Rest vs. Active Recovery

Lots of stuff happens when you’re not doing anything. It’s amazing. Your muscles rebuild (hopefully stronger than before). Your bones deteriorate less (if you’ve been doing weight bearing exercise). Britney does something silly (again). All while you’re doing nothing! Rest is an integral part of any training program. Certainly, we know that inadequate recovery is responsible for a myriad of bad things, like decreased performance, and an increased risk of injury. But what about this thing called “active recovery”?

Active recovery can be loosely defined as a low-intensity activity (such as submaximal cycling or low-intensity weight training) used to enhance the recovery process between training sessions or competitions. The theory is that by increasing blood flow (your heart rate increases, therefore your blood is making more ’rounds’ as it were), lactate and other ‘waste products’ are cleared faster, thereby minimizing their detrimental effects in tissues. This should translate practically to a faster recovery than if your blood were moving at its normal velocity. This would mean that you could train more frequently at sustained or higher intensity levels without exposing yourself to the risks of inadequate recovery. Sounds like a great idea, eh?

A recent study however, puts this translation of theory to practice into question. Its scope is somewhat limited, but worth looking at.

Andersson, H., et al. Neuromuscular fatigue and recovery in elite female soccer: Effects of active recovery. Medicine and Science in Sport and Exercise. 40(2):372-80, 2008.

Before we even get into the guts of this study, you can probably tell that we are looking at two major limitations: 1) the results of this study are only generalizable to the sport of female soccer; and 2) elite soccer, at that. So, while this study does challenge the concept that active recovery is useful, it only challenges that concept in the context of elite female soccer players, which likely excludes most of you (It definitely excludes me, on all three levels).


Soccer is a high-intensity sport, but we don’t understand a lot about recovery, particularly after games, and particularly about female soccer players. Most of the studies to date have been either inconclusive or non-demonstrable in demonstrating many changes in the biochemistry of soccer players, despite an observed performance decrease after games. Active recovery has been studied in male soccer players, but not in female ones. Until now.

These researchers wanted to know a two things: 1) what happens neuromuscularly, and biochemically to elite female soccer players after a game, and 2) does the same things happen to them if they’re on passive or active recovery?


To answer this question, they recruited 22 elite female soccer players from the highest division in Sweden and Norway. Only 17 of these players were studied, because two of them were goal-keepers (and while a very difficult and demanding position, not the same activity profile) and three of the remaining 20 were not available for testing. These 22 players played two 90-minute friendly games, 72 hours apart. The same players participated in both games and played the same positions each time. After the first game, each player was randomly assigned to either passive or active recovery, with balancing for age, height, weight, VO2 max, and field playing position.

[With that many balancing factors, one has to wonder how random it actually was]

Active recovery consisted of 2 recovery sessions, at 22 and 46 hours after the first game (20 minutes of cycling at 60% of their peak heart rate, 30 minutes of low-intensity resistance training and 10 minutes of 60% cycling again).

Prior to the first match, players were tested for 20 meter sprint time, countermovement jump, maximal isokinetic knee flexion and extension and perceived muscle soreness. Blood samples were taken 3 hours prior to the first game, immediately after the first game, and then at 21, 45, and 69 hours after the first game and again immediately after the second game.

The blood was analyzed for creatine kinase (otherwise, known as CK, a general inflammatory marker), urea, and uric acid (both waste products).

All players wore heart rate monitors during their games, and each player was filmed for the entire game. These films were later reviewed to tabulate the intensity of the game. Distance covered, running intensity as well as time spent at each running intensity was calculated to ensure that the players weren’t slacking off when compared to one another.

All players were given a meal plan to attempt to standardize diet.


The data was analysed with multiple repeated-measures two-way ANOVAs, with the Dunnett as the post hoc test.

[That’s a lot of tests!]


Work intensity: The average heart rate was significantly higher within the two groups in game 2 vs. game 1. But it was higher in both groups, so the groups remained comparable.

Physiology after the first game: All performance tests were worse after the first game. And all three biochemical markers were elevated too.

Recovery time: Almost everything was back to baseline by 69 hours after the first game, regardless of which group the subjects were in. Sprint time was the first to recover (5h). Knee extension strength recovered by 27 hours, and knee flexion strength at 51 hours. Countermovement jump (similar to vertical jump) never recovered in either group in time for the second game. CK calmed down by 69 hours, while urea and uric acid returned to baseline by 21 hours. Muscle soreness was reportedly gone by 69 hours.

Between groups: The researchers failed to find a difference between the two groups at any time point. If we consider sprint time to be the major variable of interest, at 69 hours post-game 1, which was before game 2 at 72 hours, the 20m sprint time for the active recovery group was 3.25 seconds (SE 0.03) and 3.23 seconds (SE 0.04) in the passive recovery group.

What I liked about this study is that the researchers went out of their way to determine that the groups remained comparable throughout the study, hopefully recognizing that their randomization scheme might not be enough. What I also liked about this study was that they showed that there was a detrimental effect to performance and biochemistry after the first game. We are unable to say that these soccer players were SO elite that a single game was insufficient to cause performance decreases from which they would have to recover.

Limitations (or, why some of the limitations you might think are here aren’t):

The biggest limitations to this study are the ones I’ve already mentioned. You can’t really use this study to justify why _you_ (or I) should sit on a couch–unless you happen to be an elite, female soccer player.

There were definitely some reporting issues. I definitely wonder about the quality of their randomization. If you have to balance for 5 things between 17 players, it’s not going to be that random. How many choices are you going to have if you have to find another 22 year old, 5’8, 125 pound forward with a specific VO2 max? I suppose it’s possible that elite female soccer players might be all very similar to one another…

We don’t know anything about the blinding. And we definitely don’t know about adherence. The paper doesn’t mention whether the people doing the testing knew which group each player was in, though, with most of these measurements, you’d be hard pressed to bias one way or another short of deceptively entering a false number. However, we don’t know what “passive” recovery meant for the passive recovery group. Did they sneak off to do some passive recovery on their own? In some ways this isn’t a limitation, as it simply reformats the question to ask whether adding 2 sessions of structred active recovery aids in recovery from a game, as opposed to unstructured active recovery (which the non-active group may have done, on purpose or not).

However, adherence aside, the groups did remain comparable throughout the study for most of the variables we would consider as confounders. And ultimately, the goal of randomization is to create comparable groups.

One of the criticisms that I usually make is that there were multiple tests of significance. However, that’s only really a problem if you find you have a significant result and focus on it as though you had set out to look specifically for it. In this case, there were none between the two groups. So despite the fact that the chance of seeing a statistically significant difference by random chance was higher, they failed to detect one.

Lastly, one might think that 17 players is too small a sample size and that that quality makes this a bad study. Remember that statistical significance does not dictate whether an effect size is important or not. You use a statistic to bolster the argument that the difference between two groups (which you have deemed important beforehand) is not one that you got purely by random chance alone. However, the differences observed between the two groups were always miniscule. One could argue that statistical testing is unecessary for such numbers because even if they were statistically “different”, it wouldn’t be enough of a practical difference to justify one behaviour over the other.

The argument one CAN make with a sample of 17, however, is that these 17 players are somehow not an accurate representative sample of all female elite soccer players. I can’t speak to that, not knowing what female elite soccer players are like in general. Certainly, you could make a case that this study may even only apply to Scandanvian elite female soccer players (which, I found out to my embarrassment this past summer, does not include Icelandic elite female soccer players), if you could justify why other elite female soccer players from other countries are distinctly and substantially different than Scandanavian ones.

The bottom line:

Stricly speaking, you don’t really get to use this study to change anything you do, unless you’re a Scandanvian elite female soccer player. If you are, it might be okay for you to sit on the couch between games. For the rest of us, loosely, you can probably do whatever you like best, whether it’s sitting on the couch, or getting some active recovery in, feeling relatively assured that it’s probably not going to hurt you. But certainly, this study draws attention to question whether active recovery, though theoretically sound, is actually any more beneficial than passive recovery.

Click Here to view the Full Version of our Website