J Appl Physiol 99: 1278-1285, 2005.
First published May 26, 2005; doi:10.1152/japplphysiol.01319.2004
8750-7587/05 $8.00
Left ventricular size and systolic function in Thoroughbred racehorses and their relationships to race performance
L. E. Young,1
K. Rogers,2 and
J. L. N. Wood3
1Centre for Equine Studies and 2Department of Epidemiology, Animal Health Trust, Newmarket, Suffolk; and 3Department of Veterinary Medicine, University of Cambridge, Cambridge, Cambridgeshire, United Kingdom
Submitted 23 November 2004
; accepted in final form 18 May 2005
 |
ABSTRACT
|
|---|
Cardiac morphology in human athletes is known to differ, depending on the sports-specific endurance component of their events, whereas anecdotes abound about superlative athletes with large hearts. As the heart determines stroke volume and maximum O2 uptake in mammals, we undertook a study to test the hypothesis that the morphology of the equine heart would differ between trained horses, depending on race type, and that left ventricular size would be greatest in elite performers. Echocardiography was performed in 482 race-fit Thoroughbreds engaged in either flat (1,000–2,500 m) or jump racing (3,200–6,400 m). Body weight and sex-adjusted measures of left ventricular size were largest in horses engaged in jump racing over fixed fences, compared with horses running shorter distances on the flat (range 8–16%). The observed differences in cardiac morphologies suggest that subtle differences in training and competition result in cardiac adaptations that are appropriate to the endurance component of the horses' event. Derived left ventricular mass was strongly associated with published rating (quality) in horses racing over longer distances in jump races (P 
0.001), but less so for horses in flat races. Rather, left ventricular ejection fraction and left ventricular mass combined were positively associated with race rating in older flat racehorses running over sprint (<1,408 m) and longer distances (>1,408 m), explaining 25–35% of overall variation in performance, as well as being closely associated with performance in longer races over jumps (23%). These data provide the first direct evidence that cardiac size influences athletic performance in a group of mammalian running athletes.
Thoroughbred horses; athletic performance; heart size
RESULTS FROM EXTENSIVE RESEARCH into the effects of athletic training and competition on the human heart have demonstrated differing effects of various types of exercise and training (22). Prolonged periods of high cardiac output with minimal change in arterial blood pressure, typical of long-duration endurance training (dynamic exercise), produce increased diastolic load and stimulate compensatory increases in cardiac chamber size (10). In contrast, short-duration "power" training (static exercise), associated with increased cardiac afterload and arterial blood pressure, produces increased left ventricular (LV) wall thickness due to compensatory concentric myocardial remodeling (15). Although sport-specific cardiac adaptations have been studied for many human athletic disciplines with differing static and dynamic exercise components (9, 19), most studies have compared elite human athletes competing in sports at extremes of the power/endurance spectrum. The differences in cardiac adaptation between athletes performing dynamic exercise over different distances, with subtle differences in the power/endurance components of their competitive events, have not yet been fully explored.
The role of the heart in defining athletic performance has been the subject of speculation and interest in human and equine sports medicine for many years. Postmortem examination of superlative athletes, both equine and human, have revealed substantially larger hearts than would have been predicted from lean body mass (20). The Fick principle and the dependence of cardiac output and thence O2 uptake (
O2) on stroke volume suggest that a relationship between relative heart size and athletic performance is likely (20), particularly as success in endurance competitions is inextricably linked to the ability of an individual to support a high O2 over prolonged periods (1, 2, 16). Nevertheless, despite a body of evidence supporting the concept that heart size and athletic performance are related, this hypothesis has not yet satisfactorily been proven in a population of athletes with a wide range of abilities.
As all Thoroughbred racehorses have independent objective measurements of their athletic performances readily available, they comprise a large population of highly trained running athletes that can be used to provide unique insights into the relationship between physiological variables and performance. In particular, the homogeneity of racehorse management and their closed breeding population should allow unique insights into the subtle differences in cardiac adaptations that might be associated with small differences in training horses for different race disciplines. A study was, therefore, undertaken to test the hypothesis that the morphology of the equine heart would differ between trained equine athletes, depending on their primary race type and that LV size would be larger in Thoroughbred horses racing successfully.
|
MATERIALS AND METHODS
|
|---|
Subjects.
This noninvasive epidemiological study of commercial racehorses fell out with the UK "Animals (scientific procedures) Act" and was carried out with informed consent from the trainers. The project was approved by the veterinary committee of the Horserace Betting Levy Board and management committee of the Animal Health Trust. Data were derived from 483 race-fit Thoroughbreds in a convenience sample of 5 flat and 4 National Hunt (NH) commercial training establishments. In the UK, there are two discrete types of Thoroughbred races: NH races are run over hurdles, or fences, over distances that always exceed 3,200 m. NH horses generally have a longer racing career than those running on the flat, begin racing later, at 3–5 yr of age, and reach their athletic peak between 6 and 11 yr of age. In contrast, horses that perform over shorter flat distances from 1,000–2,500 m may start racing at 2 yr of age, but the proportion remaining in training after 4 yr is very small. Although some horses, often males with little breeding value, move into NH racing having first raced on the flat, some are specifically bred for jump racing and do not undergo any athletic training until they reach 3 or 4 yr of age. The UK rules of racing allow horses that have not raced on the flat to compete in up to three NH flat races (also referred to as bumpers); these are typically 2-mile (3,200 m) competitions that provide racecourse experience before horses begin racing over jumps.
Cardiac data.
Cardiac data were derived from echocardiographic examinations taken at rest from the horses at their commercial training establishments. Data from less-fit horses and from horses that did not race were excluded from these statistical analyses. Data were also excluded from horses affected by audible murmurs of mitral or aortic valve regurgitation. In all cases, the diagnosis was later confirmed by color flow Doppler echocardiography (27).
The flat data set comprised data from 178 cardiac examinations on 125 horses made between 29 June 1998 and 31 October 2001 (Table 1). Observations were recorded from animals with, at the time of observation, a mean age of 2.6 (± 0.8) yr, weighing between 374 and 551 kg (mean 453 ± 30 kg). The NH data set comprised data from 536 examinations on 358 horses made between 2 August 1999 and 2 May 2003, recorded from 6 colts, 490 geldings, and 40 fillies or mares, aged at the time of observation between 3 and 13 yr (mean 6.5 ± 2.0 yr) and weighing between 419 and 586 kg (mean 493 ± 29 kg).
Echocardiographic methods.
A Vingmed system V ultrasound system (GE Ultrasound, Bedford, UK) and 2.5-MHz phased-array transducer were used for each echocardiographic examination. Two-dimensional (2D) and M-mode imaging was conducted at an imaging depth of 30 cm from the right hemithorax. The ultrasound probe transmission frequency was reduced to 1.5 MHz during each examination to enable the echocardiograph's second harmonic imaging mode. Use of second harmonic frequencies allowed image quality at the necessary depth of 30 cm to be optimized. An M-mode echocardiograph was obtained from a short-axis view of the LV at chordal level (23). Images were only recorded when heart rate was consistently <40 beats/min.
All 2D and M-mode data were digitized and stored on optical disks. Measurements were performed from the stored images by using a commercial software package (ECHOPAC, GE Ultrasound, Bedford, UK). Five cardiac cycles were measured from 2D and M-mode images, and an average value was obtained. A single experienced observer (LY) measured all data using manual planimetry and electronic calipers within the analysis system. Short-axis area of the LV at end diastole (SA area) was measured from at least two cine loops of cardiac images saved from chordal short-axis images obtained from two separately derived image sets. Each cine loop stored two to three heart cycles at the frame rate of 55 frames per second and imaging depth of 30 cm used in this study. M-mode data were derived from one or two M-mode images containing at least five complete cardiac cycles; the echocardiographer judged in real time the quality of the M-mode images being derived. The echocardiograph system used for this study constantly updated the 2D image during acquisition of M-mode images, so that the cursor position could be adjusted for optimal M-mode imaging.
LV internal diameter, interventricular septal, and LV free wall thickness in systole (LVIDs, IVSs, LVFWs, respectively) and diastole (LVIDd, IVSd, LVFWd, respectively) were measured from the stored images. The reliability and repeatability of measurements made by 2D and M-mode echocardiography had previously been determined for the echocardiographer performing the cardiac data acquisition (26).
An estimate of calculated LV muscle mass (LV mass) was obtained by using the formula of Devereux and Reichek (4):
This method is widely used in human patients, and it provided an acceptable (0.8) correlation to actual LV mass in an echocardiographic and postmortem study in horses (14). Mean LV wall thickness (MWT) in diastole, LV fractional shortening (FS), and LV ejection fraction (EF) were also calculated (21):
 |
 |
 |
Performance data.
Performance data were derived from two sources. Race type-specific lifetime performance indexes (average prize money per start, win-to-run ratio, and highest official rating attained) were derived for each horse from an electronic data set containing every race start made by all horses racing in Britain and overseas between 1 January 1997 and 30 June 2003 (Raceform Interactive, Raceform Limited, Newbury, UK). An official rating is given to each horse after it has run three times in any race discipline. An official handicapper employed by the race regulator assigns each rating. This numeric estimate of the horse's quality is calculated from its previous performances and the quality of its opposition. Ratings are specific for each race discipline, and the value is used to determine the weight carried by the horse in handicap races. Ratings are then continually adjusted for each race discipline throughout the horse's career, depending on all of its subsequent performances. Due to methods used in assigning ratings and to the patterns of racing over different distances undertaken by horses of different ages on the flat and over jumps, separate analyses were performed for 2-yr-old horses, older flat horses racing in "sprint" (<1,408 m) races, older horses in longer races (>1,408 m), novice NH horses racing on the flat in races of 3,200 m (bumpers), and NH horses of all ages running over flexible hurdles (hurdlers) and over solid fences (steeplechasers). For each racehorse, summary race type-specific indexes of performance were derived from their races in each of these categories during any year when ultrasonography had been performed. Annually published and adjusted Timeform ratings (Timeform, Northgate, Halifax, UK), estimated at the end of each season, were used as a measure of the overall quality of each horse on any year that ultrasound data had been obtained.
Data management and statistical analyses.
Weight-corrected variables were derived for each echocardiographic measurement or estimate from the residual values from linear regressions of weight on each variable within each data set. Following careful checking for internal consistency, the cardiac and performance data sets were merged using the unique, internationally registered official name given to each racing Thoroughbred. For flat data sets, merging was also restricted to the summary performance indexes made in each year for which there were cardiac and race performance data, when the horse was <6 yr of age. The very few (<5) observations made on older horses were not included in statistical comparisons of performance and heart size. For lifetime performance outcomes (including average prize money won per start, ratio of number of wins to race starts made, and highest official rating attained in that race type), only data from the last clinical examination at race fitness were considered. For annually varying Timeform ratings, analyses used data from each year in which echocardiographic data at race fitness had been collected. Data from one horse were not included due to problems ascertaining its precise identity. Horses could, if appropriate data were available from any year, be represented in more than one data set.
Analysis of variance models were used to estimate and compare least square means of race type-specific heart size parameters, adjusted by horse weight and sex. Pairwise statistical comparisons were made following Bonferroni correction for multiple comparisons.
Generalized linear regression modeling was used to assess the association between cardiac variables and lifetime performance indexes. Analyses were undertaken separately for each race type and for each performance outcome. For each outcome, the univariable association between dependent and independent variables was assessed, and then variables associated (P < 0.3) with the outcome were tested in least squares multivariable linear regression models. For the annual assessment of performance (Timeform rating), where there were repeated observations from individual horses, maximum likelihood mixed-effects linear regression models were developed, assuming a compound symmetry covariance structure. Variables significantly associated (P 0.05) with the dependent variable were retained in models. Quadratic terms and interaction terms between variables retained in models were also tested for inclusion. Plots of ranked normalized scores against residuals were examined to check assumptions of linear regression models (5). When plots were nonlinear, dependent variables were normalized, models were refit, and the same assumption was rechecked.
All statistical analyses were undertaken using SAS version 8.02 (SAS Institute, Cary, NC). PROC GLM was used for fitting linear regression models, and PROC MIXED was used for fitting mixed-effects linear regression models.
|
RESULTS
|
|---|
Descriptive analyses and comparison of heart size between race types.
Descriptive summaries of the flat and NH data sets are shown in Table 1. The mean values for the different heart size parameters are shown in Table 2. Generally, the heart size increased from 2-yr-old horses, through sprint and longer distance flat horses, with "bumpers" falling between the flat groups and the hurdle and steeplechase animals, with the latter being the largest in all parameters except IVSs. However, the size of the horses, as assessed by their weight, also increased from 2-yr-old horses through to sprinters and longer distance flat animals; mean weight was intermediate in hurdlers and greatest in bumpers and steeplechasers. Weight differences were significantly lower in all three flat-race groups than all three jump-race groups; the mean weight of hurdlers was significantly less than that of steeplechasers and bumpers, which had the same mean value. Differences in cardiac parameters were thus larger when unadjusted than when weight corrected (Table 2).
Weight-corrected parameters were also compared, and steeplechasers generally had the largest adjusted mean heart size parameters, with hurdlers falling between steeplechasers and bumpers, but the differences between bumpers and flat horses were much reduced or absent. Specifically, the differences in weight-corrected measures were statistically significant between steeplechasers and all other race types for LVIDs and LV mass. The differences were statistically significant between steeplechasers and all other race types except hurdlers for LVIDd (P = 0.08), IVSd (hurdlers, P = 0.16), and SA area (hurdlers, P = 0.21). Unadjusted and weight-corrected values for IVSs, LVFWd, and MWT were not significantly different between any of the groups. In contrast, unadjusted and weight-corrected EF values were significantly higher in 2 yr olds than in the three jump race types and higher in long distance flat racers compared with steeplechasers, but they were not significantly different between sprinters and any other group. EF was lowest in the steeplechasers. Weight-corrected RWT was lowest in the steeplechasers and hurdlers and highest in 2 yr olds, but the differences were only significant between steeplechasers and 2 yr olds and marginally significant between hurdlers and 2 yr olds (P = 0.07).
Male horses had significantly larger weight-adjusted heart size parameters for LV mass and LVIDd, but not for LVIDs, IVSd, IVSs, RWT, MWT, LVFWd, EF, or SA area. There was little effect of sex on the comparison between race types of weight-corrected parameter values or on their estimated (least square) mean values, indicating that sex did not explain the significant differences between steeplechasers and the other groups, nor any of the other effects noted.
Within the flat data set, the weight-corrected differences between race types were not significant other than for LV mass, which was smaller in 2-yr-old horses compared with older horses in longer races (P = 0.05). When sex was considered, however, this difference was no longer significant.
The effects of age on body weight and cardiac size were also explored. There was no significant effect of age on body weight or cardiac size in NH horses, and an average increase in body weight of only 15 kg occurred between the 2-yr-old flat horses and their older peers. The average difference in body weight of 
40 kg between flat and NH horses (Table 2) was also independent of age.
Association of cardiac parameters with performance in flat races.
The univariable associations of LV mass, the summary variable of principle interest, with performance in flat races are shown in Table 3. LV mass was not significantly (as shown by P values >0.05) associated with any index of race performance in 2-yr-old races. It was positively associated with Timeform rating in sprint races and was associated with, or nearly associated (0.1 > P > 0.05) with, average prize money per start, highest official rating, and Timeform rating in longer races. The positive regression coefficient in all except one of the LV mass comparisons with performance in older flat horses indicated that mean heart size tended to be larger in better performing horses. Nevertheless, for the most part, these relationships failed to attain statistical significance (P < 0.05).
In multivariable analyses of 2-yr-old race performances, no cardiac parameters retained in models were positively associated with any outcome (i.e., their regression coefficients were negative), with the sole exception of weight-corrected IVSd and win-to-run ratio (Table 4). No variables were significantly associated with Timeform rating in 2-yr-old races.
View this table:
[in this window]
[in a new window]
|
Table 4. Fixed-effect parameters and model R2 estimates from final multivariable models of performance in flat races
|
|
Performance of older horses in longer flat races was significantly and positively associated with one or more measurement of LV size for all of the four indexes of racing performance used (Table 4). For highest official rating, IVSs and LV EF were significant and explained 26% of the total variation in rating (Table 4; model R2 = 0.26). For both average prize money per start and win-to-run ratio, weight-corrected LVFWs was the only variable that was significant, and the model R2 for these outcomes was lower (0.06 and 0.08). For Timeform rating, LV EF and LV mass were significant. The different class of model for Timeform ratings precluded estimation of comparable R2 values.
In sprint races, performance of older horses was significantly and positively associated with LV EF and LV mass when highest official rating attained was considered (Table 4). These variables explained 35% of the total variation in highest rating (model R2 = 0.35). For average prize money per start, septal thickness in peak systole squared was significant, but, as for the longer races, model R2 was lower for this outcome. For Timeform rating, as for longer races, LV EF and LV mass were significant. There were no variables significantly associated with win-to-run ratio.
Association of cardiac parameters with performance in NH races.
The univariable associations of LV mass, the summary variable of principal interest with NH race performance are shown in Table 5. LV mass was positively (as adjudged by the value of the regression coefficients) and significantly (as shown by the P value) associated with all indexes of race performance in steeplechases, significantly associated (although slightly less strongly) in hurdle races, and was marginally associated with win-to-run ratio in bumpers. LV mass was significantly associated with Timeform rating in bumpers, but the strength of association, as adjudged by the parameter estimate, was smaller still than that in hurdle races. Official ratings are not provided for horses racing in bumpers.
In multivariable analyses, variables associated with racing performance in race-fit steeplechasers varied between the different performance variables used (Table 6), but every model included LV mass or a quadratic form of it. For both (normalized) average prize money won per start and win-to-run ratio, LV mass and EF were significant (R2 = 0.18 and 0.12, respectively). For normalized highest official rating, LV mass2, EF, and IVSs were significant (R2 = 0.23). For Timeform rating, weight-corrected LV mass alone was significant.
View this table:
[in this window]
[in a new window]
|
Table 6. Fixed-effect parameters and model R2 estimates from final multivariable models of performance in National Hunt races
|
|
Similar variables were associated with racing performance in multivariable models in race-fit hurdlers (Table 6). LV mass alone was significant in models of highest official rating. LVIDs was the only significant variable in the model of normalized average prize money per start, although LV mass was marginally significant (P = 0.06). For normalized win-to-run ratio, LVIDs2 and IVSs2 were significant (R2 = 0.11). In models of Timeform rating, LV mass and IVSd were significant. Generally, model R2 values were smaller in hurdle races than in steeplechases and flat races.
In bumper races, only LV mass was significant in models of Timeform rating and win-to-run ratio, no cardiac variable was significantly associated with average prize money won, and only LVIDs was significant for win-to-run ratio.
|
DISCUSSION
|
|---|
These results have demonstrated strong associations between objective measures of performance and LV dimensions in a large population of conditioned equine athletes. We have also shown that absolute and relative internal cardiac dimensions of equine athletes are affected by race discipline.
The differences in the weight-corrected diastolic LV dimensions LVIDd and SA area of 1 cm (8%) and 16 cm2 (16%) between race-fit 2-yr-old flat horses and seasoned NH steeplechasers are of particular interest, as they are not dissimilar in relative magnitude to that reported between sedentary humans and competitive athletes (13). These differences in chamber dimensions were consistently present between the long-distance steeplechasers and all of the other groups of racehorses studied, except for the hurdlers, who run over the same distance range, but over lower, easier fences. As LV chamber width increases in response to dynamic exercise and endurance training in both humans (8, 11) and horses (23), these data suggest that conditioned racehorses develop a cardiac morphology that is appropriate to the endurance component of their event. Previous data have also shown that racehorses adapt to commercial training with increases in both wall thickness and chamber width (23), yet we were unable to detect differences in weight-corrected wall thickness in any of the groups in the present study. This adds further weight to the assertion that the endurance component of training and racing is most likely to have influenced cardiac morphology in this group of athletes.
The magnitudes of the differences in cardiac dimensions within this group of running athletes are slightly less in percentage terms than those previously described for professional human athletes at extremes of the power and endurance spectrum (18). Nevertheless, the equine data are of interest, as differences in anthropomorphic characteristics, sex, and lifestyle have only rarely been considered in comparisons of data derived from humans (3, 9), yet, in this data set, the differences in cardiac morphology persisted after these confounding effects had been considered, even though the endurance components of the events were only subtly different. Our data also support the conclusion from human studies that sex exerts independent effects on the morphology of the LV (18) and add weight to the assertion that this effect is independent of training, body weight, or lifestyle (12). However, in contrast to previous studies from humans (6, 9), we were unable to demonstrate an independent effect of age on body weight, and thence cardiac morphology, in our group of equine athletes. It seems likely that this difference reflects the relatively short competitive career of the commercial Thoroughbred racehorse or is related to the absence of acquired cardiac, vascular, or hypertensive disease in this group of mammalian athletes (24).
In summary, these data have allowed differences in cardiac morphology to be revealed in a group of conditioned running athletes, despite relatively subtle differences in the expected endurance demands of the exercise event. It seems likely that this has been made possible because of the homogenous character of the group of athletes being studied. The extensive training and exercise data in humans have been derived from two principle sources, heterogeneous groups of naive humans subjected to exercise training, or highly motivated athletes from heterogeneous cultural and training backgrounds, who are then compared with age-matched sedentary controls. As a result, the effects of genetic and phenotypic variation on cardiac morphology are difficult to disentangle (7, 17). In contrast, each Thoroughbred racehorse is derived from a closed population that, from the early 18th century, has been selectively bred for speed and whose lineage can be derived back to one of three founding stallions. These population characteristics should markedly reduce the inherent genetic and phenotypic variation compared with that inherent in most human studies. Additionally, inevitable difficulties in controlling all aspects of the human athlete's environment, training, nutrition, and lifestyle, are also minimized in epidemiological studies of the equine population, as groups of 100 Thoroughbreds or more are collectively housed, managed, and trained under almost identical conditions within each training establishment.
Nevertheless, although our results support the hypothesis that small changes in the endurance component of athletic training might result in differences in cardiac morphology within a homogenous group of mammalian running athletes, other additional influences must still be considered. First, athletic superiority could be present in horses with a specific cardiac phenotype that differs between race types. As horse trainers tend to select horses for each race discipline based on their pedigree and previous race performances, it is possible that, in so doing, they are inadvertently selecting horses with a particular cardiac phenotype. As a result, there may still be an influence of genetics in determining the cardiac phenotype of each group, despite the horses being derived from a relatively homogeneous genetic pool. Nevertheless, our results do suggest that, within a group of running athletes competing over a relatively limited distance range of 1,000–6,400 m in races lasting between 1 and 10 min, the morphology of the equine heart differs in a way that is appropriate for the likely endurance component of the horse's principle race type.
The results also demonstrated an association between athletic ability and measurements of LV size and function, measured by echocardiography in commercial Thoroughbred racehorses in the UK. These relationships differed substantially between the different groups of Thoroughbreds studied, a finding that probably also reflects the altered importance of aerobic capacity in longer distance events. Our data showed that the derived variable, LV mass, had a significant relationship with performance in races over 1,408 m in most groups of racehorses studied. LV mass was also previously shown to have the strongest correlation (R2 = 0.79, P < 0.0001) to maximum O2 in horses exercising on a high-speed treadmill (25). As the calculation of LV mass incorporates both LV wall thickness and diastolic chamber dimension, this suggests that both characteristics are important in the determination of both racing success and maximum O2 in commercial racehorses. This supposition is indirectly supported by the data showing that commercial racehorse training results in increases in MWT, LV mass, and LVIDd (23), the same adaptive eccentric hypertrophy that is typical of human athletes engaged in sports with power and endurance elements (9, 19). The relationship between LV mass and performance was most consistent for the Timeform rating regressions. This can most likely be explained because a rating value is a better measure of racehorse ability than the other race performance outcomes used in this study. Earnings, finishing position, and win-to-run ratios are dependent on an individual's performance relative to the other horses in each race: an average horse can have excellent racing results and moderately high earnings if it runs many times successfully against poor quality horses. Despite this, the official and Timeform ratings will still reflect actual ability compared with the population as a whole.
There was a positive association between LV EF and rating (official and Timeform) in older flat (sprinters and longer distance animals) and steeplechasers, generally when combined with LV mass, but not in horses in their first year of training (2 yr olds and bumpers) and hurdlers. The R2 values derived, 35% in multivariable analysis for older sprinters, for example, represent a surprisingly high predictive value for a variable like athletic performance that has many contributing factors. The suggestion that LV EF may be an index of cardiac function that is also related to inherent athletic ability in equine athletes is novel. To our knowledge, resting EF has not been related to measures of performance in human power/endurance athletes, but such an association is plausible and represents a possible area for future investigation.
In conclusion, after consideration of the independent factors that govern cardiac size, these results have demonstrated positive associations between cardiac morphology and function and equine race performance. The different distances over which the horses raced and trained were also associated with subtle differences in cardiac morphology, consistent with the anticipated adaptations to altered endurance demands of their race disciplines. Study of this unique population of nonhuman running athletes has also allowed a relationship between athletic performance and heart size and function to be established.
|
GRANTS
|
|---|
This study was funded by a Veterinary project grant of the Horserace Betting Levy Board. The echocardiograph was funded in part by a generous donation from the Elise Pilkington Charitable Trust.
|
ACKNOWLEDGMENTS
|
|---|
The cooperation of the following racehorse trainers and breeders, without whom this study would not have been possible, is also gratefully acknowledged: Ed Dunlop, William Haggas, Sir Michael Stoute, Sir Mark Prescott, Cheveley Park Stud, Paul Nicholls, Mark Pitman, Graham McCourt, and Venetia Williams.
|
FOOTNOTES
|
|---|
Address for reprint requests and other correspondence: L. Young, Animal Health Trust, Lanwades Park, Newmarket, Suffolk, UK (E-mail: lesley.young{at}aht.org.uk)
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES
|
|---|
- Al-Hazzaa HM and Chukwuemeka AC. Echocardiographic dimensions and maximal oxygen uptake in elite soccer players. Saudi Med J 22: 320–325, 2001.[Medline]
- Blimkie CJ, Cunningham DA, and Nichol PM. Gas transport capacity and echocardiographically determined cardiac size in children. J Appl Physiol 49: 994–999, 1980.[Medline]
- Colan SD. Mechanics of left ventricular systolic and diastolic function in physiologic hypertrophy of the athlete heart. Cardiol Clin 10: 227–240, 1992.[Medline]
- Devereux RB and Reichek N. Echocardiographic estimation of ventricular mass in man: anatomic validation of the method. Circulation 55: 613–618, 1977.[Abstract/Free Full Text]
- Dohoo I, Martin W, and Stryhn H. Veterinary Epidemiologic Research. Charlottetown, Prince Edward Island, Canada: AVC, 2003, p. 706.
- Douglas PS and O'Toole M. Aging and physical activity determine cardiac structure and function in the older athlete. J Appl Physiol 72: 1969–1973, 1992.[Abstract/Free Full Text]
- Fagard RH. Impact of different sports and training on cardiac structure and function. Cardiol Clin 10: 241–256, 1992.[Medline]
- Fagard RH. Athlete's heart: a meta-analysis of the echocardiographic experience. Int J Sports Med 17, Suppl 3: S140–S144, 1996.[CrossRef][ISI][Medline]
- Fagard RH. Impact of different sports and training on cardiac structure and function. Cardiol Clin 15: 397–412, 1997.[CrossRef][Medline]
- Gilbert CA, Nutter DO, Felner JM, Perkins JV, Heymsfield SB, and Schlant RC. Echocardiographic study of cardiac dimensions and function in the endurance-trained athlete. Am J Cardiol 40: 528–533, 1977.[CrossRef][ISI][Medline]
- Huonker M, Konig D, and Keul J. Assessment of left ventricular dimensions and functions in athletes and sedentary subjects at rest and during exercise using echocardiography, Doppler sonography and radionuclide ventriculography. Int J Sports Med 17, Suppl 3: S173–S179, 1996.[Medline]
- Koenig H, Goldstone A, and Lu CY. Testosterone-mediated sexual dimorphism of the rodent heart. Ventricular lysosomes, mitochondria, and cell growth are modulated by androgens. Circ Res 50: 782–787, 1982.[Abstract/Free Full Text]
- Maron BJ. Structural features of the athlete heart as defined by echocardiography. J Am Coll Cardiol 7: 190–203, 1986.[Abstract]
- O'Callaghan MW. Comparison of echocardiographic and autopsy measurements of cardiac dimensions in the horse. Equine Vet J 17: 361–368, 1985.[Medline]
- Opie LH. Overload hypertrophy and its molecular biology. In: The Heart: Physiology, From Cell to Circulation, edited by Opie LH. Philadelphia, PA: Lippincott-Raven, 1998, p. 391–418.
- Osborne G, Wolfe LA, Burggraf GW, and Norman R. Relationships between cardiac dimensions, anthropometric characteristics and maximal aerobic power (O2 max) in young men. Int J Sports Med 13: 219–224, 1992.[Medline]
- Pelliccia A. Determinants of morphologic cardiac adaptation in elite athletes: the role of athletic training and constitutional factors. Int J Sports Med 17, Suppl 3: S157–S163, 1996.[Medline]
- Pelliccia A and Maron BJ. Outer limits of the athlete's heart, the effect of gender, and relevance to the differential diagnosis with primary cardiac diseases. Cardiol Clin 15: 381–396, 1997.[CrossRef][Medline]
- Pluim BM, Zwinderman AH, van der Laarse A, and van der Wall EE. The athlete's heart. A meta-analysis of cardiac structure and function. Circulation 101: 336–344, 2000.[Abstract/Free Full Text]
- Poole DC. Current concepts of oxygen transport during exercise. Equine and Comparative Exercise Physiology 1: 5–22, 2004.[CrossRef]
- Vuille C and Weyman AE. Left ventricle. 1. General considerations, assessment of chamber size and function. In: Principles and Practice of Echocardiography, edited by Weyman AE. Philadelphia, PA: Leas & Febiger, 1994, p. 575–624.
- Wilmore JH and Costill DL. Cardiorespiratory adaptations to training. In: Physiology of Sport and Exercise, edited by Wilmore JH and Costill DL. Champaign, IL: Human Kinetics, 1994, p. 214–238.
- Young LE. Cardiac responses to training in 2-year-old Thoroughbreds: an echocardiographic study. Equine Vet J 30: 195–198, 1999.[Medline]
- Young LE. Equine athletes, the equine athlete's heart and racing success. Exp Physiol 88: 615–663, 2003.
- Young LE, Marlin DJ, Deaton C, Brown-Feltner H, Roberts CA, and Wood JLN. Heart size estimated by echocardiography corelates with maximal oxygen uptake. Equine Vet J 34, Suppl: 467–472, 2002.
- Young LE and Scott GR. Measurement of cardiac function by transthoracic echocardiography: day to day variability and repeatability in normal Thoroughbred horses. Equine Vet J 30: 117–122, 1998.[Medline]
- Young LE and Wood JLN. The effects of age and training on murmurs of atrioventricular valvular regurgitation in young Thoroughbreds. Equine Vet J 32: 195–199, 2000.[Medline]
Copyright © 2005 by the American Physiological Society.
TOP
ABSTRACT
MATERIALS AND METHODS
