- Email: [email protected]
P2-186
Poster P2:: Monday Posters decline in both BADLs and IADLs was significantly greater (p⫽0.015 and p⫽0.002 respectively) in CIND-AD than in the CIND-Stable group. The baseline DAD score did not contribute significantly to the prediction of progression to AD (OR 1.0, 95% C.I. 0.9-1.1). Conclusion: As expected, individuals progressing to AD showed a greater decline in their functional abilities; however, the DAD total score at entry did not predict progression to AD. We anticipate that an evaluation of higher social functioning is likely required to detect the earliest significant changes in function that may herald or predict progression to AD. P2-184
CAUSES AND OUTCOMES OF COGNITIVE IMPAIRMENT NOT DEMENTIA (CIND) IN A NATIONAL SAMPLE OF U.S. ADULTS
Kenneth M. Langa1, Brenda L. Plassman2, Gwenith G. Fisher3, Guy G. Potter4, David R. Weir3, Robert B. Wallace5, 1University of Michigan, Ann Arbor, MI, USA; 2Duke University, Durham, NC, USA; 3 University of Michigan, Ann Arbor, MI, USA; 4Duke University, Durham, NC, USA; 5University of Iowa, Iowa City, Iowa, USA. Contact e-mail: [email protected] Background: Population-based studies of Cognitive Impairment that is Not Dementia (CIND) are rare. The causes (i.e., sub-types) and outcomes of CIND in population-based samples may be different than those in clinic-based samples. Objectives: 1) To identify sub-types of CIND in a nationally representative population-based sample of US adults; and 2) to determine the 18-month outcomes for progression to dementia and death among these CIND sub-types. Methods: The Aging, Demographics, and Memory Study (ADAMS), a national population-based study of 856 US adults age 70 or older was used. ADAMS CIND sub-types were defined based on clinical and neuropsychological criteria from prior studies. Extensive in-home clinical and neuropsychological assessments were performed, and an expert consensus panel used the assessments to assign a diagnosis of normal, CIND, or dementia. Sub-types of CIND were assigned to denote the cause of cognitive impairment. An 18-month follow-up in-home assessment was performed for all those diagnosed with CIND at the baseline assessment. Results: CIND was diagnosed in 23% of study participants. Prodromal Alzheimer’s disease (AD) was the most common CIND sub-type (39% of cases). Other medical conditions (e.g., heart disease, lung disease, diabetes) accounted for 23% of CIND cases, while Stroke (14%) and Vascular Cognitive Impairment (8%) were also common. Those with CIND due to Prodromal AD were significantly more likely to progress to dementia at 18 months compared to the other CIND groups (30% vs. 17%; P⬍.05). Those with CIND due to other medical conditions were most likely to die compared to the other CIND groups (28% vs. 12%; P⬍.05), but were also most likely to show significant cognitive improvement (20% vs. 12%, P⬍.05). Conclusions: Nearly one-quarter of US adults age 70 or older have CIND. CIND due to Pro-dromal AD is associated with a very high likelihood of progression to dementia (30% in 18 months). Chronic medical conditions (e.g., heart disease, lung disease, and diabetes) are important causes for CIND in the community, and are associated with both high rates of death AND cognitive improvement. This important and heterogenous group of older adults with CIND has likely been under-represented in prior studies of clinic-based samples. P2-185
METABOLIC AND COGNITIVE CHANGES WITH COMPUTER BASED COGNITIVE THERAPY FOR MCI
William Jagust1, Beth Mormino1, Charles DeCarli2, Joel Kramer3, Deborah Barnes3, Bruce Reed2, 1University of California, Berkeley, CA, USA; 2University of California, Davis, CA, USA; 3University of California, San Francisco, CA, USA. Contact e-mail: [email protected] Background: Observational studies have shown that cognitive activity may reduce the risk of development of dementia, and imaging studies suggest that changes in brain function result from cognitive training.
S289
Patients with mild cognitive impairment (MCI) might benefit from cognitive stimulation in order to reduce the risk of progression to dementia. Objective: We performed this study in order to investigate cognitive and brain metabolic changes (measured with fluorodeoxyglucose PET) resulting from a rigorous cognitive training program. Methods: Six patients with MCI (mean age 71.5, 3 men) were randomized to receive a 100 min/day computerized intervention for 4-7 weeks that was designed to improve brain processing of auditory stimuli. A control group of 9 MCI patients (mean age 71.5, 6 men) read on-line newspapers, listened to audio books, and played a computer video game for comparable times. FDG-PET was performed at baseline and following treatment, as was the Repeatable Battery for Assessment of Neuropsychological Status (RBANS). PET data were analyzed using spatially normalized scans and SPM2. Contrasts included both pre- and post-treatment comparisons for each group separately, and regressions between change in performance on the RBANS and change in glucose metabolism. Results: Groups were comparable on baseline MMSE scores (27.3 treatment and 26.9 control). Comparison of preand post-treatment PET showed reductions in glucose metabolism in right middle frontal lobes and right thalamus in the control group and no change in the treated group. In the treated group, improvements in the RBANS attention scale were correlated with increases in glucose metabolism in the right precuneus and right parietal lobe. Improvements in the RBANS visuospatial scale were correlated with increased metabolism in bilateral anterior and middle cingulate gyri and right insula. Conclusions: These results suggest the possibility that cognitive training slows the rate of metabolic decline in MCI. Individual patterns of behavioral change were related to underlying neural substrates, suggesting the possibility that this treatment approach may produce regionally specific effects on brain function. P2-186
TREATMENT OF MILD COGNITIVE IMPAIRMENT WITH VITAMIN E AND DONEPEZIL: CORRELATIONS BETWEEN SERIAL MRI AND CONVERSION, GENOTYPE, AND TREATMENT STATUS
Leon J. Thal1, Clifford R. Jack, Jr.2, Michael Grundman3, Chadwick Ward2, Drahomira Sencakova2, Shelia Jin4, Anthony Gamst4, Ronald C. Petersen2, The Alzheimer’s Disease Cooperative Study Group4, 1University of California, San Diego, San Diego, CA, USA; 2 Mayo Clinic, Rochester, MN, USA; 3Elan Pharmaceuticals, San Diego, USA; 4University of California, San Diego, San Diego, USA. Contact e-mail: [email protected] Background: The vitamin E and donepezil study for the treatment of MCI was conducted in 769 subjects at 69 North American centers. Clinical results of this study have been reported (NEJM 2005; 352(23):2379-88). Twenty-four of the recruiting sites voluntarily participated in an MRI substudy. Objective(s): We report correlations between rates of atrophy from serial MRI and clinical conversion to AD, APO E genotype, and treatment arm in those who participated in MRI. Methods: A total of 194 subjects participated in the MRI substudy and 137 completed both a baseline and followup scan allowing for computation of rates of change over time. The followup MRI scan for each subject could have been either at the end of the study (36 months) or at an interim time point associated with clinical conversion from MCI to AD. In each subject, change from baseline to followup was measured for four different brain volumes; hippocampus, entorhinal cortex (ERC), whole brain, and ventricle. Results: For each of the four brain atrophy rate measures, annualized rates of atrophy were greater in those subjects who converted to AD vs nonconverters (all p ⬍0.000). Annualized rates of hippocampal (p⬍0.05), ERC (p⬍0.05), ventricular (p⫽0.002), and whole brain (p⫽0.004) atrophy were greater in APOE 4 carriers than non-carriers. A non significant trend consistent with treatment effect was observed across all four mean MRI rate measures for both donepezil and vitamin E. The number of treated subjects for which MRI measures were available varied for different MRI measures. Conclusions: Expected and highly significant correlations were
S290
Poster P2:: Monday Posters
observed between rates of brain volume change measured with MRI and both clinical conversion status as well as APOE genotype. These results validate the technical feasibility of using MRI as an outcome measure of disease progression in multi-center trials for amnestic MCI. Failure to detect a significant treatment effect in the MRI data could reflect the absence of a biologic effect of either vitamin E or donepezil on brain volume itself. Or it could simply represent inadequate sample size in the treatment groups: donepezil (n⫽ 36 - 40, number varies with different MRI measures), vitamin E (n ⫽ 34 -43). P2-187
COMPARISON BETWEEN THE STANDARD MINI MENTAL STATE EXAMINATION AND THE MODIFIED VERSION (3MS) IN HEALTHY SUBJECTS AND PATIENTS WITH MILD COGNITIVE IMPAIRMENT OR DEMENTIA. ME´RIDA-VENEZUELA, 2005-2006
Clara I. Ramirez1, Trino J. Baptista2, Ana Serrano3, Yinet Arape3, Evelyn Thonon4, 1Department of Neurology, Los Andes University Hospital, Merida, Venezuela; 2Department of Physiology, Los Andes University Medical School, Merida, Venezuela; 3Department of Psychiatry, Los Andes University Hospital, Merida, Venezuela; 4 Department of Neurology, El Vigia Hospital, Merida, Venezuela. Contact e-mail: [email protected] Introduction: Current clinical practice and research request for precise and accessible instruments to detect early cognitive impairment (CI). Folstein’s Mini Mental State Examination (MMSE)1 is widely used in screening studies. Unfortunately, its specificity is hampered by being affected by age, education level and cultural environment. It is also very sensitive to moderate to severe-, but not to mild cognitive impairment (MCI). Teng y Chang Chui developed a more extensive version (3MS)2 which may overcome such limitations. Objectives: Since the 3MS has not been validated in Venezuela, we correlated both instruments’ scores and evaluated their sensitivity and specificity compared to the clinical diagnosis. Methods: The scales were administered in a random sequence to 93 outpatients aged 50 or more years who were assessed in the neurology or psychiatry departments. Cut-off scores for CI were 23 for the MMSE and ⬍ 79 for the 3MS. Results: We evaluated 36 healthy subjects (89% women; age (years: average ⫾ SD): 65.4 ⫾ 7.7); 25 with MCI (72% women; age: 69.4 ⫾ 10.1) and 32 demented patients (78% women; age: 73.7 ⫾ 7.3). A significant positive correlation was found between the MMSE and the 3MS: whole sample [r (93) ⫽ 0.876, p ⫽ 0.0001]; healthy subjects [r (36) ⫽ 0.466, p ⫽ 0.004]; MCI [r (25) ⫽ 0.644, p ⫽ 0.001] and dementia [r (32) ⫽ 0.854, p ⫽ 0.0001]. The MMSE specificity reached 100%, but its sensitivity was of 12% for MCI and 44 % for dementia. The 3MS specificity was of 83.3%, and the sensitivity reached 44% for MCI and 69% for dementia. Conclusions: In spite of an excellent numeric correlation between both instruments, the MMSE displayed a very low sensitivity for the MCI in our population. Since sensitivity was better for the 3MS, additional validation studies for this instrument are warranted in Venezuela. References: (1) Folstein, M., Folstein, S., McHugh P. (1975) Mini-Mental State Examination: a practical method for grading the cognitive sate of patients for the clinician. J Psychiatr Res, 12: 189-198. (1) Teng, E., Chang, H. (1987) The Modified Mini-Mental State (3MS) Examination. J Clin Psychiatry, 48: 314-318. P2-188
PREDICTION OF COGNITIVE IMPAIRMENT AMONG ELDERLY PERSONS USING A BRIEF COGNITIVE SCREENING BATTERY
Anthony LoGalbo1, Laura Baker2, Suzanne Craft1,2, The ADAPT Research Group2, 1VA Puget Sound Health Care System, Seattle, WA, USA; 2University of Washington, Seattle, WA, USA. Contact e-mail: [email protected] Background: Early detection of cognitive impairment among the elderly is becoming increasingly important. There is a need for brief measures that
can reliably predict the presence of clinically significant cognitive impairment. Objective: To decipher which brief cognitive tests most reliably predict cognitive impairment, particularly amnestic Mild Cognitive Impairment (aMCI), among elderly persons. Methods: Participants aged 70⫹ with normal baseline cognitive functioning were subsequently administered yearly 30-35 minute cognitive screening batteries, including the Modified Mini Mental State Examination (3MSE), generative verbal fluency (supermarket items), Narratives from the Rivermead Behavioral Memory Test (RBMT), Brief Visuospatial Memory Test, Revised (BVMT-R), Hopkins Verbal Learning Test, Revised (HVLT-R), digit span (DS), and the Geriatric Depression Scale (GDS). When participants’ scores on these measures fell below pre-specified cut points, a more expansive dementia evaluation was triggered to assist in diagnosing the presence and extent of cognitive impairment. Results: Scores on the cognitive screening measures were entered into a stepwise logistic regression analysis to predict outcome of the dementia evaluation. The 3MSE, RBMT delay, HVLT-R delay, DS backward, and GDS were identified as significant predictors of normal versus impaired cognitive functioning. However, when participants with normal cognitive functioning were compared only to those with aMCI, the 3MSE and RBMT delay were the only significant predictors. Furthermore, only the 3MSE was identified as a significant predictor of aMCI versus Alzheimer’s disease (AD). Conclusions: Brief cognitive screening measures are capable of differentiating between normal and impaired cognitive functioning upon more in-depth cognitive evaluation. The fact that fewer and less diverse measures reliably predicted cognitive impairment when the outcome was narrowed to normal versus aMCI supports the notion that diverse cognitive impairment can result from a variety of factors among the elderly. This finding also highlights the benefit of using certain screening measures to assist in differentiating aMCI from other forms of cognitive impairment. Results of this study further suggest that some measures are not helpful in predicting dementia evaluation outcome. Thus, their utility in triggering a dementia evaluation should be further examined. Overall results of this study have implications for the effective early evaluation and diagnosis of aMCI. P2-189
A METHODOLOGICAL APPROACH FOR ASSESSING NEUROPSYCHOLOGICAL CHANGES IN THE PRECLINICAL STAGE OF ALZHEIMER’S DISEASE (AD)
Andreas U. Monsch1, Antoinette Zehnder1, Stefan Blaesi1, Manfred Berres2, Rene´ Spiegel1, Hannes B. Staehelin1, 1Memory Clinic, Basel, Switzerland; 2RheinAhrCampus, Remagen, Germany. Contact e-mail: [email protected] Background: The characterization of cognitive changes in healthy elderly individuals is a critical prerequisite for the detection of preclinical AD. Objective(s): To develop a methodological protocol to quantify neuropsychological practice/test sophistication effects of elderly individuals and to conduct a preliminary validation with longitudinal data of early AD patients. Methods: Three-hundred seventy-four cognitively normal elderly individuals (NC) (128 women, 246 men; age [baseline] ⫽ 68.3 ⫾ 7.5 years; education ⫽ 12.8 ⫾ 3.1 years) were administered the German version of the Consortium to Establish a Registry for Alzheimer’s DiseaseNeuropsychological Assessment Battery (CERAD-NAB) at baseline (T0) and 2.4 ⫾ 0.28 years later (T1). Practice effects were calculated as differences in demographically (age, gender, education) adjusted standard scores (T1-T0), and these values were again adjusted for demographic variables and baseline performance, yielding accurate normative change scores. These results were validated against those of 95 AD patients (57 women, 38 men; age [baseline] ⫽ 74.2 ⫾ 6.4 years; education ⫽ 11.8 ⫾ 2.9 years; MMSE [baseline] ⫽ 24.1 ⫾ 3.4) who had been tested at baseline (T0) and 1.1 ⫾ 0.24 years later (T1). The apolipoprotein E (ApoE) genotype was determined in all NC subjects in order to study differential patterns of cognitive change. Conclusions: Assessment of neuropsychological change will serve as a much better basis in our effort to diagnose dementia in its preclinical stages. Results: Within the NC subjects, individuals with one or
CAUSES AND OUTCOMES OF COGNITIVE IMPAIRMENT NOT DEMENTIA (CIND) IN A NATIONAL SAMPLE OF U.S. ADULTS
Kenneth M. Langa1, Brenda L. Plassman2, Gwenith G. Fisher3, Guy G. Potter4, David R. Weir3, Robert B. Wallace5, 1University of Michigan, Ann Arbor, MI, USA; 2Duke University, Durham, NC, USA; 3 University of Michigan, Ann Arbor, MI, USA; 4Duke University, Durham, NC, USA; 5University of Iowa, Iowa City, Iowa, USA. Contact e-mail: [email protected] Background: Population-based studies of Cognitive Impairment that is Not Dementia (CIND) are rare. The causes (i.e., sub-types) and outcomes of CIND in population-based samples may be different than those in clinic-based samples. Objectives: 1) To identify sub-types of CIND in a nationally representative population-based sample of US adults; and 2) to determine the 18-month outcomes for progression to dementia and death among these CIND sub-types. Methods: The Aging, Demographics, and Memory Study (ADAMS), a national population-based study of 856 US adults age 70 or older was used. ADAMS CIND sub-types were defined based on clinical and neuropsychological criteria from prior studies. Extensive in-home clinical and neuropsychological assessments were performed, and an expert consensus panel used the assessments to assign a diagnosis of normal, CIND, or dementia. Sub-types of CIND were assigned to denote the cause of cognitive impairment. An 18-month follow-up in-home assessment was performed for all those diagnosed with CIND at the baseline assessment. Results: CIND was diagnosed in 23% of study participants. Prodromal Alzheimer’s disease (AD) was the most common CIND sub-type (39% of cases). Other medical conditions (e.g., heart disease, lung disease, diabetes) accounted for 23% of CIND cases, while Stroke (14%) and Vascular Cognitive Impairment (8%) were also common. Those with CIND due to Prodromal AD were significantly more likely to progress to dementia at 18 months compared to the other CIND groups (30% vs. 17%; P⬍.05). Those with CIND due to other medical conditions were most likely to die compared to the other CIND groups (28% vs. 12%; P⬍.05), but were also most likely to show significant cognitive improvement (20% vs. 12%, P⬍.05). Conclusions: Nearly one-quarter of US adults age 70 or older have CIND. CIND due to Pro-dromal AD is associated with a very high likelihood of progression to dementia (30% in 18 months). Chronic medical conditions (e.g., heart disease, lung disease, and diabetes) are important causes for CIND in the community, and are associated with both high rates of death AND cognitive improvement. This important and heterogenous group of older adults with CIND has likely been under-represented in prior studies of clinic-based samples. P2-185
METABOLIC AND COGNITIVE CHANGES WITH COMPUTER BASED COGNITIVE THERAPY FOR MCI
William Jagust1, Beth Mormino1, Charles DeCarli2, Joel Kramer3, Deborah Barnes3, Bruce Reed2, 1University of California, Berkeley, CA, USA; 2University of California, Davis, CA, USA; 3University of California, San Francisco, CA, USA. Contact e-mail: [email protected] Background: Observational studies have shown that cognitive activity may reduce the risk of development of dementia, and imaging studies suggest that changes in brain function result from cognitive training.
S289
Patients with mild cognitive impairment (MCI) might benefit from cognitive stimulation in order to reduce the risk of progression to dementia. Objective: We performed this study in order to investigate cognitive and brain metabolic changes (measured with fluorodeoxyglucose PET) resulting from a rigorous cognitive training program. Methods: Six patients with MCI (mean age 71.5, 3 men) were randomized to receive a 100 min/day computerized intervention for 4-7 weeks that was designed to improve brain processing of auditory stimuli. A control group of 9 MCI patients (mean age 71.5, 6 men) read on-line newspapers, listened to audio books, and played a computer video game for comparable times. FDG-PET was performed at baseline and following treatment, as was the Repeatable Battery for Assessment of Neuropsychological Status (RBANS). PET data were analyzed using spatially normalized scans and SPM2. Contrasts included both pre- and post-treatment comparisons for each group separately, and regressions between change in performance on the RBANS and change in glucose metabolism. Results: Groups were comparable on baseline MMSE scores (27.3 treatment and 26.9 control). Comparison of preand post-treatment PET showed reductions in glucose metabolism in right middle frontal lobes and right thalamus in the control group and no change in the treated group. In the treated group, improvements in the RBANS attention scale were correlated with increases in glucose metabolism in the right precuneus and right parietal lobe. Improvements in the RBANS visuospatial scale were correlated with increased metabolism in bilateral anterior and middle cingulate gyri and right insula. Conclusions: These results suggest the possibility that cognitive training slows the rate of metabolic decline in MCI. Individual patterns of behavioral change were related to underlying neural substrates, suggesting the possibility that this treatment approach may produce regionally specific effects on brain function. P2-186
TREATMENT OF MILD COGNITIVE IMPAIRMENT WITH VITAMIN E AND DONEPEZIL: CORRELATIONS BETWEEN SERIAL MRI AND CONVERSION, GENOTYPE, AND TREATMENT STATUS
Leon J. Thal1, Clifford R. Jack, Jr.2, Michael Grundman3, Chadwick Ward2, Drahomira Sencakova2, Shelia Jin4, Anthony Gamst4, Ronald C. Petersen2, The Alzheimer’s Disease Cooperative Study Group4, 1University of California, San Diego, San Diego, CA, USA; 2 Mayo Clinic, Rochester, MN, USA; 3Elan Pharmaceuticals, San Diego, USA; 4University of California, San Diego, San Diego, USA. Contact e-mail: [email protected] Background: The vitamin E and donepezil study for the treatment of MCI was conducted in 769 subjects at 69 North American centers. Clinical results of this study have been reported (NEJM 2005; 352(23):2379-88). Twenty-four of the recruiting sites voluntarily participated in an MRI substudy. Objective(s): We report correlations between rates of atrophy from serial MRI and clinical conversion to AD, APO E genotype, and treatment arm in those who participated in MRI. Methods: A total of 194 subjects participated in the MRI substudy and 137 completed both a baseline and followup scan allowing for computation of rates of change over time. The followup MRI scan for each subject could have been either at the end of the study (36 months) or at an interim time point associated with clinical conversion from MCI to AD. In each subject, change from baseline to followup was measured for four different brain volumes; hippocampus, entorhinal cortex (ERC), whole brain, and ventricle. Results: For each of the four brain atrophy rate measures, annualized rates of atrophy were greater in those subjects who converted to AD vs nonconverters (all p ⬍0.000). Annualized rates of hippocampal (p⬍0.05), ERC (p⬍0.05), ventricular (p⫽0.002), and whole brain (p⫽0.004) atrophy were greater in APOE 4 carriers than non-carriers. A non significant trend consistent with treatment effect was observed across all four mean MRI rate measures for both donepezil and vitamin E. The number of treated subjects for which MRI measures were available varied for different MRI measures. Conclusions: Expected and highly significant correlations were
S290
Poster P2:: Monday Posters
observed between rates of brain volume change measured with MRI and both clinical conversion status as well as APOE genotype. These results validate the technical feasibility of using MRI as an outcome measure of disease progression in multi-center trials for amnestic MCI. Failure to detect a significant treatment effect in the MRI data could reflect the absence of a biologic effect of either vitamin E or donepezil on brain volume itself. Or it could simply represent inadequate sample size in the treatment groups: donepezil (n⫽ 36 - 40, number varies with different MRI measures), vitamin E (n ⫽ 34 -43). P2-187
COMPARISON BETWEEN THE STANDARD MINI MENTAL STATE EXAMINATION AND THE MODIFIED VERSION (3MS) IN HEALTHY SUBJECTS AND PATIENTS WITH MILD COGNITIVE IMPAIRMENT OR DEMENTIA. ME´RIDA-VENEZUELA, 2005-2006
Clara I. Ramirez1, Trino J. Baptista2, Ana Serrano3, Yinet Arape3, Evelyn Thonon4, 1Department of Neurology, Los Andes University Hospital, Merida, Venezuela; 2Department of Physiology, Los Andes University Medical School, Merida, Venezuela; 3Department of Psychiatry, Los Andes University Hospital, Merida, Venezuela; 4 Department of Neurology, El Vigia Hospital, Merida, Venezuela. Contact e-mail: [email protected] Introduction: Current clinical practice and research request for precise and accessible instruments to detect early cognitive impairment (CI). Folstein’s Mini Mental State Examination (MMSE)1 is widely used in screening studies. Unfortunately, its specificity is hampered by being affected by age, education level and cultural environment. It is also very sensitive to moderate to severe-, but not to mild cognitive impairment (MCI). Teng y Chang Chui developed a more extensive version (3MS)2 which may overcome such limitations. Objectives: Since the 3MS has not been validated in Venezuela, we correlated both instruments’ scores and evaluated their sensitivity and specificity compared to the clinical diagnosis. Methods: The scales were administered in a random sequence to 93 outpatients aged 50 or more years who were assessed in the neurology or psychiatry departments. Cut-off scores for CI were 23 for the MMSE and ⬍ 79 for the 3MS. Results: We evaluated 36 healthy subjects (89% women; age (years: average ⫾ SD): 65.4 ⫾ 7.7); 25 with MCI (72% women; age: 69.4 ⫾ 10.1) and 32 demented patients (78% women; age: 73.7 ⫾ 7.3). A significant positive correlation was found between the MMSE and the 3MS: whole sample [r (93) ⫽ 0.876, p ⫽ 0.0001]; healthy subjects [r (36) ⫽ 0.466, p ⫽ 0.004]; MCI [r (25) ⫽ 0.644, p ⫽ 0.001] and dementia [r (32) ⫽ 0.854, p ⫽ 0.0001]. The MMSE specificity reached 100%, but its sensitivity was of 12% for MCI and 44 % for dementia. The 3MS specificity was of 83.3%, and the sensitivity reached 44% for MCI and 69% for dementia. Conclusions: In spite of an excellent numeric correlation between both instruments, the MMSE displayed a very low sensitivity for the MCI in our population. Since sensitivity was better for the 3MS, additional validation studies for this instrument are warranted in Venezuela. References: (1) Folstein, M., Folstein, S., McHugh P. (1975) Mini-Mental State Examination: a practical method for grading the cognitive sate of patients for the clinician. J Psychiatr Res, 12: 189-198. (1) Teng, E., Chang, H. (1987) The Modified Mini-Mental State (3MS) Examination. J Clin Psychiatry, 48: 314-318. P2-188
PREDICTION OF COGNITIVE IMPAIRMENT AMONG ELDERLY PERSONS USING A BRIEF COGNITIVE SCREENING BATTERY
Anthony LoGalbo1, Laura Baker2, Suzanne Craft1,2, The ADAPT Research Group2, 1VA Puget Sound Health Care System, Seattle, WA, USA; 2University of Washington, Seattle, WA, USA. Contact e-mail: [email protected] Background: Early detection of cognitive impairment among the elderly is becoming increasingly important. There is a need for brief measures that
can reliably predict the presence of clinically significant cognitive impairment. Objective: To decipher which brief cognitive tests most reliably predict cognitive impairment, particularly amnestic Mild Cognitive Impairment (aMCI), among elderly persons. Methods: Participants aged 70⫹ with normal baseline cognitive functioning were subsequently administered yearly 30-35 minute cognitive screening batteries, including the Modified Mini Mental State Examination (3MSE), generative verbal fluency (supermarket items), Narratives from the Rivermead Behavioral Memory Test (RBMT), Brief Visuospatial Memory Test, Revised (BVMT-R), Hopkins Verbal Learning Test, Revised (HVLT-R), digit span (DS), and the Geriatric Depression Scale (GDS). When participants’ scores on these measures fell below pre-specified cut points, a more expansive dementia evaluation was triggered to assist in diagnosing the presence and extent of cognitive impairment. Results: Scores on the cognitive screening measures were entered into a stepwise logistic regression analysis to predict outcome of the dementia evaluation. The 3MSE, RBMT delay, HVLT-R delay, DS backward, and GDS were identified as significant predictors of normal versus impaired cognitive functioning. However, when participants with normal cognitive functioning were compared only to those with aMCI, the 3MSE and RBMT delay were the only significant predictors. Furthermore, only the 3MSE was identified as a significant predictor of aMCI versus Alzheimer’s disease (AD). Conclusions: Brief cognitive screening measures are capable of differentiating between normal and impaired cognitive functioning upon more in-depth cognitive evaluation. The fact that fewer and less diverse measures reliably predicted cognitive impairment when the outcome was narrowed to normal versus aMCI supports the notion that diverse cognitive impairment can result from a variety of factors among the elderly. This finding also highlights the benefit of using certain screening measures to assist in differentiating aMCI from other forms of cognitive impairment. Results of this study further suggest that some measures are not helpful in predicting dementia evaluation outcome. Thus, their utility in triggering a dementia evaluation should be further examined. Overall results of this study have implications for the effective early evaluation and diagnosis of aMCI. P2-189
A METHODOLOGICAL APPROACH FOR ASSESSING NEUROPSYCHOLOGICAL CHANGES IN THE PRECLINICAL STAGE OF ALZHEIMER’S DISEASE (AD)
Andreas U. Monsch1, Antoinette Zehnder1, Stefan Blaesi1, Manfred Berres2, Rene´ Spiegel1, Hannes B. Staehelin1, 1Memory Clinic, Basel, Switzerland; 2RheinAhrCampus, Remagen, Germany. Contact e-mail: [email protected] Background: The characterization of cognitive changes in healthy elderly individuals is a critical prerequisite for the detection of preclinical AD. Objective(s): To develop a methodological protocol to quantify neuropsychological practice/test sophistication effects of elderly individuals and to conduct a preliminary validation with longitudinal data of early AD patients. Methods: Three-hundred seventy-four cognitively normal elderly individuals (NC) (128 women, 246 men; age [baseline] ⫽ 68.3 ⫾ 7.5 years; education ⫽ 12.8 ⫾ 3.1 years) were administered the German version of the Consortium to Establish a Registry for Alzheimer’s DiseaseNeuropsychological Assessment Battery (CERAD-NAB) at baseline (T0) and 2.4 ⫾ 0.28 years later (T1). Practice effects were calculated as differences in demographically (age, gender, education) adjusted standard scores (T1-T0), and these values were again adjusted for demographic variables and baseline performance, yielding accurate normative change scores. These results were validated against those of 95 AD patients (57 women, 38 men; age [baseline] ⫽ 74.2 ⫾ 6.4 years; education ⫽ 11.8 ⫾ 2.9 years; MMSE [baseline] ⫽ 24.1 ⫾ 3.4) who had been tested at baseline (T0) and 1.1 ⫾ 0.24 years later (T1). The apolipoprotein E (ApoE) genotype was determined in all NC subjects in order to study differential patterns of cognitive change. Conclusions: Assessment of neuropsychological change will serve as a much better basis in our effort to diagnose dementia in its preclinical stages. Results: Within the NC subjects, individuals with one or