hypertension . supranuclear sixth nerve palsy related to tegmentum as a

sent study. Our investigation showed a highly significant increase of CSF total tau protein in patients with probable Alzheimer's disease compared with elderly depressive patients, patients with other neurological disorders, and young and elderly controls. It seems plausible that these increases are the result of enhanced release of tau protein from damaged neurons into the extracellular space and, as a consequence, into CSF. Recent results suggested an arbitrary cut off value of 200 pg/ml, at which only a small percentage of healthy controls are positive.' Based on our data, we would propose a 10% higher cut off value, of about 225 pg/ml, to better distinguish the elderly depressive patients as well as young and elderly controls. The observed difference between patients with Alzheimer's disease and elderly depressive patients confirms a recent study and is of particular interest.' Depressive periods can be accompanied by a severe impairment of attention, receptivity, and concentration which may mimick a dementing process. Thus early neurochemical identification of a depressive syndrome is of great importance for appropriate therapeutic strategies. The reason for more tau protein in depressed patients than in young controls is unknown. On the other hand, the non-significant difference between young and elderly controls leads to the assumption that increased tau protein is not merely a marker of aging itself. There are contradictory reports regarding the correlation between MMSE and total tau protein immunoreactivity in the CSF of patients with Alzheimer's disease.' I From our data there was a significant correlation only when including the whole range of scores, which became non-significant when excluding the patients with severe dementia. These results corroborate neuropathological findings which showed a significant correlation between the number of neurofibrillary tangles and MMSE only when including severely demented patients. In conclusion, total tau protein concentration in CSF is an important candidate for discriminating patients with Alzheimer's disease from elderly depressed patients as well as from young and elderly controls without mental illness. Further studies on other neurodegenerative diseases are in progress. Moreover, different combinations of tau protein antibodies to detect exclusively normal or hyperphosphorylated tau protein may help to further evaluate the diagnostic value of tau determinations in CSF. NORBERT ROSLER ILDIKO WICHART KURT A JELLINGER Ludwig Boltzmiann Inistitute of Cliniical Neurobiology, Lainlz Hospital, Vienna, Austria