Bilateral assessment of knee muscle relationships in healthy adults

Asymmetric performance of flexor and extensor muscles of the knee may be a risk factor for knee injuries, especially the anterior cruciate ligament. Additionally, asymmetries in power and work may have correlations with fatigue and performance during functional tasks. Among untrained individuals, such asymmetries may be of potential interest for training prescription. Here, we investigated the bilateral performance of knee flexors and extensors muscle groups of untrained individuals. We quantified the torque-angle and torque-velocity relationships, as well as work, power and asymmetry indexes in 20 untrained male (25 ± 4 years old; height 1.74 ± 0.05 m; body mass 76 ± 9 kg). No significant asymmetry was observed for torque-angle and torque-velocity relationships, work and power output for knee flexor and extensor muscle groups (p < .05). Our results suggest that untrained male present symmetry in the knee flexion and extension bilateral performance. Changes in this behavior due to physical training must be monitored.

Palabras clave: pierna, lesión, rendimiento, contracción muscular, fuerza muscular, educación física y entrenamiento Here we proposed to investigate bilateral performance of knee flexors and extensors muscle groups by quantifying the torque-angle and torque-velocity relationships, as well as work, power and asymmetry indexes in untrained healthy individuals.For such purpose, performance was compared between preferred and non-preferred sides.

Design
Knee flexor and extensor torque-angle and torque-velocity relationships, were obtained bilaterally through maximal voluntary isometric and isokinetic concentric contractions (MVCs), respectively.Leg asymmetries were investigated by comparing preferred and non-preferred sides.All tests were performed in one day.Ethical approval for our study was obtained from the local Institutional Ethics Research Committee (protocol number 2007791).Participants signed an informed consent form according to the Declaration of Helsinki before starting participation in the experiments.

Participants
Twenty male individuals (mean ± standard deviation for age: 25 ± 4 years; height: 1.74 ± 0.05 m; body mass: 76 ± 9 kg) volunteered for this investigation.Participants completed a questionnaire for the purpose of identifying history of physical trauma or surgery, neuromuscular injury and previous strength training for the lower extremity, all considered exclusion criteria for the experiment.All the participants performed up to three familiarization trials for each condition in the laboratory in the day before the tests.During the time of the experiment and the six months prior to testing none of the individuals was involved in regular physical activity programs aiming at strength and power development (e.g., strength training, athletics, and gymnastics).

Introduction
Lower extremity asymmetries have been suggested as a risk factor for muscular and joint injuries.Asymmetry in knee abduction moment was 6.4 times higher in a group of women who have suffered injury of anterior cruciate ligament (ACL) compared to uninjured women (Hewett et al., 2005).Athletes who suffered ACL tear presented lower hamstrings to quadriceps concentric force in the injured side compared to the contralateral (Soderman, Alfredson, Pietila, & Werner, 2001).Additionally, asymmetry in the ratio between dorsal flexion and plantar flexion force predicted the side of the injury in college athletes (Baumhauer, Alosa, Renstrom, Trevino, & Beynnon, 1995a).It has been postulated that bilateral differences for torque, work and maximal power higher than 15% may elicit impairments in performance or predisposition to injury among athletes (Zabka, Valente, & Pacheco, 2011).The origin of these injuries may rely on the repetitive higher load experienced by one side due to the asymmetries during performance (Oliveira, 2007).Therefore, while these asymmetries most likely are resultant of continuous training and functional adaptation, in untrained individuals asymmetries should not be observed.
Skeletal muscle functional adaptation can be verified by quantifying the force production during contractions at different joint angles and angular velocities (Doheny, Lowery, FitzPatrick, & O'Malley, 2008).The force output is influenced when the muscle length changes (Gordon, Huxley, & Julian, 1966).These changes are associated with alteration in joint angles, which finally reflect differences in torque output across the joint range of motion (Doheny et al., 2008).Considering healthy untrained individuals, possible effects of skeletal muscle functional adaptation in the upper extremity due to daily preferential use in functional tasks were suggested as an explanation for asymmetries in the elbow flexor isometric peak torque (Carpes, Geremia, Karolczak, Diefenthaeler, & Vaz, 2012).
The preferential unilateral use reported for upper extremity among untrained individuals may not be observed for the lower extremity, as for most of the tasks healthy individuals recruit both legs in an equivalent way (e.g., during walking, running, stair locomotion).However, there are some evidences suggesting asymmetries indexes related with injury (Zabka et al., 2011).Asymmetry in the performance of flexor and extensor muscles of the knee is a risk factor for knee ligament injury (Knapik, Bauman, Jones, Harris, & Vaughan, 1991; Schache, Crossley, Macindoe, Fahrner, & Pandy, 2011).In other hand, asymmetries in power and work may have correlations with fatigue and performance of functional tasks (Carabello, Reid, Clark, Phillips, & Fielding, 2010).Additionally, work output is a good indicator of the muscle's capability to perform daily life activities.On the other hand, power output provides a measure of the work rate intensity and permits to infer how quickly force can be produced.From a practical point of view, evaluating such asymmetries in untrained individuals before starting participation in sport routines could significantly help to minimize risks for injuries during the time course of exercise and training.the lower leg, supported in the ankle.Afterwards, the attachment was firmly fixed to the participants' segment.Positioning was careful controlled to be consistent among all the participants.
Participants were verbally encouraged to perform their maximal effort during MVCs.At the end of each protocol, the first trial was repeated to verify fatigue effects.If the torque at the final trial decreased more than 10% with respect to the initial trial, a fatigue effect was assumed to be present and the participant would be invited to repeat the protocol on a different day.In this study no participant was requested to come back to the lab as we did not observed fatigue effects.
During testing, the active range of motion ranged from full knee extension to maximal knee flexion (0° = full active knee extension).Torque signals were sampled at 100 Hz, and were gravity corrected through the overall range of motion using the Biodex software.Leg preference was verified using the revised version of the Waterloo's inventory (Elias, Bryden, & Bulman-Fleming, 1998).After verifying the leg preference, legs were named as preferred and non-preferred (NP).

Torque measurements
Torque-angle relationship: maximal isometric flexor and extensor contractions were executed with the knee joint fixed at five different angles: 30°, 60°, 75°, 90° and 105° (Baroni et al., 2013).The order of joint angles tested was random.For each angle two repetitions were performed and the contraction with the highest peak torque produced was considered for further analysis.Each isometric contraction lasted 5 s, and a 2 min interval was observed between the consecutive contractions.
Torque-velocity relationships: torque-velocity trials were conducted for flexion-extension isokinetic contractions at five different angular velocities: 60°.s -1 , 120°.s -1 , 180°.s -1 , 240°.s -1 and 300°.s - (Baroni et al., 2013).The order of angular velocities tested was random for each participant during the isokinetic test.For each angular velocity three flexion-extension isokinetic contractions were performed and the highest peak torque produced during flexion and extension movements was considered for further analysis.A 2 min interval was observed between the consecutive contractions.
Work and power determination: total work was computed considering the amount of force output maintained throughout isometric trials.Power was computed during isokinetic trials considering the amount of total work output divided by the time to complete that total work.

Asymmetry index
An asymmetry index (AI % ) describing the relative difference in strength between the preferred and non-preferred leg was calculated using Equation 1 (Chavet, Lafortune, & Gray, 1997).This equation provides the magnitude and direction of bilateral asymmetry considering leg preference.Negative AI % values indicated higher torque for the non-preferred limb.( 1 ) where: AI % means asymmetry index, considering the ratio between measurements in the preferred (P) and non-preferred (NP) leg.

Data analysis
Torque values were expressed as mean and standarddeviation for the group.Data normality was verified using Shapiro-Wilk test.A two-way analysis of variance (5 angles/ velocities x 2 legs) was used to evaluate the effects of angles/ angular velocities and leg on peak torque, power and work.A paired t-test was used to compare torque values obtained at the first and last trials in order to observe any fatigue effect.The significance level was set at .05 for all comparisons using a statistical package (SPSS 17.0, SPSS Inc., Chicago IL, USA).

Results
No fatigue effect was observed when comparing the first and last trials for the different variables we quantified.Torque -angle results indicate that isometric flexor torque was similar between the legs (F 1,19 = 1.861; p = .189).A main effect for joint angle was observed for the peak isometric flexor torque (F 4,76 = 359.66;p < .001)(Figure 1).In the same way, asymmetries between legs were not observed for extensor isometric peak torque (F 1,19 = 2.808; p = .110).A main effect for joint angle was observed in the isometric peak extensor torque (F 4,76 = 81.063;p < .001)(Figure 1).Torque-angle relationship for knee flexors and extensors during isometric trials followed a pattern already described in the literature (Frykberg et al., 2006).
Table 1 presents the magnitudes of asymmetry indexes computed for flexor and extensor peak torques at different joint angles (isometric) and different contraction velocities (isokinetic).term training.Here we evaluated the bilateral performance in torque-angle and torque-velocity muscle relationships, work, power, compared preferred versus non-preferred leg, and computed asymmetry indexes for knee flexor and extensor muscles in non-trained participants.No statistical significant asymmetry was observed in our study.Power and work were similar between the legs.It was expected since maximal isometric peak torques and maximal isokinetic peak torques were similar between legs.Therefore, participants evaluated in our study did not present evident adaptations that influence torque production at the knee joint eliciting advantage for preferred leg over the contralateral.The differences observed between the joint angles and contraction velocities are expected in this type of assessment (Andersen et al., 2005).Additionally, discussion of such results was not among the purpose for this study and they are not further discussed here.Discussion on this topic can be found in the literature (Frykberg et al., 2006).
Limb preference involves genetic and environmental components that work towards the preferential use of a given limb.However, it is known that the environmental component has greater influence to the consolidation of limb preference (Teixeira, 2001).In this regard, training may contribute to onset of asymmetries, as the highest asymmetry index we observed was 12.6±5.3%,which is lower than the 15% suggested as a risk factor among athletes (Zabka et al., 2011).
When untrained individuals were evaluated for the torque -angle and torque-velocity relationships of elbow flexor and extensor torque, the only asymmetry reported occurred at the flexor isometric peak torque at the angle where highest peak torque was produced, in favor of the preferred limb (Carpes et al., 2012).The results were discussed considering that even among untrained individuals there is a preferential recruitment of the preferred arm for weight sustaining tasks, and for mostly of cases it happen with elbow flexed at 90º, angle where asymmetry was reported.Therefore, asymmetries observed for elbow flexors may rely on angle-specific muscle adaptations, which we have not observed for the knee joint in our study.
In general, lower extremity asymmetries have been useful to predict injury risk (Baumhauer,  ).An association between knee strength asymmetries and joint angle has been reported during jump tasks in college athletes (Kobayashi et al., 2013), for joint torques in long jumpers performing squat exercise (Kobayashi et al., 2010), and strength imbalances in athletes for both knee flexor and extensor muscle groups (Deli et al., 2011).During locomotion of healthy participants, kinematic asymmetries at knee were found (Maupas, Paysant, Datie, Martinet, & Andre, 2002), although no relationship between asymmetry and lateralization was found (Maupas et al., 2002).Additionally, isokinetic strength of knee flexors and extensors was not correlated with laterality (Maupas et al., 2002).

Discussion
Information on leg asymmetries in untrained individuals may be useful to describe a baseline condition prior to engagement in regular physical training.Such information may interest participants and coaches before starting participation in a routine of sport or physical training.The information on leg asymmetries can significantly help to minimize risk for injuries, as well as to understand changes in asymmetry indexes resultant of long- jumps, despite the similar isokinetic strength, significant correlations were observed for bilateral asymmetry index of isokinetic knee strength at 180°/s.Correlations were also observed when considering bilateral symmetry indexes for maximum flexion angle and mean knee joint torque during single-leg jumps (Kobayashi et al., 2013).Additionally, court sport players presented weaker preferred leg (determined by preference to kick a ball) when considering hamstrings working at low (60°•s −1 ) and high (300°•s −1 ) contraction velocities, whereas hamstrings to quadriceps ratio was significantly larger at preferred leg among field players (determined by preferential leg to kick a ball) at a contraction velocity of 60°•s −1 and in their non-preferred leg at 300°•s −1 (Cheung, Smith, & Wong del, 2012).These results reinforce the concept that sports training may lead to specific muscular adaptations, which may have a leg preference effect leading to asymmetries.
In summary, the main contribution of our study is to describe the characteristics of a group of untrained male considering knee isometric and isokinetic performance, to show that among untrained individual asymmetries were not observed.Such result is important when considering further engagement in regular exercises that may influence asymmetries as described in previous studies mentioned.Among the limitations in our study we highlight the small sample size and the consideration of only male individuals for the testing.

Conclusion
Bilateral torque production in healthy untrained participants show no significant asymmetries for torque-angle and torquevelocity relationships, work and power output for knee flexor and extensor muscle groups.Therefore, healthy individuals who are not involved in specific training for the lower extremity present no evidence of adaptation at the knee joint that may elicit advantage for the preferred leg over the contralateral leg.

Figure 3 .
Figure 3. Flexor (top) and extensor (bottom) work done during isometric contractions at different joint angles.