A systems approach to schizophrenia
Department of Applied Mathematics, University of Colorado at
Department of Applied Mathematics, University of Colorado
Last modified: September 19, 2007
We hypothesize that schizophrenia, a psychiatric condition with complex symptoms and dramatic impact on individuals and society, is a disease of dysregulation. To test our theory, we recorded clinical data for both schizophrenic patients and controls (brain activations and a wide variety of autonomic and endocrine variables), and researched for each individual the evolution in time of the system consisting of these measures.
Our approach is to study the regularity in this system as a whole, or as a collection of particular smaller units (such as the six-dimensional limbic subsystem, or the cardiac component), using measures such as approximate entropy and Lyapunov exponent. We use phase-space diagrams and recurrence plots to investigate the apparent manifestations of stability and sensitive dependence: a small perturbation seems to produce no change in a healthy subject, but can spin effects out of control in a vulnerable system with a pre-existing dysregulation. Under perturbation, the time-evolutions of controls appear to be attracted to stable cycles, while those of patients spin away chaotically. This suggests the presence of an “attractor” in the control systems, while in patients attractors either do not exist or have a more complicated geometry.
Chaos is difficult to detect in biological systems due to their non-stationary behavior, the large amount of noise and measurement limitations. But the struggle is worthwhile: if a system is chaotic rather then random, its behavior can be controlled by small perturbations. Identifying a diseased system as chaotic could facilitate a well-defined, etiology-based, early diagnosis, and could potentially suggest more efficient treatment options.