Previous DTI studies in schizophrenia have all found decreased white matter integrity in the patients, though the location of these differences has varied. This may be due to the use of region-of-interest methods and underpowered studies. We used voxel-based DTI to examine a much larger sample of patients with schizophrenia and controls.
Seventy-six patients with DSM-IV schizophrenia and 76 controls matched for age, gender, handedness, IQ, and education were scanned with an optimized DTI sequence at 1.5T. FA maps were co-registered using SPM2 and group differences calculated using non-parametric XBAM_v3.4. Mean FA was extracted from each significant cluster and correlated with illness duration in the patients. Cluster FA was compared between the 15 patients with a few days exposure to antipsychotics and 30 matched patients who had been treated for over a year.
At thresholds of <1 false positive (voxel p<0.01, cluster p<0.0005), there were widespread reductions in FA in the patient group. These areas included bilateral cingulum, superior & inferior longitudinal fasciculus, left uncinate and the genu of the corpus callosum. There were no areas of increased FA in patients relative to controls. In our secondary analyses, there were no significant correlations between the mean FA extracted from any of these clusters and duration of illness, and no significant differences between the briefly medicated and chronically medicated groups.
Schizophrenia is associated with FA reductions distributed widely in white matter, but these differences do not correlate with duration of illness, and do not segregate with medication.