Complex systems aren’t a new concept – it is a well-established model when it comes to monitoring and interpreting the weather and economic worlds. But they are relatively new in the rehab sphere and seem to be catching some momentum as way of visualising things like pain, disability and even human movement.

This is great! You may have figured out by now that pain, movement and disability can be confusing as hell! We were once all led to believe that pain in a joint would go if it was mobilised, strengthened, shaved down or even replaced implying a causal relationship between pain and e.g. stiffness, weakness, cartilage thinning e.t.c.

When the realisation hits you that things aren’t all sunshine and rainbows when trying to help people improve their pain it can be a bit of a dark place. You see RCT after RCT come out showing no difference in group averages for exercise versus no exercise, different types of exercise or a specific rehab approach versus reassurance and free activity. Everything seems almost a little bit pointless because, well, how much of a difference do we actually make?

Part of this thought process may well be due to peoples interpretation of the studies either on a methodological standpoint or philosophical outlook; if you think RCT’s are the most important piece of the puzzle for individual treatment plans [both things we will look at in future blogs]. But part of this might be failing to appreciate the complexity of being human, let along being human with an MSK problem seeking care from other humans within complex socioeconomic situations and healthcare systems!

This is where a complex systems approach comes into its own. Underlined throughout a complex systems approach is the appreciation of a whole, huge host of variables that are linked with each other and contribute to the ‘emergence’ of certain phenomena. Cholewicki et al (2019) had a really good go at this in their recent paper [HERE], you may have seen this picture knocking about in a few places.

Before we really dive into this I think it is helpful to provide you with a visual that you can refer back to, as this stuff ain’t the easiest thing to wrap your head around! This diagram is from Bittencourt et al (2016), which is a paper I highly recommend you go and read at some point.

I want to start with the main, key component of any complex system; the ‘emergence’ of certain phenomena. In our MSK world, a common phenomenon that emerges is pain. It comes about from a huge number of variables and when these variables link together and hit just right… ouch, it can be fucking painful. Without getting too ahead of ourselves, ‘emergence’ is unpredictable, and it cannot be reduced down to just the sum of its parts! We have all had experiences of this in the past, in fact I have had one just recently as I am starting to take up jogging a bit more and last week was getting some horrible shin splints. Within a complex system framework, pain (that we diagnose as medial tibial stress syndrome) emerges from a huge web of variables; these can be my recent increase in mileage, my bony profile, my nervous system function, my conceptualising of discomfort / pain (in other words, what does it mean to me?), my footwear, my calf strength / endurance, how much sleep I had the night before, what my diet had consisted of in the previous month… the list can go on and on and on! If you take one of those things and make it a net positive, for example my mileage; reduce it to 80% of what I had actually ran in the previous week and there are scenarios where I still end up with shin splints, conversely, this week I have increased my mileage more [a kind of experiment if you will!] and my shins weirdly feel so much better than they have.

The emergent phenomena of pain has changed as has what disability it caused me, and these constantly change over time! It is wrong to view pain / disability as a black and white phenomena or a number out of 10, which is actually what all research papers start to make you believe.

Whilst we are talking about the variables that cause an emergent phenomena, it is absolutely crucial that we acknowledge the non-linear relationship. Here we go with strange terminology again hey? Well, let’s reverse engineer this a little bit.

A linear relationship (the opposite) assumes that the outcome or emergent phenomena is in some way related to the sum of its parts (variables). For example, if someone has the below traits and osteoarthritis pain was equal to the sum of it’s parts you would expect the variables to include a lot of the following:

• Had a busy job on their feet all their life

• Lived on a diet of pro-inflammatory foods, smoked and drank leading to a metabolic syndrome

• Sustained loss of cartilage from their hip joint leading to contact of subchondral bone surfaces and inflammation

• All of the above when added together will cause pain which we would diagnose as hip osteoarthritis.

Sounds like quite a nice model doesn’t it? That’s because our brains crave simplicity! Simple answers we tend to like, uncertainty and complexity… not so much. We all know people who tick all of the first three points above but not the fourth. These people punctuate the point that the variables don’t have a linear relationship. You can’t add them together and get a consistent outcome or ‘emergence’. Sure, some might be necessary for someone to have pain, but it doesn’t make it linear or simple!

Lets take it away from pain for a second and consider something else that I think is a really nice example and blows my mind every time I think about it… consciousness.

Within a complex systems approach, consciousness is our emergent phenomena. It comes about due to a complex interaction of many, many neurosciency things. Of these you can throw in 86 billion neurons, more than 40 different neurotransmitters, many cerebral arteries and the cells / nutrition that get carried along in the blood. This of course is nowhere near an exhaustive list but I think you get the point. It is very complicated, and even the world’s leading neuroscientists haven’t figured out how we have become conscious.

Let’s look at these variables in the web of determinants a bit closer; a neuron is critical to conduction of nerve impulses which in turn are important for brain function and consciousness but in and of itself… a neuron does not hold the power of consciousness. Neither does a neuron plus another neuron and some neurotransmitters; they are crucial to the overall emergent phenomena, but you can’t add them together and expect to have a linear process leading to consciousness.

Hopefully you are starting to see a pattern here; one that depicts relationships between variables as non-linear.

A next key feature of a complex system is that of feedback loops. By this we mean that the output of something becomes new information to be processed and becomes part of the web of determinants. Remember, if we are saying pain is the emergent phenomena it is important not to visualise this as a black and white thing of ‘Has pain’ or ‘Has no pain’; that is your cavewo/man brain trying to take over and you mustn’t let it! But to appease it a little bit let’s think of the emergent phenomena as the intensity of pain someone has.

Peggy has sciatica [fine, radicular pain… but this isn’t the thing to waste brain power on right now!]. Her sciatica is an emergent phenomenon of gazillions of variables ranging from how she perceived her mother reacted to pain as a child through to how strong the presence of neuroinflammation is within her nervous system. She has noticed that going for a walk more than half an hour really kicks off her leg pain and only gets better when she is able to sit down for a bit. This really is a kick in the teeth because Peggy absolutely loves to walk, but now she starts to get anxious when she leaves the house and avoids longer walks as a result out of fear of her leg pain becoming excruciating.

Peggy’s emotional suffering through anxiety and fear now enter back into the web of determinants which may affect the emergent outcome (which we have decided for this is severity of pain). The outcome / emergent phenomena have fed back into the complex system by way of circular causality. And this happens all the time, for all of us.

Have you ever had someone with back pain go to pick something up in the gym with you, and it actually was a lot easier than they anticipated? If we think about their behaviour, or kinesiophobia, then this violation of their expectations is likely to factor into their movement behaviours going forwards, it is a feedback loop. It would be great if this was a linear/predictable process going forwards, but we all know from experience there are those that will go on to do great after this… and those that will push themselves a bit too much and end up cursing the day they ever met you.

This brings us onto our final point about self-organisation. This is essentially why we have non-linearity (let’s just call this, you can’t predict what is going to happen); and links closely with everything we have mentioned above); self-organisation is the method through which spontaneous order is achieved within a system (you/me/us) and the overall behaviour of the system is no proportional to the individual behaviour of the variables.

All of these aforementioned characteristics bring about uncertainty. And uncertainty is pretty much the thing we are programmed to dislike, but also have to learn to live and get along with just like a disgruntled neighbour you don’t really like but have to be nice to and take the bins out for every now and then.

The reason I like working within a complex systems approach is that, if we are always considering the complexity and depth of a person’s presentation, then I don’t expect a few quad sets or massages are going to ‘fix’ their problem.

If uncertainty is the expected characteristic of a complex system, then the patients I see (as complex systems themselves) are probably going to have quite a lot of variety in their outcomes.

As our science in the field of pain and rehabilitation evolves, we may well see advances in treatment. For example, seeing that a change in certain variables can mediate (cause) a positive change in a patients presentation. Some mediating variables proposed already are psychological in nature (e.g. kinesiophobia) which is great, it gives us some guidance at least but how exactly we target that with the individual in front of us is going to differ a hell of a lot from patient to patient. Kinesiophobia (or whatever the variable we are targeting is) is probably an emergent phenomena of its own complex system where we HAVE to consider the persons own individual history.

Is all reductionism bad?

Huntington’s disease emerges from a fault in a gene that is responsible for a protein called huntingtin. The disease itself is both life shortening and can in the end stages have devastating effects on a person’s independence, cognition and quality of life. Applying a reductionist approach by zooming in on this faulty gene and the potential treatments for it is likely a positive thing as if this gene can be treated through gene therapy approaches, it could stop the disease altogether. In this case, the causal link between the faulty gene and the eventual disease is so strong that focusing on one area is necessary.

It would be reasonable to assume that certain variables will hold greater significance in certain MSK presentations also. But the difference is we very rarely will see single variables that have such a strong causal link to a condition, that are modifiable. We mentioned reductionism at the start of the blog in the form of focusing on just one aspect of treatment for an MSK problem.

For me, the difference between reductionism that is beneficial and reductionism that just oversimplifies the problem in the MSK world is the wider context that it is placed. Someone after an ACL reconstruction may be really struggling with their knee function in the return to play phase; here the emergent phenomenon is perceived knee function. Some of the variables in the ‘web of determinants’ may be some of the below [not exclusive]:

The number of decelerations / tight change of directions required by the sport per match

The level of play and how this influences the intensity

The length of the match

Pain

Biomechanical factors (Peak strength, rate of force development, endurance, range of motion)

Psychological factors (anxiety, kinesiophobia, confidence)

Social factors (Partner in the back of their mind saying ‘don’t you dare get injured again’ or the worry about not being able to play with their kids in future if they get injured again)

Within complex systems you will often hear the terms ‘zooming in’ and ‘zooming out’; this means, for example, we could take lack of perceived knee function as the problem and zoom in on the psychological reasons for this, complete and assessment or screening and then, if necessary, refer onto a sports psychologist. We could then zoom out to a wider perspective then choose to zoom in on the biomechanical factors associated with higher level sporting activity such as how much impulse the person can generate over short spaces of time or what the efficacy of their Quadriceps (which take a lot of the brunt of loading) is like.

This is technically a reductionist approach as we are homing in on one aspect of the problem. The difference between those of us who adopt a systems approach and those who don’t is that within the systems-based approach, all relevant variables are considered and even though the interventions may look the same, the reasoning behind this is more calculated. This matters!

Someone after an ACL-R can go to a physiotherapist struggling with the problem mentioned above and get quads strengthening exercises. They could attain the strongest quads in the game but still struggle with knee function and if the physio is applying a reductionist approach outside a systems approach… there could be not a lot else to offer them!

I truly believe the value of a good MSK therapist doesn’t 100% lie in the actual intervention that they give; the value comes from the knowledge and reasoning BEHIND giving the recommendation for treatment.

Something to bear in mind.

Wrapping it up

There we have a whistle stop tour in a complex systems approach. I recommend have a read of Bittencourt et al (2016) for sure and reach out with any questions or thoughts you might have on the topic! Whilst it might seem like a lot of terminology and hard-thinking for something you may already consider in clinical practice, I would argue a set framework is beneficial for us to discuss rehab ideas and to build upon as the years go by.

Thanks for reading,

Jeff