The mystery of the taped ankle…

I’m back from a one-month vacation in Sweden, and thanks to a horrible delay at the Frankfurt airport was able to get a review done for this week. Thanks for being patient with me and as soon as I install the rest of my regular software on this new hard drive, I’ll be fully up and running…if only I could remember which box I packed it all in…

This week, I give you a review on a study about ankle taping.

It’s not often that I see a study whose main intention is to investigate a placebo effect. Athletic taping has long been disputed in terms of whether it actually has a physiological benefit with respect to injury prevention. The body of evidence in the literature tells us that taping does, in fact, reduce the incidence of injury, but is unclear as to the mechanism by which that happens.

This study aims to look at one of those possible mechanisms, while at the same time, investigating whether there are any actual physiological parameters that can be measured to show that ankle taping alters the way in which specific tasks are performed. The premise of this study is that ankle taping, in fact, doesn’t have any physiological benefits, but that the way it might prevent injury is, in fact, the psychological effect that believing that something will help you, will, in fact, help you.

In browsing the study, I think there are only two obvious weaknesses to it, but I guess we’re going to find out.

Sawkins K, Refshauge K, Kilbreath S, and Raymond J. The placebo effect of ankle taping in ankle instability. Medicine and Science in Sport and Exercise, 39(5): 781-787, 2007.

Introduction:

I think one of the strengths of this paper is the acknowledgment of the preceeding literature. We know that ankle taping does reducing ankle injuries, but we don’t understand why. One of the theories is that the athlete’s expectation that the taping will help them, is sufficient to change their performance such that it does.

This study differs from most randomized control trials in that the authors used deception to blind their subjects. This is not an often-seen tactic, but can be very useful in isolating the effect of interest. In this case, we are interested in seeing whether “real” taping is different from “placebo” taping in terms of objective testing and in terms of subjective reporting of experience. To isolate this difference, the subjects needed to have the same expectation of the two taping methods. This way, everything stays the same, including the psychological factors, and any differences we observe between the taping methods should be attributable to the taping methods themselves.

Methods:

It’s not entirely clear where the subjects in this study came from. Presumably, they were recruited at the University of Sydney. People were eligible for the study if they had ankle instability from previous ankle sprains. Ankle instability was defined by a cut-off score on the Cumberland ankle instability tool (CAIT). People were excluded from the study if they had had an ankle sprain within 3 weeks of testing, if they had a past history of fracture or surgery to the lower limb, or if they had pain or palpable swelling of the ankle, or if they had any neurological, visual or vestibular deficits.

[I’m not personally familiar with the CAIT, and I’m a bit nervous with the reference not being a specific validation of the CAIT. However, even if the CAIT is not valid and there were some people with stable ankles who were included in this study, it might not change the actual conclusion that we can make from this study.]

Subjects were not informed as to the true intent of the study–only that the researchers were interested in comparing two types of taping–a mechanical one and a “proprioceptive” one. They were not aware that the relabelled “proprioceptive” taping method was, in fact, a placebo.

This study was a cross-over randomized control trial. That is, each subject acted as his/her own control and the order in which treatments were used was random.

[Taping is a great example of a crossover trial that works because there’s no real carry-over from one treatment to the next. The effects of one type of taping disappear once the tape is removed and there’s no need to have an extensive “wash-out” period between treatments. Within-subject trials have two major benefits: 1) You don’t need as many subjects (in fact, you need only half the number you would normally need!) and 2) there is less variability between treatment groups because whatever intrinsic variables that exist are the same between the groups because the person is the same. This means that you have the ability to either detect a smaller (and hopefully practical) difference between two treatments (you have better resolution), or you can detect a larger difference with a smaller number of people.]

Subjects were tested in three conditions: mechanical taping, “proprioceptive” taping and no tape. They did 2 physical tests and a questionnaire.

The first physical test was the hopping test, in which the subject had to hop onto 8 squares, one at a time and then back. Four of the squares were level, 1 square had a 15 degree incline, 1 square had a 15 degree decline, and 2 squares with a 15 degree lateral inclination. The outcome measured with the hop test was the time it took to complete the 16 hops (8 forward, turn around, 8 back). There was a time penalty of 1 second if a subject landed outside a square, or if they did the square in the incorrect sequence (the squares were not arranged in a line).

The second test was the modified star excursion balance test. The setup for the this test is a star, made up of 4 strips of tape–two strips to form a cross, and 2 strips to make up the X through the cross to form an 8-ray star. Subjects started the test by standing on both feet in the middle of the star. They were then instructed to stand on their test leg, and to move the other leg along one of the rays as far as possible without losing balance, or touching down with the non-test foot. Start and finish positions had to be maintained for at least 1 second. The outcome measured in this case was the distance travelled by the non-test foot. Three measurements were take (one for the anterior ray, one for the lateral ray and one for the posterior ray.)

The questionnaire asked the subjects how they felt during the tests under the different taping conditions. It asked how stable they felt, how confident they felt and how much reassurance the taping provided. Stability referred to how steady and controlled the subject felt. Confidence referred to how well the subject felt they could perform the test and reassurance referred to how confidence they were that they could do the test without spraining their ankle.

Taping

The mechanical taping was a standard ankle tape job: An anchor, three stirrups, a low anchor, a figure of six and a heel lock. Subjects were told that this method of taping increased mechanical support to stabilize the ankle, and that it would help them perform better in the tests.

The proprioceptive taping was a single strip of tape (10 cm long) applied on the lateral aspect of the leg just above the lateral malleolus. It was aligned with the peroneus longus. Subjects were told that this method of taping would increase cutaneous input and improve proprioception to prevent ankle sprains, and that it would help them perform better in the tests.

Subjects were additionally blindfolded during taping and the actual taping was covered up by a cloth skirt to prevent the subject from seeing the tape job.

[Why they did this is not _entirely_ clear, as you can presumably feel when one or more strips of tap are applied and whether you can see your ankle or not probably doesn’t affect whether you can feel like there’s one or more pieces of tape on your leg. But okay, I can accept that they did it.]

Statistics:

The taping methods were compared using a one-way ANOVA for the hopping test, and a two-way ANOVA for the modified star test. The questionnaire was qualitatively analyzed to identify whether subjects felt improvements in any of the three categories (stability, confidence, reassurance) and then a chi-square test was used to see if the rate at which people reported improvements was different between the two taping methods.

Results:

The researchers failed to detect a significant difference between groups with respect to their primary outcome–the Hopping test. When using the real tape job, subjects completed the test in 10.5 (SD 3.6) seconds. When using the placebo tape job, subjects completed the test in 10.5 (SD 3.6) seconds. And when using no tape at all, subjects completed the test in 10.5 (SD 3.7) seconds.

The authors also did not detect a significant difference on the modified star excursion test. There were some differences detected within subjects, between excursion directions, but examination of the actual numbers shows that despite detecting a statistical difference, the actual difference was millimeters of difference, and thus, it is not meaningful to say that one tape job produced a relevant different effect that the other.

What is interesting is that despite the lack of objective differences between groups, almost all of the subjects reported feeling more stable during the hopping test (29 out of 30 subjects), more confident (24 out of 30 subjects) and more reassured (23 out of 30 subjects) with the real tape job, compared to only 8, 7 and 7 subjects reporting better stability, confidence and reassurance respectively with the placebo tape job.

This pattern was mirrored in the modified star excursion test, but to a lesser extent.

Discussion:

The authors spend a bit of their discussion talking about how the star test was limited because they didn’t adjust for leg-length, but I’m not nearly as concerned about this, because the modified star test isn’t fully validated and wasn’t their main outcome of interest. The modified star test also tests stability during a static activity, and thus is less useful in drawing conclusions about things like injury prevention during activity or dynamic activity (which is most often when ankle injuries occur in sport.)

What we have here is a study that shows that the placebo effect of ankle taping in terms of improving _feeling_ of performance is incredibly high, but that it is most definitely a placebo effect because it doesn’t actually alter the performance of the subject when they have no tape on at all (subjects performed the same, regardless of whether they had tape of any kind on or not.) This is in the context of a previously injured population as well, in which we would expect to see differences if they existed. So, the question is, whether the power of belief is SO strong, that it actually prevents injuries despite not changing any of the physical factors of performance, and by what mechanism this might occur, because intuitively, this makes no sense. It would be almost like saying if you believe a helmet will protect you from injury in a parachuting accident, it will.

I think what I would have liked to have seen in this paper was an analysis of actual within-subject differences. Although the means are the same between the two tape jobs, it’s possible that they appear that way because some subjects performed with such drastic difference that they evened each other out. The mean of thirty 20’s, is 20. The mean of fifteen 30’s and fifteen 10’s is also 20. Obviously the standard deviation is totally different (this is an example for drama purposes), but it is possible to have a similar spread (which is what the standard deviation is a measure of) with large differences in performance.

However, it still doesn’t change the fact perception of confidence, stability and reassurance were so drastically different. So basically, if anyone is thinking of inventing a new method of preventing ankle injury and comparing it to taping, they’re going to have to overcome a MASSIVE placebo effect to derive a bigger benefit in terms of perceptions of security, but perhaps not in actually improving stability.

One last component missing from this paper is a comparison to people who didn’t have ankle instability at all. What this paper lacks is a bit of context (which may be in the preceding literature). How did the ankle-unstable subjects compare with their healthy counterparts with respect to the Hopping test?

The bottom line:

This is a fairly strong paper showing that ankle taping improves the feeling of stability, but actually doesn’t improve athletes performance in tests designed to test ankle stability in both dynamic and static activities.