Nonlinear Dynamical Systems in Psychiatry

Comment on Thomasson

A most interesting paper. I am wondering if the author's are familiar with the work of Gottschalk, Bauer, Whybrow, (1995) Arch Gen Psych 52(11)947-59. These authors carried out a nonlinear time series study of 10 manic depressive patients observed over a period of, I belive, 2 years, together with a small number of controls. Their study was notabl for two reasons. First of all, they demonstrated that patients exhibited reduced tempora variability in mood although the extent of the variation in mood was greater than for controls. They suggested that patients were governed by a chaotic dynamics. Normal controls exhibited a marked temporal variability identical with that expected from a stochastic process, effectively a random walk. They also demonstrated that treated patients, although they had reduced magnitude of mood shifts, exhibited less variability than normal controls. This is important because it suggests that from a dynamical perspective, treatment does not restore normal mood function. Instead it merely dampens and constrains mood variability. The patient is controlled but not normalized.

The result of your paper is interesting because you demonstrate a decline in entropy of the EEG recordings over time and treatment, suggesting an increase in order in the EEG, which agrees with the increased order obsevered in the treated patients by Gotschalk et. al.
This suggests a physiological correlate to the behavioral changes previously observed.

Wed Dec 9