Evidence from the present investigation suggest that skeletal muscle shear elastic modulus is adaptable in marathon runners; in that it appears modifiable prior to, in response to, and in the recovery from marathon running. Consistent with our hypothesis, we observe that pre-marathon muscle shear elastic modulus is relatively lower in both competitively experienced runners and runners with higher exercise performance capacity, as evidenced by faster marathon finish times. Second, while muscle shear elastic modulus increases following long-distance marathon running appear independent of experience level, sex, or course, this response is significantly blunted in runners who wore highly cushioned shoes with plates. Lastly, recovery time following marathon running is longer in runners who have the largest positive pre-to-post marathon change in muscle shear elastic modulus. Taken together, a change in skeletal muscle shear elastic modulus may result as a function of acute and chronic exercise insults over time and drive a shared cycle of adaptation. Future work involving longitudinal interventions could help elucidate potential practical applications for monitoring and intentionally shifting the dynamic nature of passive muscle elastic properties and tolerances before, during and after exercise insult in humans.

We first sought to better understand measures of in-vivo skeletal muscle shear elastic modulus in marathon runners prior to any insult that would occur from racing a marathon. Explicitly, whether runners with greater experience or exercise performance capacity have a distinct presentation of quiescent skeletal muscle stiffness. Accordingly, the present study shows that shear elastic modulus of the rectus femoris muscle is lower in competitive vs recreational marathoners and that runners spanning a wide range of exercise performance capacity have markedly different pre-marathon shear modulus (Fig. 2). Moreover, faster marathon finish times are inversely correlated with a greater pre-marathon muscle shear elastic modulus across all participants (Fig. 2D), suggesting a fundamental relationship between these variables that outweighs experience level per se. When statistically controlling for experience, sex, age or body weight on pre-marathon shear modulus, average marathon running speed was the most potent variable. Racing a marathon is effectively a time trial running performance and therefore an index of exercise capacity. Whether low pre-marathon muscle shear elastic modulus acutely mediates a faster average speed in marathon distance races, or whether participants arrive to their race prepared to run faster marathons and are accompanied by lower elastic modulus by way of adaptations accrued beforehand is unclear. Regarding the latter however, there is some existing rationale for the presentation of low shear elastic modulus (i.e. low muscle stiffness) in fast marathon runners who have undergone very large and prolonged volume of exercise training coupled to arriving at a marathon in a well-rested state.

Reports as early as the 1920s from Gasser and Hill (Gasser and Hill 1924), as well as Ranson (Ranson 1928), began to suggest that persistent stress on muscle dissipates strain in a manner such that muscle tension re-equilibrates in a more permanent elongated state. Recent evidence shows eight weeks of eccentric loading via plyometric training resulted in a decrease in passive skeletal muscle stiffness in humans (Ando et al. 2021), while rodent models suggest a reduction in passive skeletal muscle stiffness associated with slow-twitch muscle presence after 4 weeks of exercise (Noonan et al. 2020). In cardiac muscle, while single acute bouts of exercise are known to augment titin-related passive muscle stiffness, chronic exercise programs result in the opposite response and appear to reduce passive stiffness (Slater et al. 2017; Lalande et al. 2017). Interestingly, this impulse-response-adaptation cycle observed in cardiac tissue is not unlike what could be surmised from the collective observations of the present study in skeletal muscle.

How low passive muscle stiffness could mechanistically enable higher sustained speeds for very long running durations is less clear, however we postulate that lower tissue elastic modulus may offset soft tissue requirements consequent to long distance running. One initial interpretation is that low muscle stiffness is a foundational adaptation useful for energy dissipation upon repeated ground contacts of high force. For example, reduced muscle fascicle elongation consequent to endurance running (Abe et al. 2000) could attenuate the advancement of muscle damage, wherein higher passive muscle elasticity (i.e. low stiffness) might spare the muscle from the degree of required eccentric force generation throughout the bout of exercise. Alternatively, low muscle stiffness could in series modulate tendon elasticity thus promoting tendon stretch and elastic energy storage. Lastly, it could be that changes in muscle shear elastic modulus enable enhancement of mechanical vertical leg stiffness and therefore augmented running economy. Nonetheless, a greater understanding on the mechanistic basis for the contribution of muscle elasticity to endurance running is warranted.

Unlike other endurance exercise modalities (i.e. cycling), the demands for success in long distance running are exclusive and require tissue storage of elastic energy, as well as a continuous revolution of eccentric-to-concentric muscle contraction transitions, all occurring at forces many times body weight (Alexander 1984; Cavagna 2006; Cavagna et al. 2011). High volume (eccentric) negative work from marathon running elicits prominent muscular microtrauma (i.e., muscle damage) and inflammation (Fridén et al. 1981; Hikida et al. 1983; Froeling et al. 2015; Hooijmans et al. 2020), and isolated small muscle eccentric contractions influence the extent of in vivo muscle damage in accordance to muscle strain (Guilhem et al. 2016). Shear wave elastography provides an index for a muscle’s resistance to non-permanent elastic deformation (i.e. the shear modulus) and can indicate early detection of exercise-induced muscle damage (Lacourpaille et al. 2017), as well as muscle stretch-related muscle soreness and declines in muscle force (Guilhem et al. 2016). From a functional standpoint, muscle swelling, muscle soreness, and muscle force declines are expected consequences of muscle structural damage (Fridén et al. 1981; Brooks et al. 1995; Chalchat et al. 2022), of which is seen in response to marathon running (Hikida et al. 1983; Froeling et al. 2015; Hooijmans et al. 2020). A second fundamental observation of the present study is that rectus femoris shear elastic modulus is on average significantly increased following marathon running (~ 23% for the present cohort) (Fig. 3A and C). Acute marathon-mediated increases in muscle shear elastic modulus were definitively widespread and not attenuated by experience or sex, nor was the change in shear elastic modulus from pre- to post-marathon different across course. These data are consistent with the central premise that exercising above an existing tissue tolerance (i.e. high magnitude of unaccustomed exercise, such as that consequent to racing a marathon vs a long training run) elicits a significant strain to human muscle via stretch and/or negative fascicle work that results in prominent changes in muscle elastic shear modulus and muscle damage (Guilhem et al. 2016; Shoji et al. 2021; Chalchat et al. 2022).

In contrast to previously mentioned biological factors, the external factor of underfoot cushioning, meaning the in-race wearing of highly cushioned footwear via Vaporfly vs Other during the marathon, mitigated marathon-induced elevations in shear elastic modulus (Fig. 3D). To further rule out that attenuation in shear modulus in the Vaporfly group was not simply due to a bias for faster runners/faster finish times wearing Vaporfly footwear, we proceeded to generate matched finish time groups (Table 1), yet the Vaporfly protection effect was retained in this sub-group analysis despite controlling for marathon finish times (Fig. 3E–F). Both external cushioning and surface compliance during running (via either surface type or footwear worn) appear to directly influence mechanical and metabolic responses to exercise (McMahon and Greene 1978; Kerdok et al. 2002; Tung et al. 2014; Brown et al. 2017), as well as modify leg compression and muscle fascicle stretch (Hollville et al. 2020). While multiple lines of evidence suggest that humans control landing strategies (to a drop landing or running) to modulate the degree of mechanical impact on the body (Skinner et al. 2015), particularly against negative muscle work, leg compression, and muscle fascicle lengthening (McMahon and Greene 1978; McMahon et al. 1987; Ferris et al. 1998; Kerdok et al. 2002; Hollville et al. 2020), the addition of cushioning within the system either by a surface type change or by footwear can decrease the absorption of the load through negative muscular work and muscle fascicle stretch (Kerdok et al. 2002; Tung et al. 2014; Skinner et al. 2015; Hollville et al. 2020). Such data are consistent with reports of reduced exercise-induced muscle damage and muscle soreness when running on softer surfaces (Williams et al. 2016; Brown et al. 2017). We recently reported preliminary observations indicating that highly cushioned footwear might modify the exercise-induced response to marathon running as evidenced through lower blood markers of muscle damage and inflammation, and associated muscle soreness (Kirby et al. 2019). Accordingly, if suppressed elevations in shear elastic modulus are accepted as an early indication of less tissue stiffness and skeletal muscle damage (Lacourpaille et al. 2017), footwear selection may be one simple, practical and effective method to circumnavigate such consequences, and thus ultimately minimizing downtime away from running and increasing the potential total training opportunity for runners.

It is estimated that approximately 1 million people each year run a marathon, perceiving the personal payoff greater than any anticipated penalty of musculoskeletal burden. We show that muscle soreness continues to be significantly elevated even after 3 days (Fig. 4A), in line with other reports suggesting that marathon running evokes significant muscle soreness in the days following the event (Fridén et al. 1981; Hikida et al. 1983; Tokinoya et al. 2020). Muscle soreness is often considered a prominent lagging indicator that signals the extent of a prior exercise load. However, the degree of momentary muscle soreness could also be reframed and considered as a leading indicator on the potential of future exercise load. Therefore, muscle soreness may act as a simple, real-world, impeding factor, or at least a delaying factor, to a runner’s ability or desire to run again following completion of an intense long-distance run. Accordingly, perceived readiness to run again (even just two miles) was suppressed still after 3 days following marathon running (Fig. 4B), and next run readiness was strongly correlated to the magnitude of muscle soreness, wherein on average, a 50% increase in soreness cut the likelihood that a person could perform a short segment of running at moderate intensity in half (Fig. 4C). Collectively, these data widen our view on how exercise responses to past insult may also act as potentially powerful indicators of future exercise capacity (or capability).

One practical conundrum presented for long distance runners is: that while eccentric contraction-mediated consequences, such as muscle soreness, are well accepted (Guilhem et al. 2016; Tokinoya et al. 2020), consistent running behavior is an essential prerequisite for high running performance. However, the extent of muscle soreness itself could dampen the likelihood of runners to run again (i.e. next run readiness). Given previous evidence highlighting time course patterns of muscle shear elastic modulus during isolated elbow flexor eccentric exercise (Lacourpaille et al. 2014), we hypothesized that a quicker muscle soreness recovery time might be observed in those with the least increase in elastic modulus, and the longest recovery for those with more prominent increases in shear modulus. We observed that runners who reported the fastest decay in muscle soreness (of 24 h) demonstrated a reduction in muscle shear elastic modulus, wherein those who reported a continuation of muscle soreness extending beyond 48 or 72 h had greater increases in elastic modulus in response to the marathon (Fig. 4D). These findings align with Shoji et al, (2021) who present data showing smaller increases in rectus femoris shear modulus in participants classified as ‘fast recovery’ following eccentric exercise. Taken together, these data suggest that early identification of passive muscle shear elastic modulus may describe an individual’s future performance capacity. In addition, the change in shear elastic modulus could act as a surrogate index to describe the potential recovery time window for muscle soreness and next run readiness (Fig. 4A–D).

Many studies examining muscle elastic properties focus on the change in muscle stiffness in response to a single acute stimulus (for example a bout or session of exercise). The present study however shows that muscle shear elastic modulus prior to an acute stimulus (i.e. marathon) was associated with exercise performance capacity, as evidenced by marathon finish time, whereas a delta change in muscle stiffness in response to the marathon run was not (Fig. 5). As a function of this, we raise awareness to the fact that absolute values of unperturbed resting shear elastic modulus (i.e. independent of marathon-elicited changes) are presumably highly influenced by the extent of prior exposures to mechanical stress. Meaning, that long term exercise training with elements of repeated (eccentric-biased) mechanical stress exposure generates a chronic adaptation towards lower basal skeletal muscle shear elastic modulus following an acute period of elevated shear elastic modulus. This concept mimics the classic biological dose–response-tolerance relationship and is consistent with data from elite athletes (Avrillon et al. 2020), with titin related changes across repeated exercise bouts (Lalande et al. 2017; Noonan et al. 2020), and recent evidence demonstrating that muscle shear elastic modulus is sensitive to the repeated bout effect of muscle damaging exercise (Chalchat et al. 2022). Importantly to the latter, Chalchat et al., (2022) showed that not only does an initial insult of downhill walking induce notable exercise-mediated elevations in rectus femoris shear elastic modulus that (a) lasts > 7 days and (b) is coupled to impaired muscular performance and neuromuscular function, but remarkably after a second bout of matched insult, the shear modulus becomes lower than pre-exercise levels in < 3 days (see data supplement of (Chalchat et al. 2022)). Moreover, additional reports (Ochi et al. 2018; Ando et al. 2021) further suggest the potential for training-induced decreases in muscle shear modulus. These data emphasize the collective need for extra attention when considering transitions between acute and continually repeated stimuli on net muscle shear elastic modulus. The collective findings of the present study spotlight that the muscle shear elastic modulus response to exercise may follow the all-too-familiar physiological phenomena whereby a specific stress elicits an apparent bi-directional response, depending on whether it is an acute insult or it is layered repetitively into a chronic adaptation response.

Summary graphic depicting the cycle of change in skeletal muscle shear elastic modulus to marathon distance running. The upper left quadrant depicts baseline difference in muscle shear modulus as function of pre-marathon exercise performance capacity, presumably from some combination of experience and exercise training. The upper right quadrant depicts how an eccentric dominant exercise insult or unaccustomed mechanical stress beyond present day tolerance (i.e. such as in a marathon race) is expected to acutely increase shear elastic modulus. The bottom right quadrant depicts how highly cushioned footwear with plates can blunt marathon-induced rises in muscle stiffness (an index of muscle damage). The bottom left quadrant depicts how muscle soreness recovery time is longer in marathoners who have greater changes in muscle stiffness the preceding running exercise bout, and that this has associative influence on the time a person can run again. Hypothetically, the cycle could continue with extended time and long-term adaptation, where muscle shear elastic modulus returns to baseline or continues to progress lower in kPa, depending on the future magnitude and frequency of insult (i.e. repeated endurance exercise training)

Although still relatively novel, ultrasound-based shear imaging can be obtained painlessly and relatively quickly (Bercoff et al. 2004), is reliable for resting muscle shear elastic modulus (Lacourpaille et al. 2012), tracks muscle activity (Nordez and Hug 2010), and can reflect early detection of exercise-induced muscle damage (Lacourpaille et al. 2014, 2017). More common protocols exploring exercise-induced muscle damage have historically focused on more potentially complex and expensive techniques (such as MRI or specialized assays) (Hooijmans et al. 2020), invasive muscle biopsy (Fridén et al. 1981), or blood markers that are susceptible to high variability and somewhat poor specificity and sensitivity (Brancaccio et al. 2007).

The observational, cross-sectional, and uncontrolled ‘field’ study design of the present study limits causal inference. Longitudinal interventions are required to resolutely conclude on specific training-induced changes in passive muscle stiffness. Neither detailed outline of runner experience nor the exercise program training load was experimentally controlled through direct monitoring in the present study, which while intricately complex in study design, could lead to deeper causal understanding on muscle shear elastic modulus changes across single and repeated bouts of exercise. While certainly of interest, further examination amongst the interactions of muscle shear modulus and post-marathon recovery time specific to footwear selection was not possible in the present study. This was largely due to rapidly diminishing participant numbers when partitioned by (1) participants who completed the recovery timeline surveys, (2) participant footwear selection, (3) recovery time decay bins, and (4) within footwear comparisons for a given day of recovery (i.e. recovery in 24 h in shoe X vs recovery in 72 h in shoe X or Y). More controlled and targeted study designs would yield further welcomed mechanistic insights.