- Email: [email protected]
Depression predicts progressive brain atrophy in mild cognitive impairment
Poster Presentations P1
P1-323
COMBINATION OF CEREBRAL METABOLIC AND COGNITIVE INFORMATION AS A BETTER PREDICTOR OF THE CONVERSION TO ALZHEIMER’S DISEASE IN MILD COGNITIVE IMPAIRMENT: A TWO-YEAR FOLLOW-UP STUDY
Dong Young Lee1, Il Hann Choo1, Eun Hyun Seo1, Bo Kyung Sohn1, Jae Hwa Park1, Jee Wook Kim2, Jong Inn Woo1, 1Seoul National University Hospital, Seoul, South Korea; 2St. Andrew’s Neuropsychiatric Hospital, Icheon, South Korea. Background: This study aimed to identify characteristic regional patterns of cerebral glucose metabolism at baseline in patients converting from mild cognitive impairment (MCI) to Alzheimer’s disease (AD) after two years of follow-up, and to compare the ability of AD prediction between cerebral metabolic information alone and combination of metabolic and cognitive information. Methods: Forty seven elderly subjects with MCI and 30 cognitively normal (CN) elderly individuals were evaluated at baseline with a [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) scan and CERAD neuropsychological tests and followed up annually. Voxel-based statistical comparison of FDG-PET data with CN group and between MCI subgroups was performed using statistical parametric mapping 2 (SPM2). Logistic regression analyses were conducted to examine the ability of regional cerebral glucose metabolism (rCMglc) alone and combination of rCMglc and cognitive test scores to predict those who would progress to AD. Results: Among the overall MCI subjects, 14 (29.8%) were converted to clinically evident AD (referred to as MCIAD) and 33 (70.2%) were still in the MCI state (referred to as MCIMCI) after the 2-year follow-up. Compared with CN, MCIAD showed reduced rCMglc in the precuneus, posterior cingulate, temporoparietal, and frontal cortex at baseline. Relative baseline hypometabolism in the middle frontal, precuneus, and inferior parietal region was also found in MCIAD in comparison with MCIMCI. In terms of AD prediction ability, combination of frontal rCMglc and mini-mental state examination score (accuracy: 92.8%) was significantly better than frontal rCMglc alone (accuracy: 78.6%). However, the addition of other CERAD neuropsychological test scores did not improve AD prediction ability further. Conclusions: Our findings support that MCI patients who progress to AD within two years already have typical hypometabolism pattern of AD, even including frontal hypometabolism, at baseline. The addition of a global cognitive test can improve the AD prediction ability of FDG-PET in MCI.
then compared across the three groups. Results: Regional analyses were performed and direct comparisons between groups (corrected for multiple comparisons using permutation tests) revealed significantly higher rates of temporal (p ¼ .03), parietal (p ¼ .03), and left frontal (p ¼ .02) white matter atrophy in the DEP group compared to the NOSYMP group. The DEP group also demonstrated greater atrophy compared to the OTHER group in frontal and parietal white matter regions, though permutation tests did not reach significance. Comparisons between the OTHER and NOSYMP groups revealed no significant differences in rates of atrophy over 2 years.
Conclusions: The presence of depressive symptoms in MCI subjects was associated with increased white matter atrophy in regions known to be affected as AD progresses. These findings suggest that depression in individuals with MCI may reflect underlying pathological changes that represent prodromal AD. Thus, assessment of depressive symptoms may be a potentially useful clinical marker in identifying MCI patients who are most likely to progress to AD. P1-325
P1-324
DEPRESSION PREDICTS PROGRESSIVE BRAIN ATROPHY IN MILD COGNITIVE IMPAIRMENT 1
1
1
1
1
Grace Lee , Po-Haong Lu , Xue Hua , Suh Lee , Alex Leow , Stephanie Wu1, Ken Nguyen1, Edmond Teng2, George Bartzokis1, Arthur Toga1, Clifford Jack3, Michael Weiner4, Paul Thompson1, 1UCLA, Los Angeles, California, United States; 2Greater Los Angeles VA/UCLA, Los Angeles, California, United States; 3Mayo Clinic, Rochester, Minnesota, United States; 4VA/UCSF, San Francisco, California, United States. Background: Depression has been shown to predict higher rates of progression from mild cognitive impairment (MCI) to Alzheimer’s disease (AD). The current study examined the underlying neuroanatomical changes associated with depression in MCI patients. Tensor-based morphometry (TBM) was used to compare the longitudinal progression of brain atrophy in MCI patients with and without depression. Methods: 245 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) who were diagnosed with MCI and had MRI scans at baseline and 2-year follow-up were included in the present study. The subjects, ranging in age from 55 to 90 (mean age ¼ 74.9 years, sd ¼ 7.0), were divided into three groups based on their scores from the Neuropsychiatric Inventory Questionnaire (NPI-Q): individuals with reported depression (DEP, n ¼ 47), with any other neuropsychiatric symptom except depression (OTHER, n ¼ 92), and with no neuropsychiatric symptoms (NOSYMP, n ¼ 106). TBM was used to create 3D Jacobian maps of local brain atrophy rates for individual participants, which were
S215
AGE-ASSOCIATED MEMORY IMPAIRMENT IS ASSOCIATED WITH GREATER RATE OF LONGITUDINAL BRAIN ATROPHY
Po-Haong Lu1, Grace Lee1, Xue Hua1, Alex Leow1, Stephanie Melchor1, Arthur Toga1, Clifford Jack2, Michael Weiner3, Paul Thompson1, 1UCLA, Los Angeles, California, United States; 2Mayo Clinic, Rochester, Minnesota, United States; 3VA/UCSF, San Francisco, California, United States. Background: Age-Associated Memory Impairment (AAMI) was initially conceptualized to describe a feature of cognitive aging. It is defined as memory performance that is consistent with age-matched peers but significantly lower than younger adults. However, there is mounting evidence that AAMI may not be characteristic of healthy aging but is instead a harbinger of future cognitive decline and eventual development of dementia. We examined and compared the longitudinal progression of brain volume changes between individuals with AAMI and normal cognition. Methods: 151 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) who were initially diagnosed with normal cognition (NC) and had MRI scans at baseline and 2-year follow-up were included in the present study. Baseline performance on the delayed recall trial of the Rey Auditory Verbal Learning Test (RAVLT) was examined, and individuals who scored more than 1 standard deviation (SD) below the mean established for young adults (ages 20-29) were reclassified as having AAMI. Ninety subjects met criteria for AAMI while 61 subjects remained in the NC group. Tensor-based morphometry (TBM) was used to create 3D Jacobian maps of local brain atrophy rates for individual participants; these spatially detailed 3D maps were compared
P1-323
COMBINATION OF CEREBRAL METABOLIC AND COGNITIVE INFORMATION AS A BETTER PREDICTOR OF THE CONVERSION TO ALZHEIMER’S DISEASE IN MILD COGNITIVE IMPAIRMENT: A TWO-YEAR FOLLOW-UP STUDY
Dong Young Lee1, Il Hann Choo1, Eun Hyun Seo1, Bo Kyung Sohn1, Jae Hwa Park1, Jee Wook Kim2, Jong Inn Woo1, 1Seoul National University Hospital, Seoul, South Korea; 2St. Andrew’s Neuropsychiatric Hospital, Icheon, South Korea. Background: This study aimed to identify characteristic regional patterns of cerebral glucose metabolism at baseline in patients converting from mild cognitive impairment (MCI) to Alzheimer’s disease (AD) after two years of follow-up, and to compare the ability of AD prediction between cerebral metabolic information alone and combination of metabolic and cognitive information. Methods: Forty seven elderly subjects with MCI and 30 cognitively normal (CN) elderly individuals were evaluated at baseline with a [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) scan and CERAD neuropsychological tests and followed up annually. Voxel-based statistical comparison of FDG-PET data with CN group and between MCI subgroups was performed using statistical parametric mapping 2 (SPM2). Logistic regression analyses were conducted to examine the ability of regional cerebral glucose metabolism (rCMglc) alone and combination of rCMglc and cognitive test scores to predict those who would progress to AD. Results: Among the overall MCI subjects, 14 (29.8%) were converted to clinically evident AD (referred to as MCIAD) and 33 (70.2%) were still in the MCI state (referred to as MCIMCI) after the 2-year follow-up. Compared with CN, MCIAD showed reduced rCMglc in the precuneus, posterior cingulate, temporoparietal, and frontal cortex at baseline. Relative baseline hypometabolism in the middle frontal, precuneus, and inferior parietal region was also found in MCIAD in comparison with MCIMCI. In terms of AD prediction ability, combination of frontal rCMglc and mini-mental state examination score (accuracy: 92.8%) was significantly better than frontal rCMglc alone (accuracy: 78.6%). However, the addition of other CERAD neuropsychological test scores did not improve AD prediction ability further. Conclusions: Our findings support that MCI patients who progress to AD within two years already have typical hypometabolism pattern of AD, even including frontal hypometabolism, at baseline. The addition of a global cognitive test can improve the AD prediction ability of FDG-PET in MCI.
then compared across the three groups. Results: Regional analyses were performed and direct comparisons between groups (corrected for multiple comparisons using permutation tests) revealed significantly higher rates of temporal (p ¼ .03), parietal (p ¼ .03), and left frontal (p ¼ .02) white matter atrophy in the DEP group compared to the NOSYMP group. The DEP group also demonstrated greater atrophy compared to the OTHER group in frontal and parietal white matter regions, though permutation tests did not reach significance. Comparisons between the OTHER and NOSYMP groups revealed no significant differences in rates of atrophy over 2 years.
Conclusions: The presence of depressive symptoms in MCI subjects was associated with increased white matter atrophy in regions known to be affected as AD progresses. These findings suggest that depression in individuals with MCI may reflect underlying pathological changes that represent prodromal AD. Thus, assessment of depressive symptoms may be a potentially useful clinical marker in identifying MCI patients who are most likely to progress to AD. P1-325
P1-324
DEPRESSION PREDICTS PROGRESSIVE BRAIN ATROPHY IN MILD COGNITIVE IMPAIRMENT 1
1
1
1
1
Grace Lee , Po-Haong Lu , Xue Hua , Suh Lee , Alex Leow , Stephanie Wu1, Ken Nguyen1, Edmond Teng2, George Bartzokis1, Arthur Toga1, Clifford Jack3, Michael Weiner4, Paul Thompson1, 1UCLA, Los Angeles, California, United States; 2Greater Los Angeles VA/UCLA, Los Angeles, California, United States; 3Mayo Clinic, Rochester, Minnesota, United States; 4VA/UCSF, San Francisco, California, United States. Background: Depression has been shown to predict higher rates of progression from mild cognitive impairment (MCI) to Alzheimer’s disease (AD). The current study examined the underlying neuroanatomical changes associated with depression in MCI patients. Tensor-based morphometry (TBM) was used to compare the longitudinal progression of brain atrophy in MCI patients with and without depression. Methods: 245 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) who were diagnosed with MCI and had MRI scans at baseline and 2-year follow-up were included in the present study. The subjects, ranging in age from 55 to 90 (mean age ¼ 74.9 years, sd ¼ 7.0), were divided into three groups based on their scores from the Neuropsychiatric Inventory Questionnaire (NPI-Q): individuals with reported depression (DEP, n ¼ 47), with any other neuropsychiatric symptom except depression (OTHER, n ¼ 92), and with no neuropsychiatric symptoms (NOSYMP, n ¼ 106). TBM was used to create 3D Jacobian maps of local brain atrophy rates for individual participants, which were
S215
AGE-ASSOCIATED MEMORY IMPAIRMENT IS ASSOCIATED WITH GREATER RATE OF LONGITUDINAL BRAIN ATROPHY
Po-Haong Lu1, Grace Lee1, Xue Hua1, Alex Leow1, Stephanie Melchor1, Arthur Toga1, Clifford Jack2, Michael Weiner3, Paul Thompson1, 1UCLA, Los Angeles, California, United States; 2Mayo Clinic, Rochester, Minnesota, United States; 3VA/UCSF, San Francisco, California, United States. Background: Age-Associated Memory Impairment (AAMI) was initially conceptualized to describe a feature of cognitive aging. It is defined as memory performance that is consistent with age-matched peers but significantly lower than younger adults. However, there is mounting evidence that AAMI may not be characteristic of healthy aging but is instead a harbinger of future cognitive decline and eventual development of dementia. We examined and compared the longitudinal progression of brain volume changes between individuals with AAMI and normal cognition. Methods: 151 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) who were initially diagnosed with normal cognition (NC) and had MRI scans at baseline and 2-year follow-up were included in the present study. Baseline performance on the delayed recall trial of the Rey Auditory Verbal Learning Test (RAVLT) was examined, and individuals who scored more than 1 standard deviation (SD) below the mean established for young adults (ages 20-29) were reclassified as having AAMI. Ninety subjects met criteria for AAMI while 61 subjects remained in the NC group. Tensor-based morphometry (TBM) was used to create 3D Jacobian maps of local brain atrophy rates for individual participants; these spatially detailed 3D maps were compared