Journal of Applied Physiology

Exercise training increases oxidative capacity and attenuates exercise-induced ultrastructural damage in skeletal muscle of aged horses

Jeong-su Kim, Kenneth W. Hinchcliff, Mamoru Yamaguchi, Laurie A. Beard, Chad D. Markert, Steven T. Devor


Exercise training improves functional capacity in aged individuals. Whether such training reduces the severity of exercise-induced muscle damage is unknown. The purpose of the present study was to determine the effect of 10 wk of treadmill exercise training on skeletal muscle oxidative capacity and exercise-induced ultrastructural damage in six aged female Quarter horses (>23 yr of age). The magnitude of ultrastructural muscle damage induced by an incremental exercise test before and after training was determined by electron microscopic examination of samples of triceps, semimembranosus, and masseter (control) muscles. Maximal aerobic capacity increased 22% after 10 wk of exercise training. The percentage of type IIa myosin heavy chain increased in semimembranosus muscle, whereas the percentage of type IIx myosin heavy chain decreased in triceps muscle. After training, triceps muscle showed significant increases in activities of both citrate synthase and 3-hydroxyacyl-CoA-dehydrogenase. Attenuation of exercise-induced ultrastructural muscle damage occurred in the semimembranosus muscle at both the same absolute and the same relative workloads after the 10-wk conditioning period. We conclude that aged horses adapt readily to intense aerobic exercise training with improvements in endurance, whole body aerobic capacity, and muscle oxidative capacity, and heightened resistance to exercise-induced ultrastructural muscle cell damage. However, adaptations may be muscle-group specific.

  • citrate synthase
  • 3-hydroxyacyl-CoA-dehydrogenase
  • exercise-induced ultrastructural muscle damage
  • electron microscope
  • myosin heavy chain

the ability of mammalian skeletal muscle to produce and sustain power output decreases during the aging process (37). Decreases in muscle mass and the number of motor units are partially responsible for this age-related reduction in the functional capacity of muscle (21). Because physical inactivity is also partly responsible for age-related changes in skeletal muscle, exercise training induces positive adaptations and has been shown to attenuate the age-related reduction in functional capacity of the muscle cell (10, 11, 15). Specifically, endurance exercise training increases activities of aerobic enzymes and partially attenuates the decline in maximal aerobic capacity (V̇o2 max) with aging (19, 20).

In concert with the metabolic adaptations that occur during the aging process, fibers of older rats are markedly more prone to contraction-induced injury (6). The damage-induced muscle fiber degeneration and regeneration cascade, which typically follows vigorous exercise training, may enhance resistance to contraction-induced cellular damage in skeletal muscle of older individuals. Therefore, exercise training may assist older individuals in preventing the abrupt injury associated with age-related impairments, such as a reduction in muscle force-generating capacity, which leads to impaired mobility, and tendency to fall (10). Although potential mechanisms are not clear, a progressive exercise-training program will induce a heightened recruitment of type II muscle fibers by increasing mechanical load, which may slow an age-related transformation from type II to type I fibers by diminishing motor unit remodeling. Thus exercise training may attenuate an age-related reduction in muscle contractile function. Consequently, even though there are inherent age-associated alterations in skeletal muscle, we hypothesized that relatively intense aerobic conditioning of aged horses would diminish the severity of the exercise-induced muscle damage after acute exercise and that this reduction in damage would be associated with conditioning-induced increases in activity of metabolic enzymes associated with muscle oxidative capacity.

The purpose of the present investigation was to evaluate the effects of 10 wk of treadmill exercise training on the 1) magnitude of ultrastructural skeletal muscle damage occurring in response to incremental treadmill exercise tests performed before and after aerobic conditioning, 2) expression of skeletal muscle myosin heavy chain (MHC) isoforms, and 3) activities of skeletal muscle enzymes indicative of oxidative and glycolytic metabolism in aged horses. In the present investigation, exhaustive incremental exercise tests (V̇o2 max test) played a role in creating exercise-induced ultrastructural muscle damage and were used for the determination of exercise training intensity for each horse.


Experimental Design

The present study was longitudinal in design. To assure a similar level of fitness at the beginning of the study, horses were deconditioned by 12 wk of stall rest before being conditioned by 10 wk of regular exercise on a treadmill. After the deconditioning period, all horses performed an incremental treadmill exercise test. Two subsequent incremental exercise tests were performed during the conditioning phase of the study at weeks 8 and 10. Muscle samples were collected from locomotor and nonlocomotor muscles before and after all three of the exercise tests, and tissue samples were examined for evidence of exercise-induced damage and changes in oxidative capacity, as well as training-induced shifts in myosin isoform expression.


Six old, healthy, untrained Quarter horse mares, from 23 to 30 yr of age, were utilized. The study was approved by the Institutional Laboratory Animal Care and Use Committee at The Ohio State University. The overnight dexamethasone suppression test, a highly sensitive test for equine Cushing's disease (7), was normal in all mares. Body weight was 487 ± 24 kg. Horses were housed in box stalls and fed grass hay and mixed grain sufficient to meet the National Research Council recommendations for the nutrient requirements of horses and to maintain an ideal body condition score (5–6 on a scale of 9).

Experimental Procedures

During the first month of the 12-wk deconditioning period, horses undertook one or two treadmill familiarization sessions per week. During the familiarization sessions, horses wore a respiratory gas-collection mask and walked or trotted on the treadmill for <5 min/session. At the conclusion of the deconditioning period, muscle biopsy samples (Pre-1) were taken from triceps, semimembranosus, and masseter muscles using an open-biopsy technique. The Pre-1 muscle biopsy samples were collected 3 days before the horses underwent the first incremental exercise test. The second set of muscle biopsy samples (Post-1) were collected immediately (<15 min) after the first incremental exercise test. After the first exercise test, all of the horses began a 10-wk exercise training program consisting of treadmill exercise 3 days/wk for 10 wk (see Exercise Training Protocols below for details) (8).

After 8 wk of aerobic conditioning, all horses performed a second incremental exercise test that exactly duplicated the first test in final speed attained, duration of each stage, and total time spent on the treadmill. Therefore, the second treadmill exercise test required horses to perform the same absolute workload (speed, treadmill incline, and duration) as the first exercise test that was conducted before the initiation of conditioning. Muscle samples were taken in the same manner (i.e., pre- and postexercise samples) from the same muscle groups. Biopsy samples collected after 8 wk of conditioning were labeled Pre-2 (i.e., collected before the same absolute treadmill exercise test) and Post-2 (i.e., collected after the same absolute treadmill exercise test).

After another 2 wk of aerobic conditioning (i.e., the end of 10-wk aerobic conditioning period), collection of muscle samples and the incremental exercise tests were repeated. The additional 2 wk of training allowed the horses to recover from potential exercise-induced ultrastructural damage resulting from the second exercise test. The final incremental treadmill exercise test was not performed at the same absolute workload; rather, it was another exhaustive maximal exercise test that permitted determination of the effect of training on V̇o2 max. The muscle samples collected before and after the final exercise test were labeled Pre-3 and Post-3, respectively, and permitted determination of training-induced skeletal muscle adaptation for ultrastructural damage to the same relative stress after aerobic conditioning.

For all three incremental tests, horses were positioned on the treadmill for measurement of oxygen (V̇o2) and carbon dioxide uptake (V̇co2) and calculation of respiratory exchange ratio using an open-circuit calorimeter (8). Respiratory gas exchange was measured throughout exercise.

As illustrated in Fig. 1, Pre-1 biopsy specimens were analyzed to determine the characteristics of skeletal muscle of old sedentary horses and served as baseline data for the training adaptation. Using light and electron microscopy, Post-1 muscle samples were compared with Pre-1 to observe the magnitude of ultrastructural damage after one bout of maximal exercise (incremental exercise test) on a treadmill, and the same comparisons were made between Pre-2 and Post-2 and between Pre-3 and Post-3 muscle samples.

Fig. 1.

Experimental protocol. Same absolute, an incremental exercise (Ex.) test at the same absolute workloads; Same relative, an incremental exercise test at the same relative workloads; Pre-1, muscle sampling before the 1st incremental exercise test; Post-1, muscle sampling immediately after the 1st incremental exercise test; Pre-2, muscle sampling before the 2nd incremental exercise test; Post-2, muscle sampling immediately after the 2nd incremental exercise test; Pre-3, muscle sampling before the 3rd incremental exercise test; Post-3, muscle sampling immediately after the 3rd incremental exercise test; V̇o2 max, maximal aerobic consumption.

The differences in the degree of ultrastructural damage between Pre-1 and Post-1 muscle samples were compared with differences between Pre-2 and Post-2 to determine the effect of training on ultrastructural damage responses to the same absolute workload. Additionally, Pre-1 and Post-1 samples were compared with differences between Pre-3 and Post-3 muscle samples to evaluate the extent of ultrastructural damage after the same relative workload after the conditioning. Last, Pre-1 and Pre-3 muscle samples were analyzed and compared to evaluate the effects of exercise training on MHC isoform expression as well as oxidative and glycolytic capacities of exercise trained muscles in association with a change in V̇o2 max after training.

Exercise Training Protocols

The 10-wk exercise training program was accomplished with an interval exercise training protocol performed on the treadmill (8). The exercise training intensity for the first training protocol was determined based on the results of the first incremental exercise (V̇o2 max) test for each horse, and the training protocol was updated every 1–2 wk based on performance and response to given workloads. Exercise consisted of walking (2 m/s) for 2 min on a treadmill with a 4° incline. Treadmill speed was increased to a speed calculated to require 40% V̇o2 max and maintained for 2 min. Then, treadmill speed was increased rapidly and maintained at a speed calculated to require 60% V̇o2 max for 5 min, at which point treadmill speed was increased to a speed calculated to require 80% V̇o2 max for 5 min. The horses then ran at 60% V̇o2 max for 5 min before walking at 2 m/s for 1 min. Total exercise time was 20–22 min, with 15 min being spent at a relatively high intensity. When not training, horses were confined to box stalls at all times.

Muscle Biopsy Sampling Procedure

Muscle samples were taken from the triceps brachii, semimembranosus, and masseter (control) muscles by open biopsy. All pre- and postexercise muscle samples were collected from the midbelly region of the right and left muscle group, respectively. Horses were sedated by intravenous injection of detomidine (40 μg/kg), and the area for biopsy was desensitized by injection of mepivacaine (2% solution, 10 ml) and aseptically prepared. A 3-cm skin incision was made, and all biopsy samples were collected from ∼7 ± 2 mm depth under the skin. Approximate biopsy sample size was ∼500 mg total. Immediately after muscle sample collection, the inner portion of each sample was trimmed to avoid subsequent laboratory examination of muscle fibers injured during the biopsy process. Samples were dissected longitudinally in approximate sizes of 2-mm length and 1-mm thickness. The isolated tissue samples were placed in 5% phosphate-buffered glutaraldehyde fixative solution (5% glutaraldehyde, 0.05 M sodium phosphate), pH 7.4, overnight at 4°C for electron microscopic (EM) analyses. The remaining portions of the muscle samples were immediately frozen in liquid nitrogen and stored at −80°C for subsequent biochemical analyses.

Blood Sampling Procedure

Catheters for blood collection were inserted into the right jugular vein after desensitization of overlying skin. For lactate analysis, blood samples were obtained immediately before, during the last 10 s of each stage of the incremental exercise test, and immediately after (within 2 min) the conclusion of each incremental test. Blood samples for creatine kinase (CK) analysis were taken before, within 5 min of the end, and 4 and 24 h after each incremental test. Plasma and serum were harvested within 30 min of blood collections for determination of plasma lactate concentration and serum CK activity, respectively. All collected blood samples were kept on ice during the process.

Measurement of V̇o2 max (Indirect Calorimetry)

For the incremental exercise test, each horse ran on a treadmill (Sato, BIAB Industrial, Uppsala, Sweden) inclined at 4° for 90 s at 3 m/s, and the treadmill speed was increased by 1 m/s every 90 s until the horse was unable to maintain its position at the front of the treadmill. V̇o2 was measured at 10-s intervals during the exercise test with an open-circuit calorimeter (Oxymax-XL, Columbus Instruments, Columbus, OH) as previously described (8). V̇o2 max was defined as the point at which V̇o2 reached a plateau despite further increases in speed. A plateau was defined as a change in V̇o2 of <4 ml·kg−1·min−1 with an increase in speed.

Histological Analyses

Muscle fixation and analysis.

Damage to muscle fibers was evaluated by light microscopic (LM) and EM examination of longitudinal sections of muscle fibers, as described by Yamaguchi et al. (38). For EM analysis, after the first overnight fixation in 5% phosphate-buffered glutaraldehyde fixative solution, muscle sections were washed in 0.05 M sodium phosphate buffer at least three times and postfixed in a buffered solution of 1% osmium tetroxide for 24 h. After the second fixation, muscle samples were rewashed in phosphate buffer (3×) and dehydrated through a graded ethanol series and acetone washes. All histological muscle preparation procedures were conducted using cold liquids.

After fixation, each specimen was embedded in epoxy resin and polymerized for 5 days at 60°C. Thick sections (3 μm) were cut on an ultramicrotome (LKB BROMMA 2088, Ultrotome) with a premade glass knife and stained on a glass slide for LM examination. Thick sections were also observed under a light microscope (Leitz Wetzlar, Dolan Scientific Instruments, Dallas, TX) to determine the region of specimen for thin sectioning. Ultrathin sections (60–90 nm) were cut with a diamond knife on the same ultramicrotome and stained with uranyl acetate and lead citrate. The sections were examined with a Philips EM-300 transmission electron microscope (Philips Electronic Instruments) operating at 60 kV. Approximately 6–10 pictures were taken under the electron microscope for each muscle at magnifications of ×2,800, ×5,000, ×18,000, and ×35,000.

Ultrastructural muscle damage.

The inherent characteristics of EM analysis are qualitative in nature (10). Although the present EM analysis utilized consistent numerical data applied to all EM pictures in an attempt to provide some level of semiquantitative analysis, the EM data in the present study must still be viewed as descriptive and largely qualitative, as previously described by Devor and Faulkner (6). Therefore, in the present study, no statistical test was conducted on ultrastructural muscle damage differences between pre- and postexercise training.

All EM picture analyses for the ultrastructural cell changes were carefully assessed by an investigator blinded to the identity of the horse and time of collection (i.e., before or after training, and before or after exercise). A total of approximately two to four EM pictures (×5,000 magnification) were grouped for each muscle group and each biopsy time point for the horses and used for the analysis. To minimize a possibility of misreading ultrastructural changes, ranking scales were created for picture quality (from 0 as worst to 10 as best) and oblique section levels (from 0 as least to 10 as most). EM pictures with picture-quality scale of <7 and/or oblique section level scale of >3 were eliminated before our EM image analysis. The signs of ultrastructural damage for each muscle sample were carefully characterized based on our observations and defined as “Z-disc streaming” and “Z-disc disruption.” These morphological alterations have been previously described by our group (6) as signs of contraction-induced skeletal muscle damage after intense lengthening contractions. The number of sarcomeres (per ×5,000 EM picture) with these observed damage signs were counted in pre- and postexercise samples for all three incremental exercise tests at weeks 0, 8, and 10 of conditioning to examine exercise training-induced ultrastructural adaptations.

Biochemical Analyses

MHC electrophoresis.

MHC electrophoresis was performed after the protocol of Talmadge and Roy (35). A standard sample of adult rat costal diaphragm muscle (Sigma, St. Louis, MO) was included in gels and used as a control, which has been shown to contain four different MHC isoforms identified as type I, IIa, IIx, and IIb. Approximately 10 mg of isolated muscle (triceps, semimembranosus, and masseter) sample were homogenized in a 1:100 dilution in Tris buffer (62.5 mM Tris, pH 6.8). Polyacrylamide gels were comprised of 8% separating gel with 30% (wt/vol) glycerol and 4% stacking gel with 30% glycerol. Based on predetermined total protein concentrations by protein assay, 10-μl aliquots of diluted myosin, which contained 5 μg of protein, were subjected to electrophoresis for 26 h at 285 V and 4°C in 20-cm-long vertical gels. Separating gels were stained with Coomasie blue, cleared, and scanned with the Mode GS-700 Imaging Densitometer (Bio-Rad, Hercules, CA). MHC protein isoforms were identified as three bands: I, IIx, and IIa (see Fig. 7) from the fastest to the slowest migrating band, as described by Talmadge and Roy. The relative MHC protein isoform density (optical density) was quantified via Bio-Rad Multi-Analyst version 1.0 software. The relative percentages of MHC isoforms in each sample were transformed by utilizing an arcsine transformation to allow for statistical analysis.

Enzyme analyses.

Muscle tissue stored at −80°C was weighed once it had warmed to −20°C. For the citrate synthase (CS) assay, tissue samples were homogenized in a solution of 100 mM Tris buffer and 0.1% Triton X-100 in distilled H2O. The activity of CS as a measure of citric acid cycle activity was determined at 30°C using spectrophotometric techniques (4). For 3-OH-acyl-CoA-dehydrogenase (ACDH) and lactate dehydrogenase (LDH) assays, tissue was homogenized in a solution of 50% glycerol, 20 mM K2HPO4, 5 mM 2-mercaptoethanol, 0.5 mM EDTA, and 0.05% BSA in distilled H2O. The activities of ACDH as a measure of lipid oxidation and LDH as a measure of glycolytic capacity were analyzed at 25°C using fluorimetric techniques as described by Chi et al. (4). All assay mixtures were described in detail by Passonneau and Lowry (26).

Blood analyses.

Plasma lactate concentrations were measured using commercial kits (Sigma Chemical Kits, St. Louis, MO) on a microplate reader (Power Wave X, Bio-Tek Instruments, Winooski, VT). The activity of CK in serum was measured by The Ohio State University Veterinary Hospital Clinical Laboratory using an automated analyzer (Hitachi, Hialeah, FL).

Statistical Analysis

MHC isoform expression, enzymatic activities, and V̇o2 max data were analyzed using dependent t-tests to determine whether significant effects of exercise training existed. The effects of training were considered significant at P < 0.05. Plasma lactate concentrations were compared at the same speeds on each exercise trial by using a one-way repeated-measures ANOVA with a Tukey's post hoc test. Although this comparison tested for training-induced adaptations in plasma lactate at a given workload, it was limited to analyzing only those speeds that all horses were able to achieve. Therefore, to examine the effect of exercise training on plasma lactate concentrations in association with the improved exercise duration of aged horses during the same relative incremental exercise tests, the area under the lactate vs. time curve was calculated for each horse on each occasion by using the trapezoidal rule, and then a one-way repeated-measures ANOVA was used to compare these values.


Exercise Tests

Absolute (l/min) and relative (ml·kg−1·min−1) V̇o2 maxof the horses increased by 16 and 22%, respectively, after 10 wk of conditioning. The lesser increase in absolute V̇o2 max was due to a significant decrease (4.2%, P < 0.05) in body weight of horses after conditioning (Table 1).

View this table:
Table 1.

Effect of exercise training on V̇o2max and body weight

After 10 wk of exercise training, there was a 24% increase in run time to exhaustion during the incremental exercise test (P < 0.05; from 9.1 ± 0.4 to 11.3 ± 0.5 min). Maximum treadmill speed attained during the exercise trials increased 26% from week 0 (6.5 ± 0.3 m/s) to week 10 (8.2 ± 0.4 m/s) (P < 0.05). The conditioning-induced increase in maximum aerobic capacity of aged horses was accompanied by an increase in the peak plasma lactate concentration at their new peak treadmill speed after conditioning (7.7 ± 0.6 mM pretraining vs. 11.4 ± 0.6 mM posttraining) (Fig. 2). Training also resulted in a significant reduction in plasma lactate concentrations at the same absolute workloads of 6 and 7 m/s (Fig. 2). The calculated area under the lactate vs. time curve (s·mM) revealed there was an increase in total lactate response during the same relative incremental exercise test after 10 wk of training (P < 0.05; from 1,174 ± 234 s·mM at week 0 to 17,206 ± 1,501 s·mM at week 10). It also indicated a reduction in the area under the plasma lactate concentration-time curve during the same absolute incremental exercise test after 8 wk of training (P < 0.05; from 11,704 ± 235 s·mM at week 0 to 8,475 ± 193 s·mM at week 8).

Fig. 2.

Effect of exercise training on plasma lactate concentration. Week 0, an incremental exercise test before exercise training; week 8, an incremental exercise test after 8 wk of exercise training; week 10, an incremental exercise test after 10 wk of exercise training. Values are means ± SE; n = 6. *Significant difference between pre- and postexercise training lactate values (P < 0.05).

MHC Composition and Total Protein Concentration

There was no alteration in MHC composition in the masseter (control) muscle after 10 wk of exercise conditioning (P > 0.05) (Table 2). In contrast, there was a significant increase (10.9%; P < 0.05) in percentage of MHC IIa expression of the semimembranosus muscle after conditioning (Table 2). Furthermore, percentage of MHC IIx expression of triceps decreased significantly (3.1%; P < 0.05) after conditioning (Table 2). No other statistically significant changes in MHC isoform levels were found (Table 2).

View this table:
Table 2.

Effect of exercise training on percent MHC isoform expression

Muscle Oxidative and Glycolytic Capacity

Muscle CS.

The activity of CS (Table 3) in triceps muscle increased by 41% (P < 0.05) after 10 wk of conditioning, whereas CS activities in semimembranosus and masseter muscles were not changed significantly after conditioning (P > 0.05). In addition, the masseter muscle showed significantly (∼200%) greater CS activity compared with triceps and semimembranosus muscles (P < 0.05).

View this table:
Table 3.

Effect of exercise training on muscle enzyme activity

Muscle ACDH.

Similar to CS activity, ACDH activity in triceps muscle (Table 3) increased by 72% (P < 0.05) after 10 wk of exercise conditioning. There were no significant changes in ACDH activity in either the semimembranosus or masseter muscles. In addition, ACDH activities in the masseter muscles were significantly (∼3-fold) greater compared with those in semimembranosus and triceps muscles (P < 0.05).

Muscle LDH.

There was no significant change in the activity of LDH (Table 3) after 10 wk of exercise conditioning in any of the three muscles (P > 0.05). However, LDH activity in the masseter was significantly lower (50%; P < 0.05) than that in semimembranosus and triceps muscles.

Ultrastructural Muscle Damage and CK Response

Examples of our two categorical observations of ultrastructural damage are shown in Fig. 5. Compared with each preexercise CK value, there was a significant increase in CK activity (P < 0.05) after all three incremental exercise tests (Fig. 3). Exercise conditioning did not, however, reduce the magnitude of rise in serum CK activity after either the 8- or 10-wk incremental exercise tests (P > 0.05).

Fig. 3.

Effect of exercise training on serum creatine kinase (CK) activity. Values are means ± SE; n = 6. Pre-0, immediately before an incremental exercise test; Post-0, immediately after an incremental exercise test; Post-4hrs, 4 h after an incremental exercise test; Post-24hrs, 24 h after incremental exercise test. There was no statistically significant difference between pre- and postexercise training CK values (P > 0.05). *Significant difference in postincremental exercise test CK values compared with preexercise test CK values (P < 0.05).

There were signs of minor damage observed in all three muscle groups (Pre-1 semimembranosus, triceps, and masseter) before the first incremental exercise test. After each incremental exercise test at weeks 0, 8, and 10 of conditioning, triceps and semimembranosus muscles appeared to suffer more severe muscle structural damage on both LM (Fig. 4) and EM (Fig. 5) picture examinations, whereas the prevalence of damage appeared unchanged in control (masseter) muscle (Fig. 6). As shown in Fig. 5, cell damage in semimembranosus muscle from the same horse appeared to be more localized (Post-3) after 10 wk of exercise conditioning compared with damage in response to the first incremental exercise trial before conditioning, which was widespread (Post-1). However, before incremental exercise tests after training, triceps muscle (Pre-2 and Pre-3) appeared to exhibit more visible damage than semimembranosus muscle; thus comparisons of ultrastructural muscle damage between pre- and postexercise training were difficult to assess.

Fig. 4.

Exercise-induced muscle damage in semimembranosus at light microscopic level (×805). Pre-1, before 1st incremental exercise test; Post-1, after 1st incremental exercise test.

Fig. 5.

Ultrastructural muscle damage in semimembranosus at electron microscopic (EM) levels after each incremental exercise test before and after exercise training at ×5,000 (left) and ×18,000 (right) magnification. Post-3, after 3rd incremental exercise test; a, widely spread Z-disc streaming before training; b, Z-line disruption; c, focalized Z-disc streaming after training.

Fig. 6.

No signs of ultrastructural muscle damage in masseter at EM levels after the first incremental exercise test before exercise training.

Fig. 7.

Sodium dodecyl sulfate-polyacrylamide gel of horse semimembranosus, triceps, and masseter muscle samples showing myosin heavy chain isoform separation before and after exercise training. Semimem, semimembranosus muscle; Pre, preexercise training; Post, postexercise training.

Based on our previous findings (6), the average counts of sarcomeres with damage signs per each ×5,000 EM picture were assessed for each biopsy sampling time point. In the semimembranosus muscle (Post-1) after the first incremental exercise trial before exercise conditioning, the average counts of sarcomeres with damage signs were 23.6 ± 2.7 (mean ± SE per ×5,000 EM picture), whereas Pre-1 muscle sections contained 5.2 ± 0.6 sarcomeres with these damage signs. Because each ×5,000 EM picture showed ∼80 to 120 total sarcomeres, we observed signs of damage in ∼15–40% of sarcomeres in the Post-1 semimembranosus muscle sections after the first incremental exercise test before training. This represented a nearly fivefold increase in the prevalence of damage signs observed compared with Pre-1 semimembranosus muscle sections. In response to the same absolute incremental exercise trial after 8 wk of conditioning, signs of damage in Post-2 muscle sections (15.1 ± 1.8) were observed in ∼10–20% of sarcomeres, which represented an approximate twofold increase in damage prevalence relative to Pre-2 muscle sections (7.6 ± 1.1). A similar trend was noted in response to the same relative incremental exercise trial after 10 wk of conditioning, because the number of observed damage signs in Post-3 muscle sections (16.1 ± 2.1) was about twice the number observed in Pre-3 muscle sections (7.9 ± 0.9). We interpret these observations in semimembranosus muscle as an indication that the number of damaged sarcomeres in response to incremental exercise tests was reduced after exercise conditioning. This apparent conditioning response was not observed in the triceps muscle, however, as the number of observed acute exercise-induced ultrastructural damage signs appeared to increase similarly (∼2.5-fold) before and after the exercise conditioning program. The triceps muscle, however, showed similar exercise-induced ultrastructural cell damage after exercise conditioning, with ∼15–55% total section area with damage in Post-1, Post-2, and Post-3. There were ∼2.5-fold more sarcomeres with damage after both 8 wk of training (Pre-2: 11.2 ± 1.4; Post-2: 25.1 ± 2.5) and after 10 wk of training (Pre-3: 11.5 ± 1.5; Post-3: 26.4 ± 2.3), whereas there were approximately threefold more damaged sarcomeres in pretraining muscle sections (Pre-1: 7.3 ± 0.8; Post-1: 23.6 ± 3.0).

In summary, no evidence was found in triceps muscle for reduced ultrastructural cell damage after exercise conditioning (images not shown). By contrast, there was an apparent reduction in the severity of ultrastructural damage in semimembranosus muscle at the same absolute and the same relative workloads, suggesting a protective effect of training in semimembranosus muscle only.


We demonstrated that 10 wk of exercise conditioning by aged horses resulted in marked increases in variables indicative of enhanced aerobic capacity and an increase in proportion of MHC IIa, although adaptations appeared to be muscle-group specific (16). Exercise conditioning of aged horses increased V̇o2 max, total treadmill run time, maximum treadmill run speed, and total lactate response during the incremental exercise trials. We also observed that 10 wk of aerobic exercise conditioning appeared to attenuate the severity of acute exercise-induced muscle ultrastructural damage in the semimembranosus muscle. Furthermore, the adaptations in muscle metabolic properties of exercise-trained animals are well coordinated with the conditioning-induced transition in MHC isoforms (33, 34) of aged skeletal muscle. Our findings suggest that aerobic exercise training cannot only improve the metabolic properties but also improve the resistance to injury of aged skeletal muscle cells. The conditioning-induced cellular adaptations were concomitant with the improvement in whole body aerobic capacity and performance in aged horses (1).

Older horses exhibit many of the changes in function (25, 28) that are characteristic of aged humans (5, 16, 37). However, the responses of older horses to exercise training have not been demonstrated previously. Our findings that older horses are capable of aerobic exercise training-induced increases in V̇o2 max and exercise capacity, as indicated by lower blood lactate concentrations at similar speeds and prolonged endurance and higher maximal speeds during incremental exercise testing, have been demonstrated in humans (2). That these changes are at least partially attributable to conditioning-induced changes in skeletal muscle properties is indicated by the increased proportion of MHC IIa and activities of CS and ACDH.

Changes in MHC isoforms in response to training in the aged horses are similar to those previously observed in younger horses (29, 33). We observed the increased percentage of MHC IIa in semimembranosus muscle and the decreased percentage of MHC IIx in triceps muscle of exercise-trained aged horses. There was no alteration in MHC composition of the nonlocomotor (masseter) muscle after the aerobic exercise training, demonstrating that the conditioning-induced responses are associated with local muscle activity and are not attributable to the systemic effects of aerobic exercise conditioning.

The results of the present study demonstrate that relatively intense aerobic exercise training improves the muscle oxidative capacity of aged horses, similar to the effect in other species. The activities of mitochondrial CS and ACDH in the triceps muscle of aged mares were markedly enhanced by 10 wk of treadmill exercise training, whereas there was no alteration observed in muscle LDH activity after training. Regardless of mammalian age and species, endurance exercise training is known to induce an increase in the activity of the key aerobic enzymes that are markers of muscle oxidative capacity (4, 9, 14, 33, 34). Although exercise training is also known to induce an increase in glycolytic enzyme activities (30, 36), the majority of previous findings suggest no conditioning-induced alterations in muscle LDH activity (9, 14), which is similar to our finding in the present study.

Our LM and EM observations suggest that 10 wk of relatively intense treadmill exercise training in aged horses reduces the magnitude of acute exercise-induced widespread damage compared with the pretrained muscle. After a period of conditioning, signs of acute exercise-induced damage persisted but appeared to be more focal. This appearance of localized (focal) damage has been previously identified by Manfredi et al. (23) as indicative of reduced ultrastructural muscle damage (vs. widespread areas of damage), which was based on their observations in EM pictures of human vastus lateralis. This localization of ultrastructural damage after exercise conditioning was also supported by our observations of reduced numbers of sarcomeres containing signs of cell damage in mechanically loaded muscles of aged horses.

Roth et al. (31, 32) recently attempted to examine the effects of heavy resistance exercise training on ultrastructural muscle damage in young and aged individuals via EM analysis. They demonstrated that 9 wk of high-volume, heavy-resistance training resulted in similar levels of ultrastructural muscle damage in the vastus lateralis of young and older men (32). Furthermore, a follow-up report from the same research group suggests that the muscle damage is greater in older women than young women, whereas a similar improvement in muscle strength was observed in both young and older groups after the same training regimen (31). No age or sex comparison was made in the present study, but our findings reveal attenuation of exercise-induced ultrastructural muscle damage after aerobic exercise training in aged horses.

The conditioning-induced increase in resistance to muscle damage in response to one bout of exhaustive treadmill exercise might be due to increased recruitment and contribution of fast oxidative fibers after an intense aerobic exercise training program compared with fast glycolytic fibers. Our hypothesis is based on a previous theory by Lieber and Friden (22) who presented potential mechanisms of muscle damage in fast glycolytic muscle fibers that fatigued in the first few minutes of the exercise period. They hypothesized that, during the intense exercise, fast glycolytic fibers may become unable to regenerate ATP and enter a rigor state. As stiffness of the rigor fibers increases, mechanical disruption would occur in these fibers due to subsequent stretch.

In previous EM studies in equine skeletal muscles, similar observations of ultrastructural cell changes have been reported in response to intense exercise tests (12, 24). McCutcheon and colleagues (24) have demonstrated an increased area of mitochondria and sarcoplasmic reticulum after exhaustive exercise in horses. They reported swollen mitochondria after intense exercise, which was also observed in highly oxidative masseter muscle (Fig. 6) in the present study. In gastrocnemius muscles of human marathon runners, Hikida et al. (13) have reported similar morphological observations, but the damage appeared more severe. As described earlier, the characteristics of ultrastructural damage including Z-disc streaming and Z-disc disruption in the present study were similar to those previously reported characteristics of the contraction-induced damage initiated by high-intensity resistance exercise (lengthening contractions) (6, 31, 32). Although not typically discussed as signs of exercise-induced ultrastructural damage, other phenomena frequently observed in biopsy samples after incremental exercise tests include splitting of myofibrils within the sarcomere or Z-disc and an oblique appearance due to individual myofibrils not returning to resting length. These unique observations from the present study may be characteristic of aerobic exercise-induced ultrastructural cell damage and may provide new evidence of the early morphological damage consequences in mammalian skeletal muscle after exhaustive treadmill running.

Based on our observation of the splitting of myofibrils within the sarcomere or Z-disc, we hypothesize that training-induced attenuation of muscle cell damage may be due to exercise training-induced muscle fiber hypertrophy (17). The splitting of sarcomeric elements may represent a hypertrophic stimulus following numerous lengthening muscle contractions during the treadmill running. This concept is supported by Radak et al. (27), who demonstrated that aerobic exercise training increases the activity of DNA repair and resistance against oxidative stress in proteins. As Devor and Faulkner previously demonstrated (6), the attenuation in ultrastructural muscle damage we observed may be due to training-induced increases in satellite cell activity (3, 18), which may have increased resistance to contraction-induced muscle injury.

The present results indicate that the conditioning-induced adaptations in oxidative capacity and attenuating effect on the severity of exercise-induced muscle damage may have been dependent on the relative workload imposed on individual muscles. The training-induced attenuation of the exercise-induced ultrastructural muscle damage was only observed in the semimembranosus muscle, but not in the triceps muscle, whereas mitochondrial enzyme activity increased after training in the triceps but not in the semimembranosus muscle.

There is greater mechanical stress applied to the smaller triceps muscle compared with the semimembranosus muscle during treadmill running because ∼60% of the body mass of a horse is centered over the forelimbs. Unlike the triceps muscle, which acts as the primary elbow extensor, the semimembranosus muscle acts in concert with both biceps femoris and semitendinosus muscles. Greater damage might therefore be expected in the triceps compared with the semimembranosus muscle with each bout of treadmill exercise training. Based on the disparate prevalence of ultrastructural damage in baseline triceps and semimembranosus muscle samples at Pre-2 and Pre-3, 1 wk of recovery time before these samples may not have been sufficient for complete recovery of cell damage in the triceps muscle but appears to have been adequate in semimembranosus muscle. If recovery in the triceps was in fact incomplete in preexercise biopsy samples at each timepoint, this obviously would have precluded us from assessing damage resulting solely from the incremental exercise tests after conditioning. On the other hand, the relatively greater workload imposed on triceps during treadmill exercise training may have led to greater adaptations in mitochondrial enzyme activity compared with semimembranosus muscle. For future morphological experiments, the recovery time may need to be extended to achieve appropriate pre-/postexercise comparisons in ultrastructural muscle damage. However, the optimal recovery period before detraining adaptations is not known and may be muscle-group specific.

Blood CK activity was higher after all incremental exercise trials (Fig. 3). As previously reported (23), our CK data also reveal that the measurement of blood CK activity as an indirect cell damage marker is not sensitive enough to support our direct morphological evidence under LM and EM analyses. It appeared to be mainly due to noticeable individual differences in blood CK response after incremental exercise test.


The results of the present study suggest that relatively intense aerobic exercise training can overcome the age-related reduction in the muscle oxidative capacity of horses. This finding is consistent with an exercise training-induced transition of MHC isoforms. Furthermore, aerobic exercise training can attenuate the severity of exercise-induced ultrastructural cell damage in aged skeletal muscle at given absolute and/or the same relative workloads. Therefore, aerobic exercise training can improve endurance (oxidative capacity and plasticity) as well as strength (resistance to cell damage) of aged skeletal muscle cells, and these cellular adaptations are in concert with the improvement in whole body aerobic capacity.

Exercise training may increase the resistance of older individuals to exercise-induced muscle injury. Consequently, exercise training may assist older individuals in preventing abrupt injury associated with age-related impairments such as a reduction in muscle strength, impaired mobility, and tendency of falling. However, these adaptations in mammalian skeletal muscle appear to be muscle group and/or load specific.


This study was supported by the Equine Research Fund from the College of Veterinary Medicine, Council for Research, at The Ohio State University.


The authors thank Dr. Wuthichai Klomkleaw for time and technical support with the EM analyses, William J. Fink at Ball State University for providing detailed enzymatic analysis protocols, and Dr. Marcas M. Bamman at the University of Alabama at Birmingham for support of data presentation at the 2003 American College of Sports Medicine annual meeting.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract