Lack of association between volumetric hippocampal measurements and episodic memory in schizophrenia but not in healthy controls

Lack of association between volumetric hippocampal measurements and episodic memory in schizophrenia but not in healthy controls

118 Corresponding author: Antonio Vita, MD, PhD, Institute of Psychiatry, Pad. Guardia II, Scientific Research and Care Institute, IRCCS, Ospedale Ma...

100KB Sizes 1 Downloads 7 Views


Corresponding author: Antonio Vita, MD, PhD, Institute of Psychiatry, Pad. Guardia II, Scientific Research and Care Institute, IRCCS, Ospedale Maggiore, via F. SJbrza, 35, 20122 Milan, Italy In the last 15 years a large series of Magnetic Resonance Imaging (MRI) studies have addressed the question of the existence of specific abnormalities in the brain of schizophrenic patients. Despite the number of positive findings, these have frequently been dismissed as insubstantial or controversial, and qualitative reviews have likewise failed to lead to definite conclusions. We considered the articles on MRI studies published in English language that reported measurements of cerebral structures in schizophrenic patients and control subjects. 23 meta-analyses were performed on a total of 56 studies. Significant effect sizes were demonstrated for results relative to a decrease in the volumes of frontal lobes (effect size: right = -0.28, left= -0.24), temporal lobes (effect size: right= -0.23, l e f t = - 0 . 1 8 ) , amygdala (effect size: r i g h t = - 0 . 5 4 , left= -0.43 ), hippocampus (effect size: right = - 0.37, left = -0.28) and an increase in ventricular volumes (effect size: 0.32) in schizophrenia. There is no evidence to support a reduction in volume of the superior temporal gyrus (effect size: right= -0.14, l e f t = - 0 . 1 2 ) . The left but not right caudate nucleus proved to be increased in size. As for frontal lobe, the only region for which available data allowed differential analysis for gray and white matter, no evidence emerged supporting a more severe reduction of one of the two components. For lateral ventricles only, a statistically significant difference emerged between patients' and controls' laterality indexes (p = 0.02). The available published data strongly indicate the presence of multisite brain pathomorphology in schizophrenia. Results concerning involvement of the limbic structures appear particularly solid.



K.R. Laws

Department of P3Tchology, University of Hertfordshire, College Lane. Hatfield, Herffbrdshire, ALIO 9AB, UK Cognitive and functional imaging studies have implicated an executive dysfunction and an accompanying hypofrontal profile in schizophrenics. A recent recta-analysis of the cognitive studies has, however, cast doubt on the presence of an executive dysfunction in schizophrenia and the typicality of the patients examined in functional imaging studies (K.R. Laws, 1999, Cognitive Neuropsychiatry, 4, 1 35). Structural MRI scanning studies analysing the frontal lobes have also produced somewhat contradictory findings. Hence, a meta-analysis was conducted across the eight MRI studies that

have compared the right and left frontal lobe volumes of schizophrenics and normal controls (n=493). This revealed a small but significant effect size (ES) for right frontal volume (ES=0.28, p=0.04); the left frontal volume ES was nonsignificant (ES=0.22). These ESs are small and given that the fail-safe for the right frontal lobe was 6 studies, this may be susceptible to the file-drawer problem. Correlational analyses showed that right frontal ESs correlated inversely with slice thickness ( r = - 0 . 7 1 ) , and positively with Telsa magnet strength (r =0.78) and patient age (r =0.69). The left frontal lobe also correlated significantly with these variables. There were no significant correlations for: illness duration, age of onset, year of study, sample size or the reliability of measurements. These findings show that any right frontal reduction is quite subtle being picked up by stronger magnets and narrower slices and being more pronounced in elderly schizophrenic patients.


Cognitive Neurop6Tchiatrv Unit, The University of Melbourne, and Applied Schizophrenia Division, Mental Health Research Institute o[' Victoria, Locked Bag 11, Parkville, Victoria 3052, Australia There is evidence from symptomatology, functional neuroimaging and neuropsychological studies that the parietal lobe is involved in the pathophysiology of schizophrenia, especially where there are high levels of delusions of control. We investigated whether structural brain changes also occur in schizophrenia characterised predominantly by delusions. We scanned 23 patients with chronic schizophrenia using double-echo MRI sequences (3 mm T2/proton density images) which were analysed using the Structural Magnetic Resonance Toolkit (SmaRT). SmaRT allows comparison of the relative distribution of grey and whte matter between groups at the single voxel level (Suckling et al., 1999). Using SAPS/ PANNS/clinical data we identified patients with delusions of control (n = 12) and patients without delusions (n = 11 ). The patients with delusional symptoms exhibited significantly less (p<0.005) tissue volume in the cerebellum bilaterally, the left post-central gyrus, left anterior cingulate gyrus and the left inferior parietal lobe. These data suggest that specific volume reductions in a network of brain structures, which include the inferior parietal lobe, are associated with high levels of delusions in patients with chronic schizophrenia. This finding is consistent with recent neuropsychological and functional neuroimaging evidence for the involvement of the inferior parietal lobe in schizophrenia and suggests that neuroanatomical models of schizophrenia must include the parietal lobe.