The characterization of the well-being of athletic horses has been determined through invasive and non-invasive methodologies. Recent approaches aim to better understand the processes involved in the acute and chronic stress of these animals through the evaluation of different biomarkers [1] - [5] . Equestrian sports that derive from rural activities of different countries, such as those developed in cattle farms, can benefit from such methodologies as they can provide new insights into metabolism during and after the physical activity in this large group of athletes.

In recent years, the frequency and intensity of training and competitions have significantly increased, which may compromise the well-being of athletic horses if the horses are not regularly evaluated. Vaquejada horses, for example, are trained 4 to 6 days a week for up to 60 minutes in many training centers [6] , because the trainers believe that this is the best way to prepare these horses for the championships. Moreover, in a vaquejada competition, horses must run at least five times to qualify for the final race, which takes place one or two days after the initial qualifications. Then they will run five or more times until the result is known. In these cases, regular evaluation of different blood biomarkers can serve as an indicator of metabolic and physiologic changes induced by the competitions and training, providing important information about the health and well-being of these animals. Thus, a regular assessment can contribute to athletic longevity and produce a positive impact on legislations and regulations related to equestrian operations and horses’ welfare.

Previous studies with vaquejada horses have reported that these equine athletes, when competing and training under official rules [7] [8] [9] , show transient changes in different blood biomarkers associated with stress, health, and well-being [3] [10] [11] . These variations in properly trained horses occur physiologically, with an acute increase in different biomarkers such as lactate, hematocrit, creatine kinase (CK), and cortisol, among others, with the recovery of baseline values in less than 24 hours. Such studies used several indicators and signals which can included in the different models for assessing equine welfare [4] .

More recently, the concomitant assessment of biomarkers of stress and inflammation has been indicated for the characterization of animal welfare, as they provide information regarding the effects of exercise on different tissues [12] [13] [14] . Thus, stress and inflammation indicators, such as cortisol, interleukin (IL)-6 and IL-1β, were proposed as a method to assess the animal behavior and possible impact, positive or negative, on different tissues. Additionally, acute phase proteins (APP) can also be used as inflammatory biomarkers, such as transferrin/iron, since APP are influenced by variations in IL-6 [15] [16] . Ferritin/Iron is a negative acute phase reactant, as its concentration is reduced in the acute phase of inflammation. In horses, the normal concentration for serum iron ranges from 80 to 277 µg/dL, but concentrations below 64 - 85 µg/dL are considered indicative of some degree of systemic inflammation [16] [17] [18] [19] . Another biomarker that can also be used is gamma-glutamyl transferase (GGT). This enzyme has been associated with overtraining and thus can be used to assess the health status of equine athletes [20] [21] [22] .

Thus, we hypothesized that vaquejada horses would present an increase in blood biomarkers following physical exercise which would return to baseline values after a period of recovery. For this purpose, the concentrations of blood cortisol, IL-6, IL-1β, iron, GGT, urea and creatinine before and sequentially after vaquejada simulation tests (VqST) were assessed. The same evaluation protocol was performed in a group of animals that was not submitted to VqST. Such information may contribute to the assessment of the welfare in equine athletes of this equestrian sport, particularly during regular training for vaquejada races.

All procedures were approved by the Committee for Ethics in the Use of Animals of Centro Universitário Cesmac (CEUA/CESMAC) under the registration number 01200702819/2016-97.

Animals and breeding system: Fourteen Quarter Horses (age: 4 - 8 years old; weight: ~450 Kg; body score: 5 - 6/9) were housed in individual stalls, measuring 15 m² and with good views of other horses, at a training center for vaquejada races (−7.973887, −35.021494), located in the Pernambuco’s Atlantic Forest zone, Brazil. All male horses used during this study were pull vaquejada horses and had not participated in a vaquejada race in the last 5 days. Vaquejada is a high-intensity and short-duration exercise in which helper horses (HH) are responsible to keep a bull running in a line while pull horses (PH) work to put the bull down after 100 m of running.

The 14 animals were randomly distributed into two groups: the exercise group (n = 10) and the control group (n = 4, without exercise). The exercise group underwent the vaquejada simulation test (VqST) as indicated in the literature [2] [23] , whilst the control group remained stabled and did not perform the exercise test.

Both groups of horse were fed approximately 20.0 kg of fresh chopped elephant grass (Pennisetum purpureum Schum.), which was divided into three daily meals (5 a.m., 2 p.m., 6 p.m.). Each animal also received an additional supplementation with 6.0 kg/day of concentrate, divided into three meals of ~2.0 kg/animal (6 a.m., 12 p.m., 5 p.m.) (MaxEquinos Trabalho Mix, CP: 14% (min); EE: 5% (min); FB: 10% (max); MM 10% (max); ADF 15% (max), DuRacho Nutrição Animal, Brazil). Salt and water were offered ad libitum to all horses.

All animals had the same level of physical conditioning and a similar training, which comprised of trot and gallop exercises, performed 3 to 4 times a week, for 30 to 40 minutes, on an official track for the sport. This exercise routine could be performed with or without running with cattle. At least twice a week, the animals were ridden for 60 minutes at walk in areas with varied topography.

Vaquejada simulation test:

The VqST was performed according to previous citations and in accordance with the official regulation [2] [7] [23] . In the exercise group, the test comprised of 10 minutes of warm-up, at trot and gallop, followed by three races with cattle on a 130m long sand track, lasting 15 to 20 seconds per race with 1-minute interval between them. In the end, the horses walked during 15 minutes for recovery. The races were held between 7 a.m. and 8 a.m. and the duration of the horse’s races was measured with a digital stopwatch. Cattle and horses were evaluated 24 hours before, during, and 24 hours after performing the VqST through the “Five Domains” system [24] . As previously mentioned, the control group animals did not perform the VqST and remained in the stables during this period.

Collection of blood samples:

Pre-test blood samples were collected with the animals in their stalls, following a fasting period of 8 hours. The remaining samples were collected immediately after the VqST, and at 1, 4, and 24 hours of recovery. For the animals of the control group, blood samples were obtained in the same moments except the phase immediately after the exercises. Blood samples were obtained by jugular venipuncture using negative pressure tubes with EDTA and tubes without anticoagulant, to obtain plasma and serum, respectively. Samples were then stored at −20˚C. The analysis of serum urea, creatinine, iron and GGT were performed with semi-automatic equipment (Doles D250, Doles Equipamentos Laboratoriais, São Paulo, Brazil) using commercial kits of the same brand. Cortisol, IL-6, and IL-1β were measured using ELISA in semi-automatic equipment (Bioclin® Minddray MR-96A, 837, Minas Gerais, Brazil) and the commercially available kits (Cortisol: Fine-EH0641-96T, Wuhan Fine Biotech Co., Wuhan, China; IL-1beta: BOST-EK0410-96SW, Boster Biological Tech, Pleasanton, California, USA; IL-6: BOST-EK0410-96SW, Boster Biological Tech, Pleasanton, California, USA).

Statistical Analysis:

The results were analyzed using One-way ANOVA and the Tukey test, to assess the possible influence of sampling time and to compare groups. In all cases, the significance level was set at 5% and the results are expressed as means ± mean standard error.

The analyzed biomarkers did not present significant variations between both groups of horses (p > 0.05) (Table 1). Also, no significant differences were found when comparing the two groups regardless of sampling moment (p > 0.05) (Table 2).

The speed of the vaquejada horses in the exercise group was set between 7.5 and 9.0 m/s and each race lasted on average of 12 to 20 seconds.

All animals of the exercised group were evaluated 24 hours after the races, and they were clinically healthy and showed no signs of lameness. In the animals of the control group, there were also no significant changes during the analyzed period.