Good hydration = good race time

Sounds interesting enough.

What is pre-hydration levels? 55% of what?

The article forgets the difference between correlation and causation. Obviously, being well hydrated will help any athlete during an Ironman. But, there could be underlying factors that cause the correlation between faster times and hydration. For example, more experienced athletes, whom have performed numerous Ironmans, will likely have learned that hydration is important and will emphasize the need to hydrtate leading up to a race. Researched as shown that the finish times of Ironman athletes improves after the first Ironman; thus, faster times and better hydration....just one example of many possible causations.

Well that's quite a "study."

Shane

here is the link to the more in depth analysis - http://www.tanita.com/en/hydration-data/

according to them muscle is normally approximately 50-70% hydrated whereas fat is only 10-15% hydrated - so i'm assuming the 55% falls into the muscle measurement as the following was included in the analysis - amount of lean and fat tissue within the body, bone mass and hydration

while I agree on the comment about correlation does not equal causation - with the amount of data that they analyzed - 5000+ athletes - then predictive analysis using multiple regression analytic techniques is a recognized method of predictive behavior/performance (I loved stats in college) - you see this in employment studies...like a predictive model of success in this career field with these variables - and then there are multiple variables, each one weighed differently depending on importance

My causation vs correlation angle has to do with athletes who hydrate well are also athletes who train well, race smart, and pay attention to all the other things that could derail their race. The hydration is just one piece, but someone who pays attention to hydration is likely paying attention to a lot of the other things as well.

true - but with 5000+ athletes and the wide varying ranges - over the 3 year period - then a good portion of the data is likely going to encompass a wide variety of ppl and circumstances which increases the generalizability of it - would be interesting to take into account training, previous performance, number of IM etc to see if that has an effect - but the data collection on that, the tracking, especially if you were to factor in training (self-logged vs. monitored) would be insane and likely highly variable in its accuracy (especially if relying on self reports)

While there could certainly be data collected that shows a causal relationship between hydration and finishing time, these data do not do that.

Elements that the study would need to address in order to actually have some value:

- Error estimate of the scales ability to take measurements

- Error bars on their graphs

- p-values

Shane