Really? NYTimes: "Even Tylenol has been shown to boost endurance performance by 2 percent."

What was the placebo effect? If that was itself significant, then wouldn't the tylenol effect or beet juice or caffeine effect be quite large over non-placebo, non-tylenol, etc.?

Beet Juice =/=> Komen 7:11-7:16

Caffeine =/=> Komen 7:11-716

Slightly possible that Caffeine => Komen 7:20 instead of 7:21.

Didn't Komen test positive for caffeine at some point? The effects probably don't stack all that well (i.e. you don't get a 2% improvement for caffeine, then another 2% from Tylenol).

rojo wrote:

I'm catching up on last week's news as I write the Week That Was.

I was reading a NY Times article on doping which casually had this line in there, "Even Tylenol has been shown to boost endurance performance by 2 percent."

Is that true? Anyone know of the study? I've never heard of that and if it was true, you'd think I'd have heard of other athletes, college coaches using it, but I've never heard of it.

Article is here:

http://www.nytimes.com/2014/11/30/opinion/sunday/when-doping-isnt-cheating.html?_r=1

Haven't read the studies, but I very much doubt the veracity of that claim. 2% boost in performance means, what, exactly? a reduction in a PR by 2%? thats huge. thats going from a 14:00 5K to a 13:43 JUST from tylenol. No effing way, you wouldn't even get that from real PEDs the day of the race.

My guess is they're quoting some poorly designed, small sample size, non-controlled study which has some obscure primary outcome which they falsely equivocate to performance in their conclusions.

I am a doc wrote:

Haven't read the studies, but I very much doubt the veracity of that claim.

Quintessential letsrun quote.

To the point, the study cited was controlled and blinded and did directly measure performance (time in a bicycling time trial). It did have a small sample size (13).

How general this result would be would obviously need to be tested in different settings and with higher sample sizes, but to dismiss the effect outright shows a pretty clear lack of curiosity and the sort of arrogance you expect from a typical doc.

Just read the study. Much better than I thought it would be. Still have massive reservations though.

sample size of 13, 6 per group when you break it into controls. This is at best a starting point for further research. I would need a much better powered study to convince me, like around 400 per group.

The results in the control and treatment groups aren't statistically significant. The overlap between the control and treatment groups is much too great to derive any meaningful conclusions, you have to accept the null here. I like how all the data points on the graphs use either the low or high deviation from the mean for each group only, to clean it up and make it look as though a trend exists which may or may not actually exist. That is very suspect. I bet you add in all the data points and the vertical bars would overlap so much that it would eliminate any perceived trends. Why did they choose to leave half the data out for each group on the graphs?

As a style point, they spend a ton of time rambling on in the discussion section. Classic physiology. light on data and methodology, heavy on speculation. And thats what this paper is, scant and poorly analyzed data with a ton of hand waving at the end.

If they were measuring disease rather than performance, no medical journal would ever consider publishing this in a million years. what an amazing difference in standards.

two things:

absolutely no cross over from english university students riding as bike to exhaustion and a 30:30 minute 10km boy trying to get to 29:15.

and

british studies are shunned across the world of ex. physiology, big skeptical eyebrow raised on every work.

those in the community know that.

(how are the beets working out for you?)

egun wrote:

absolutely no cross over from english university students riding as bike to exhaustion and a 30:30 minute 10km boy trying to get to 29:15.

Exactly, these are just hobbyjoggers on the treadmill

They can knock 2% off their PR just by picking a better flavor GU

egun wrote:

two things:

absolutely no cross over from english university students riding as bike to exhaustion and a 30:30 minute 10km boy trying to get to 29:15.

and

british studies are shunned across the world of ex. physiology, big skeptical eyebrow raised on every work.

those in the community know that.

(how are the beets working out for you?)

Again, it was not an exhaustion trial but a ten mile time trial. Unless you have a compelling reason why there wouldn't be crossover, I'm inclined to think you pulled that out of your butt.

Ho Hum wrote:

Again, it was not an exhaustion trial but a ten mile time trial. Unless you have a compelling reason why there wouldn't be crossover, I'm inclined to think you pulled that out of your butt.

There may be crossover, but you have to consider the law of diminishing returns.... what helps a 5 hour marathoner improve 2% probably won't help Kipsang the same amount

You misread the article.

They actually have a sample size of 13 if you accept that the two trials are independent. Each participant was tested with the control and the Tylenol with 2-7 days between the trials. This means that some did their first trial with tylenol and some did their second with tylenol.

The figure shows the mean times of all the rides and that is why there is such a big overlap in the error bars. Much of that is variation in rider ability. There was no statistical test on mean time. The statistical test was a paired t test which removes that inter biker variation. The verbiage in the figure and text is a bit fuzzy on what the test done on time was.

As to the quality of research, you have to be kidding. My father is suffering from a relatively rare disease and there are 10s of papers published on it in decent medical journals with similar sample sizes to this paper. Research pretty much sucks in all fields.

I think the point is being missed. Forget about the actual percentage, or whether or not benefits are "stackable" or if elites benefit as much as non-elites.

What this really shows is that the fight on doping is pointless. Just like the war on drugs. What this shows (and what I've said before on here and in other places), is that the term "performance enhancing" is too nebulous and hard to pin down. Lets just say there is "some" benefit to Tylenol.

You will then have a situation where there are "super responders." That's one of the arguments people use against Lance Armstrong...he was a super responder to EPO, otherwise he had no talent. Maybe or maybe not. Same argument can be made for people that are "super responders" to interval training, high volume miles, carrots, sleep, altitude.

rojo wrote:

I'm catching up on last week's news as I write the Week That Was.

I was reading a NY Times article on doping which casually had this line in there, "Even Tylenol has been shown to boost endurance performance by 2 percent."

Is that true? Anyone know of the study? I've never heard of that and if it was true, you'd think I'd have heard of other athletes, college coaches using it, but I've never heard of it.

Article is here:

http://www.nytimes.com/2014/11/30/opinion/sunday/when-doping-isnt-cheating.html?_r=1

next time try google. Original study:

Also referenced in:

If you really know this stuff can you please provide the backup calculation for how you arrived at a sample size of 400 per group in order to achieve an appropriately powered study?

K THANKS BYE

know this stuff... wrote:

The results in the control and treatment groups aren't statistically significant. The overlap between the control and treatment groups is much too great to derive any meaningful conclusions, you have to accept the null here.

What are you basing this on? They found a statistically significant difference between the means of the placebo and acetaminophen treatments. The mean completion time difference, as stated below, was 30 seconds with a confidence interval of 4.3 to 55.8 seconds. They used a pair t-test which was appropriate in this case given that each rider was given both treatments.

They also referenced two studies to support their claim that a sample size of 12 is appropriately sized. It doesn't really appear that you read the study...

A significant difference in completion time between conditions was found (t12 = 2.54, P < 0.05) (95% CI of the difference = 4.3–55.8), with participants completing the TT in significantly less time during the ACT condition (26 min 15 s ± 1 min 36 s) than during the PLA condition (26 min 45 s ± 2 min 2 s), as shown in Fig. 1.

The full article states the average VO2 Max was 65 +/- 5. How does this equate to a "5 hour marathoner". VO2 max of 65 for runners = low 2:31 marathon. The were not out of shape "hobby cyclists".

Tylenol, unless overdosed, has by far the least amount of side effects of pain control medicines. No way a dose of Tylenol, every month or so prior to a big race, would cause any long term issues.

Is it cheating- possibly. But how many of us drink coffee, put on aspercreme, etc. prior to race

Avg Vo2 /Tylenol is very safe wrote:

The full article states the average VO2 Max was 65 +/- 5. How does this equate to a "5 hour marathoner". VO2 max of 65 for runners = low 2:31 marathon. The were not out of shape "hobby cyclists".

Tylenol, unless overdosed, has by far the least amount of side effects of pain control medicines. No way a dose of Tylenol, every month or so prior to a big race, would cause any long term issues.

Is it cheating- possibly. But how many of us drink coffee, put on aspercreme, etc. prior to race

Eh, not exactly. My buddy has a 65 VO2 max tested. 2:58 marathon w/ taper and 80 mpw.

No.

So I did misread the sample sizes, however 13 is no better than 6. As far as times go, I think you're misreading me. I think its great to use the mean times for all the rides. But include the positive and negative standard deviations for each point on every graph. Not the negative SD on the lower line of best fit, and the positive SD on the higher line of best fit. There is a significant amount of overlap between tylenol and no-tylenol group as reported, but not shown, in the figures.

Sorry about your dad. I intended to say all medical research is quality. With many rare diseases, there is little $ in it for companies, and therefore little research done. If you want to look at any NEJM papers on drugs for common diseases though, they analyze 10's of thousands of patients to really eek out the benefits of therapy. Most of that research is straight forward. Most has little hand waving and relatively short discussion sections, as the data and methodology speaks for itself, unlike this paper.

13 is much better than 6. Please stop exaggerating.