slvrblltday wrote:

From the Thoroughbred Safety Committee on Soundness Issues:
Charge number one: The training and racing of 2-year-old Thoroughbreds is predisposing these horses to accelerated rates of injury and prematurely shortened careers.

This charge is leveled by some people in and out of the horse industry, especially people outside of racing. It is a very popular theme with animal welfare organizations that are ill informed on the topic of racing and the horse; it is also parroted frequently in the popular press.

To examine these data The Jockey Club Information Systems extracted one-year windows at five-year intervals, using the years 1975 through 2000 as data sets. Horses were divided into the categories "raced as two-year-olds" and "raced, but not as two-year-olds." The data shows a definitive answer to this charge.

The first category of data examined was average starts per starter lifetime. The data shows that horses that raced as 2-year-olds raced many more times in their lifetime in each of the years examined when compared to horses that did not race until after their 2-year-old season. Some of these starts were made in the 2-year-old year for the horses that raced at 2, but the difference was more marked than the 2-year-old year alone would account for.

Average lifetime earnings per starter for horses that raced as 2-year-olds are almost twice the amount earned by horses that did not race as 2-year-olds. Career average earnings per start for horses that raced as 2-year-olds exceeded average earnings per start for horses that did not race as 2-year-olds in every one of the years from 1975 to 2000 examined. Lastly, the percent stakes winners in horses that raced as 2-year-olds is nearly three times higher than in horses that did not race until their 3-year-old year or later.

This data is definitive. It shows that horses that began racing as 2-year-olds are much more successful, have much longer careers, and, by extrapolation, show less predisposition to injury than horses that did not begin racing until their 3-year-old year. It is absolute on all the data sets that the training and racing of 2-year-old Thoroughbreds has no ill effect on the horses' race-career longevity or quality. In fact, the data would indicate that the ability to make at least one start as a 2-year-old has a very strong positive affect on the longevity and success of a racehorse. This strong positive effect on the quality and quantity of performance would make it impossible to argue that these horses that race as 2-year-olds are compromised.

These data strongly support the physiologic premise that it is easier for a horse to adapt to training when training begins at the end of skeletal growth. Initiation of training at the end of growth takes advantage of the established blood supply and cell populations that are then converted from growth to the adaptation to training. It is much more difficult for a horse to adapt to training after the musculoskeletal system is allowed to atrophy at the end of growth because the bone formation support system that is still present in the adolescent horse must be re-created in the skeletally mature horse that initiates training.


I'll outline some of the issues with this article.

1. the use of loaded language "charge" that racing two-years olds (neutral = hypothesis or theory), emphasis on "animal welfare organizations" and "people outside racing" - suggesting that only outsiders have concerns about this issue is a red flag that this is already a biased article.

2. there are no "definitive answers" or "definitive conclusions" in science.

3. First, the claim that the horses started more - of course they did. THEY RACED AS TWO YEAR OLDS. This is circular reasoning, and totally inappropriate. The appropriate method would be to remove all 2 year old starts from the equation, and proceed from there. But even given that, this is actually a *terrible* measure for the outcome of interest - at best it's a weak proxy, and at worst it's confounded. For example, starting less often is not necessarily a marker for unsoundness - it may also be a marker for an owner/ trainer's unwillingness to run a horse more often - owners concerned about long-term soundness would be likely to a. start a horse later, and b. run them less often. Using age at first start and time between starts as continuous variables would provide a much clearer picture of career trajectory, and the dependent variables should be acute injury and long term chronic injury/ osteoarthritis. Another important consideration would be use of NSAIDS and injectables.

4. Lifetime earnings is, again, a poor proxy for soundness. We already know that racing 2 year olds is encouraged by the industry and as a result, lucrative - otherwise, why do it?

5. OMG, you CANNOT extrapolate physiologic outcomes from the information provided. !!!!!!!!!!!!!!!! IF YOU WANT TO KNOW ABOUT RATES OF PHYSICAL DETERIORATION, YOU HAVE TO LOOK AT PHYSICAL DETERIORATION. If you must use a proxy, there have better be an ironclad defense of that proxy with excellent controls in the analysis. For example, in our study, we measure blood flow through uterine arteries in pregnancy. Other people have already done work that shows that this blood flow is strongly related to placental function. Our outcome of interest is fetal growth - so we can use blood flow as a measure of placental function as it relates to placental growth. BUT YOU CAN'T SKIP STEPS BECAUSE IT'S CONVENIENT FOR YOU.

6. And then they jump right to their conclusions, which were the same ones they had when they started. And hope that no one reads this who can call them out on their BS. It is entirely possible - in fact likely - that some level of strenuous work improves bone quality in young horses. That has already been shown. But to conclude that racing - and the level of training involved - at the start of a two-year old year as a result of that information and the heaps of crap in this article is ABJECTLY IRRESPONSIBLE AND INTELLECTUALLY DISHONEST.