Abstracts /International Journal of Pgrhophysiology 25 (1997) 17-84 63 in mid-dorsolateral prefrontal cortex, in parietal cortex and in A/3 and increased cholinergic deficit in brains of ApoE ~4 the basal ganglia during estimation of five seconds. homozygous AD patients compared to those without ~4. TEMPORAL RESPONSE CHARACTERISTICS OF SAC- CADIC EYE MOVEMENTS TO AUDITORY AND VI- SUAL STIMULI V. Haase’, E. Ruhnau’ ‘Belo Horixonte, Brazil; ‘Juelich, Germany It has been found in many studies that histograms of choice reaction time show a multimodal characteristic, the modes being separated by 30 to 40 ms. Multimodal responses are, however, only observed if stationarity conditions for the exper- imental observations are strictly observed. There are mainly two sources violating stationarity, not allowing a creative look into the data-generating mechanism. Subjects often have not gamed a plateau of performance. In choice reaction time experiments, a satisfactory level of performance with only statistical fluctuations is only reached after several thousand trials. The other source of instationarity comes from an inap- propriate choice of the statistics of interstimulus intervals which leads to an undesirable modulation of subjective probability of stimulus occurrence. However, given the control of these and the other factors, multimodality of response distributions in choice reaction time experiments is a typical result. We asked ourselves whether such multimodality can also be seen in a choice situation in which saccadic eye movements are recorded. Visual and auditory stimuli were presented in random order at different eccentricities both left and right of the vertical meridian. The task of the subject was to look as fast as possible towards the position where either the visual or the auditory stimulus was presented. It turned out that both far visually or acoustically initiated saccades multimodal response distributions were observed, the modes being separated by approx. 30 ms. We conclude that a visual or an auditory stimulus initiates an oscillatory response in a neuronal assembly and that on this basis saccades to visual or auditory stimuli are preferentially triggered in a time-locked fashion at distinct intervals after stimulus onset. Thus, tem- poral processing with respect to external events is apparently discontinuous. AD initially presents with memory loss and recent MRI studies have indicated a pronounced decline in hippocampal volume as a sensitive and early sign of AD. Hippocampal atrophy has been shown to correlate with severity of memory loss in AD. A study in AD patients showed that patients with two apoE ~4 alleles have more severe hippocampal volume loss in MRI scans and more impairment in tests assessing delayed memory, than AD patients without the c4 allele in spite of similar global severity of dementia. Minor changes in hippocampal MRI volumetry can also be detected in nonde- mented elderly, particularly in those with an l 4/4 genotype. Data from a population based study showed that elderly subjects carrying the e4 allele had worse learning ability than those with e2/2 or 2/3 phenotypes, whereas these groups did not differ in other cognitive domains. These data suggest that the ApoE l 4 allele may signi& candy enhance severity of hippocampal atrophy, is associated with severe memory impairment in AD and is related to impaired learning ability even in nondemented elderly individ- uals. These data suggest that ApoE ~4 allele is a significant contributor to memory loss in the elderly. CORRELATES OF MEMORY IN PATIENTS WITH ALZHEIMER’S DISEASE AND AGE-ASSOCIATED ME- MORY IMPAIRMENT: A SPECT STUDY E-L Helkala’, J. Kuikka’, T. Hlnninen’, M. Hallikainen’, K. Koivisto’*3 and P. Riekkinen Srzs4 Department of Neurology’ and Clinical Physiology’ of Kuo- pio University Hospital, University of Kuopio’ and Seinajoki Central Hospita13, Finland, A.I. Virtanen Institute4, Univer- sity of Kuopio, 70211 Kuopio, Finland We studied regional cerebral blood flow measured with 99m Tc-ECD single photon emission tomography in ten healthy elderly subjects (C-group), ten Alxheimer disease patients (AD group) and ten subjects with age-associated memory impairment @AM1 group) during two similar tasks involving three-letter word beginnings (stems). APOE AND MEMORY IN OLD AGE H.S. Soininen* and P.J. Riekkinen Sr Department of Neurology, Kuopio University and University Hospital, A.I. Virtanen Institute, Kuopio, Finland Apolipoprotein E (ApoE) ~4 is a well-documented risk factor for late-onset sporadic and familial Alzheimer’s disease (AD). AD patients show neuronal damage particularly in the medial temporal lobe structures involved in memory processing. ApoE has been implicated in nerve regeneration following injury and synaptogenesis in the hippocampus of experimental ani- mals. Recent studies have shown increased accumulation of Prior to each task, subjects studied a list of words. A dose of 99mTc-ECD was intravenously injected during presentation of the third word stem while subjects a> completed stems to form the first. words to come to mind, but the stems were not beginnings of any previously presented study words (baseline), or b) tried to recall study words using the stems presented, half of which were contained in the previously presented words (memory). The memory task decreased the activation of left parietal as well as right and left occipital cortices com- pared to the baseline condition in the AAMI and the control groups, whereas there were no changes in the AD group. There was also decreased activation in the combined region of right amygdala and hippocampus in the AAMI group. All subject groups produced fewer words in the memory task than in the practice task, which was the priming task, although