Massage Research: Are We Asking The Right Questions?

By Jenny Richardson

What do YOU think massage research should focus on?

  • If massage helps people with back pain?
  • If massage helps people with depression or anxiety?
  • If massage can help ease morning sickness?
  • Something else?

There are so many things that we as therapists think that massage can help to some degree.

Overall it gets worded in terms of “effectiveness”, “efficacy” or “compared to” but there are a lot of studies now looking at whether massage helps “x”.

The answer from research is “not much”. That varies, based on the study, for a whole range of reasons, but generally the clinical research isn’t singing our praises. It’s not “no” but it’s not the resounding “yes” that we think it should be. Yet our clients still come, often saying massage is the only thing that helps them.

We all know that our clients come to us for a range of reasons, and a large proportion of them leave us feeling better than when they walked in. I’m not going to even start on the question of why that might be. There are lots of theories about that. Whatever the reason is that massage “works”, shouldn’t that still be showing up more clearly in clinical trials?

What if we’re not asking the right questions in research? What if it’s the questions that are the problem, not the poor evidence?

Research usually starts with a question. A big part of the initial preparation for a research study is making that question very clear and very specific. Everything that happens after that is based on “answering” the question with evidence.

If a study asks “does massage help back pain”, I immediately think “well, what do you mean by massage?” What I would do is different to what Sharon would do, which is different to what Dave would do. Two main things then happen in studies:

1. “Massage” is applied. There is no description, so we don’t really know what the therapist did or, for that matter, what training the therapist had. Perhaps to the researchers “massage” means just rubbing the sore area but we don’t know. So how do we know what it is that did or didn’t work?


2. A specific protocol is outlined for the therapist to follow e.g. 5 strokes of effleurage, followed by two strokes of kneading, followed by 1 minute of tapotement on a specific location or over the whole body. Is this how you do massage? It’s not how I do massage. If someone with back pain/depression/morning sickness comes in, I customise what I do for that person. I would guess almost all therapists who have been practising for a little while do NOT do the same thing for every client.

I am going to go out on a limb and suggest that everyone with, for example, back pain has something different going on (which is why massage therapists customise their treatments). This complicates things further. A protocol that does work for one person probably won’t work for the next person. That’s not because the protocol fails but because each person needs something different. However, another huge issue in research is that we try to treat a symptom as having the same cause in all cases. One day I’ll write about that too.

Here’s the dilemma in research: in clinical studies, the aim is to control as many factors (“variables”) as possible, while changing just one thing to see if it has an effect in some measurable way. This is called “internal validity” i.e. can the researcher be sure they are measuring the effect of that one thing because everything else is accounted for. BUT in studies that don’t involve a test tube i.e. in all massage research, the greater the internal validity, the less like real life the experiment is.

On the other hand, real life has many variables. Designing a study that can be relatable to real life allows it to have “external validity” i.e. the researcher can have a better idea that what they find in the study will also be true in real life.

Within an hour long massage session, very few therapists would do just one thing, in one body location. So testing one thing might tell us whether that works on its own (on the assumption that all clients with that symptom are the same) but it doesn’t tell us whether it works as part of a whole range of things that happen. Those things include physical treatment, interaction, environment, education etc.

The more research I read, the more I believe that we, as a massage community, need to think carefully about the research we do to make sure that the research questions make sense in the first place.

Not all researchers are massage therapists themselves, therefore they don’t know what MTs actually do. We don’t all do the same thing. Maybe that might be the first set of questions we need to ask.

  • What do massage therapists do?
  • How do they decide what to do?
  • Which clients benefit (can we figure out which part of the treatment gave that benefit or why some clients benefit and others don’t)?
  • Are some massage therapists more effective than others (given a specific client “issue”)? Why? What do they do differently? Is that different for different issues (eg back pain vs morning sickness vs anxiety)?

So many big picture questions that would help provide guidance for what needs to be considered in clinical research.

“Does massage help” is not an unreasonable question and it’s one that the medical system, and clients, need answered. As a tightly controlled study, it doesn’t really make sense. There are too many variables that can never be controlled and when some of them are controlled, we lose the reality of a massage therapist working with an individual client.

There are other forms of research that suit this better, in my opinion. There was a study published in 2017 “Real-world massage therapy produces meaningful effectiveness signal for primary care patients with chronic low back pain: Results of a repeated measures cohort study“ (Elder et al)1 that used a research design that let massage therapists treat in the way they usually would. Not surprisingly, the results were strongly positive.

Recruiting guinea pigs for massage research.

An early study – “Development of a taxonomy to describe massage treatments for musculoskeletal pain” (Sherman, Dixon, Thompson & Cherkin, 2006)2 – allowed massage therapists to determine the treatment within some guidelines but asked them to record what they did so that researchers could evaluate what happened in a session. These studies do not give experimental evidence but they do give a much clearer answer to “does massage help”. Then it becomes a separate set of questions to find out what helps and why it helps. This might be where experimental research is more useful.

Maybe we could start to understand the aspects of massage that are important to help us with decision-making. We all have opinions but there are very few studies to help us know more about things like:

  • What do different levels of pressure achieve?
  • Is there any measurable tissue change if you stay on a spot for 5 seconds or for 60 seconds?
  • How far apart should sessions be if someone has “x” or “y” or “z”?
  • How do pain and the amount of pressure relate in massage, and does that mean anything in terms of treatment?
  • Does the angle of pressure have any effect?

We’re still going to run into some of the same problems of the dichotomy between internal and external validity but if we’re trying to understand something and gain insight into each variable rather than come up with evidence for/against something, maybe more specific questions will work better in a tightly controlled study. This may help us ask the broad questions because we know what variables have what effects.

To me, asking “does massage help back pain?” is a bit like asking “does seeing a doctor help back pain?”

“Massage” has too many options, aspects and nuances to be a useful concept for a controlled experiment. If it is defined as a particular thing, it can’t be generalised to all massage. A big picture question needs a big picture research approach, with the understanding that it will raise more questions than it answers. And that’s ok. Those other questions are perhaps the ones we need to be asking sooner rather than later.

What would you like to see explored through research? Do you have ideas on how that could be done in a way that is good research but still reflects “massage” as we know it?

  1. Elder, W., Munk, N., Love, M., Bruckner, G., Stewart, K., & Pearce, K. (2017). Real-world massage therapy produces meaningful effectiveness signal for primary care patients with chronic low back pain: Results of a repeated measures cohort study. Pain Medicine, 18(7), 1394-1405. doi: 10.1093/pm/pnw347
  2. Sherman, K., Dixon, M., Thompson, D., & Cherkin, D. (2006). Development of a taxonomy to describe massage treatments for musculoskeletal pain. BMC Complementary And Alternative Medicine, 6(1). doi: 10.1186/1472-6882-6-24
About the Author

Jenny Richardson is the owner of Canberra Myotherapy and has been practising massage and myotherapy for over 12 years. Jenny is passionate about understanding how the body works and using this to help clients with chronic pain. On top of this, Jenny is a current member of the AMT Board. Jenny is speaking on “Massage Research: Are We Asking The Right Questions” at the 2018 AMT National Conference on Saturday 13 October. Book your conference spot today.

Share this post:
The curious incident of the space suit in the storeroom
Announcement: Launch of the AMT mentoring scheme


  1. Great article Jenny.

    • Sharon Livingstone
      16/08/2018 - 2:41 pm

      We agree, Jeff! Can’t wait to hear Jenny expand on this topic at the AMT Conference.

Leave a Reply

Your email address will not be published / Required fields are marked *