Metabolic changes in four beat gaited horses after field marcha simulation



    1. Núcleo de Pesquisa Equina and Professor do Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Rua Dom Manuel de Medeiros, Bairro de Dois Irmãos, Recife, Pernambuco, Brazil; and Equine Science Center, Departmente of Animal Sciences, Rutgers University the State University of New Jersey, New Brunswick, USA.
    Search for more papers by this author

    Corresponding author
    1. Núcleo de Pesquisa Equina and Professor do Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Rua Dom Manuel de Medeiros, Bairro de Dois Irmãos, Recife, Pernambuco, Brazil; and Equine Science Center, Departmente of Animal Sciences, Rutgers University the State University of New Jersey, New Brunswick, USA.
    Search for more papers by this author
  • H. E. C. C. C. MANSO,

    1. Núcleo de Pesquisa Equina and Professor do Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Rua Dom Manuel de Medeiros, Bairro de Dois Irmãos, Recife, Pernambuco, Brazil; and Equine Science Center, Departmente of Animal Sciences, Rutgers University the State University of New Jersey, New Brunswick, USA.
    Search for more papers by this author

    1. Núcleo de Pesquisa Equina and Professor do Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Rua Dom Manuel de Medeiros, Bairro de Dois Irmãos, Recife, Pernambuco, Brazil; and Equine Science Center, Departmente of Animal Sciences, Rutgers University the State University of New Jersey, New Brunswick, USA.
    Search for more papers by this author
  • K. H. McKEEVER

    1. Núcleo de Pesquisa Equina and Professor do Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Rua Dom Manuel de Medeiros, Bairro de Dois Irmãos, Recife, Pernambuco, Brazil; and Equine Science Center, Departmente of Animal Sciences, Rutgers University the State University of New Jersey, New Brunswick, USA.
    Search for more papers by this author



Reasons for performing study: Mangalarga-Marchador is a popular 4-gaited Brazilian horse breed; however, there is little information about their metabolic and physiological response to exercise.

Objectives: To measure physiological and metabolic responses of the Mangalarga-Marchador to a simulated marcha field test and to compare these responses between 2 types of marcha gaits (picada and batida).

Methods: Thirteen horses were used in the study and randomly assigned to either the picada or batida gait for the simulated marcha field test (speed ∼3.2 m/s; 30 min; load ∼80 kg).

Measurements: Included body composition, heart rate (HR), respiratory rate (RR), glucose (GLUC), lactate (LACT), packed cell volume (PCV), total plasma protein (TPP), albumin, urea, creatinine, total and HDL cholesterol, triglycerides, creatine kinase, alanine, glutamate and glutamine (GLN). Measurements were obtained pretest (control/fasting), immediately after simulation (T0), and 15 (T15), 30 (T30) and 240 (T240) min after the simulation. Lactate (LACT) was measured at T0, T15 and T30. Data were analysed using ANOVA, Tukey's test and t tests with significance set at P<0.05.

Results: Significant acute changes were observed in HR, RR, [GLUC], [LACT], [TPP], PCV and [GLN] (P<0.05) relative to control. Heart rate fell below 60 beats/min at T15 and RR recovered to pretest values by T240. Significant increases in [GLUC], [LACT], PCV and [TPP] and a decrease in [GLN] were observed at T0. Treatment and interaction effects were also observed between marcha types and time of sampling for HR, RF, PCV, and [LACT] (P<0.05). These parameters were large in picada.

Conclusion: The simulation of field-test produced changes in some physiological and blood parameters in marcha horses, with some degree of dehydration during recovery period. Also, it was demonstrated that picada horses spend more energy when compared with batida horses at the the same speed.

Potential relevance: Batida horses expend less energy when compared with picada horses, which will need special attention in their training and nutritional management.


Four-gaited or marcha breeds were selectively bred in Brazil to facilitate rides between farms and cities. The smooth nature of their gait made them ideal for people that are not skillful at riding horses at a trot or gallop. In Brazil, there are 4 different marcha horse breeds (Campolina, Mangalarga, Mangalarga-Marchador and Piquira) and one donkey breed (Pêga). Other breeds around the world have gaits with similar characteristics (Robilliard et al. 2007), such as ponies and Paso-Fino horses. The Brazilian breeds have at least 2 natural types of 4 beat gaits. The picada gait is a broken pace with little vertical movement of the rider, which is similar to tölt by ponies. The batida gait is a broken trot with significant movements of the legs in a diagonal pattern and with a vertical rider's movement. The ideal range of speed of the marcha gaits is 3.0–4.0 m/s because this speed allows maximum stability, with significant moments of triple support and a 4 beat sequence and the absence of aerial phase (Procópio 2003).

There are approximately 350,000 registered Mangalarga-Marchador making them the most popular purebred breed of horse in Brazil. Unfortunately, there is no information about the physiological and metabolic responses associated with 4 beat gaits during and/or after exercise. Recently, it has been shown that different breeds of fit marcha horses have a large amount of body fat accumulation (∼14.8%) compared to other regional equine sport breeds (Manso Filho et al. 2009a). This suggests that they may have bred specific differences in their metabolism. Unfortunately, more precise information about their physiological responses and metabolic adaptations during acute exercise and training is absent.

Therefore, the objective of this study was to characterise the physiological and metabolic responses in Mangalarga-Machador horses after a field test, where they walk using one of their 2 typical gaits: marcha picada or marcha batida. It was hypothesised that there would be changes in the physiological and metabolic associated with a field test and that there would be differences due to the 2 types of marcha gaits.

Materials and methods

The Universidade Federal Rural de Pernambuco Institutional Animal Care and Use Committee approved all methods and procedures used in this study (protocol # 62/2007-CTA/DZ).


Thirteen fit Mangalarga-Machador horses (mean ± s.e. age 6.4 ± 0.8 years; weight 378.1 ± 8.1 kg; gaits: 9 horses with natural picada and 4 horses with natural batida) were assigned randomly to do the marcha simulation test (MST). Because it was not possible to test all animals at same time, they were divided randomly in 4 groups. Three groups had 3 horses and one group had 4, but all groups had animals with picada and batida gaits. All horses used their natural gait only.

Field simulation

The horses completed a marcha simulation test (MST) by walking their natural gait, at the same speed (∼3.2 m/s), for 30 min. During the test the horses carried a similar load (∼80 kg, rider plus saddle) while walking over a flat grass field (∼180 m length). All animals had a 5 min warm-up walk, followed by 30 min doing marcha followed by a 5 min walk to cool down. Horses were well familiarised with the exercise field. Climatic conditions (temperature and humidity) were measured in the shade during the 2 days of MST. Finally, fit marcha horse was defined when the heart rate was <64 beats/min at the end of 15 min of recovery, after 30 min at marcha training.

Body composition evaluation

Body composition was determined 5 days before the marcha field simulation test using rump fat thickness by ultrasound (Scanner 450, probe 5.0 MHz1) followed by characterisation of the percentage of body fat (Westervelt et al. 1976; Manso Filho et al. 2009b). The body mass was determined using an electronic scale2 (Toledo 5000).

Physiological parameters

Heart rate (HR) was measured for 1 min using a stethoscope. Respiratory rate (RR) was measured for 1 min by visual observation of chest movement. These parameters were recorded 5 times: at pretest (control/fasting), immediately after MST (T0), and 15 (T15), 30 (T30) and 240 (T240) min after MST.

Blood metabolites and enzymes

Samples of blood (20 ml) were collected by using needles, with minimal manual pressure on the left or right jugular vein, for the measurement of plasma glucose (GLUC), lactate (LACT), total plasma protein (TPP), alanine (ALA), glutamate (GLU) and glutamine (GLN) concentrations and packed cell volume (PCV), serum albumin (ALB), urea (UREA), creatinine (CREAT), total cholesterol (T-CHOL), triglycerides (TRYG), high-density cholesterol (HDL-CHOL), creatine kinase (CK), alanine aminotransferase (ALT) and aspartate aminotransferase (AST). Blood samples were collected from the left jugular vein and put into ice-chilled, vacuum tubes at pretest (control/fasting), immediately after MST (T0), and 15 (T15), 30 (T30) and 240 (T240) min after MST. These tubes contained either the anticoagulant fluoride/oxalate (for GLUC, LACT, TPP and ALB samples), or the anticoagulant sodium heparin (for ALA, GLU and GLN), or no anticoagulant (for UREA, CREAT, T-CHOL, HDL-CHOL, TRYG, CK, AST and ALT samples). On the day prior to the MST, food was withheld after completion of the afternoon meal (18.00 h) and horses did not receive grass or concentrate before or during the recovery period (4 h) of the MST, but they were allowed free access to water and mineral salt throughout.

Immediately after blood collection at pretest and at T0, T15, T30 and T240 min, one aliquot of blood from fluoride tubes was centrifuged (15 min at 1200 g) to obtain plasma, which was used immediately to measure plasma glucose (Accu-check Advantage II)3, lactate (Accu-trend)3, total protein (refractometry), while an additional aliquot was used to determine PCV (microhaematocrit method). These analyses were made in the laboratory at the horse farm. [LACT] was measured only at T0, T15 and T30 because their concentrations in pretest and T240 were lower than detection levels of the portable device used.

Plasma samples for amino acid analysis were prepared using method described by Manso Filho et al. (2009a), where heparinised plasma, obtained by centrifugation was deproteinised with an equal volume of perchloric acid 10%. The supernatant, isolated by centrifugation (15 min at 1200 g), was then neutralised with potassium hydroxide and neutralised samples were stored frozen (-20°C) until analysis. [GLN], [GLU] and [ALA] were analysed using enzyme based methods as previously described (Lund 1979). Serum samples were analysed for [ALB], [UREA], [CREAT], [T-CHOL], [HDL-CHOL], [TRYG], [CK], [ALT] and [AST] using different commercial kits (LABORLAB)4 and spectrophotometer (Genesys 20)5.

Statistical analysis

Results are expressed as means ± s.e. The data from physiological and metabolic results were analysed using repeated measures one-way ANOVA. A 2-way ANOVA with repeated measures for time and picada vs. batida horses were used to determine if there were main effects or interactions. Post hoc tests were performed using Tukey's test. The t test was used to compare body composition between horses with different marcha gaits. An a priori level of statistical significance was set at P<0.05 for all tests. SigmaStat 3.0 for Windows6 was used for all analyses.


The weather during the marcha simulation tests (MST) was uniform and calm. The temperature and air humidity averaged 27.9°C and 61.4% on the first day and 26.8°C and 79.3% on the second day. However, the temperature and air humidity evaluation were made in the shade and do not represent these values in the open field. Values of temperature and air humidity were within the normal variation for the time of the year at this location (latitude: 7.9742° south; longitude 34.9976° west; 109 m above sea level) in this tropical region of Brazil.

Evaluations of body composition showed that Mangalarga-Marchador horses used in this research had ∼13.4% of body fat. However, it was observed that there were no differences between picada and batida horses in their body mass (P = 0.44), fat-free mass (P = 0.32) and fat mass (P = 0.84) (Table 1).

Table 1. Body composition of experimental animals by type of marcha
Type of marchanBody mass (kg)*Fat-free mass (kg)*Fat mass (kg)*Fat (%)*Fat thickness (cm)*
  1. Observation: * P>0.05, t test.

Picada9373.7 ± 10.4322.5 ± 9.051.2 ± 2.513.7 ± 0.51.08 ± 0.10
Batida4388.0 ± 12.8337.8 ± 8,650.2 ± 2.712.8 ± 0.80.90 ± 0.17

Significant exercise-related changes were observed in the physiological and blood analyses in experimental horses for HR (P<0.001), RR (P<0.001), [GLUC] (P<0.001), [LACT] (P<0.001), [TPP] (P<0.001), PCV (P<0.001) and [GLN] (P<0.001) (Table 2). The other parameters did not change after MST. Finally, a significant interaction was observed between marcha type (picada vs. batida gait) and time of sampling for HR, RR, PCV and [LACT] (P<0.05) (Fig 1). Animals used during MST did not show any lameness or increase in [CK] and other enzymes in T240.

Table 2. Results of physiological and metabolic measured in blood from Mangalarga-Marchador before and after marcha simulation test
 Pre-exercise evaluationPost exercise evaluation
  1. Obs: different letters at same row represents P<0.05, by Tukey test. T0= immediately after MST, T15= 15 min after MST, T30= 30 min after MST and T240= 240 min after MST; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

Physiological parameters     
 Heart rate, beats/min46.07 ± 1.56C96.69 ± 3.81A54.69 ± 2.78B45.76 ± 2.18C42.84 ± 2.53C
 Respiratory rate, breaths/min21.92 ± 2.00C97.69 ± 5.42A49.23 ± 5.97B37.53 ± 4.90B22.69 ± 1.85C
 Glucose, mmol/l4.50 ± 0.17C5.46 ± 0.26A5.20 ± 0.19AB4.90 ± 0.13ABC4.67 ± 0.14BC
 Lactate, mmol/l2.63 ± 0.43A1.69 ± 0.31B1.13 ± 0.26C
 Packed cell volume, %31.15 ± 1.04C39.76 ± 0.71A33.38 ± 0.72B30.84 ± 0.69C34.61 ± 0.70B
 Total protein, g/l64.69 ± 0.84C71.07 ± 1.38A68.53 ± 1.48B67.53 ± 1.15B71.07 ± 0.99A
 Albumin, g/l27.23 ± 2.3730.53 ± 1.7124.61 ± 2.4526.30 ± 1.7027.15 ± 1.24
 Alanine, mmol/l323.22 ± 21.64320.53 ± 14.49296.27 ± 18.36309.31 ± 16.43296.92 ± 16.94
 Glutamate, mmol/l81.40 ± 8.60100.89 ± 9.1295.25 ± 9.09105.21 ± 9.60103.79 ± 9.25
 Glutamine, mmol/l512.99 ± 22.74A432.68 ± 17.74B426.63 ± 19.99B407.77 ± 15.01B389.42 ± 17.96B
 Urea, mmol/l4.14 ± 0.394.40 ± 0.294.36 ± 0.254.04 ± 0.274.17 ± 0.33
 Creatinine, µmol/l94.06 ± 8.4796.49 ± 8.95117.27 ± 10.9399.41 ± 7.99105.94 ± 6.94
 Total cholesterol, mmol/l1.77 ± 0.101.79 ± 0.121.80 ± 0.101.88 ± 0.081.89 ± 0.08
 HDL cholesterol, mmol/l1.10 ± 0.071.17 ± 0.061.08 ± 0.061.16 ± 0.051.27 ± 0.07
 Triglycerides, mmol/l0.20 ± 0.050.27 ± 0.020.31 ± 0.090.22 ± 0.030.29 ± 0.09
 Creatine kinase, U/l42.75 ± 8.5642.63 ± 8.1453.89 ± 17.8475.09 ± 11.2652.87 ± 7.60
 AST, U/l50.6 ± 5.0548.2 ± 5.2938.02 ± 4.7247.01 ± 5.4942.87 ± 3.77
 ALT, U/l5.22 ± 0.955.05 ± 1.046.65 ± 2.024.63 ± 1.324.93 ± 1.82
Figure 1.

Results of interaction between marcha type (picada vs. batida gait) and time of sampling for heart rate, respiratory rate, Packed cell volume and concentration of lactate.


The present paper presents new information that documents exercise-related induced changes in physiological and metabolic parameters in Brazilian 4 beat gaited Mangalarga-Marchador horses. The values of HR and RR were higher at T0 in these horses and then dropped rapidly to pre-exercise values after 15 min of recovery, showing that horses are fit and trained for this equine sport. Recently, Prates et al. (2009) showed that maximum HR in Mangalarga-Marchador horses during MST was ∼185 beats/min and after 20 min of recovery the HR dropped to ∼70 beats/min, but they did not measure RR. The RR returned to pretest values only 240 min after cessation of exercise, showing that horses have some difficulty in their thermoregulation process during their recovery period. Comparing data from marcha horses with other types of horse activities is not easy because of the relationship between the rider and animal during MST, when riders control horses and only allow them to move at their special 4 beat gait during the whole period without trotting or pacing. Regarding the relationships between rider and horse, Mottini et al. (2006) showed that when a trotting horse is mounted, the rider modifies some gait related variables, thus producing changes in the metabolic and physiological demand when compared with an unmounted trotting horse.

During the MST, significant changes (P<0.05) were observed in [GLUC] and [LACT], and when associated with heart and respiratory rates, should be used to characterise this equine sport with low accumulation of lactate in fit horses. [GLUC] was higher at T0 of MST contributing to the general energy availability but returned to pretest levels in the horses 30 min after the MST. The largest [LACT] measured was <3.0 mmol/l, showing that animals developed a typical low intensity and mid duration exercise with lower accumulation of lactate. Glucose and lactate results in this MST were measured with a portable clinical analyser that produced accurate measurements when values were above 0.7 mmol/l and, principally, when a PCV below 50% is used (Evans and Golland 1996). Lactate concentration pretest and at T240 were below detection levels for the device used in this research.

The lowest [TPP] was measured pretest and the highest immediately after the MST. Also, at the end of the 4 h recovery period, we observed another elevation in [TPP] (∼71 g/l) and PCV (∼35%) when compared with pretest levels (TPP: ∼65 g/l; PCV: ∼31%) (Table 2), indicating a substantial shift of body water during recovery since they had access to water and mineral salt inside the stalls. In endurance horses different degrees of dehydration are present when HR is >70 beats/min after a 30 min recovery period, which was not observed in marcha horses in this study. The term ‘involuntary dehydration’, which represents an incomplete voluntary replacement of body water fluids and electrolytes (Butudom et al. 2002) may explain changes in [TPP] observed in marcha horses at T240, which had access to water and salt but apparently had not drunk enough water or/and eaten salt to replace fluids and salts lost during exercise. Sampieri et al. (2006) observed that calculation of possible changes in plasma volume and dehydration should be associated with increases in [TPP] and [ALB] rather than [TPP] and PCV. In this experiment [ALB] did not change, but the measured changes in TPP and PCV are probably due to haemoconcentration after the MST and during the recovery period. Detection of some degree of dehydration at the recovery period (T240) represents one important finding in this study because some marcha horses take part in several competitions within a few days. Ideally, during this period, marcha horses would replace their body water and electrolyte deficits. The supplementation with electrolytes and water may improve horse recuperation after rides (Düsterdieck et al. 1999; Butudom et al. 2002; Sampieri et al. 2006).

Alanine and GLN may be released from skeletal muscle during exercise (Wagenmakers 1998) and contribute to the energetic metabolism. During MST changes in [GLN] measurement (P<0.05) were observed, with a large reduction of [GLN] after the MST when compared with concentration inthe pretest phase. A decrease in plasma [GLN] after MST, if persistent, may increase risk of infection or impair tissue repair (Routledge et al. 1999; Castell and Newsholme 2001), especially if the horse takes part in more than one competition in a few days. Results of plasma [GLN] pretest were higher than our previous observations ([GLN]∼330 mmol/l) but lower than glutamate concentration (∼120 mmol/l) in another experiment with mares (Manso Filho et al. 2008). Concentration of GLN in pretest represents basal levels and were not associated with overnight fasting, because fasting does not produce changes in blood amino acids concentration in horses (Russell et al. 1986; Routledge et al. 1999). In addition, during this research [ALA] was evaluated but there was no modification in [ALA] after the MST. In contrast to data from Standardbred horses, during submaximal exercise to fatigue where changes were observed in [GLN] and [ALA] (Pösöet al. 1991), MST did not produce changes in plasma [ALA].

Comparison of the 2 types of marcha (picada vs. batida) at same speed using a 2-way ANOVA, detected an interaction (P<0.05) marcha type and time (Fig 1). Mangalarga-Marchador with picada gait had large average of HR (102.5 ± 2.2 beats/min), RR (106.2 ± 3.6 breaths/min), PCV (40.4 ± 0.5%) and [LACT] (3.2 ± 0.1 mmol/l) when compared with batida gait (HR = 83.7 ± 3.3 beats/min; RR = 78.5 ± 5.4 breaths/min; PCV = 38.2 ± 0.7%; [LACT]= 1.3 ± 0.2 mmol/l) immediately after MST. Also, there is a tendency to interactions in [TPP] (P = 0.07) was observed. Those differences between picada and batida marcha horses were not expected but they showed that picada gait was not well adapted to the speed used in this study (∼3.2 m/s) probably because picada gait produces more friction/contact with the grass surface during locomotion and spending more energy. Different studies (Hoyt and Taylor 1981; Wickler et al. 2002) showed that horses have speeds where they spend less energy and, when their gaits are ‘shortened’ or ‘extended’, the cost of locomotion increased, as a curvilinear function; however, more studies need to be done in marcha breeds to characterise their energy expenditures and their real cost of transport. Finally, the difference between the number of picada and batida horses in this research reflect on the regional preference for picada horses.

The ideal speed for gaited Mangalarga-Marchador horses is between 3.0 and 4.0 m/s (Procópio 2003), where they preserve their typical 4 beat gaits; speeds >4.0 m/s produce elimination of the horses from the most important competitions. Actually both picada and batida horses compete together in most competitions in Brazil; however, if both groups of animals continue to be selected at the same speeds, batida horses will have an advantage over picada horses because apparently they are more adapted to the ideal speed variation. Finally, because picada horses spend more energy when compared with batida horses, they need special attention in their training and nutritional management, principally to improve their aerobic capacity.


The marcha simulation test produced significant changes in physiological and blood parameters similar to other forms of low intensity and mid duration exercise test, with some degree of dehydration during the recovery period. After the test it was determined that picada horses had larger cost of transport when compared with batida horses when they move at the same speed.


This work was supported financially by: the IRCA Nutrição Animal (Carpina-PE, Brazil), CNPq - National Council for Scientific and Technological Development, and Haras Cascatinha Horse Training Center (Camaragibe-PE, Brazil).

Conflicts of interest

None declared.

Manufacturers' addresses

1 Pie Medical, Maastricht, Holland.

2 Toledo Balanças Eletrônicas, Brazil.

3 Roche Diagnostic GmbH, Mannheim, Germany.

4 Laborlab, Guarulhos, São Paulo, Brazil.

5 Thermo Fisher Scientific Inc, Waltham, Massachusetts, USA.

6 Jandel Scientific, San Rafael, California. USA.