We would like to thank Drs Krystal and Greenside for their interest in our recent article1 and for their comments, which further elucidate the caution that must be applied when interpreting data in this fashion. They raise an important issue regarding the analysis of the power spectra in our article. Our presentation of the power spectra in Figure 4 of the article was in a log-log coordinate system, and the parameter α of the power-law relationship (1/fα) was a natural means of quantifying the increase in power that was clearly present toward the lower frequencies (f). We did not present any material that considered alternate models for these power spectra. Krystal and Greenside point out recent work2 where it is asserted that the presence of a finite region of exponential decay (e−kf) in the power spectrum of a process is sufficient to characterize the process as one with deterministic chaotic dynamics.
Gottschalk A, Bauer MS, Whybrow PC. Low-Dimensional Chaos in Bipolar Disorder?—Reply. Arch Gen Psychiatry. 1998;55(3):275-276. doi: