Athletes use caffeine strategically:
In 2009, Ganio and his colleagues published a systematic review of 21 studies on caffeine in timed performance. Most of the researchers looked at subjects cycling, but some also studied running, rowing, and cross-country skiing, and most of the tests were in the 15-minute to two-hour range. Looking across all the results, Ganio found consistent improvements in performance.
The improvements can be substantial, he told me, often as much as 3 percent. To put that into context, a 3 percent improvement would mean an 18-minute boost in a 10-hour race. Eighteen minutes was all that separated the top eight finishers in both the men’s and women’s pro races at Kona.
[...]
Ganio said it is important to take the right dose, which shakes out to about three to six milligrams per kilogram of body mass. That is a lot of caffeine. An 80-kilo (176-pound) athlete taking six milligrams per kilogram would need 480 milligrams of caffeine.
“That’s four strong cups of coffee,” said Ganio. “If you can tolerate it, it seems to be the upper end of what you can have to improve performance.”
Since “cups of coffee” is a notoriously imprecise measure of caffeine, it may help to think of it this way: 480 milligrams would be six 8-ounce Red Bulls, two and a half NoDoz tablets, or two Extra Strength 5-hour Energy shots. A more moderate dose for a smaller athlete, say, a 65-kilo (143-pound) athlete taking three milligrams per kilo, is still an impressive amount of caffeine: equal to one NoDoz tablet, one 5-hour Energy shot, or two and a half Red Bulls. Even this amount of caffeine is difficult to obtain using caffeinated sodas like Coca-Cola. A 65-kilo athlete would need to chug nearly six cans of Coke at once to get a caffeine dose of three milligrams per kilogram.
Also, researchers found no evidence of dehydration from using caffeine.
I am very skeptical of their claimed finding that caffeine isn’t a dehydrating diuretic.
There are two kinds of dehydrating agents: relative, and absolute. A relative agent messes with homeostatic processes, lowering the hydration set point. An absolute agent takes a set amount of water per gram.
Excess salt is an absolute dehydrating agent. The kidneys filter it out and dilute it to a safe concentration.
Caffeine is usually consumed with more than enough water to offset absolute dehydration. If by “no evidence of dehydration” they mean caffeine-takers don’t end up dehydrated, this might be what they’re talking about.
Alrenous, the relevant passage is here:
The claim is that coffee is indistinguishable from water and challenges the notion that caffeine is a diuretic at all, which is a much, much stronger statement than that caffeine consumption is a net-neutral on hydration levels because drinkers tend to offset the diuretic effects with just enough additional liquid volume.
My bodily functions definitely respond differently to a large (cup and a half) coffee than to a large water, so I find the claim highly suspect.
Sounds to me like the researchers prefer saying things to thinking about what they said. Paragraph one and two don’t match. They also commit a scientific sin of non-explicitness; the content is all in vague implications.
Whenever I do intense exercise, I sweat. When I exercise for a long time, I have to re-hydrate.
In my case, any water lost to caffeine-bladder effects during exercise is trivial compared to water lost through sweat glands.
So heavily sweating athletes are probably a noisy data source.