Subjects

Two hundred non-professional athletes were recruited for the present study. After the initial meeting, the inclusion criteria were explained. These criteria consisted of engaging in physical activity for at least 30 minutes per day, three days per week, and having no history of knee or ankle injury in the past 12 months. Following this meeting, 38 athletes withdrew from the study, resulting in one hundred sixty-two participants (75% males, 25% females). All participants completed surveys to assess their weekly and daily levels of physical activity, as well as their height and body mass (Table 1). The potential risks of the tests were verbally communicated to the participants, and they provided signed informed consent. The Ethics Committee of the University of Seville approved the study.

Procedures

The participants completed a Maximum Voluntary Contraction (MVC) test using the Biodex Multi-joint System (Shirley, New York). The exercise was knee flexion at 45° to assess the strength of the posterior thigh musculature and knee extension at 45° to assess the anterior thigh musculature. Following this, they underwent a directional change test with the dominant foot, which included two open cuts at 45° (SC₄₅) and 30° (SC₃₀), as well as a closed cut (SC-_45cl) on a force platform (Kistler 9260 AA6, Winterthur, Switzerland). Prior to the test, all participants underwent a standardized warm-up, which involved a 5-minute cycling session on a cycloergometer (Ergoline 900, Ergometrics, Bitz, Germany) at an intensity of 60 W (60 rpm). Additionally, they familiarized themselves with sports-specific sidecutting maneuvers through five to eight practice attempts. The velocity of the movements was regulated using a metronome set at 4-5.5 m/s, and the designated direction for the movements was indicated on the floor using tape.

Kinetics

A force plate (Kistler 9260 AA6, Winterthur, Switzerland) assessed the Ground Reaction Forces (GRF) in the vertical, mediolateral, and anteroposterior axes. Additionally, triaxial accelerometers (xyz-PLUX, PLUX-Wireless Biosignals, S.A., Lisbon, Portugal) were employed to measure accelerations at the knee and ankle joints. The knee accelerometer was positioned on the lateral condyle. In contrast, the ankle accelerometer was placed on the malleolus (Fig. 1). Antero-posterior (AP), Medial-Lateral (ML), and Longitudinal (Z) axes were calculated, and all signals were recorded at a sampling rate of 1000 Hz. Contact Time was determined during side-cutting by measuring the duration from the initial ground contact until the foot entered the flight phase.

Fig. 1 Location of accelerometers. 

Kinematics

An analysis of knee, ankle, hip, and trunk angles was conducted. Reflective markers were positioned on the lateral malleolus, lateral condyle, greater trochanter, and acromion to capture the movements precisely. The angles were calculated at two specific time points: first, during the initial contact between the foot and the ground, and second, when the Vertical Ground Reaction Force (VGRF) was recorded. Three video cameras (240 Fps) were positioned perpendicularly, one for each type of change of direction, at a distance of two meters from the central point of the force platform. Subsequently, the Kinovea software was employed for digitization and angle extraction by 2D analysis.

Statistical analysis

Statistical analysis was performed using the SPSS 22.0 software package (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test assessed the data distribution, and parametric variables were identified. Baseline data were compared using an independent-sample t-test. Potential statistical differences were evaluated using a repeated-measures two-way ANOVA (time x group). Mean values are reported with standard deviation (SD), and statistical significance was set at p ≤ 0.05.

The effect size (Cohen’s d) was calculated by dividing the difference between the means of the groups by the combined standard deviation of both groups, considering the sample size. Effect sizes of 0.2 were considered small, 0.5 moderate, and 0.8 large.

Kinetic

During the SC₃₀, subjects exhibited a higher value of VGRF of 2135.49 ± 633.14 N, which corresponds to approximately three times their body weight, while ML Force (625.44 ± 170.99 N) and AP Force (945.62 ± 421.53 N) were lower. For the SC₄₅, VGRF was lower compared to SC₃₀, with a value of 2015.83 (± 687.27 N), while ML Force was 566.87 (± 189.22 N) and AP Force was 765.75 (± 417.06 N). In SC_45cl, VGRF was 1790.60 (± 517.55 N), and ML Force exhibited the smallest force value among all measured changes of directions (515.41 ± 181.53 N), as well as in AP Force (670.37 ± 368.42 N). For more specific results, the force was normalized to each subject’s body weight for each directional change, as shown in Table 2.

Regarding the SC₃₀, the results showed the highest applied force in all axes, including VGRF, AP Force, and ML Force. The gender analysis in Table 2 revealed that during the SC₃₀ directional change, women exhibited significantly higher force values than men in ML Force (p = 0.045). On the other hand, in the SC₄₅, men demonstrated higher values than women in both VGRF (p = 0.025) and AP Force (p = 0.020). Finally, in SC_45cl, significant differences were only observed in the force exerted in AP Force, where women displayed a higher force value than males (p = 0.020).

The analysis of partial knee and ankle accelerations during changes of directions is presented in Table 3. The SC₃₀ exhibited the lowest ACC_KNEEAP (1.72 ± 1.48 g). Similarly, ACC_ANKLEZ had the lowest values compared to the other changes of directions (4.11 ± 0.81). For the SC₄₅, both ACC_KNEEML (2.39 ± 1.58 g) and ACC_KNEEZ (1.53 ± 1.30 g) showed the lowest values. However, both ACC_ANKLEML and ACC_ANKLEAP were higher tan in the other changes of directions (4.23 ± 0.75 g and 4.05 ± 0.92 g, respectively). During the SC_45cl, the highest ACC_KNEEML (2.49 ± 1.49 g), ACC_KNEEZ (1.69 ± 1.40 g), and ACC_ANKLEZ (4.20 ± 0.79 g) were found. Conversely, ACC_ANKLEML and ACC_ANKLEAP were lower during this directional change (4.14 ± 0.88 g and 3.98 ± 0.99 g, respectively).

On the other hand, gender-specific acceleration was also evaluated (Table 4). During the SC₃₀, although no significant differences were observed, partial knee accelerations were higher in women than men. In the ankle, only ACC_ANKLEZ showed higher values in women than men (4.20 ± 0.76 g). In the case of SC₄₅, there was a trend towards higher acceleration in women compared to men, both in the knee and ankle. Regarding ACC_KNEEZ, the difference between women and men was significant (p=0.002), with women displaying an acceleration of 1.82 ± 1.54 g compared to 1.44 ± 1.21 g exhibited by men.

During the SC_45cl, no significant gender differences were found, except for ACC_KNEEML where men exhibited higher values (2.49 ± 1.48 g). In the remaining partial knee accelerations, women showed higher values for this parameter, although they were not statistically significant. Regarding the ankle, only a trend towards higher values in men than women was observed for ACC_ANKLEZ (4.22 ± 0.81 g), although statistical significance was not reached.

Kinematics

Kinematic factors were also studied during changes of direction. Knee flexion angles at initial contact (KneeFlexInit) and maximum knee flexion angle (KneeFlexMax) were analyzed. The range of ankle dorsiflexion (AnkleDorsiRange), initial hip flexion angle (HipFlexInit), maximum hip flexion angle (HipFlexMax), initial trunk flexion angle (TrunkFlexInit), and maximum trunk flexion angle (TrunkFlexMax) were also analyzed.

Regarding the changes of directions (Table 5), the highest value of KneeFlexInit was observed in SC₃₀ (144.6° ± 7.6°), while the lowest value of KneeFlexMax was found during SC₄₅ (125.2° ± 8.9°). For the hip, the maximum value of HipFlexMax was 60.6° during SC₃₀, while the trunk flexed 92.9°. Finally, a range of ankle dorsiflexion of 39.4° was found between the initial contact of the directional change and the maximum plantar flexion during the change.

In SC₄₅, KneeFlexMax was 125.2° (extension), similar to the value of 123.3° in SC₃₀. At that moment, HipFlexMax was 51.21°, while TrunkFlexMax showed a flexion of 95.1°. Additionally, the AnkleDorsiRange was 41.7°, defined as the amount of flexion from the first contact until maximum flexion occurred.

In SC_45cl, KneeFlexMax was 119.9°, representing the most significant flexion of the three change types. From initiating contact to maximum knee flexion, the hip flexed by 13.5°. As for the trunk, it exhibited a flexion of 77.8° at that moment. Finally, the AnkleDorsiRange was 41.8°.

Finally, a gender analysis of kinematics during direction changes was performed (Table 6). It was observed that during SC₃₀, women exhibited a significantly higher AnkleFlexInit than men (p=0.040). However, gender differences during SC⁴⁵ were found in AnkleFlexMax (p=0.004) and AnkleDorsiRange (p=0.007). Lastly, during SC_45cl, gender differences were found in KneeFlexInit (p<0.001) and KneeFlexMax (p=0.006), where men displayed a lower angle compared to women.

Kinetics

During changes of directions, the VGRF relative to body weight was significantly higher in SC₃₀ (29.49 N/kg ± 7.75) compared to SC₄₅ (27.70 N/kg ± 7.81; p=0.004) and SC_45cl (25.13 N/kg ± 7.59; p<0.001). One possible explanation for these differences is that greater force needs to be exerted in the medial-lateral (ML) axis to perform

more forceful or larger amplitude changes (remember that 30° is measured above the horizontal plane). According to Nigg’s paradigm ¹¹, which compiled studies spanning over 25 years, it became evident that VGRF plays a particularly important role in injuries among individuals engaged in physical activities such as running, especially in situations where pronation is accentuated, such as during changes of directions. This possibility is also supported by the values of ML force found in our study, as the analyzed subjects displayed significantly higher values during SC₃₀ (8.64 N/kg ± 2.08) compared to SC₄₅ (7.78 N/kg ± 2.08, p<0.001) and SC_45cl (7.17 N/kg ± 2.40, p<0.001). Therefore, higher levels of force, especially in the ML axis, which need to be attenuated by the lower extremities, suggest the need for increased preventive strategies in these types of actions common in various sports. It is essential to highlight that these 30°-changes of directions do not only increase the ML force but also the AP force (12.9 N/kg ± 5.32) when compared to SC₄₅ (10.54 N/kg ± 5.39, p<0.001) or SC_45cl (9.35 N/kg ± 5.09, p<0.001), suggesting that biomechanical modifications in the different involved joints occur in response to these increments, as will be discussed later.

Despite numerous studies regarding the biomechanical and kinetic factors manifested during direction changes, the results are contradictory ¹². The aspect that seems most relevant in explaining this disparity of results may be the variety of procedures carried out in different studies. Brughelli et al.¹³ differentiate between anticipated and unanticipated changes of directions, changes towards the dominant or non-dominant foot, as elements that influence the final results. Regardless of the methodological variety, it is evident that both VGRF, ML -force, and AP-force are variables closely related to the on-set of injury processes ² and should be controlled and attenuated as much as possible.

While it is evident that higher forces in both axes are associated with a greater risk of injury in this population group, it is essential to mention the differences observed in the different variables according to gender. Sigward and Powers ¹⁴ suggested that an increased risk of knee injury through increased valgus can occur due to greater forces in either axis in both men and women. In fact, their work reflected values that were very similar to those found in the present study regarding VGRF during changes of directions. Our results show that during SC₃₀, where women sustained higher relative loads than men (p=0.045), there were significant differences in the force exerted in the ML axis. This finding has been widely observed in studies with athletes, where women have shown significantly greater valgus than men ¹⁵. Differences in the ML axis between men and women could be related to an increased risk of injury in females due to the biomechanical position of the knee in abduction or adduction during ground contact ^1,16.

On the other hand, during SC₄₅, the differences were found in VGRF (p=0.025) and AP force (p=0.020), where men significantly exceeded women. Our results are consistent with those of Liu et al. ¹⁷, who related a reduction of VGRF and ForceAP with a reduction of the load supported by the knee during open direction changes. However, authors such as James et al. ¹⁸ found the opposite during maximum speed changes of directions. One possible explanation for these differences could be the openness of the directional change at a non-maximal speed. SC₄₅ may involve less demand than SC₃₀, so being a more vertical change and closer to straight-line running, it is possible that AP force is increased, as well as VGRF in men.

Another important finding was that women exhibited a significantly higher AP force during SC_45cl (p=0.020) than men. Our results align with those presented by McLean et al. ², who found higher AP forcé in women compared to men during a directional change with an angle of exit between 35° and 55° (women = 1.80BW [± 0.54]; men 1.54BW [± 0.76]). Moreover, although not statistically significant, women in our study also showed higher VGRF during SC_45cl. Recently, de Hoyo et al. ¹⁹ reported very similar data in male athletes who performed a closed change with a 60° opening (24.1 N/kg ± 8.4). Unfortunately, these authors only conducted their intervention with male subjects, making a comparison impossible. Nevertheless, there is limited research on the role of gender in VGRF during closed changes of directions. This is the first study to assess vertical force in men and women during closed changes.

What seems evident is the increased risk of knee injury through ML force values during changes of directions based on gender ¹. Authors like Sigward and Powers ¹⁴ indicated that women who exerted greater ML force also exhibited a greater valgus moment, which increased the risk of a knee injury. Although our results do not provide data on impulse outcomes, a possible contribution can be inferred from the partial knee and ankle acceleration. As described previously, women reported significantly higher ML force during SC₃₀ than men and a slightly higher trend was observed in this variable during SC_45cl. In this regard, it can also be observed that women experience 18.53% more ML knee acceleration than men during SC₃₀ and 19.93% during SC₄₅, with almost no difference (0.4%) during SC_45cl. This ML knee acceleration could represent valgus moments, although this fact should be corroborated in future studies through a kinematic analysis in the frontal plane.

There were significant differences between the different changes of direction. ACC_KNEEAP was significantly higher in SC_45cl compared to SC₃₀ (p=0.006). This finding suggests an increased risk in closed changes compared to open changes, as the load is mobilized more rapidly in the sagittal axis, potentially leading to an increase in tibial translation. However, ForceAP was significantly lower in SC_45cl than SC₃₀ (p<0.001). These differences also existed between open changes, as ACC_KNEEAP was significantly higher in SC₄₅ than SC₃₀ (p=0.023). One posible explanation for these differences could be the biomechanical demand, with SC₄₅ being less demanding than SC₃₀, resulting in a faster strategy during less open changes. Analyzing ACC_KNEEML between changes of directions, SC₃₀ showed a significantly higher acceleration than SC₄₅ (p<0.001) and SC_45cl (p<0.001). Sigward and Powers ¹⁴ suggest that frontal plane control is necessary to reduce the relative risk of knee injury. Our data indicate that SC₃₀ poses an increased risk compared to the other changes of directions, characterized by higher ForceML and ACC_KNEEML, which are associated with an increase in knee valgus.

Regarding gender differences during changes of directions, significant differences were found only in SC₄₅ (p≤0.05), with women exhibiting higher ACC_KNEEZ than men. One possible explanation could be that Chapell et al.²⁰ suggested that women experience greater shear force in the tibia during ground contact. This negative strategy indicates that energy absorption by women is less efficient than in men, or at least faster, which may be related to an increased risk of knee injury. Surprisingly, this trend is not observed in SC₃₀, which would be expected to be more demanding at this level. However, the trend still favored women, who presented 14.67% more ACC_KNEEZ during SC₃₀ and 11.35% more during SC_45cl.

Kinematics

It is evident that a knee angle close to the extension increases the risk of injury ^20-22. Our results indicate a higher KneeAnglein during SC₃₀ compared to the other changes of directions (144.63° ±7.65). However, the maximum KneeAngle reached is highest during SC₄₅ (125.16° ±8.91). Concerning the 180° representing full knee extension, the KneeAngle during SC₃₀ averaged 35.37°. Markolf et al. ²³ mentioned that knee flexion of 40º or less increases tibial translation supported by the ACL and, consequently, the tension sustained by the joint complex, leading to an increased risk of injury. A preventive strategy should involve increasing knee flexion to absorb impact. Cochrane et al.²⁴⁾ went further by identifying angles below 30º of flexion during initial contact as a critical factor in ACL injuries during changes of directions. Consistent with Markolf et al. ²³, our results showed that for all changes of directions, the KneeAnglein exceeded 30° (SC₃₀= 35.37°; SC₄₅= 36.55°; SC_45cl = 38.57°) but remained below 40°, suggesting that participants subjected the joint to excessive stress, which could partially explain the previously reported increases in force.

Despite the relationship between knee flexion angle and the relative risk of joint injury, it is evident that the biomechanical strategy during direction changes is not exclusive to the knee. Several studies have attributed a determining role to other joints reflecting greater or lesser knee flexion. Imwalle et al.²⁵ confirmed a direct influence of the hip on the knee during direction changes. Potter et al.²⁶ stated that open and closed changes of directions influence the biomechanical strategies manifested by subjects, with significantly greater hip flexion in open changes than in closed changes. However, our results do not align with those reported by these authors, as hip flexion was practically the same regardless of the type of directional change. One possible explanation could be that kinematic aspects were evaluated in an agility circuit with open and closed changes of direction, where athletes performed different direction changes depending on their own technique.

Moreover, gender-based differentiation in knee kinematics has been observed, showing different biomechanical strategies related to an increased risk of knee injury between men and women ²⁷. In our study, during SC₃₀, women exhibited a KneeAnglemax (in extension) of 33.4°, although without significant differences compared to men (p = .080), who exhibited an angle close to 36°. Significant differences were found during SC_45cl, where women had a KneeAnglein close to 146° and men to 140°. Additionally, significant gender differences were found in Knee-Anglemax (p = 0.006), with women exhibiting more significant knee extension than men, indicating a higher injury risk. These results are consistent with those reported by James et al. ¹⁸, who conducted a comparative study by gender during closed direction changes to 60° and found a KneeAnglein in women 5.8° lower than in men. In our study, during SC_45cl, women flexed 6.4° less than men, with significant differences compared to men. This indicates that the directional change technique employed by women may be associated with an increased risk of knee injury during closed direction changes. Furthermore, in the study mentioned above, the researchers found that women significantly exceeded men in KneeAnglemax, with greater extension (64.1° and 61.4°, respectively). Our results are very similar, with men exhibiting an average of 61.8° and women 54.4°.

Regarding open changes of directions, a slightly higher, but not significant, trend in extension can be observed in women compared to men during SC₃₀. Additionally, during SC₄₅, the KneeAnglein values are practically the same in men and women, with an increase in knee extension in women at the moment of VGRF. Dai et al. ²⁸ evaluated knee kinematics in men and women during SC₄₅ and also found greater extension in women compared to men at the moment of VGRF. A greater range of knee flexion during the first contact could serve as a biomechanical strategy to absorb impact and reduce the load. In this sense, women have a higher relative risk of knee injury than men, as they exhibit more significant knee extension during changes of direction ²⁹.

In conclusion, the study highlights the importance of controlling GRF and biomechanical strategies during changes in direction to reduce the risk of knee injuries. Significant gender differences in forces and accelerations during changes of directions were observed, underscoring the need to consider specific gender factors in injury prevention. Variables such as VGRF, ForceML, and ForceAP, as well as acceleration in the vertical axis, appear to be deterministic kinetic variables that explain these gender differences and knee flexion angles. The results can be helpful in informing the design of preventive strategies in the sports field and improving the understanding of the underlying mechanisms of knee injuries during changes of direction.