The Role of Stiffness, Part 1: Understanding a Mechanical Property

The term stiffness is used in two fairly different ways in sport science. In biomechanics it refers to a specific and measurable quantity with an established definition. In everyday practice, in conversations about sprinting, plyometrics, tendon health or movement quality, it more often functions as a general description of how an athlete moves and how their ground contacts appear. Both usages have their place, but they do not carry the same information, and the second is not a substitute for the first when the question is what to train and how much. The mechanical definition is the more useful starting point, since it describes behaviour that can be measured, predicted within limits, and influenced through training.

In biomechanical terms, stiffness describes the resistance of a body to deformation under load. It is expressed as the ratio between the change in applied force and the resulting change in length or angle, ΔF/Δx. A stiffer structure deforms less for a given force; a more compliant structure deforms more. That definition is deliberately narrow, and its narrowness is what makes it useful. It allows us to distinguish between the stiffness of a tendon, the stiffness of a joint, and the stiffness of the whole body behaving as a mechanical system, and to recognise that these are related but not interchangeable quantities. It also allows us to separate mechanical stiffness from perceived stiffness, which is a sensory and clinical phenomenon that may or may not correspond to what is happening mechanically.

The Force–Deformation Curve and What It Describes

The most direct way to represent stiffness is through the force–deformation curve, which describes how a tissue or a system responds to progressively increasing load. In most biological tissues the curve has a characteristic shape. There is an initial region of relatively high compliance, sometimes described as the toe region, where small forces produce comparatively large deformations as collagen fibres align. This is followed by a region that is approximately linear, where the relationship between force and deformation is more stable and where stiffness can reasonably be approximated by the slope of the curve. Beyond that point the tissue enters a plastic region, where deformation is no longer fully recoverable, and eventually reaches failure.

Stiffness is the local slope of the curve, not a single fixed value: the same tissue is more compliant in the toe region than in the linear region.

Stiffness, then, is the local slope of this curve rather than a single fixed number. A healthy tendon has a wide elastic region and a slope appropriate to storing and returning energy without entering plastic deformation. A tendon with degenerative change may show a flatter or more irregular curve, with reduced capacity to store energy and greater susceptibility to accumulated micro-damage. This is one reason why two athletes with similar maximal force outputs can respond very differently to the same plyometric exposure.

There is a further consideration that is easy to overlook in a static representation. Biological tissues are viscoelastic, which means their mechanical behaviour depends on the rate at which load is applied. At high loading velocities the tissue responds in a stiffer manner than it does at low velocities. This has a practical consequence: a tendon may tolerate slow, heavy loading well while struggling with explosive loading of a comparable peak magnitude, because the rate of loading changes the effective mechanical demand.

Three Levels of Stiffness, and Why They Are Not the Same Thing

When the analysis moves from isolated tissue to the whole athlete, the literature generally distinguishes three levels of measurement (Brughelli and Cronin, 2008; Maloney and Fletcher, 2021).

Vertical stiffness (K_vert) is the ratio between peak vertical force and the vertical displacement of the centre of mass during ground contact. It is a global measure of how much the body descends relative to the force it applies. Leg stiffness (K_leg) treats the lower limb as a spring that compresses and extends, and relates force to the effective shortening of the limb between the hip projection and the point of contact. Joint stiffness refers to individual articulations, most often the ankle and knee, and is expressed as the change in joint moment divided by the change in joint angle. It is influenced by muscular co-contraction, passive tissue properties and neuromuscular control.

These three levels are related but not reducible to one another, and an athlete may present with a favourable value at one level and an unfavourable one at another. It is possible to observe acceptable leg stiffness alongside insufficient ankle joint stiffness, which tends to produce a distal yielding that the more proximal segments then have to compensate for. The reverse pattern also occurs, with a very rigid ankle combined with poor coordination of the limb as a whole. Reliability also differs between measures: vertical stiffness tends to be the most reproducible across tasks, while joint stiffness is the least, with ankle measurements generally more stable than knee measurements (Maloney and Fletcher, 2021). This matters when interpreting any single test result.

These quantities have historically required force plate analysis, which restricted their use to laboratory settings. Validated field-based equations now allow whole-body stiffness to be estimated from spatiotemporal variables without a force platform, which has made the measures considerably more accessible in applied environments (Maloney and Fletcher, 2021). Accessibility, however, does not remove the interpretive problem. A stiffness value obtained from a hopping protocol, a drop jump and a running trial does not describe the same underlying construct, and values are not directly comparable across tasks or across testing conditions.

A further point concerns the environment. The effective stiffness of the system is not determined by the athlete alone. Human runners adjust their leg stiffness in response to surface stiffness so that the combined stiffness of leg and surface remains relatively constant (Ferris, Louie and Farley, 1998). Part of this adjustment appears to be automatic, and part of it is likely trainable, which has implications for how we interpret data collected on different surfaces.

"A healthy tendon has a wide elastic region and a slope appropriate to storing and returning energy without entering plastic deformation. A tendon with degenerative change may show a flatter or more irregular curve, with reduced capacity to store energy and greater susceptibility to accumulated micro-damage. This is one reason why two athletes with similar maximal force outputs can respond very differently to the same plyometric exposure."

The Spring-Mass Model and the Idea of an Optimum

The spring-mass model represents the body as a single mass resting on a linear spring. During ground contact the spring compresses, stores elastic potential energy, and releases part of it during propulsion. Despite its simplicity, the model captures a substantial proportion of the variance in running and jumping mechanics and remains a reasonable starting point for practical reasoning.

Within this framework the consequences of insufficient and excessive stiffness are both visible. When stiffness is too low, the spring compresses excessively: the centre of mass descends further, ground contact times lengthen, and a greater proportion of the stored energy is dissipated rather than returned. When stiffness is too high, compression is minimal, peak forces rise, and the mechanical demand placed on tendons and articular structures increases. The implication is that the objective is not to maximise stiffness but to identify a value appropriate to the task and the velocity at which it is performed. That value is considerably higher for maximal-velocity sprinting than for slow running or for a change of direction involving a large braking component.

Running velocity itself modulates these variables in a non-uniform way. As speed increases from slow to moderate values, leg stiffness tends to remain relatively constant while vertical stiffness and joint stiffness increase (Brughelli and Cronin, 2008). Longitudinal observations point in a similar direction: improvements in maximal sprinting speed over a training macrocycle have been accompanied by increases in vertical and ankle joint stiffness, with leg and knee stiffness largely unchanged (Nagahara et al., 2017).

From One Mass to Two

The single-mass representation has a structural limitation. By treating the body as one mass on one spring, it describes global phenomena well but cannot account for the differential behaviour of body segments during the very short contacts that characterise high-speed running. The two-mass model addresses this by dividing the body into a lower mass, corresponding approximately to the contacting foot and shank, and an upper mass comprising the remainder (Clark, Ryan and Weyand, 2017). The two are coupled through the stiffness of the intervening joint and muscle–tendon structures.

The two-mass model separates the contacting foot and shank from the remainder of the body, which is where most of the deformation and elastic return during contact occurs (Clark, Ryan and Weyand, 2017).

The two masses do not move identically during contact. The lower mass undergoes a rapid deceleration and an equally rapid re-acceleration, while the upper mass follows a slower and more damped trajectory. Vertical ground reaction force waveforms can be reconstructed with high accuracy from body mass, contact and aerial times, and the vertical acceleration of the lower limb during impact, across a wide range of speeds, foot-strike patterns and footwear conditions (Clark, Ryan and Weyand, 2017; Udofa et al., 2019).

The practical reading of this is that most of the deformation and elastic return during contact is associated with the lower mass, while the upper mass contributes principally to stability and to the orientation of the movement. Ankle and Achilles tendon stiffness therefore have a disproportionate influence on how quickly the lower mass can reverse direction, while trunk position and control determine whether the resulting force is transmitted effectively or partially dispersed. This helps explain a pattern that is common in practice: athletes who are strong and mechanically capable at the distal level but appear unreactive often present a problem of coordination and timing between segments rather than a deficit in force production or in stiffness considered in isolation.

Stiffness Within the Stretch-Shortening Cycle

The stretch-shortening cycle describes the sequence in which a muscle–tendon unit is first lengthened and then shortened, with a brief transition between the two phases. Stiffness intervenes at two points in this sequence: it influences how much elastic energy is stored during lengthening, and how much of that energy is returned rather than dissipated during shortening.

In fast stretch-shortening cycle actions, conventionally defined by contact times below approximately 250 ms, the tendinous component and the stiffness of the system as a whole are particularly influential, while maximal voluntary force contributes relatively less than it does in slower actions (Komi, 2000). Adequate stiffness limits joint yielding during the eccentric phase, preserves favourable joint angles, and shortens the transition between phases. Where stiffness is insufficient, energy is lost to excessive deformation. Where it is excessive, the capacity to attenuate load is reduced and the mechanical demand on passive structures rises.

The speed of the eccentric-to-concentric transition depends on the interaction between neuromuscular coordination, rate of force development and viscoelastic tissue properties, and this interaction is sensitive to fatigue. Repeated stretch-shortening cycle work produces a characteristic bimodal deterioration in neuromechanical performance, with an immediate reduction, partial recovery within a few hours, and a secondary decline coinciding with peak muscle soreness (Nicol, Avela and Komi, 2006).

"Adequate stiffness limits joint yielding during the eccentric phase, preserves favourable joint angles, and shortens the transition between phases. Where stiffness is insufficient, energy is lost to excessive deformation. Where it is excessive, the capacity to attenuate load is reduced and the mechanical demand on passive structures rises."

Three Related But Distinct Constructs

Rate of force development, stiffness and tendon elasticity are frequently discussed together and are sometimes treated as approximate synonyms. They are not. Rate of force development describes how quickly force can be generated. Stiffness describes resistance to deformation. Tendon elasticity describes the capacity of the tendon to lengthen and shorten while returning energy.

The three interact in a fairly direct way. A high rate of force development allows tendon loading to be achieved within very short time windows. Favourable tendon mechanical properties allow a meaningful proportion of that load to be stored and returned. Adequate system stiffness ensures that the returned energy contributes to propulsion rather than being absorbed by joint yielding. An athlete can nevertheless present with a high rate of force development and suboptimal system stiffness, typically because of limitations in intersegmental coordination or joint control. Fatigue alters the relationship further: rate of force development declines, leg stiffness tends to decrease, contact times lengthen, and tendon loading may shift toward less favourable conditions (Darch et al., 2022; Masson et al., 2026). This is one plausible mechanistic account of why contact quality deteriorates in the closing stages of a match or a demanding session.

The Ankle as the Primary Spring

Across sprinting and many explosive actions, the ankle functions as the primary spring of the lower limb. The triceps surae complex and the Achilles tendon account for a substantial proportion of the elastic energy stored and returned during contact. Mechanically, the ankle is required to resist excessive dorsiflexion during the eccentric phase, to permit a rapid transition into an effective concentric phase, and to transmit force efficiently toward the ground.

The triceps surae and Achilles tendon account for a substantial proportion of the elastic energy stored and returned during ground contact.

The available data suggest a degree of specialisation in how this is regulated. Across increasing sprint velocities, average ankle joint stiffness has been reported to remain approximately constant while knee joint stiffness increases, which has been interpreted as indicating that the spring-like behaviour of the limb is adjusted principally at the knee, with the ankle constrained by the dominant role of tendon stiffness within the triceps surae unit (Kuitunen, Komi and Kyröläinen, 2002). Dynamic ankle stiffness also appears to be more a property of the task than of the individual, with higher values in sprinting than in submaximal running (Stefanyshyn and Nigg, 1998). At the same time, ankle stiffness during the push-off phase of the sprint start relates to performance variables in a phase-dependent manner, with opposite associations during the negative and positive power phases (Charalambous et al., 2012).

In applied terms, athletes who use the ankle as the principal generator of stiffness and propulsion tend to display shorter contacts and a more direct transmission of force, whereas those who compensate with large hip excursions tend toward longer contacts and lower elastic efficiency. The distinction is a matter of degree rather than category, and it is influenced by anthropometry, training history and the specific demands of the sport.

The structures distal to the ankle also contribute more than is often assumed. Metatarsophalangeal flexion torque, passive foot stiffness and foot–ankle reactive strength together account for a meaningful proportion of the variance in effective vertical impulse and contact time during maximal-velocity sprinting (Tourillon et al., 2024). This suggests that the elastic behaviour observed at the level of the limb depends on a chain of structures rather than on a single joint, and that assessment restricted to the ankle joint may miss part of the picture.

Where This Leads

Understanding stiffness as a measurable property, distributed across tissue, joint and system levels, and modulated by velocity, surface and fatigue, provides the conceptual basis for any subsequent decision about training. It does not, by itself, tell us what a given athlete needs. That requires an additional layer of reasoning about the mechanical demands of the specific task, the trade-off between performance and tissue tolerance, and the progressions available for influencing stiffness in a controlled way. Those questions are addressed in the second part of this series.

---

Frequently Asked Questions

Q: Is higher stiffness always better for performance?

No. Within a certain range, greater stiffness is associated with improved performance in rapid actions, because it limits energy dissipation and shortens ground contact. Beyond that range, peak forces rise and the capacity to attenuate load declines, which increases the mechanical demand on tendons and bone. The relevant question is not how much stiffness an athlete has in absolute terms, but whether the value is appropriate for the task being performed and tolerable given that athlete's history and current training load.

Q: What is the practical difference between vertical, leg and joint stiffness?

Vertical stiffness relates peak vertical force to the vertical displacement of the centre of mass and describes the behaviour of the whole body. Leg stiffness models the lower limb as a single spring and relates force to the effective shortening of that limb. Joint stiffness describes individual articulations and is the most locally specific of the three. They are related but can diverge, so an athlete may present adequately at one level and poorly at another. Vertical stiffness is generally the most reliable measure, while joint stiffness is the least.

Q: Why does the two-mass model matter if the spring-mass model already works?

The spring-mass model describes global behaviour well but treats the body as a single mass, which limits its ability to explain what happens during very short contacts. The two-mass model separates the contacting lower limb from the remainder of the body and shows that most of the deformation and elastic return is associated with the lower mass. This reframes trunk control as a determinant of force transmission rather than a separate concern, and helps explain why some strong athletes remain unreactive.

Q: Does stiffness change with fatigue, and how quickly?

Leg stiffness tends to decrease with accumulated fatigue, accompanied by longer ground contact times and greater step-to-step variability. Recovery of stretch-shortening cycle function is not linear: an immediate decline is typically followed by partial recovery within one to two hours and a secondary decline coinciding with peak muscle soreness, with full restoration sometimes requiring several days depending on the severity of the exposure.

Q: Can tendon stiffness itself be modified through training?

Yes. Mechanical loading produces moderate increases in tendon stiffness, with the effect mediated primarily by changes in tendon material properties rather than cross-sectional area (Lazarczuk et al., 2022). Protocols involving higher localised tendon strain appear to produce greater adaptation than low-strain protocols, and resistance training has been associated with larger increases in modulus than other training modalities. Adaptation of this kind develops over months rather than weeks, which is one reason why stiffness is better treated as a medium-term training objective than as something responsive to short interventions.

References

• Brughelli, M. and Cronin, J. (2008) 'Influence of running velocity on vertical, leg and joint stiffness: modelling and recommendations for future research', Sports Medicine, 38(8), pp. 647–657.

• Charalambous, L., Irwin, G., Bezodis, I.N. and Kerwin, D. (2012) 'Lower limb joint kinetics and ankle joint stiffness in the sprint start push-off', Journal of Sports Sciences, 30(1), pp. 1–9.

• Clark, K.P., Ryan, L.J. and Weyand, P.G. (2017) 'A general relationship links gait mechanics and running ground reaction forces', Journal of Experimental Biology, 220(2), pp. 247–258.

• Darch, L., Chalmers, S., Wiltshire, J., Causby, R. and Arnold, J. (2022) 'Running-induced fatigue and impact loading in runners: a systematic review and meta-analysis', Journal of Sports Sciences, 40(12), pp. 1338–1352.

• Ferris, D.P., Louie, M. and Farley, C.T. (1998) 'Running in the real world: adjusting leg stiffness for different surfaces', Proceedings of the Royal Society B: Biological Sciences, 265(1400), pp. 989–994.

• Komi, P.V. (2000) 'Stretch-shortening cycle: a powerful model to study normal and fatigued muscle', Journal of Biomechanics, 33(10), pp. 1197–1206.

• Kuitunen, S., Komi, P.V. and Kyröläinen, H. (2002) 'Knee and ankle joint stiffness in sprint running', Medicine and Science in Sports and Exercise, 34(1), pp. 166–173.

• Lazarczuk, S.L., Maniar, N., Opar, D.A., Duhig, S.J., Shield, A., Barrett, R.S. and Bourne, M.N. (2022) 'Mechanical, material and morphological adaptations of healthy lower limb tendons to mechanical loading: a systematic review and meta-analysis', Sports Medicine, 52(10), pp. 2405–2429.

• Maloney, S.J. and Fletcher, I.M. (2021) 'Lower limb stiffness testing in athletic performance: a critical review', Sports Biomechanics, 20(1), pp. 109–130.

• Masson, L., Morin, J.B., Millet, G.Y. and Giandolini, M. (2026) 'Stiffness and running performance: from the tissue to the body. A narrative review', Sports Medicine, advance online publication.

• Nagahara, R., Naito, H., Miyashiro, K., Morin, J.B. and Zushi, K. (2017) 'Development of maximal speed sprinting performance with changes in vertical, leg and joint stiffness', The Journal of Sports Medicine and Physical Fitness, 57(12), pp. 1572–1578.

• Nicol, C., Avela, J. and Komi, P.V. (2006) 'The stretch-shortening cycle: a model to study naturally occurring neuromuscular fatigue', Sports Medicine, 36(11), pp. 977–999.

• Stefanyshyn, D.J. and Nigg, B.M. (1998) 'Dynamic angular stiffness of the ankle joint during running and sprinting', Journal of Applied Biomechanics, 14(3), pp. 292–299.

• Tourillon, R., Bothorel, H., Gojanovic, B., Fourchet, F. and McKeon, P.O. (2024) 'Human foot muscle strength and its association with sprint acceleration, cutting and jumping performance, and kinetics in high-level athletes', Journal of Sports Sciences, 42(11), pp. 1013–1024.

• Udofa, A.B., Clark, K.P., Ryan, L.J. and Weyand, P.G. (2019) 'Running ground reaction forces across footwear conditions are predicted from the motion of two body mass components', Journal of Applied Physiology, 126(5), pp. 1315–1325.

Antonio Robustelli is the founder of Omniathlete. He is an international high performance consultant and sought-after speaker in the area of Sport Science and Sports Medicine, working all over the world with individual athletes (including participation in the last 5 Olympics) as well as professional teams in soccer, basketball, rugby, baseball since 24 years. Currently serving as Faculty Member and Programme Leader at the National Institute of Sports in India (SAI-NSNIS).