Diffusion tensor imaging of anterior commissural fibers in patients with schizophrenia

Abstract

Introduction: Alterations in white matter connections in schizophrenia have been investigated using diffusion tensor imaging (DTI). There is also evidence from post-mortem studies as well as from magnetic resonance imaging morphometry studies that the anterior commissure (AC) might be implicated in schizophrenia, but no studies, to date, have investigated the AC using DTI or tractography.

Method: DTI scans were analyzed from 25 patients and 23 controls. Mean fractional anisotropy (FA) and trace were measured from the AC tracts. SANS and SAPS were used to evaluate clinical symptoms, and the Iowa Gambling Task, related to decision making, was also examined.

Results: Results revealed a significant decrease in mean FA and a significant increase in mean trace of AC tracts in patients compared with controls. In addition, patients, but not controls, showed a negative correlation between age and AC integrity. Statistically significant positive correlations were also found between AC FA and total positive symptom score. Decision making was negatively correlated with FA in patients on the Iowa Gambling Task, but not in controls.

Conclusion: This study provides quantitative evidence for a reduction of interhemispheric connectivity in schizophrenia within the AC. Negative correlation between age and AC FA in the patients is consistent with the idea that schizophrenia may be a disorder of white matter maturation. Positive correlation between FA and positive symptom is discussed in the context of white matter's established role in modulating neural conduction velocity.

Fig. 1

ROIs to draw the AC and Tractography. (A) In the color-by-orientation images, the ROI (indicated in green) was drawn on the red voxels (i.e., representing right-left diffusion). (B) Because anterior commissure fibers pass through the midline, a second ROI (indicated in blue) was drawn in the midsagittal plane (C). (D) Anterior commissure fibers were extracted from the two ROIs.

Fig. 2

Results of FA and Trace comparison.

Fig. 3

A scatter plot between mean FA and age. Blue dots represent patients with schizophrenia and green dots represent normal controls. Solid line indicates a regression line for schizophrenia (r=0.484; p<0.01) and the dashed line indicates a regression line for normal controls (r=0.19; p=0.40).

Fig. 4

Correlation analyses between Iowa Gambling Tasks and FA of AC. In schizophrenia group (blue circles), FA was significantly correlated with net gambling score [(C+D)-(A+B)] (r=−0.46; p=0.04), total amount won (r=0.51; p=0.02), total amount lost (r=0.52; p=0.02), and total amount earned (r=−0.50; p=0.02). In normal controls (green squares), FA was not significantly correlated with all parameters.