Poster Presentations: Wednesday, July 27, 2016
participating in the Harvard Aging Brain Study underwent T807 (aka AV1451) and PiB PET imaging within 6 months of responding to three SCC questionnaires. An SCC composite was created using z-transformed memory subscales from the Memory Functioning Questionnaire, the Everyday Cognition Scale, and the adapted Structured Telephone Inventory for Dementia Assessment. Results: Greater EC tau was associated with greater SCC (b¼0.39, p<.001) after taking into account age, depressive symptoms, sex and education. IT tau was marginally associated (b¼0.18, p¼.16). After interactions between tau and Ab were included, no interactive effect of EC tau and Ab was found on SCC (p¼.93), but an IT and Ab interaction was now marginally associated with greater SCC (p¼.07). Conclusions: This is the first study to demonstrate regionally-specific patterns of association between tau burden and SCC. Specifically, EC tau was strongly associated with greater SCC, but the effect was independent of Ab burden. Increasing IT tau deposition, on the other hand, was related to greater SCC only within the context of increasing Ab burden. As such, both tau and Ab are associated with SCC and appear to become multiplicative under certain circumstances. It is possible that IT tau-driven SCC reflects individuals likely on the AD trajectory, whereas EC-related SCC captures a range of phenomena including normal aging, in which EC tauopathy is very common. It remains to be seen whether regional tau underlies qualitatively different SCC phenotypes.
P4-326
AUTOMATED MACHINE LEARNING METHODS TO DECTECT UNDIAGNOSED COGNITIVE IMPAIRMENT USING ELECTRONIC MEDICAL RECORDS
Debby Tsuang1,2, Eugene Shao3, Kathryn Chen4, Stephen Thielke1,2, Qing Zeng3, 1VAPSHCS/GRECC, Seattle, WA, USA; 2University of Washington, Seattle, WA, USA; 3George Washington University, Washington, DC, USA; 4Unversity of Washington, Seattle, WA, USA. Contact e-mail:
[email protected] Background: Elderly individuals are at risk for developing dementia and when undiagnosed, significantly increases health-care costs and reduces quality of life. However, screening for cognitive impairment is impractical and time-consuming in the clinical setting. Therefore, we seek to develop an automated method of topic modeling to detect words and phrases—referred to as topics—associated with dementia and thereby identify undiagnosed dementia. Methods: We used a Latent Dirichlet Allocation (LDA) topic-modeling algorithm with a stable topic identification method (sLDA) to identify topics (i.e. recurring themes) in a collection of notes without manual annotations. To capture topics related to early dementia, we applied sLDA to clinic and home-visit notes (n¼871,236) of 1,861 cases (i.e., ICD-9 dementia diagnoses) and 9,305 controls (i.e., no dementia diagnoses or not on anti-dementia medications (matched [5:1] to cases by age, gender, and comorbidity) in VA EMR. For cases, notes were analyzed from 6 months to 3 years prior to diagnosis and for controls during the same approximate timeframe. Dementia specialists subsequently reviewed the same notes in ten cases and ten controls (blinded), diagnosing dementia using DSM-V. Results: In the period prior toICD-9 dementia diagnosis, sLDA identified many topics in cases that indicated dementia: e.g. a topic containing the terms “dementia,” “cognitive impairment,” and names of anti-dementia medications appeared in notes of 74.6% of cases versus 13.46% of controls
P1159
(OR¼18.88; 95% CI 16.74, 21,28). A total of 877 unique and stable topics were generated with nine topic had OR>3.0. Agreement between ICD-9 dementia diagnosis and clinician-assigned diagnosis was high (Kappa¼0.810, 95% C.I. 0.571, 1.0). Conclusions: We used topic modeling to detect topics in clinical notes that are more likely to be associated with dementia cases than controls. Dementia symptoms were documented in notes many months/years prior to ICD-9 dementia diagnosis. In a subset of subjects, we demonstrated strong agreement between previously assigned ICD-9 dementia diagnoses and manual chart reviews by dementia experts. These studies demonstrate the feasibility of using these topics in a separate group of subjects without previous ICD-9/10 dementia diagnoses, to identify individuals who may have a high probability of undiagnosed dementia. These studies will have implications for clinical practice.
P4-327
BODY MASS INDEX (BMI) EFFECTS ON REGIONAL BRAIN VOLUMES IN MILD COGNITIVE IMPAIRMENT (MCI)
Ashley H. Sanderlin, David Todem, Andrea C. Bozoki, Michigan State University, East Lansing, MI, USA. Contact e-mail:
[email protected] Background: Obesity has been linked to lower brain volumes, cogni-
tive deficits and the future development of dementia in middle age adults with normal cognition. Yet, the relationship of obesity and brain volume in MCI is not well known. This study investigated the effect of BMI on regional brain volumes in MCI, and the relationship of brain volume with age and cognitive scores. Methods: We analyzed baseline neuroimaging, behavioral and clinical data from MCI subjects in the ADNI dataset (Phases-1/Go/2). Thirtysix cortical and subcortical regional brain volumes were selected due to previously reported relationships with either BMI or MCI, averaged between hemispheres and corrected by the total intracranial volume. A multivariate analysis of variance model compared brain volumes across three BMI groups: normal weight (NW), overweight (OW) and obese (OB). We also examined the relationship between BMI and MMSE, CDR–sum of boxes (SB) and Geriatric Depression Scale (GDS). Results: The MCI sample consisted of 635 subjects; 216(34%) NW, 282(44%) OW and 137(22%) OB. The mean BMI was 27.1kg/m2, age 71.9 years, education 15.9 years, and 43% were female. Obese subjects were significantly younger than NW, with lower educational attainment and a higher mean GDS score. There was a main effect of BMI on brain volume in 14 regions from frontal, cingulum, occipital, parietal and temporal regions as well as the hippocampus and amygdala. Surprisingly, regional brain volumes were significantly lower in NW subjects for all comparisons. A follow-up correlation analysis confirmed a positive association of BMI with brain volume while age was negatively associated with volume in each region. Neither MMSE nor CDR-SB differed by BMI, however all 14 significant region volumes were positively correlated with the MMSE and 12/14 were negatively correlated with the CDR-SB. Conclusions: In this study, NW MCI subjects were unexpectedly older with lower regional brain volumes than OW/OB subjects. This study may provide neurological evidence for recent studies that have demonstrated a protective effect of an OW BMI on MCI and increased likelihood of NW conversion to dementia. Understanding the interactions of weight, age and cognition may be important in assessing neurologic vulnerability and dementia risk in individuals with MCI.