Available online at www.sciencedirect.

com

Journal of Science and Medicine in Sport 14 (2011) 168.e1–168.e5

Original paper

Stroke phases responses around maximal lactate steady state in front crawl
Jailton G. Pelarigo, Benedito S. Denadai ∗ , Camila C. Greco
Human Performance Laboratory, UNESP, Rio Claro, Brazil
Received 25 April 2010; received in revised form 19 July 2010; accepted 19 August 2010

Abstract
The objective of this study was to analyze changes in stroke rate (SR), stroke length (SL) and stroke phases (entry and catch, pull, push and
recovery) when swimming at (MLSS) and above (102.5% MLSS) the maximal lactate steady state. Twelve endurance swimmers (21 ± 8 year,
1.77 ± 0.10 m and 71.6 ± 7.7 kg) performed in different days the following tests: (1) 200- and 400-m all-out tests, to determine critical speed
(CS), and; (2) 2–4 30-min sub-maximal constant-speed tests, to determine the MLSS and 102.5% MLSS. There was significant difference
among MLSS (1.22 ± 0.05 m s−1 ), 102.5% MLSS (1.25 ± 0.04 m s−1 ) and CS (1.30 ± 0.08 m s−1 ). SR and SL were maintained between the
10th and 30th minute of the test swum at MLSS and have modified significantly at 102.5% MLSS (SR – 30.9 ± 3.4 and 32.2 ± 3.5 cycles min−1
and SL – 2.47 ± 0.2 and 2.38 ± 0.2 m cycle−1 , respectively). All stroke phases were maintained at 10th and 30th minute at MLSS. However, the
relative duration of propulsive phase B (pull) increased significantly at 102.5% MLSS (21.7 ± 3.4% and 22.9 ± 3.9%, respectively). Therefore,
the metabolic condition may influence the stroke parameters (SR and SL) and stroke strategy to maintain the speed during swim tests lasting
30 min.
© 2010 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.

Keywords: Swimming technique; Stroke length; Aerobic capacity; Aerobic training

1. Introduction 100-m front crawl races. The authors found that the reduc-
tion (24%) in power-generating capacity (i.e., fatigue) led to
In swimming, both biomechanical (i.e., the level of appli- a 12.4% decrease in speed. In addition, SR declined through-
cation of propulsive force and passive and active drag) and out the race (10.6%) and was related with the decrease in
physiological aspects (i.e., the energy production systems) swim speed. As the reduction in swimming speed leads to a
are essential to the locomotion and significantly contribute to reduction in drag, the authors proposed that the SR adjusts to
performance.1 The swimming technique has been frequently the reduced propulsion requirements.
assessed using both stroking parameters (stroke rate – SR and During imposed paced, the swimmer can increase the
stroke length – SL) and the index of coordination (IdC). The time relative to the whole stroke cycle over which forces are
IdC represents how the swimmers organize the propulsive applied to obtain the propulsive impulse necessary to main-
(pull and push) and non-propulsive (glide + catch and recov- tain the speed, as fatigue develops.8 Indeed, Alberty et al.3
ery) phases of right and left arms.2 Some factors which may have demonstrated that the duration of non-propulsive phases
interfere on the stroking parameters and arm coordination are decreased, whereas the duration of the propulsive phases
speed, swimming ability and fatigue.2,3 remained constant during imposed swim speed tests until
It has been demonstrated reduction in SL and increase in exhaustion. Since the SR and IdC were higher, the contri-
SR during all-out and imposed swim paces,3–5 which may bution of propulsive phases to propulsion per distance unit
result from the reduced capacity to generate force to over- was increased. It is important to note that these data were
come drag.6 Toussaint et al.7 have analyzed the speed, SR, obtained at 95, 100 and 110% of the mean speed attained in
SL and power output (MAD system) during two maximal a 400-m race (S400), where no blood lactate steady state is
observed.9,10
∗ Corresponding author.
Some authors have found significant changes in SR and
E-mail address: [email protected] (B.S. Denadai). SL only when swimming above the anaerobic threshold11

1440-2440/$ – see front matter © 2010 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2010.08.004
168.e2 J.G. Pelarigo et al. / Journal of Science and Medicine in Sport 14 (2011) 168.e1–168.e5

and maximal lactate steady state (MLSS).9 These findings bottom of the pool. The swimmers were instructed to keep
led the authors to suggest a possible relationship between their feet above the red marks at each signal. The difference
the development of peripheral fatigue and the degradation in between predict and actual swim speed was lower than 2% in
swimming technique.9,11 To the best of our knowledge, there all tests. Based on the results of Dekerle et al.,9 the first trial
is no data referring to the stroke strategies to maintain the was performed at 88.5% S400. If during the first trial a steady
speed during long-distance swims performed at conditions state or a decrease in lactate was observed, further subsequent
of metabolic equilibrium (at or below MLSS). These con- trial with 2.5% higher speeds was performed on separate days
ditions are found during some swim training sessions and until no blood lactate concentration ([La]) steady state could
long-distance events (open water swims and long-distance be maintained. If the first trial resulted in a clearly identi-
triathlons) and are important to promote specific adap- fiable increase of the [La] and/or could not be completed
tations (i.e., oxidative capacity) related to long duration due to exhaustion, further trials were conducted with sub-
performance.12 sequently reduced speeds. Earlobe capillary blood samples
This study has been designed in an attempt to analyze (25 ␮L) were collected in capillary tubes at the 10th and the
changes in stroke parameters (i.e., SL and SR) and arm coor- 30th minute of each test, and subsequently analyzed for cap-
dination (i.e., propulsive and non-propulsive phases and IdC) illary blood [La] using an automated analyzer (YSI 2300,
when swimming at (blood lactate steady state) and above Yellow Springs, OH, USA). Blood samples collection lasted
(non blood lactate steady state) MLSS. Based on previous 30 s. MLSS was defined as the highest [La] that increased by
findings,3,9 it was hypothesized that (a) there would be reduc- no more than 1 mmol L−1 between the 10th and 30th minute
tion of SL and increase of SR when swimming above MLSS, of the sub-maximal test, and were calculated as the average
and (b) these changes would be concomitant with an increase of the two [La] values.
in the relative duration of propulsive phases (pull and push) The swimmers were filmed during all tests by two cameras
and IdC. with rapid shutter speed (1/1000s); one above the surface of
the water (Panasonic – NV-GS 180, operating at 30 Hz) and
one below (Panasonic – NV-GS 320, operating at 30 Hz and
contained in a waterproof box). They were fixed on a trolley
2. Methods and pushed by an operator. This apparatus allows the analysis
of each stroke cycle in a sagittal plane. Views of both cameras
Twelve middle-distance and long-distance male were postsynchronized with a visual signal that was visible
swimmers (mean ± SD; 21 ± 8 year, 1.77 ± 0.10 m and on recordings for both cameras. The camera above the water
71.6 ± 7.7 kg) volunteered and gave written informed measured the time over a distance of 12.5 m (from 10 m to
consent to participate in the present study, which was 22.5 m) to obtain the average velocity and SR, from which
approved by the university’s ethics committee. Participants SL was calculated. The reference to register the time was
were undergoing training for at least 5 years (8 training the swimmer’s head. The determination of stroke phases was
sessions a week; 50 km per week during the 2 weeks prior conducted by an expertise operator (i.e., more than 30 h of
testing), and were competing at regional and national experience with images analysis).14 Dvideow software15 was
level (400-m performance equal to 289 s, corresponding to used for the video analysis of each stroke with an accuracy
74.1 ± 2.5% word record). The participants were instructed of 0.017 s.
to refrain from intense training sessions at least 24 h before Arm coordination was quantified using the IdC defined
the experimental sessions. by Chollet et al.2 Each arm movement was divided into four
The tests were performed in a 25-m outdoor swimming- distinct phases (one phase corresponded to an action between
pool (26 ◦ C). All tests were swum in front crawl, initiated two times), defined as phases A (glide + catch), B (pull), C
with a push start. They were all conducted within a 14-day (push) and D (recovery). The duration of each phase was
period. Testing occurred at the same time of the day (±2 h) to determined as proposed by Chollet et al.2 and expressed as a
minimize the effect of circadian variation on performance.13 percentage of the duration of a complete stroke. The duration
Swimmers firstly performed two all-out tests at the dis- of a complete stroke was the sum of the propulsive (B and C)
tances of 200 (S200) and 400 m (S400), for the determination and non-propulsive phases (A and D). The lag time between
of critical speed (CS). Participants swam one event per day the beginning of propulsion in the first right arm stroke and
in random order. CS was determined using the slope of the the end of propulsion in the first left arm stroke defined IdC1
linear regression between swimming distances and the time (i.e., arm coordination to the left side), which was expressed
taken to swim. as a percentage of the duration of a complete stroke. The lag
Then, all swimmers performed 2–4 30-min sub-maximal time between the beginning of propulsion in the second left
tests at imposed pace (one test per day) for the determina- arm stroke and the end of propulsion in the first right arm
tion of MLSS and 102.5% MLSS. The swimming speed was stroke defined IdC2 (i.e., arm coordination to the right side),
controlled using an mp3 player attached to the goggles of the which was also expressed as a percentage of the duration of
swimmer. A regular audible signal gave the target pace to a complete stroke. IdC was thus the mean of IdC1 + IdC2.2
the swimmer. Four red marks were placed every 5 m at the The arm coordination was classified as “catch-up”, if the IdC
 J.G. Pelarigo et al. / Journal of Science and Medicine in Sport 14 (2011) 168.e1–168.e5 168.e3

Fig. 1. Mean ± SD values of stroke rate (SR) and stroke length (SL) obtained at 10th and 30th min of the exercise swum at MLSS and 102.5% MLSS. #p < 0.05
significantly different with the value at 10th min, N = 12.

value was lower than 0, “opposition” if the value was equal 30th min (p = 0.477 and p = 0.977, respectively) between two
to 0 and “superposition” if the value was higher than 0.2 exercise intensities (Fig. 1).
The values were expressed as mean ± SD. For each set Stroke phases A (p = 0.313), B (p = 0.075), C (p = 0.952)
of data, normal distribution (Shapiro–Wilk test) and homo- and D (p = 0.374) were maintained at 10th and 30th minute
geneity of variance were checked. A 2-way ANOVA with at MLSS. However, phase B increased significantly at
repeated measures (Bonferroni correction post hoc test using 102.5% MLSS (p = 0.035), while the phases A (p = 0.233),
the Greenhouse-Geisser procedure) was used to determine C (p = 0.674) and D (p = 0.400) were maintained (Table 1).
the effect of exercise intensity (at and above MLSS) and There was no significant effect of time on IdC values at
time (minutes 10 vs. 30) on SL and SR. The changes in the MLSS (−4.68 ± 6.6% and −3.84 ± 6.2%) and above MLSS
stroking phases and IdC over the constant-speed tests were (−3.85 ± 6.2% and −3.15 ± 6.1%). In all conditions, the arm
compared using the Wilcoxon test. A significance level of coordination adopted was catch-up.
5% was accepted (p ≤ 0.05).

4. Discussion
3. Results
The main findings of this study were that: (a) the SL
The values of S200 and S400 were 1.45 ± 0.05 m s−1 and decreased and the relative duration of pull phase increased
1.37 ± 0.05 m s−1 , respectively. There was significant differ- throughout the exercise when swimming above (102.5%)
ence among MLSS (1.22 ± 0.05 m s−1 , 88.6 ± 1.1% S400) MLSS, and; (b) the IdC and non-propulsive phases were
and 102.5% MLSS (1.25 ± 0.04 m s−1 , 91.3 ± 1.1% S400) not modified throughout the time, irrespectively of the swim
(p = 0.002), MLSS and CS (1.30 ± 0.08 m s−1 , 93.7 ± 2.7% speed analyzed. Thus, when swimming at heavy intensity
S400) (p = 0.007) and 102.5% MLSS and CS (p = 0.028). domain (i.e., below CS – intensities where the steady state in
The [La] at MLSS (3.28 ± 0.97 mmol L−1 ) was signifi- VO2 is delayed, and an additional slow component of VO2
cantly lower than at 102.5% MLSS (4.59 ± 1.36 mmol L−1 ) causes an eventual and elevated steady state),10 the metabolic
(p = 0.001). condition may influence the stroke technique to maintain the
SR was maintained between the 10th and 30th minute speed during swim tests lasting 30 min.
of the test swum at MLSS (p = 0.068) and increased sig- Studies analyzing all-out distance trials3,6 and imposed
nificantly at 102.5% MLSS (p = 0.015). Similarly, SL was pace9,11 have found reduction in SL and increase in SR,
maintained (p = 0.107) between the 10th and 30th minute of which may result from the reduced capacity to generate force
the test swum at MLSS and decreased significantly at 102.5% to overcome drag.6 Similar to our study, some authors have
MLSS (p = 0.038). There was no significant difference on SR observed that the condition of [La] stability may be impor-
and SL at 10th (p = 0.898 and p = 0.736, respectively) and tant for the maintenance of the stroke parameters during
168.e4 J.G. Pelarigo et al. / Journal of Science and Medicine in Sport 14 (2011) 168.e1–168.e5

Table 1
Mean (SD) values of phases A, B, C and D (%) at 10th and 30th minute of the swim tests performed at MLSS and 102.5% MLSS, N = 12.
Phase A (%) Phase B (%) Phase C (%) Phase D (%)

10th min 30th min 10th min 30th min 10th min 30th min 10th min 30th min
MLSS 38.1 (6.2) 37.6 (6.9) 21.3 (3.1) 22.1 (3.8) 24.1 (3.1) 24.1 (3.2) 16.4 (0.8) 16.0 (1.3)
102.5% MLSS 38.3 (5.9) 36.9 (6.7) 21.7 (3.4) 22.9 (3.9)# 23.5 (2.9) 23.9 (2.5) 16.3 (1.4) 16.2 (1.3)
# p < 0.05 significantly different with the value at 10th min.

sub-maximal imposed pace tests lasting 30 min.9 Dekerle et our study) or not of the non-propulsive phases and IdC (as
al.9 reported stability in SL (from 2.55 to 2.46 m cycles−1 ) observed by Alberty et al.3 ) during imposed pace cannot be
at MLSS throughout time for all athletes, but a reduction explained only by the metabolic condition, but mainly by
in SL above MLSS for the athletes who could not main- the differences of exercise intensity between studies. It is
tain the pace up to 30 min (from 2.76 to 2.39 m cycles−1 ). important to note that in swimming the relationship between
In the former study, the athletes had similar aerobic per- energetic cost and speed is not linear,19 and therefore, a small
formance level (MLSS = 1.22 m s−1 ; S400 = 1.37 m s−1 ) of increment in the speed requires a much greater metabolic
the swimmers used in the present study. Otherwise, in a energy turnover, and probably, different stroke strategies to
recent study conducted in our laboratory16 with less experi- maintain the pace. In accordance with the studies conducted
enced athletes (MLSS = 1.13 m s−1 ; S400 = 1.30 m s−1 ), SL by Craig and Pendergast5 and confirmed by Alberty et al.,3
was reduced throughout the exercise at and above MLSS the magnitude of variation on the stroke parameters at fatigu-
(102.5%), with an interaction effect between intensity and ing conditions depends on the imposed intensity. In the study
time (greater decrease at 102.5% MLSS). Differences in tech- conducted by Alberty et al.,3 with the development of fatigue,
nical skill of the swimmers analyzed in these studies can the swimmers had less freedom to change the magnitude and
explain the different results. In fact, some studies have ver- distribution of the propulsive force within the stroke cycle,
ified at conditions of all-out distance trials at supramaximal leading to reduced non-propulsive phases and increased IdC
intensities, that swimmers with higher technical skill would to maintain the imposed pace. In our study, a greater freedom
be able to maintain greater SL for longer.17,18 Therefore, at to change the stroke strategies during the fatigue development
least for more experienced swimmers, MLSS represents an allowed the swimmers to change the relative duration of only
upper limit for maintenance of technical and physiological pull phase to maintain the swim speed. Finally, it is impor-
responses during tests lasting 30-min. tant to note that our swimmers did not swim until exhaustion,
The effect of fatigue on the stroke phases and arm coor- which may have also an influence in the magnitude of changes
dination has been analyzed during all-out distance trials4 in stroke strategies.
and imposed pace.3 Alberty et al.4 analyzing all-out 200-
m swim trials have verified increase of the relative duration
of the propulsive phases B and C. Since SR decreased, the 5. Conclusion
longer pull and push phases found in this study4 suggest
The changes in swimming technique (i.e., reduction in SL
that the swimmers spent more time in propulsive phases
and increase in SR and relative duration of pull phase) of
under fatigue, meaning that their capacity to generate propul-
well trained swimmers during long-distance imposed speed
sive force was compromised. Analyzing swim conditions of
tests performed at heavy intensity domain, occurs only at
imposed pace and lower exercise intensity performed until
condition of non-metabolic equilibrium. Therefore, MLSS
exhaustion (95% and 100% S400), Alberty et al.3 verified
seems to determine the upper boundary beyond which an
decrease in the duration of non-propulsive phases and main-
increase in the relative duration of pull phase is necessary to
tenance of the propulsive phases, with fatigue development.
maintain the swim speed for a long period of time.
Since in the study conducted by Alberty et al.3 the SR and
IdC were higher, the time allotted to propulsion per distance
unit was increased. Although the conditions analyzed in both 6. Practical implications
studies3,4 were different (all-out or imposed pace), it can be
verified that the fatigue determines an increase in the par- • For well trained swimmers performing continuous training
ticipation of the propulsive phases aiming to maintain the sessions, the MLSS seems to be the optimal swimming
propulsive impulse. technique speed.
In the present study, the exercise intensity (MLSS vs. • Different spatio-temporal and arm coordination modifica-
102.5%MLSS) and, consequently, the metabolic condition, tions in the crawl-stroke cycle can be found at the same
were important to determine different responses only in one exercise intensity domain.
propulsive phase. The values of non-propulsive phases and • The control of stroke phases and arm coordination dur-
IdC were not modified, irrespectively of the swim inten- ing long-distance swims may add important information
sity analyzed. Therefore, the maintenance (as observed in regarding stroke strategy.
 J.G. Pelarigo et al. / Journal of Science and Medicine in Sport 14 (2011) 168.e1–168.e5 168.e5

Acknowledgements 9. Dekerle J, Nesi X, Lefevre T, et al. Stroking parameters in front crawl

swimming and maximal lactate steady state speed. Int J Sports Med
2005;26:53–8.
This research was supported by grants from CNPq,
10. Dekerle J, Brickley G, Alberty M, et al. Characterizing the slope
FAPESP and FUNDUNESP. of the distance–time relationship in swimming. J Sci Med Sport
2010;13:365–70.
11. Keskinen KL, Komi PV. Stroking characteristics of front crawl swim-
References ming during exercise. J Appl Biomech 1993;9:219–26.
12. Pelayo P, Alberty M, Sidney M, et al. Aerobic potential, stroke param-
eters, and coordination in swimming front-crawl performance. Int J
1. Hollander AP, De Groot G, van Ingen Schenau GJ, et al. Measure-
Sports Physiol Perform 2007;2:347–59.
ment of active drag during crawl arm stroke swimming. J Sports Sci
13. Atkinson G, Reilly T. Circadian variation in sports performance. Sports
1986;4:21–30.
Med 1996;21:292–312.
2. Chollet D, Chalies S, Chatard JC. A new index of coordination for the
14. Seifert L, Schnitzler C, Aujouannet Y, et al. Comparison of subjective
crawl: description and usefulness. Int J Sports Med 2000;21:54–9.
and objective methods of determination of stroke phases to analyze arm
3. Alberty M, Sidney M, Pelayo P, et al. Stroking characteristics during
coordination in front-crawl. Port J Sport Sci 2006;6:92–4.
time to exhaustion tests. Med Sci Sports Exerc 2009;41:637–44.
15. Figueroa P, Barros RML. A flexible software for tracking of markers
4. Alberty M, Sidney M, Huot-Marchand F, et al. Intracyclic velocity
used in human motion analysis. Comput Methods Programs Biomed
variations and arm coordination during exhaustive exercise in front
2003;72:155–65.
crawl stroke. Int J Sports Med 2005;26:471–5.
16. Oliveira MFM. Metabolic and swim technique responses during the
5. Craig AB, Pendergast DR. Relationships of stroke rate, distance per
exercise at the speed corresponding to maximal lactate steady state
stroke and velocity in competitive swimming. Med Sci Sports Exerc
determined continuously and intermittently. Master Thesis. São Paulo
1979;1:278–83.
State University, Rio Claro, Brazil, 2008.
6. Craig AB, Skehan PL, Pawelczyk JA, et al. Velocity, stroke rate, and
17. Chollet D, Pelayo P, Delaplace C, et al. Stroking characteristic vari-
distance per stroke during elite swimming competition. Med Sci Sports
ations in the 100-m freestyle for male swimmers of differing skill.
Exerc 1985;17:625–34.
Percept Mot Skills 1997;85:167–77.
7. Toussaint HM, Carol A, Kranenborg H, et al. Effect of fatigue on
18. Vorontsov AR, Binevsky DA. Swimming speed, stroke rate and stroke
stroking characteristics in an arms-only 100-m front-crawl race. Med
length during maximal 100 in freestyle of boys 11–16 years of age. In:
Sci Sports Exerc 2006;38:1635–42.
Biomechanics and Medicine in Swimming IX. 2003.
8. Sanders RH. New analysis procedures for giving feedback to swimming
19. Capelli C, Pendergast DR, Termin B. Energetics of swimming at max-
coaches and swimmers. In: Gianikellis KE, Mason BR, Toussaint HM,
imal speeds. Eur J Appl Physiol 1998;78:385–93.
et al., editors. Scientific Proceedings – Applied Program – XXth Inter-
national Symposium on Biomechanics in Sports – Swimming. Caceres:
University of Extramedura; 2002.