Hippocampal volume is reduced in patients with unipolar depression, maybe as a consequence of repeated periods of major depressive disorder. Bipolar patients did not seem to show a reduction in hippocampal volume, but this has been much less investigated.

Major depression is a mood disorder that is often accompanied by the impairment of cognitive functions. Although suggestive, the large range of existing neuropsychological, neuropsychiatric, and, lately, neuroimaging investigations have not yet given a consistent picture of the psychological and biological disturbances involved in this psychiatric disorder. The present study of the cognitive functions in depression was part of an extensive investigation, including neuropsychological testing, psychiatric examination, and neuroimaging. A representative sample of 40 severely depressed hospitalized patients and a group of 49 closely matched control subjects were tested with an extensive neuropsychological test battery. Results, corrected for various confounding factors, confirmed the current notion that depressed patients suffer from wide-spread cognitive impairments. The group analysis did not allow any hypothesis on a possible pattern to the dysfunctions, but heterogeneity in the test performances calls for further analysis of the data in patient subgroups in relation to neuroimaging results.

Our findings support the notion that depressed patients have disturbances in the loops connecting the frontal lobes, limbic system, basal ganglia, and cerebellum.

Our main finding is increased blood flow to the hippocampus, even when controlling for a number of confounders. This is in accordance with a rapidly expanding literature suggesting an important role for this structure in major depression.

Stroop's test and the Verbal Fluency test are commonly argued to be measures of the integrity of the prefrontal cortex. This assumption has only to some degree been con®rmed by lesion studies. In the present study, Positron Emission Tomography (PET) with H 2 15O was used to further validate Stroop's test and the Verbal Fluency as measures of frontal lobe function; both tests were implemented as activation paradigms during scanning of normal middleaged individuals. Stroop interference was found to activate the left anterior cingulate cortex, the supplementary motor cortex, thalamus, and the cerebellum. Although the prominent anterior cingulate activation is in the frontal lobe, it is not prefrontal. Verbal Fluency activated the left inferior frontal cortex and the left dorsolateral prefrontal cortex, the supplementary motor cortex, the anterior cingulate cortex and the cerebellum. These results bring this latter test closer to being a speci®c test of prefrontal function.