The muscle damaging protocol
Muscle damage was induced through the use of a drop jump protocol similar to that used by Miyama and Nosaka, 2004a; subjects dropped from a 0.6 m box and upon landing jumped up maximally, landing on the same surface. Five sets of twenty drop jumps were performed on a concrete based floor, in time with a set of recorded beeps allowing 10 s rest between each jump and 2 minutes rest between each set.

Treatments
The cryotherapy group underwent a seated immersion (up to iliac crest) in an inflatable ice bath for 12 min; legs were kept apart in the cold water to ensure a maximum surface area exposure. The water was maintained at the recommended temperature of 15 ± 1°C (by adding crushed ice), Meeusen and Lievens, 1986 predict this to elicit a drop in intramuscular temperature of approximately 5-8 °C. This treatment was applied immediately post-exercise and every 24 h thereafter for the following 3 days. The control group remained seated for the same amount of time as the immersions were administered for and received no treatment.

Criterion measurements
Dependent variables to indicate damage were maximal voluntary contraction (MVC) of the knee extensors, creatine kinase activity (CK), muscle soreness (DOMS), range of motion (ROM) and swelling which have been used in previous research (Byrne et al., 2004; Howatson and van Someren, 2003; Warren et al., 1999). MVC measurements were recorded immediately pre and post-exercise and for the following 96 h at 24 h increments. All other variables were measured immediately pre-exercise and then for the following 96 h at 24 h increments. All measures were obtained before the cryotherapy treatments to avoid any potential effects such as analgesia.

Plasma creatine kinase activity (CK): Plasma CK was sampled from the subject's earlobe using capillary puncture. A sample of whole blood ~30 μL was collected into a heparinised capillary tube and examined immediately using an automated reagent test strip analyser (Reflotron Plus Analyser, Bio Stat Ltd. Stockport, UK). The normal reference ranges of plasma CK activity for this method are 24-195 IU.L^-1 (according to the analyser instruction manual).

Maximal Voluntary Contraction (MVC): MVC was assessed using an isokinetic dynamometer (Kin-Com, Chattecx Corporation, Chattanooga Group, Tennessee, USA). The device was set up according to the manufacturer's recommendations to exercise the knee extensors of the non-dominant leg. Anatomical zero was set at a knee angle of 0° (full extension) and MVC was determined by setting the joint angle at 70° of flexion and then locked in place and marked to ensure consistency during subsequent testing sessions. Prior to the commencement of each test the limb weight and moment acting upon the dynamometer power head were corrected for gravity. Each repetition lasted 3 s interspersed with 60 s rest, the peak torque generated from three trials was recorded as the MVC.

Delayed Onset Muscle Soreness (DOMS): Subjects rated muscle soreness whilst standing and during a 90° squat movement (DOMS-SQ) using a 200 mm visual analogue scale (VAS) with the far-left end point representing 'no pain' (0) and the far-right end point representing 'extremely painful' (200) (Howatson et al., 2007).

Knee Flexion - Range of Motion (ROM): For the measurement of knee flexion ROM, subjects laid prone on a massage bed with both knees fully extended. From this position subjects were asked to fully flex their non-dominant knee. The knee joint angle was determined by using a goniometer (Bodycare Products, Warwickshire, UK) and universal landmarks (lateral epicondyle of the femur, lateral malleous and greater trochanter) to ensure alignment (Tokmakidis et al., 2003). Landmarks were marked with a semi permanent pen to ensure consistency on subsequent measures, 3 measurements were performed and the average was reported.

Swelling: thigh circumference: An anthropometric tape measure (Bodycare Products, Warwickshire, UK) was used to determine thigh circumference. Mid-thigh circumference was determined mid-way between the greater trochanter and the lateral epicondyle of the femur. The site on the subject's non-dominant leg was measured three times and the averages were reported. The skin was marked with a semi permanent marker for consistency on subsequent days.

Statistical analysis
The CK data were log transformed and values for ROM, swelling and MVC were normalised and expressed as a percentage change relative to baseline. Difference in the measured variables among conditions and trials were analyzed with two-way ANOVA for repeated measures (Treatment, 2 x Time, 5), using treatment as the between subject factor and time as the within-subjects factors. Data were analysed using SPSS for Windows (15.0 software package) and significance was set at p < 0.05.

Values for CK, MVC, DOMS, DOMS-SQ and thigh circumference showed significant time effects (p < 0.01), although there were no interaction or group effect (p > 0.05), baseline data for both conditions are shown in Table 1.

In both groups, CK activity peaked 24 h post-exercise and returned to baseline values by 96 h post-exercise (Figure 1). Difference between the groups were approaching significance (F_1,16 = 4.16, p = 0.058). No differences were found between conditions in MVC (F_1,16 = 0.08, P = 0.783) and the peak percentage decrement immediately post-exercise from baseline measures for the control and cryotherapy group was 81% and 78% respectively (Figure 2). Muscle soreness peaked 48 h post-exercise and returned to baseline levels by 96 h post-exercise, soreness during a 90° squat movement followed the same time course (Figure 3 and 4). Muscle soreness in a static position and muscle soreness during a squat movement did not significantly differ between groups. ROM did not significantly change over time and was not significantly different between the 2 groups. An increase was shown in thigh circumference for both groups (Figure 5) although differences between the conditions were not significant (p > 0.05).

This investigation examined the efficacy of repeated CWI in enhancing recovery from EIMD. This is the first investigation to examine the effects of serial CWI on the recovery from a sport specific bout of heavy plyometric exercise. The results clearly demonstrate that repeated CWI did not significantly enhance the recovery process following damaging exercise to a greater extent than an experimental control. These findings add to a growing body of literature that demonstrates cryotherapy to be an ineffective strategy when recovering from EIMD (Howatson and van Someren, 2003; Howatson et al., 2005; Isabell et al., 1992; Sellwood et al., 2007).

The exercise bout was successful in inducing muscle damage, which was evident from the significant change of dependent variables and concurs with previous literature that reported similar trends following a like mode of exercise (Miyama and Nosaka, 2004b). Of particular note is muscle soreness peaking at 48h post-exercise and MVC, which showed a large decline immediately post-exercise and a general recovery towards pre-exercise levels in the subsequent 96h and concur with other previous literature (Isabell et al., 1992; Eston and Peters, 1999; Miyama and Nosaka, 2004a; 2004b; Howatson et al., 2007). Intracellular release of CK has been used as an indirect marker of EIMD for many years (Manfredi et al., 1991; Howatson et al., 2005). The CK response in this investigation peaked at 24h post-exercise, which is the same response as previous data using a similar protocol to induce damage (Miyama and Nosaka, 2004a; 2004b). It would appear that other lower limb exercise such as downhill running (Eston et al., 1995) also follows a similar trend to this investigation. However, most CK responses following damaging eccentric exercise in the upper limb tend to be slightly more delayed and peak 24 h later (Howatson et al., 2007; Nosaka et al., 2002); although the reason for this is unclear (Miyama and Nosaka, 2004a), it may be speculated that the upper limb is more unaccustomed to eccentric loading and hence has a greater susceptibility to damage than the lower limb; consequently CK is more delayed and of a greater magnitude in the upper limb (Jamurtas et al., 2005).

Cryotherapy is documented to constrict capillaries, reduce capillary permeability and blood flow (Meeusen and Lievens, 1986) thereby attenuating swelling and the inflammatory response (Smith, 1991) which may reduce the negative effects associated with damaging exercise. Limb girth showed no discernable difference between groups and consequently provided indirect evidence that the intervention was unsuccessful in bringing about these vascular changes, a result that has been demonstrated elsewhere (Sellwood et al., 2007). In addition, cryotherapy has been speculated to reduce membrane permeability, thereby reducing CK efflux (Eston and Peters, 1999) and alter nerve conduction velocity and hence pain tolerance (Algafly and George, 2007), however, these changes were not observed in this investigation.

Contrary to the positive effects that have been found in some studies (summarised by Wilcock et al., 2006), the repeated applications of cryotherapy that were administered in the present study did not enhance the recovery of any dependent variable. These findings concur with a number of other cryotherapy investigations (Isabell et al., 1992; Howatson et al., 2005). Perhaps a positive point is that the immersions do not seem to have a detrimental effect on recovery, although the impact on adaptation or repeated bout effect has yet to be elucidated. It is important to note that some aspects of fatigue, not measured in this study, may have improved as an effect of the CWI, which may include psychological benefits.

Another factor that has the ability to affect the intramuscular cooling from cryotherapy is the amount of adipose tissue (particularly subcutaneous) in the area being cooled (Merrick et al., 1999). Myrer et al., 2001 found a significant inverse relationship between overlying adipose tissue and intramuscular temperature changes during and after cryotherapy treatment of the calf muscle at different skin fold thicknesses (<8 mm, 10-18 mm and >20 mm). Myrer and colleagues (2001) concluded that the greater amount of adipose thickness lead to a greater amount of time for maximum cooling to be obtained. This is further emphasised by the results of Otte et al., 2002, who displayed dramatic effects of skin fold thickness on cooling rates of the anterior thigh. Skin fold thickness of 20 mm or less required 20 min cooling, whereas skinfolds of 20-30 mm required 38 min and skinfolds of 30-40 mm required 59 min to decrease intramuscular temperature 7 °C from baseline. Although these skin fold thicknesses are quite large in comparison to active populations, like those used in this investigation, the observations of Otte et al., 2002 and Myrer et al., 2001 highlight the importance of determining skin fold thickness when administering cryotherapy treatments, something future investigations might like to consider.

• Cryotherapy, particularly cold water immersions are one of the most common interventions used in order to enhance recovery post-exercise.
• There is little empirical evidence demonstrating benefits from cold water immersions. Research evidence is equivocal, probably due to methodological inconsistencies.
• Our results show that the cryotherapy administered did not attenuate any markers of EIMD or enhance the recovery of function.
• We conclude that repeated cold water immersions are ineffective in the recovery from heavy plyometric exercise and suggest athletes and coaches should use caution before using this intervention as a recovery strategy