In the May 26 article by Dave Briggs entitled Why the AGCO Lowered the TCO2 Test Threshold there are a number of points that need to be addressed and corrected.
There is no doubt that the main reason for the decision to lower the testing threshold was “mostly a matter of aligning itself” with other racing jurisdictions. That does not mean they have the science right – they are merely doing what everyone else is doing – rightly or wrongly.
The AGCO release regarding the ‘natural level’ of TCO2 in blood is inaccurate, and based on averages, not the actual values in scientific research studies, including those studies supported by horse racing jurisdictions. The normal range, as reported in scientific studies that have been peer reviewed by other scientists, is 23 to 38 mmol/L, not 26 to 32 mmol/l (see below for a selection of these studies). The range for the ~99% of horses is from 25 to 35 mmol/l. There are no scientific studies, reporting values from individual horses, that have reported a range as narrow at 26 to 32 mmol/l.
The threshold for the TCO2 test set by the IFHA is not based on physiological data, but rather statistically manipulated data obtained from thousands of racehorses on race day, in varying conditions of health, and in varying conditions of hydration. Most of these studies assume, as opposed to actual knowledge, that horses that tested with a TCO2 greater than 35 mmol/l had been administered an alkalizing substance. This is only an assumption, not fact. This is a very poor justification for a threshold test of a naturally occurring substance. It is wishful thinking, not science.
Locke is quoted as stating, “There is considerable empirical evidence to suggest that the use of alkalinizing agents in excess of normal feed supplementation artificially enhances the performance of horses.” On the surface this is correct – but when examined physiologically it is incomplete. Scientific evidence supports that normal feeding regimen of racehorses tends to be acidic and contribute to systemic acidosis – something that horsemen have known for generations. Acidosis is not good for the gut and not good for the health of the horse. Scientific evidence supports the addition of minerals to the feed to increase the dietary cation-anion difference so that systemic pH is better maintained – grain feeding causes acidosis.
The scientific evidence in many species of domesticated animals, including horses, is that health is better when pH and TCO2 are prevented from getting too low. There needs to be a balance – a balance to prevent performance-enhancing effects while at the same time allowing for health-enhancing effects. A TCO2-only test is not capable of being used to identify horses that are health-optimized versus those that have been intentionally manipulated to enhance performance. Such testing does exist and is straightforward to implement.
Scientific evidence supports that normal feeding regimen of racehorses tends to be acidic and contribute to systemic acidosis – something that horsemen have known for generations.
I was asked by the ORC (AGCO predecessor) in 1994 to analyze data on thousands of testing results of Standardbred racehorses. At the time testing comprised pH, sodium and bicarbonate (bicarbonate is ~95% of the total CO2). My analysis confirmed that testing bicarbonate (or TCO2) identified some horses as ‘over the threshold’, when in fact sodium was low and pH was low. Such horses could not have been administered an alkalizing substance, yet had elevated bicarbonate. My advice was then, and continues to be, that this testing was better than TCO2-only testing. It is honest, and it is fair to the horsemen.
Seary stated, “The current instrumentation used to determine TCO2 levels in Ontario and its associated reference ranges have been extensively validated and documented for forensic use in Canadian and international studies. This instrumentation operates with the same analytical approach that was used to establish the TCO2 international threshold. Data collected from Canadian racehorses using this instrument support the implementation of the international threshold.” This is true, but it is not scientific justification.
Seary stated Lindinger’s citations, “offer generalized results without any interpretation. The wide range of TCO2 values need to be evaluated in the light that the type of equipment and, consequently the analytical approach to calculate TCO2 concentration was not consistent in all studies. Some of these instruments measure pH and partial pressure of CO2 and from these values compute the TCO2 concentration. Others, like the one used in Canada, measure TCO2 directly. Thus, TCO2 values will vary from instrument to instrument depending on the analytical approach used.”
What is apparent from these remarks it that Seary did not read deeply and does not understand the physiology – there is considerable interpretation in the document referred to.
The instruments currently used by Racing Forensics in Ontario have been obsolete for 15 years. Regardless, the instrument variation is very low when calibrated correctly and standards verified.
Seary is also incorrect putting the onus of TCO2 range on the testing equipment used. Properly calibrated and validated instruments of any make will report the correct values within 1mmol/l. This is because corroboration of the calibration procedure is supposed to be confirmed by the use of standards of precisely known TCO2 content. It is also completely erroneous that the Beckman EL-ISE instrument used for many years by Racing Forensics, and its predecessor laboratory, measure TCO2 directly. They do not. These instruments calculate the TCO2 from measurements of pH and the partial pressure of CO2 in the blood sample, the same way as the other instruments. This instrument was not designed to measure TCO2 above 35 mmol/l, has not been produced since about the year 2000; Beckman stopped providing reagents and instrument support in 1995. The instruments currently used by Racing Forensics in Ontario have been obsolete for 15 years. Regardless, the instrument variation is very low when calibrated correctly and standards verified.
Racing Forensics does not provide the actual instrument calibration and verification results when a blood sample tests high for TCO2, something that the ORC used to do. It is an inappropriate test and provides no recourse for those horsemen that have not done anything wrong.
On a final note, unfortunately, an editorial error in the 2007 paper by Waller and Lindinger is used as justification by Seary for saying instrument readings can vary as much as 2 – 3 mmol/l. This can only occur in improperly calibrated and validated equipment. This has been corrected in the recent paper published in the Journal of Equine Veterinary Sciences.
Related peer-reviewed studies:
Foreman J, Foreman C. Positive” plasma tCO2 tests in racehorses in New York, Maine, and California result from day-to-day instrument variability. Equine Vet J 2014;46. doi:10.1111/evj.12267_52. 17–17.