Photo by Bernard Walker.

Tennis Note #31b

“Proof” from an Elusive Meldonium Study

Five days ago, I published Tennis Note #31 on the science of meldonium and the somewhat questionable meldonium studies being cited as evidence. Recall, the following part of my long analysis:

They claim that there are 25 articles directly related to the effects of mildronate on physical work capabilities but only cite one from 2002 — Effect of Mildronate on physical working capacity of highly qualified judokas. I have tried to track down this paper but I still cannot find it. Other recent papers continue to cite it but a little insider information about work cited in research papers: many times authors will see how a citation is used in another article and just copy and paste the citation without reading the article. Especially if you write only one sentence about mildronate in the context of sports performance.

The one sentence from this ‘WADA’ paper that caught my attention and made me a bit suspicious states:

“Under sport-physiological aspects, reports on positive effects on the physical working capacity of elite athletes were published and dosages of Mildronate (per os between 0.25 and 1.0 g twice a day over 2–3 weeks during the training period and 10–14 days before competition) were discussed.

A student of meldonium’s inventor [Prof. Kalvins] wrote an abstract that was a literature review presented at a conference a few years before, citing the same judoka study with the following sentence:

“Mildronate is usually administered to athletes perorally in dose 0.5–1.0 g twice a day before training, as 14–21 day course during training period and 10–14 days before competitions.”
Kakhabrishvili Z, Chabashvili N, Akhalkatsi V, Skhirtladze T, Chutkerashvili T. Mildronate effect on physical working capacity among highly qualified judokas. Ann. Biomed. Res. Edu. 2002, 2, 551. What in the world does the 551 refer to? Is this another propagating error?

My main concern was that if I could not even find this paper, how could I draw any conclusions? Well, one of the readers of Tennis Note #31, a student at the same university as this paper, contacted me and helped me track it down so I can now dissect it for you. The title is shown in the image and I already have a critique — was this cited incorrectly because I cannot figure out what the 551 stands for [it is not the issue or page number] when everything else matches. Everyone has cited it the same way.

How was this study conducted?

Fourteen members of the Georgian National Team of Judo participated in this study. There were two groups: Experimental and Control. As I mentioned in Tennis Note #31, control groups are crucial in order to draw good conclusions. For this study, the image below describes each group.

Why is there such a massive discrepancy between the experimental and control group weight and age? This is the first thing that I noticed and as soon as I saw it, I was not sure if I could even proceed with the data. There is always variability between people so the least you can do is separate your group of fourteen athletes in a way that is more even.

The experimental group took 0.25 mg of mildronate in capsules per os four times a day during 20 days. In case you are wondering, per os means oral administration. Recall what the two other studies stated in their introduction, and that Kalvins’ abstract is referenced in the drug testing paper. I think you know what I am thinking.

Moving on, both groups trained through the same routine and all subjects underwent several different tests including:

  1. General Medical Exam — my guess is for weight, height, heart rate?
  2. Electrocardiogram — they do not really mention the data output
  3. Echocardiogram — to measure cardiac output, stroke volume, ejection fraction, shortening fraction and left ventricular mass
  4. Functional Tests: JMG test and PWC 170 test

The physical working capacity was assessed with a PWC 170 test, described in the video. The other major test is the JMG test, developed in 1982, which specifically looks at the aerobic-anaerobic capacity for work of judokas based on a formula and specifically designed for judo. Remember, mildronate helps the production of energy in oxygen depleted environments like anaerobic exercise, which causes lactic acid buildup or fatigue. There are three different exercise tests (described in the link) and the athletes have one minute per exercise to perform as many repetitions as they can. After they finish the test, their heart rate is measured the moment they finish and a minute afterwards.

Based on the number of repetitions, the heart rate, weight and age, there is a formula for a JMG ratio and it scales from -50 to 50, where -50 is excellent and 50 is very bad in terms of this ability to work in an oxygen depleted state and oxygen favorable state. If you are still uncomfortable with this measure, do not worry, you are not alone. However, since it seems to be established in judo, I will let it go and explore their data.

Before I explain anything further, for those not familiar with some simple statistics terms, make sure you know the definition of the average [mean] value and the standard deviation, which will appear as +/- some value. The measurements are essentially before and after taking mildronate. I imagine the control group took a placebo but this was not explicitly mentioned. Assuming there was a placebo, they also did not mention if the study was double-blind. This is important if you want to remove any bias in your study. Last thing to mention, this paper does not have any visuals, just two tables — one for each group, provided on the 3rd page (of the 3 page article).

What does the data demonstrate?

The first piece of data they mention is the ejection fraction[EF] and shortening fraction [SF]. The ejection fraction describes the amount of blood pumped out of the ventricles with each contraction. Shortening fraction looks at the how the left ventricular diameter changes between end-diastole and end-systole. If you want to know more about some of the cardiac terminology, I just taught a lecture on it last week. The normal range for EF and SF is 50%-75% and above 30%, respectively, and is something not mentioned in this paper.

In the control group, EF increased from 63.8+/- 2.6% to 65.8+/- 2.1%. The experimental group increased from 60.7+/- 0.85% to 64.8+/- 0.8%. Thus, there is a subtle difference in increase from the average EF for each group but both groups fall within the normal range.

For the control group they found that the SF increased from 35.0+/- 2.16% to 36.57+/- 1.7%. The experimental group increased from 32.8+/- 0.6% to 35.8+/- 0.59%. Again, a subtle difference in increase between groups but still within the normal range.

It should be noted that this paper stated the percent change between before and after mildronate for both groups, but I think the actual numbers tells you more, right? You might notice a few other things about the “before the drug” values for each group. We will get back to this in a minute.

The next part looks at the general and relative PWC 170 data. The PWC test predicts the power output at a projected heart rate of 170 bpm. Based on the units, it seems ‘relative’ means relative to weight. Because these groups were uneven in weight, I really do not know what to make of the general data.

The control group increased from 1148 +/- 62.8 to 1228.36+/-73.9 [kg*m/min, not the units for power] in the general PWC test or a 7% increase (which is on the order of the standard deviation in these measurements). The experimental group increased from 1555 +/- 88.2 to 1741.6 +/- 66.1 or a 12% increase (which is on the order of twice the standard deviation in these measurements). For me, the fact that the control group and the experimental group start with drastically different values even before the intake of mildronate/placebo, caught my attention.

Relative PWC 170 data. Before is in light gray. After is in dark gray.

However, if you look at the relative PWC data points, the control increased from 15.8+/- 1.06 to 16.8 +/-0.5 and the experimental increased form 15.8+/- 0.97 to 18.02 +/- 0.67. Since relative PWC appears to be relative to weight, this could explain the drastic differences in the normal data. I made this data visual so you can see the change between before [light gray] and after [dark gray] for control and experimental. This is a good time to mention this study did not perform any statistical tests to determine if there is a significant difference between before and after or control and experimental results. Again, they simply provide the percent change, as if this is is a valid measure.

The last piece of data was from that JMG ratio. Remember the scale is -50 to 50 where the negative range is more favorable so we are looking for a decrease in this value.

JMG ratio data. Before is in light gray. After is in dark gray.

The control group went from 18.14 +/- 2.5 to 17.05 +/-1.7. The experimental group went from 15.57 +/-1.8 to 13.08 +/-1.4 and the visual illustrates this in another way. All the values are positive so I guess on the rating scale, no one is excellent. The start value between both groups is different although there is some overlap — again this could do with age/weight discrepancy between groups, especially since it is part of the JMG formula. Just looking at this data visually, I can tell it is not statistically significant based on the overlap between the before and after for each group. I highly doubt a difference of 2.49 is much much better than a difference of 1.09. This is ridiculous.

I will not even mention any subjective statements because an athlete stating they feel good means nothing.

All I got from this data was a possibly biased study design with zero statistical analysis and differences in value provided in percentages instead of raw numbers, when in reality many measurements are within a normal range or have minute differences. I do not even know how many times this experiment was repeated. Basically I conclude nothing.

However, the paper concludes with four statements.

  1. This drug does not violate the doping criteria. [Remember, in 2002 this was not banned.]
  2. Mildronate can be used to increase physical capacity. [Does it though?]
  3. No side effects.
  4. They recommend the use of mildronate while doing aerobic training.

Remember, this paper is cited repeatedly as proof that mildronate increases the physical working capacity for athletes. How convinced are you based on this data and the study design?

Special thanks to the reader who had access to the paper. If you enjoy reading these tennis notes, make sure to follow the publication, ‘Recommend’ and share! Check us out on Facebook! Made a cool observation? Interested in certain topics and writing? Are you a tennis photographer? Comment, add notes, and check out the submission guideline. Let me know which visuals are good and which are not so great. Cheers!