Abstract
Objective: The thalamus and caudate nucleus are key subcortical structures in the fronto-striato-thalamic pathways that have been implicated in the pathogenesis of schizophrenia. Previous studies have been inconsistent in identifying structural and functional abnormalities in these structures. However, methodologies in these studies have been unreliable and some have not adequately matched patients and controls.
Methods: Using algebraically-manipulated double-echo magnetic resonance (MR) images, we developed a reliable method to estimate caudate and thalamic volumes in a group of 13 monozygotic (MZ) twins; eight discordant for schizophrenia and five normal. Initially, volumes were measured on four image types: proton density (PD), T2-weighted, summed (PD + T2) and subtracted (PD-T2) to determine the most reliable method.
Results: There was a significant method by region interaction, where caudate volumes measured on PD images were significantly larger than those measured on T2-weighted images, while the opposite was found for thalamic volumes. However, there was no interaction of method by diagnosis. Test-retest reliability was highest for the summed images. Using summed images to measure the volumes of the caudate and thalamus in each twin, we found significantly increased caudate volumes in affected twins compared to their unaffected cotwins, but no significant difference in thalamic volume.
Conclusions: Our results in a small sample of MZ twins discordant for schizophrenia do not support the presence of structural abnormalities in the thalamus. The findings in the caudate are consistent with previously reported effects of antipsychotic medication. We also report a reliable method for assessing the volumes of subcortical structures. However, volumetric estimates of brain structures may be dependent on which method is used and the structure being assessed. Such interactions need to be considered in future studies investigating brain structural abnormalities in schizophrenia and other disorders.