Knee Joint Effusion and Cryotherapy Alter Lower Chain Kinetics and Muscle Activity

Abstract

Context: Cryotherapy has been shown to disinhibit the quadriceps muscle after joint effusion by a resting measure (Hoffmann reflex) of motor recruitment. I sought to determine whether cryotherapy-induced motor recruitment changes resulted in subsequent changes in functional movement.

Objective: To quantify muscle recruitment changes and knee joint function after joint effusion and subsequent joint cryotherapy.

Design: A 3 × 4 multivariate mixed-model design was used to compare groups (normative, effusion/control, effusion/cryotherapy) across time (preinjection, postinjection, 30 minutes postinjection, and 60 minutes postinjection).

Setting: Human performance laboratory.

Patients or Other Participants: Forty-five volunteers (26 males, 19 females; age = 21 ± 2 years, height = 174.8 ± 10.2 cm, mass = 78.1 ± 15.4 kg).

Intervention(s): Experimental joint effusion was used to elicit inhibition of the quadriceps muscle. Cryotherapy was a treatment intervention.

Main Outcome Measure(s): Lower chain peak joint torque, peak and average power, knee anterior joint reaction force, and average and peak vastus medialis, vastus lateralis, medial hamstrings, and gastrocnemius muscle normalized electromyographic activity were collected during the extension phase of a seated, recumbent stepping motion with a resistance of 36% of 1-repetition maximum and a controlled cadence of 1.5 Hz.

Results: Decreases in peak torque and peak power were observed after effusion, whereas no decrease was observed over time in the cryotherapy or normative groups. A decrease in peak vastus lateralis activity was also noted after effusion relative to other groups. Also, the effusion/cryotherapy group had a greater knee anterior joint reaction force relative to the effusion/control and normative groups after effusion.

Conclusions: Joint cryotherapy negated movement deficiencies represented by knee peak torque and power decreases. This could be due to facilitated vastus lateralis activation relative to other groups.

Arthrogenic muscle response is an ongoing reflex inhibition or facilitation of joint musculature after distension or damage to structures of the joint. ^{1 2} A natural response, it is designed to protect the joint from further damage. However, the presence of arthrogenic muscle response creates gait deficiencies ³ and potentially retards the rehabilitation process despite complete muscle integrity. ²

If the affected joint can be protected from further damage, a decrease in arthrogenic muscle response might allow for early active exercise to expedite the rehabilitation process. Early active exercise during rehabilitation after joint injury is essential for decreased healing time, ⁴ increased structural strength and stiffness of ligaments, ⁵ increased collagen synthesis in tendons, ⁶ increased proteoglycan content in articular cartilage, ⁷ and periosteal expansion of bone tissue. ⁸ Uninhibited motor recruitment would also help to restore normal movement patterns around the injured joint during active exercise, assisting in the restoration of neuromuscular control and the overall rehabilitation process.

Neuromuscular consequences have been examined in an experimentally induced joint effusion model. ^{3 9–13} Cryotherapy may disinhibit or undo the inhibition of the quadriceps muscle after knee joint effusion. ¹⁰ Previous authors ¹⁰ have demonstrated that joint cryotherapy after artificial knee joint effusion resulted in disinhibition or facilitation of the quadriceps. In other words, the inhibition due to the knee joint effusion was removed after joint cryotherapy treatment. The measurement used in this study, however, was a resting or involuntary measure of motor recruitment (Hoffmann reflex).

During voluntary recruitment, different motor strategies may be used to accomplish closed chain movement. Therefore, the results of disinhibition after joint cooling, as previously reported, ¹⁰ may not be generalizable to voluntary contraction and subsequent functional movement. Muscle inhibition and/ or cryotherapy-induced disinhibition could be masked by altered descending volleys, and potential changes to the motoneuron pools of antagonists and/or synergists may have an effect on the overall movement pattern. ¹⁰ Essentially, functional movement could be affected by compensation from other joints of the lower chain or by changes in recruitment of other muscle groups crossing the joint. More data are needed to determine if cryotherapy might have beneficial effects on voluntary, closed chain movement after joint injury.

My purpose was to quantify muscle recruitment changes and knee joint kinetics during a voluntary, closed chain activity after joint effusion and subsequent joint cryotherapy. This study will provide some evidence as to whether joint cryotherapy restores normal motor function about the knee after joint effusion.

METHODS

A 3 × 4 multivariate mixed-model design was used to compare groups (normative, effusion/control, effusion/cryotherapy) across time (preinjection, postinjection, 30 minutes postinjection, and 60 minutes postinjection) for all dependent variables of the dominant ankle, knee, and hip. Dependent variables included peak joint torque (Nm), peak and average joint power (W), peak and integrated vastus medialis (VM), vastus lateralis (VL), medial hamstrings (MH), and gastrocnemius (G) processed electromyographic (EMG) activity and knee anterior joint reaction force (N).

Subjects

Subjects were 45 healthy, physically active volunteers (26 males, 19 females; age = 21 ± 2 years, height = 174.8 ± 10.2 cm, mass = 78.1 ± 15.4 kg) randomly assigned to 1 of 3 equal groups (normative, effusion/control, or effusion/cryotherapy). Volunteers were considered physically active if they participated in moderate exercise for 30 minutes at least 3 times per week. Subjects had no lower extremity conditions resulting in surgery in the last 2 years, nor did they have any lower extremity injuries in the 6 months before data collection. Furthermore, subjects had no history of any neurologic or systemic disorder that interferes with normal sensory or motor activity. Informed consent was obtained from all subjects before participation in this project. The University Institutional Review Board for Human Subjects granted approval for this project.

Instrumentation

Kinetic variables (peak joint torque; peak and average joint power for the ankle, knee, and hip; knee anterior joint reaction force) were measured using the Ominkinetic dynamometer (Interactive Performance Monitoring, Inc, Pullman, WA). The Omnikinetic device is a lower extremity dynamometer in which a resisted semirecumbent stepping motion is performed. (For a detailed description of this instrument, see Davis and Dolny. ¹⁴) Briefly, the Omnikinetic closed chain dynamometer uses a modified Hanavan anthropometric model to calculate individual body segments and model lower extremity kinetics. ¹⁵ Assumptions include the following: motion is constrained geometrically as a 2 degrees-of-freedom kinematic chain, the hip is fixed throughout movement, joints are frictionless hinge joints, and the limb segments are rigid bodies. ¹⁶ The Omnikinetic dynamometer employs a concentric (quadriceps) extension and eccentric (quadriceps) flexion pattern, ¹⁶ and a pneumatic resistance mechanism is used to reduce the inertial effects during exercise. ¹⁴ Rotary position encoders are positioned at the crank and pedal axis, a linear encoder measures seat position, and a 2-dimensional force transducer at the axis of the pedal measures force. The Omnikinetic also provides a limited estimation of anterior tibial shear force (knee anterior joint reaction force) by assessing joint moments and reaction forces through a cadaveric and radiographic model by Nisell. ¹⁷ Osternig et al ¹⁸ provided a description of how this model is used to estimate anterior tibial shear force at the knee. Signals are sampled at 65 Hz with a 16-bit A/D converter (CIO-Terminal/DST model 1402/16; Computer Boards, Inc, Middleboro, MA).

During the semirecumbent stepping motion, EMG data were collected using the MP100 system (BIOPAC Systems, Inc, Santa Barbara, CA). Signals were amplified (TEL100M; BIOPAC Systems Inc) from disposable, pregelled Ag-AgCl electrodes. The EMG measurements were collected at 1000 Hz. The input impedance of the amplifier was 1.0 megaohm, with a common mode rejection ratio of 90 dB, high- and low-pass filters of 20 and 400 Hz, a signal-to-noise ratio of 70 dB, and a gain of 1000. Raw EMG signals were processed using a root mean square (RMS) algorithm with a 10-millisecond moving time window.

Surface temperature measurements were collected using 30-gauge, exposed junction thermocouples with Kapton insulated leads (TX-31; Columbus Instruments, Columbus, OH) connected to a portable Datalogger (MSS-3000; Commtest Instruments, Christchurch, New Zealand).

Orientation

A 30-minute orientation and screening, 1 week before testing, allowed for a general explanation of the study and measurement protocol. All risks involved in the study were described, and subjects signed informed consent forms. Subjects were randomly assigned to a group (normative, effusion/control, or effusion/cryotherapy), and descriptive data (age, height, and mass) were recorded. Anthropometric measurements were taken to construct a model for each subject: distance from the tuberosity of the fifth metatarsal to the lateral malleolus, foot circumference at the fifth metatarsal, circumference from the navicular tuberosity to the calcaneus, barefoot distance from floor to lateral malleolus, shoed distance from floor to lateral malleolus, distance from lateral malleolus to the center of rotation of the knee (lateral joint line), circumference immediately above malleolus, circumference at the thickest diameter of the calf, lateral condylar notch to the peak of the greater trochanter, circumference immediately above the knee, and circumference at the thickest diameter of the thigh. Measurements were collected with a standard medical tape measure and recorded in cm. Subjects then practiced on the Omnikinetic machine at a self-selected maximal speed to “get a feel” for the pneumatic resistance mechanism. They then practiced at a controlled cadence of 1.5 Hz (metronome set at 90 beats/min), which was the cadence used for testing. After this warm-up period, a 1-repetition maximum for the Omnikinetic dynamometer was established by beginning with a resistance equal to body weight and increasing the resistance 40 N for each subsequent attempt. Intervals of 40 N allowed the subject to reach 1-repetition maximum relatively quickly. Pilot work provided evidence that smaller intervals (less resistance) might lead to fatigue, in turn affecting the value. The motion was similar to that of a seated leg press with the dominant leg. Two minutes' rest separated attempts, and the greatest resistance that could be fully extended marked the subject's 1-repetition maximum.

Subject Preparation

Locations over the VM, VL, MH, and G of the dominant leg were shaved, debrided (abraded with fine sandpaper), and cleaned with isopropyl alcohol for application of the EMG electrodes (10 mm Ag-AgCl; BIOPAC Systems Inc) for each subject. Leg dominance was defined as the stance leg or the preferred jumping leg. The locations for EMG electrodes were determined according to Basmajian and Deluca, ¹⁹ who suggested that electrodes be placed approximately halfway between the estimated innervation zone and the distal tendon insertion. Electrodes were placed in approximate line with the longitudinal axis of the muscle, spaced 2 cm apart center to center. The ground location was on the ipsilateral medial malleolus. Placement of the electrodes was confirmed by manual muscle testing and visual EMG inspection of real-time muscle activity on the computer. Electrodes remained in place for the duration of the testing session.

Before each measurement, thermocouples were temporarily placed on the center of the patella and on the skin lateral to surface EMG electrodes at the VM and MH. The thermocouples were held in place with a 4-cm strip of athletic tape. Surface temperature data were collected before each measurement interval.

Omnikinetic Procedure

The Omnikinetic device allows for measurement of kinetic data during a semirecumbent stepping motion against a set resistance. The subject was seated in the dynamometer with his or her back resting against the back of the seat and the fifth metatarsal of the foot lined up with a mark on the foot-plate ( Figure 1). The subject's shoe was marked with a permanent marker in line with the mark on the footplate to ensure identical placement during subsequent testing sessions. The foot was also fixed to the footplate with a strap around the toes and another around the back of the heel. Using the previously entered anthropometric data, the Omnikinetic software specified a seat location for the subject. The seat was located to that position, which was recorded to ensure identical placement during subsequent measurement sessions.

Figure 1. Positioning of a subject on the Omnikinetic closed chain dynamometer.

Once the subject was positioned and secured, he or she performed several practice trials against light to moderate resistance (approximately 15% of the 1-repetition maximum). At least 10 practice trials were performed, and in some cases, more were performed until the subject felt ready to begin testing. Kinematic data were visually confirmed during this time by inspecting the real-time graphic display. The subject then maintained a position of 40° of knee flexion, measured goniometrically, at 70% of the 1-repetition maximum for normalization of EMG data. A goniometer was used to position the knee joint angle, and the subject held the load in place for the 5-second period while muscle activity was recorded.

After a 10-minute rest period, each subject performed the recumbent stepping exercise at 36% of the 1-repetition maximum and a controlled cadence of 1.5 Hz. The stepping motion cadence was matched to a metronome to control the speed, and 8 seconds of data were collected (5 repetitions). The subject was allowed to begin the exercise, and after a few repetitions for synchronization with the metronome, I initiated data collection from the Omnikinetic. An output from the Omnikinetic allowed for simultaneous data collection from the EMG. The data collection procedure was repeated after joint effusion and at 30 and 60 minutes postinjection.

Joint Effusion Procedure

An area inferomedial to the patella was cleaned with alcohol and Betadine (The Purdue Frederick Co, Norwalk, CT). Using a sterile, disposable syringe and under sterile conditions, a physician injected 55 mL of sterile saline into the knee joint capsule. No lidocaine was used to anesthetize the injection area. The physician then performed an effusion wave and ballotable patella test to ensure that the effusion was within the knee joint capsule.

Testing Procedure

Three subjects, one from each group, reported for each testing session. I was able to accommodate 3 subjects at a time by staggering the preinjection measurements. Strict adherence to the time intervals was observed. Electrode placement sites were prepared as previously described. Each subject warmed up on a stationary exercise bicycle at a submaximal intensity and performed general stretching of major lower extremity muscle groups for 5 minutes before the initial measurement.

As previously discussed, normalization data for EMG were established, and baseline (preinjection) measures were recorded. Each subject in the effusion/control or effusion/cryotherapy groups was then injected with saline. Once the effusion was confirmed, subjects were posttested immediately. Subjects in the normative group were pretested and posttested at the same time intervals as other groups, except they received no injection of saline. Subjects in the effusion/cryotherapy group then received 1.5 L of crushed ice applied directly to the anterior surface of the knee and secured with an elastic wrap during this time immediately after the postinjection measurement. The ice bag was placed so that the most proximal part of the ice bag covered the most proximal part of the patella to reduce any potential cooling of the vasti muscles. Subjects in the effusion/control group received a sham ice bag applied to the knee in the same fashion. The bag was filled with a room-temperature substance of similar mass and texture (cat litter). Subjects in the normative group received no intervention. All subjects sat quietly between testing sessions. After 30 minutes of treatment, the ice bag or sham bag was removed and measurements (30 minutes postinjection) were recorded for all subjects, followed by another set of measurements at 60 minutes (60 minutes postinjection). After data collection, subjects were instructed not to participate in any weight-bearing activity other than walking for 24 hours.

Surface temperature measurements were recorded from the knee and EMG electrode sites at each measurement interval. This was done to ensure that cooling was not taking place at the EMG electrode sites, which could potentially affect the EMG measurements. ²⁰

Using custom software, matched data from the Omnikinetic and EMG were rectified and processed using an RMS algorithm with a 10-millisecond time window. The EMG data were matched to the Omnikinetic by identifying the initiation of the knee extension phase from the crank position sensor. Figure 2 shows an example of the raw EMG output with the Omnikinetic data to identify the extension phase of movement. The knee extension phase of each repetition was identified for the EMG and Omnikinetic files, and the mean peak and integral from the extension phase of 5 repetitions were used for analysis of each dependent variable. The EMG data were normalized to the mean processed EMG data collected during isometric contractions at 40° of knee flexion.

Figure 2. Raw electromyographic activity data with input from the Omnikinetic dynamometer. The first vertical line represents the beginning and the second vertical line the end of the extension phase of a single repetition.

Statistical Analysis

I performed a 2-way multiple analysis of variance with repeated measures on time intervals to determine overall group differences for joint torques, power, and knee anterior joint reaction force. Additional 2-way multiple analyses of variance were performed on integrated and peak RMS EMG amplitude to determine between-group differences over time for each muscle. Univariate F tests and the Sidak multiple comparison procedure were used to make post hoc comparisons. The a priori α level was set at P ≤ .05.

RESULTS

Means and standard deviations are provided in Tables 1– 3. An overall time × group effect was detected for knee kinetic variables (F _24,504 = 2.228, P = .001) and peak EMG measures (F _24,504 = 2.062, P = .002). Deficiencies were found in the effusion/control group after effusion relative to the effusion/ cryotherapy and normative groups. The effusion/control group showed a 25.0% decrease in peak torque (univariate F _6,126 = 2.226, P = .045, Sidak P = .005), a 22.7% decrease in peak power (univariate F _6,126 = 4.527, P = .0001, Sidak P = .001), and an 11.9% decrease in peak VL activity (univariate F _6,126 = 2.552, P = .023, Sidak P = .001) at 30 minutes postinjection relative to the effusion/cryotherapy and normative groups. Further, the effusion/control group extended this finding at 60 minutes postinjection for peak power (28.9% decrease, Sidak P = .023,) and peak VL activity (11.3% decrease, Sidak P = .047). Conversely, knee anterior joint reaction force increased (F _6,126 = 2.741, P = .017) in the effusion/control group at 30 (23% increase, Sidak P = .008) and 60 minutes (27.4% increase, Sidak P = .006) postinjection relative to the other groups ( Figure 3). Means and standard deviations for surface temperature measurements (control variable) are presented in Table 4. Temperatures did not change at the electrode sites, but surface temperature at the anterior knee decreased.

Table 1. Results for Kinetic Variables (Mean ± SD).

Table 3. Normalized (Percentage of Isometric Reference Position) Integrated Electromyographic Activity (Mean ± SD).

Figure 3. Knee anterior joint reaction force data represented as percent change from baseline. Eff/con indicates effusion/control group; eff/ice, effusion/cryotherapy; norm, normative. *Greater than baseline relative to effusion/control, normative groups ( P < .05) .

Table 4. Surface Temperature (°C) at the Anterior Knee, Vastus Medialis, and Medial Hamstrings Electrode Sites (Mean ± SD).

DISCUSSION

The finding of deficiencies in quadriceps recruitment and knee joint torque is consistent with previous work demonstrating quadriceps inhibition during gait after knee joint effusion. ³ Torry et al ³ reported that subjects with knee joint effusion walked with a quadriceps-avoidance gait, mediated by a decrease in quadriceps muscle activation. Arthrogenic muscle inhibition of the quadriceps has often been observed in subjects with pathologic or experimentally induced knee joint effusion. ^{10 11 21–25} The one difference observed with these data is that I did not find a significant decrease in VM activity, as has often been reported. This could have stemmed from variability inherent to EMG measurements during dynamic movement. ²⁶ Another possibility for observing changes in VL activity while not seeing changes in VM activation is the nature of the movement (seated leg-press motion) used in this study compared with other authors who have used walking, ³ open chain knee extension torque, ^{27 28} or a resting measure of motor recruitment. ^{10 11 25} Considering the linear nature of the movement used in this study and the load applied, demand on the VL may have been greater. This could have provided a larger window of motoneuron pool exposure for detection of changes in recruitment.

Although the findings support those of authors reporting quadriceps inhibition after knee joint effusion, I did not observe significant compensation at the hip or ankle in terms of kinetic variables. Further, significant increases were not noted in hamstring or triceps surae EMG activity, which have previously been reported. ^{3 11 13} Certainly, measurement variability could play a role, and interpretation of EMG measurement during dynamic movement is limited. ²⁹ After experimental knee joint effusion, Torry et al ³ found a decrease in knee joint torque and quadriceps EMG activity during gait. They also reported an accompanying increase in hip extensor work and hamstring EMG activity. These movement patterns (walking) seem similar to the patterns used in this experiment (recumbent stepping motion), but the intensity or resistance is quite different. In this experiment, I used a resistance of 36% of a previously recorded 1-repetition maximum, and subjects stepped to a cadence of 90 steps/minute. Further, trunk flexion during the stepping motion could play a significant role in the potential for hip compensation during the movement in this experiment.

Consistent with previous work, ¹⁰ I observed that joint cryotherapy applied after knee joint effusion resulted in no inhibition, whereas the effusion/control group demonstrated signs of inhibition at 30- and/or 60-minute intervals. These findings suggest that knee joint cryotherapy had a disinhibitory effect on the quadriceps. Disinhibition is defined here as a return of some measure of recruitment to baseline or preinjury levels. Using an involuntary measure of motor recruitment (Hoffmann reflex), a similar result was observed previously. ¹⁰ Although the current kinetic data appear to support the use of joint cryotherapy to return joint torque and power to levels observed before knee joint effusion, some evidence suggests that joint cryotherapy may cause facilitation above baseline levels. Peak VL activity demonstrated a trend toward facilitation after joint cryotherapy, yet there was no significant difference between the normative and effusion/cryotherapy groups. Additionally, knee anterior joint reaction force increased in the effusion/cryotherapy group (approximately 27% above baseline), indirectly suggesting increases in quadriceps activation during extension relative to other group values. Overall, these data suggest that joint cryotherapy may return quadriceps recruitment to more active levels or even above baseline levels. This may seem to be advantageous, but more data are needed to determine if this “overshoot” in quadriceps activation could result in altered movement patterns.

The possible mechanisms responsible for quadriceps inhibition after knee joint effusion have been thoroughly discussed in the literature. ^{2 10 11 13 30–32} However, it should be noted that these mechanisms are largely theoretic. Increased afferent output from slowly adapting joint mechanoreceptors appears to initiate the inhibitory mechanism, ^{9 25 33} which may then be mediated through Ib inhibitory interneurons. ³⁴ Activity through this type of interneuron would produce a patterned response consisting of inhibition of the quadriceps and facilitation of the hamstrings and triceps surae muscle groups. ^{3 11} However, recent findings suggest that the arthrogenic effects observed in the quadriceps and triceps surae groups after knee joint effusion are likely mediated by presynaptic and postsynaptic mechanisms. ^{13 30} These data ^{13 30} indicate that the mechanism responsible for quadriceps arthrogenic inhibition is not as clear as previously assumed and that descending activity plays a large role in the excitability of the motoneuron pool.

The process by which knee joint cryotherapy disinhibits the quadriceps is unknown. The disinhibitory effects of joint cryotherapy have previously been observed after knee joint effusion with an involuntary measure of motoneuron recruitment (Hoffmann reflex) ¹⁰ and in healthy subjects. ^{35 36} In all cases, the quadriceps motoneuron pool was facilitated above baseline levels. ^{10 35 36} Investigators ^{10 36} have proposed that facilitation of motoneuron pools above baseline levels after joint cooling suggests that supraspinal influences are largely responsible. Indeed, supraspinal pathways generally modulate spinal reflexes to allow for controlled movement. ³⁷ Perhaps this tonic activity is reduced during cryotherapy treatment, allowing for less supraspinal control over reflexive activity and recruitment within the quadriceps motoneuron pool. As previously discussed, this may seem to be beneficial as clinicians try to overcome inhibition in the rehabilitation setting. However, without modulation from supraspinal centers, movement could become uncontrolled with excessive recruitment. So, although quadriceps activation and joint torque and power are increased due to joint cryotherapy, the quality of the movement remains a question.

These data are consistent with the concept of cryokinetics, in which an injured joint is cooled to allow for exercise and its therapeutic effects. ³⁸ Clinical observation provides support for the concept that cryotherapy facilitates the motoneuron pool such that injured athletes seem able to perform exercises that were not possible before a cryotherapy treatment. Knight ³⁸ attributed this finding to a decrease in residual pain from the injury. In previous work using this model of joint injury, pain was not found to be a factor, as the experimental effusion model resulted in no or minimal pain. ⁹ However, subjects in this study did qualify the sensation after effusion as “tight.” This tightness could produce discomfort when muscles surrounding the effused knee contract, potentially increasing pressure. If this were the case, cryotherapy could have an effect on the perception of the potential discomfort. Although the extent to which pain contributes to arthrogenic muscle inhibition is unknown, the mechanical contribution demonstrated with joint effusion seems clear.

The effusion model provided an excellent controlled model for injury, but its limitations must be mentioned. I did not investigate pain and other inflammatory factors, yet pain certainly must play some role in the inhibitory mechanism. However, it is a difficult variable to control. Future authors should focus on pain and pathologic conditions. Another limitation with the effusion model was the resorption of the sterile saline during testing. From past experience using this model, I understood that the saline would be in the capsule for a limited period of time, and active movement would likely result in enhanced movement of the saline from the capsule. Although I did observe some effect at the 60-minute interval for some dependent variables, it was likely that the saline was almost completely resorbed by this time. Finally, the effusion model does have some variability due to the fact that different-size subjects will have different-size capsules, creating various pressures within the capsule. This factor could lead to different levels of inhibition and/or facilitation.

Another limitation to this study was the use of a semirecumbent stepping motion to represent functional movement. Indeed the movement pattern is similar to that of walking and running, but the position and demand are far different. It should be noted, however, that muscle recruitment patterning is very similar to the stance phase of gait, and although no direct comparisons should be made, some information may be helpful to future authors in this area. The Omnikinetic dynamometer is an instrument that relies on theoretic modeling to estimate force and power around the lower extremity joints. A level of error is inherent with this type of modeling. In this study, foot position was another factor that may have added to some of the variability. Because foot position between subjects was controlled by placement of the fifth metatarsal on a marked spot distal to the force transducer, the distance between the force transducer and the lateral malleolus likely varied among subjects with different-size feet. However, moderate to strong reliability has been reported previously for this instrument using these exact methods. ³⁹

In conclusion, these data are further evidence that joint cooling may assist in restoring neuromuscular deficiencies after joint injury. Joint cryotherapy negated movement deficiencies represented by knee peak torque and power decreases. This effect was likely the result of enhanced activity of the VL after joint cryotherapy. With enhanced quadriceps activity, an increase in anterior tibial shear force was also observed. These data are consistent with the idea that cryokinetics may be an effective tool during rehabilitation to restore motor deficiencies. More data are needed to determine if the observed increases in muscle activation and knee anterior joint reaction force might be detrimental to the quality of movement during therapeutic exercise and other physical activity.

Table 2. Normalized (Percentage of Isometric Reference Position) Peak Electromyographic Activity (Mean ± SD).