A systems approach to prefrontal-limbic dysregulation in schizophrenia

Abstract

Introduction: Using a prefrontal-limbic dysregulation model for schizophrenia, we tested whether a dynamic control systems approach in conjunction with neuroimaging might increase detection sensitivity in characterizing the illness. Our analyses were modeled upon diagnostic tests for other dysregulatory diseases, such as diabetes, in which trajectories for the excitatory and inhibitory components of the negative feedback loops that reestablish homeostasis are measured after system perturbation. We hypothesized that these components would show distinct coupling dynamics within the patient population, as compared to healthy controls, and that these coupling dynamics could be quantified statistically using cross-correlations between excitatory and inhibitory time series using fMRI.

Methods: As our perturbation, we activated neural regions associated with the emotional arousal response, using affect-valent facial stimuli presented to 11 schizophrenic patients (all under psychotropic medication) and 65 healthy controls (including 11 individuals age- and sex-matched to the patients) during fMRI scanning. We first performed a random-effects analysis of the fMRI data to identify activated regions. Those regions were then analyzed for group differences, using both standard analyses with respect to the time series peaks, as well as a dynamic analysis that looked at cross-correlations between excitatory and inhibitory time series and group differences over the entire time series.

Results: Patients and controls showed significant differences in signal dynamics between excitatory and inhibitory components of the negative feedback loop that controls emotional arousal, specifically between the right amygdala and Brodmann area 9 (BA9), when viewing angry facial expressions (p = 0.002). Further analyses were performed with respect to activation amplitudes for these areas in response to angry faces, both over the entire time series as well as for each time point along the time series. While the amygdala responses were not significantly different between groups, patients showed significantly lower BA9 activation during the beginning of the response (0.000<or= p<or= 0.021) and significantly higher BA9 activation towards the end of the response (0.008<or= p<or= 0.025), suggesting longer time-lags between patients' excitatory responses and the inhibitory activation that modulates it.

Conclusions: Our results capture a significant dysregulation between the excitatory (amygdala) and inhibitory (prefrontal) limbic regions in medicated schizophrenic patients versus healthy controls. They suggest that, analogously to diagnostic tests used in other physiological diseases, quantifying dysregulation using a control systems approach may provide an appropriate model to investigate further in developing presymptomatic neurobiological assessments of risk, or illness severity in symptomatic patients.