- Email: [email protected]
MESIAL TEMPORAL TAU BURDEN IN OLDER ADULTS WITH SUBJECTIVE MEMORY COMPLAINTS RELATES TO POORER MEMORY FUNCTION
P344
Podium Presentations: Wednesday, July 27, 2016
Rating – Sum of Boxes within 128 weeks after randomization. Multiple cut-offs on the FCSRT were explored for their sensitivity and specificity in identification of ‘progressors’ versus ‘non-progressors’. In addition, possible impact on recruitment/screen failure rate was evaluated by calculating the percentage of patients above and below each cut-off. Results: To achieve a high degree of specificity (0.8), excluding most non-progressors, a very low cut-off was necessary (0.37). However, this also reduced sensitivity to a low level (0.44) and resulted in a large potential increase in screen failure (71% of patients did not meet the criterion). A high degree of sensitivity (0.84) was possible and most progressors would be included, but with low specificity (0.34) and a more modest impact on screen failure (28% of patients did not meet the criterion). Conclusions: A cut-off of 0.67 on the FCSRT Index of Cueing was identified as providing the best balance between sensitivity, specificity and impact on screen failure. FCSRT-based cut-offs may be valuable both in identifying pAD patients for enrollment in clinical trials and as part of enrichment strategies to identify patients more likely to experience disease progression during the trial.
O4-05-06
LIMITATIONS OF THE FDA-APPROVED ‘AD CLINICAL TRIAL SIMULATION TOOL’
Richard E. Kennedy1, Gary R. Cutter1, Guoqiao Wang2, Lon S. Schneider3, 1University of Alabama at Birmingham, Birmingham, AL, USA; 2Division of Biostatistics, Washington University School of Medicine, St. Louis, MO, USA; 3Keck School of Medicine of USC, Los Angeles, CA, USA. Contact e-mail: [email protected] Background: Recently, the FDA approved a simulation tool as a “fit-
for-purpose” drug development tool for modeling clinical trials for mild to moderate Alzheimer’s disease (AD). Simulations offer the opportunity to explore different clinical trial designs much more rapidly than “real world” implementations. However, validation of data generated by the ‘AD Clinical Trial Simulation Tool’ (Rogers et al., 2012) against historical clinical trials remains limited to only a small number of studies. Methods: We used a metadatabase of 18 ADCS studies and ADNI to compare predicted values from the AD Clinical Trial Simulation Tool to actual clinical trial results. We included 8 studies with data on the ADAS-cog, baseline MMSE severity, and ApoE genotype, with a total of 2,276 participants with mild AD. We compared the observed values of the ADAS-cog for participants in the placebo arm of each study to the simulated ADAS-cog values for a trial with identical length and age, gender, baseline MMSE, and ApoE distributions using the AD Clinical Trial Simulation Tool. Results: For 7 of the 8 ADCS trials examined, the mean ADAS-cog score in the placebo arm at the conclusion of the trial differed from the predicted ADAS-cog score by more than 2 points, which is the difference that is often used for declaring an intervention to be more effective than placebo. No consistent pattern was observed, with 4 trials over-estimating the placebo effect and 3 trials under-estimating the placebo effect. Conclusions: Historical clinical trials often show significant deviations from model predictions that can exceed the clinical effect of an intervention. Development of models for AD clinical trials should examine additional covariates that may be required to more accurately predict outcomes. Although modelbased simulations are useful for studying the general behavior of AD clinical trials, simulations for a single clinical trial may not accurately reflect trial outcome. Simulated results of a single clinical trial must be interpreted cautiously, particularly for deciding whether to proceed with a “real world” trial implementation.
WEDNESDAY, JULY 27, 2016 ORAL SESSIONS O4-06 CLINICAL (NEUROPSYCHIATRY AND BEHAVIORAL NEUROLOGY): SUBJECTIVE COGNITIVE IMPAIRMENT AND MILD COGNITIVE IMPAIRMENT O4-06-01
MESIAL TEMPORAL TAU BURDEN IN OLDER ADULTS WITH SUBJECTIVE MEMORY COMPLAINTS RELATES TO POORER MEMORY FUNCTION
Rachel F. Buckley1,2, Christopher C. Rowe3, Vincent Dore4, Simon M. Laws5, Pierrick Bourgeat6, Samantha Burnham7, Olivier Salvado6, Paul Maruff2,8, S. Lance Macaulay9, Ralph N. Martins5,10, Colin L. Masters1, Victor L. Villemagne11,12, 1University of Melbourne, Melbourne, Australia; 2The Florey Institutes of Neurosciences and Mental Health, Melbourne, Australia; 3Austin Health, Melbourne, Australia; 4 CSIRO, Melbourne, Australia; 5Edith Cowan University, Perth, Australia; 6 CSIRO, Brisbane, Australia; 7CSIRO, Perth, Australia; 8Cogstate Ltd., Melbourne, Australia; 9CSIRO Food and Nutrition, Parkville, Australia; 10 Sir James McCusker Alzheimer’s Disease Research Unit (Hollywood Private Hospital), Perth, Australia; 11Department of Molecular Imaging and Therapy, Centre for PET, Austin Health, Heidelberg, Australia; 12The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia. Contact e-mail: [email protected]
Background: Tau aggregation in the brain is related with neurode-
generative changes and subsequent declines in cognitive function. Little is known about the relationship between subjective memory complaints (SMC) and evidence of global or regional tau deposition. Our aim was to elucidate this relationship, and determine whether these factors would work in concert with amyloid (Ab) and apolipoprotein ε4 genotype (APOEε4) to influence memory and executive functioning in healthy older adults. Methods: Healthy participants from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study were dichotomised into those with memory complaints (HC-SMC: n ¼ 64) and without (HC-NMC: n ¼ 23). Participants underwent Ab (18F-florbetapir and 18F-flutemetamol) and tau PET neuroimaging (18F-AV1451; n ¼ 60, 18F-THK5351; n ¼ 27) as well as APOEε4 genotyping. Memory complaints were measured with the MAC-Q questionnaire. Verbal and non-verbal memory, executive function and language were the cognitive domains of interest. Ab and tau burden were divided into high and low according to predetermined SUVR cut-offs for each tracer. Tau was further separated into three composite regions (mesial temporal (Me), temporoparietal (Te), and rest of neocortex(R)). Results: Relative to HC-NMC, HC-SMC exhibited greater tau burden in Me and Te regions, but not in R. In those with high Me tau burden, HCSMC showed significantly poorer verbal delayed recall in comparison with HC-NMC (hp2 ¼ 0.08). By contrast, executive function was better in HC-SMC compared with HC-NMC (hp2 ¼ 0.06). These medium-sized effects were sustained after adjusting for Ab and APOEε4 status, and were not evident in the low Me tau burden group. There was no effect on non-verbal memory, language or depression, and no effects were found for other tau regions of interest. In HC-SMC with high Me burden, poorer executive function correlated with greater SMC severity. Conclusions: These are the first results to report regional tau burden relationships with SMC in healthy older adults. The findings suggest an interaction between SMC and Me tau burden, such that those with evidence of both show poorer verbal memory performance but better executive function. Executive function underlies metacognitive monitoring, so it
Podium Presentations: Wednesday, July 27, 2016
is possible that better performance may well drive better detection of memory change. O4-06-02
FRAILTY IS ASSOCIATED WITH SUBJECTIVE COGNITIVE DECLINE IN OLDER FEMALE ADULTS WITHOUT DEMENTIA: THE VANDERBILT MEMORY & AGING PROJECT
Susan Bell, Dandan Liu, Jacquelyn E. Neal, Katherine A. Gifford, Timothy J. Hohman, Angela L. Jefferson, Vanderbilt University School of Medicine, Nashville, TN, USA. Contact e-mail: susan.p.bell@vanderbilt. edu Background: Subjective cognitive decline (SCD) predicts cognitive progression and diagnostic conversion and may represent early changes across the cognitive spectrum. Physical frailty represents a loss of physiological reserve and is associated with cognitive impairment and incident dementia. We hypothesized that early frailty markers are associated with increased SCD. Methods: Participants with normal cognition (n¼164, 60% male) and mild cognitive impairment (n¼165, 59% male) were drawn from the Vanderbilt Memory & Aging Project, a case-control cohort. Markers of frailty (gait speed, grip strength, exhaustion, physical activity) were measured using standard methods and a composite frailty score was calculated by averaging the individual component z-scores. SCD was quantified using a multi-questionnaire protocol, including a total SCD inventory and the Everyday Cognition Scale (ECog). Proportional odds models, stratified by sex, related gait speed, grip strength, and composite frailty score to total SCD score and total ECog scale adjusting for age, education, body mass index, cognitive diagnosis, depression, Framingham Stroke Risk Profile (including age, sex, systolic blood pressure, anti-hypertensive medication usage, diabetes, left ventricular hypertrophy, atrial fibrillation, and prevalent cardiovascular disease), and height (in gait speed models only). Results: Mean age of participants was 7367 years; gait speed was 1.1160.22 m/sec; grip strength was 31611 kg; and total SCD score was 297697. 46% of women and 32% of men were categorized as pre-frail using standard criteria. Using a false discovery rate of 0.1 for multiple comparisons, gait speed, grip strength and composite frailty were not associated with total SCD in men. In comparison, increasing gait speed was associated with lower total SCD (OR¼0.09, 95%CI 0.01-0.75, p¼0.025) and total ECog score (OR¼0.07, 95%CI 0.01-0.44, p¼0.0049). Increasing composite frailty was associated only with the total ECog score (OR¼2.01, 95%CI 1.05-3.85, p¼0.0365). Conclusions: Among older women without dementia, early markers of frailty (gait speed, composite frailty score) are associated with more SCD. Findings suggest frailty (an independent predictor of the development and progression of cognitive impairment) is related to other early markers of dementia. Further studies should investigate underlying mechanisms linking early changes in frailty, SCD, and cognition. O4-06-03
DIFFERENCES IN QUANTITATIVE METHODS FOR MEASURING SUBJECTIVE COGNITIVE DECLINE: RESULTS FROM A PROSPECTIVE MEMORY CLINIC STUDY
Asmus Vogel1,2, Lise Cronberg Salem2, Birgitte Bo Andersen2, Gunhild Waldemar2, 1University of Copenhagen, Copenhagen, Denmark; 2 Danish Dementia Research Centre, Rigshospitalet, Copenhagen, Denmark. Contact e-mail: [email protected] Background: Cognitive complaints occur frequently in elderly
people and may be a risk factor for dementia and cognitive
P345
decline. Results from studies on subjective cognitive decline are difficult to compare due to variability in assessment methods, and little is known about how different methods influence reports of cognitive decline. Methods: The Subjective Memory Complaints Scale (SMC) and The Memory Complaint Questionnaire (MAC-Q) were applied in 121 mixed memory clinic patients with mild cognitive symptoms (mean MMSE ¼ 26.8, SD¼2.7). The scales were applied independently and raters were blinded to results from the other scale. Scales were not used for diagnostic classification. Cognitive performances were measured by MMSE and Addenbrooke’s Cognitive Examination. Depressive symptoms were rated using Major Depression Inventory. We studied the association between the two measures and investigated the scales’ relation to depressive symptoms, age and cognitive status. Results: SMC and MAC-Q were significantly associated (r¼0.44, N¼121, p¼0.015) and both scales had a wide range of scores. In this mixed memory clinic cohort lower age was significantly associated to higher subjective scores as opposed to findings from population based studies where subjective symptoms increase with age. There were no significant correlations between cognitive test performances and scales measuring subjective decline. Depression scores were significantly correlated to both scales measuring subjective decline. Linear regression models showed that age did not have a significant contribution to the variance in subjective memory beyond that of depressive symptoms. Conclusions: Measures for subjective cognitive decline are not interchangeable when used in memory clinics and the application of different scales in previous studies is an important factor as to why studies show variability in the association between subjective cognitive decline and background data and/or clinical results. Careful consideration should be taken as to which questions are relevant and have validity when operationalizing subjective cognitive decline.
O4-06-04
NEUROIMAGING CORRELATES OF ANOSOGNOSIA IN MILD COGNITIVE IMPAIRMENT
Patrizia Vannini1,2,3, Bernard J. Hanseeuw1,2, Catherine E. Munro2, Rebecca Amariglio4, Gad A. Marshall3,5, Dorene M. Rentz4,6, Alvaro Pascual-Leone7, Keith Johnson1,3,5,8, Reisa A. Sperling4,5,9, 1 Athinoula A. Martinos Center for Biomedical Imaging and the Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA; 2Massachusetts General Hospital, Charlestown, MA, USA; 3Center for Alzheimer Research and Treatment, Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 4Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 5 Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 6Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA; 7Berenson-Allen Center for Noninvasive Brain Stimulation, Division of Cognitive Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA; 8Department of Radiology, Division of Molecular Imaging and Nuclear Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 9Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA. Contact e-mail: [email protected] Background: Anosognosia, or loss of insight of memory deficits is a common and striking symptom in Alzheimer’s disease (AD). Previous findings in AD patients suggest that anosognosia is due to both
Podium Presentations: Wednesday, July 27, 2016
Rating – Sum of Boxes within 128 weeks after randomization. Multiple cut-offs on the FCSRT were explored for their sensitivity and specificity in identification of ‘progressors’ versus ‘non-progressors’. In addition, possible impact on recruitment/screen failure rate was evaluated by calculating the percentage of patients above and below each cut-off. Results: To achieve a high degree of specificity (0.8), excluding most non-progressors, a very low cut-off was necessary (0.37). However, this also reduced sensitivity to a low level (0.44) and resulted in a large potential increase in screen failure (71% of patients did not meet the criterion). A high degree of sensitivity (0.84) was possible and most progressors would be included, but with low specificity (0.34) and a more modest impact on screen failure (28% of patients did not meet the criterion). Conclusions: A cut-off of 0.67 on the FCSRT Index of Cueing was identified as providing the best balance between sensitivity, specificity and impact on screen failure. FCSRT-based cut-offs may be valuable both in identifying pAD patients for enrollment in clinical trials and as part of enrichment strategies to identify patients more likely to experience disease progression during the trial.
O4-05-06
LIMITATIONS OF THE FDA-APPROVED ‘AD CLINICAL TRIAL SIMULATION TOOL’
Richard E. Kennedy1, Gary R. Cutter1, Guoqiao Wang2, Lon S. Schneider3, 1University of Alabama at Birmingham, Birmingham, AL, USA; 2Division of Biostatistics, Washington University School of Medicine, St. Louis, MO, USA; 3Keck School of Medicine of USC, Los Angeles, CA, USA. Contact e-mail: [email protected] Background: Recently, the FDA approved a simulation tool as a “fit-
for-purpose” drug development tool for modeling clinical trials for mild to moderate Alzheimer’s disease (AD). Simulations offer the opportunity to explore different clinical trial designs much more rapidly than “real world” implementations. However, validation of data generated by the ‘AD Clinical Trial Simulation Tool’ (Rogers et al., 2012) against historical clinical trials remains limited to only a small number of studies. Methods: We used a metadatabase of 18 ADCS studies and ADNI to compare predicted values from the AD Clinical Trial Simulation Tool to actual clinical trial results. We included 8 studies with data on the ADAS-cog, baseline MMSE severity, and ApoE genotype, with a total of 2,276 participants with mild AD. We compared the observed values of the ADAS-cog for participants in the placebo arm of each study to the simulated ADAS-cog values for a trial with identical length and age, gender, baseline MMSE, and ApoE distributions using the AD Clinical Trial Simulation Tool. Results: For 7 of the 8 ADCS trials examined, the mean ADAS-cog score in the placebo arm at the conclusion of the trial differed from the predicted ADAS-cog score by more than 2 points, which is the difference that is often used for declaring an intervention to be more effective than placebo. No consistent pattern was observed, with 4 trials over-estimating the placebo effect and 3 trials under-estimating the placebo effect. Conclusions: Historical clinical trials often show significant deviations from model predictions that can exceed the clinical effect of an intervention. Development of models for AD clinical trials should examine additional covariates that may be required to more accurately predict outcomes. Although modelbased simulations are useful for studying the general behavior of AD clinical trials, simulations for a single clinical trial may not accurately reflect trial outcome. Simulated results of a single clinical trial must be interpreted cautiously, particularly for deciding whether to proceed with a “real world” trial implementation.
WEDNESDAY, JULY 27, 2016 ORAL SESSIONS O4-06 CLINICAL (NEUROPSYCHIATRY AND BEHAVIORAL NEUROLOGY): SUBJECTIVE COGNITIVE IMPAIRMENT AND MILD COGNITIVE IMPAIRMENT O4-06-01
MESIAL TEMPORAL TAU BURDEN IN OLDER ADULTS WITH SUBJECTIVE MEMORY COMPLAINTS RELATES TO POORER MEMORY FUNCTION
Rachel F. Buckley1,2, Christopher C. Rowe3, Vincent Dore4, Simon M. Laws5, Pierrick Bourgeat6, Samantha Burnham7, Olivier Salvado6, Paul Maruff2,8, S. Lance Macaulay9, Ralph N. Martins5,10, Colin L. Masters1, Victor L. Villemagne11,12, 1University of Melbourne, Melbourne, Australia; 2The Florey Institutes of Neurosciences and Mental Health, Melbourne, Australia; 3Austin Health, Melbourne, Australia; 4 CSIRO, Melbourne, Australia; 5Edith Cowan University, Perth, Australia; 6 CSIRO, Brisbane, Australia; 7CSIRO, Perth, Australia; 8Cogstate Ltd., Melbourne, Australia; 9CSIRO Food and Nutrition, Parkville, Australia; 10 Sir James McCusker Alzheimer’s Disease Research Unit (Hollywood Private Hospital), Perth, Australia; 11Department of Molecular Imaging and Therapy, Centre for PET, Austin Health, Heidelberg, Australia; 12The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia. Contact e-mail: [email protected]
Background: Tau aggregation in the brain is related with neurode-
generative changes and subsequent declines in cognitive function. Little is known about the relationship between subjective memory complaints (SMC) and evidence of global or regional tau deposition. Our aim was to elucidate this relationship, and determine whether these factors would work in concert with amyloid (Ab) and apolipoprotein ε4 genotype (APOEε4) to influence memory and executive functioning in healthy older adults. Methods: Healthy participants from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study were dichotomised into those with memory complaints (HC-SMC: n ¼ 64) and without (HC-NMC: n ¼ 23). Participants underwent Ab (18F-florbetapir and 18F-flutemetamol) and tau PET neuroimaging (18F-AV1451; n ¼ 60, 18F-THK5351; n ¼ 27) as well as APOEε4 genotyping. Memory complaints were measured with the MAC-Q questionnaire. Verbal and non-verbal memory, executive function and language were the cognitive domains of interest. Ab and tau burden were divided into high and low according to predetermined SUVR cut-offs for each tracer. Tau was further separated into three composite regions (mesial temporal (Me), temporoparietal (Te), and rest of neocortex(R)). Results: Relative to HC-NMC, HC-SMC exhibited greater tau burden in Me and Te regions, but not in R. In those with high Me tau burden, HCSMC showed significantly poorer verbal delayed recall in comparison with HC-NMC (hp2 ¼ 0.08). By contrast, executive function was better in HC-SMC compared with HC-NMC (hp2 ¼ 0.06). These medium-sized effects were sustained after adjusting for Ab and APOEε4 status, and were not evident in the low Me tau burden group. There was no effect on non-verbal memory, language or depression, and no effects were found for other tau regions of interest. In HC-SMC with high Me burden, poorer executive function correlated with greater SMC severity. Conclusions: These are the first results to report regional tau burden relationships with SMC in healthy older adults. The findings suggest an interaction between SMC and Me tau burden, such that those with evidence of both show poorer verbal memory performance but better executive function. Executive function underlies metacognitive monitoring, so it
Podium Presentations: Wednesday, July 27, 2016
is possible that better performance may well drive better detection of memory change. O4-06-02
FRAILTY IS ASSOCIATED WITH SUBJECTIVE COGNITIVE DECLINE IN OLDER FEMALE ADULTS WITHOUT DEMENTIA: THE VANDERBILT MEMORY & AGING PROJECT
Susan Bell, Dandan Liu, Jacquelyn E. Neal, Katherine A. Gifford, Timothy J. Hohman, Angela L. Jefferson, Vanderbilt University School of Medicine, Nashville, TN, USA. Contact e-mail: susan.p.bell@vanderbilt. edu Background: Subjective cognitive decline (SCD) predicts cognitive progression and diagnostic conversion and may represent early changes across the cognitive spectrum. Physical frailty represents a loss of physiological reserve and is associated with cognitive impairment and incident dementia. We hypothesized that early frailty markers are associated with increased SCD. Methods: Participants with normal cognition (n¼164, 60% male) and mild cognitive impairment (n¼165, 59% male) were drawn from the Vanderbilt Memory & Aging Project, a case-control cohort. Markers of frailty (gait speed, grip strength, exhaustion, physical activity) were measured using standard methods and a composite frailty score was calculated by averaging the individual component z-scores. SCD was quantified using a multi-questionnaire protocol, including a total SCD inventory and the Everyday Cognition Scale (ECog). Proportional odds models, stratified by sex, related gait speed, grip strength, and composite frailty score to total SCD score and total ECog scale adjusting for age, education, body mass index, cognitive diagnosis, depression, Framingham Stroke Risk Profile (including age, sex, systolic blood pressure, anti-hypertensive medication usage, diabetes, left ventricular hypertrophy, atrial fibrillation, and prevalent cardiovascular disease), and height (in gait speed models only). Results: Mean age of participants was 7367 years; gait speed was 1.1160.22 m/sec; grip strength was 31611 kg; and total SCD score was 297697. 46% of women and 32% of men were categorized as pre-frail using standard criteria. Using a false discovery rate of 0.1 for multiple comparisons, gait speed, grip strength and composite frailty were not associated with total SCD in men. In comparison, increasing gait speed was associated with lower total SCD (OR¼0.09, 95%CI 0.01-0.75, p¼0.025) and total ECog score (OR¼0.07, 95%CI 0.01-0.44, p¼0.0049). Increasing composite frailty was associated only with the total ECog score (OR¼2.01, 95%CI 1.05-3.85, p¼0.0365). Conclusions: Among older women without dementia, early markers of frailty (gait speed, composite frailty score) are associated with more SCD. Findings suggest frailty (an independent predictor of the development and progression of cognitive impairment) is related to other early markers of dementia. Further studies should investigate underlying mechanisms linking early changes in frailty, SCD, and cognition. O4-06-03
DIFFERENCES IN QUANTITATIVE METHODS FOR MEASURING SUBJECTIVE COGNITIVE DECLINE: RESULTS FROM A PROSPECTIVE MEMORY CLINIC STUDY
Asmus Vogel1,2, Lise Cronberg Salem2, Birgitte Bo Andersen2, Gunhild Waldemar2, 1University of Copenhagen, Copenhagen, Denmark; 2 Danish Dementia Research Centre, Rigshospitalet, Copenhagen, Denmark. Contact e-mail: [email protected] Background: Cognitive complaints occur frequently in elderly
people and may be a risk factor for dementia and cognitive
P345
decline. Results from studies on subjective cognitive decline are difficult to compare due to variability in assessment methods, and little is known about how different methods influence reports of cognitive decline. Methods: The Subjective Memory Complaints Scale (SMC) and The Memory Complaint Questionnaire (MAC-Q) were applied in 121 mixed memory clinic patients with mild cognitive symptoms (mean MMSE ¼ 26.8, SD¼2.7). The scales were applied independently and raters were blinded to results from the other scale. Scales were not used for diagnostic classification. Cognitive performances were measured by MMSE and Addenbrooke’s Cognitive Examination. Depressive symptoms were rated using Major Depression Inventory. We studied the association between the two measures and investigated the scales’ relation to depressive symptoms, age and cognitive status. Results: SMC and MAC-Q were significantly associated (r¼0.44, N¼121, p¼0.015) and both scales had a wide range of scores. In this mixed memory clinic cohort lower age was significantly associated to higher subjective scores as opposed to findings from population based studies where subjective symptoms increase with age. There were no significant correlations between cognitive test performances and scales measuring subjective decline. Depression scores were significantly correlated to both scales measuring subjective decline. Linear regression models showed that age did not have a significant contribution to the variance in subjective memory beyond that of depressive symptoms. Conclusions: Measures for subjective cognitive decline are not interchangeable when used in memory clinics and the application of different scales in previous studies is an important factor as to why studies show variability in the association between subjective cognitive decline and background data and/or clinical results. Careful consideration should be taken as to which questions are relevant and have validity when operationalizing subjective cognitive decline.
O4-06-04
NEUROIMAGING CORRELATES OF ANOSOGNOSIA IN MILD COGNITIVE IMPAIRMENT
Patrizia Vannini1,2,3, Bernard J. Hanseeuw1,2, Catherine E. Munro2, Rebecca Amariglio4, Gad A. Marshall3,5, Dorene M. Rentz4,6, Alvaro Pascual-Leone7, Keith Johnson1,3,5,8, Reisa A. Sperling4,5,9, 1 Athinoula A. Martinos Center for Biomedical Imaging and the Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA; 2Massachusetts General Hospital, Charlestown, MA, USA; 3Center for Alzheimer Research and Treatment, Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 4Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 5 Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 6Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA; 7Berenson-Allen Center for Noninvasive Brain Stimulation, Division of Cognitive Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA; 8Department of Radiology, Division of Molecular Imaging and Nuclear Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 9Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA. Contact e-mail: [email protected] Background: Anosognosia, or loss of insight of memory deficits is a common and striking symptom in Alzheimer’s disease (AD). Previous findings in AD patients suggest that anosognosia is due to both