Hand force symmetry during breaststroke swimming

The aim of this study was to analyze the hand force symmetry during the breaststroke and its relationship with swimmers’ performance. Seventeen breaststroke and/or medley specialists participated (12 men and 5 women, 19.5 ± 5.2 years and average performance of 73.4 ± 7% of the 50 m breaststroke world record). Each swimmer performed three repetitions of 25 m breaststroke swimming at maximal speed. Pressure sensors from Aquanex acquisition system were placed in both swimmers’ hands and Mean Force (Fmean) and Maximal Force (Fmax) were measured. The symmetry index proposed by Sanders was calculated and the time of a 50-m breaststroke trial at maximum speed (T50m) was used as a performance indicator. The variables were compared between hands using tests for dependent samples, and the relationship between variables were investigated using Spearman correlation test (p<0.05). The Fmean applied was 47.9 ± 16.7 N and 47.9 ± 14.5 N for right and left hands, respectively. The Fmax corresponded to 120.7 ± 43.6 N and 112.8 ± 35.7 N for right and left hand, respectively. No signifi cant differences were observed for none of the variables between right and left hands. Analyzing the subjects individually, it was possible to observe asymmetries levels up to 30.6% for Fmean and 35.9% for Fmax, however the relationship between symmetry indexes and the 50 m breaststroke performance was not statistically signifi cant.


Introduction
Hand force symmetry during breaststroke swimming Rafaela Grübel WERLANG * Suzana Matheus PEREIRA * Caroline RUSCHEL * Gustavo Soares PEREIRA * Ana Paula Moratelli PRADO * Gustavo Ricardo SCHÜTZ ** Helio ROESLER * Th e goal of competitive swimmers is to travel the swimming race distance in the shortest time possible, with the highest average speed, considering the traveled distance, as well as the total time to complete the race, always respecting its specifi c rules 1 .To this goal, it is necessary to improve technical and physical aspects aiming to increase the swimmer's propulsive capacity and reducing the water resistance.
Th erefore, it seems obvious that the propulsive force is an important determinant of swimming performance 2 , and its assessment may provide indicatives of the eff ectiveness of the swimming technique 3 .Estimation of propulsive force has been investigated by sports biomechanical analysts throughout the years by several methods, as tethered swimming 2,[4][5] , tridimensional video analysis 6 , computational fl uid dynamics 7 and with pressure sensors attached in swimmer's hands 8-10.When analyzing propulsive force production, researchers are interested in the incidence of asymmetries between the right and left limbs, and its relationship with the performance 5,11 .Even though a certain level of asymmetry is acceptable/ tolerable, due to diff erences inherent to human body 12 , the bilateral diff erence in force production during swimming activities may be a limiting factor for optimal performance 13 .According to Sanders et al. 14 , in order to accomplish a high performance level, both arms and legs have to contribute in an http://dx.doi.org/10.11606/1807-5509201700010051optimal manner to maximize propulsion and reduce swimming drag.
The studies presented in literature analyzed the propulsive asymmetry during the crawl stroke 5,11,15 .Marinho et al. 7 investigated the relationship between propulsive force and swimming performance of competitive swimmers, however, the authors did not analyze the bilateral symmetry.Data on the diff erence between limbs, with regards to the force production in symmetrical swimming such as breaststroke, are scarce.
Jaszczak and Zatón 16 pointed out that the symmetry of movements, typical of the breaststroke, makes it an interesting tool in the treatment of postural problems in children's.However, Jaszczak 12

Subjects
showed asymmetries between the upper limbs during a simulation of the breaststroke swimming in an ergometer.Years later, in studies carried out in the swimming pool with youth and adult amateur swimmers, authors from the same research group also verifi ed the occurrence of force asymmetries in the stroke 9,16 .We believe it is necessary to investigate if there are asymmetries in real swimming conditions performed by breaststroke specialists, as according to Maglischo 17 , swimmers seems to have an arm more effective than the other, which could influence the performance.In this sense, the present study aimed to analyze the hand force symmetry during the breaststroke and its relationship with swimmers' performance.
Seventeen subjects (12 men and 5 women), specialists in breaststroke and individual medley races, participated in this study.Swimmers were from teams of the region of Grande Florianopolis, Santa Catarina, and were regularly participating in continental, national and state competitions.All

Instruments and variables analysis
Swimmers' anthropometric measures (body mass and height) were obtained with a digital scale (Techline, model BAL-150 PA) and a stadiometer with accuracy of 0.01m (Sanny, professional model).Subjects responded to anamneses with information regarding laterality, training duration, frequency and the most relevant performance in swimming competitions.
Dynamometric data were collected using the Aquanex system (Swimming Technology Research, United States of America), composed by two pressure * R e l a t i o n s h i p , i n percentage points, between the performance in the 50 m breaststroke at maximum speed and the world's record on the 25 m swimming pool, respecting the men and women categories.
subjects were informed about the aims, procedures and possible risks associated with the present study and gave their informed consent prior to enrolment in the study.Th e present study was approved by the Ethics Committee of the institution.Subjects' characteristics are depicted in TABLE 1. sensors, one A/D converser and a software for data acquisition.Th e sensors allowed estimating the force applied by the arms during the stroke with an error of 0.2%.Th e system was connected with a portable computer and the data acquisition frequency was 100 Hz.Th e following variables were analyzed: • Mean Force (F med ): mean force value, expressed in N, measured in the swimmer's hand during the propulsive phase of the arm stroke.
• Maximum Force (F max ): maximum force value, expressed in N, measured in the swimmer's hand

Data analyses and processing
Where: F med R: mean force applied by the swimmer's right hand; F med L: mean force applied by the swimmer's left hand; max(R,L): maximum value obtained for the mean force, irrespectively of the side (right or left).
• Symmetry index for F max (IS max %): percentage diff erence between the maximum force measured in the swimmer's left and right hand.Th e symmetry index was calculated as follows 3 : Where: F max R: maximum force applied by the swimmer's right hand; F max L: maximum force applied by the swimmer's left hand; max(R,L): maximum value obtained for the maximum force, irrespectively of the side (right or left).
Th e total time to complete the breaststroke 50 m test (T 50m ) at maximum speed was recorded with a digital chronometer (Casio, model HS-3V), with a precision of 0.01s.

Data Collection
Data were acquired in a 28º C heated swimming pool with 25 m width, before the swimmer's regular training session.After the anamneses, each subject warmed up for approximately 10 minutes according to the coaches' instructions.Th en, the pressure sensors were positioned between the middle and ring fi ngers of both hands and their cables were fi xed with elastics in the forearms, shoulders and hip.Subjects were allowed to familiarize with Data acquired by anamnesis were digitalized in a Microsoft Offi ce Excel 2010 spreadsheet (Microsoft Inc., United States of America) and were analyzed with using a descriptive statistical procedure (mean, standard deviation and percentage).Th e analyzis of dynamometric data were performed using Matlab For the statistical analyzes, the mean value of 18 stroke cycles (six stroke cycles in each 25 m attempt) was used.Regarding the symmetry index, IS% positive values indicate that the absolute value for the variable (F max or F mean ) was greater for the right hand, whereas IS% negative values indicate that the absolute values for the variable (F max or F mean ) was greater for the left hand.Th e mean and standard deviation for IS med % and IS max % were calculated considering the module of individual values.
Data normality was verifi ed by the Shapiro-Wilk test.When the normal distribution was assumed, comparison between the right and left hands was performed using a Student t test; in cases where normal distribution was not assumed, data were compared with the Wilcoxon test.Th e relationship between independent variables (F med , F max , IS med % the equipment as they were instructed to freely swimming with it before data acquisition.
After this, each swimmer performed three repetitions of 25 m swimming breaststroke at maximum speed, with an interval of fi ve minutes between trials.Then, swimmers performed an active period of recovery, according to the coaches' instructions.Lastly, subjects were instructed to perform a 50-m breaststroke trial at maximum speed in order to obtain the T 50m , used in this study as a performance indicator.
during the propulsive phase of the arm stroke.
• Symmetry index for F med (IS med %): percentage diff erence between the mean force measured in the swimmer's left and right hand.Th e symmetry index was calculated as follows 3 :

Results
The main results showed no significant statistical differences for the mean force, as well as for the maximum force, between the right and left hands.However, when inspecting data individually (TABLE 3), it was observed that some individuals presented an asymmetry index higher than 20% both for the mean and the maximum forces.The red rectangle i n d i c a t e s t h e s i x stroke cycles selected for analysis in this trial, excluding the underwater phase and the fi rst stroke cycle.TABLE 2 shows the mean values, standard deviation and confi dence interval (95%) for F med and F max , as well as the mean diff erence and p-value for right and left hand comparison.

Variables
and IS max %) and performance (T 50m ) was investigated through the Spearman correlation coeffi cient.All data were analyzed with the SPSS 20.0 for Windows package software (SPSS Inc., United States of America).Th e level of signifi cance adopted was p<0.05.

Discussion
Th e present study analyzed the maximum force and the mean force of the right and left hands during the breaststroke using a pressure sensor system.Additionally, swimming performance, measured as the total time to complete one trial of 50m at maximum speed, was correlated with the mean force values and maximum force values applied by each hand, and with the hand force symmetry index, calculated from these values.

Th e maximum force values for the breaststroke
The mean values for IS med % and IS max % were 8.7 ± 7.7% e 11.4 ± 9.6%, respectively.Regarding the swimmers' performance, mean time for the 50 m (T 50m ) was 36.06 ± 3.44s.TABLE 4 shows the results for the correlation between T 50m and the other variables of the study.
observed by Morouço et al. 2 are fi ve times greater as compared with those presented in this study (513.0 ± 153.9 N); and the mean force values are more than the double as observed herein (115.6 ± 30.5 N).However, it is important to highlight that those authors measured the force using the tethered swimming method, as such, these values represent the total force produced by the strokes of the arms and legs, without any distinction between the right and left limbs.Th e analysis of force production by swimmers during breaststroke in ergometers showed maximum force values of 159.9 ± 37.6 N and 113.1 ± 15.5 N for men and women, respectively 12 .According to the authors, the symmetry analysis revealed mean diff erences of 13.9 ± 4.9 N between hands for men and of 8.7 ± 3.1 N for women, but there is no mention if the diff erence between limbs was statistically signifi cant.
In the present study it was not possible to observe statistically signifi cant diff erences in force application between right and left hands during breaststroke swimming.However, individual data presented in TABLE 3 point to a high variability within the group.Th ese variations, which could be due to factors as gender, physical characteristics (e.g.maximal strength and the length of the segments involved in propulsion) and technical aspects (e.g.stroke mechanics and performance level), among others, may have "masked" possible diff erences of force between right and left hands.
A great variability within group was also observed for the symmetry indexes.Although the average IS% was 8.7% for F med and 11.4% for F max , individual values were up to 30.6% for IS med % and 35.9% for IS max %.In the current literature, asymmetries up to 10% between limbs are considered acceptable for activities such as human walk 18 .Th ese values are also adopted for swimming studies that aim to investigate coordination analysis, force applied in swimming ergometer and tethered swimming [19][20][21][22] .In this study, six of the 17 analyzed swimmers presented values higher than 10% for the IS med %, and seven for the IS max %.
The occurrence of force asymmetry has been showed in crawl stoke 4,15,22 , and it is generally attributed to the unilateral alternation of arm movements and to the breathing mechanics 14,22 .However, substantial asymmetries between limbs are also found in strokes with bilateral characteristics, such as breaststroke and butterfly, even in elite swimmers 13 .
Recently, studies involving breaststroke swimmers with diff erent technical levels sought to understand the magnitude of asymmetries, its possible causes and also the consequences for performance 9,12,16,23 .Jaszczak 12 compared the force asymmetry between crawl stroke and breaststroke performed in a ergometer by swimmers of both genders and with a "moderate" performance level (as defi ned by the author).Th e mean values of asymmetry found for the breaststroke was of 8% for women and 9% for men.For the crawl stroke, average values were of 14% and 12% for female and male subjects, respectively.For the male swimmers, there were no signifi cant diff erences between swimming strokes (crawl or breaststroke).It is important to highlight that there are technical differences between swimming in an ergometer and in a real swimming scenario, and this could cause some confusion in the interpretation of the values, which were probably underestimated for the ergometer.
Aiming to analyze the hand force symmetry during breaststroke in situations with a higher ecological validity, Jaszczak 9 and Jaszczak and Zatón 16 investigated maximal repetitions of 15 m in a swimming pool, using pressure sensors attached to the swimmers' hands.In both studies, the authors analyzed the effect of the leg kick technique ("correct" -symmetrical movements, "incorrect" -asymmetrical movements) in hand force symmetry, in college students 9 and 11-year old boys 16 with "moderate" performance level, and with no expertise in swimming competition.For both groups, the upper limb symmetry was registered during isolated movement of arms (using a pull-buoy between legs) and during breaststroke (whole stroke).In both groups, swimmers that performed asymmetrical movement of legs also presented an increase of asymmetries in upper limbs, when the isolated upper arm condition was compared to the whole stroke condition: 12% versus 14% for college students 9 and ~25% versus ~30% for 11-year old boys 16 .According to the authors, the results point that the increase of asymmetries in upper limbs may be a compensatory response to asymmetrical movements of the lower limbs.Sanders 24 corroborates with the previous hypothesis, indicating that in bilateral strokes, diff erences in force production in the lower limbs could cause body misalignment and, consequently a compensation in the hand's trajectory; this compensation, in turn causes a technical asymmetry that could become a habit, leading the swimmer to produce more force in one side of the body than in another.
Asymmetrical movements of upper and/or lower limbs are not only present in swimmers with low and moderate levels of performance.Sanders et al. 23 , in a recent case study with an elite female swimmer specialist in breaststroke, observed asymmetries in the kinematic path of both lower and upper limbs during 100-m trials, which lead to the occurrence of a momentum in the antero-posterior axis (vertical in relation to the bottom of the swimming pool).Sanders et al. 14 point the possibility that the technical asymmetries could reinforce and perpetuate other types of asymmetries, for example, muscular imbalance.In the case analyzed by Sanders et al. 23 , the authors attribute part of magnitude of the technical asymmetries to the fact that the muscles from the right side of the swimmer were stronger than the ones from the left side, both for fl exion and internal rotation.Consequently, the stroke pull is faster for the right side, producing higher propulsive force, which causes the momentum in the antero-posterior axis.
Sanders et al. 14 state that, although the breaststroke is a bilateral demand, and the activity by itself stimulates symmetry, it is not possible to neglect the eff ect of other daily and sportive activities of asymmetrical nature.In this context, it is important to highlight that swimmers, in general, train a huge amount of distances swimming crawl stroke, even being specialists in other strokes.
In relation with the asymmetry direction (IS positive or negative, depending on the side that the swimmer applies more force), it seems important to consider aspects like laterality.In the present study, the laterality was determined by questioning the preferential hand of the swimmers for performing daily actions.Fifteen of the 17 swimmers chose the preferential side as the right one.In this way, the individual data (TABLE 2) do not seem to present any relation between the laterality -in the way it was measured -and the direction of asymmetry.It is important to consider that other strategies could be used to the determination of laterality (e.g. the preferential side for breathing in crawl stroke, or the analysis of the unilateral strength of muscles involved in propulsion).
Th e data found on the present study corroborates with findings in the literature in respect to the presence of asymmetries in breaststroke.Considering the information previously debated, it is possible that the IS% values during the arm stroke might be infl uenced by some associated factors, such as the kinematical asymmetry of upper and lower limbs, the asymmetry of force production in the lower limbs, and the asymmetries of strength and fl exibility of the main muscular groups involved in propulsion.
Since the causes of symmetry were already discussed, it is interesting to analyze its relationship with swimming performance.In bilateral strokes, the asymmetry in force production could cause a rotation in the antrero-posterior axis, increasing active drag 13,24 .Th e imbalanced force contribution could also compromise the propulsive effi ciency 24 .Although it is expected that these aspects may directly affect performance, a lack of studies regarding this topic is found.It is not clear if the presence of asymmetries of diff erent natures is necessarily related with a decrease in performance 24 .
According to Morouço et al. 22 , it is not yet defined if the asymmetries of force, velocity, trajectory and coordination reported by the literature for the crawl stroke could alter the optimal function, or simply are inside a normal boundary of variation.In order to contribute with the discussion, the authors analyzed the relationship between the symmetry index of crawl stroke on tethered swimming and the total time in a 50 m performance, in male swimmers of diff erent performance levels.Th e authors found values of IS% between 3.3% and 48.5%, with the majority of the swimmers presenting values above 10%.Interestingly, the analysis of the relation between force and performance, when controlled by the IS% magnitude, showed that higher values of asymmetry did not lead to worse performances.Based on that, the authors assumed that, until a certain limit, the asymmetry may not be a critical factor to achieve maximal swimming velocities.
On the present study, the analysis of relationship between IS% and swimming performance (T 50m ) resulted in a moderate, non-statistically signifi cant correlation coefficient for mean force (ϱ=0.424,p=0.090), and a weak, non-statistically signifi cant correlation coeffi cient for maximal force (ϱ=0.159,p=0.542).In this way, the upper limb asymmetry seems to not infl uence the breaststroke 50 m performance, considering the way the variables were measured.It is important to emphasize that, although the asymmetries may increase the active drag and reduce the propulsion, it is possible that the subjects analyzed -aligned with other scientifi c fi ndings -adopted strategies of compensation, reducing its eff ects in the global stroke performance.Additionally, it is possible that the 50 m performance is infl uenced in a greater scale by other factors, such as the leg kick effi ciency (most propulsive element in the stroke 17 ), the coordination between lower and upper limbs, and the individual capacity of energy consumption and resistance to fatigue.Th e force applied by the upper limb, by itself, seems to have low relevance in the 50 m performance, based on the low correlation coeffi cients presented in TABLE 4.
When the total force produced by the swimmer (using upper and lower limbs propulsion) is considered, its relationship with swimming velocity is substantially stronger (r-0,94) 2 .Also, according with Havriluk 10 , for all competitive strokes, the technique adopted by the swimmer to reduce the active drag appears to have a better eff ect on performance, compared with the quantity of force applied.Th e author also empathize the necessity of analyzing the way the swimmers apply the force during the entire stroke cycle.
According with Sanders at al. 23 , the severity of asymmetries and the impact that they have on performance vary from swimmer to swimmer.More information is needed to assist the observation of the eff ectiveness of interventions to correct the issue.As reported by the authors, some asymmetries may not aff ect performance and, in this case, drills performed to correct them may negatively interfere in the training program of the swimmer.In contrast, if asymmetries of force, fl exibility or technique would produce a momentum that could cause bad alignment and increased drag, it may be valid to consider strategies to correct them.It is important to consider that it could be necessary to correct asymmetries that although not bring signifi cant reduction in performance, may expose the swimmer to a higher incidence of musculoskeletal issues.Th e need of force or/ and kinematical compensations in function of asymmetries, combined with the high number of repetitions and the training overload, could lead to organic adaptations, causing muscular imbalance and increasing the risk of injuries 25 .
Th is exploratory study allowed to analyze the hand force symmetry during the breaststroke and, although the aim was not to examine the eventual causes of asymmetries, it sought to contribute in the whole understating of the relation between the asymmetries and swimming performance.Th e results showed that, although the diff erences of mean and maximum forces between left and right hands are not statistically signifi cant for the analyzed swimmers, individual asymmetries could reach 35%.However, it seems not to exist a relation between this asymmetry and the total time of a 50 m performance.Some limitations are recognized in regard of the measurement method used on the study for estimation of hand force, since it provides an indirect measure of force, and does not represent propulsive force by its entirety.However, the instrument allowed the swimmers to perform freely during the analysis, increasing the ecological validity in comparison with tethered swimming.Finally, the authors suggest that further studies analyze the relationship between asymmetries in hand force and other performance indicators, such as stroke frequency and stroke length.Additionally, it could be relevant to investigate the relationship between hand force asymmetry and/or swimming technique with other types of asymmetries (e.g.morphological and/or functional).
7.1 environment (Mathworks Inc., United States of America), with a processing routine composed by the following steps: (1) fi ltering (low-pass 3 rd order Butterworth digital fi lter with 10 Hz cutoff frequency); (2) visual inspection of curves for the right and left hands, separately, in order to manually select six stroke cycles for each attempt (discarding the underwater phase and the fi rst stroke cycle); (3) extraction of the variables of the study and (4) exportation of the variables values of each curve in *.txt format.FIGURE 1 presents an example of force versus time curves of the right and left hands obtained during one 25-m breaststroke trial performed by one of the participants of the study, with the indication of the six stroke cycles analyzed.

Force
versus time curves of right hand (gray solid line) and left hand (dashed black line) obtained during a 25-m breaststroke trial performed by one of the participants of the study.

TABLE 1 -
Mean (standard deviation) of the swimmers' characteristics from the present study.

TABLE 2 -
Mean ± standard deviation (95% confi dence interval) for mean force (F med ) and for maximum force (F max ), and the results for between right and left hand comparison.
*: P-value obtained by Student t test; ϯ : P-value obtained by Wilcoxon test.

TABLE 3 -TABLE 4 -
Spearman coeffi cient correlation (ρ) obtained from the correlation analysis between the swimmer's performance indicator (T50m) and the other variables of the study.Mean ± standard deviation of individual values and symmetry index (IS%) for mean force (F med ) and maximum force (F max ).