The influence of the lower limb components on genu varum in football players: a full leg length magnetic resonance imaging study

Abstract

Background

This study aimed to evaluate lower extremity alignment in football players with and without genu varum using magnetic resonance imaging (MRI) and to investigate the underlying mechanisms and contributing factors to malalignment.

Methods

This prospective case-control study included 36 male football players aged 16–19 years, divided into two groups: 18 with genu varum and 18 controls with normal lower extremity alignment. Full-length lower extremity MRI was used to assess alignment parameters. The isokinetic strength of the concentric knee extensor-flexor and concentric hip abductor-adductor muscles was measured using an isokinetic dynamometer at angular velocities of 60°/sec and 180°/sec. Logistic regression was used to evaluate the risk factors for genu varum.

Results

Genu varum group had a mean mechanical axis deviation (MAD) of 14 ± 5 mm (p < 0.001), with 11 players exceeding the clinical cutoff of 15 mm. Significant differences were observed in the lateral distal tibial angle (LDTA) (p = 0.014), lateral proximal femoral angle (LPFA) (p = 0.017), and medial distal femoral angle (mLDFA) (p = 0.002) between the groups. Muscle strength values were comparable between the groups, except for the hip adductor-abductor strength ratio at 60°/sec, which was significantly lower in the genu varum group (p = 0.008), while all other comparisons were non-significant (p > 0.05). The regression analysis demonstrated that the mechanisms responsible for varus alignment in football players differ between the dominant and non-dominant leg.

Conclusions

The findings in this study suggest that the proximal tibial deformity is a key factor in malalignment among football players with genu varum. Differences in alignment were observed between dominant and non-dominant legs. Strength values were similar between players with and without varus alignment, except for the 60˚/sec angular velocity Add/Abd ratio.

Clinical trial registration

NCT06606964 / 16.09.2024.

Level of evidence

Level III.

Introduction

Genu varum, a common malalignment among both youth [1] and veteran football players [2], is associated with a higher prevalence of knee injuries [3,4,5] such as anterior cruciate ligament (ACL) disruptions [6], meniscal tears, patellofemoral pain syndrome [7], and shin splints [8]. Prior research has demonstrated that athletes engaged in football are at an elevated risk of developing genu varum compared to their counterparts in other sporting disciplines [1, 3, 9]. Genu varum can have a deleterious effect on an athlete’s physical capabilities, disrupting both static and dynamic balance through alterations in the gravitational axis of the lower extremities. Furthermore, genu varum can result in damage to the tibiofemoral articular cartilage, thereby increasing the risk of developing knee osteoarthritis in later life [9, 10]. It is, therefore, important to understand the etiology of genu varum and implement protective and corrective measures for football players.

In recent years, researchers have conducted studies to investigate the relationship between genu varum and participation in football. Various measurement techniques have been employed to analyze the lower extremity alignment, including calipers, goniometers, and photographic methods. However, these methods have inherent limitations, which make it challenging to examine the underlying etiological factors and identify differences in anthropometric components [3, 11,12,13]. In a study conducted by Colyn et al. [4], a comparison was made between football players and other athletes and non-athletes using digital radiology measurements. The findings indicated that male football players exhibited genu varum, with the proximal tibia identified as the primary factor. Similarly, Krajnc and Drobnič [14] investigated lower limb alignment in asymptomatic adult professional football players and identified the proximal tibia as the primary source of genu varum. It is currently unclear which specific movement patterns in football are responsible for malalignment between football players with similar years of participation in the sport. However, directional changes [15], cleated shoes increasing varus stress [16], and neuromuscular fatigue reducing knee stability and raising injury risk [17, 18] are potential contributing factors. Nevertheless, it has been proposed that further research is necessary to gain a comprehensive understanding of the underlying mechanisms. In this context, advanced imaging techniques play a crucial role in obtaining detailed insights into anatomical and functional characteristics. Evaluation of the lower extremity for genu varum using magnetic resonance imaging (MRI) enables a comprehensive assessment of all components without exposing individuals to any radiation, making it a valuable tool in both clinical and research settings.

Moreover, Witvrouw et al. [13] postulated that an imbalanced strength distribution between the adductor and abductor muscles may precipitate genu varum in football players. This is because kicking, a frequent activity in football, often results in increased strength of the adductor muscles, which can lead to a change in the average adductor/abductor strength ratio. However, the authors highlight that no data on the adductor/abductor strength of football players is currently available in the literature. The authors recommend that further research be conducted to address this knowledge gap.

The asymmetrical movements in football, resulting from the inherent difference between the kicking and supporting legs, may lead to the formation of genu varum in the dominant and non-dominant legs through different mechanisms. It is important to note that not all football players demonstrate varus alignment of the knees. Consequently, to ascertain the mechanisms that contribute to the formation of genu varum, it is essential to conduct a comparative analysis between football players exhibiting and those lacking a varus deformity. Such a comparison can assist in identifying the mechanisms that cause genu varum in football players and the factors that are likely to contribute to these mechanisms. The study had three objectives: (i) to evaluate lower extremity alignments in football players with and without genu varum using MRI, (ii) to determine the underlying mechanisms and contributing factors of malalignment, and (iii) to investigate the relationship between lower extremity alignment and isokinetic strength.

Materials and methods

Patients and study design

The study was a prospective case-control study conducted between 2021 and 2023 at Akdeniz University in collaboration with the Faculty of Sport Sciences and the Faculty of Medicine. The study was post-registered to ClinicalTrials.gov (NCT06606964).

The study participants were selected from otherwise healthy male football players aged 16–19 who currently play for a professional football club, have at least five years of football experience, train five days a week, and participate in official matches. Subjects with a history of musculoskeletal disease, postural disorders, previous surgery for lower extremity sports injuries, or a history of fractures of the long bones of the lower extremity that could cause lower extremity malalignment were excluded from the study. The present study focused exclusively on male soccer players in order to minimize the confounding effects of sex- and age-related changes in bone structure and hormonal influences, muscle strength, lower limb profile, and size. In addition, subjects with a body mass index (BMI) outside the normal range were excluded (18.5 < BMI < 24.9). Players who met these criteria and volunteered to participate were assessed for lower extremity alignment using differences in tibiofemoral joint distance (ICD-IMD). Football players with an ICD-IMD of 40 mm or greater were considered to have genu varum. Participants were matched for age, height, body weight, and year of participation in football and divided into two groups: those with genu varum (Genu varum group; n: 18) and those with normal lower extremity alignment (Control group; n: 18). The study was approved by the Akdeniz University Clinical Research Ethics Committee (KAEK-917 / 09.12.2020). The study adhered to the principles outlined in the Declaration of Helsinki. Volunteers who agreed to participate in the study were provided verbal and written information about the research protocol. Informed consent was obtained from the volunteers and their families.

Sample size calculation

A priori power analysis was conducted using G*Power 3.1 [19]., based on the study by Krajnc and Drobnič [14], to determine the required sample size. The effect sizes employed in the sample size calculation were derived from the aforementioned study, specifically for the medial proximal tibial angle (Cohen’s f = 1.694) and the hip-knee-ankle (HKA) angle (Cohen’s f = 1.749). These effect sizes indicated a minimum of 9 and 8 participants per group, respectively, to achieve a margin of error of 0.05 and a power of 0.95. To account for the potential non-parametric distribution of the data and possible participant dropout, the sample size was increased to 18 participants per group.

Clinical assessments

Weight, height and body density were measured using air displacement plethysmography (BOD POD) and a stadiometer (Holtain 98.602VR stadiometer, UK). The Brozek [20] equation (percent fat equation: %fat = (457 / body density) − 414.2.) was used to convert each participant’s body density to body fat percentage. Body mass index (BMI) was calculated as body weight in kg divided by height in meters squared (kg/m²). Measurements were performed according to the American College of Sports Medicine (ACSM) protocols [21].

To calculate the ICD-IMD, the distance between the medial edges of the condyles and malleoli was measured in millimeters using a digital caliper with a precision of 0.01 mm. The measurement was taken in a standing position with the hips and knees fully extended and the medial condyle or malleoli touching [3]. If the femoral condyles and medial malleoli touched, then ICD-IMD was defined as zero. However, if a measurable intermalleolar distance was present, it was expressed as a negative value in mm [9].

Radiological assessment of the alignment of the lower limbs

Several methods are available to evaluate lower extremity alignment parameters [3, 13, 22, 23], including standing radiographic evaluation (ortho-roentgenography), which is considered the most accurate method. However, due to ethical concerns, it is not appropriate to expose healthy children and young people to radiation [1, 22, 24]. As an alternative, full-length lower extremity MRI offers a radiation-free and non-invasive alternative with superior visualization of soft tissues and bony structures compared to radiographs, making it ideal for evaluating lower limb alignment in young athletes. Consequently, the MRI was utilized in this study to assess lower limb alignment parameters.

To maintain a neutral foot and ankle position and prevent natural foot drop during imaging, a sponge cushion was used to stabilize the ankle, ensuring proper alignment. This setup also ensured that the patellae were parallel to the imaging plane, similar to standing ortho-roentgenography.

To assess lower extremity alignment, several measurements were taken, including mechanical axis deviation (MAD), mechanical lateral proximal femoral angle (mLPFA), mechanical lateral distal femoral angle (mLDFA), medial proximal tibial angle (MPTA), lateral distal tibial angle (LDTA), and joint line convergence angle (JLCA). The study measured various alignment parameters, including hip-knee-ankle angle (HKA), quadriceps femoris angle (Q angle), tibial torsion (TT), femoral anteversion (FA), tibial tubercle-trochlear groove (TT-TG) distance, and tibial tubercle-posterior cruciate ligament (TT-PCL) values [25] (Fig. 1). Separate measurements were taken for each leg, as shown in Table 1. A radiologist assessed the MRI scans with 12 years of MRI experience who was blinded to the participants’ group assignments.

Fig. 1

MRI measurements

Table 1 Definition of lower limb alignment measurements

MRI protocol

The MRI examinations were conducted using a 1.5 Tesla closed MRI system (Magnetom Avanto, Siemens, Erlangen, Germany) with a body and spine coil. The participants’ knees were fully extended, and their feet were neutral during imaging. The scan was performed from the anterior superior iliac spine to the tip of the toes, and both legs were included in the field of view, as described by Diederichs et al. [26]. A turbo-spin echo (TSE) sequence was used for image acquisition due to its lower sensitivity to susceptibility artifacts and shorter acquisition time. The bandwidth was set to 500 Hz/pixel to reduce chemical shift artifacts that can distort measurements of bony structures’ length and angle. This also reduced the echo time (TE) and repetition time (TR) to 9 ms and 872 ms, respectively. The TSE factor was set to 8 to increase acquisition speed. The slice thickness (SL) was 8 mm, and a 128 × 256 matrix was used to produce the image. The in-plane resolution was 3.1 × 1.6 mm. The phase encoding direction was from foot to head, and the frequency encoding direction for the field of view (FOV) was set to 400 mm. The phase encoding direction was reduced to 200 mm, and 100% phase oversampling was used to avoid wraparound artifacts [27]. To create true coronal images, 20 images were used, each covering 176 mm with a 10% gap in the slice encoding direction. The imaging process involved obtaining three consecutive imaging positions from the hip joint downwards to the ankle level, with incremental and automatic table repositioning to avoid overlap or gaps. The parameters were selected to ensure an adequate signal-to-noise ratio with short data acquisition. The total imaging time for all coronal images was 2 min and 57 s. Additional axial images were acquired along the femoral head, neck, knee, and ankle using the same sequence parameters as the coronal images. The total imaging time, including proper patient positioning, was 10 min.

Isokinetic strength measurements

The strength of the concentric knee extensor-flexor and concentric hip abductor-adductor muscles was measured using an isokinetic dynamometer (IsoMed 2000; D&R Ferstl, Hemnau, Germany) at two different angular velocities: 60˚/sec and 180˚/sec. Three repetitions were performed at 60˚/sec, and five repetitions were performed at 180˚/sec. To calculate the H/Q ratio, we divided the peak concentric torque of the hamstrings by that of the quadriceps during the same contraction velocity. Similarly, to determine the Add/Abd ratio, we divided the peak concentric torque of the adductors by that of the abductors during the same contraction velocity. Before the testing, participants completed a general warm-up phase. This phase included a 10-minute stationary bicycle ergometer warm-up (TechnoGym, Italy) at a pedal speed of 70 rpm, followed by a stretching exercise specifically designed for lower limb muscles [28].

During the isokinetic test, participants received verbal encouragement to perform to the best of their ability. Before each measurement session, the dynamometer was calibrated according to the study manual. All tests were conducted on the participants’ dominant leg, which was considered a better indicator of their maximum muscle strength. The dominant leg was defined as the limb they preferred to kick a ball as hard as possible, as described by de Marche Baldon et al. [29]. To standardize the data based on body weight, the peak torques were divided by body weight and presented as (peak torque/body weight) Nm/kg.

Statistical analysis

Continuous variables were presented as means and standard deviations. Categorical variables were summarized as frequencies and percentages. Various methods were used to check if the data had a normal distribution, including analyzing kurtosis-skewness values, histograms, normal Q-Q plots, and box plots, followed by a Shapiro-Wilk test. Based on these analyses, it was concluded that the data exhibited a normal distribution. The independent sample t-test was used to analyze the differences between the groups. Additionally, Cohen’s d effect size was calculated. The effect sizes were classified as small (d = 0.20), medium (d = 0.50), and large (d = 0.80) based on Cohen’s [30] recommendations.

To investigate whether the alignment parameters used to evaluate lower extremity alignment in football players are risk factors for the development of genu varum, we created a logistic regression model. Logistic regression model assumptions, including multicollinearity (variance inflation factor < 10) and linearity of the logit, were validated. The model included the variables that were found to be correlated in the Pearson correlation analysis. All statistical analyses were performed using Jamovi software (version 2.3.18). The statistical significance level was set at p < 0.05, p < 0.01, and p < 0.001. Intraclass correlation coefficients (ICC) were calculated using three repeated measurements for each leg in 10 randomly selected participants. Both intra-observer and inter-observer reliability were assessed. Koo and Li [31] defined poor reliability as values below 0.5, moderate reliability as values between 0.5 and 0.75, good reliability as values between 0.75 and 0.9, and excellent reliability as values above 0.9.

Results

There were no statistically significant differences between the genu varum group and the control group since they were matched in terms of age (16.83±1.14 vs. 16.96±1.06), height (179.97±6.25 vs. 179.50±6.86), body weight (69.39±7.62 vs. 69.50±5.76), and year of participation in football (8.83±1.54 vs. 8.61±1.38) p > 0.05.

Table 2 Comparison of lower extremity alignment evaluation parameters between genu varum and control groups

The LDTA of the genu varum group was lower in both the dominant (86.21 vs. 87.91) and non-dominant (86.17 vs. 89.11) leg than in the control group, while HKA was higher (4.20 vs. 1.56; 4.10 vs. 1.95, respectively for the dominant and non-dominant leg). Similarly, the MAD values were higher in both the dominant (14.09 mm vs. 1.26 mm) and non-dominant (14.29 mm vs. 4.62 mm) legs compared to the control group. Notably, 11 out of the 18 football players exhibited MAD values exceeding the clinical threshold of 15 mm.

It was found that some of the leg alignment variables of the genu varum group and control group differ regionally for the dominant and nondominant legs. The mLPFA of the genu varum group was significantly lower in the dominant leg (89.18 vs. 91.60) than in the control group, whereas it was similar (90.03 vs. 90.44) in the non-dominant leg. Furthermore, the mLDFA of the genu varum group was significantly higher in the dominant leg (88.24 vs. 86.47) than in the control group. In contrast, it was similar (87.82 vs. 87.11) in the non-dominant leg. However, the MPTA of the genu varum group was lower (83.51 vs. 85.35) in the non-dominant leg than in the control group, whereas it was similar (84.56 vs. 85.54) in the dominant leg (Table 2).

No significant differences were found between the two groups regarding knee flexion and extension at both 60°/sec and 180°/sec angular velocities. Additionally, no significant difference was observed in hip abduction and adduction isokinetic strength at 180°/sec angular velocity (p > 0.05). However, the hip abduction muscle strength of the genu varum group was found to be significantly higher than that of the control group at an angular velocity of 60°/sec (p < 0.05; ES: 0.76) (Fig. 2). Moreover, the Add/Abd ratio of the genu varum group was significantly lower than that of the control group (p < 0.01; ES: -1.01) (Fig. 3).

Fig. 2

Isokinetic strength values of knee and hip

Fig. 3

Isokinetic strength ratios of knee and hip

Table 3 Logistic regression model analyses

Table 4 Predictive accuracy of regression models

The logistic regression model demonstrated excellent goodness of fit, with Nagelkerke’s R² = 0.80. (Fig. 4.) The ROC curve analysis revealed an AUC of 0.96, sensitivity of 83%, and specificity of 89% for Model I and an AUC of 0.94, sensitivity of 89%, and specificity of 89% for Model II. An increase of one mm in MAD in the dominant leg leads to a 1.63 times higher probability of genu varum. Similarly, a one-degree increase in mLPFA and mLDFA increases the probability of genu varum by 1.90 and 2.17 times, respectively. The logistic regression analysis revealed that Model II was statistically significant (Pseudo R² = 0.72, p < 0.001). The analysis indicates that a one-degree increase in MPTA in the non-dominant leg increases the probability of genu varum by 1.28 times. Similarly, a one-degree increase in LDTA increases the probability of genu varum by 1.90 times (Table 3). When analyzing Table 4, Model I had a prediction rate of 83.3% for varus malalignment, while Model II had a rate of 88.9%. (Table 4).

Fig. 4

The logistic regression models

Discussion

The principal objective of the study was to achieve a comprehensive understanding of lower limb alignments in football players, with a specific focus on the assessment of alignment in players with genu varum and those without it. Moreover, the study sought to ascertain the etiology of malalignment and its contributory factors and examine the relationship between lower limb alignment and isokinetic strength. The findings of this study suggest that the proximal tibia is the primary cause of malalignment in football players. This agreed with previous research findings indicating that the development of genu varum is primarily associated with proximal tibial deformities [1, 14]. However, the results of our study indicate that football players with genu varum exhibit a different alignment in their dominant leg compared to their non-dominant leg. Although the femoral angles were within the normal range for both the control and genu varum groups, there were statistically significant differences between the two groups concerning their dominant leg. On the contrary, Krajnc and Drobnič [14] did not find angular differences between right and left or dominant and non-dominant legs in professional football players. However, studies demonstrate that there are significant differences in alignment, strength and biomechanics between dominant and non-dominant legs. Nakahira et al. [32] demonstrated kinematic differences in the knee valgus angles of the dominant and non-dominant legs in female football players during a single-leg vertical jump. Furthermore, the dominant leg is generally more flexible and stable [33], whereas the non-dominant leg may have less range of motion and stability [34]. The asymmetrical loading of the dominant leg during kicking and stance mechanisms in football may contribute to the observed differences in alignment parameters. Therefore, angular differences between dominant and non-dominant legs can be expected.

In the genu varum group, both legs demonstrated MPTA values below the clinical cutoff, indicating a significant difference, while the control group angles measured within the reference range reported by Paley [25]. Hence, only the non-dominant leg demonstrated a statistically significant discrepancy compared to the control group. Additionally, the regression analysis demonstrated that the mechanisms responsible for varus alignment in football players differ between the dominant and non-dominant leg. Our findings indicate that the magnitude of tibial deformation varies depending on the leg used for stance during kicking. The uneven distribution of body weight on the knee compartment of the stance leg during kicking may account for the differences observed in the non-dominant leg, rather than specific football actions.

Colyn et al. [4] suggested that playing football, especially during adolescence, significantly increased the level of genu varum, and this increase was primarily due to changes in the proximal tibia with lower MPTA values in football players compared to non-athletes and other athletes. Krajnc et al. [35] used diffusion-weighted MRI to examine the growth plates in young football players and concluded that playing football has a significant effect on the growth plates, especially the medial proximal tibia, which may result in more frequent knee varus deformities in football players. However, there was no distinction made between the dominant and nondominant legs in either study. Further research is necessary to establish a causal relationship between the non-dominant leg and the development of genu varum in football players. In this context, a more comprehensive perspective is needed, considering the potential impact of bilateral strength asymmetry on lower extremity malalignment.

In football, shooting is a distinct and precise action. Biomechanical research suggests that the kicking motion involves components of adduction, hip flexion, and knee extension. This repetitive movement in football has been proposed as a potential contributing factor to the development of genu varum [1, 3]. Additionally, some studies [36,37,38] have suggested that the significant amount of shooting in football may contribute to stronger adductor muscles in football players. However, these studies do not provide any clear evidence of strength differences among football players and any definitive association with genu varum. Therefore, it remains unclear whether the adduction/abduction strength imbalance disrupts varus alignment or if the strength imbalance occurs because of disrupted varus alignment. The findings of our study demonstrated that the hip adductor strength at an angular velocity of 60°/s was statistically lower in the genu varum group compared to the control group, whereas no significant differences were observed in isokinetic strength measurements at 180°/s. The significant differences at 60°/s but not 180°/s suggest that lower extremity alignment affects maximal strength more than power or dynamic performance. 60°/s primarily measures maximal strength, which is more influenced by musculoskeletal alignment, while 180°/s assesses power and endurance, relying more on neuromuscular efficiency. At higher speeds (≥180°/s), players may compensate for alignment deficits through muscle recruitment and movement adjustments, whereas at lower speeds, these compensatory mechanisms are less effective, making structural strength differences more evident [39, 40]. Due to the lack of sufficient evidence in the literature regarding the relationship between strength and genu varum, further studies are needed on this topic, as these findings suggest that the role of strength imbalances in contributing to malalignment remains a subject of ongoing debate.

Strengths and limitations

Previous studies have investigated the correlation between football participation and genu varum, or the differences in lower limb alignment between football players and non-athletes. These studies have found that genu varum is highly prevalent in football players, but there are also football players who do not have this knee malalignment. However, to the best of our knowledge, there is no research that evaluates lower leg alignment in football players using MRI according to their dominant and non-dominant legs and examines the impact of isokinetic strength on this alignment. It is important to note that this study is limited in scope, as it only focuses on malalignment in young male football players and the factors contributing to this condition. Further research is required to examine whether female football players have comparable misalignments due to pelvic structural variations. Although the sample size was determined through power analysis, its relatively small size may limit the statistical power and generalizability of the findings. The influence of playing position on lower limb loading and alignment was not controlled and remains a potential confounding factor. One of the limitations is that, in contrast to the standing position, we were unable to evaluate the impact of body weight on the lower limb sequence while using MRI in the supine position. Using MRI in a supine position may not replicate weight-bearing alignment, but it ensures consistent foot positioning and eliminates variability due to improper stance. Nevertheless, a study that evaluated leg alignment using full-leg standing X-rays reported that it was more appropriate to use tomography or MRI in the lying position for patients with particularly high varus values, as they were unable to get the proper lower extremity position during imaging [14]. Therefore, further investigation into this limitation of the imaging approach is required. MRI should be prioritized due to the potential harmful effects of X-ray imaging, particularly if there are no limitations on imaging in a supine position.

Conclusion

Our findings suggest that genu varum in football players results from proximal tibial deformation, as indicated by differences in MPTA and LDTA values compared to the control group. Additionally, the results of our study indicate that football players with genu varum exhibit a different alignment in their dominant leg compared to their non-dominant leg. During skeletal maturation, knee alignment is influenced by the growth of the distal femoral and proximal tibial epiphysis, highlighting the need to design training plans that account for factors contributing to genu varum development in football players whose skeletal-muscular system is not yet fully mature. Training plans should incorporate exercises to strengthen the muscles surrounding the proximal tibia to improve alignment and reduce varus stress, while also considering potential differences between the dominant and non-dominant leg. Future studies should examine the effects of training on the prevention and correction of genu varum. Additionally, considering the differences in women’s skeletal structures and the biomechanical loads in other sports, research should investigate whether similar mechanisms contribute to the development of genu varum in female football players and athletes in other sports.