By Alex Hutchinson;
In pharmaceutical research, one of the big problems is that negative studies sometimes don't get published. This may be partly to due nefarious conspiracies, but it's also a simple function of human nature: "We failed to find any effects" just isn't that interesting in most cases. As a result, when you aggregate the results of different studies, you get a falsely rosy picture of the effects, since only positive studies have been published.
Of course, the same problems apply in just about any field of science, and in science journalism too. It's more interesting to write about an exciting new breakthrough than about a study that found nothing, especially since it's very difficult to distinguish between a study that found nothing because there was nothing to find, and a study that found nothing because it didn't have enough subjects or was looking at the wrong outcome variable or was otherwise poorly designed.
All of which is by way of introduction to a study just published in the European Journal of Applied Physiology, from researchers at the University of New Hampshire. They tested ice baths on a group of 20 recreationally active males (half got ice baths, half got nothing) after a 40-minute downhill treadmill run on a 10% grade to induce muscle damage. They measured a bunch of parameters, including perceived soreness, muscle torque, swelling, and something called plasma chemokine ligand 2, a marker of inflammation that circulates in the blood. And, in short, they found no differences between the groups.
So what do we do with this information? As I've written before, there are tons of ice baths studies out there with endlessly conflicting results. My overall feeling, based on the biggest meta-analysis and the reports of athletes, is that there probably is some effect. But it's very hard to nail down this effect. One possibility here is that the damage and recovery protocols don't have good "ecological validity" -- that running downhill on a treadmill doesn't produce the same damage/fatigue characteristics that a real-world hard workout does, and that seeing how well your peak isometric torque recovers isn't the same as seeing how well you perform in a race or game or workout the next day.
Another possibility is the ice bath protocol wasn't optimal. A common criticism of some previous negative studies is that they weren't cold and/or long enough to produce enough tissue cooling (e.g. studies with 3 x 1:00 in the ice bath). That's not a problem here: they used 20 minutes (that's looong!) at 5 C (that's coooold!). In fact, it's tempting to wonder whether that's actually too long and too cold -- the practical recommendations I've heard from, for example, recovery specialists at the Australian Institute of Sport, is 10 minutes at 15 C, which is a big difference.
And the final possibility, of course, is that there really is no effect. The truth is, we simply don't know for sure at this point. There's enough positive evidence to justify using ice baths if you like them and find they help; and there's enough negative evidence to justify a decision not to use them if you're not that into them. The one thing I think we can say for sure is that -- like many of the "supplementary" things we debate on this blog -- any effect they have is relatively small, otherwise the studies wouldn't be so ambiguous. Ice baths may help, but they're meaningless if you're not also getting the big things right: training hard, resting hard, sleeping and eating well.