Cardio evaluation (almost) comes full circle
For some time now, cardiovascular evaluation of thoroughbreds using echocardiography has been one way to evaluate the performance potential of an unraced thoroughbred. Much of the work on this relationship was initially done many years ago by Dr. Norm Rantanen, Dr. David Lambert and Jeff Seder in America, while Dr. Lesley Young in the United Kingdom has also done a lot of work in this field.
Echocardiography has emerged to become a standard diagnostic procedure in equine cardiology. Development of M-mode echocardiography in the mid-1970s and introduction of 2D real-time echocardiography in the 1980s allowed evaluation of internal cardiac structure, size and function in horses. This is a completely different process to the use of electrocardiography, and the generation of Heart Scores (HS), which is what Steele and Stewart used in the 1960's and 70's and what Marianna Haun erroneously pinned her "X-Factor" research on. We could go on quite extensively here about how the "X-Factor" is a complete waste of time, but it will suffice to say that there have been papers that have shown that electrocardiographic Heart Scores which the "X-Factor" was built on has no relationship to V02Max and racetrack performance in thoroughbreds The X-Factor was a great story but that is all it was.
It is a different story for echocardiography however. Cardiovascular parameters measured by echocardiography have been of interest to exercise physiologists due to the relationship between the capacity of the cardio to deliver oxygenated blood to the muscle. The theory to prove was that a strong relationship exists between cardiac dimensions and peak oxygen uptake (V02Max) in thoroughbreds with the latter being a measure of performance (that is horses with high peak Vo2Max scores are as a group better racehorses with higher earnings than those with lower peak Vo2 Max scores). Back in 1999 there was a good paper authored by Dr Bayley of Washington State University that looked at the relationship between Vo2Max and echocardiographic measurements, while in 2002, Lesley Young and others published a paper with similar findings, that is, V02Max had a strong correlation to echocardiographic measurements of left ventricular size. Both of these studies were performed on mature horses.
A few years later in 2006, Dr. David Evans also authored some very interesting papers with others looking at the relationship between velocity at maximal heart rate (VHRMax) and performance. Evans' studies were based on the concepts above - running ability of thoroughbred racehorses is correlated with Vo2Max, and the velocity at Vo2Max is highly correlated to velocity at maximal heart rate (VHRMax), thus VHRMax, which can be very easily measured using a heart rate monitor with a GPS on a horse in work, is a good measure of both fitness and performance potential of a thoroughbred. Evans found that not only was VHRMax highly correlated to general fitness of the horse, but more importantly horses with higher VHRMax earned significantly more dollars per race start (r=0.41).
While Evans' studies established the link between V02Max and VHRMax, and the papers by Young and Bayley established the relationship between Vo2Max and cardiovascular parameters, there were a couple of gaps in published research to bring this 'full circle'. Firstly, there wasn't a study that tied the relationship between echocardiographic measurements and VHRMax and secondly all the studies on Vo2Max and VHRMax had (rightly) been done on mature horses and thus had limited value in prediction of performance when looking at young horses. The missing link as far as those using cardio measurements on young horses to predict performance was establishing the relationship between young, untested horses and subsequent VHRMax/Vo2Max when those horses in training.
This week in the Journal of Equine Veterinary Science a paper was published that helps close that gap. The paper, authored by Drs Gur and Matur, outlines the relationship between the VHRmax of a horse at 28 months of age, and cardiovascular measurements taken at 4, 8,12,16,20 and 24 months. Now there are a couple of caveats on the study here that need to be addressed. Firstly, the VHRMax was evaluated by using the software program within E-Trakka, the device used to measure heart rate and velocity. While Andrew Stewart at E-Trakka has done a sterling job creating a great product (it can now gather accelerometer data which is really interesting), as far as I am aware, the E-Trakka formula for creating the VHRMax is not the same as that developed by Evans, et al (Evans helped create E-Trakka and among other things I understand that the VHRMax calculation used by E-Trakka was a point of disagreement between Evans and Stewart), and E-Trakka has not published under peer-review its method of measuring VHRMax within its software so we are not 100% sure that the VHRMax generated by E-Trakka is indeed the actual VHRMax of the horse. Secondly, the sample size was rather small (21 horses). That said, with those caveats aside, the paper had some very interesting finds.
Left Ventricular internal diameter in diastole (LVDd) measured was inversely related to VHRMax at 28 months. Thus horses with low LVDd values as yearlings would have high VHRMax values at 28 months of age. This was a rather surprising finding in some ways as Young (2002) had found that LVDd had significant correlation with maximal oxygen uptake(VO2Max) in mature horses. This lack of agreement could point to the rapid growth that a young foal goes through and the changes that the cardio undergoes during training making LVDd a poor predictor as a young horse but a good one for measuring older horses in training.
Left Ventricular Mass (LVM) was very lowly correlated to VHRMax at 28 months and in some months inversely correlated. This finding is in agreement with Young (2005) who found that LV Mass was not correlated to race performance in 2yo horses (a similar time measurement of 28 months). Young (2002 and 2005) did find that LV Mass was correlated to performance when measured as older horses so again maturity and training response may have an influence here.
Ejection Fraction (EF%) and Fractional Shortening (FS%) values tended to be positively correlated to VHRMax at 28 months, but only those measurements that took place after 12 months of age. The r-squared values for EF% and FS% were particularly high (r=0.45) at age 20 months indicating that those scores used at two year old in training sales might be the most effective. This is in agreement with Young (2005) who found that EF% was correlated to performance as an older horse in both sprint and longer distance races, but not in agreement with Seder (2003) who found no relationship between FS% and performance.
Anomalies in the data seemed to occur around the age of 16 months. This is slightly different to Seder (2003) who found that these anomalies tended to be found at 20 months but the explanation for this can be found in the fact that in this study the horses were broken-in and started training in September right at the time that most of the horses would be 16 months of age, where the group in Seder's study would most likely have been sold as yearlings at the same time and not been broken in until closer to 20 months of age, thus initial response to training may have had an effect as may age related growth (puberty).
This latest paper shows that at differing ages various measures of the heart are moderately correlated to performance. It is also interesting to note that the direct morphologic measurements (LVDd and LVM) had either a negative or very low correlation to VHRMax at age 28 months, but that the functional cardiac features with indices such as FS% and EF%, were more highly correlated. You will notice that we used the term "moderately correlated". In general the positive correlations that were found in this paper fell in the range of 0.26 and 0.49 which is considered "weak", but not "low" in statistical terms. As an example, at 16 months the FS% saw an r2=0.402. This means that the FS% explained about 40% of the VHRMax at 28 months. This is a decent figure, but it still means that 60% of the VHRMax at 28 months was explained by other factors such as blood volume (the spleen, which is another thing we have talked about measures of), mitochondrial activity, lactate buffering, other genetic influences, etc.
These finding are somewhat in keeping with the great cardio study done by Rikke Buhl et al on standardbreds. Dr Buhl and his colleagues measured 103 standardbreds over two years starting at 20 months of age, prior to any training and finishing at 3.5 years of age. They found that in standardbreds, as Young found in thoroughbreds, there was significant changes in the cardio in response to training. Some horses were high responders to training, others were not, but that in general, horses that were trained had significant cardiovascular changes than those that were not trained. Buhl found that training intensity was highly correlated to LVDd (which may further explain the anomalies mentioned in point 1 above) with standardbreds that were trained more intensely having greater LVDd changes. They also found that both LV Mass and LVDd was positively correlated to performance which Young found in older thoroughbreds, but differs to Gur and Matur who found negative correlations in younger horses of 1-24 months. Again this looks to be age and response to training related.