Complex Systems and Running Injuries: Part 1 - We're Trying to be Less Wrong

Running injuries are multi-factorial in nature. We hear this statement touted constantly and read it at the start of every paper. It's worth showing these graphics again (which I frequently share in blog posts and presentations) because of how striking it is to visualize what "inputs" are involved for the "output" of a specific injury:

Bittencourt et al. 2016

The Web of Determinants model, where the circles represent both intrinsic and extrinsic risk/protective factors for injury. Factors are weighted differently and interact with each other in different ways.

Bertelsen et al. 2017

The Bertelsen model, where intrinsic and extrinsic factors relate on a session by session and stride by stride basis.

If these models look complicated... you're right in thinking so. It becomes apparent pretty quickly how reductionist it is to put your finger on one factor and think you've found the answer. Now, that said, these are models. Models have flaws. In fact, a very smart statistician has some great insight on that:

"Since all models are wrong the scientist must be alert to what is importantly wrong. It is inappropriate to be concerned about mice when there are tigers abroad." - George Box, 1976

Science is the process of trying to be less wrong. If we minimize how wrong we are, we maximize our ability to help others. From the clinical perspective, to find the tiger is to identify the least-wrong running injury etiology model, informing our interventions that reduce injury risk. There's a lot of steps involved there. The process looks something like this: 

This takes an incredible amount of time and resources. A fraction of each of those stages could be the life's work of a researcher. Besides time, an even bigger issue is that with each extra variable added to a study (look at the models above... there's a lot of variables), we need larger sample sizes to decrease our chances of making inappropriate statistical conclusions. It's a big task to take on, even if you have access to the Garmin data of 20,000 runners. Enter, an alternative approach:

Complex systems modeling

Bender 2001

There are a lot of things that need to be defined before talking about their applicability, so the rest of this post will be a run-down on what complex systems are with practical examples to make it a little more digestible.

Complex systems:
Networks of interacting variables that are notoriously difficult to predict. Bittencourt describes key characteristics of the complex system as having "self-organisation, non-linearity and emergent properties."

Common example: an ecosystem.

Emergent properties:
An unpredictable characteristic that arises from the interactions of the components in a system.

Common example: secondary colors; green is produced through mixing blue and yellow.

Self-organization:
Creating spontaneous order within a system. The process that creates emergent properties.

Common example: motor learning; acquiring a concrete skill through self-regulation.

Nonlinearity:
The input of the system isn't directly proportional to the output. Large or small changes in specific variables may not predictably affect the end outcome.

Common example: acceleration; time increases as a constant and velocity increases disproportionately.

I get it, at face value it's hard to imagine how this applies to running. It does end up making sense, though. Running injury etiology is a complex system. It's a network of interacting variables (training characteristics, running biomechanics, genetics, etc) that sends you out of a loop with a sore Achilles tendon and no individual explanation for how you got there. Reducing the risk of running injury requires understanding how we got there in the first place and developing realistic prevention strategies. Complex systems modeling can help reconcile our current research methodology with a more theoretical approach. As for the reductionist approach?

Mazzocchi 2008

Yeah, I was going to say that at the same time, if not a little bit before he did. Part two of this post will be specifically on how complex systems are being used in running injury research. If you want to know more until then, follow Adam Hulme (his Twitter handle is literally "@system_complex") and read some of his work.

Sorry. I have a headache too.

Jason Tuori, PT, DPT, CSCS



References:

1. Bittencourt NFN, Meeuwisse WH, Mendonça LD, Nettel-aguirre A, Ocarino JM, Fonseca ST. Complex systems approach for sports injuries: moving from risk factor identification to injury pattern recognition-narrative review and new concept. Br J Sports Med. 2016;50(21):1309-1314.
2. Bertelsen ML, Hulme A, Petersen J, et al. A framework for the etiology of running-related injuries. Scand J Med Sci Sports. 2017;27(11):1170-1180.
3. Bahr R, Krosshaug T. Understanding injury mechanisms: a key component of preventing injuries in sport. Br J Sports Med. 2005;39(6):324-9.
4. Finch C. A new framework for research leading to sports injury prevention. J Sci Med Sport. 2006;9(1-2):3-9.
5. Ioannidis JP. Why most published research findings are false. PLoS Med. 2005;2(8):e124.
6. Nielsen RØ, Bertelsen ML, Ramskov D, et al. The Garmin-RUNSAFE Running Health Study on the aetiology of running-related injuries: rationale and design of an 18-month prospective cohort study including runners worldwide. BMJ Open. 2019;9(9):e032627.
7. Renfree A, Casado A. Athletic Races Represent Complex Systems, and Pacing Behavior Should Be Viewed as an Emergent Phenomenon. Front Physiol. 2018;9:1432.
8. Mazzocchi F. Complexity in biology. Exceeding the limits of reductionism and determinism using complexity theory. EMBO Rep. 2008;9(1):10-4.
9. Hulme A, Mclean S, Salmon PM, Thompson J, Lane BR, Nielsen RO. Computational methods to model complex systems in sports injury research: agent-based modelling (ABM) and systems dynamics (SD) modelling. Br J Sports Med. 2019;53(24):1507-1510.