Do Flywheel Exercises Provide Eccentric-Overload Training?

Abstract

Interests in eccentric resistance exercises have been increasing in both research and practice. However, implementing eccentric resistance exercise training is often challenging due to the mechanical limitations of traditional training equipment. To address this, flywheel (FW) devices emerged as a practical alternative. FW devices are commonly considered to provide eccentric-overload training, in which the load is greater in eccentric than concentric phase. However, this is not always the case. In this article, we summarize the mechanical and physiological factors influencing the effectiveness of FW devices in achieving eccentric overload. Then, we discuss a significant limitation of FW resistance exercise in accurately quantifying the load, since eccentric mechanical load is constrained by preceding concentric phase. Lastly, we explore potential practical solutions and improvements in research methods for FW resistance exercises. FW resistance exercises become eccentric-overload exercises only when higher mechanical quantities are achieved and confirmed during the eccentric than in the concentric phase of repetitions. It is important to examine if eccentric overload is actually achieved during training and testing, which can clarify if the eccentric overload is a key factor for the neuromuscular adaptations observed following a FW resistance training.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40798-025-00974-y.

Key Points

• Flywheel resistance exercises do not necessary provide eccentric overload.

• The tempo or technique of exercise execution plays a critical role in achieving eccentric overload in flywheel resistance exercises.

• Eccentric overload is limited to a short time interval and is influenced by a combination of physiological factors such as the participant’s age, sex, training history, movement velocity (inertia magnitude), fatigue tolerance, tempo of execution, joint angle, muscle length, and movement complexity.

• Future studies should monitor mechanical outputs to confirm whether eccentric overload is the primary factor driving the adaptations observed in FW resistance training.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40798-025-00974-y.

Introduction

Eccentric exercises focus on lengthening muscle (eccentric) actions (contractions) that occur when the force exerted on the muscle exceeds the instantaneous force generated by the muscle [1]. In eccentric contractions, the muscle-tendon system is loaded with low energy expenditure [2], and by unique muscle activation patterns such as less motor unit recruitment and greater brain activity in comparison to isometric or shortening muscle (concentric) actions (contractions) [3, 4]. Emphasizing eccentric rather than concentric contractions in resistance exercises has been shown to be more effective for increasing muscle strength, power, and muscle mass, due to greater neural, morphological, and architectural adaptations [5]. The primary stimulus for the adaptations is considered to be mechanical stress to the muscle-tendon complex, which is capable of generating more than 40% greater force in eccentric than concentric contractions [6]. A recent meta-analysis [7] showed that eccentric training is more effective in improving maximal voluntary eccentric contraction strength, while both eccentric and concentric muscle strength training equally increase maximal voluntary concentric and isometric contraction strength. Therefore, the effects of eccentric resistance exercise training are more comprehensive, improving muscle strength in different contraction modes.

However, achieving a high eccentric load is not always possible due to the mechanical limitations of conventional training devices [8]. Consequently, practitioners employ strategies to induce eccentric overload, such as two-limb concentric and one-limb eccentric combinations (2−1 method), additional partner resistance for eccentric phase, partner assistance for concentric phase, the use of weight releasers, or increasing the moment arm relative to the joint’s axis of rotation during the eccentric phase [8]. Moreover, dedicated devices, mostly motorized machines [9], that enable the delivery of eccentric-only or accentuated eccentric loading are available on the market, but they are generally not used in practice [8]. Passive (non-motorized) flywheel (FW) devices are one of the viable alternatives that can potentially provide greater eccentric than concentric peak loading. They are considered to provide iso-inertial movements, are mostly portable and versatile, and have therefore gained both practical and research interest, demonstrating a variety of positive effects on muscle strength and power, including improvements in jumping, sprinting, and change-of-direction abilities in athletes [10], as well as on muscle hypertrophy [11]. They have been also shown to be effective for treatment of tendinopathies [12] and improving quality of life of older adults and people with neurological disorders [13].

Eccentric overload is defined as a production of greater mechanical quantities (force, velocity or power) in the eccentric than in the concentric phase in resistance exercise [14, 15]. FW devices provide a combination of concentric and eccentric contractions, allowing smooth transitions between consecutive eccentric–concentric repetitions within a set, rather than requiring the isolation of eccentric contractions. The release of stored elastic energy during an eccentric contraction results in improvement of subsequent concentric contraction force and power generating capability [15, 16], closely mimicking sport-specific movements such as change of direction [17], jumping [16, 18] and sprinting [19].

Although FW exercises are generally characterized as eccentric-overload modality [11, 20, 21], a recent meta-analysis revealed that only 17 of 79 studies provided sufficient data to prove the existence of eccentric overload. Most studies did not demonstrate that eccentric overload was achieved. The primary metrics indicating eccentric overload were peak power and peak velocity, rather than peak force [22]. Therefore, it is argued that FW resistance exercise cannot be always classified as eccentric-overload modality [14, 22–24]. It is crucial to consider the neuromuscular prerequisites for achieving higher muscle forces during the eccentric than concentric phase in FW resistance exercises. However, these are rarely addressed in the research papers, which predominantly focused on the mechanical characteristics of devices [22, 23, 25]. Furthermore, we identified a lack of studies on the role of eccentric overload in neuromuscular adaptations over time. This raises the question of whether eccentric overload is the primary factor driving the adaptations observed in FW resistance training studies. It is possible that improvements in performance after FW resistance training may not necessarily be attributable to eccentric overload.

Mechanical Factors Underpinning Eccentric Overload Using Flywheel Devices

In FW resistance exercise, the inertia of the FW, combined with the dimensions of its shaft, typically shaped as a cylinder or a cone, provides resistance. The FW is mounted on a force frame via this shaft, allowing it to rotate around its longitudinal axis. At the start of each repetition, a strap or a rope is wound around the shaft of the FW. Then, the rope unwinds through the concentric contraction of the muscles. When the rope/strap reaches its length, the FW keeps spinning by virtue of its inertia in the same direction, but the strap rewinds around the shaft in the other direction, which causes the rope to shorten. The aim of the exercise is then to decelerate (or brake) the spinning of the FW by resisting the pulling rope with eccentric contraction.

True eccentric overload can theoretically be achieved by the shorter duration or smaller range of motion (RoM) of the eccentric phase in comparison to the concentric phase of the repetitions (please see the Supplementary material for the details). Martínez-Hernández [24] introduced key practical methods to achieve eccentric overload; (a) increasing the RoM during the concentric phase or reducing it during the eccentric phase, (b) using assisted overload, and (c) alternating exercises (e.g., performing a squat during the lifting phase and a deadlift during the lowering phase) or varying limb involvement between phases (e.g., lifting with both limbs and lowering with one). In practice, elongated straps or ropes are also used to delay rope recoil at the end of the concentric phase. While these approaches are highly practical, there is limited research supporting their effectiveness.

It is important to note that the metric (peak or average) and the variable being assessed (force, velocity, or power) can influence the effectiveness of achieving eccentric overload. Peak power and peak velocity, which have been highlighted as primary metrics for eccentric overload in a previous study [22], represent brief, time-limited events within an unstable signal during the eccentric phase of each repetition. Additionally, type of metric used to assess eccentric-overload can be influenced by execution technique. From a physical perspective, catching the FW in the final third of the eccentric phase (at high velocity) may be more effective for maximizing eccentric overload in terms of peak power. In contrast, smoothly transitioning into braking with minimal resistance at the start of the eccentric phase, followed by forcefully stopping the rotation near the end of RoM, may lead to greater eccentric overload in peak force, due to smaller changes in movement velocity. This, based on practical experience, is easier to achieve with higher FW inertias, where the movement through the RoM is slower. However, these theoretical frameworks and technical variations require validation in future studies. In our opinion, delaying the braking action (e.g., in the first third of the eccentric phase) appears to be the most effective method to induce eccentric overload. This approach does not negatively impact the RoM of the subsequent concentric (shortening) phase, maintains stability, and therefore allows for high FW kinetic energy (W_k) production during concentric phase which transfers to the subsequent eccentric phase, while also ensuring a smooth transfer from the eccentric to the concentric phase [26].

Influence of Flywheel Device Structure on Eccentric Overload

FW device shaft shape (cylinder or cone), pulley systems (fixed pulley or moveable pulley) used with the rotating FW, rope/strap compliance, friction of the rotary bearings, and additionally moving device body parts with its own mass (also named as parasitic inertia) all influence attainment of eccentric overload [27]. The latter three could negatively influence W_k transition from concentric to eccentric phase of the repetition, while the energy stored in the FW during the concentric phase of the repetitions can be wasted as a heat or transformed to other forms of energy.

Special focus should be placed on the FW device shaft shape. Using a cone shaped shaft, the instantaneous radius of the rope coil changes along the exercise RoM [23], typically being the largest at the transition between the eccentric (braking) and concentric (pulling) phases of repetitions. This allows participants to brake or pull with less effort. Smaller diameters of a cone shaped FW [28] and cylindrically shaped shafts [22] enable achievement of greater eccentric overload values. Using a strap recoiling on itself around the cylindrical shaft also increases the instantaneous radius along the exercise RoM which negatively influences eccentric overload production.

In addition, the use of a movable pulley, which is usually attached to the lifting harness, increases the intensity of the exercise without altering the FW inertia. A pulley system theoretically reduces the velocity of the movement by half, allowing muscles to generate higher forces during the concentric contraction phase [29]. However, this is not true for the eccentric contraction phase, where muscle force production capacity theoretically increases with greater contraction velocity [30]. This suggests that shaft shape [31] and the use of a movable pulley can affect eccentric overload differently, depending on the magnitude of FW inertia used. The unique designs and mechanical properties of FW devices likely vary in eccentric overload, and between-device comparisons should therefore be made with caution.

Physiological Factors Underpinning Eccentric Overload Using Flywheel Devices

It is known that men have a lower ratio of maximal eccentric-to-concentric strength than women [6]. The reasons are unclear, but it is speculated that men may engage more in traditional strength training, which tends to stimulate concentric strength more than eccentric strength [6]. Specifically, data from FW resistance training show that when the magnitude of the FW inertia increases (reduced movement velocity), eccentric overload tends to increase in men but not in women [32].

Eccentric overload generation is conditioned by resistance training experience. A minimum of 2–3 familiarization sessions were suggested for participants to become acquainted with the FW device [33, 34] and to stabilize mechanical output variables [34]. Additionally, force-producing capacity during high load eccentric contractions is highly influenced by activation of protective mechanisms, such as Golgi tendon organ inhibitory feedback via Ia afferent fibers [35]. Strength training reduces the inhibition of protective mechanisms during eccentric contractions and therefore eccentric overload is different between strength trained and untrained participants and changes over the training period [36–39].

The ratio between eccentric and concentric strength is largely affected by velocity, with a 0.2% increase in the ratio for every 1°/s increase in isokinetic single joint angular velocity [6]. An increase in FW inertia lowers concentric and eccentric velocity and increases concentric and eccentric force in a linear [26, 40] or a curvilinear [41, 42] relationship. However, the use of lower or higher inertias yields inconclusive results regarding their favorability in achieving higher eccentric overload [34, 41, 42]. A recent review article [22] found that to attain higher peak power or peak velocity in the eccentric phase of the exercise it is more suitable to use lower inertias. However, it has to be emphasized that eccentric overload in terms of instantaneous velocity cannot be considered a physiologically relevant stimulus for muscle adaptations, because the primary driver of neuromuscular adaptations during eccentric contractions is generally regarded as mechanical stress on the muscle-tendon complex [43]. Furthermore, it is most likely that the peak velocity during the eccentric phase is reached while unloading the pulling rope or strap at the initiation of eccentric phase, at which point no resistance from the FW is applied to the agonist muscles (please see Fig. 3 in the study of Spudić et al. [26]). Therefore, this peak velocity reflects the movement quality rather than the actual mechanical stress imposed on the muscles.

Furthermore, since eccentric force and power production are constrained by the W_k generated during the concentric phase, lower inertias (higher velocity movements) limit the ability to achieve high absolute eccentric forces, although its contrary might be expected from the well-known shape of eccentric F-V relationship [6, 30]. This constraint could result in a disproportionately greater stimulus for increasing concentric than eccentric muscle strength.

Age could also influence eccentric overload manifestation, because eccentric strength is better preserved with aging than concentric strength due to neurological, mechanical and cellular mechanisms [44]. This suggests that FW training might be particularly beneficial for the elderly population [45, 46]. Young adults exhibit higher fatigue tolerance during eccentric than concentric contractions, while older adults experience similar levels of fatigability between the two [47]. This indicates distinct prerequisites for attaining eccentric overload within a set or between sets when fatigue accumulates during FW resistance training in older adults.

Joint angle and consequently muscle length may also influence the achievement of eccentric overload [48]. Eccentric muscle force production relies not only on contractile elements but also on noncontractile elements, which are stretched more at longer muscle lengths. In contrast, concentric muscle force generation may be impaired at these longer lengths. The resistance provided by a FW device is highly variable throughout the exercise’s RoM. Resistance is the greatest at the initiation of the FW spin and decreases as the RoM increases due to the nature of FW inertia. For example, during a FW squat with thighs parallel to the ground, maximal resistance occurs at the start of the lifting phase when the knees and hips are flexed. In this position, the knee and hip extensor muscles are stretched, which impairs concentric force generation but favors eccentric force production due to the relatively high contribution of passive muscle force, making eccentric overload easier to achieve. In contrast, during a quarter squat, where the transition between the lowering and lifting phases occurs at more extended knee and hip angles, the muscles are in a more optimal position for concentric force generation, but the contribution of passive muscle force during the eccentric phase may be lower.

FW resistance exercises can be performed with maximal or submaximal effort in the concentric phase of the repetitions. In case of submaximal execution, less W_k is transferred from the concentric to the eccentric phase. By the law of conservation of energy, this results is lower absolute eccentric muscle loading in comparison to maximal repetition execution [49]. Nevertheless, based on eccentric exercise training studies, submaximal load in the eccentric phase of the repetition appears to be still sufficient for strength and muscle mass gains in the general population [50, 51] but not for trained athletes [5].

Lastly, the studies comparing concentric and eccentric muscle strength are limited to single joint movements [6], and the ratio between the concentric and eccentric muscle strength during multi-joint exercises has not been widely investigated in free-weight exercises [30, 52]. In contrast, the squat, which is one of typical multi-joint exercises, is among the most studied exercise in the context of FW resistance exercise [11, 53, 54]. Recent research showed that peak eccentric forces were only about 10% greater than isometric forces in maximal iso-velocity squatting [30]. This is consistent with studies on eccentric overload in multi-joint FW resistance training, which showed that eccentric peak force did not surpass 20% of the concentric peak force [22, 26]. This corresponds to the speculations of Núñez et al. [14], suggesting that eccentric overload may be lower in multi-joint exercises when compared to single-joint exercises, but further studies are needed to validate this assumption.

Based on physiological prerequisites, it could be speculated that there is a significant intra-individual variation in the ability to achieve eccentric overload. Moreover, the interaction between the factors discussed above plays a role in influencing eccentric overload. As a result, the effects of FW exercise training vary such that some participants may derive greater benefits than others. Reasons for individual differences in response to FW exercise training should be investigated further.

Challenges in Maximizing Eccentric Contractions with Flywheel Resistance Training

Only a small portion of the previous FW studies reported that they actually achieved eccentric overload [14, 22–24], which may indicate a general lack of understanding in previous research on how to effectively attain eccentric overload [24]. To address this, we have summarized the currently known prerequisites for achieving true eccentric overload in Fig. 1.

Fig. 1.

A graphical overview of our discussion showing the factors and metrics responsible for achieving higher eccentric load using flywheel devices

Increase in muscle strength is greater with the existence of eccentric overload during the FW resistance exercise [54]. This may be attributed to the enhanced morphological and architectural adaptations in the muscle-tendon unit that occur when training with higher eccentric loads [36, 38, 55]. Nevertheless, it remains unclear to what extent eccentric force generation capacities are utilized in FW resistance training. In the meta-analysis by Muñoz-López and colleagues [22], eccentric peak force reached a maximum of 120% of concentric force in a single-leg extension task, which aligns with the study by Spudić et al. [26] showing 17% greater peak forces for eccentric than concentric phase during FW squats. It is likely that eccentric loading during FW exercises does not fully exploit the maximal eccentric force generation capabilities, which are known to be approximately 40% higher than maximal concentric contraction force under isokinetic conditions [6]. The fact that arm assistance during the concentric phase enhances eccentric overload supports the assumption that eccentric force production in FW squats is limited by concentric W_k production [49].

It is generally accepted that neuromuscular adaptation mechanisms underpinning increases in muscle strength are velocity specific [56]. It is important that not only the magnitude of the eccentric overload but also the load across the entire RoM defines the specificity of adaptation. For example, comparing training with a 10% eccentric overload at higher FW inertias (characterized by low movement velocity and high absolute forces) and training with lower inertia (high movement velocity and lower absolute forces) would provide limited insights without considering other training variables (tempo, training volume, rest). However, there is currently little evidence whether velocity specificity extends to FW modality [57], while quantifying inertial load magnitude remains a limitation of FW resistance training. The definition of high and low FW inertial loads remains an open question, since both the mass moment of inertia and at least the shaft diameter should be considered for the definition of inertial load magnitude. Different device characteristics limit this generalizability. Spudić and colleagues [40] have recently proposed the F-V testing procedure for FW squats. This procedure provides insight into the lower extremity mechanical output abilities of the neuromuscular system across a range of FW inertias and presents a potential method for individualizing FW load in resistance training [40, 57], as it accounts for actual force and velocity production during movement. Nevertheless, the eccentric phase of the exercise is neglected, despite FW being considered an eccentric-overload training modality. Defining the FW inertia that elicits peak power within a spectrum of loading conditions might serve as a reference point for distinguishing between low (below peak-power load) and high (above peak-power load) FW inertial loads, although this applies only to a specific FW device. Furthermore, load magnitude could also be interpreted through the movement velocity of a given FW exercise (e.g., FW squat) in relation to a functional task (e.g., a slalom skiing turn), where faster-than-functional execution represents a lower load, and slower execution represents a higher inertial load. Regardless of the eccentric overload, higher FW inertias result in greater overall force demands and longer time under tension within a single repetition. Drawing parallels with traditional resistance exercise [58, 59], it could be speculated that higher FW inertias might be more beneficial for improving maximal strength and hypertrophy due to the greater mechanical stimulus [60], while lower FW inertias might be more effective for enhancing peak power production [61]. The addition of eccentric overload could further amplify these benefits [54]. Nevertheless, the transfer of load-specific adaptations achieved under FW loading conditions to real-life performance remains largely unexplored.

Fatigue resilience in the eccentric phase of repetitions is much higher than in the concentric phase [62, 63]. Therefore, it can be speculated that failure during consecutive repetitions using FW device is primarily limited by concentric fatigue, similarly to traditional weightlifting exercises [6]. Since eccentric force generation is constrained by concentric force production, the accumulation of fatigue over consecutive FW repetitions in the concentric phase further limits force production in the eccentric phase. Consequently, training under fatigued conditions might provide a greater stimulus for enhancing concentric rather than eccentric muscle capabilities, although this speculation requires confirmation in future studies.

The effects of resistance exercise training protocols are not determined solely by the load and number of repetitions in the exercise but also by reaching to failure or not, RoM, muscle length, time under tension, internal/external focus, rest and mode of contraction [64]. We believe that comparisons of training effects between FW and traditional resistance training modalities (as well as eccentric-focused modalities [65]) are not valid unless at least the relative intensity is clearly defined and equalized between them, given that the intention behind moving the load and the tempo of exercise execution often differ across modalities. For example, comparing an “all-out” execution using the FW modality with a slow concentric–eccentric execution using traditional weights involves fundamentally different neuromuscular control, and consequently distinct mechanical and physiological responses. These differences also reflect divergent training purposes, for instance, power development versus hypertrophy, regardless of the training modality or device used. A more meaningful approach might be to compare the most effective training protocols for enhancing specific neuromuscular characteristics (e.g., strength, power or endurance) within each modality. Nevertheless, such comparisons remain limited in their interpretability until the load-dependent adaptations and optimal dose–response relationships for specific FW resistance training methods are clearly established. Thus, more high-quality studies are warranted to address this issue. Finally, while FW devices primarily enable the achievement of instantaneous eccentric overload, it is important for future practice and research to distinguish FW device eccentric overload from eccentric overload occurring across the entire eccentric RoM (in terms of force, velocity, or power).

Conclusion

FW resistance training is a practical and highly effective modality for improving physical performance and quality of life. However, it can only be considered an eccentric-overload training method if higher mechanical values are produced and measured during the eccentric phase of repetitions. It may be that the load in the eccentric phase during FW training is submaximal, and limited to a very short time window rather than the entire eccentric RoM. This may lead to a disproportionately greater stimulus for increasing strength and power in concentric than in eccentric contraction.

Proper exercise technique is critical for trainers to emphasize when familiarizing participants with FW equipment. In our opinion, eccentric overload in terms of force is most effectively achieved by delaying the braking action at the beginning of the eccentric phase. However, individual characteristics such as age, sex, training history, movement velocity (inertia magnitude), fatigue tolerance, joint angle, muscle length, and movement complexity all affect eccentric strength. Therefore, eccentric overload may not always be attainable, even when mechanical prerequisites are met. Furthermore, according to the laws of physics, the achievement of eccentric overload is not unique to FW devices, since it is primarily dependent on exercise tempo, which can also be achieved by traditional weights. It is necessary for researchers to clarify if eccentric overload is actually provided during training and testing to establish it as a key factor in driving training adaptations in future FW resistance training studies. If FW resistance exercises are properly performed by understanding the characteristics of FW, this results in eccentric overload in resistance exercises, which is likely to contribute to greater muscle adaptations and improvements of exercise and sport performance.

Author contributions

DS performed the analyses, visualized the data, and wrote the first draft of the manuscript. DS and KN contributed equally to the conception and design of the study, interpretation of the data, drafting and critical revision of the manuscript. The authors read and approved the final manuscript.

Funding

The research was conducted under the research programme Kinesiology of Monostructural, Polystructural and Conventional Sports, code: P5-0147, financed by the Public Research Agency of the Republic of Slovenia for author DS. Additionally, DS received financial support for a visiting professorship at the School of Medical and Health Sciences, Edith Cowan University in Australia for three months, through the ‘[RSF] Internal Call for Co-Financing Mobility of Assistants, Assistants with Doctorates, and Higher Education Teachers (Educational Staff) at Higher Education Institutions Abroad for 2023–2024 (B.II.3)’ from the University of Ljubljana, Slovenia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no conflicts of interest relevant to the content of this review.