Auditory verbal hallucinations (AVH) is a common and stressful symptom of schizophrenia. Disrupted connectivity between frontal and temporo-parietal language areas, giving rise to the misattribution of inner speech, is speculated to underlie this phenomenon. Disrupted connectivity should be reflected in the microstructure of the arcuate fasciculi (AF); the main connection between frontal and temporo-parietal language areas.
In this study we compared microstructural properties of the AF and three other fiber tracts (cortical spinal tract, cingulum and uncinate fasciculus), between 44 schizophrenia patients with chronic severe hallucinations and 42 control subjects using diffusion tensor imaging (DTI) and magnetic transfer imaging (MTI).
The DTI scans were used to compute fractional anisotropy (FA) and to reconstruct the fiber bundles of interest, while the MTI scans were used to compute magnetic transfer ratio (MTR) values.
The patient group showed a general decrease in FA for all bundles. In the arcuate fasciculus this decreased FA was coupled to a significant increase in MTR values. A correlation was found between mean MTR values in both arcuate fasciculi and the severity of positive symptoms.
The combination of decreased FA and increased MTR values observed in the arcuate fasciculi in patients suggests increased free water concentrations, probably caused by degraded integrity of the axons or the supportive glia cells. This suggests that disintegrated fiber integrity in the connection between frontal and temporo-parietal language areas in the schizophrenia patients is associated with their liability for auditory verbal hallucinations.
Auditory verbal hallucinations (AVH) are a core symptom of schizophrenia, which underlying neurobiology is still largely unclear. Unraveling its pathophysiology may provide clues for new treatment strategies for the 25% of patients who fail to respond to current antipsychotic medication (Shergill et al., 1998).
Recent functional magnetic resonance imaging (fMRI) studies have revealed important aspects of the neurobiology of AVH. Several brain areas consistently showed activation during AVH (Jardri et al., 2010), such as the right and left temporo-parietal cortices and Broca's area and its right-sided homologue (Shergill et al., 2000, Sommer et al., 2003, Hoffman et al., 2007, Sommer et al., 2007, Sommer et al., 2008 and Diederen et al., 2010). Temporo-parietal activation during AVH probably reflects the perception of speech, while activity in the bilateral inferior frontal areas suggests the production of language. Words produced in these areas may be experienced as AVH. It is unclear, however, why patients do not recognize these words as self-produced, but rather attribute them to an external source. Ford et al. (2007) pointed to malfunction of the corollary discharge mechanism: a neuronal circuit that suppresses the sensory consequences of self-generated actions. Such systems are well known in the visual system, but also serve the auditory language circuit ( Paus et al., 1996). EEG and ERP studies showed that where healthy individuals suppress auditory perception areas during speech, this suppression is decreased in schizophrenia patients ( Heinks-Maldonado et al., 2007 and Ford et al., 2007). Insufficient corollary discharge in the language system could result from disrupted communication between frontal and temporo-parietal areas ( Whitford et al., 2010). Such disturbed connectivity could result from microstructural alterations in the arcuate fasciculi, the most important fiber bundle between Broca's area and Wernicke's area ( Lichtheim, 1885).
Diffusion tensor imaging (DTI) can measure structural connectivity in the human brain for specific white matter bundles. Fractional anisotropy (FA) is a measure of directionality of the axons forming the fiber bundles and is often used as an index of fiber integrity and, to a lesser extent, myelination (Beaulieu, 2002). Decreased FA, along with increased diffusivity within prefrontal and temporal lobes, as well as abnormalities within the fiber bundles connecting these regions are frequent findings in schizophrenia (Kubicki et al., 2007). Three studies have specifically measured the arcuate fasciculus in hallucinating schizophrenia patients using DTI (Hubl et al., 2004, Seok et al., 2007 and Shergill et al., 2007), reporting decreased FA values of this bundle.
More specific information about abnormalities of white matter tracts can be obtained by combining DTI with magnetization transfer imaging (MTI) (Kubicki et al., 2005). MTI is sensitive to macro molecules in tissue including myelin (Wolff and Balaban, 1994, van Buchem et al., 1999, Barkovich, 2000 and Henkelman et al., 2001). A measure of this magnetization transfer is usually expressed as magnetic transfer ratio (MTR). However, sequences with relatively high T1 weighting (as used in this study) are also sensitive to free water concentrations (Henkelman et al., 2001). By combining the MTR results with the FA results we can differentiate between these two possible explanations. With increasing myelin concentration one would expect an increase in MTR as well as a decrease in radial diffusivity and an increase in FA ( Gulani et al., 2001). However if the level of free water increases (e.g. as a result of degradation of its microstructure) then also an increase in MTR may be expected but no decreases in radial and axial diffusivity. In fact, if the increase in free water is for instance due to a less dense packing of axons then an increase in radial diffusivity and hence a reduction in FA may be expected. Thus, low values of both FA and MTR would point to decreased myelin, while low FA values associated with high MTR suggest increased free water concentrations, probably caused by degraded integrity of the axons or the supportive glia cells.
In this study we investigated microstructural connectivity between the frontal and temporo-parietal language areas in schizophrenia patients with chronic hallucinations and matched controls. Mean FA values and mean MTR were determined along the left and right arcuate fasciculi. In order to make inferences about specificity of our findings, the same measures were taken from three other tracts, namely the cortico spinal tract (CST), cingulum (CGL) and uncinate fasciculus (UF). These different tracts were chosen for anatomical reasons, because they share directional components with the AF, rather than their functional role in schizophrenia. We hypothesize that FA and MTR values integrated along the arcuate fasciculus are specifically affected in the patient group, which may underlie their predisposition for AVH
