The effect of relaxing massage on heart rate and heart rate variability in purebred Arabian racehorses
Abstract
The objective of this study was to assess the effect of relaxing massage on the heart rate (HR) and heart rate variability (HRV) in young racehorses during their first racing season. In the study, 72 Purebred Arabian racehorses were included. The study was implemented during the full race season. The horses from control and experimental groups were included in regular race training 6 days a week. The horses from the experimental group were additionally subject to the relaxing massage 3 days a week during the whole study. HR and HRV were assumed as indicators of the emotional state of the horses. The measurements were taken six times, every 4-5 weeks. The HRV parameters were measured at rest, during grooming and saddling the horse and during warm-up walking under a rider. The changes of the parameters throughout the season suggest that the relaxing massage may be effectively used to make the racehorses more relaxed and calm. Moreover, the horses from the experimental group had better race performance records.
Introduction
The Purebred Arabian horses bred in Poland are routinely submitted to race training at the average age of 2.5 years. That procedure can involve unfavorable effects in young horses. The first stressors are loading and transport to the new training facility, new people (caretakers), new living conditions, feeding and so on (Janczarek & Kędzierski 2011a; Kędzierski et al. 2012). Moreover, keeping racehorses in boxes during a greater part of the day results in the restriction of free movement (Henderson 2007). In fact, the factors can negatively influence the physical performance of the horse and affect the horse's emotional state (Janczarek & Kędzierski 2011a,2011b). Emotional stress, due to mental overload, is one of the most important factors causing a decrease of performance and consequently, exclusion of horses from races (Evans 2007). Unfortunately, the generally accepted stressful training procedures cannot be eliminated, and then their effects should be mitigated (Evans 2007).
Thus, in practice it is a challenge to find and develop effective treatments which reduce such stress reactions. According to recent studies massage has turned out to be an interesting alternative treatment of not only musculoskeletal disorders but increasingly also a method of stress reduction in horses (Hinds et al. 2004; McBride et al. 2004; Scott & Swenson 2009). The authors have reported that a relaxing massage effectively decreases the concentration of catecholamines and cortisol, reduces arterial blood pressure as well as heart rate (HR) in horses. Moreover, the relaxing therapy has a significant effect on promoting behavioral changes, increasing the sense of well-being in horses (Hinds et al. 2004; Haussler 2009; Scott & Swenson 2009).
During the massage, the horse remains in close contact with the masseur. It is known that the horse response to humans depends on the horse's earlier experience. For example, horses which have been handled or trained in a gentle manner performed easier in grooming test or person test than those non-handled or trained using force (Jezierski et al. 1998; Baragli et al. 2009). As adults, experienced horses subjected to a temperament test, HR remained independent from human influence, whereas heart rate variability (HRV) was under strong influence of the human presence (König von Borstel et al. 2011).
One of the first symptoms of emotional disturbances in horses are changes in HR which reflect the influence of dynamic interactions and balance between vagal and sympathetic activity (Rietmann et al. 2004; Janczarek & Kędzierski 2011a). Different stressors can induce a shift of the autonomic balance toward either a sympathetic or a parasympathetic dominance. Therefore, it is difficult to assess the functional regulatory characteristics of the autonomic nervous system (ANS) only with simple measurement of HR (Visser et al. 2002; vonBorell et al. 2007). The analysis of HRV is recently considered as a much more accurate and detailed determination of the functional regulatory characteristics of ANS (vonBorell et al. 2007; Matsuura et al. 2010). Based on recent studies, the analysis of HRV has been suggested as an appropriate and, furthermore, easy-to-use method for the detection of imbalance between ANS function, training and recovery which may result in a decrease of maximal output (Cottin et al. 2006; Kinnunen et al. 2006; Matsuura et al. 2010; Munsters et al. 2013). We hypothesize that the horse's response to stress during race training can be effectively reduced by the introduction of relaxing massage. Thus, the aim of this study was to evaluate the effects of relaxing massage on HR and HRV in young racehorses, during their first competition season.
Material and Methods
Animal care and experimental procedures were in accordance with the European Committee regulations on Protection of Experimental Animals and were approved by the Local Ethics Review Committee for Animal Experiments of the University of Life Sciences in Lublin, Poland.
Horses
Seventy-two clinically sound Purebred Arabian racehorses were included in the study. The emotional state and behavior of the horses evaluated by a qualified trainer were normal. The study was conducted during 2 consecutive years and in each year, 36 horses were studied. Each horse participated in the study only in 1 year during its first racing season. At the beginning of the study, the horse age varied between 28 and 31 months. All the horses were brought to the Sluzewiec Horse Race Track (Warsaw, Poland) from their mother stud farms 3 months earlier to be able to acclimatize to the new conditions. The animals were randomly divided into two groups and the only criterion for the division was to form groups comprising an equal number of mares and stallions. The horses from control (C) group (n = 24, 12 mares and 12 stallions) and experimental (E) group (n = 48, 24 mares and 24 stallions) were included in regular race training 6 days a week. Additionally, the horses from E group were subject to the relaxing massage. During the whole experimental period the horses were housed in two stables under the same social and environmental conditions (Table 1). The horses were kept in clean and spacious boxes, which allowed them to lie down on comfortable dry bedding. The boxes were also large enough to allow the horses to move freely. All horses were fed according to dietary guidelines and cared for in a manner which was typical for racehorses. To reduce the influence of nonspecific factors on the animals’ emotional reactions, all the horses were fed the same diet and were handled by the same caretakers. All the caretakers were adult men with extensive experience in handling racehorses.
| Firtst year of the study | Second year of the study | |||
|---|---|---|---|---|
| Stable A | Stable B | Stable A | Stable B | |
| Control group | n = 6 | n = 6 | n = 6 | n = 6 |
| Experimental group | n = 14 | n = 10 | n = 12 | n = 12 |
Training routine
During the study, the horses were subject to classical training designed for racehorses. Training sessions of 1 h a day were conducted for 6 days a week. On the routine training day, the training session was performed in the morning, from 06.00 to 09.00 hours. After saddling, the horses were warmed up walking under the rider for approximately 10 min on a sand patch leading to the race track. The proper race training was conducted on the sand track with the speed and duration adapted to individual abilities and condition of the horse. Finally, the horses were brought walking to the stable, unsaddled and put into a horse walker for 45 min. In the weeks when the horses took part in a race, the training was modified. A day before and 2 days after the race, the horses were given 60 min exercise in the automatic horse walker.
At the end of both race seasons, the performance of the horses was estimated according to five parameters, based on official race records: (i) the number of races in which a horse took part; (ii) the number of races in which a horse won; (iii) winnings per race, that is sum of prizes won to the number of races a horse took part; (iv) success coefficient, that is sum of winnings won by a horse to sum of winnings won by all horses at the same age in a respective race season; and (v) general handicap, that is theoretical weight (in kg) a horse should carry in a race to equal the horse's chance of winning the race, with the chances of other horses at the same age.
Massage routine
Additional to the training routine, the horses from E group were subjected to a long-lasting, relaxing physiotherapy of their body. The massage was manual, hence it involved the human influence on the horse. Possible differentiation of the influence was minimized because the massage program was performed by only four masseurs, all well-trained specialists in the area of horse physiotherapy. On the consecutive days of the experiment, the masseurs worked with randomly chosen horses. The massage was applied for 25–30 min 3 days a week during the whole study. Prior to the introduction of the primary massage, the horses were accustomed to the technique for 1 week. The primary massage session included: (i) laying on of hands, which is not, in fact, a proper massage technique but is more a way of connecting with the horse; (2) friction, that is massage with a circular stroking movement applied across the horse's muscle fibers, instead of restricting it to the direction of the muscle, tendons and ligaments; (iii) petrissage, that is kneading some areas of the body to increase the circulation in the tissues, to help soften the horse's muscle tissue and to prepare it for deeper equine therapy movements; (v) shaking used for increasing local circulation and for improving effectiveness of stretching the horse's limbs; and (v) tapotement applied for stimulating the horses at the end of the massage session (Palmer 2012). Four body areas were treated by the massage: (i) neck area (M. trapezius, M. rhomboideus, M. splenius, M. serratus ventralisthoracis and M. brachiocephalicus); (ii) scapulas, forearms and back (M. supraspinatus, M. infraspinatus, M. triceps brachii and M. latissimus dorsi); (iii) buttocks (M. gluteus superficjalis, M. gluteus medius, M. biceps femoris, M. semitendinosus and M. semimembranosus); and (iv) hind limbs (M. extensor digitorumlongus and M. flexor digitorumsuperficjalis).
HR and HRV measurement
To collect the HRV data, a telemetric device was mounted on the horse's body. An elastic belt with a transmitter for telemetric HR registration (Polar RS800CX, Polar Electro Oy, Kempele, Finland) was put around the horse's chest. During the first week of the study, to acclimate the horses to the HRV measuring equipment, the elastic belt with non-active transmitter was mounted on each horse staying in its box for 5 to 15 min a day. After the acclimation time, first measurements were conducted to estimate the baseline values of HR and HRV parameters in each horse. The data were collected in five successive measurements, taken once in 4-5 week intervals. In the instance when the term of the measurement covered the term of the race, the examination was performed 1 week later. As a result, six measurements were collected in 25 weeks of the race season. Each day of the data collection, the telemetric sets were mounted early in the morning, before any routine horse handling activities. After activation of the set, the horses were left alone in boxes for about 10 min to achieve the resting values of HR and then the HR was recorded for 10 min to obtain the resting HR and HRV levels. Recording was continued during 10 min of grooming and saddling the horse and the next 10 min of warm-up walking. The experimental data was analyzed by the method described by Kuwahara et al. (1996) using Kubios HRV software (version 2.0, released 2008, Biomedical Signal Analysis Group, Department of Applied Physics, University of Kuopio, Finland).
- time domain analysis, including the square root of the mean of the squared successive differences in R-R intervals (RMSSD) and
- frequency domain analysis covering low frequency component (LF), high frequency component (HF), with low to high frequency ratio (LF/HF).
The magnitude of power was integrated in both the low-frequency band (LF: 0.01-0.07 Hz) and high-frequency (HF: 0.07-0.6 Hz). LF was used as a representative of sympathetic modulation activity, while HF and RMSSD were used as representatives of vagal modulation activity. LF/HF ratio was used as a reflection of sympatho-vagal balance.
Statistical methods
The results are presented as means with standard deviations (SD). In the statistical analyses, the SAS software package, version 9.1.3 (SAS Institute Inc., Cary NC, USA) was used. To determine the differences between HR and HRV parameters, the repeated measures design was applied (sigma-restricted model, active hypothesis decomposition), considering six successive measurements and the following factors: horse group (C and E), year of the study (first, second), horse sex (stallions, mares) as well as common interactions between the factors. The significance of differences between the mean values of the cardiac activity parameters and between race performance parameters was determined by Tukey's test. Statistical significance was accepted at the level of P < 0.05.
Results
Heart rate (HR)
Mean resting HR scores for C and E groups of horses are shown in Figure 1. The statistically significant difference between the groups was found in the sixth stage of the experiment (P < 0.05). Within both groups, HR parameters differed significantly between the first two stages and the next four stages of the study (P < 0.05).
Mean HR scores recorded during saddling of the horses are presented in Figure 2. Mean values of HR in C and E groups were statistically different (P < 0.05) from the third stage of the experiment. Interestingly, considering HR values within the horse groups in consecutive stages, the significant difference was noted only in the C group (P < 0.05).
The measurement of HR during the horse-walking activity revealed significant differences between C and E groups starting from the third stage of the experiment (P < 0.05). HR was lower in the E group compared to C during the whole experimental time. Moreover, in contrast to the C group, there was no statistically significant increase of HR within the E group (Fig. 3.).
Root mean square of successive differences (RMSSD)
Resting values of RMSSD in horses from the E group reached statistically higher values during the experimental time, regarding data obtained in the first stage (P < 0.05). Considering the successive measurements within groups, the first measurement significantly differed from the next stages (P < 0.05) (Fig. 1.).
Saddling activities significantly affected RMSSD values obtained in the E group from the beginning to the fourth stage of the experiment (P < 0.05). After that, the values of RMSSD did not differ significantly. There were also significant differences in RMSSD between the horse groups starting from the third stage of the experiment (Fig. 2.).
Mean RMSSD recorded during horse-walking significantly differed (P < 0.05) between C and E groups from the fourth stage of the experiment (Fig. 3). Considering the successive measurements, E group showed significant increase of RMSSD starting from the second measurement, whereas in the C group the fluctuation of this parameter reached statistical significance only at the third measurement.
Low frequency (LF)
Comparison of resting LF revealed similar values in horses from both groups, up to the third measurement (Fig. 1.). Since then, significantly higher (P < 0.05) LF was found in the C group. LF differed within E and C groups starting from the fifth and fourth stages of the experiment, respectively.
In response to the saddling, horses from the E group had significantly lower LF (P < 0.05) starting from the fourth measurement of the experiment. The parameter statistically decreased (P < 0.05) in the E group from the second measurement, whereas in C, the changes were less regular (Fig. 2).
During the horse-walking, significantly higher values (P < 0.05) of LF were found in the C group, beginning at the third stage of the experiment (Fig. 3). The intra-group differences were less pronounced.
High frequency (HF)
The resting analysis of HF revealed that physiotherapy applied to the horses significantly (P < 0.05) increased the parameter starting from the second stage of the study (Fig. 1.). A similar effect of the physiotherapy on HF was recorded during saddling and horse-walking (Figs 2 and 3, respectively).
LF/HF ratio
At rest, LF/HF ratio was significantly higher (P ˂ 0.05) in the C group compared to the E group, starting from the second stage of the experiment (Fig. 1). In the E group, the significant decrease in LF/HF ratio appeared between the third and fourth measurements. Within the C group, significant increase occurred after the third measurement.
LF/HF ratio obtained during saddling revealed that the horses subject to massage had significantly lower (P < 0.05) scores when compared with the controls. These substantial variances were identified between the third and sixth stages of the study.
The saddling activities increased LF/HF ratio even more than threefold (Fig. 2). The changes in LF/HF ratio in the E group of horses were inverse, although less marked.
During walking, statistically significant differences in the parameter also occurred between the groups, starting from the second month of the experiment (P < 0.05). Similar to saddling, walking particularly influenced LF/HF ratio in horses not subject to physiotherapy (Fig. 3).
Race performance parameters
The race performance parameters are presented in Table 2. The horses in E group achieved better results than those in C group with regard to the number of races won, winnings per race and success coefficient.
| Group of horses | Number of races | Number of races won | Winnings per race | Success coefficient | General handicap |
|---|---|---|---|---|---|
| Control | 5.06 ± 2.23 a | 0.54 ± 0.07 a | 299 ± 132 a | 0.81 ± 0.09 a | 55.9 ± 10.7 a |
| Experimental | 4.11 ± 1.82 a | 0.90 ± 0.15 b | 668 ± 91.9 b | 1.28 ± 0.20 b | 63.8 ± 11.3 a |
- a, b, means in columns marked with different letters differ significantly at P < 0.05. Winnings per race are given in Euros. Success coefficient, sum of winnings won by a horse to sum of winnings won by all horses at the same age in respective race seasons; General handicap, theoretical weight (in kg) a horse should carry in a race to equal the horse's chance to win the race, with the chances of other horses at the same age.
Discussion
The main focus of this study was to assess the influence of relaxing massage of the horse's body on ANS response evaluated by HR and HRV parameters. In this study, the cardiac parameters were analyzed during 7 months of race training procedures, at rest, saddling and warm-up walking, before proper race training on the sand track.
In our study, HR at rest in both groups decreased throughout the period of the experiment. Knowledge on the decrease of resting HR in horses during training is scarce (vonBorell et al. 2007). As it is generally accepted, physical activity is the most important HR-increasing factor; however, HR is also related to emotional stress (Kędzierski & Janczarek 2009). The data obtained in our study have shown that relaxing massage significantly decreased HR at rest, saddling and warm-up walking and the fact took place after 2 months of massaging the horses. Generally, HR depends on the balance between two branches of ANS: the sympathetic (SNS) and parasympathetic (PNS) nervous systems. Increased SNS or diminished PNS activity results in acceleration of HR. Conversely, low SNS activity or a high PNS activity causes deceleration of HR (Acharya et al. 2006).
During the horse's saddling, HR showed higher values than at rest. This result may be due to grooming (brushing, hoof cleaning etc.) and other activities related to saddling performed by caretakers in boxes. These activities are more or less stressful for horses (Janczarek & Kędzierski 2011a). HR in the C group significantly increased, starting from the scond stage of the experiment, while in the E group that parameter did not change during the entire experiment. Therefore, it can be suggested that relaxing massage plays a crucial role as an arousal stabilizer of trained horses. Our results are in line with observations on horses described by Łojek et al. (2013) and on humans as shown by Lindgren et al. (2010).
In this study, warming-up lasted approximately 10 min. During this time, the mean values of HR in C and E groups of horses reached the highest values in comparison to the discussed pre-training activities. Our results are similar to the finding recently described by Munsters et al. (2013). These authors proved that the normal horse walk increases HR even up to 106 bpm. This increase is due to changes both in physical activity and emotional reactions (Mastorakos et al. 2005). However, similarly to results from Munsters et al. (2013) and Mastorakos et al. (2005), we found that relaxing massage reduced walking HR during the entire experimental period, which was significant beginning from the third stage of the study.
While increase in HR is mainly caused by physical activity, changes in HRV also indicate a stress response (vonLewinski et al. 2013). Our analysis concerning the influence of relaxing massage on HRV parameters in horses was performed similar to the method described by Gehrke et al. (2011). Based on the detailed analysis of HRV, we could demonstrate ANS response to stress, independent from the effect of the physical activity (Wolff et al. 1997; Visser et al. 2002; Acharya et al. 2006).
RMSSD reports mainly parasympathetic activity (vonBorell et al. 2007; Becker-Brick et al. 2013). According to the recommendations of the European Cardiology Society, RMSSD is more useful for long-term HR recordings (longer than 5 min intervals). However, we decided to use this parameter in the study mainly because it may be better statistically interpreted than other time-domain parameters (Malik 1996). The results of RMSSD showed that relaxing massage strongly influenced this parameter at rest, during saddling and walking. While comparing mean RMSSD in groups, the significant increase of the parameter is pronounced in the E group. This result documents that relaxing massage increases PNS tone. In consequence, stress is reduced and the effectiveness of training is enhanced. This finding is in agreement with the study by Becker-Brick et al. (2013), Schmidt et al. (2010a,b) and Visser et al. (2002) and may suggest that the physical activity, as a separate factor, can reduce sympathetic tension. However, a previous study by Mohr et al. (2000) indicated that RMSSD component might be slightly related to the physical effort.
Hence, to eliminate a potential influence of exercise on PNS in the E group, HF component was also estimated. The results proved that the massage techniques used in this study significantly increased PNS activity in comparison to resting HF values in the E group. This is in agreement with the findings by Buttagat et al. (2011) and Diego and Field (2009) carried out on humans. Our simultaneous analysis of HF in C and E groups revealed only minor changes in that parameter during the entire experimental period. The fact shows that HF power is an indicator of parasympathetic tone independent of physical activity. In addition, our finding is in line with the study by Matsuura et al. (2010) who did not find any significant differences between HF in horses performing the same type of exercise.
To find the answer to the question whether relaxing massage affects sympathetic tone, LF component of HRV was analyzed. As is generally accepted, LF is strictly attributable to sympathetic activity and its increase reflects higher tension of SNS (Rietmann et al. 2004; vonBorell et al. 2007; Gehrke et al. 2011). In a stressful situation, the parasympathetic nervous activity decreases in favor of a higher sympathetic activity, enabling the animal to react in a biological manner, for example during the fight or flight decision (Kinnunen et al. 2006). The results of the present study demonstrate that the power values of the LF band are significantly higher in the C group starting from the second stage of the experiment. It is worth underlining that the basal values of LF were similar in horse groups and seemed to be dependent on the horse activity. Hence, the highest LF values were found during walking, the intermediate values during saddling and the lowest values at rest. These results are in agreement with earlier studies by Mohr et al. (2000) and Rietmann et al. (2004) who found that LF was strictly dependent on physical activity. Contrary to the controls, the massaged horses had significantly lower LF power component of HRV. That suggests that relaxing massage can be used to reduce SNS tension. Buttagat et al. (2011) reports there is a lack of clinical studies and the interpretation of LF power is still unclear. However, our results seem to be consistent enough to suggest that relaxing massage may be an effective anti-stress factor.
Many studies have demonstrated the usefulness of LF/HF ratio of the power spectrum as an indicator of sympathetic activity during a number of physical and/or psychological stresses (Miličević 2005; Billman 2013). The increase of LF/HF ratio is interpreted as a regulatory shift toward sympathetic dominance and a decrease of this index corresponds to the parasympathetic dominance (Rietmann et al. 2004; Matsuura et al. 2010). This hypothesis seems to be proven by the results of the present study. Namely, the sympatho-vagal balance expressed by LF/HF ratio was significantly elevated only in the C group. These results of non-massaged horses suggest that elevated LF/HF reflects the alteration of autonomic nervous activity from the rested to the stressed state, and that race training is a strong stressor. The finding is in agreement with other studies showing that physical effort increases the LF/HF component in exercised horses (Cottin et al. 2006; Matsuura et al. 2010). Contrary to the controls, the massaged horses in our experiment showed constant decrease of this ratio during the entire experiment. It should be highlighted that LF/HF ratio at the end of the study was significantly lower than at the beginning at rest, saddling and walking. The results are similar to those obtained in humans showing HF and LF/HF components in healthy volunteers subjected to light and moderate massage (Lindgren et al. 2010). The authors concluded that moderate massage increased parasympathetic nervous activity and decreased LF/HF ratio. The results indicate that relaxing massage both in humans and in horses can be used as a relaxation technique, preventing exercise-induced stress.
The benefits of relaxing massage were also noted in the race records. The performance of the horses from the E group was higher, which in turn had economic importance. This fact shows that applying relaxing massage is desired because of the positive effect on a racehorse's welfare and performance.
Conclusion
In conclusion, the present study shows that relaxing massage is profitable for the horse's psychical as well as physical state. This effect made the racehorses more relaxed and calm and in consequence resulted in better performance. Hence, we suggest that massage should be introduced to the routine care of racehorses. The massage may be beneficial not only to the horse's welfare but also can be a valuable tool for trainers in early prevention of overtraining syndrome. Further investigation assessing the practical use of the massage in horses involved in various equestrian disciplines would be warranted.
Acknowledgments
The study was supported by the National Centre for Research and Development, Poland (grant No. N180061).
