Background: Activation of the peroxisome proliferator activated receptor gamma (PPARg) has been found to improve behavioural deficits and neuro- pathological changes in murine models of Alzheimer‘s disease by suppress- ing neuroinflammation, increasing microglial Ab clearance and direct modulation of beta-secretase-1 promotor activity. Pioglitazone is a PPARg agonist, widely used for the treatment of non-insulin dependent type II dia- betes. Based on the above preclinical data, we hypothesized that sustained medication with pioglitazone would reduce the risk of dementia. Methods: Using routine data of the German largest public sickness fund of the years 2004 until 2010 we studied the association of pioglitazone and the onset of dementia in a prospective cohort study of 145,717 subjects aged 60 years or above who were free of dementia at baseline. The information on pre- scriptions of pioglitazone on a quarterly basis were implemented as a linear variable covering the time-dependent number of quarters of prescriptions which range between zero and 28 quarters. We performed a Cox propor- tional hazard model to explore the transition into dementia and to calculate the relative risk of dementia dependent on the use of pioglitazone adjusted for sex, age, use of NSAIDs, use of rosiglitazone, use of metformin and car- diovascular comorbidities including diabetes mellitus, cerebrovascular dis- eases, hypertension, ischemic heart diseases, atrial fibrillation and hypercholesterolemia. Results: During the follow-up period 13,841 sub- jects developed dementia. The relative risk of dementia reduced signifi- cantly (HR¼0.94, p¼0.004) with each additional quarter of pioglitazone prescription. Further analysis are done by converting numbers of quarters with prescriptions into prescribed doses of pioglitazone using the daily defined dose (DDD). Also these results support the initial hypothesis. Con- clusions: The long-term use of pioglitazone reduces the risk of dementia incidence in health claims data. One possible pathway is via suppressing neuroinflammation. P2-296 BILINGUALISM, EXECUTIVE CONTROL, AND COGNITIVE RESERVE: IMPLICATIONS FOR PREVENTION OF ALZHEIMER’S DISEASE Anthony Martyr 1 , John V. Hindle 1 , Christopher J. Whitaker 1 , Fergus I.M. Craik 2 , Ellen Bialystok 3 , Pamela A. Martin-Forbes 4 , Alexandra J.M. Bastable 1 , Kirstie L. Pye 1 , Catherine Quinn 5 , Enlli M. Thomas 1 , Virginia C. Mueller Gathercole 6 , Linda Clare 7 , 1 Bangor University, Bangor, United Kingdom; 2 Rotman Research Institute, Toronto, Ontario, Canada; 3 York University, Toronto, Ontario, Canada; 4 NISCHR CRC North Wales Research Network, Bangor, United Kingdom; 5 Bangor University, Gwynedd, United Kingdom; 6 Florida International University, Miami, Florida, United States; 7 Bangor University, Bangor, United Kingdom. Contact e-mail: a.martyr@bangor.ac.uk Background: Recent evidence for a bilingual advantage in non-linguistic executive control tasks involving inhibition and management of response conflict has led to the suggestion that being bilingual contributes to increased cognitive reserve, hence preventing or delaying the onset of Alz- heimer’s disease (AD). Further investigation may shed more light on the relevance of bilingualism for AD prevention. Methods: We assessed the performance of older ( 60 years) bilingual Welsh/English speakers (n ¼ 50) and monolingual English speakers (n ¼ 49) living in North Wales, UK, on a range of executive control tasks covering the domains of mental generativity and speed, working memory, set-shifting and switching, and in- hibition and management of response conflict. Results: There were no dif- ferences between monolingual and bilingual groups in key demographic variables or in health and functional status, but monolinguals performed bet- ter on tests assessing English language ability. Across the 40 executive func- tion indices assessed, bilinguals performed better on 9, mostly with small effect sizes, while monolinguals performed better on 31, with a mix of mainly small and medium effect sizes. Univariate analyses suggested a monolingual advantage in mental generativity and speed, working memory, and set-shifting and switching, which was greatly reduced in the domain of inhibition and management of response conflict. Only three differences were significant after correction for multiple comparisons, with monolin- guals performing significantly better on spatial span forwards, elevator counting with reversal and design fluency switching tasks. Multivariate analysis confirmed the monolingual advantage in working memory and in set-shifting and switching. Variations in degree of daily use of the two lan- guages in the bilingual group had no significant effect on performance. The two groups did not differ on a measure of cognitive reserve. Conclusions: Being bilingual did not result in better performance on executive control tasks or contribute to enhanced cognitive reserve among older participants in Wales, and thus is unlikely to confer benefits with regard to preventing AD in the Welsh context. A possible explanation for the unexpected obser- vation of a monolingual advantage may lie in the nature of the sociolinguis- tic context and its influence on cognitive processing in the bilingual group. P2-297 HERPESVIRIDAE INCREASES THE RISK OF ALZHEIMER’S DISEASE WHEN MEDIATED BY APOE E4: A META-ANALYSIS Ariah J. Steel 1 , Guy D. Eslick 1 , 1 The University of Sydney, Penrith, Australia. Contact e-mail: aste3972@uni.sydney.edu.au Background: Alzheimer’s disease (AD) is one of the most challenging dis- orders of this century and a looming public health crisis. Despite its preva- lence, the etiology of AD remains largely unknown. The role of infectious agents in the development of AD has long been debated, in particular, the herpesviridae family. Herpes simplex virus 1 in particular is thought to interact with APOE e4 to confer a risk for AD, however, the evidence sur- rounding a viral etiology for AD is conflicting. We therefore conducted a meta-analysis to quantitatively assess all published data to establish whether there is an association. Methods: We identified studies that looked for the presence of viral DNA (brain or blood) of people with AD from four elec- tronic databases searched to 21 December 2013, with no language restric- tions. From 50,739 citations identified, a total of 23 studies met our inclusion criteria that included 1294 AD cases and 3059 controls. Subgroups analyzed were the number of AD cases, APOE e4 carriers, subspecies of herpesviridae, Trigeminal ganglion (TG) latency and continent. Results: The meta-analysis showed an increased risk for AD when herpesviridae is present in the brain compared to controls [OR 1.42; 95% CI 1.20-1.69]. A subgroup analysis showed that a number of herpesviridae family members conferred an increased risk for AD including; CMV [OR 1.20; 95% CI 0.68-2.11]; EBV [OR 1.55; 95% CI 1.12-2.13]; HHV-6 [OR 2.25; 95% CI 0.95-5.23]; HSV-1 [OR 1.47; 95% CI 1.15-1.89], while HSV-2 [OR 0.31; 95% CI 0.17-1.57] and VZV [OR 0.93; 95% CI 0.34 - 2.53] showed no increased risk. Further sub-analysis showed that APOE e4 and HSV-1 together increased the risk of AD development [OR 4.29; 95% CI 1.72 - 10.69] and the presence of latent HSV-1 in TG increased the risk of AD [OR 2.34; 95% CI 0.68 - 8.10]. Conclusions: The results show that the her- persviridae family of viruses and in particular HSV-1 together with the pres- ence of the APOE e4 allele increase the risk of the development of AD. Additional research is required to determine if treatment strategies to pre- vent the onset of AD are viable in the future. P2-298 CAIDE DEMENTIA RISK SCORE, ALZHEIMER, ANDVASCULAR NEUROPATHOLOGY IN THE VANTAA 85+ STUDY Babak Hooshmand 1 , Tuomo Polvikoski 2 , Miia Kivipelto 3 , Maarit Tanskanen 4 , Liisa Myllykangas 5 , Mira M€ akel€ a 6 , Anders Paetau 6 , Raimo Sulkava 7 , Alina Solomon 1 , 1 Karolinska Institutet, Stockholm, Sweden; 2 Newcastle University, Newcastle upon Tyne, United Kingdom; 3 Karolinska Institutet, Stockholm, Sweden; 4 University of Helsinki and Helsinki University Central Hospital, Helsinki, Finland; 5 Folkh€ alsan Institute of Genetics and Department of Pathology, University of Helsinki and HUSLAB, Helsinki, Finland; 6 Department of Pathology, Helsinki University Central Hospital, Helsinki, Finland; 7 Institute of Public Health and Clinical Nutrition, University of Eastern Finland, Kuopio, Finland. Contact e-mail: babak.hooshmand@ki.se Poster Presentations: P2 P586