P3-272 SELF-RATED HEALTH AND RISK OF INCIDENT DEMENTIA: A COMMUNITY-BASED ELDERLY COHORT, THE 3C STUDY Claire Montlahuc 1 , Aicha Soumare 1 , Carole Dufouil 1 , Claudine Berr 2 , Jean-Francois Dartigues 3 , Michel Poncet 4 , Christophe Tzourio 1 , Annick Alp erovitch 1 , 1 Inserm U708, Paris, France; 2 Inserm U 888, Montpellier, France; 3 Inserm U897, Bordeaux, France; 4 H^ opital Universitaire la Timone, Marseilles, France. Background: Few studies have investigated the relationship between self-rated health and dementia. We examined the relationship between self-rated health and incident dementia, and investigated the impact of cognitive complaints, depressive symptoms, and functional status on this relationship. Methods: Participants of the 3C study, a prospective French cohort study composed of 8,169 non-demented community-dwell- ing men and women aged ¼65 years, were asked to rate their health at the baseline examination in 1999-2001. They were followed over an 8 year pe- riod during which three extensive follow-up examinations were performed (at years 2, 4, and 8) and dementia was screened and diagnosed. Hazard ra- tios of dementia according to baseline self-rated health (good, fair, or poor) were estimated with a Cox model adjusted for demographic factors, vascu- lar risk factors or diseases, and non-vascular chronic diseases. Stratified analyses according to the presence or absence of cognitive complaints, de- pressive symptoms, or functional disability were also realized. Results: During the 46,990 person-years of follow-up, 618 participants developed dementia. Risk of dementia was increased in participants with poor (ad- justed hazard ratio: 1.70, 95% CI 1.22 to 2.37) or fair (adjusted hazard ra- tio: 1.34, 95% CI 1.13 to 1.59) self-rated health compared to those with good self-rated health. Poor self-rated health was associated with both Alz- heimer’s disease (1.48, 1.00 to 2.24) and vascular dementia (3.38, 1.25 to 9.17). Self-rated health was a stronger predictor of dementia in participants without cognitive complaints [risk of dementia in those rating their health as poor but without cognitive complaints: 1.96 (1.24 to 3.09), P¼0.004] and in those without functional disability (1.76, 1.14 to 2.71). Moreover, the predictive value of self-rated health was independent of depressive symptoms. Conclusions: There are few recognized predictive factors for dementia that can be assessed easily. Our findings suggest that a simple self-rated health question could help raise awareness of medical doctors about patient’s risk of dementia, especially in those without conditions in- dicative of potential cognitive impairment, such as cognitive complaints or disabilities. P3-273 DEMENTIA PREVALENCE IN AN INDIGENOUS POPULATION FROM BRAZILIAN AMAZON Leonardo Caixeta, Federal University of Goias, Goi^ ania, Brazil. Background: Indigenous peoples live in many places around the world, but very few live in villages far from civilization and keeping a fishing- colecting culture like his ancestors from 10,000 years ago. The Karaja from Brazilian Amazon are an example of this indigenous people with such a life style. Isolated human populations have provided a natural ex- perimental laboratory for the ongoing study of human disease. Methods: One hundred and eight indigenous elderly from Karaja tribe in the Brazilian Amazon were assessed using a modified version of the MMSE, as well as a survey of activities of daily living (ADL - Pfeffer) and a neuropsychiatric clinic evaluation. Results: 68.5% of the sample were female. The mean age was 72.4 years (-8.7), with a confidence interval of 70.8 - 74.1. 100% were illiterate. The average MMSE was 16.2 (-5.6), with a confi- dence interval of 15.1 - 17.3. The mean ADL was 5.4 (-8.0), with a con- fidence interval of 3.5-7.0. Smoking was present in 60.2% and alcoholism in 15.7%. Mental problems were reported by relatives in 16.7% of the sample. The prevalence of dementia in this sample was 6.4%. Conclu- sions: The rate found is below the one reported to the Brazilian non-indian population (7.1%). The most prevalent form of dementia has features of possible Alzheimer’s disease. Healthy lifestyles and the absence of several cardiovascular risk factors might have contributed to the lower prevalence rate of dementia in this sample. P3-274 ALZHEIMER’S DISEASE AND ALCOHOLIC ENCEPHALOPATHY Leopold Liss, Ohio State University, Columbus, Ohio, United States. Background: The higher risk of developing Alzheimer disease (AD) is caused by events which injure the brain and facilitate or create a locus minoris resitentiae which accelerates the time of onset of pathological changes and subsequent clinical symptoms. The cerebral trauma and im- pairment of the cerebral perfusion are recognized as risk factors. This study identifies alcoholic encephalopathy (AE) as a predisposing factor for development of AD. Methods: Material and Methods The brains from 268 autopsies were examined in the Division of Neuropathology and correlated with case histories obtained from the Veteran’s Psychiatric Hospital. A detailed photographic record and a detailed histological exam- ination was routinely used. The histories of 142 patients who had diagnosis of dementia indicated also alcohol abuse. The clinical information were not detailed and most frequently were limited to a statement such as: alcohol- ism, heavy alcohol uptake, binge drinking, alcohol addiction, excessive so- cial alcohol consumption, heavy alcohol uptake in the past until., continuous moderate drinking. These statements in the history gave only indication that alcohol was ingested in great quantities and further inquiries provided only sketchy clinical data. Results: The gross and histological ex- amination provided several distinct diagnoses: posttraumatic dementia: Five brains had gross and histological evidence traumatic alcoholic demen- tia: Twenty one brains had evidence of severe thiamine deficiency enceph- alopathy. Alzheimer’s Disease: one hundred and five had cerebral atrophy with ventricular dilatation, hippocampal atrophy and histological confir- mation of plaques and tangles. There was varying degree of mammillary bodies pathology, the most severe showing total shrinkage with fibrosis, while others had almost preserved form and size. The surprising feature was presence of neurofibrillary tangles and plaques in the mammillary bod- ies. In the remaining eleven brains we found mixtures of AE and trauma. Conclusions: Alcoholic brain injury has to be considered as one of the risk factors to develop AD. The early presence of AD pathognomonic changes in mammillary bodies provides proof that creation of a locus mi- noris resitentiaeby a noxious agent or event, accelerates the onset of AD. The rarity of vascular dementia in alcoholic population demands further study. P3-275 ONE-YEAR SURVIVAL OF DEMENTED STROKE PATIENTS IN THE DIJON STROKE REGISTRY (1985-2008) Yannick B ejot 1 , Agn es Jacquin 2 , Olivier Rouaud 1 , Corine Aboa-Eboul e 1 , J er^ ome Durier 1 , Marie Hervieu-B egue 1 , Guy-Victor Osseby 1 , Maurice Giroud 1 , 1 Dijon Stroke Registry, EA4184, University Hospital and Medical School of Dijon, France, Dijon, France; 2 Department of Neurology, University hospital of Dijon, France, Dijon, France. Background: Our aim was to evaluate one-year survival of stroke patients with dementia and contributing factors in a population-based study. Methods: From 1985 to 2008, overall first-ever strokes (ischemic strokes, intracerebral haemorrhages and subarachnoid haemorrhages) occurring within the population of the city of Dijon, France (150,000 inhabitants) were recorded. To ensure the completeness of case-ascertainment, multiple overlapping sources of information were used to identify hospitalized and non-hospitalized, fatal and non-fatal strokes. Dementia was diagnosed dur- ing the first month following stroke, according to DSM-III and DSM-IV cri- teria. Survival was estimated one year after stroke and multivariate analysis was performed using Cox regression to identify contributing factors. Re- sults: Over the 24-year study-period, 3948 first-ever strokes were recorded. Among stroke patients, 3201 (81%) were testable of whom 653 (20.4%) had dementia (337 women and 316 men). The inability to test the cognitive sta- tus of patients (n¼747, 19%) was due to either severe aphasia (n¼66; 8.7%) or death (n¼681; n¼91.3%) in the acute phase of stroke. One-year survival Poster Presentations P3 S604