P 40 Correlation between hypothalamus and third ventricle of patients with affective disorders

Abstract

Objectives Former studies proved the key role of the hypothalamus in the pathophysiology of affective disorders. Because of limited imaging methods and difficult delineation, attested volumetric changes has only been verified post-mortem, so in vivo studies made use of the adjacent third ventricle as an indirect marker of hypothalamic changes. Considering that these former ventricle studies leave sufficient scope for methodical improvements, it was our aim to replicate and complement these findings by using a high-resolution 7T-MRI for the first time. Method 84 subjects, aged 20–62 years, with major depressive or bipolar disorder or no history of psychiatric disorders, participated in the study. The hypothalamus volumetry was performed by using the method of Schindler et al. For ascertaining the volume of the ventricle, the software SPM segmented the entire cerebrospinal fluid fully automatically. Afterwards a rater applied a new developed algorithm for the segmentation of the hypothalamic part of the third ventricle. Results The two-factorial ANOVA showed, that patients with bipolar disorder have significant larger ventricles than healthy controls (p = 0.013). For patients with major depressive disorder the effect was not significant. The correlation analysis revealed no relation between hypothalamus and third ventricle after correcting for confounding variables as age, intracranial volume and outliers. Conclusions To our knowledge, this is the first study which investigated the volume of the third ventricle in high-resolution 7T images. Our result, that patients with bipolar disorder have enlarged ventricles, is in line with previous results. The formerly presumed correlation between hypothalamus and ventricle has not been verified, which could be due to the fact that our sample is not large enough to reach the necessary statistical power. A potential influence of psychotropic medication is currently the subject of analysis.