Exercise-induced muscle damage: multi-parametric MRI quantitative assessment

Background To explore the value of magnetic resonance quantitative analysis using diffusion tensor imaging, T2 mapping, and intravoxel incoherent motion in the evaluation of eccentric exercise-induced muscle damage and to compare the effects of various eccentric exercise modes at different time points in rats. Methods A total of 174 Sprague-Dawley male rats were randomly divided into five groups: control, once-only exercise, continuous exercise, intermittent exercise, and once-fatigue exercise groups. Each experimental group was divided into seven time-subgroups: 0.5 h, 24 h, 48 h, 72 h, 96 h, 120 h and 168 h after exercise. The quadriceps femoris muscles were then scanned using magnetic resonance imaging. The apparent diffusion coefficient and fractional anisotropy values of diffusion tensor imaging, T2 values of T2 mapping, D and D* values of intravoxel incoherent motion and optical density values of desmin were measured. Associations among different eccentric exercise programmes, magnetic resonance imaging findings, and histopathological results were evaluated. Dunnett’s test, two-way repeated measures analysis of variance, and Pearson correlation analysis were used for statistical analysis. Results Diffusion tensor imaging showed that the number of muscle fibre bundles decreased to varying degrees with different time points and eccentric exercises. Apparent diffusion coefficient values of the exercise groups showed a trend that first increased and then decreased, the opposite of fractional anisotropy. The specimens in all eccentric exercise programmes showed high signal T2 values after exercise, the highest among which was in the once-fatigue exercise group. D and D* in the experimental groups were significantly higher than those in the control group at 0.5–48 h after exercise. The apparent diffusion coefficient, fractional anisotropy, T2, D and D* values correlated with the optical density values of desmin. Conclusions Diffusion tensor imaging, T2 mapping, and intravoxel incoherent motion technology accurately reflect the degree of skeletal muscle damage and recovery associated with eccentric exercise. The degree of muscle damage was the lowest in the continuous exercise group and the highest in the once-fatigue exercise group, which may provide more information and guidance for the formulation of physical and athletic training programmes.


Background
Exercise-induced muscle damage (EIMD) commonly occurs after intensive and unaccustomed centrifugal exercises, such as strenuous endurance and strength training, and is characterised by delayed-onset muscle soreness (DOMS), muscle tenderness, muscle stiffness, muscle weakening, and volume enlargement [1][2][3]. EIMD can seriously affect living, working, and performing fitness exercise in the general population, as well as in sports training and professional sports competitions [4,5]. Currently, the primary mechanisms of EIMD include mechanical injury, metabolism disorder, inflammation, and connective tissue injury theory [6][7][8][9]; the international mainstream view is that EIMD is attributed to the lack of homogeneity in the stretching of the sarcomeres, which is the mechanical damage of physical stress upon the muscle fibre [7]. This mechanical damage subsequently triggers the loss of calcium homeostasis, the production of reactive oxygen species, pro-inflammatory responses, oxidative stress and immune responses [10]. If the muscle tension remains the same while the muscle length changes, then the muscle contraction is isotonic; the muscle length can increase to produce eccentric contractions or shorten to produce concentric contractions. Eccentric exercises are more effective than concentric exercises [11] and are consequently the main mode that results in EIMD [12], the model for which was created by Armstrong [13].
Desmin is a type of muscle-specific intermediate fibrin that is mainly distributed in the cytoplasm, which is the primary component of the skeletal muscle, myocardium, and smooth muscle cells of vertebrates. It is a crucial index for research on the cytoskeleton and repair of the skeletal muscle damage, as well as an important structure to maintain the integrity of muscle cells. The loss of desmin after eccentric exercises has been proven to be a prelude to skeletal muscle damage [14,15]. Although histological techniques can be used to determine muscle damage, the invasiveness of biopsies limits their applicability.
Owing to its better soft-tissue resolution and multi parameter imaging characteristics, magnetic resonance imaging (MRI) provides important information for evaluating the extent and degree of muscle damage. However, the conventional imaging examination lacks sensitivity to early muscle lesions. Diffusion tensor imaging (DTI) is a technology to quantitatively evaluate diffusion anisotropy of water molecules from multiple directions, which can provide information on the organisational proliferation and direction of the fibre bundles [16,17]. Currently, DTI has been widely used in the field of study of the skeletal muscle and the nervous system [18,19]. T2 mapping can reflect specificity through the detection of transverse magnetisation attenuation of tissues. T2 values are obtained by measuring the MR signal intensity of different echo times. T2 mapping is primarily applied to the bone and heart; more research has been performed on the joints and cartilage [20,21], and less research has been conducted on muscles. Intravoxel incoherent movement (IVIM) is an imaging technique [22] that measures changes in blood perfusion [23]. However, limited reports exist on the application of IVIM for the diagnosis of muscle damage. Few studies have been conducted on exercise patterns and MRI evaluation methods for EIMD, and no studies have reported on MRI and clinical assessments under different eccentric exercise modes.
We hypothesised that different eccentric exercise modes would result in varying degrees of muscle damage. Hence, four eccentric exercise models based on human habitual motion patterns were applied in Sprague-Dawley rats, which is the most used animal model for studying EIMD. The aim of this study was to compare the effects of different time points, different eccentric modes, and relevance among variables using MRI quantitative parameters and histopathological expressions.

Group design
The experimental procedures and animal research were approved by the Ethics Committee of Shengjing Hospital of China Medical University (2019PS022K). A total of 174 male Sprague-Dawley rats, weighing 200 ± 20 g (specific-pathogen-free grade, aged 8 ± 0.6 weeks) were purchased from Changsheng Biotechnology Co., Ltd. (Liaoning, China). The sample sizes were determined using power analysis and the principles of 3Rs for animal research [24]. All rats were allocated individual cages with standard food and water in suitable environmental conditions (Animal Department, Benxi R & D Center, Shengjing Hospital, China). The rats were randomly divided into five groups: 1) control group (CTL, n = 6); 2) once-only exercise group (OEG, n = 42) which exercised once in 24 h; 3) continuous exercise group (CEG, n = 42) which exercised once in 24 h for 3 days (three bouts in total); 4) intermittent exercise group (IEG, n = 42) which exercised once, and then once more after a 6-day break (two bouts in total); and 5) once-fatigue exercise group (FEG, n = 42) which exercised only once with a different exercise protocol. Each experimental group was divided into seven subgroups according to the time points of the last exercise [10,25]: 0.5 h, 24 h, 48 h, 72 h, 96 h, 120 h, and 168 h (Fig. 1a).

Exercise programmes
Rats in CTL did not exercise, whereas those in the experimental groups were exercised to become familiar with treadmill running for 3 days at 09:00 am (10 min/day, 8-10 m/min). Three groups (OEG, CEG, and IEG) underwent the following eccentric exercise program: downhill treadmill running at a slope of − 16, a 5-min adaptation exercise at a speed of 8 m/min, and 5 min of running at a speed of 16 m/min with a 2-min rest, repeated 18 times. FEG performed treadmill running at a speed of 16 m/ min for 20 min with a 5-min rest, repeated four times, and then running for another 10 min. Rats in all experimental groups exercised for a total of 90 min (Fig. 1b). Rats were stimulated with a weak electrical or photic charge to initiate running.
Three-dimensional images of the quadriceps muscle were obtained via a morphological assessment: diffusion tensor tractography (DTT), and which is derived from DTI [27]. Two observers (S.N.P., with 29 years of experience in muscle MRI diagnosis; and Y.G., with 4 years of experience in muscle MRI diagnosis, including 5 months of MRI post-processing training) were trained to measure all the quantitative data twice, with a measurement interval of 4 weeks. They independently measured the whole data. The measurement of the maximum region of quadriceps femoris injury was determined using T1WI axial anatomical and T2WI coronary positions. The regions of interest (ROI) were manually plotted around the lesions; the bones and blood vessels were excluded. The postprocessing software automatically generated quantitative data and three-dimensional images of the muscle fibre within the scope of ROI. Measurements in the control group were conducted at the same location. Three measurements were conducted, and the results averaged.

Histopathological examinations
After MRI, tissue specimens from the quadriceps muscles were taken, promptly put into 4% paraformaldehyde to avoid destruction and fixed at room temperature (20 ± 5°C) for 1 week. The muscle specimens were then dehydrated, removed, and embedded in paraffin. The paraffin blocks were sectioned and dewaxed, and haematoxylin-eosin (HE) and desmin immunohistochemical staining were performed. HE staining was used to observe morphological changes and degree of damage to the skeletal muscle fibres. Image-Pro Plus 6.0 (Media Cybernetics, MD, USA) was used for optical density (OD) values of desmin. Statistical analysis SPSS 19.0 software (IBM, Chicago, IL, USA) was used for the analysis. Homogeneity and consistency tests were conducted on the data. Descriptive data were expressed as mean ± standard deviation (SD). Inter-and intra-observer reliabilities for the measured data were calculated using the intraclass correlation coefficient (ICC). Dunnett's test was performed to compare the ADC, FA, T2, D, D*, and OD values at each time point of the control versus experimental group. Two-way analysis of variance (ANOVA) was used to analyse differences of variables between the groups and different time points of the exercise mode, before which the Shapiro-Wilk test was performed for the data homogeneity test of variance. Finally, Pearson correlation analysis was performed, with OD values of desmin as the independent variable and ADC, FA, T2, D, and D* values as the dependent variables. Data were considered significant for p-value < 0.05.

Inter-and intra-observer repeatability measures
The ICCs for the parameters (ADC, FA, T2, D, and D*) were higher than 0.75, indicating substantial or excellent measurement reliability (Table 1).

ADC, FA values of DTI and three-dimensional muscle fibre bundle of DTT in different groups
Morphological assessment of DTT revealed that the injured muscle fibres were disorganised, distorted, and decreased in number compared with those in CTL (Fig. 3a,  b, c). ADC values increased at 0.5 h after exercise, and the peak time and degree differed between exercise programmes. The values then declined to varying degrees (Fig. 4), and the trend of FA values was the opposite of that of ADC (Fig. 5). The highest ADC value and the lowest FA value appeared at 48 h of FEG. The amplitude and degree of variation in the ADC and FA values were the least in CEG, and the peak values appeared at 24 h.

T2 values of T2 mapping in different groups
Compared with CTL ( Fig. 6a, b), the colour of the pseudo colour chart of quadriceps femoris changed (Fig.  6c); the signal intensity and T2 values increased at 0.5 h and 24 h in the FEG (Fig. 6d, e), OEG, and IEG peaking at 48 h (Fig. 6f) varying differing degrees (FEG > OEG > IEG, Fig. 7). At 120 h after exercise, the T2 values were close to normal in IEG, and after 168 h, the values approached normal in the OEG and those in the FEG were still slightly higher than in the controls. The trend was different in the CEG: the T2WI signal intensity and T2 values increased slightly after exercise, peaked at 24 h, and then declined (Figs. 6g, i, 7), to approach normal values at 96-120 h.    Table 2 shows the resultant values of each variable.

HE staining changes
HE staining showed that the muscular spaces were uniform without inflammatory infiltration and that the fibres were regularly arranged in CTL (Fig. 10a). The muscle fibres in the experimental groups were distorted, disorganized, and decreased in number; the extracellular spaces were enlarged, along with inflammatory infiltration. The changes were most obvious with part of the muscle fibres broken at 48 h after exercise in the FEG (Fig. 10b, c). At 120 h after exercise, the inflammatory infiltration decreased and the muscle fibres regenerated in the experimental groups (Fig. 10d). The recovery times were different in the CEG, IEG, OEG, and FEG-120 h, 120-144 h, 144 h, and more than 144 h, respectively.

Immuno-histochemical results
Desmin staining of CTL was brown and distributed around the muscle fibres in groups, with an OD value of 0.453 ± 0.012. The lowest OD value was 0.391 ± 0.011, which appeared in the FEG at 48 h. The staining was light yellow, scattered, and absent around the muscle fibres. When comparing the lowest OD values in each experimental group, the highest value appeared in the CEG (0.413 ± 0.016); desmin staining was yellowish brown, sparsely distributed, and partially absent around the muscle fibres. The OD values increased by varying degrees during 96-168 h in all the experimental groups (Fig. 11).

Discussion
The present study demonstrated that the quantitative parameters of DTI, T2 mapping, and IVIM were consistent with changes in HE staining and desmin expressions in the muscle tissues. The baseline values of our ADC, FA, and D* in our study was 1.22 ± 0.03 × 10 − 3 mm 2 /s, 0.31 ± 0.02, and 13.16 ± 1.26 × 10 − 3 mm 2 /s, respectively, which are consistent with the results of previous studies [28,29]. Baseline D value was 0.99 ± 0.11 × 10 − 3 mm 2/ s, which is slightly lower than that reported in Jungmann's study [29]. Baseline T2 value (35.22 ± 1.15 ms), as well as the maximum values in the OEG, were higher than those reported in Fu's study [30]. We considered that the differences are due to the different technical parameters of magnetic resonance and modelling. The diffusion encoding parameters (b values and diffusion gradient directions), signal-to-noise ratio, and T2 changes can impact the quantitative indices [31]. In addition, the difference of animal species, fibre type [32], injury modelling (exercise programme), feeding pattern, and adaptive training affects the degree of muscle injury, thereby affecting the maximum values of the quantitative indices. EIMD can trigger a temporary increase in microcirculation and water content within the muscle [33,34], as well as a decrease or disruption of the muscle fibres. DTI and fibre tracing technology can define the structure of muscle fibres in detail. Sinha et al. [35] found that DTI can show the direction and measure the length and cross-sectional area of muscle fibres. The structure of normal skeletal muscle fibre is orderly and complete.
The influence of the cell membrane, transmembrane concentration gradient, and free diffusion in a direction perpendicular to the long axis of the muscle fibres makes the diffusion motion of water molecules slower than that of the parallel direction. Hence, the skeletal muscle is anisotropic. When the skeletal muscle is damaged, the diffusion of water molecules inside and outside the skeletal muscle cells changes; the oedema of the interstitium affects changes in DTI parameters such as ADC and FA values, reflecting anisotropy [18,36]. Our study found that the ADC values increased at 0.5 h after eccentric exercises, peaked at 24-48 h, and then declined; the trend of FA values was the opposite of that of ADC. During repeated eccentric exercises, the diameter of the muscle fibre increased and the extracellular matrix and muscle fibre bundle were remodelled. Furthermore, destruction of the muscle fibres led to reduction of the diffusion restriction of water molecules, which increased the extracellular stroma space and the ADC value [37]. Yoon et al. [38] suggested that the FA value of skeletal muscle was related to fibre diameter, structural integrity, and density. The FA value increases when the arrangement order of the anisotropic tissue structure is regular and parallel, whereas it decreases when anisotropy of the fibres increases [39], which is consistent with our results. Fibre tracing showed that fibre shape changed in many directions after skeletal muscle injury, which can allow visual observation of changes in muscle fibre bundles. Clearly, DTI technology can objectively and accurately reflect the damage and recovery of skeletal muscle after exercise. T2 mapping relies on increasing MRI transverse relaxation time [21,40]. After muscle exercise, due to the increase of water molecules in and out of muscle cells, the relaxation time of protons and T2 values change [41]. Our study found that all eccentric exercises showed a delayed increase in the T2 values, which may relate to the delayed increase of enzymes in the muscle or increased concentration of serum myosin heavy chain fragments. The increase in T2 values is closely related to serum fsTnI level [30], cell lysis, necrosis, inflammatory reaction and myocyte regeneration in muscle; the peak T2 value is usually the period when there is more serious oedema and inflammatory reaction. Esposito et al. [42] demonstrated a statistical correlation between the injury score used to evaluate the degree of inflammatory response in the muscle damage and T2 values. Our study confirmed that the T2 values and trends are correlated with the expression of cytoskeletal protein desmin, the loss of which is a prelude to sarcomere changes triggered by eccentric exercises [15]. In addition, peaks of the T2 values, were the lowest in the CEG and the highest in the FEG, respectively, consistent with the changes in HE staining and DTI in the muscle tissue. This means that the quantification of T2 values within a certain range can help in objectively measuring the relaxation time of T2 muscle tissue, exclusive of the influence of subjective factors. To some extent, T2 mapping can reflect the expression of inflammatory response in muscles and indirectly indicate the severity of muscle damage.
IVIM is equivalent to diffusion weighted imaging of multiple b values. The parameters D and D* values represent the response relative to the real state of organization and diffusion of water molecules and blood capillary net of microcirculation perfusion-related situation [29,32], respectively. The current study showed the D and D* values of the four experimental groups had a delayed increase after eccentric exercise, and the highest values were found in CEG. Morvan et al. [43] believed increase in the D and D* values after exercise were due to temperature. However, we consider the increase to be due to interstitial oedema and increased microcirculation perfusion. The peak times of perfusion parameters in the experimental groups were different, as it takes a different time for the muscles to adapt to local requirements [44]. Moreover, along with muscle contractions, the blood vessel will be compressed and will take longer to fill. The present study considered the reason for higher D and D* values in the CEG and IEG to be that repeated centrifugal exercises make the muscles adapt and promote microvascular circulation. In contrast, muscle injuries in the FEG were serious, which affected the recovery time of blood vessel filling; however, it showed less statistical difference in the D* values in the late stage of EIMD, which indicated that microcirculation perfusion was not obvious. Above all, IVIM can reflect the blood perfusion and diffusion function in EIMD.
The trends in skeletal muscle damage were similar but with varying degrees in the four groups, showing an immediate onset, reaching a peak at 24-48 h, and alleviating or recovering at 120-168 h. Comparing the four eccentric exercise models, the FEG had the highest peak damage degree with long duration and recovery time, whereas the CEG showed earlier damage peaking time and lowest damage degree. Overall damage in the IEG was slightly better than that in the OEG. However, the CEG having a lower degree of damage is in disagreement with some previous studies [5]. We believe this is related to differences in adaptive training, total exercise time, rest time, and animal modelling. The reason for damage changes in the IEG and CEG at each time point being less than that in the OEG is the "repeated bout effect" (RBE). This means that following a second bout of exercise, the muscle recovery is faster than in the initial bout [45]. Pincheira at al [44]. reported that the RBE causes adaptations in the non-contractile elements of muscles. Lynn et al. [46] suggested that repeated exercise leads to remodelling of the skeletal muscle sarcomeres. Along with injury repair, the muscles enhance the ability to resist the damage, and the regenerated muscle fibres strengthen the ability to resist the tension, which may even help to protect the spinal cord [47]. Consistent with Ziaaldini [48], H Zhao, a member of our team, reported increased the mRNA expressions of superoxide dismutase isoenzyme in the continuous eccentric programmes [25], which may associate with enhancement of muscle oxidative adaptation. Moreover, our study also indicates that high-intensity fatigue training results in more obvious injuries and an extension of recovery time. One important reason for the increasing incidence of sports-related injuries is the lack of standardised healthy exercise and training modes [49]. We considered that eccentric exercise can improve muscle function and neural adaptation to some extent [10,50,51]. Especially, RBE in the CEG and IEG demonstrated that properly repeating the eccentric session is a useful recovery strategy in EIMD. Our research technically verified that continuous exercise without fatigue can strengthen the muscles, promote health, and avoid damage, providing a more reasonable exercise mode as a potential means of muscle injury prevention. The limitations of this study were as follows. First, as this was an animal experiment, clinical human studies should be conducted before extrapolating these results to application. Second, we did not study the threshold of the relevant parameters; whether the imaging markers can make a quantitative diagnosis of the grading for EIMD will be the core of further studies. Another limitation was that the relatively long imaging time of the functional MRI sequences may limit its clinical application.

Conclusions
In conclusion, the findings confirm that quantitative MRI accurately reflects the histopathological abnormalities, as well as effectively responds to the degree of skeletal muscle damage induced by eccentric exercise. DTI, T2 mapping, and IVIM techniques can be applied to analyse the EIMD through water molecule diffusion, anisotropy, and muscle inflammation and microcirculation perfusion, respectively. The degree of damage was the least in the CEG, which may provide valuable information for the formulation of physical and athletic training programmes.