Effects of Flywheel Leg Curls on Muscle Structure and Function in Athletes with a History of Hamstring Strain Injury: A Randomized Controlled Trial

Abstract

Background:

Hamstring strain injuries (HSIs) have a high rate of recurrence, highlighting the need for effective tertiary prevention strategies. Flywheel resistance training appears effective in targeting muscular risk factors for HSI.

Hypothesis:

Flywheel leg curls will result in greater improvements in eccentric knee flexor strength and biceps femoris long head (BF_LH) fascicle length compared with conventional leg curls.

Study Design:

Randomized controlled trial.

Level of Evidence:

Level 2.

Methods:

Twenty-six athletes with a history of HSI within the past 18 months participated in an 8-week preventive training program. Athletes were assigned randomly to perform leg curls using either conventional (CON) or flywheel (FLY) equipment. Primary outcomes included eccentric knee flexor strength and BF_LH fascicle length. Secondary outcomes included isometric strength, maximal hip flexion active knee extension (MHFAKE), passive straight leg raise (PSLR), BF_LH muscle thickness, and pennation angle. Reinjury occurrence was monitored over a 6-month period.

Results:

The FLY group showed significantly greater improvements in eccentric strength (19 ± 19% vs 6 ± 9%; P = 0.04) and BF_LH fascicle length (9 ± 8% vs 2 ± 3%; P = 0.01) compared with the CON group. The FLY group also exhibited superior gains in isometric strength and flexibility (MHFAKE), whereas the CON group showed a larger increase in BF_LH pennation angle. During follow-up, the CON group sustained 3 reinjuries, whereas the FLY group reported only 1 (reinjury risk ratio 3.0, 95% CI, 0.31-28.9).

Conclusion:

Athletes with a history of HSI who participated in a preventive training program including flywheel leg curls showed greater improvements in knee flexor strength, fascicle length, and flexibility, as well as a lower reinjury occurrence compared with those trained with conventional leg curls.

Clinical Relevance:

Flywheel resistance training offers enhanced muscular adaptations that may reduce the risk of reinjury in athletes with a history of HSI.

Hamstring strain injury (HSI) is reported consistently as one of the most common noncontact injuries in sports that involve high-speed running, such as football (soccer), rugby, American football, Australian rules football, and sprint events in track and field. ²⁶ For example, European elite men’s football clubs can expect 1 out of 5 athletes in a team squad to miss a training session or match due to an HSI in each season. ⁹ Most HSIs are considered mild to moderate in severity, with the athlete losing up to 4 weeks of sports practice. ¹⁰ HSIs account for approximately 14% of total injury-related lay-off days, ⁹ negatively impacting football clubs’ performance and finances. ¹¹ Therefore, optimizing prevention strategies against HSI is a primary goal for healthcare practitioners.

The primary risk factor for an athlete to sustain an HSI is having previously experienced this type of injury, especially if the index HSI occurred recently. ¹³ There is a notable recurrence rate of HSIs, varying from 18% to 38%. ⁹ Interestingly, more than two-thirds of HSI recurrences occur within 2 months after the athlete’s return to full participation from the index injury.^8,26 This high early reinjury rate suggests a premature return to sport, an inadequate rehabilitation program, or a combination of both. ¹⁴ Despite commendable advances in the rehabilitation of athletes with HSI, ²¹ evidence suggests that most athletes have strength deficits at return to sport after HSI. ¹³ Even more impressive are findings that hamstring eccentric strength deficits seem to persist up to 18 months after a HSI.^26,29,35 Moreover, athletes with a history of HSI appear to have different muscle architecture compared with those without such a history, particularly characterized by shorter fascicles in the biceps femoris long head (BF_LH).^6,35

Considering that both poor eccentric knee flexor strength and reduced BF_LH fascicle length are factors associated with a greater risk of sustaining an HSI,^4,20,32 it seems advisable to engage athletes with a history of HSI in tertiary prevention programs targeting these factors. Among the available options, flywheel resistance training has garnered significant attention from the scientific community and has become increasingly popular among healthcare practitioners. According to a current guideline on this topic ³ : “Flywheel resistance training is characterized by the use of rotating flywheel discs or cones to provide resistance. The concentric action is initiated by pulling the strap connected to the shaft of the device, spinning (accelerating) the flywheel/cone. Once the strap rewinds around the shaft, an eccentric action is performed to decelerate the flywheel/cone rotation.” If executed correctly, this enables a greater application of force during the eccentric action, resulting in a mechanical eccentric overload. Consequently, training programs incorporating flywheel resistance exercises have been shown to increase eccentric knee flexor strength, enlarge BF_LH fascicle length, and reduce the risk of index HSI.^2,28,33

In summary, there is a scientific rationale and preliminary evidence for using flywheel resistance training to prevent HSI. Research into the architecture and function adaptations of the hamstring muscles has been ongoing in healthy people. However, there remains a gap in the literature concerning the responses to flywheel resistance training observed in athletes with a recent history of HSI. Therefore, the aim of the current study was to determine whether incorporating flywheel leg curls into a preventive training program leads to greater improvements in muscle structure and function, and a lower reinjury occurrence, compared with conventional leg curls in athletes with a history of HSI.

Methods

Study Design

This is a randomized controlled trial with 2 parallel intervention groups. Athletes with a history of HSI in the past 18 months were assigned to an 8-week preventive training program, performing leg curl exercises on either conventional (CON group) or flywheel (FLY group) equipment. Data were collected between March 2020 and November 2022. Muscle structure and function outcomes were assessed at baseline and at 1 week after intervention. Reinjury occurrence was assessed during a 6-month follow-up period. The study was approved by the Fundación Oulton Ethics Committee (CIEIS 1413/17, study no. 3828) Ethics Committee and registered at clinicaltrials.gov (NCT04050813). This report adheres to CONSORT and TIDieR guidelines.^18,30

Participants

This study recruited professional or semiprofessional football and rugby athletes with a history of HSI. Eligible participants had to be registered with their sport’s federation, actively involved in their team’s full training routine, and have sustained a unilateral HSI within the past 18 months. They must have undergone physiotherapy (either at their club or privately) and played ≥1 match after returning to the sport. All participants had a structural hamstring injury (Munich Consensus type 3B), clinically confirmed by a medical doctor through imaging examinations. Exclusion criteria included: (1) absence of an imaging examination confirming HSI with structural damage; (2) history of hamstring or gluteal tendinopathies, or other injuries with posterior thigh symptoms, within the last 18 months; (3) previous knee or hip surgery; and (4) previous administration of platelet-rich plasma or other injectable therapies (eg, corticosteroids) in the posterior thigh within the last 3 months. Injury history and injectable therapies were verified through medical records and participant self-report. All participants provided informed consent.

Sample Size, Randomization, and Blinding

The sample size was calculated using G*Power software (Version 3.1.9.2; Universität Kiel). Based on effect sizes of d = 1.21 for eccentric knee flexor strength and d = 1.4 for BF_LH fascicle length—both primary outcomes of this study—a statistical power of 80% and an alpha level of 0.05 yielded required sample sizes of 12 and 10 participants per group, respectively.^2,27 Therefore, a minimum sample of 24 athletes was considered for this 2-arm RCT.

An investigator not involved in recruitment, assessment, or intervention assigned participants to CON or FLY groups randomly using a free online randomization tool. Stratification by age (under or >30 years) and sport (football or rugby) was applied using 4-participant blocks. The investigator delivering the training program was unblinded to group allocation. Participants were blinded and instructed not to disclose their group during assessments. Assessors remained blinded until data collection was completed.

Outcome Measures

Given the reported association between eccentric knee flexor strength and BF_LH fascicle length with HSI in prospective cohort studies,^4,20,32 both were selected as primary outcomes in the present study. Some evidence suggests that hamstring flexibility and isometric strength may be linked to HSI risk,^7,15,39 which is why maximal hip flexion active knee extension (MHFAKE) and passive straight leg raise (PSLR) tests, along with isometric strength assessments at 45°/45° (ISO45) and 90°/90° (ISO90) of hip/knee flexion, were included as secondary outcomes. In addition, results from BFLH muscle thickness and pennation angle were included as muscle architecture parameters not related directly to HSI. The data analyzed were collected from the previously injured limb for all muscle structure and function outcomes. Finally, the occurrence of reinjuries during the 6-month follow-up period was treated as a secondary outcome, as it was not feasible to include as a primary outcome due to the large number of reinjuries needed for a statistically robust analysis.

Muscle Architecture Assessment

A single specialist performed all assessments using a B-mode ultrasound system (MyLab Class C; Esaote) with a 6-cm linear array transducer (LA523; 4-13 MHz; Esaote). After 10 minutes of rest, participants were assessed in a prone position with hips neutral, knees fully extended, and feet hanging off the bed. Participants kept muscles relaxed during image acquisition.

The extended field of view (EFOV) ultrasound technique provided panoramic images of the entire muscle to measure BF_LH fascicle length without extrapolation. ¹¹ The ultrasound transducer was positioned initially at the midpoint of the femur length, between the medial and lateral borders of the BF_LH. The transducer was then aligned with the fascicle plane, and the path of the BF_LH was determined by following its fascicles in the superficial compartment proximodistally. The transducer was manipulated to ensure the fascicles remained continuous and visible while the aponeuroses remained parallel. The optimal ultrasound imaging path was then marked on the skin. For acquiring EFOV images, the transducer was moved slowly and continuously along the marked path with constant pressure from the distal to the proximal musculotendinous junctions, while continuously adjusting the transducer orientation to stay within the fascicle plane. ¹² Transmission gel was used to enhance acoustic contact and minimize transducer pressure on the skin.

The scans were digitized and later analyzed using a free image-processing program (ImageJ Version 1.48; National Institutes of Health). Fascicle length was defined as the length from the intermediate to the superficial aponeuroses and was measured directly using the segmented line tool, which accounted for fascicle and superficial aponeurosis curvature. Pennation angle was measured as the angle between the drawn fascicle and the intermediate aponeurosis. Muscle thickness, defined as the distance between the superficial and intermediate aponeuroses, was measured 5 times proximodistally, with the average value used for analysis. The test-retest reliability of this procedure is supported in the literature: fascicle length (intraclass correlation coefficient [ICC], 0.96), pennation angle (ICC, 0.84), and muscle thickness (ICC, 0.94). ¹²

Muscle Strength Assessment

Eccentric knee flexor strength was measured using a hand-held dynamometer (Mark-10; 2500 N, 1000 Hz; EK Basic Ergonomics Kits) following the methodology described by Whiteley et al, ³⁷ which demonstrated consistent reliability (ICC, 0.90). Participants were positioned in prone with hips neutral and knees flexed to 45° (0° = full extension). The HHD was placed against the distal shank near the malleoli. Participants contracted their knee flexors isometrically, aiming for maximum intensity within 2 seconds. During peak contraction, the evaluator applied a gradual force to “break” the participant’s isometric hold, inducing an eccentric contraction. This method, also known as “break test,” was repeated 3 times with 30-second rest intervals between trials. The highest strength value was recorded for analysis.

Isometric strength was tested in 2 joint positions for a more comprehensive assessment. Using a setup similar to Hickey et al, ¹⁶ who found consistent reliability values (ICC, 0.89-0.92), isometric strength was assessed at ISO45 and ISO90. Participants were supine, with the pelvis secured to the plinth and the heel attached to a load cell (Valkyria Trainer Push Pull; 2000 N, 1000 Hz; Ivolution). Tests were performed with the leg parallel to the ground. Participants performed 2 to 3 submaximal familiarization attempts, followed by 3 maximal attempts lasting 3 to 5 seconds each, with 30-second rest intervals. The highest strength values for each position were recorded.

Flexibility Assessment

The PSLR and MHFAKE tests were included to assess hamstring flexibility during both passive and active tasks. ³⁸ A smartphone inclinometer (Clinometer; Plain Code) was placed on the mid shin to measure the range of motion in both tests. For the PSLR test, participants were in a supine position with the contralateral limb fixed. The tested limb was raised passively, and the angle was recorded at maximum hip flexion or when pain was reported. For the MHFAKE test, participants, also in supine, kept the hip of the tested limb in maximal flexion by clutching the thigh to the chest and actively extended the knee to the point of maximal stretch or pain. Whiteley et al ³⁷ reported consistent reliability values for both tests (ICC, 0.83-0.88).

Reinjury Follow-Up

At 6 months after completing the preventive training program, the researchers reached out to participants to monitor the occurrence of new injuries. Time-loss injuries affecting the posterior thigh of the same limb involved in the index injury—clinically confirmed by a medical doctor as a structural hamstring injury (Munich Consensus type 3B) - were classified as reinjuries.

Intervention

Athletes in the CON and FLY groups completed an 8-week, twice-weekly training program at the Oulton Institute. Each session began with a warm-up consisting of 10 minutes of treadmill running at 8 km/h followed by dynamic flexibility exercises. This was followed by 3 lumbopelvic stability exercises (lateral bridge, dead bug, and bird dog),^1,18,19 4 resistance exercises targeting the lower limb muscles (leg extension, hip thrust, clamshell, and heel raise), ²³ and conventional or flywheel leg curls (Figure 1). A 30-second self-myofascial release using a foam roller was performed as part of the cool-down.

Figure 1.

Exercises included in the preventive training program. (a) Lateral bridge. (b) Dead bug. (c) Bird dog. (d) Leg extension. (e) Hip thrust. (f) Clamshell. (g) Hell rise. (h) Conventional leg curl. (i) Flywheel leg curl. The bottom panel illustrates the training periodization over the 8 weeks. rep, repetition.

The only difference between groups was that athletes in the CON group used a traditional leg curl machine (Cybex International), while athletes in the FLY group used flywheel equipment (Ivolution) (Figure 1). The resistance exercises were periodized over 4 blocks during the 8-week program (Figure 1). Participants completed a familiarization session to learn the exercises and establish their initial load (ie, the weight they could lift for 12 repetitions). Throughout the training program, load progression was tailored to each athlete’s individual response. Specifically for the flywheel leg curl, the load progression was organized as follows: weeks 1 to 2, 0.05 kg m⁻²; weeks 3 to 4, 0.08 kg m⁻²; weeks 5 to 8, 0.13 kg m⁻². During weeks 7 to 8, concentric repetitions were performed with both legs, followed by eccentric repetitions with 1 leg.

Statistical Analysis

Analysis of muscle structure and function data was conducted following intention-to-treat principles. Missing values were substituted with the arithmetic mean of the group value. The data distribution was assessed through visual inspection of Q-Q plots and histograms. The between-group differences and their respective CIs at baseline and post-training were calculated using mixed linear models. The outcomes percent change from baseline to post-training were used for between-group comparison through mixed linear models. Within-group analysis was performed with effect size (ES) calculation through Cohen’s d, classified as “trivial” (ES < 0.2), “small” (ES > 0.2), “moderate” (ES > 0.5), or “large” (ES > 0.8). Analyses were performed using the SPSS 18.0 software (Statistical Package for the Social Sciences Inc), with a significance of α < 0.05.

Results

A total of 30 athletes volunteered to participate in the study (Figure 2); 26 of these volunteers met the inclusion criteria and were assigned randomly to CON and FLY groups (Table 1). All participants completed the assessment and training schedule successfully and were evaluated at the 6-month follow-up. Adherence to training sessions was ≥80%. Missing data accounted for 1% of all outcomes.

Figure 2.

Flowchart of the study. CON, conventional equipment; FLY, flywheel equipment.

Table 1.

Characteristics of athletes

	CON group (n = 13)	FLY group (n = 13)
Age, y	27.77 (5.59)	24.38 (4.57)
Weight, kg	73.46 (12.35)	81.00 (8.50)
Height, cm	175.00 (5.55)	178.77 (5.54)
Sport
Soccer	10; 77%	9; 69%
Rugby	3; 23%	4; 31%
Experience, y	17.85 (3.76)	16.77 (4.55)
Training volume, h/week	10.62 (3.40)	12.92 (2.50)
Injured limb
Dominant	6; 46%	8; 62%
Nondominant	7; 54%	5; 38%
Time since injury, months	6.15 (2.54)	7.62 (5.33)

Results presented as mean (SD) or absolute; percent distribution. CON, conventional equipment; FLY, flywheel equipment.

Table 2 displays the baseline and post-training values for all outcomes. At baseline, the FLY group exhibited significant higher values in ISO90 strength (P = 0.02) and BF_LH muscle thickness (P < 0.001) compared with the CON group. After training, the FLY group exhibited significantly higher values in ISO45 strength (P = 0.01), ISO90 strength (P = 0.003), eccentric strength (P = 0.03), and BF_LH muscle thickness (P < 0.002), whereas the CON group had significantly higher values of BF_LH pennation angle (P = 0.05).

Table 2.

Baseline and post-training values for primary and secondary outcomes

	Unadjusted mean (SD)		Adjusted mean between-group differences [95% CI]
	CON group	FLY group
ECC strength, N
Baseline	229.31 (42.18)	248.54 (77.14)	19.23 [−31.10 to 69.56]
Post-training	241.62 (39.92)	285.00 (56.28)	43.38 [3.89 to 82.89]
Effect size	0.31	0.56
BFLH fascicle length, mm
Baseline	109.17 (17.61)	108.19 (13.97)	–0.98 [−13.84 to 11.89]
Post-training	111.67 (16.85)	117.05 (11.24)	5.37 [−6.22 to 16.97]
Effect size	0.15	0.57
ISO45 strength, N
Baseline	264.62 (62.60)	280.00 (56.97)	15.38 [−33.07 to 63.83]
Post-training	285.92 (65.95)	350.38 (62.75)	64.46 [12.35 to 116.57]
Effect size	0.34	1.22
ISO90 strength, N
Baseline	270.31 (42.93)	320.00 (61.53)	49.69 [6.75 to 92.64]
Post-training	299.08 (45.27)	379.54 (77.14)	80.46 [29.26 to 131.66]
Effect size	0.68	0.89
PSLR test, deg
Baseline	76.77 (9.60)	77.92 (5.51)	1.15 [−5.18 to 7.49]
Post-training	76.69 (10.55)	80.54 (6.68)	4.00 [−3.15 to 11.15]
Effect size	-0.01	0.45
MHFAKE test, deg
Baseline	71.15 (12.93)	74.00 (9.23)	2.85 [−6.25 to 11.94]
Post-training	71.62 (13.45)	79.46 (7.80)	7.85 [−1.05 to 16.74]
Effect size	0.01	0.67
BFLH muscle thickness, mm
Baseline	27.40 (2.52)	33.72 (4.90)	6.32 [3.16 to 9.48]
Post-training	28.69 (2.99)	35.08 (5.79)	6.38 [2.65 to 10.11]
Effect size	0.49	0.26
BFLH pennation angle, deg
Baseline	12.85 (2.69)	12.11 (2.28)	−.75 [−2.77 to 1.27]
Post-training	13.38 (2.29)	10.70 (1.76)	−1.68 [−3.34 to −0.30]
Effect size	0.22	0.72

BF_LH, biceps femoris long head; ECC, eccentric test; ISO45, isometric test at 45°/45° of hip/knee flexion; ISO90, isometric test at 90°/90° of hip/knee flexion; MHFAKE, maximum hip flexion with active knee extension; PSLR, passive straight leg raise.

The conventional training program produced a moderate effect size for ISO90 strength and small effect sizes for ISO45 strength, eccentric strength, BF_LH muscle thickness, and BF_LH pennation angle, whereas the effects for PSLR test, MHFAKE test, and BF_LH fascicle length were negligible (Table 2). Conversely, the FLY training program yielded large effect sizes for ISO45 and ISO90 strength, along with moderate effects for MHFAKE test, eccentric strength, BF_LH fascicle length, and BF_LH pennation angle, and small effect sizes for PSLR test and BF_LH muscle thickness (Table 2).

Between-group comparisons revealed that the FLY group exhibited significantly greater baseline-to-post-training changes than the CON group for the primary outcomes (eccentric strength and BF_LH fascicle length), as well as for ISO45 strength and MHFAKE test (Table 3). The CON group experienced greater percentage changes for BF_LH pennation angle (Table 3). No significant between-group differences were observed for ISO90 strength, PSLR test, and BF_LH muscle thickness (Table 3).

Table 3.

Percent change from baseline to post-training for primary and secondary outcomes

	CON group		FLY group		P value
	Mean (SD)	95% CI	Mean (SD)	95% CI
ECC strength	6.17 (8.78)	−2.58 to 14.92	19.12 (19.75)	10.37 to 27.87	0.04*
BFLH fascicle length	2.48 (3.22)	−0.99 to 5.94	8.85 (7.92)	5.39 to 12.31	0.01*
ISO45 strength	8.99 (8.58)	−1.00 to 18.98	27.70 (23.15)	17.70 to 37.69	0.01*
ISO90 strength	11.29 (10.82)	0.54 to 22.04	20.93 (24.24)	10.19 to 31.68	0.20
PSLR test	−0.15 (4.98)	−3.99 to 3.70	3.73 (8.08)	−0.11 to 7.58	0.15
MHFAKE test	0.68 (6.60)	−3.70 to 5.06	7.98 (8.57)	3.60 to 12.36	0.02*
BFLH muscle thickness	4.70 (5.04)	1.18 to 8.22	3.93 (7.08)	0.41 to 7.45	0.75
BFLH pennation angle	−2.00 (12.54)	−10.88 to 6.87	−9.45 (17.98)	−18.32 to −0.58	0.05*

BF_LH, biceps femoris long head; ECC, eccentric test; ISO45, isometric test at 45°/45° of hip/knee flexion; ISO90, isometric test at 90°/90° of hip/knee flexion; MHFAKE, maximum hip flexion with active knee extension; PSLR, passive straight leg raise.

^*Significant difference.

Over the 6-month follow-up, 3 athletes in the CON group (23.1%) and 1 athlete in the FLY group (7.7%) experienced reinjuries. Notably, all reinjuries occurred >3 months after completing the preventive training program. The reinjury risk in the CON group was 3.0 (95% CI, 0.31 to 28.9) relative to the FLY group.

Discussion

An innovative aspect of this study was the incorporation of flywheel leg curls into a preventive training program aimed at reducing risk factors for HSI in athletes with a previous history of this injury, offering a tertiary prevention approach. Our findings revealed that athletes who performed flywheel leg curls experienced greater improvements in eccentric knee flexor strength and BF_LH fascicle length, as well as in secondary outcomes potentially associated with the risk of HSI, such as isometric strength and hamstring flexibility. In addition, flywheel-trained athletes presented a lower reinjury occurrence compared with those who trained with conventional leg curls.

Although there is no consensus on the role of modifiable factors in the risk of HSI, ¹³ the selection of our primary outcomes was guided by prospective studies indicating that poor eccentric strength and reduced fascicle length are associated with an increased risk of HSI.^4,20,32 In addition, previous investigations have evidenced deficits in both eccentric strength and fascicle length in athletes who sustained HSI in the previous 18 months,^6,29,31 further supporting the relevance of our primary outcomes. Given the naturally increased risk of HSI for players with a recent injury history, ¹³ optimizing modifiable risk factors should be a primary focus. Therefore, it is reasonable to suggest that an intervention capable of simultaneously increasing eccentric knee flexor strength and lengthening BF_LH muscle fascicles could be beneficial in mitigating incidence and/or severity of hamstring reinjuries.

The eccentric strength gains experienced by athletes in the FLY group were 3 times higher than those observed in the CON group. This result was expected, given the specificity of the stimulus provided by each type of equipment. Conventional leg curls apply the same resistance to both the concentric and eccentric phases of the movement. Since the muscle naturally has a lower force production capacity during the concentric phase compared with the eccentric phase, the load used in conventional leg curls is limited by the individual athlete’s concentric strength, leading to a submaximal stimulus during the eccentric phase of the movement. Conversely, flywheel leg curls enable greater resistance during the eccentric phase than the concentric phase. ³ This mechanical eccentric overload appears to be the key factor behind the enhanced eccentric strength typically observed after flywheel training programs.^2,33

Significant increases in eccentric knee flexor strength have been reported in athletes undergoing flywheel resistance training with the leg curl and Nordic exercises.^2,33 Askling et al ² observed an average increase of 19% in eccentric isokinetic strength among professional football players after 10 weeks (16 sessions) of preseason training with the flywheel leg curl, which matches the number of sessions and type of exercise used in our study. Caution is needed when comparing trials, as eccentric strength assessments were conducted using different methods: isokinetic dynamometry versus hand-held dynamometer testing. However, athletes with a history of HSI allocated to the FLY group in our study experienced the same 19% average increase in eccentric knee flexor strength. This similar magnitude of muscle strengthening through flywheel resistance training is encouraging, particularly given the previously reported impaired responsiveness to strengthening programs in athletes with a history of HSI. ²⁷

Our findings further support the role of eccentric overload as a crucial factor for enhancing fascicle length, as previously evidenced by studies conducted in animal models and in humans.^5,34 After the 8-week training period, only athletes allocated to the FLY group experienced an increase in their BF_LH fascicle lengths. The 0.9-cm average increase found in the FLY group falls within the range of 0.5 cm to 1.4 cm reported after 5- to 6-week flywheel resistance training in athletes and recreationally active people, respectively.^28,33 According to a univariate logistic regression from a cohort of 152 professional football athletes, each 0.5-cm increase in BF_LH fascicle length corresponded to a 74% reduction in the risk of HSI. ³² In theory, a muscle with longer fascicles contains a higher number of in-series aligned sarcomeres, which would prevent the muscle from damage due to overlengthening. ²⁴ Conversely, a muscle with reduced fascicle length presents an increased susceptibility to eccentrically induced microscopic muscle damage, which could facilitate the macroscopic injury. ²⁵ However, it is important to note that the relationship between increased fascicle length and the risk of HSI has yet to be demonstrated in randomized controlled trials.

Among the secondary outcomes, the FLY group showed greater adaptations than the CON group in the ISO45 and in the MHFAKE test, but not in the ISO90 or PSLR test. Generally, isometric strength and flexibility are factors often considered to have minor relevance to the risk of HSI. Interestingly, deficits in maximal isometric strength at 15° (but not at 90° of knee flexion) and in active flexibility (but not in passive flexibility) upon return to sport after a HSI have been associated with a higher risk of reinjury within 12 months. ⁷ Furthermore, a secondary prevention strategy through weekly monitoring of isometric knee flexor strength in youth footballers reduced HSI rate and time lost to injury, ³⁹ whereas a prospective study found an association between flexibility levels and the risk of HSI in professional football players. ¹⁵ Therefore, the potential impacts of adaptations promoted by flywheel resistance training on these secondary outcomes should not be overlooked. On the other hand, the larger changes observed in the BF_LH pennation angle in athletes allocated to the CON group appear to have no plausible effect on the risk of HSI.

The 3-fold increased risk of reinjury in the CON group during the follow-up period is a notable finding; however, it should be interpreted with caution. The wide risk ratio CI, likely due to the small number of events observed in this study, denotes high uncertainty related to this outcome. It is also important to note that, as seen in most studies aimed at assessing the impact of interventions on athlete susceptibility to hamstring reinjury,^2,17,22 the exposure of athletes was not monitored during the follow-up period. Given that the number of hours and type of exposure (training vs matches) are linked closely to the risk of new muscle injuries, any differences in exposure between athletes assigned to the CON and FLY groups could have acted as a confounding factor influencing reinjury outcome. Thus, the reduced incidence of reinjuries among athletes who underwent flywheel leg curl training should not be regarded as conclusive evidence of the exercise’s superiority in preventing HSI. Rather, it suggests that adaptations on muscle structure and function observed in this study may potentially lead to a reduced likelihood of experiencing reinjury.

From a practical standpoint, the findings of the present study support the inclusion of the flywheel leg curl in preventive training programs designed to address risk factors for HSI in athletes with a history of this injury. The Nordic hamstring exercise is considered a primary exercise-based strategy for preventing HSIs, reducing the incidence of these injuries by approximately half across various athletic populations. ³⁶ This effectiveness has been attributed partly to the positive muscular adaptations induced by the Nordic hamstring exercise, which improve eccentric knee flexor strength and BF_LH fascicle length. ²² Therefore, it is interesting to note that the FLY group in the present study achieved higher eccentric strength gains than those reported in some trials comprising 8 to 12 weeks of training with the Nordic hamstring exercise in athletes,^22,24 concurrent with similar enhancement of fascicle length. ²²

Some limitations of this study should be acknowledged. First, we were unable to monitor participant exposure to training and matches during the 6-month follow-up, and we did not have access to further details regarding the severity of reinjuries (eg, time-loss). These limitations hinder our ability, as well as that of previous trials,^2,17,22 to calculate reinjury risk using relative incidence rates (ie, injuries per 1000 hours of exposure) and to analyze the impact of the intervention on reinjury severity. Second, there was no standardization of field training performed alongside the 8-week preventive program, which could have led to potential variations in outcomes due to differences in athletes’ field training. Third, although the hand-held dynamometer used for eccentric strength assessment is a reliable clinical tool, it is more dependent on operator technique than gold-standard methods such as isokinetic dynamometry or Nordic hamstring exercise testing. Finally, the lack of a control group that did not receive any intervention is a limitation. Although the comparison between the flywheel and conventional leg curl groups provides valuable insights into the relative effectiveness of these interventions, the absence of a true control group prevents us from determining the absolute effect of the training programs.

At the same time, some notable strengths should be emphasized. First, the concurrent implementation of the preventive intervention alongside the athletes’ training routines enhances its ecological validity in the context of high-performance sport. The high level of participant adherence further underscores the feasibility of the protocol. Second, direct supervision by a physiotherapist throughout the sessions ensured strict adherence to the prescribed intervention, enabling effective control over key variables such as exercise technique and appropriate load progression. Third, the use of the EFOV imaging technique provided highly accurate measurements of fascicle length, thereby enhancing the reliability of muscle architecture assessments.

Conclusion

Male football and rugby athletes with a history of HSI who participated in a preventive training program including flywheel leg curls showed greater improvements in eccentric knee flexor strength and BF_LH fascicle length compared with those trained with conventional leg curls. Flywheel resistance training also led to greater improvements in secondary outcomes potentially associated with the risk of HSI, such as isometric strength and hamstring flexibility. Finally, athletes who underwent flywheel leg curl training experienced fewer reinjuries over the following 6 months compared with those who underwent conventional training.