The Effect of Functional Training on Dynamic Knee Valgus of Youth Soccer Players: A Pilot Study

Abstract

Knee injuries are very common in soccer players. High loads on the knee during landing or changes of direction can lead to a valgus shift of the knee, known as Dynamic Knee Valgus (DKV). Many studies have shown that a large shift in DKV is a predisposing factor for knee injuries and disease, such as anterior cruciate ligament injury (ACL), patellofemoral pain syndrome or osteoarthritis. Functional training could be a great tool to reduce DKV. Thus, in this pilot study, nine male youth soccer players (mean age: 16.4 ± 0.2 years) took part in six weeks of functional training program. DKV was measured in pre- and post-measurements during single leg squats using Kinect Azure camera with DynaKnee software. In addition, maximal voluntary muscle contraction (MVC) was measured using a dynamometer and muscle activation amplitude (MAA) was measured using electromyography. Data were analysed using a non-parametric Wilcoxon signed rank test to assess differences in DKV, MAA and MVC at a 5% significance level. Data showed a small improvement of DKV in the dominant leg. MVC increased slightly in all variables. The MAA of the involved muscles in the single-leg squat did not show a higher value, but rather a decreasing trend. However, none of the analyses showed significant changes. The small number of participants and the short duration of the training might have been a limiting factor. Further studies should repeat the procedure with a larger group of participants and a longer training period.

Introduction

The prevalence of sports injuries in elite sport is high, thus research on injury prevention in sport has become important in the recent past. Especially, the impact of injuries in youth can have serious consequences on the health, career, and physical activity levels of youth athletes.1 In youth soccer there is high incidence of injuries, and the knee especially is one of the most frequently injured areas.2 The present study focused on one of the most important injury risk factors of the lower limb which can cause anterior cruciate ligament (ACL) rupture3 and other knee joint damage in non-contact situation.4,5 During physical activity in case of shifting, change of direction, jumping and landing the knee collapses to the medial side of the body which is called dynamic knee valgus (DKV). DKV is described by Wilczynski et al6 as an “incorrect movement pattern”. DKV is a combination of adduction and internal rotation of the femur in the hip joint7–9, abduction and external rotation of the tibia in the knee joint,7–9 as well as a tibial anterior translation7 and ankle eversion.7 This results in a medial displacement of the knee in relation to the foot-thigh-line.10 Uneven weight distribution and sheer force can overload the cartilage and the ligaments within the knee,11 thus causing knee injuries.

It has been shown, that DKV is common during ball games, e.g.: basketball and soccer, which themselves are already predisposing factor for non-contact knee injuries.12 Females having a higher risk of DKV due to a wider pelvic.9 Previous studies have reported conflicting findings regarding DKV in the dominant (kicking) versus the non-dominant (standing) legs. One study found greater DKV in the dominant leg,13 while another showed greater DKV in the non-dominant leg.14 To identify the DKV, Uhlár et al15 suggested, that the range of normal medial knee displacement at 15% squat depth in a single leg squat using the DynaKnee software was 2%.

Several studies have suggested that DKV may be an important predisposing factor for acute and chronic injury and disease, such as ACL injuries,11 knee osteoarthritis5 and patellofemoral pain syndrome.4 To reduce the risk of injuries, there is a need to create an effective injury prevention program, which focuses on modifiable factors of DKV and general lower limb injury risk reduction. Programmes such as FIFA 11+ have been primarily designed for recreational and semi-professional warm-ups and especially focus less on the DKV.16 The program in this study therefore places a special emphasis on strengthening the muscle groups involved in DKV control. Wilczynski et al6 identified modifiable factors of DKV, highlighting reduced strength in hip abductors, extensors, external rotators and limited mobility in higher midfoot.6,17 These findings were further supported by other studies, which emphasized the importance of injury prevention training focused on strengthening and activating gluteal muscles, improving trunk lateral flexion strength, increasing ROM dorsiflexion ankle and midfoot mobility.6 Additionally, Hollman et al18 found significant correlations between higher DKV and an increased medial hip rotation and abduction, as well as decreased gluteus maximus recruitment.8

Effects of strength training on the DKV have been demonstrated and therefore the comprehensive functional training program includes strength training exercises focusing on the hip muscles,19–23 the quadriceps19 and hamstrings19,22 to reduce DKV. Another risk factor for lower limb injuries that gets addressed in the program is an imbalanced hamstring to quadriceps ratio.24,25 In addition to muscle strengthening, plyometrics, balance and core stability improvement have been reported to control DKV movement.6 Studies have also shown that core strength is associated with lower extremity injuries, particularly knee injuries.19 Kovács and Pucsok26 mentioned that soccer injuries may result from deficits in the core musculature and therefore an inefficient kinetic chain. Core strengthening plays a crucial role in preventing lower limb injuries and was therefore included in the developed programme.27,28 Evidence shows that balance has an important effect on general injury prevention for the lower extremities.28 Especially when it comes to neuromuscular control and stabilisation of the knee during dynamic activities, such as landings, direction shifts or decelerations.29 Excellent balance is positively associated with injury prevention in DKV and so part of the program.19 A Study by Saki et al30 found positive effects of plyometric training on dynamic balance and movement performance in DKV participants and thus plyometric exercises were also included in the program. Therefore, this study aims to fill the gap in the research literature investigating the effects of a comprehensive training programme on DKV in youth soccer players.

The aim of the present study was to 1) develop and manage a comprehensive functional training program twice a week for 45 minutes to reduce the risk of lower limb injury due to a decrease in DKV, 2) discover the effect of the training on the neuromuscular performance in form of muscular voluntary contraction (MVC) and muscular activation amplitude (MAA) in youth soccer players. The comprehensive intervention program included pre-activation before focusing on strengthening exercises for the gluteus medius, gluteus maximus, quadriceps and hamstrings, as well as balancing, plyometric and core exercises.

Methods

Participants

Nine male participants (age: 16,35 ± 0,19 yrs) were recruited from the same under-17 team competing in a national elite youth soccer league (MLSZ Országos U17 Alap csoprt). Measurements and training were executed in-season. The number of participants was naturally set to correspond with the size of the participating soccer team. Out of 23 invited players, 14 players were excluded due to low participation in less than six training sessions. Descriptive data of the athletes are shown in Table 1. Participants provided a written informed consent signed by a parent. This research was carried out fully in accordance to the ethical standards of the International Journal of Exercise Science31 and approved by the Ethics Committee (Ethical License Number: TE-KEB/33/2022).

Table 1.

Descriptive Data

Anthropometric parameters		Mean ± SD
Age (years)		16,35 ± 0,19
Height (cm)		177,56 ± 4,04
Weight (kg)	Pre	66,11 ± 7,24
	Post	66,11 ± 7,87
BMI	Pre	20,92 ± 1,55
	Post	20,90 ± 1,69
Leg Dominance		5 r / 4 l
Players Position	Defender	3
	Midfielder	3
	Forward	3

Note: SD = standard deviation; Pre = pre-measurement; Post = post-measurement; r = right side; l = left side

Protocol

A standardised measurement protocol was developed in order to maintain the same conditions for all participants and both measurements, including the order and execution of the test, as well as clothing, room temperature, questionnaires on additional data and condition of the subjects.

Measurements were performed in the same room in a rotating order starting with evaluating the DKV, then the MAA and finally the MVC. Participants were asked to wear sport shirts and shorts without covering knees. The measurement was performed barefoot, and without knee braces or kinesiotape. As a warm-up, the subject performed ten normal squats.

DKV was measured using Microsoft Azure Kinect camera system (Microsoft Corp. Redmond, WA, USA). The setup for this measurement included the Microsoft Kinect Azure camera on a tripod, the connected laptop to manage, record and analyse the data using DynaKnee software (Orthosera Kft, Budapest, Hungary), and a free area to perform the movements. Participants were asked to remove their shirt to ensure that the software could identify the 18 biomarker points on the body.

The starting position was standing upright with hands on hips to obtain baseline data for the length of the lower extremities. For the movement excursion the subjects bent their left knee to a 90-degree angle so that the left lower leg was lifted off and parallel to the ground. To measure the DKV shift, the participants performed ten well executed single leg squats as low as possible while keeping their heel and foot in contact with the ground. In this way the knee is placed under a heavy load and a DKV under tension is provoked. Squats were considered valid if participants maintained their balance throughout the repetition with their hands on their hips and did not step away during the examination. In case of an improperly executed trial, the subjects were asked to repeat it (see Figure 1).

Figure 1.

Illustration of the setup and the procedure during the measurements of DKV (left), MAA (middle-left & -right) and MVC (right).

MAA was measured using a surface electromyography device, Telemyo Mini 16 channel telemetry unit (Telemyo Mini, Neurodata, Germany). The setup consisted of a laptop for data recording, the EMG device with a portable connection station to which the bipolar surface electrodes were attached, a place for skin preparation and an area for performing the movements. The skin was prepared by shaving and disinfecting it to reduce the impendence between the electrodes.32 The bipolar non-invasive electrodes were then placed at the proximal and distal ends of the muscles. The measured muscles were rectus femoris, vastus medialis, vastus lateralis, gluteus medius, gluteus maximus and the biceps femoris. In addition, an electrode was placed on the patella to measure a baseline, which had to be placed on a bony surface with little soft tissue so that no muscle activity could interfere (see figure 1). The electrode localisation by palpation and procedure of the MAA measurement and was followed by SEMG guidelines.33

MVC was measured using a hand-held DynaMo strength and range of motion dynamometer (VALD Performance, Newstead, Australia). The setup of the measurement consisted of the device itself with a palm and a curved pad for the attachment on the body, a connected tablet for recording, and a physiotherapist’s bench. In addition, hamstring-quadriceps ratio was calculated by dividing the maximum hamstring contraction by the maximum quadriceps contraction in each patient. The standardized measurement protocol in all cases of muscle strength testing followed the VALD protocol guide recommendations including patient position, contact point, device orientation, position of examiner and instructions.34 The subject was instructed in the movement of pushing against the pad as hard as the participant could and the MVC was measured in three attempts after the verbal start signal until the verbal relax signal. First knee extension was measured, where the subject was placed in a seated position with the back of the knee against the edge of the bench. The leg was hanging down, the device was placed on the anterior lower leg, just proximal to the ankle and the arms were relaxed on the thighs to avoid additional use of other muscles. Next, knee flexion was measured, where subject lay in the prone position with one leg extended. The other leg to be measured was bent at a 90-degree angle. Arms were bent and placed behind the back and subject was instructed to not lift the hips of the bench. Dynamometer was placed on the posterior aspect of the subject’s lower leg, just proximal to the ankle. Lastly, hip abduction was measured subject lay in the supine position, hands were laid on the belly to avoid the involvement of other muscle groups. The device was place on the lateral aspect of the patient’s lower leg, just proximal to the lateral malleolus.

Six weeks of Training program held twice a week for 45 minutes. The programme was based on five parts, including a general warm-up, pre-activations for the following exercises and the main part with plyometrics, balance, strength, and core stability exercises (Table 2). It varied slightly from training day to day with the exercises becoming increasingly difficult. On one day of the week the warm-up was reduced due to a previous football training session. The post-measurement happened 24–48h after the last training session.

Table 2.

Functional Training Programme

Training parts	Exercises	Amount	Duration	Relation
Warm-up	Jumping Jacks	5	60s	10%
	High Knee Skipping	5	30s
	Tapping with Rotation	5	60s
	Mountain Climber	5	30s
Pre-Activation	Knee Hug	11	1rep	15%
	Hip bending to Toe Touch	11	1rep
	Hip Opening to Side Lunge	10	1rep
	Front Lunge to Upwards Rotation and Downwards Bend	11	1rep
	Inch Worm	10	1rep
Plyometrics	Ice-Skater	12	30s	25%
	Lateral Single Leg Jump	12	30s/30s
Balance	Balance Clock	12	30s/30s
	Rotational Jumps	5	30s
	Single Leg Stance with External Interference	11	30s/30s
Strength	Hamstring Curl	12	<7rep	30%
	Step Back Lunge	7	60s
	Jumping Lunge	11	30s
	Hip Abduction	12	30s/30s
	Clam Exercise	11	30s/30s
	Quadruped Leg Extension	12	30s/30s
Core Stability	Plank	12	60s	20%
	Side Plank	12	45s/45s
	Bridging	12	60s(30/30)

Note: s = seconds; rep = repetitions; % = percent

Statistical Analysis

Data from the DynaKnee software and the VALD Hub were analysed and collected. MAA data were sent to the department in a rectified format. A five hertz low pass filter was applied to smooth the data at a sampling rate of 1000/sec. Squat period was determined by viewing and the amplitude per trace in each muscle was analysed using the amplitude mean. LabChart Por 8.1.24 (AD Instruments, New Zealand) was used to process the MAA data. Final analyses were performed using IBM SPSS Statistics software (International Business Machines Corporation, Armonk, New York, USA). The Shapiro-Wilc test results did not meet the assumption of normality in all samples. Therefore, non-parametric analyses were used in the present study. To evaluate the effects of the independent variable training on the dependent variables DKV, MAA and MVC non-parametric Wilcoxon signed-rank test was used. To examine the magnitude of the relationship, effect sizes were calculated using the rank-biserial correlation based on the Wilcoxon signed-rank test. The following guidelines were used for interpretation: r < 0.1 indicating no effect, r between 0.1 and 0.3 indicating a small effect, r between 0.3 and 0.5 indicating a medium effect, and r > 0.5 indicating a large effect.35 Correlations between all performance data and protocols were analysed using the Spearman’s correlation coefficient. The significance level was set to p < 0.05. Statistical significance was highlighted by an asterisk (*).

Results

Results of DKV in the non-dominant side were made with data of 8 participants due to software problems in the data processing of one participant. DKV value decreased in the dominant leg (difference: −0,84 ± 2,31) and did not change or even slightly increased in the non-dominant leg (difference: 0,04 ± 1,50) (Figure 2). Six participants were able to reduce their DKV in at least one side. Three of them in both the non-dominant and dominant side. Furthermore, four subjects were able to reduce their DKV on one side from above the higher risk limit of 2% to below this reference value. The analysis concluded that the influence of the functional training however had no significant effect on the DKV with a medium effect size in the dominant leg (z = −1,01; p = 0,314; r = 0.3) and no effect size in the non-dominant leg (z = 0,14; p = 0,889; r = 0.0).

Figure 2.

The percentage of DKV measured before (pre) and after (post) six weeks of functional training in the nondominant (ND) and dominant (D) leg. Values are means ± SD.

MVC increased in gluteus, hamstrings, and quadriceps muscles of dominant side with a medium or large effect size. In the non-dominant side, all measured muscle groups increased with a medium or large effect size as well, except of the quadriceps muscles (Table 3). MAA decreased in almost all measured muscle groups (Table 4). However, analysis concluded that both the results of MVC and MAA did not show any significant changes.

Table 3.

Maximum Voluntary Contraction

Muscle	Side	N	Pre Mean ± SD	Post Mean ± SD	Difference Mean ± SD	Significance	Effect Size
Gluteus	ND	8	209,38 ± 43,08	224,50 ± 32,88	15,13 ± 57,20	0,208	0.4
	D	8	202,88 ± 37,37	225,00 ± 35,63	22,13 ± 51,87	0,161	0.5
Hamstrings	ND	8	140,38 ± 36,22	169,75 ± 15,21	29,38 ± 44,38	0,093	0.6
	D	8	147,63 ± 24,18	166,63 ± 23,86	19,00 ± 41,97	0,263	0.4
Quadriceps	ND	8	212,50 ± 51,82	226,50 ± 25,23	14,00 ± 39,78	0,327	0.3
	D	8	226 ± 29,21	223,00 ± 28,39	−3,00 ± 38,385	0,779	0.1

Note: SD = standard deviation; ND = non-dominant; D = dominant

Table 4.

Muscle Amplitude Activation

Muscle	Side	N	Pre Mean ± SD	Post Mean ± SD	Difference Mean ± SD	Significance	Effect Size
Rectus Femoris	ND	9	113,78 ± 47,84	104,15 ± 44,15	−9,63 ± 27,77	0,260	0.4
	D	9	116,86 ± 53,55	95,79 ± 46,08	−21,07 ± 33,20	0,086	0.6
VMO	ND	9	163,77 ± 47,76	190,35 ± 81,35	26,59 ± 84,42	0,314	0.3
	D	9	185,30 ± 96, 97	158, 18 ± 89,96	−27,12 ± 47,90	0,110	0.5
VLO	ND	9	160,14 ± 62,44	167,18 ± 60,3	7,04 ± 34,79	0,859	0.1
	D	9	218,8 ± 115,12	162,73 ± 33,87	−56,06 ± 101,04	0,173	0.5
Gluteus Medialis	ND	9	137,77 ± 38,92	118,26 ± 54,26	−19,51 ± 51,70	0,260	0.4
	D	9	158,28 ± 52,74	142,97 ± 54,23	−15,31 ± 40,20	0,374	0.3
Gluteus Maximus	ND	9	136,22 ± 62,44	122,15 ± 57,02	−14,07 ± 56,53	0,594	0.2
	D	9	148,99 ± 74,50	112,44 ± 51,94	−36,55 ± 53,55	0,110	0.5
Biceps Femoris	ND	9	67,22 ± 25,43	66,53 ± 23,91	−0.69 ± 29,35	0,953	0.0
	D	9	71,98 ± 31,18	59,55 ± 20,8	−12,43 ± 23,73	0,260	0.4

Note: ND = non-dominant, D = dominant; VMO = Vastus Medialis Oblique; VLO = Vastus Lateralis Oblique

Moreover, the differences between the non-dominant and dominant leg in terms of DKV, MVC and MAA were analysed and showed no significant differences. However, in the dominant leg there were large effect sizes in decrement of MAA in Rectus Femoris, VMO, VLO and Gluteus Maximus muscles. Results based on the development from pre- to post-measurement in DKV, MVC and MAA data did not show consistent significance. Maximum and minimum values of the hamstring-quadriceps ratio are as well as the mean and its difference between pre- and post-measurement shown in Table 5.

Table 5.

Hamstring-Quadriceps-Ratio

		Minimum		Maximum		Mean ± SD		Differences	Significance	Effect Size
		pre	post	Pre	post	Pre	Post
H-Q-ratio	ND	0,44	0,56	1,18	0,84	0,705 ± 0,27	0,75 ± 0,09	0,052 ± 0,29	0.484	0.2
	D	0,46	0,56	0,88	0,86	0,663 ± 0,14	0,747 ± 0,09	0,089 ± 0,20	0.263	0.4

Note: H-Q-ratio: Hamstring-Quadriceps-ratio; ND = non-dominant; D = dominant; SD = Standard Deviation

Analyses were performed on additional subject data including age, height, weight, BMI, leg dominance, on field position, training participation and rate of perceived exertion of the overall training intervention and subject condition on the measurement date were made. No significant correlations were found. However, the DKV showed a trend towards improvement with higher participation rates.

Discussion

The aim of this study was to investigate the effect of a developed functional training program on the DKV in elite youth football players. Results showed that there was a tendency for improvement in DKV, especially in dominant side with a medium effect size. In terms of individual results, the training programme was successful in four participants in the dominant leg, reducing the DKV value below the critical 2% threshold to reduce the risk of knee injury.15 However, due to the lack of significant differences between pre- and post-measurement, the power of this finding is low. The hypothesis that the developed functional training reduces the risk of knee injury due to reduced DKV cannot be supported, further investigation should be done. The results of the MAA and MVC were also inconsistent. MVC results showed a tendency to increase in all muscle groups with large effect size in the gluteus muscles in the dominant side and the hamstrings in the non-dominant side. Most of muscle groups showed a decrement in MAA. The lack of significant differences and the high standard deviations also led to low statistical power to draw conclusions. The hypothesis of an effect of the functional training on the neuromuscular performance in form of MVC and MAA cannot be supported.

Previous research has shown the importance of improving DKV4 and implementing injury prevention programmes in the daily life of athletes.6 Exercises and training methods have been investigated to find ways to improve injury risk prediction.16,22 Programmes such as FIFA 11+ are well researched but have been developed for use in warm-up sequences during training and on match days in the majority of grassroots and amateur sports16. In contrast the comprehensive functional training programme in this study aimed to reach elite level youth soccer players. The exercises and training methods selected for this comprehensive training programme demonstrated positive effects on muscle development and injury prevention.16,19,21,23 However, the results of this study did not support the hypotheses. Improvement of the developed program or the measurement methods in order to obtain more consistent and significant results are needed. The results are therefore not consistent with previous research, most probably due to study limitation, such as low number of participants.

Previous studies have highlighted the effects of biofeedback on DKV and recommended this method as an important part of injury prevention programmes.20,36 In another study, Zamankhanpour et al37 found that that dual tasking has an impact on the ability to control the movement of DKV and should therefore be considered in prevention programmes for DKV. Furthermore, this study did not focus much on the influence of the ankle eversion mechanism of DKV. Although the evidence for the influence of ankle-focused exercises is unclear. Malloy et al3 highlighted the correlation of ankle dorsiflexion flexibility with DKV and recommended the implementation of these exercises in clinical practice. Dischiavi et al38, who criticized the overestimation of DKV influences, also stated that the focus should be more on the holistic kinetic chain system. The influence of endurance, psychological readiness, and self efficacy, which Dischiavi et al38 identified as important influences, could not be debiased in the present study.

The study did not show any significant correlations of DKV with MAA and MVC. Appearance on only one side or one day of measurement is not strong enough to draw causal or generalisation conclusions for the peer group or athletes in general. The development of the hamstring-quadriceps ratio was also tested and no significant differences and correlations with DKV were found. However, the results showed an increase in both legs with a small effect size in the non-dominant and a medium effect size in the dominant leg. The ratio is more balanced correlating with a lower risk of injury.24 The minimum values in both non-dominant and dominant leg were increased by at least ten percent. This being close to the 60% limit associated with a lower risk of injury.25 The higher mean difference for the hamstrings than for quadriceps indicates that the functional training exercises which are specifically targeting hamstrings were effective.

Differences between non-dominant and dominant side were also not significant for DKV, MVC and MAA. The literature review showed a similar picture. Studies supported greater DKV in the dominant leg,13 whereas another showed greater DKV in the non-dominant leg.14 Therefore, this study supports the inconsistent scientific situation and indicates the need for further scientific investigation on this topic.

Results on the differences in training participation, ranging from 8 to 12 participations, did not show a significant correlation. However, there was a tendency for greater improvements in the 5 players with full training participation. Some studies have shown effects of training after six weeks,20 on the other hand, neuromuscular and physiological adaptations usually require longer duration to show more significant effects.39 A longer period was planned, however due to the club availability had to change. Missed training sessions are likely to have a greater impact on the results.

Several limitations must be addressed when discussing the results of this study: Firstly, the study design with the initial expected number of 23 participants was considered as sufficient. The cross over or control group design is emphasised as a more powerful option for scientific conclusions. However, recruitment did not allow such design to be implemented. Chosen design analysed the effects of the training by comparing the performance of their normal training with the performance following the period of the developed preventive training. Secondly, the intervention took place in-season and there was no bias control of training load in their practice or additional individual strength training outside the study. The small sample size made the data vulnerable. The medium to large effect sizes in some variable changes coupled with the differences being non-significant implies that the study might be underpowered due to the small sample size. Notably, outliers have a large effect on the results. Subjects with a previous knee injury, foot abnormalities, or minor pain or fatigue on testing days could not be excluded due to the already small number of participants. Participants may have been susceptible to personal daily fatigue, volition, and motivation to perform the tasks, as it is common in adolescents. Further questionnaires would help identify these to create more robust inclusion and exclusion criteria.

Muscle activation based on peak activation could be a viable alternative to the MAA used in this study. The binary electrode measurement only measures a representative number of motor units. High-density EMG measures muscle activation from multiple closely spaced electrodes40. Therefore, high-density EMG could improve the accuracy of the MAA measurement and provide more precise information about training adaptations.41

A comprehensive training programme can further include biofeedback, kinetic chain load distribution and movement quality improvement to increase the influence on DKV. A larger group and a longer training period could lead to larger and more significant effects. In addition, comparisons at different ages and levels could be adapted to this study. Further research into the effects of the dominant and non-dominant leg would address uncertainties.

In conclusion, results of the present study did not support the stated hypotheses. The current literature showed that an improvement in DKV was highly expected. However, there was no significant effect of the functional training on the DKV in youth soccer players. Some of the participants still showed results suggesting a reduced injury risk of the knee. The development of the MVC supported by medium to large effect sizes and the change in hamstring quadriceps ratio also showed a tendency towards the effectiveness of the developed injury prevention programme. Correlations between DKV and MAA, leg dominance, subject data or the condition at measurement have been shown in previous studies but cannot be confirmed by the data in this study. Limitations such as small number of participants and short duration of the training may have led to the inconsistent data and an underpowered study. Further studies should therefore repeat the procedure with a larger group of participants and a longer training period. In addition, biofeedback, ankle-focused exercises, dual-tasking, and movement quality improvement could be added to the developed training programme.