THE EFFECT OF ACTIVE RECOVERY , COLD WATER IMMERSION AND PASSIVE RECOVERY ON SUBSEQUENT KNEE EXTENSION AND FLEXION STRENGTH

BACKGROUND: Recovery is an important aspect of every physical activity. Many athletes train hard without giving their body time to recover which can lead to overreaching, burnout or poor performance. Currently cold-water immersion recovery and active recovery have emerged as some of the most popular interventions enabling faster recovery. OBJECTIVE: To assess the eff ect of three kinds of recovery (active recovery, cold water immersion, passive recovery) on medium-term knee strength in the extension and fl exion. METHODS: Fourteen athletes at the age of 26.6 ± 4.4 years performed, in a random cross-over design, 3 sessions with 3 repeated medium-term isokinetic tests. The eff ect of active recovery, passive rest and cold water immersion were assessed by 3 × 3 (time × recovery) repeated-measure ANOVA, respectively. The dependent variables were – peak torque, total work and average power. RESULTS: We found signifi cantly lower absolute diff erences between the fi rst and third trial in knee extension for peak torque after the active recovery (↑ 0.9 N × m) than after the cold water immersion (↓ 14.6 N × m) or the passive recovery (↓ 13.9 N × m). The decrease of the average power was signifi cantly lower diff erences after the active recovery (↓ 5 W) than after the cold water immersion (↓ 23.7 W) or passive recovery (↓ 25.9 W). The changes in total work were not signifi cant. We did not found any changes in the isokinetic strength for the knee fl exors after diff erent kinds of recovery. Maximal heart rate (HR max ) was signifi cantly higher during the active recovery than during the cold water immersion and the passive recovery (173 ± 14, 166 ± 14 and 167 ± 14 rpm). We have found signifi cant diff erences in the average heart rates (HR avg ) during active recovery, cold water immersion and passive recovery (124 ± 8, 97 ± 9 and 107 ± 12 rpm). CONCLUSION: We found the positive eff ect of the active recovery on the subsequent medium-term performance for knee extension. That was the only method which showed lower decrease of knee strength in extension in comparison with passive recovery and cold water immersion. We have found the signifi cant diff erences of heart rate which was recovery dependent.


INTRODUCTION
It is generally accepted that good recovery enables better performance and decreases the number of injuries in athletes.Recovery is an important aspect of any physical activity.Many athletes train extremely hard without giving their body time to recover which can lead to overreaching, burnout or poor performance (Cochrane, 2004).According to Hargreaves and Spriet (2006) recovery is simply defi ned as a biological process of removing fatigue.Astrand, Rodahl, Dahl, and Stromme (2003) defi ned fatigue as a complex of physiological and psychological process that leads to pre-load condition.It is a routine for athletes to employ post exercise strategies or practices in a bid to speed up recovery (Bleakley & Davison, 2010).
Recently, cold-water immersion recovery has emerged as one of the most popular interventions.
Despite its popularity, the evidence from clinical trials remains ambiguous, and there is also little evidence to design an optimal treatment protocol (Cochrane, 2004).Anecdotal reports from coaches, medical personnel and athletes suggest that this method has positive eff ects on subsequent performance (Enwemeka et al., 2002;Myrer, Draper, & Durrant, 1994).There is general consensus that the application of cold water immersion decreases skin, subcutaneous and muscle temperature.Subsequently, the decrease in tissue temperature is thought to stimulate the cutaneous receptors causing vasoconstriction which decreases the swelling and possible infl ammatory processes by slowing the metabolism and production of metabolites and thereby limiting the degree of the injury (Enwemeka et al., 2002).We can quantify the degree of the recovery by such indicators as reduction of HR, respiratory rate and ventilation parameters, restoration of energy reserves and ion balance, removal of waste products of metabolism, decrease of muscle tension and reduction of the activity of the central nervous system (Bleakley & Davison, 2010).Various processes of recovery improve and accelerate these indicators and decrease fatigue.In our study, we focused on the application of cold water immersion and passive and active recovery.
Especially active recovery and cold water immersion off er very useful and low-cost recovery procedure.All the activities such as walking, running or cycling can be used as active recovery.The start of recovery processes while using active procedures needs to be kept at a moderate intensity of the selected exercises (50-65% maximum HR or 35% VO 2 max) which usually takes around 20 min.(Cochrane, 2004;McArdle, Katch, & Katch, 2001).The idea of this kind of recovery is to accelerate the supply of oxygen into the tissue and lactate and other muscle metabolites removal (Barnett, 2006;Draper, Bird, Coleman, & Hodgson, 2006).Therefore the intensity of the exercise has to be set at certain threshold level.
Hydrotherapy belongs to the popular and widely used recovery methods.Water immersion reduces muscle oedema and increases heart rate.It also increases blood fl ow which further speeds up the removal of waste products after muscle work (Bleakley & Davison, 2010;Ingram, Dawson, Goodman, Wallman, & Beilby, 2009).Interestingly, there is not such decrease of performance after cold water immersion application in combination with warm water bath compared to passive recovery (Viitasalo, Niemela, & Kaappola, 1995).
Finally passive recovery in the sitting or lying position without any other activities and any other aids belongs to one of the most easily performed recovery.
By all means, diff erent types of recovery are dependent on the sportʼs loads.Therefore the aim of the study was to clarify the infl uence of diff erent recovery process (active and passive recovery, cold water immersion) on the subsequent medium-term knee strength during extension and fl exion.

Participants
The group of participants consisted of 14 men (mean ± standard deviation) in the age of 26.6 ± 4.4 years, body height of 1.80 ± 0.06 m and body weight 74.6 ± 5.2 kg (fat 11.5 ± 1.9 %, lean body mass 65.9 ± 4.5 kg), determined by bioimpedance method (Bunc, 1995).We chose students of physical education who were somehow active in sports.None of the participants stated anything that could infl uence the results.During the last two years, none of the students suff ered from any injuries or other pathologies of the lower limbs.Meas-urements were performed only on the dominant lower extremity.The dominant leg was defi ned as the leg that the participant uses as a take off leg for the long jump (Miyaguchi & Demura, 2010).Twelve of the 14 participants identifi ed right leg as their dominant leg.
The study received approval from the institutions Ethics Advisory Committee of Charles University.

Experimental procedure Isokinetic knee strength measurement
The measurements were performed under constant laboratory conditions.Measurement of the concentric strength in knee extension and fl exion was realized on the isokinetic dynamometer Cybex Humac Norm (Cybex NORM ®, Humac, CA, USA).
We measured isokinetic knee strength in extension and fl exion with three trials in one day defi ned by constant angular velocity of 150° × s -1 (50 repetitions) with a duration of 70-80 s. Between the trials (15 min.rest with recoveries) of the isokinetic strength measurement we applied the recoveries (passive recovery, active recovery and cold water immersion).The strength parameters were peak torque, total work and average power.
The loaded limb was fi xed above the knee strap to stabilize it.The chair and dynamometer were set to the settings of dynamometer for knee extension and fl exion measurement.The knee axis was parallel with the dynamometer arm.Range of the knee motion was approximately 90°.
During the isokinetic strength measurements we motivated the students verbally by audible instructions.The subjects were also motivated visually by seeing their immediate results.

Recovery strategies
Immediately after isokinetic knee strength we applied passive recovery in one group, active recovery in the second group and cold water immersion in the third group.The other day, the groups were exchanged so that everyone would undergo all the kinds of recovery.The schedule and times were as follows: 1.The passive recovery was performed in the sitting position on the chair of dynamometer for 15 min.at roomʼs temperature 22 ± 1° C. 2. As active recovery we used walking for 10 minutes on the treadmill of 5.5 km × h -1 speed and a gradient -was depended on the HR to achieve 60-65% of individual HR max .HR max was determined during the vita maxima test on a treadmill ergometer.Individuals HR max reached 194 ± 10 rpm. 3.As cold water therapy the participants were immersed up to their hips in the cold bath (13 ± 1° C) for 3 × 2, 5 min.with 2 min.out of the bath, repeated twice as intermittent protocol.In the room there was average air temperature 22 ± 1° C while the par-ticipants were in the sitting position (Vaile, Halson, Gill, & Dawson, 2008a).The water temperature was determined according to the protocol of Malkinson, Martin, Simper, and Cooper (1981).

Experimental design
This was a randomized cross-over experiment design study.Each individual was at the same time in the group with the experimental factors (active recovery, cold water) in the control group (passive recovery).The dependent variables were peak torque, total work and average power.
The signifi cance of the results was verifi ed by 3 × 3 (time × recovery) analysis of variance with repeated measures.The signifi cance of the results was assessed by the coeffi cient partial η², which represents percentage of the total variance explained as independent variable.Statistical signifi cance of diff erences was in the level p ≤ 0.05.The SPSS Statistical program for Windows (version 18.0) was used to evaluate the results of the study.

RESULTS
We observed signifi cantly smaller absolute changes between fi rst and third trial for knee extension, for peak torque after active recovery (↑ 0.9 N × m) compared with cold water (↓ 14.6 N × m) and passive recovery (↓ 13.9N × m).We assessed signifi cant decreases of average power after cold water immersion (↓ 23.7 W) and after passive recovery (↓ 25.9 W).After active recovery (↓ 5 W) we did not notice a signifi cant diff erence of average power.The changes of total work for knee extension and fl exion were insignifi cant.The percentage changes of peak torque, total work and average power between fi rst and third trial are in the TABLE 1.The decrease of strength performance for knee extension is shown in Fig. 1-3 and for knee fl exion in Fig. 4-6.The decrease of dependent variables was recorded (Fig. 1-6) after recoveries.After active recovery and cold water immersion we found no signifi cant diff erences in strength performance (peak torque, total work, average power).The signifi cant diff erences of knee extension was assessed for peak torque F 2, 26 = 9.0 (p = 0.00), η² = 0.41 for both recovery F 2, 26 = 3.5 (p = 0.04), η² = 0.21 and for average power F 2, 26 = 6.8 (p = 0.00), η² = 0.34 for time factor.
For knee extension recovery was in the interaction with the order of measurements statistically and substantively signifi cant for peak torque, where was F 2, 52 = 5.9 (p = 0.00), η² = 0.31 and the average power was F 2, 52 = 2.7 (p = 0.04), η² = 0.10.Recovery in the interaction with the time had no signifi cant eff ects on the overall work of knee extension, where F 2, 52 = 1.4 (p = 0.25), η² = 0.10.Fig. 1-3 for knee fl exion there is not obvious decrease of the peak torque, total work or average power after three types of recovery.The three types of recovery did not have any eff ect on strength indicators.The strength diff erences for knee fl exion occurred in the interval of the standard error of measurement (SEM).We did not fi nd signifi cant diff erences of knee fl exion measurement after recovery.

DISCUSSION
We found only minimal changes in repeated strength performance after active recovery and cold water immersion compared to passive recovery.A signifi cant eff ect of active recovery was observed only in case of repeated

Fig. 2
Means and standard deviation of total work for knee extension for subsequent measurement for three types of recovery (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion)

Fig. 4
Means and standard deviation of peak torque for knee fl exion for subsequent measurement for three types of recovery (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion) knee extension for peak torque and average power.For total work, the eff ect of recovery was not signifi cant.For knee fl exion we found signifi cant changes for repeated measurements in active recovery and cold water immersion compared to passive recovery.Signifi cant eff ect of active recovery and cold water immersion was observed only for knee extension for repeated measurement on peak torque and average power.
However, no signifi cant diff erences in knee strength during fl exion were observed among any of the cold water immersion and active recovery protocols.
Walking (65% intensity of maximal HR) as active recovery was chosen because it doesn't require such complex coordination and it is a widely accessible activity.Active recovery, for example running or cycling, could have similar eff ects as walking.The intensity of certain

Fig. 5
Means and standard deviation of total work for knee extension for subsequent measurement for three types of recovery (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion)

Fig. 6
Means and standard deviation of average power for knee fl exion for subsequent measurement for three types of recovery (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion) activity is more important than the kind of this activity itself (Cochrane, 2004).Walking (maximum intensity of 65% HR max ) has been already used in the study of Baláš, Chovan, and Martin (2010).The authors used active recovery after repeated climbing until exhaustion and stated 14% decrease of performance in comparison with 41% after passive recovery.However, the study was focused only on the upper extremities after the climbing performance.In our case while focusing on lower extremities, we noticed decreased peak torque (0.5%) after walking about the same intensity and the decrease of the total work and the average power (by 1.3% and 2.0%).We found that walking is an adequate type of recovery for repeated isokinetic strength activity of the lower limbs.In the other studies (Heyman, De Geus, Mertens, & Meeusen, 2009;Watts, Daggett, Gallagher, Wilkins, 2000) the bicycle ergometer seems to provide the activity for acceleration the recovery process compared to the passive one.Active recovery plays an important role when we want to get the optimal decrease of fatigue symptoms.We have to keep in mind that the recovery we chose should match with the intensity of loading processes (Bielik, 2010;Vanderthommen, Makrof, & Demoulin, 2010).

Fig. 7
Means and standard deviation of maximal (HR max ) and average heart rates (HR avg ) during knee strength measurement and recovery application (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion) Legend: * Signifi cant diff erence of HR max PAS and HR max ACT, p = 0.02 + Signifi cant diff erence of HR avg PAS and HR avg ACT, p = 0.00 ++ Signifi cant diff erence of HR avg PAS and HR avg CWI, p = 0.03 Vanderthommen, Makrof, and Demoulin (2010) reported the insignifi cant increase of maximal power after active recovery (intensity was 50% HR max , for 25 min.)on bicycle ergometer then after passive rest (105.3 ± 12.2 vs. 99.1 ± 10.7%).Nevertheless, we could not compare specific cycling performance because we evaluated the extension and fl exion separately.All the same Bielik (2010) applied longer recovery time (20 min.),after which he recorded smaller decline in maximal and average power after active recovery process than after passive rest (970.2 ± 68.9 vs. 875.5 ± 56.2 W, p < 0.05 and 746.1 ± 47.0 vs. 678.4± 45.2 W).The eff ect of recovery on the fatigue (active vs. passive recovery) was not signifi cant (35.2 ± 7.7 vs. 33.6 ± 8.4%).The concentration of blood lactate was signifi cantly lower after active recovery than the passive rest (7.4 ± 3.9 vs. 13.3 ± 2.9 mmol × l -1 , p < 0.01) (Bielik, 2010).In our case we confi rmed the signifi cant eff ect of recovery on peak torque and average power for extensors (p = 0.001).We found signifi cant changes for the repeated measurements of knee strength fl exor that was not infl uenced by recovery.The training process is very often neglected and knee fl exors are often injured due to inadequate muscle preparation (Greig & Siegler, 2009).The eff ect of active recovery on blood lactate after exercise is well documented but the exact mechanism of lactate fatigue in a repeated performance needs to be discussed further (Barnett, 2006;Dodd & Alvar, 2007;Watts et al., 2000).
Cold water immersion has the potential to minimize performance reduction of knee extension and fl exion.The eff ect of cold water immersion showed the smallest decrease by % of peak torque, total work and average power of knee fl exion strength.Cold water immersion was changed into higher room temperature repeatedly not to hold up the recovery processes.In our case cold water immersion was only partial diving (to the hip) for the lower extremities to avoid the negative eff ects of the whole body cold bathing to the urinary tract.The individuals had to stand in a barrel submereged to their hips.The water temperature was increased up to 15° C after one submersion procedure.In the other studies (Parouty et al., 2010;Peiff er et al., 2010), they used a water temperature of 14° C and the subjects were to the neck in cold water for 5 min.and 5 min.in room temperature in a sitting position, there was a signifi cant eff ect on repetitive strength performance in the group of sprint swimmers.Water temperature and the duration of submersion seemed to be optimal in our study too.Similar positive eff ect of water temperature and duration of cold water immersion was confi rmed in a group of climbers when applied only on the forearm (Baláš et al., 2010).
There are certain recovering processes occurring after the application of such cold water -local vasoconstriction of blood vessels, increase local blood fl ow avoiding cell necrosis, swelling, slowing of cellular metabolism, slowing of nerve transmission and means of reducing pain by proprioceptors of loaded muscles.Whether performance recovery after cold water immersion is primarily linked to the return of fl uid from muscles to the blood requires further investigation (Wilcock, Cronin, & Hing, 2006).The positive eff ect of hydrotherapy can be also caused by hydrostatic pressure, which aff ects blood circulation.Hydrostatic pressure during water immersion might be the other mechanism linked to performance recovery (Ingram et al., 2009;Vaile, Halson, Gill, & Dawson, 2008b).Barnett (2006) noted that some applications of cold water immersion have a contradictory eff ect of deceleration the recovery processes.This may explain the insignifi cant eff ect of cold water on total work of knee extension and all variables of knee fl exion in our study.The temperature of water and the duration of the water immersion should be discussed in this case.
In our work we observed signifi cantly higher HR max (173 ± 14 rpm, p = 0.02) during active recovery than during passive recovery (167 ± 14 rpm).Active recovery also led to signifi cantly higher values of HR avg (124 ± 8 rpm, p = 0.00) than passive recovery (107 ± 12 rpm).Bielik (2010) recorded signifi cantly increase of HR avg (125 ± 12.4 rpm, p < 0.01) after active recovery on bicycle ergometer than in passive rest (105.1 ± 8.2 rpm).We agree with the work of Draper, Bird, Coleman, and Hodgson (2006), who applied walking recovery after climbing performance.HR avg was increased after active recovery between maximal climbing performance tests.Authors Draper, Bird, Coleman, and Hodgson (2006) found steeper increase of HR avg than with passive recovery but the diff erences were not statistically signifi cant.
HR avg during passive recovery was 107 ± 12 rpm and during using cold water immersion was 97 ± 9 rpm.The diff erences were signifi cant.Klugar et al. (2009) reported signifi cant decrease of HR avg after scuba diving (on average from 63.2 to 58.8 rpm) where the temperature of the water reached an average of 29° C.
We can fi nd diff erent opinion on this problem in the current literature, which is caused mainly by the diff erent load localization and its eff ect on the blood circulation (Baláš et al., 2010;Draper, Bird, Coleman, & Hodgson, 2006).

CONCLUSIONS
We found signifi cantly smaller changes of the peak torque after active recovery (↑ 0.9 N × m) compared with cold water immersion (↓ 14.6 N × m) and passive recovery (↓ 13.9 N × m) in subsequent knee mediumterm muscular performance.We noticed a signifi cant decrease of average power after active recovery (↓ 5 W) and after cold water immersion (↓ 23.7 W) and after passive recovery (↓ 25.9 W).There was no signifi cant diff erence in total work by using the other recoveries.Active recovery and cold water immersion had a significant eff ect on total work of knee extension and fl exion.
HR max signifi cantly increased during active recovery than passive recovery (173 ± 14 vs. 167 ± 14 rpm) and HR avg (124 ± 8 vs. 107 ± 12 rpm).The cold water immersion did not have any eff ect on HR max compared to passive and active recovery on the contrary there were signifi cant changes in HR avg during passive recovery and active recovery (97 ± 9 vs. 107 ± 12 rpm).
In our study we found a positive eff ect of walking recovery on lower limbs strength performance.After passive recovery we found larger decline of muscle performance in medium term activities.

Fig. 1
Fig. 1Means and standard deviation of peak torque for knee extension for subsequent measurement for three types of recovery (PAS -passive recovery, ACT -active recovery, CWI -cold water immersion)