Evaluation of neuromuscular electrical stimulation on fibre characteristics and oxidative capacity in equine skeletal muscles


email: Anna.Bergh@afb.slu.se


Reasons for performing study: Neuromuscular electrical stimulation (NMES) is used to increase or maintain muscle strength during rehabilitation. Human studies investigating different protocols show that some treatments induce changes in muscle characteristics. Despite the frequent use of NMES in horses, no studies have been published describing its efficacy.

Objectives: To investigate the effects of a NMES protocol on equine fibre types and areas, glycogen concentrations and enzyme activities.

Methods: NMES was administrated to m. gluteus medius and m. longissimus dorsi, on one side of 6 healthy Standardbred horses. The contralateral side of each muscle served as a nonstimulated control. The horses were stimulated at 50 Hz a day, with 21–39 mA, for 45–60 min, 5 days a week for 4 weeks. Needle biopsies were obtained from the muscles on both sides before and after the experimental period. Muscle samples were analysed for fibre type proportions and area using histochemical methods and for glycogen and enzyme activities (citrate synthase, 3-OH-acyl CoA dehydrogenase, hexokinase and lactate dehydrogenase) using biochemical methods. Muscle contractions at the location and depth of the muscle biopsy were confirmed by diagnostic ultrasound. Nonparametric tests (Mann-Whitney, Wilcoxon sign-rank) were used for statistical analysis.

Results: No significant differences were observed in the percentage of types I, IIA or IIX fibres, fibre areas, glycogen levels or enzyme activities either when comparing stimulated and nonstimulated muscles before and after the NMES treatment, or when comparing the left and right muscle samples.

Conclusions: The NMES treatment was well tolerated by the horses, but the present protocol did not induce significant muscle adaptations. Further studies are needed to describe the effect of more intense and/or prolonged NMES treatment protocols on muscles of healthy horses, and to describe if stimulation protocols induce positive changes in atrophied muscles.


Physical rehabilitation of equine musculoskeletal injuries is an expanding field (McGowan et al. 2007). Methods used in man are often extrapolated to veterinary medicine without sufficient scientific evidence of their effects. One modality used is neuromuscular electrical stimulation (NMES), where electrodes are placed on the skin and contractions in the muscle are produced by electrical stimulation (Lake 1992). With NMES, it is possible to induce contraction patterns different from those during voluntary contraction (Delitto and Snyder-Mackler 1990). It is generally accepted that NMES first recruits force-producing type II muscle fibres and then type I fibres, which is the reverse of the muscle recruitment pattern in a voluntary contraction (Hainaut and Duchateau 1992; Requena-Sánchez et al. 2005).

The main purpose of using NMES is to improve the maximal voluntary strength of weakened and/or healthy muscles. Human reviews report increased muscle strength due to NMES in athletic training, training during immobilisation and pathology treatments (Laughman et al. 1983; Gibson et al. 1988; Wigerstad-Lossing et al. 1988; Hainaut and Duchateau 1992; Parker et al. 2003; Gondin et al. 2006; Herrero et al. 2006; Colson et al. 2009). It can be concluded that the effect of NMES on healthy subjects is equal to but not superior to voluntary strength training, that the protocols that associate electrostimulation with voluntary muscle contractions seem to have more effect than electrostimulation alone, and that NMES is effective in reducing muscle wasting during immobilisation (Bax et al. 2005; Requena-Sánchez et al. 2005; Vanderthommen and Duchateau 2007). In man, NMES has been shown to cause fibre type changes towards fewer fast contracting type II fibres, increased cross-sectional area and increased citrate synthase (CS) activity indicating a higher oxidative capacity in the muscle (Gauthier et al. 1992; Thériault et al. 1994; Pérez et al. 2002; Nuhr et al. 2003, 2004; Del Corso et al. 2007). The above results are contradicted by other human studies where no significant differences in muscle strength, muscle mass, fibre types or CS activity were seen after electrical stimulation (Eriksson et al. 1981; Arvidsson et al. 1986; Suetta et al. 2004, 2008).

To the best of our knowledge, there is nothing reported in the literature on NMES and its effect on horse muscle tissue. Changes in strength cannot be measured directly in horses as in man, but alterations in muscle histochemical and fibre type properties can be studied with the muscle biopsy technique. Therefore, the aim of present study was to use this technique to investigate if a NMES protocol has any effect on muscle characteristics in healthy untrained horses.

Materials and methods

Horses and environment

The study comprised 6 healthy Standardbred trotters (3 mares and 3 geldings), with a mean age of 6 years (range 4–13 years) and a mean weight of 471 kg (range 420–492 kg). During the treatment period, the horses were out on pasture for 4–5 h/day. The experiments were approved by the Ethical Committee on Animal Experiments in Uppsala, Sweden.


The muscles chosen for stimulation were m. gluteus medius and m. longissimus dorsi on one side, while the muscles on the other side did not receive any intervention and thereby served as within-subject controls.


Neuromuscular electrical stimulation was performed with the muscle stimulator CEFAR MYO1. The electrodes were 50 × 100 mm carbon-impregnated rubber electrodes1. Blue Gel1 was used to create good contact between the electrode and the skin.

Electrode placement:  After identifying the motor points (the area where the motor nerve enters the muscle), the positions of the 4 electrodes (2 for each muscle) were shaved to improve the contact between the electrodes and the skin and to assist the investigator in finding the motor point during subsequent stimulation sessions.

Stimulation protocol:  The horses were stimulated once a day, between 45 and 60 min, 5 days a week for 4 weeks. Both muscles were stimulated simultaneously. The number of contractions was increased from 3 × 10 during week 1, 3 × 15 during weeks 2 and 3, to 3 × 20 contractions week 4. There was a two-minute rest between the sets of contractions. The stimulation was a biphasic rectangular pulse form with a frequency of 50 Hz and with a pulse width set to 300 µs. The protocol included a 3 s rise to the full contraction, which was held for 10 s followed by a 2 s decrease to 0 mA. The total stimulation time was therefore 15 s. In order to achieve a comparable degree of muscle contraction, the stimulation current was set individually for each horse (Table 1) with such intensity that it produced an obvious visual and palpable contraction in the muscle, without any apparent discomfort for the horse.

Table 1. Amperage interval (mA) for the horses at the beginning and the end of the weeks of (NMES) neuromuscular electrical stimulation treatment
 Week 1Week 2Week 3Week 4
Mean RangeMean RangeMean RangeMean RangeMean RangeMean RangeMean RangeMean Range
m. gluteus2528.5293333.534.533.536
m. longissimus22.526.52731.531.5323234

Muscle biopsies: Muscle biopsies (Lindholm and Piehl 1974) were performed with a Bergström biopsy needle2 (external diameter 5 mm). Biopsy samples were taken by the same investigator from both the stimulated and nonstimulated m. gluteus medius and m. longissimus dorsi before and after the treatment period, at a site halfway between the 2 electrodes for each muscle. The biopsy needle was inserted to a depth of 4–5 cm and the collected muscle sample was frozen in liquid nitrogen and then stored at -80°C until analysed. The muscle sample for histochemical analysis was rolled in talcum powder before being frozen in liquid nitrogen.

Histochemical analysis: Transverse serial sections (10 µm) of muscle biopsy specimen were cut in a cryostat at -20°C. These sections were stained for myosin ATPase after both acid (pH 4.3 and 4.6) and alkaline (pH 10.3) pre-incubation. Muscle fibres were identified as type I, type II A, type II B or type II C (Brooke and Kaiser 1970; Essén et al. 1980). Type I fibres stained black, type II A fibres white and type II B fibres grey or brown after pre-incubation at pH 4.6. Type II C fibres stained black or grey after pre-incubation at pH 4.3, 4.6 and 10.3. Type II B fibres identified with the ATPase staining method will be called type II X fibres in the present study. The reason for this is to not confuse with the nomenclature used when 4 fibre types are identified due to different MHC isoforms (MHCI, MHCIIA, MHCIIX, MHCIIB). Since MHCIIB fibres, which are expressed in small mammals, do not exist in horse muscle type II B fibres are more related to MHCIIX fibres (Rivero and Serrano 1999).

Muscle composition was determined by typing at least 200 fibres. In a few cases, the samples contained a lower number of fibres (the range was 125–286 fibres). In a computerised image analysis system (BIO-RAD)3, fibre type distribution (%) and mean fibre type area (µm2) for each fibre type were investigated. All measurements were carried out blind, and the codes were broken only after collection of all the data.

Biochemical analysis: Whole muscle biochemical analyses were performed on freeze-dried muscle, weighing 1–3 mg, after blood, fat and connective tissues had been dissected free under a microscope. One muscle sample was homogenised in ice cooled 0.1 mol/l phosphate buffer at pH 7.3 and the activities of citrate synthase (CS), 3-OH-acyl-CoA-dehydrogenase (HAD), hexokinase (HK) and lactate dehydrogenase (LDH) were analysed as previously described (Essén et al. 1980; Essén-Gustavsson et al. 1984). CS and HAD were analysed to evaluate the muscle oxidative capacity, HK to evaluate the capacity for phosphorylation of glucose and LDH was analysed as a marker for glycolytic capacity as this enzyme is needed for lactate production. The muscle sample for glycogen determination was boiled for 2 h in 1 mol/l HCl and the glucose residues determined fluorometrically (Lowry and Passonneau 1972).

Additional study on an anaesthetised horse: On one occasion, one anaesthetised healthy Standardbred trotter was placed in lateral recumbency and stimulated with the identical programme as the standing horses. The intensity required to produce an obvious visual and palpable contraction comparable with those of the standing horses was 34.5–39.0 mA. During stimulation, an experienced radiologist confirmed the localisation and depth of the muscle contraction with a diagnostic ultrasound (Mindray)4, using a linear probe at the frequency of 7 MHz.

Statistics: SPSS for Windows version 10 was used for data analysis. The Wilcoxon sign-rank tests were applied for comparison of the differences between pre- and post values for both stimulated side and control side. The Mann-Whitney test was used for comparisons between horses. Statistical significance was set to P<0.05.


The stimulation intensity to the horses is reported in Table 1, results on muscle fibre characteristics are shown in Table 2 and results on enzyme activities and glycogen in Table 3. Type II C fibres were found in only some samples and, since there were so few fibres, they were added to the type II A fibre pool, since II C fibre areas mostly resembled this fibre type. Before the treatment period started, there were no significant differences in muscle characteristics between the NMES and the control side. NMES treatment for 4 weeks did not induce any significant changes in fibre type composition, fibre area, enzyme activities and glycogen content in any of the muscles. Furthermore, no differences were seen in these parameters when comparing the left and right side of the muscles. No apparent discomfort or skin irritation was detected in any horse. The result from the evaluation with diagnostic ultrasound shows a clear visual muscle contraction down to a depth of 8 cm in the m. gluteus medius muscle and a depth of 6 cm in the m. longissimus dorsi muscle.

Table 2. Mean ± s.d. for fibre type composition and fibre type areas in m. gluteus medius and m. longissimus dorsi of 6 horses without (control) and with (NMES) neuromuscular electrical stimulation for 4 weeks
 M. gluteus mediusM. longissimus dorsi
BeforeAfter 4 weeksBeforeAfter 4 weeks
Fibre type (%)    
  Control14 ± 513 ± 220 ± 518 ± 3
  NMES17 ± 515 ± 721 ± 220 ± 3
 II A    
  Control42 ± 343 ± 441 ± 340 ± 4
  NMES42 ± 843 ± 843 ± 642 ± 6
 II X    
  Control44 ± 644 ± 339 ± 442 ± 4
  NMES41 ± 1042 ± 936 ± 638 ± 5
Fibre area (µm2)    
  Control2190 ± 4572075 ± 7092690 ± 7162725 ± 841
  NMES2014 ± 3391787 ± 5752733 ± 3912611 ± 327
 II A    
  Control2635 ± 3922430 ± 7552866 ± 6232786 ± 584
  NMES2772 ± 3162138 ± 4192962 ± 4702999 ± 476
 II X    
  Control4494 ± 6734476 ± 12655057 ± 17145161 ± 984
  NMES4552 ± 11553622 ± 10765465 ± 9895357 ± 594
Mean area (µm2)    
  Control3446 ± 5053389 ± 8963654 ± 9223799 ± 840
  NMES3436 ± 4042735 ± 5543792 ± 5573819 ± 428
Table 3. Mean ± s.d. for citrate synthase (CS), 3-OH acyl CoA dehydrogenase (HAD), lactate dehydrogenase (LDH) and hexokinase (HK) and glycogen in m. gluteus medius and m. longissimus dorsi activities of 6 horses without (control) and with (NMES) neuromuscular electrical stimulation for 4 weeks
 M. gluteus mediusM. longissimus dorsi
BeforeAfter 4 weeksBeforeAfter 4 weeks
CS (mmol/kg/min)    
 Control53 ± 1352 ± 1054 ± 1249 ± 9
 NMES53 ± 1053 ± 1353 ± 1551 ± 13
HAD (mmol/kg/min)    
 Control22 ± 524 ± 432 ± 530 ± 5
 NMES23 ± 424 ± 428 ± 331 ± 5
LDH (mmol/kg/min)    
 Control1300 ± 1741284 ± 1892015 ± 1771887 ± 114
 NMES1291 ± 951258 ± 591936 ± 951991 ± 95
HK (mmol/kg/min)    
 Control3.9 ± 0.73.6 ± 0.62.7 ± 0.73.2 ± 1.3
 NMES4.1 ± 0.83.9 ± 0.62.7 ± 0.52.6 ± 1.3
Glycogen (mmol/kg)    
 Control495 ± 62500 ± 101522 ± 84535 ± 57
 NMES493 ± 70500 ± 93546 ± 41567 ± 50


Results from the present study demonstrate that daily treatment with the current NMES-protocol for 5 days a week over 4 weeks does not cause any significant changes in muscle fibre types, fibre areas, glycogen content or enzyme activities in healthy horses.

Data on muscle characteristics in the present study are in good agreement with earlier findings on Standardbred trotters (Ronéus et al. 1992; Karlström et al. 1994). Data obtained from the right side of the muscle did not differ from those obtained on the left side, which also agrees with results from an earlier study (Essén-Gustavsson et al. 1989). In man, NMES may influence the muscle properties in both legs, when stimulation was performed only in one leg (Carrol et al. 2006). However, such a crossover effect is not likely in the present study as no changes were seen in either stimulated or nonstimulated side, before or after stimulation.

Our results diverge from some results found in human studies, where NMES has been shown to induce fibre type changes. The amount of type II A fibres increased while the amount of type II X and type I fibres decreased after electrical stimulation for 30 min, 3 days/week for 6 weeks (Pérez et al. 2002) and stimulation for 10 weeks, 4 h/day and 7 days/week caused a decrease in II X fibres and an increase in type I fibres (Nuhr et al. 2003, 2004). These studies also showed changes in enzyme activities that indicated an increased oxidative capacity in the muscle.

Other studies have shown increased CS activity after electrical stimulation; 3 h/day, 6 days/week for 6 weeks (Gauthier et al. 1992) and after 8 h/day, 6 days/week for 4 weeks (Thériault et al. 1994). A slight increase in fibre cross-sectional area has been reported (Pérez et al. 2002; Del Corso et al. 2007). There are, however, studies that show no changes in enzymes and muscle fibre characteristics after similar stimulation (Eriksson et al. 1981; Arvidsson et al. 1986; Suetta et al. 2004, 2008). The influence of NMES on muscle phenotype depends primarily on the total number of impulses delivered and is less dependent on the stimulation frequency (Salmons 2009). The fact that neither fibre type and fibre area changes nor any alterations in enzymes were observed in the present study may be related to the fact that NMES was used for 4 weeks in contrast to the longer durations employed in the previously described studies.

An important methodological question is whether the muscle specimens obtained with the biopsy needle had actually been activated. The depth of biopsy corresponded to the depth taken in human studies on m. quadriceps, which shows that a depletion of glycogen and phosphagen stores occurs in the muscle after electrical stimulation (Eriksson et al. 1981; Söderlund et al. 1992). Further, the evaluation with diagnostic ultrasound on the anaesthetised horse showed a clear visible contraction in the muscles at the depth of the muscle biopsy, thus indicating muscle activation. The intensity used on the anaesthetised horse produced contractions of the same degree as those in standing horses, estimated by visualisation and palpation.

Training studies on horses show that changes can occur in fibre characteristics and oxidative capacity in muscles, evaluated by muscle biopsies taken at the same depth as in the present study. No changes in fibre type composition but an increase in CS activity and a decrease in type II A fibre area was seen after 5 weeks training (Essén-Gustavsson et al. 1989). Fibre type changes with a reduction in type II X fibres and a concomitant increase in type II A fibres, as well as an increase in oxidative capacity and glycogen levels has been shown in horses after 3 months of endurance training (Serrano et al. 2000). This indicates that changes in fibre types occurred after a longer duration of training, while metabolic changes occurred earlier.

Although the NMES protocol used in the present study was designed to simulate strength-training with an increasing number of contractions over the experimental period, it was used on the back muscles of healthy horses - muscles that are more or less constantly engaged where an animal stands for 18–22 h/day (Dallaire 1986). It is likely that the stimulation level given to the healthy horses in the present study may have been lower than their normal work load and therefore not sufficient to induce any significant changes in muscle characteristics.

One of the limitations of NMES is the lack of evidence-based knowledge on how to maximise elicited torque output and to quantify optimal parameter settings. It has been indicated that an increase in current amplitude and pulse duration increases both the maximum voluntary isometric torque, as well as the percentage of muscle group activated (Han et al. 2007; Gorgey and Dudley 2008). However, a rise of stimulus duration was connected to an increase in patient discomfort (Kaczmarek et al. 2009). The stimulation characteristics of the present NMES protocol were selected according to previous recommendations (Hainaut and Duchateau 1992). It was chosen to mimic strength training used in clinical practices. It is generally accepted that the skin produces resistance to current flow by patient dependent capacitive impedance (Dorgan and Reilly 1999). Therefore, the current intensity was individually set in order to create the comparable highest work load and muscle contraction possible, without causing serious discomfort. If the protocol had used the same intensity for each horse, there had been a major risk of having some horses not showing any muscle contractions and some experiencing discomfort during the stimulation.

On the basis of present results, it may be of little clinical use to stimulate a healthy horse less than according to the protocol described. In conclusion, healthy horses tolerated the NMES protocol used in the present study and no changes in muscle characteristics were seen after 4 weeks. If muscle adaptations occur after more intense and/or longer stimulation periods needs to be investigated in future studies and also if there is an effect of NMES in rehabilitation of horses with muscle hypotrophy.


The authors wish to express their sincere gratitude to AB Trav och Galopp (ATG) for financial support, Cefar Medical AB for the use of equipment, Dr Nostell for taking the muscle biopsies and Dr Uhlhorn for ultrasound evaluation.

Manufacturers' addresses

1 Cefar Medical AB, Lund, Sweden.

2 Dixons surgical instruments Ltd, Wickford, Essex, UK.

3 Scan-Beam, Hadsun, Denmark.

4 Scandivet AB, Enköping, Sweden.