Sodium and quantitative hydrogen parameter changes in muscle tissue after eccentric exercise and in delayed-onset muscle soreness assessed with magnetic resonance imaging

1 INTRODUCTION

Muscle injuries and overexertion disorders account for 10%–55% of all sporting injuries.^1-3 According to the Munich Consensus Statement, delayed-onset muscle soreness (DOMS)—an entity of ultrastructural muscle injury—is classified as an overexertion-functional muscle disorder type 1b.⁴ It is caused by unaccustomed exercises of high-force eccentric muscle contractions.⁵ Through biopsies, it has been shown that eccentric training causes Z-band streaming and broadening, destroys sarcomeres in the myofibrils,⁶ and leads to inflammation and myofiber necrosis.⁷ DOMS is one of the most common reasons for reduced muscle performance in sports. It is associated with muscle soreness, as well as reduced muscle strength, range of motion (ROM), and neuromuscular coordination, and is frequently observed in both professional and recreational athletes.^8-10 Signs and symptoms begin 6–12 h after exercise and increase progressively until they reach a peak between 48 and 72 h.¹¹ Primarily eccentric muscle actions often result in perforations in the sarcolemma and damage to sarcomeres when loading exceeds the limits that the muscle is accustomed to. Thus, a resulting increase in membrane permeability is observed, which allows intracellular molecules such as creatine kinase (CK) to leak into the interstitial fluid.¹² An injury of the sarcomeres leads to a local inflammatory response, which is accompanied by edema. The breakdown products of injured tissues sensitize nociceptors.¹³ A better understanding of the underlying pathophysiological effects is desirable to optimize training and reduce associated injuries.

Magnetic resonance imaging (MRI) has been used to describe changes in muscle tissue after exercise.^14-19 In particular, T₂-weighted MR images and mapping of T₂ relaxation time have frequently been utilized to evaluate muscle damage resulting from repeated eccentric exercise.^17-19 Repetitive eccentric muscle contraction often results in damage-related muscle edema.²⁰ Fluid accumulation results in prolonged T₂ relaxation times of exercised muscles and increases signal intensity in T₂-weighted images.²¹

Besides fluid accumulation, the sodium (Na⁺) concentration increases in muscle tissue after exercise.²² The predominant cause of this is damage to the sarcolemma,²³ which leads to a Na⁺ inward current and extracellular edema.¹⁹ The extracellular Na⁺ concentration is approximately 10 times higher than the intracellular Na⁺ concentration. Na⁺/K⁺-ATPase is responsible for this gradient over the membrane and pumps Na⁺ ions from the intracellular to the extracellular space and K⁺ ions in the reverse direction at a ratio of 3:2. Muscle contraction-initiating action potentials cause a sudden influx of Na⁺ ions and an efflux of K⁺ ions. During exercise, these ion shifts degrade the transmembrane gradients, leading to increasing loss of membrane excitability and muscle contractility, which are believed to be the main causes of muscle fatigue.²⁴

Recent studies have shown the potential of sodium (²³Na) MRI for the quantification of total sodium concentration in human skeletal muscle.^{14, 19, 25-29 23}Na MRI is particularly challenging because the in vivo signal-to-noise ratio (SNR) is roughly four orders of magnitude smaller than the SNR of standard proton (hydrogen [¹H]) MRI.³⁰ Additionally, quantitative ²³Na MRI measurements benefit from pulse sequences with ultrashort echo times (UTE)³¹ because of the short T₂ relaxation times of ²³Na in (muscular) tissue.³² Meanwhile, it is feasible to provide a partial suppression of signal from sodium ions in the extracellular compartment (e.g., edema) by using an inversion recovery (IR) ²³Na MRI sequence.^{27, 33, 34} Furthermore, investigations of muscular channelopathies have indicated that IR ²³Na MRI can provide an intracellular-weighted sodium signal (ICwS).²⁷ In this work, the tissue sodium concentration (TSC) and the ICwS MRI signal were measured before exercise, after exercise, and during DOMS. Additionally, we characterized water T₁, water T₂ relaxation times, and fat fraction (FF). By assessing these parameters, we attempted to further understand the physiological processes of muscle damage.

2 MATERIALS AND METHODS

2.9 Image analysis

2.9.1 Anatomical images

Image analysis was performed by an experienced radiologist with a professional image processing program (syngo.via VB10; Siemens Healthineers, Erlangen, Germany). The compartments of the lower leg were differentiated by defining specific regions of interest (ROIs) in the T₁-weighted images according to the anatomic margin of each muscle (medial gastrocnemius muscle [MGM], soleus muscle [SM], and tibialis anterior muscle [TA]) (Figure 1) and copying those over the T₂-TIRM and T₂-mapping sequence images. These muscles were selected because they exhibit sufficient volume for the analysis of the low-resolution ²³Na MR images. The largest changes were expected in the MGM and SM, because they are heavily recruited during the performed eccentric exercise. By contrast, hardly any load is placed on the TA during the entire exercise intervention. Thus, the TA can serve as an internal control. Analysis of the axial MR images was performed at the maximum circumference of the proximal lower leg, which was marked with permanent markers and specific MRI markers during the entire trial to obtain identical slice positions. To analyze images of the sodium MRI, ROIs were delineated in the same muscles as in the ¹H MRI protocol on anatomical reference images and interpolated in the sodium maps.⁴³

Using a free software tool (www.itksnap.org), ROIs were manually drawn on the multislice multiecho images of the MGM, SM, and TA. The ROIs were limited to the interior of the muscle, thereby avoiding fasciae and large blood vessels.⁴⁴

2.9.2 Water T₁ quantification

From the MRF T₁-FF acquisition, a nonuniform fast Fourier transform was applied to reconstruct a series of 175 highly undersampled images. Then FF, water T₁, B1, and off-resonance values were obtained by applying matching pursuit to the individual pixel signal evolution with a dictionary of fingerprints simulated by the multicomponent time propagation of Bloch equations.⁴¹

2.9.3 Sodium quantification

As described in previous research,⁴³ all the images were acquired with tube phantoms of the same calibration (four cylinders filled with 20 and 40 mmol/l NaCl with and without agarose gel), which were placed in a phantom holder next to the calf muscles and included in the field of view. Magnitude images were reconstructed using a Hamming filter to reduce Gibbs' ringing artifacts.^{45, 46} Sodium concentrations were then calculated for the muscles using linear regression in MATLAB. Therefore, we computed the average signal intensity of two calibration phantoms (20 and 40 mmol/l NaCl in 5% agarose) and an ROI drawn in the background signal that was used as a reference for 0 mmol/l sodium concentration. The resulting linear regression curve was set to explore the sodium maps of the entire leg. TSC values were evaluated on the central slice of this map in three manually segmented muscles: the MGM, SM, and TA.²⁸

2.9.4 Volumetry

To calculate the volume of the MGM, ROIs were manually drawn below the fascia of the MGM in each slice, starting from five slices above to five slices below the isocenter in the coronal T₂ TIRM using free software (www.slicer.org). This software then calculated the volume by multiplying the drawn areas by the height.

3 RESULTS

DOMS was successfully induced in all 14 participants. They all performed five sets of eccentric calf raises as previously described (repetitions of the last set: 178 ± 68; range: 100–351). ROM decreased significantly after 48 h compared with baseline (p ≤ 0.01). Participants experienced a significant increase in soreness after 48 h (p ≤ 0.01) when compared with baseline (Table 2). Although measured CK levels increased from a physiological range at baseline, the changes were not statistically significant (Table 2). At follow-up, the CK levels of three participants increased to above 2000 U/l, which implies a more than 10-fold increase from the initial values (denoted as “high CK”). Eleven participants showed CK levels below 2000 U/l (denoted as “low CK”). Figure 2 presents the TSC, ICwS, water T₁, and water T₂ maps of a “low CK” and a “high CK” participant. The postintervention MRI investigation revealed intramuscular edema in the MGM for all participants (i.e., hyperintensities in T₂-weighted imaging).

The TSC and ICwS of exercised muscles (MGM, SM) increased significantly from t0 to t1 and decreased significantly from t1 to t2 (Figures 3 and 4; Table 3). The TSC and ICwS in the TA remained at baseline levels at each measurement point (Figure 5). In the MGM of low-CK participants, maximum TSC and ICwS values were measured at t1 and then dropped below baseline levels at t2 in all but two low-CK participants (Figure 6). In high-CK participants, maximum TSC values were measured at t2 and at least doubled in comparison with baseline values. ICwS in MGM decreased below baseline levels at t2 in all participants except for two high-CK participants (Figure 6). In the SM, the highest TSC values were measured at t1 and then returned to baseline levels at t2. ICwS levels in the SM dropped below initial values at t2 in all participants (Figure 6). In contrast to the sodium parameters, no significant changes in T₁ or T₂ relaxation times were observed in any muscles from baseline to t1 (Figures 3, 4, and 5; Table 4). The T₁ relaxation times of the MGM and SM increased significantly from t0 to t2 (Table 4). Moreover, T₂ relaxation times increased significantly in the MGM from t0 to t2 (Figures 3 and 4; Table 4). Areas with high T₁ and T₂ relaxation times also show a high TSC but not ICwS (Figure 2). Relaxation times in TA remained at baseline levels. Moreover, MGM volume significantly increased 48 h after exercise. In high-CK participants, MGM volume increased more than in low-CK participants at t2 (Figure 2). Measured fat fractions of examined muscle groups remained stable at all time points.

In the MGM, the relative increases of TSC and ICwS at t2 are highly correlated (r = 0.88, p < 0.01). The relative increase of ICwS in the MGM at t1 correlates with an increase at t2 (r = −0.626, p = 0.017). Relative increases of TSC in the MGM and SM correlate at t1 (r = 0.566, p = 0.035) and t2 (r = 0.766, p ≤ 0.01). At t1, relative increases of ICwS in the SM correlate with a relative increase of TSC in the MGM (r = 0.793, p ≤ 0.01). At t2, the relative increase of TSC in the SM correlates with the relative increase of ICwS in the MGM (r = 0.759, p ≤ 0.01). After 48 h, the relative increase from baseline in T₁ correlates highly with the relative increase in T₂ (r = 0.981, p ≤ 0.01). Additionally, at t2, the relative increase in muscle volume correlates with the relative increase of T₁ (r = 0.932, p ≤ 0.01) and the relative increase of T₂ relaxation time (r = 0.921, p ≤ 0.01) (Figures 7 and 8).

4 DISCUSSION

In this study, sodium and quantitative ¹H MRI parameter changes in muscular tissue after exhausting eccentric sports and in DOMS were assessed. After exercise, TSC increased significantly in the affected muscle groups. Compared with postexercise values, TSC and ICwS values decreased significantly after 48 h. Moreover, after 48 h, ICwS was even significantly below baseline levels in the SM.

Previous studies that investigated DOMS mainly applied T₁- and T₂-weighted sequences and focused on ¹H MRI.^{47, 48} To date, sodium MRI has so far only been applied directly after exercise but not in DOMS.^{22, 25} Studies assessing changes in sodium concentrations after muscle contraction have used sodium MRI to investigate short-term changes in TSC after repetitive plantarflexion¹⁴ or after participants hopped on one leg.¹⁹ In this study, the muscle load was much higher, with changes in TSC and ICwS, as well as T₁ and T₂ relaxation times in muscle tissue after eccentric exercise, and with DOMS also being investigated. Also, longitudinal relaxation times have not yet been studied in muscles after contraction or in DOMS. By applying our multiparametric MRI protocol—including water T₂, water T₁, FF, TSC, and ICwS—we were able to characterize different processes observed in the muscle directly after eccentric exercise and with DOMS.

In affected muscle groups, TSC increased significantly at t1. These results confirm previous studies.^{14, 19, 22, 25, 49} Compared with postexercise values, TSC and ICwS values significantly decreased after 48 h (Figures 2, 3, and 5). ICwS was significantly below baseline levels in the SM after 48 h (Figures 3 and 5). Hammon et al. observed a decline in sodium concentration to at least baseline in all but one exercised muscle after 50 min of recovery.²² Furthermore, Bansal et al. reported the return of sodium signal to baseline with a half-life of approximately 0.5 h.¹⁴ Additionally, Chang et al. calculated a half-life of 22 min after measuring an increased signal intensity following exercise.¹⁹ Also, Sjørgaard et al. noted an increase from 6 to 24 mM of intracellular sodium at the end of intense exercise through biopsies, while extracellular sodium increased from 133 to 135 mM. After 3 min of recovery, intracellular and extracellular sodium levels returned to 14 and 129 mM, respectively.⁴⁹ Because of the short half-life of the sodium signal increase, the period of time that passed between the end of the exercise and data acquisition must be considered. TSC acquisition began about less than 5 min after the end of the exercise, while approximately 15 min passed until ICwS was measured. Thus, both TSC and, in particular, ICwS concentrations might have originally been much higher.

In this study, three participants stood out with a greater than 30-fold increase in CK and a markedly higher ascent than the other participants. Large interindividual variation in response to eccentric exercise is commonly reported.⁷ This may be because of motivation, posture, training status, or genetic predisposition. TSC values in MGM for low-CK participants were highest at t1, while the values of high-CK participants were highest at t2 and at least doubled compared with baseline.

T₁ relaxation times increased significantly after 48 h in the MGM and SM (12.6% and 3.6%, respectively). T₂ relaxation times and volume increased in the MGM at t2 (36% and 15.4%, respectively). T₁ and T₂ values of TA remained at baseline levels (0.6% and 1.64%, respectively). These results agree with those of other studies.^{14, 19, 22, 25, 40, 44, 49, 50} Moreover, this is expected because barely any load is put on the TA during the performed exercise. The nonuniformity among triceps surae muscles is likely attributable to variation in muscle strain and thus muscle damage.²¹ Notably, the MGM is the most heavily recruited muscle in calf raises.⁵¹ To meet important metabolic demands during exercise, the intravascular volume expands, which is associated with a short-term increase in both T₁ and T₂ relaxation times.⁴⁴ However, studies of T₂ changes in exercising muscles showed that the predominant contribution was intracellular and resulted from an accumulation of the end products of anaerobic metabolism within muscle cells, which causes a decrease in pH and an increase in intracellular volume.^{52, 53} Later, the infiltration of inflammatory cells is accompanied by fluid effusion. Cytokines in damaged tissue increase vascular permeability.⁵⁴ Thus, the elevation of osmotic pressure and/or increased permeability of blood vessels may also be associated with the formation of edema, especially soon after exercise.²¹

Compared with the sodium MRI parameters, the water relaxation times showed different time courses. For instance, in the MGM, sodium parameters were highest at t1, on average, while water relaxation times were highest at t2. The cause of these discrepancies may be that different mechanisms are at play in the re-establishment of the equilibrium conditions for water and sodium ion concentrations. Clearance from metabolites that caused the osmotic shift of water from the extracellular to the intracellular space might be faster than the clearance of sodium ions, which depends on Na⁺/K⁺ pump activity as well as the ability of blood to remove extruded sodium ions from the extracellular space, which may require more time, because Na⁺-K⁺ pumps are depleted after intense exercise.^{19, 22} T₂ values in the MGM only increased by 4% at t1, whereas TSC and ICwS increased by 39% and 35%, respectively. One explanation for the observed increased ICwS and TSC values despite relatively normal water T₂ and T₁ values might be that sodium MRI is more sensitive to detecting osmotic shifts in muscle tissue. Sodium MRI is highly sensitive to shifts in the relative sizes of the intracellular and extracellular compartments because these well-defined tissue compartments have widely differing sodium concentrations.^{14, 55} Because there were no changes in the ICwS/TSC ratio at t1, this observation could be compatible with two concurrent phenomena: an increase in intracellular sodium (e.g., sodium inward current because of a ruptured sarcolemma¹²) and an increase in the extracellular volume fraction (e.g., because of increased perfusion). A significant change in the ICwS/TSC ratio at t2 (0.27) when compared with baseline (0.38) might be caused by a decrease of intracellular Na⁺ concentrations below baseline levels. Previous research demonstrated that the ICwS sequence showed a weighting towards the intracellular Na⁺.³³ An explanation for the different time courses of ¹H and ²³Na MRI parameters is that early ¹H changes could not be detected in this study because of the protocol design. Notably, the ¹H protocol started approximately 25–30 min postexercise. In a study conducted by Chang et al., the ¹H T₂ values of subjects who hopped on one leg until fatigue first increased and then again returned to baseline values 35 min after the end of the exercise.¹⁹ Thus, potential early changes were not detected in our study and might have been much higher. As such, further studies are needed to assess this process.

Different sodium and potassium concentrations in the intracellular and extracellular space cause an electrochemical gradient along the sarcolemma that is maintained by Na⁺-K⁺-ATPase. At high stimulation frequencies, the K⁺ efflux can exceed the capacity of the ion pumps,⁵⁶ while sodium accumulates intracellularly.^{22, 50} Thus, fatiguing contractions cause ion perturbations.⁵⁰ To re-establish the membrane potential, Na⁺/K⁺-ATPase increases its capacity, which might explain the observed sodium signal decrease at t2. For low-CK participants, TSC and ICwS values decreased below baseline levels in the MGM and SM (see Figure 6). Causal for these decreases might be a decrease in the intracellular Na⁺ concentration. Additionally to ion perturbations, eccentric muscle actions often result in perforations of the sarcolemma and a subsequent increase in membrane permeability,¹² which enhances the ions streaming along their electrochemical gradient.⁵⁷ Thus, even more sodium ions can stream into the muscle cell and K⁺ can diffuse into the extracellular compartment unhindered.⁴⁹ These shifts lead to depolarization and result in a loss of excitability, and are also thought to be a cause of muscle fatigue.²⁴ Our results are in accordance with these physiological processes.

Subject #2 stood out with a remarkably high TSC of 65.7 mM and ICwS of 13.0 mM after 48 h (Figure 6). The MGM volume of this subject increased by 64% at t2. To estimate if the observed increase of TSC is compatible with extracellular edema, we assumed a simple two-compartment model consisting of extracellular and intracellular space, with volume fractions of 10% and 90% at baseline, respectively. A pure extracellular increase of the volume and physiological Na⁺ concentrations for the extracellular (140 mmol/l) and intracellular space (10 mmol/l) would result in a TSC of 68.7 mmol/l, which is close to the measured result.

5 LIMITATIONS

The interpretation of these findings is limited to the small sample size of this study. It must be emphasized that it is not currently possible to measure intracellular and extracellular ²³Na separately with MRI. The IR preparation enables the weighting of measurement towards intracellular ²³Na based on the assumption of different relaxation properties; however, significant contributions from the extracellular pool cannot be excluded. Also, decreasing ICwS should be further assessed alongside volume changes in the intracellular and extracellular compartments to further understand physiological and pathological processes.

Directly after exercise, sodium MRI measurements were performed before water relaxation times were assessed. The postexercise state of sodium concentrations is highly dynamic and short half-lives have been reported.^{14, 19} First, TSC was acquired. Then, approximately 15 min passed from the end of the exercise until the ICwS measurement began. Thus, ICwS likely differed substantially from the actual values directly after exercise. Moreover, approximately 25 min passed from the end of the exercise until the ¹H protocol began. Chang et al. measured ¹H half-lives of 12–15 min in subjects after they hopped on one leg until reaching fatigue.¹⁹ These early changes would not have been detected with the applied study protocol because quantitative ¹H MRI parameters may have already normalized and may have also originally been higher. Thus, the results at t1 must be interpreted with caution given the long acquisition times and highly dynamic postexercise state. In future studies, the time gap between the acquisition of ²³Na and ¹H MRI data might be reduced by applying techniques that enable interleaved⁵⁸ or simultaneous^{59, 60} acquisition of ²³Na and ¹H MRI data.

6 CONCLUSION

In conclusion, ²³Na MRI depicts increased Na⁺ in muscles after eccentric exercise. Sodium MRI parameters and water relaxation times peaked at different points in time. On average, TSC and ICwS in exercised muscles were highest at t1, while water relaxation times were highest at t2 when participants experienced DOMS. ICwS levels in the SM dropped below initial values at t2 in all participants. Our results suggest that ²³Na MRI could be used as a sensitive marker to characterize physiological processes during DOMS. Its use should lead to a better comprehension of pathophysiology, enhancing diagnostics and potentially improving therapeutic strategies for muscle injuries.