomechanics, 25, 1007–1015.Wagenaar, R. C., & van Emmerik, R. E. (2000). Resonant frequencies of arms and legs identify different walking patterns. Journal

of Biomechanics, 33, 853–861.Wannier, T., Bastiaanse, C., Colombo, G., & Dietz, V. (2001). Arm to leg coordination in humans during walking, creeping and

swimming activities. Experimental Brain Research, 141, 375–379.Wheat, J. S., & Glazier, P. (2006). Measuring coordination and variability in coordination. In K. Davids, S. Bennett, & K. Newell

(Eds.), Movement system variability (pp. 167–181). Champaign, IL: Human Kinetics.Wilson, B. D., & Howard, A. (1983). The use of cluster analysis in movement description and classification of the backstroke

swim start. In H. Matsui & K. Kobayashi (Eds.), Biomechanics VIII-B (pp. 1223–1230). Human Kinetics Publisher.Yanai, T. (2003). Stroke frequency in front crawl: Its mechanical link to the fluid forces required in non-propulsive directions.

Journal of Biomechanics, 36, 53–62.

Documents

Individual profiles of spatio-temporal coordination in high intensity swimming