Ever wonder how much the very, very best marathoners in the world drink during marathons, and how much weight they lose to sweat? A cool new study in the Clinical Journal of Sports Medicine, from an international team led by the University of Glasgow, collects some very neat data from some really, really big names: Haile Gebrselassie, current world-record holder Patrick Makau, 2008 Olympic champion Sammy Wanjiru, 2004 Olympic champion Stefano Baldini, plus other stars like Kenenisa Bekele, Marilson dos Santos, Martin Lel, Felix Limo and Evans Cheruiyot.
There are several parts to the study. In the "headline" part, they attempt to estimate total fluid consumed by the athletes above in winning major marathons (the Olympics, Berlin, Dubai, Chicago, London, Fukuoka, New York, Rotterdam) based on video footage from the races provided by the BBC. Inevitably, this is a very crude estimate, since the footage is imperfect and they have to guess the actual drinking rate. The major finding from this is that there's huge variability from person to person, and it doesn't just depend on environmental conditions. For example, Felix Limo appeared to drink far more in winning London in 2006 than Martin Lel did the following year, even though it was way hotter in 2007.
Far more interesting (and reliable) is this nugget:
they measured Geb before and after he won Dubai in 2009, and despite drinking at a fairly high rate, he lost 9.8% of his starting body weight, dropping from 58.2 kg to 52.5 kg. He apparently drank 1.7 liters during the race, but sweated out 3.6 liters per hour. Moreover, this is apparently similar to his weight loss in the Berlin marathon when he set the world record (twice). So much for the idea that losing more than 2% of your body weight hurts performance!
The researchers also did a series of acclimatization studies leading up to the 2008 Olympics with athletes including Geb, dos Santos and Bekele (there are others, but I can't deduce from the information given who they are). On average, the runners sweated at a rate of 2.3 liters per hour.
So what's the overall message here? When the researchers attempted to look for patterns (e.g. plotting fluid consumed versus temperature for all the marathons analyzed), the data is all over the place. Among this
elite group of marathoners, drinking behavior seems to come down to individual taste: the amount they drink depends on how thirsty they are and how much their stomachs can handle. They're definitely not making any attempt to replace all the fluid they lose, or to stay within 2%.