gender. MCI participants spoke longer (p¼0.037) by 105 minutes on average per day during daily life than normal subjects in GEE models con- trolling for age, education, gender and living arrangement. Conclusions: MCI subjects may exhibit subtle language processing deficits that affect so- cial cognitive abilities required for smooth social interactions, leading to more words spoken during conversations or longer daily talking time. The assessment approach used here may provide an ecologically valid behav- ioral marker sensitive to transitions to MCI. P4-349 A MULTICENTER, RANDOMIZED TRIAL TO ASSESS EFFICACY OF HOME-BASED AND GROUP COGNITIVE INTERVENTION PROGRAMS FOR AMNESTIC MILD COGNITIVE IMPAIRMENT Hae Ri Na 1 , Seong H. Choi 2 , Jee-Hyang Jeong 3 , JungEun Kim 4 , Duk L. Na 5 , Sang Won Seo 5 , Juhee Chin 6 , Sun A. Park 7 , Eun-Joo Kim 8 , Hyun Jeong Han 9 , Seol-Heui Han 10 , Jae-Hong Lee 11 , Soo Jin Yoon 12 , Kyung Won Park 13 , So Young Moon 14 , Moon-Ho Park 15 , Mun Seong Choi 16 , Il-Woo Han 17 , Jun Hong Lee 18 , Jung Seok Lee 19 , 1 Bobath Memorial Hospital, Seongnam, South Korea; 2 Inha University School of Medicine, Incheon, South Korea; 3 Ewha Womans University Mokdong Hospital, Seoul, South Korea; 4 College of Medicine, Seoul National University, Seongnam-si, South Korea; 5 Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea; 6 Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea; 7 Soonchunhyang University Bucheon Hospital, Bucheon-si, South Korea; 8 Pusan National University Hospital, Pusan, South Korea; 9 Kwandong University College of Medicine, Koyang, South Korea; 10 Konkuk University Medical Center, Seoul, South Korea; 11 Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea; 12 Eulji University School of Medicine, Daejeon, North Korea; 13 Dong-A University College of Medicine, Busan, South Korea; 14 Ajou University School of Medicine, Suwon, North Korea; 15 Korea University College of Medicine, Ansan, North Korea; 16 Hallym Hospital, Seoul, South Korea; 17 Yong-in Hyoja Geriatric Hospital, Yongin, North Korea; 18 National Health Insurance Servic Ilsan Hospital, Goyang-si, South Korea; 19 Jeju National University School of Medicine, Jeju, South Korea. Contact e-mail: neuna102@paran.com Background: The previous studies have failed to show consistently that cognitive intervention is efficacious in amnestic mild cognitive impairment (MCI). There is a need for good quality trials examining the efficacy of cognitive interventions in MCI. The objectives of this study are to assess the efficacy of home-based and group cognitive intervention programs in amnestic MCI. Methods: Randomized and single-blind trial for a 12- week cognitive intervention conducted at 18 hospitals in South Korea from March 2011 to January 2013. A total of 293 participants with amnestic MCI were randomly assigned to group cognitive intervention (n¼95), home-based cognitive intervention (n¼98), or waitlist control (n¼100). Par- ticipants were assessed at post-intervention, 3 months ’ follow-up, and 6 months ’ follow-up. Primary efficacy measure was a change from baseline in the modified Alzheimer Disease Assessment Scale-cognitive subscale (ADAS-Cog) after 12 weeks of intervention. Secondary efficacy measures evaluated cognition, behavior, Quality of life (QOL), and function. Results: A total 25 of 95 participants (26.3%) from the group cognitive intervention group and 21 of 98 participants (21.4%) from the home-based cognitive intervention group withdrew the intervention and 23 of 100 participants (23.0%) withdrew from the control group. The participants in the group cognitive intervention and the home-based cognitive intervention groups improved 2.24 points (Standard Deviation (SD) ¼4.66, p¼0.02 vs. controls) and 2.43 points (SD¼ 4.45, p¼0.03 vs. controls), respectively and those in the control group improved 0.84 points (SD¼4.79) on the modified ADAS- cog at the end of the intervention. At 6 months ’ follow-up, modified ADAS- cog, immediate story recall test, Prospective Memory test, Prospective and Retrospective Memory Questionnaire (PRMQ), and Caregiver-Adminis- tered Neuropsychiatric Inventory scores significantly favored group cogni- tive intervention compared with controls. At 6 months ’ follow-up, modified ADAS-cog, Sum of Boxes of Clinical Dementia Rating scale, PRMQ, and QOL-Alzheimer ’ s disease scores significantly favored home-based cogni- tive intervention compared with controls. Conclusions: The group and home-based cognitive intervention programs were effective to improve cognition, QOL, and behavior in amnestic MCI. P4-350 WORD RETRIEVAL IN PRIMARY PROGRESSIVE APHASIA FOLLOWING LANGUAGE THERAPY Theresa Raiser 1 , Karen Croot 2 , Lyndsey Nickels 3 , Cathleen Taylor 4 , Nibal Ackl 5 , Elisabeth Wlasich 6 , Gisela Stenglein-Krapf 5 , Axel Rominger 5 , Adrian Danek 6 , 1 Universit€ atsklinikum LMU, Munich, Germany; 2 University of Sydney, Sydney, Australia; 3 Macquarie Universtity, Sydney, Australia; 4 War Memorial Hospital, Waverley, Australia; 5 Universit€ atsklinikum LMU, Munich, Germany; 6 Universit€ atsklinikum LMU, M€ unchen, Germany. Contact e-mail: theresa.raiser@med.uni-muenchen.de Background: Primary Progressive Aphasia (PPA) is a neurodegenerative disease with its most prominent symptom being impaired language at the early stage. As the syndrome progresses the clinical symptoms of PPA worsen. Impaired word retrieval is one of the core symptoms of PPA and consequently communication is often restricted. However, there are few ev- idence-based treatments available for PPA. Aim: The aim of this study was to evaluate the efficacy of a therapy for word retrieval in PPA using a series of single case experimental designs. We predicted that treated items should be retrieved more easily following therapy, in comparison to untreated items, bearing in mind that the default pattern in a neurodegenerative con- dition is for a worsening of performance. Treatment should also result in improved word retrieval for the treated items in the patient’s spontaneous speech in comparison to untreated words. Methods: Three patients with PPAwere recruited through the German Research Consortium of frontotem- poral lobar degeneration. They were enrolled in a treatment programme, which used an adaptation of the Croot, Taylor and Nickels Sydney protocol. Items for treatment (n120) were selected individually for each participant. Two matched sets of words (n30 each) were treated with an errorless learning technique (repetition and reading in the presence of a picture, RRIPP) in two successive treatment phases. The remaining stimuli (n60) served as a control set. The therapy was carried out independently (without a clinician) using delivery via a computer for two (one set) and four weeks (other set). One patient underwent further six months of training of all items. Outcome measures were accuracy of confrontation naming for treated items and untreated items and carry over into spontaneous speech measured using a semi- structured interview. Results: In one out of three pa- tients significant improvement in naming performance was recorded for treated items but not for untreated items. The other two patients’ showed no change in performance over the therapy period, despite comprehension worsening in one of these patients. These results are interpreted as a positive effect of treatment in the context of a neurodegenerative disease where per- formance (without treatment) is expected to decline. The patient who improved significantly kept practising for another six months and further improved his naming skills. The other two patients stopped the training independently. Conclusions: No change was shown in untreated items over the same time period, which is interpreted positively. Transfer to spon- taneous speech was absent. Nonetheless, positive results for the patients’ mood were present due to therapy measured with the Beck Depression In- ventory. In summary, despite the progressive nature of PPA language ther- apy can produce improved performance for treated stimuli and hence is appropriate for patients with PPA. P4-351 A PLASMA PHOSPHOLIPID BIOMARKER PROFILE FOR DETECTING PRECLINICAL ALZHEIMER’S DISEASE CAN BE MODIFIED BY ORAL INTAKE OF NUTRIENTS THAT INCREASE PHOSPHOLIPID SYNTHESIS Tobias Hartmann 1 , Nick van Wijk 2 , Richard J. Wurtman 3 , Marcel G.M. Olde Rikkert 4 , John W.C. Sijben 2 , Hilkka Soininen 5 , Bruno Vellas 6 , Philip Scheltens 7 , 1 Saarland University, Homburg, Germany; 2 Nutricia Research, Utrecht, Netherlands; 3 Massachusetts Institute of Technology, Cambridge, Massachusetts, United States; Poster Presentations: P4 P916