The central message of complex-systems theory is that migraine is an inherently dynamical disease — a network of interdependent events with tipping points. That reframing changes what we should look for, and when.
Ask most people what a migraine is and they will describe the pain. Ask a physicist and they will want to know about the transition: the moment a brain tips from a normal state into an attack. Migraine is, in the language of complex systems, a dynamical disease — its symptoms are not a fixed defect but the behaviour of a network of interdependent processes crossing a threshold.
The climate analogy
The most useful comparison turned out to be an unlikely one: climate. Large dynamical systems — the climate, an ecosystem, a brain — can sit in a stable state for a long time and then, near a tipping point, flip abruptly into another. Just before such a transition, the system tends to give itself away. It becomes slower to recover from small perturbations; its fluctuations grow and become more correlated. These early-warning signals have been used to anticipate abrupt shifts in ecosystems and in the climate record — think of the anomalous, unusually severe winters that punctuated the 2000s as a system straining near a threshold.
We proposed that the premonitory symptoms of migraine — the yawning, the food cravings, the mood and concentration changes that many patients feel hours before pain begins — play the same role as those early-warning signals. They are not the disease; they are the sound of a system approaching its tipping point. And migraine pain driven by central sensitisation can be described, by analogy, as an abrupt overturning — a change in the traffic of the pain-signalling network rather than a simple “too much input.”
Dynamical network biomarkers
If that picture is right, it has a concrete payoff. Instead of hunting for a single molecule or a single misbehaving region, we should look for a dynamical network biomarker: a small set of signals whose collective statistics — their variance, their correlation — shift characteristically as an attack approaches. The most upstream events in an episode are the ones most likely to reveal root causes, and here human data has to replace the speculative extrapolation from animal models. This is the theoretical foundation a collaboration with Kazuyuki Aihara’s group in Japan set out to build.
The practical dream is early, personalised warning: not “you have migraine,” but “your network is drifting toward a transition — now is the moment to act.” That is a very different, and much more hopeful, way to think about a chronic disease.
References
- Dahlem MA, Rode S, May A, Fujiwara N, Hirata Y, Aihara K, Kurths J. Towards dynamical network biomarkers in neuromodulation of episodic migraine. Translational Neuroscience 4: 282–294 (2013).
- Dahlem MA, Kurths J, Ferrari MD, Aihara K, Scheffer M, May A. Understanding migraine using dynamic network biomarkers. Cephalalgia (2014).
A full publication list is on Google Scholar ↗.