A prospective cohort study on the acute:chronic workload ratio in relation to injuries in high level eventing horses: A comprehensive 3-year study.

In human sport science, the acute:chronic workload (ACWR) ratio is used to monitor an athlete's preparedness for competition and to assess injury risks. The aim of this study was to investigate whether acute and chronic workload calculations for external and internal loads (e.g. high-speed work distance and associated exertional effort) were associated with injury risk in elite eventing horses and to identify workloads performed by horses competing in different competition and at different fitness levels. Training load and injury data were collected from 58 international eventing horses (CCI2*-CCI5* level) over 1-3 years. A total of 94 individual competition seasons were monitored. During this period, heart rate (HR; beat/min) and GPS data were collected of all their conditional training sessions and competitions. External load was determined as the distance (m) covered at high speed (HS1 ; velocity between 6.6 and 9.5 m/s), and sprint speed (SS2 ; velocity>9.5 m/s). Internal load was calculated for HS and SS, using individualized training impulses (TRIMP3 ;AU). For internal and external workload HS and SS the acute (1-week) and chronic (4-week) workloads were calculated and ACWR4 determined. The injury data in relation to ACWR was modelled with a multilevel logistic regression. Akaike's information criterion was used for model reduction. Sixty-four soft tissue injuries were registered from a total of 2300 training sessions and competitions. External and internal workload at HS and SS were significantly affected by the year and fitness level of horses. Competition level and year significantly affected the distances covered at SS. The ACWR of high-speed distance of the present week (OR; 0.133, 95 % CI; 0.032, 0.484) and the previous week (OR 3.951, 95 % CI; 1.390, 12.498) were significantly associated with injury risk. Competition level and chronic workload had no significant effect on injuries. In agreement with findings in human athletes, acute spikes of workload in eventing horses increased the risk of injury. Evaluation of horses' workload can be used to design and effectively monitor training programs and can help to improve equine welfare by reducing injury risk.

Exercise testing in Warmblood sport horses under field conditions.

Regular exercise testing in Warmblood sport horses may, as in racing, potentially help to characterise fitness indices in different disciplines and at various competition levels and assist in understanding when a horse is 'fit to compete'. In this review an overview is given of the current state of the art of exercise testing in the Olympic disciplines of eventing, show jumping and dressage, and areas for further development are defined. In event horses, a simple four-step incremental exercise test measuring heart rate (HR), lactate concentration (LA) and velocity (V) is most often used. In dressage and riding horses, a wide variety of exercise tests have been developed, including incremental exercise tests, indoor riding tests and lunging tests. In show jumping, the use of a five-step incremental exercise test and exercise tests evaluating technical skills and fatigue of the horse has been reported. The velocity at a plasma LA of 4 mmol/L (VLA4) and HR recovery during submaximal exercise intensity have been shown to be the best parameters in event horses for predicting performance and impending injuries. In riding horses, the fitness level of horses is also an important determinant of injuries. Implementation of regular exercise testing and monitoring of training sessions may have important added value in the assessment of performance ability and potential future injuries in Warmblood sport horses. However, there is an urgent need to standardise methodologies and outcome parameters in order to make results comparable.

A prospective study on a cohort of horses and ponies selected for participation in the European Eventing Championship: reasons for withdrawal and predictive value of fitness tests.

BACKGROUND: Eventing is generally recognized as a challenging equestrian discipline and wastage figures for this discipline are relatively high. There is a need for information that provides insight into the causes of wastage and withdrawal from competition, for animal welfare and economic reasons. The aim of the present investigation was to conduct a prospective study following the entire national selection of event horses (n = 20) and ponies (n = 9) in the Netherlands that prepared for the European Championship in 2010 (ponies) and 2011 (horses), noting causes of withdrawal and monitoring fitness using standardized exercise tests (SETs), with heart rate (HR; beats/min), speed (V; m/s) and plasma lactate concentrations (LA; mmol/L) as measured parameters. RESULTS: In SET-I, performed at the beginning of the season, horses (n = 17) had a mean VLA4 (V at LA 4 mmol/L) of 10.3 ± 0.4 m/s with a mean V200 (V at 200 beats/min) of 11.4 ± 0.8 m/s and ponies (n = 9) a mean VLA4 of 7.8 ± 0.9 m/s and V200 of 9.6 ± 0.7 m/s. Before SET-II, performed six weeks before the European Championship, 16/20 horses and 6/9 ponies were withdrawn. The most common reason for withdrawal was locomotor injury (9/16 horses, 4/6 ponies; P < 0.001 and P = 0.011, respectively). Other reasons included an animal 'not meeting the competition criteria' (4/16 horses, 2/6 ponies) and being sold (3/16 horses). Animals were divided on the basis of VLA4 and recovery-HR during SET-I into good and average performers. Average performers were significantly more likely to be injured (50.0%) than good performers (0%, P = 0.05). In a subpopulation of ten horses, in which all condition training sessions were evaluated for HR and speed, HRpeak was significantly lower in horses that stayed sound (186 ± 9 beats/min) compared with horses withdrawn from training and competition because of injury (201 ± 5 beats/min; P = 0.016). CONCLUSIONS: Of the national selection, 45% of all animals were unavailable for the European Championship because of locomotor injuries. Field tests were useful in assessing the potential injury risk, as individuals with better fitness indices (good performers) were less likely to become injured than average performers. Furthermore, monitoring of training sessions showed predictive value for future injuries, as horses withdrawn because of injury later on showed already higher peak HRs during condition training than horses that stayed sound. Therefore the increase in peak HR seemed to precede visible lameness in a horse.

Young Friesian horses show familial aggregation in fitness response to a 7-week performance test.

The aim of this study was to monitor the fitness level of young Friesian horses and to assess whether fitness data are predictive for final performance score and whether familial aggregation of response to training could be detected. Sixty-six young Friesian horses, the offspring of six different stallions (A, B, C, D, E and F), underwent a 7-week performance test. The horses were given a performance score for their ability for dressage (in weeks 5 and 7; 0-110 points) and were evaluated for fitness using standardised exercise tests (SETs) at the beginning (week 2, SET-I) and the end (week 6, SET-II) of the period. Heart rate (HR, beats/min) was measured in both SETs, and plasma lactate concentration (LA, mmol/L) was measured only in SET-II. Fitness of the horses improved moderately but significantly between SET-I and SET-II (P=0.015). There was a large heterogeneity in responsiveness to training; some horses were high responders, whereas others were non- or low responders. There was a familial aggregation of HR canter-1 response to training (P=0.039), while the HR of stallion C's offspring decreased significantly more than those of stallions A (P=0.09), D (P=0.013) and F (P=0.009). Horses that were reluctant to exercise did not differ in HR or LA concentrations compared to those that completed the SET, which may have been a sign of overreaching. HR had no predictive value for the performance score, but horses that did not reach the anaerobic threshold in SET-II scored significantly better (73.8 ± 5.6 points) than horses that did not (69.9 ± 5.9 points, P=0.025). The findings demonstrate for the first time in the horse a familial aggregation of HR response to training, as has been reported previously in humans. Familial aggregation suggests a genetic influence on the effect of training on fitness in horses. HR could not predict final performance score, but LA concentrations during SET-II were predictive.

A prospective study on fitness, workload and reasons for premature training ends and temporary training breaks in two groups of riding horses.

Little is known about wastage in riding horses and the factors like fitness and workload that may reduce injuries and maximise welfare. To evaluate fitness, workload and reasons for premature training ends (PTEs) and temporary training breaks (TTBs) during a nine week training period, two groups of riding horses were used: Group A consisting of 58 horses used for student equitation courses (32 with training prior to admission and 26 without) and Group B consisting of 26 horses owned by two riding schools (school-I and school-II). To assess fitness, all horses performed a standardised exercise test (SET) at the start (SET-I) and end of the training period (SET-II) measuring heart rate (HR bpm) and speed (m/s). In addition, all horses were monitored daily during the training period for their health and workload. In Group A, trained horses had significantly lower HRs in SET-I (P=0.05) compared to untrained horses and in SET-II, trained horses tended to have lower HRs than untrained horses, though this was not statistically significant (P=0.057). During the training period all horses received an identical workload. A total of 19.0% of Group A horses ended the training period prematurely for veterinary reasons (PTEV); of those untrained horses had earlier a PTEV in the training period (after 2.8 ± 1.3 weeks) than trained horses (after 4.1 ± 1.5 weeks, P=0.030). In Group B, school-I and school-II horses did not differ significantly in fitness level nor in workload. More school-II horses ended the training period prematurely for veterinary reasons (n=7; 70%) compared to school-I horses (n=4; 25%, P=0.032), although seven (63.6%) of these horses were still continuously used in riding lessons. In both groups (A and B), small injuries (without a temporary training break) were significantly associated with premature training ends for veterinary reasons later on: in Group A small injuries preceded 27.3% of the PTEVs (P=0.005) and in Group B small injuries preceded 54.5% of the PTEVs (P=0.030). In conclusion, as all horses in each subgroup had the same workload, the occurrence of PTEV seemed not associated with the workload. In Group A horses, level of fitness seems to be an important factor for the point in time injuries will occur during the training period. In all horses, injuries were more likely when a temporary training break was not taken following seemingly minor injuries. Since a lot of injured Group B horses were used in riding lessons against veterinary advice, this may indicate that riding school owners have different perception on welfare and if true this may cause serious welfare problems.

The influence of challenging objects and horse-rider matching on heart rate, heart rate variability and behavioural score in riding horses.

A good horse-rider 'match' is important in the context of equine welfare. To quantify the influence of repetition and horse-rider matching on the stress of horses encountering challenging objects, 16 Warmblood horses were ridden in a test-setting on three occasions. On each occasion the horse was ridden by a different rider and was challenged by three objects (A-C). Heart rate (HR), heart rate variability (HRV) of horse and rider, and behaviour score (BS) of the horse were obtained for each object and as a total for each test. The horse-rider interaction was evaluated with each combination and assessed as 'matching' or 'mismatching', and the horses were categorised as 'compliant', 'partly-compliant' or 'non-compliant'. Horses exhibited a decreased HR (P=0.015) and a decreased BS (P=0.004) within and across different tests. 'Matching' horse-rider combinations exhibited less stress as indicated by reduced HR ('match' 69±10 vs. 'mismatch' 72±9, P=0.001) and BS ('match' 1.9±1.1 vs. 'mismatch' 3.8±1.4, P=0.017) of the horse. 'Compliant' (68±8, P<0.001) and 'partly-compliant' (71±9, P=0.002) horses had significantly lower HR than 'non-compliant' (75±9) animals. The findings of the study indicate that HR and BS measurements support a subjective 'match' diagnosis and HR measurement may be a valuable tool in assessing horse compliance.