There has been a lot of opinions flying around regarding Michael Phelps’ use of cupping at the Olympics in Rio. Dynamic Tape developer and Australian Musculoskeletal Physiotherapist, Ryan Kendrick raises 10 key points regarding the validity, not of the interventions but of these opinions.
Is the cup half full?
There has been a lot of recent activity on social media regarding the use of cupping and also kinesiotaping given some of the high profile athletes using it at the Olympics. I don’t usually comment on such things but given the mocking tone in which these methods have been criticised I felt compelled to pen something.
Let me start by saying that personally I do not use either method. My treatment approach tends to be based on managing load, movement and function, and the interaction between them and trying to identify the various contributions from neural, muscular and articular systems, the pain processes at play, the various drivers and how they are driven and addressing these taking into consideration the patient’s aims, values, co-morbidities and the situation.
That said, I am not going to offer an opinion on the validity of cupping or kinesiotaping one way or the other. I do however want to offer some points for reflection regarding the validity of the opinions of those who see it as their duty to mock any approach.
1. While ‘research’ may show lack of a statistically significant effect, they are generally looking at a group effect. Within any cohort there are generally some subjects who do well, some who may get worse, lots who might not do much so statistically there is no significance. That does not mean that some people within the group do not benefit significantly.
2. The natural argument is that those who do respond significantly are responding to the placebo effect (whatever that is exactly). This may be true but it is worth noting that the placebo effect is not an imagined effect. It creates real, physiological changes. This may be reversal of ECG changes in some subjects undergoing sham cardiac surgery for example (Cobb et al, 1959). They are effects worth having. Hashish et al, 1988 when examining ultrasound intensities post wisdom tooth extraction showed the best response when the ultrasound was stationary and disconnected internally but both the patient and therapist believed it to be functioning. Not only was pain reduced but improvements in swelling and trismus were also observed.
It is also worth noting that an intervention is required to create this effect. No intervention = no effect. I am not suggesting that our treatment should be aimed solely on creating placebo effects with methods that we ‘know’ do not create significant therapeutic benefits but rather that we should be maximising the effect of the (previously considered) non-therapeutic aspects of our management during the administration of our evidenced based interventions. The work on communication styles and the doctor-patient interaction is interesting as is our understanding of the impact of beliefs, expectations, past experience, social and cultural factors.
3. We do not know the aim of the treatment in these situations. Michael Phelps might just like the feeling of the cups. He might find the whole process relaxing. The aim may not be to ‘cure’ anything. Point 1 or 2 above may apply. Having worked with athletes at the elite level it is very apparent that the smallest thing to us can have a seemingly profound effect on an athlete and its absence may result in them entering the sporting arena with less than 110% confidence. Things like staying in the same hotel or using the same locker as a previously successful year clearly don’t have a direct physiological effect but may be important for the athlete and his or her ability to reproduce their best performance. A certain intervention may in a sense be anchored to a certain physiological state.
4. It is very unlikely that the research that you are referring to was conducted on the athlete in front of you. As such, there may be other co-morbidities that confound or complicate the decision making process (and would have excluded them from any study). The situation is likely to be different as the research was probably not conducted on 20 x gold medal winning Olympic swimmers heading into another Olympic final (if it was you would have a subject cohort of n = 1 so might only be applicable to Phelps himself). We always need to combine what we can draw from the research with our clinical expertise and assessment findings again against the backdrop of the given situation, the patient in front of us and their aims and values. Does research on chronic tendon problems in 50 year old males apply to a female Olympic triple jumper with first episode, reactive tendinopathy two weeks out from the games? How would this vary again to a marathon runner with a degenerative tendinopathy?
5. The way research is conducted is very different to the clinical situation. In research, we bring in a group of subjects, do the same thing on them and look for an effect on the group. In clinic, we start with the end in mind. We know the outcome that we are looking for before we start our intervention and we continue to work or modify our intervention until we achieve it. Then we re-evaluate the influence that has had and determine whether it supports or refutes our primary hypothesis. This applies to our exercise, manual therapy or taping (at least when it comes to biomechanical taping for movement control). For example, I may be using Dynamic Tape to reduce the magnitude of collapse of the hip into adduction and internal rotation in a patient with patellofemoral pain. If I re-evaluate their gait or squat after the application and I have not achieved the result I was after I need to reconsider the technique. Perhaps, I did not have them in the shortened position such that there is no tension or resistance produced by the tape. Perhaps my line of pull was too high relative to the axis or too vertical in orientation resulting in absence of any force vector into hip extension, external rotation and abduction. Perhaps they just needed more force and a double layer PowerBand is required. We may have seen a good change in peak adduction and velocity but no corresponding change in the symptoms and this might suggest that this movement is not a major contributing factor in this individual. Similarly, our research subjects may vary in their needs but the technique is generally kept the same to reduce variables thereby not giving a true indication of its efficacy.
6. Does the person in front of us need this application? While some studies show a correlation between dynamic valgus at the knee and patellofemoral joint pain, it does not mean that everyone with patellofemoral joint pain exhibits that movement pattern or that it is a major factor in their case. We may have evidence to show this correlation, and that Dynamic Taping can reduce the peak adduction and the velocity of adduction but if the person in front of us is not collapsing in excessively, they do not need this taping application. Trying to use research to create some sort of box checking algorithm rarely works well. In many ways it seems that the focus on ‘the evidence’ has resulted in a reduction in the ability to assess and clinically reason, and many studies due to either methodological flaws or incorrect application and interpretation actually cloud our understanding.
7. Was the research conducted on people who need the intervention and how well were they classified? If we wish to examine the influence of Dynamic Taping on hip adduction and internal rotation of the hip but only a few of the subjects are actually dropping in excessively, we will get a large washout effect because many subjects do not need it, will not be moving into a position that creates tension and resistance to motion. This is like saying that we will see how effective the firemen are at putting out fires by measuring the temperature of all the houses in the city before and after they have been hosed. If only a few are on fire, we will not see much effect statistically because most will not change much. If however we only include houses that are on fire in our study, we will see a very big change in temperature. Our research and subject classification needs to be designed so that we are examining the houses that are on fire. The subjects all need the intervention being offered.
Just saying that subjects have had back pain for greater than three months and it doesn’t extend past their buttock does not make them homogenous. The lack of diagnosis (e.g. nonspecific low back pain) does not itself a diagnosis make. Doing exactly the same thing on 150 people in this group is unlikely to yield any statistically significant result and is not indicative of the merits of the intervention. This may account for the results that we see as described in point 1.
8. Do not just read the abstract and conclusions (for the reason above) as many are wide ranging and go far beyond what the study itself tells us. Even systematic reviews do this. Most reviews on kinesiotaping (and several manual therapy interventions) say that there are insufficient high quality studies to pool data to really draw meaningful conclusions. This is often interpreted as being evidence against it. That is not the case. They are just saying that there are not enough studies that look at the same patient population, same treatment technique, same outcome measures to be able to pool the data. Most are looking at e.g. two studies of the neck, three of the knee etc. In research our desire is to find something new and not reproduce someone else’s work but that is exactly what is required.
9. Conclusions are often not representative of clinical practice. For example several reviews on kinesiotaping suggest that there is a small, short term effect on pain relief. Even in reviews that show this they state that as the subjects are all the same at four weeks, there is no benefit. If we take the case of Michael Phelps going in to yet another final at the Olympics, he is not concerned with where he will be in four weeks. Short term pain relief could be of significant benefit and allow him again to enter the arena at 110% confidence. It may mean that your patient requires less analgesia in the first four weeks – another benefit. It may mean that there is less pain inhibition, less compensation strategies, fear of movement etc. during the first four weeks which although the group effect is similar at four weeks, may be a factor in a small percentage of people going on to chronicity.
10. Our understanding has a way of going full circle or at least doing 180º turns. Not long ago, people would have argued intensely that eccentric exercise was the way to go for tendinopathies (again incorrectly throwing all tendinopathies in the same group). Ebony Rio and co. are now showing great pain relieving effects with isometric exercise. People (through extrapolation and misrepresentation of research on transversus abdominis and multifidus) would have argued that this was all we needed to do for all of our back pain subjects. Again, the original research never suggested that but at least now there is greater understanding of where and when this may fit in. We have better classification of subjects thanks to some of the work of Peter O’Sullivan for example. This also highlights another area that has done an about turn. Many people in the past have been told by well meaning friends, doctors or therapists that they must not bend their backs as this is damaging or potentially damaging. We were not made to function in a neutral spine position all day, every day. Backs are made to bend and the failure to re-educate this or reinforcing the concept that bending is dangerous may lead to fear avoidance behaviours, splinting and muscle guarding and perhaps set up peripheral drivers that help to maintain abnormal processing and pain outputs. The way that the great pain physiology research is misinterpreted and propagated is the subject of an entirely separate soapbox session altogether.
In summary, our understanding is not all encompassing. I, for one, do not believe that I understand everything (sometimes anything) with regards to this field and certainly not enough to dismiss all interventions for all people at all times, particularly if those interventions do not have any significant adverse effects. This is where the evidence (properly developed, investigated and interpreted) needs to be combined with the clinical expertise of the clinician, a thorough and relevant assessment and giving due consideration to the patient and their particular circumstances.
Sometimes, cupping just might feel good and that might be the aim. If you are selling it as a cure for cancer, that might be a little different.
Some people these days are very quick to criticise things that have been done in the past but are doing so with the benefit of hindsight, hindsight that has only been created by the work of those whom they now seek to criticise (and whose inquiry required more effort than a quick google search i.e. going to the library, scrolling through the microfiche, searching the aisles for the journal, photocopying hundreds of pages before you could even start to read them).
Academic discussion and debate when conducted professionally, respectfully and constructively is always welcome. However, shocking, mocking and inappropriate posts often based on someone’s own opinion and handpicking a supportive study or part of a study seem to be less about finding solutions for our patients and more about gaining notoriety and twitter followers.
For more on thought provoking perspectives on the current direction of EBM visit Evidence based medicine: a movement in crisis?
Ryan Kendrick BPhty, MPhtySt. (MSK)
Ryan earned a Bachelor of Physiotherapy from the University of Queensland in 1994 and a Masters in Musculoskeletal Physiotherapy from the same university in 2000. He has worked extensively in elite sport including forming part of the ‘Dream Team’ responsible for the resurrection of former world number 4, Greg Rusesdki’s career following a string of injuries. Ryan has also worked as Physiotherapist for the Essex County Cricket Club in the UK and with several professional, Commonwealth and Olympic athletes. His private practice work focused on chronic musculoskeletal pain, primarily spinal. He has been clinical tutor on the Bachelor and Masters programme at Griffith University, Gold Coast, Australia. Ryan developed Dynamic Tape to allow the genuine input of an external mechanical force into the kinetic chain (without limiting range) to permit a biomechanical approach to managing load, movement and function. He has taught in over 20 countries.