Risk for dementia in subjective memory impairment, early mild cognitive impairment and late mild cognitive impairment

Risk for dementia in subjective memory impairment, early mild cognitive impairment and late mild cognitive impairment

Poster Presentations: P2 Having higher high-density lipoprotein (HDL) appeared to be protective, tracking with improved performance on verbal learning...

121KB Sizes 1 Downloads 54 Views

Poster Presentations: P2 Having higher high-density lipoprotein (HDL) appeared to be protective, tracking with improved performance on verbal learning (<0.05) while higher low-density lipoprotein (LDL) was associated with lower performance on composite episodic memory. Those healthy participants who had higher PROCAM scores had statistically significant lower cognitive performance on composite episodic memory (<0.05). Conclusions: Vascular factors have an important role in the clinical manifestation of cognitive disease. This study examines a combined cohort of 939 Australians who underwent full cognitive testing and vascular risk profile and had normal cognition. Even in the normal cognitive range of performance, vascular risk presents a significant influence on cognitive performance. Further work examining the follow-up of these 938 individuals at repeat testing is currently underway. P2-174


Bryan James1, Sue Leurgans1, David Bennett2, 1Rush Alzheimer’s Disease Center, Chicago, Illinois, United States; 2Rush Alzheimer’s Disease Center, Chicago, Illinois, United States. Background: Alzheimer’s disease (AD) is listed by the CDC as the sixthleading cause of death in the United States. However, there is a gap between the number of persons who die with AD and the number of deaths attributed to AD based on death certificates. An accurate assessment of the attributable risk of AD on mortality from prospective cohort studies would aid in a more accurate assessment of the burden of mortality due to AD. Population-based cohort studies with intervals of several years between diagnostic assessments may underestimate attributable risk of AD on mortality by failing to account for the impact of persons with rapidly progressive AD. Methods: We used longitudinal data (n ¼ 2,214) from two longitudinal studies of aging and incident dementia with identical inclusion criteria and diagnostic evaluations, the Rush Memory and Aging Project and the Religious Orders Study. Both studies have annual diagnostic assessments. Follow-up among living participants exceeds 90%. Ascertainment of mortality is essentially 100%. Hazard ratio (HR) and population attributable risk (PAR) estimates over 10 years for different ages were derived from Cox proportional hazards models with age as the time scale, left truncation, and time-varying AD status. In addition the full dataset, we created 4 reduced datasets in which waves of data were removed to calculate PARs from assessments every 2, 3, 4, or 5 years. Results: Over an average of 8.19 years, 515 (21%) participants without dementia at baseline developed AD and 899 (37%) died. Median time to death from Kaplan-Meier curves for participants who developed AD was 2.7 years. The mortality HR for AD was 3.16 (CI ¼ 2.37, 4.21). The population attributable risk of AD on mortality over 10 years was 2.9% for ages 70-80; 14.7% for ages 80-90; and 27.7% for ages 90-100. When models were run on the reduced datasets, PAR was dramatically attenuated. Conclusions: AD accounts for a large proportion of the burden of mortality in older persons, and this proportion increases dramatically with age. Cohort studies with more than 1 year between assessments may miss deaths due to AD and significantly underestimate the attributable risk of AD on mortality.



Cedric Annweiler1, Manuel Montero-Odasso2, Robert Bartha3, Olivier Beauchet4, 1Angers University Hospital, Angers University Hospital, France; 2The University of Western Ontario, London, Ontario, Canada; 3Robarts Research Institute, London, Ontario, Canada; 4Angers University Hospital, Angers, France. Background: Vitamin D insufficiency is very common among Alzheimer’s patients, with a prevalence ranging between 70-90%. Although brain atrophy and ventricle enlargement have been reported in rodents deprived in vitamin D, the impact of low serum 25-hydroxyvitamin D (25OHD) on brain trophicity has not been studied yet in humans. The purpose of this cross-sectional study was to determine whether vitamin D insufficiency among older adults could be associated with an increase in lateral cerebral ventricle vol-


ume, a measure of brain atrophy. Methods: Ninety Caucasian communitydwellers with subjective memory complaint and no hydrocephalus (mean age 72.2 6 6.3 years; 45.6% female) from the GAIT study, were divided into 2 groups according to serum 25OHD concentration (either insufficiency 50nmol/L, or normal status >50nmol/L). Cerebral ventricular volume was quantified using semi-automated software from three-dimensional T1weighted Magnetic Resonance Images. Age, gender, education level, use of vitamin D supplements, periventricular white matter lesions and serum calcium concentration were considered as potential confounders. Results: Compared to subjects with a normal 25OHD status (n ¼ 58), subjects with 25OHD insufficiency (n ¼ 32) had enlarged lateral ventricles (47.0 6 27.3 mL versus 36.9 6 16.4 mL respectively, P ¼ 0.032). Vitamin D insufficiency was associated with an increase in cerebral ventricular volume (fully adjusted beta ¼ 10.46 [95CI: 1.30;19.62] with P ¼ 0.026). There was a significant linear association between serum 25OHD concentration and lateral ventricle volume (fully adjusted beta ¼ -0.16 [95CI: -0.31; -0.01] with P ¼ 0.045). The ventricular enlargement specifically involved the main ventricle bodies (P ¼ 0.030) but not the temporal horns (P ¼ 0.245). Conclusions: Serum 25OHD insufficiency was associated with larger lateral cerebral ventricles in the studied cohort of community-dwelling older adults with subjective memory complaint. Continued work is needed toward understanding the role of vitamin D in brain and the impact of vitamin D insufficiency in the course of Alzheimer’s disease. P2-176


Chengxuan Qiu1, Alina Solomon1, Laura Fratiglioni2, Hilkka Soininen3, Jaakko Tuomilehto4, Miia Kivipelto1, 1Karolinska Institutet, Stockholm, Sweden; 2ARC-Karolinska Institutet, Stockholm, Sweden; 3University of Eastern Finland, Kuopio, Finland; 4University of Helsinki, Helsinki, Finland. Background: Systematic review reveals an age-dependent relationship between blood pressure and dementia. We seek to (1) describe variations in blood pressure (BP) over 30 years from middle age to late life by late-life cognitive status, and (2) examine the cross-sectional and longitudinal relation of changes in BP over time to the risk of dementia. Methods: The study population consisted of 1440 participants in the Finnish Cardiovascular Risk Factors, Aging and Dementia (CAIDE) who were initially examined during 1972-1987 as part of the national FINRISK study. Survivors of the CAIDE cohort were reexamined in 1998 and 2005-2008 to detect subjects with dementia following the DSM-IV diagnostic criteria. Arterial BP was measured at each wave of examination. Data were analyzed using the mixed models and logistic regression models controlling for major confounders. Results: The mean age at baseline (1972-1987) was 50.4 (SD, 6.0) years. Dementia was diagnosed in 61 subjects in 1998 and in 47 subjects in 2005-2008. There was no significant association between changes in BP from baseline to 1998 and the risk of dementia diagnosed in 1998. Compared with no decline over the time period, the adjusted odds ratio of dementia diagnosed in 2005-2008 was 4.23 (95% confidence interval, 1.93-9.30) for a decline of more than 15 mm Hg in systolic pressure and 3.50 (1.26-9.72) for a decline of more than 15 mm Hg in diastolic pressure. Conclusions: Substantial decline in blood pressure from middle age to late life is associated with an increased risk of subsequently development of dementia. P2-177


Frank Jessen1, Steffen Wolfsgruber1, Birgit Wiese2, Horst Bickel3, Hanna Kaduszkiewicz4, Michael Pentzek5, Steffi Riedel-Heller6, Siegfried Weyerer7, Martin Scherer4, Wolfgang Maier1, Michael Wagner1, 1 University of Bonn, Bonn, Germany; 2Medical University Hannover, Hannover, Germany; 3Technical University, Munich, Germany; 4University of Hamburg, Hamburg, Germany; 5University of D€usseldorf, D€usseldorf,


Poster Presentations: P2

Germany; 6University of Leipzig, Leipzig, Germany; 7Central Institute for Mental Health, Mannheim, Germany. Background: In the attempt to define earliest at-risk stages for Alzheimer’s Disease (AD) dementia, mild cognitive impairment (MCI) has been subdivided into late MCI (LMCI) and early MCI (eMCI). LMCI is defined by a cognitive test performance of 1.5 standard deviations (SD) below the norm, whereas eMCI refers to a performance of only 1.0 SD below the norm. Epidemiological studies further suggest that pure subjective memory impairment with normal test performance (SMI) is also associated with an increased risk for dementia. We tested the risk for AD dementia for each of these categories in 2892 subjects of the German study on Aging, Cognition and Dementia (AgeCoDe). Methods: Subjects were divided into LMCI, eMCI, SMI and controls by delayed recall performance. In addition, these groups were subdivided by the presence of concerns associated with experienced memory impairment. Risk for AD dementia over six years followup was assessed. Results: The risk for dementia was greatest in LMCI, followed by eMCI and SMI. In those subjects with concerns regarding memory impairment, SMI and eMCI were associated with the same risk for AD dementia, whereas LMCI was associated with a greater risk. In those subgroups without concerns, SMI was not associated with increased risk for AD dementia, whereas eMCI and LMCI were. Conclusions: We confirm LMCI and eMCI as risk conditions of AD dementia. SMI with concerns is as predictive for AD dementia as eMCI. Pre-LMCI risk conditions should be extended to SMI with concerns as eMCI itself may be of limited sensitivity as a risk indicator.

was 87 6 10, 62% are female, and 10% are carrying the ApoEe4 allele. Two distinct nutrient biomarker patterns were associated with rates of cognitive decline: a vitamin pattern high in vitamins B, C, E and D (P ¼ 0.001) and a high trans fat pattern (P <0.001) associated with less and more decline over two years, respectively. These findings were independent of age, gender, education years, APOE4, vascular risk factors, MMSE and multivitamin use. Conclusions: A dietary pattern high in certain vitamins and low in trans-fat may promote cognitive and functional health in older populations at risk for dementia. Unstandardized Coefficients

95.0% Confidence Interval for B



Std. Error


Lower Bound

Upper Bound

Nutrient Biomarker Pattern 1 REGR factor score Nutrient Biomarker Pattern 2 REGR factor score Nutrient Biomarker Pattern 3 REGR factor score Nutrient Biomarker Pattern 4 REGR factor score Nutrient Biomarker Pattern 5 REGR factor score Nutrient Biomarker Pattern 6 REGR factor score Nutrient Biomarker Pattern 7 REGR factor score Nutrient Biomarker Pattern 8 REGR factor score Age at baseline Gender (1¼male) Education years Mini Mental State Exam Multivitamin use (0¼no)









































.003 .132 .029 -.085

.013 .228 .045 .044

.822 .564 .530 .057

-.023 -.324 -.062 -.172

.028 .588 .119 .003






Dependent Variable: Clinical Dementia Rating sum of boxes, annual change P2-178


Gene Bowman1, Jeffrey Kaye2, Joseph Quinn3, 1OHSU, Portland, Oregon, United States; 2Oregon Health & Science University, Portland, Oregon, United States; 3Oregon Health and Sciences University, Portland, Oregon, United States. Background: We previously identified three distinct nutrient biomarker patterns associated with both psychometric and neuroimaging indices in a cross-sectional analysis. The objective of this study was to examine the relationship between the nutrient biomarker patterns and cognitive decline over 2 years. Methods: Thirty biological markers of diet were assayed in plasma drawn from 104 non-demented participants of the Oregon Brain Aging Study. Principal component analysis constructed 8 distinct nutrient biomarker patterns. A linear regression model was used to examine the change in the Clinical Dementia Rating - sum of box score over two years explained by the nutrient biomarker patterns at baseline. Results: Mean age



Gene Bowman1, Hiroko Dodge2, Lynne Shinto3, Lisa Silbert3, Jeffrey Kaye3, Joseph Quinn3, 1OHSU, Portland, Oregon, United States; 2 Oregon Health & Science University, Portland, Oregon, United States; 3 OHSU, Portland, Oregon, United States. Background: Brain white matter hyperintensities (WMH) identified on T2 weighted MRI scans are remarkably prevalent in the elderly and increase the risk for all cause dementia. We previously reported a cross-sectional association of a nutrient biomarker pattern representing the omega-3 polyunsaturated fatty acids (PUFAs, eicosapentaenoic acid and decosahexaenoic acid) and executive function and WMH volume in non-demented elders. The objective of this study was to assess the longitudinal relationship between these pufas and decline in psychometric indices historically sensitive to