- Email: [email protected]
REGIONAL BRAIN METABOLISM AND CORTICAL THICKNESS IN F18-FLUTEMETAMOL AMYLOID-POSITIVE VERSUS -NEGATIVE MILD COGNITIVE IMPAIRMENT PATIENTS
Oral Sessions: O2-03: Neuroimaging: Imaging in Mild Cognitive Impairment and Subjective Memory Complaint
control groups during the first year, and the proportion of energy from fat increased (paired t-test for groups separately, p<0.01). Intakes of vitamin E, vitamin D, and dietary fiber increased in the intervention group (p<0.05), and decreased in the control group (table 1). Furthermore, quality of fat improved in the intervention group: intake of saturated fatty acids decreased while the intake of unsaturated fatty acids increased. Participants were categorized into 3 groups based on their level of cognitive function, and dietary changes were compared across the tertiles. In the control group, those with highest cognitive scores had less decrease in vitamin D (p for trend across tertiles 0.032) and in vitamin E (p¼ 0.012) during the 1 st year. In the intervention group the positive dietary changes were observed irrespective of cognitive function. The FINGER dietary intervention shows that beneficial dietary changes can be achieved in a group of older, high-risk individuals. The lack of association between baseline cognition and dietary changes in the intervention group suggests that participants with different levels of baseline cognitive function benefit from a dietary intervention. The future results of the FINGER study will show, if these positive changes in the diet will also prevent cognitive decline and dementia. Table 1 Nutrient intakes at baseline and the change during the first year of intervention in the FINGER study. Data are presented as mean (sd) and stars indicate statistically significant difference between the groups (***p<0.001)
Energy intake (MJ) Fat (E%) Saturated fat (E%) Unsaturated fat (mono- and polycombined, E%) Protein (E%) Carbohydrate (E%) Fiber (g/MJ) Vitamin E (mg/MJ) Vitamin D (mg/MJ) Folate (mg/MJ)
Intake at baseline
Change during the 1st year of intervention
Intervention group
Control group
Intervention group
7.6 (2.3) 33.7 (6.5) 12.4 (3.5) 16.6 (3.9)
7.6 (2.2) 33.6 (6.6) 12.3 (3.6) 16.7 (4.1)
-0.2 (1.8) 1.0 (6.8) -0.6 (3.4) 1.4 (4.7)
-0.2 (1.8) 1.0 (6.9) 0.6 (3.4) *** 0.4 (4.5) ***
17.4 (3.2) 46.0 (7.3) 2.87 (0.90) 1.45 (0.92) 1.85 (1.52) 36.0 (18.4)
17.3 (3.4) 46.2 (7.2) 2.91 (0.91) 1.60 (2.70) 1.88 (1.62) 36.4 (19.4)
0.1 (3.7) -0.5 (6.8) 0.14 (0.93) 0.10 (0.96) 0.34 (1.89) 1.6 (20.8)
-0.1 (3.6) -0.7 (6.8) -0.13 (0.84) *** -0.14 (1.42) *** -0.09 (1.68) *** -0.5 (25.9)
Control group
P167
Our goal was to evaluate amyloid deposition, glucose metabolism, and medial temporal lobe (MTL) atrophy in SMC participants from ADNI. Methods: 569 participants were selected from the ADNI cohort, including 177 healthy controls (HC), 93 participants with SMC, and 299 patients with early mild cognitive impairment (EMCI). The HC participants were further divided into those with genetic risk (APOE ε4 positive and/or family history of AD (HC-risk)) and those without genetic risk (HC). The SMC participants were also further divided by the presence or absence of informant complaints about the participant’s memory. SMC participants were considered to have self-only concerns (SMC) if the informant did not endorse >1.5 SD above the HC mean on the ECog Memory domain (w62% of items) and to have self-plus-informant concerns if the informant endorsed > 62% (SMC-plus). Florbetapir and structural MRI scans were downloaded from the ADNI site for the baseline visit and processed as previously described [1]. FDG PET scans were also downloaded from this timepoint and processed using standard techniques. Average Florbetapir SUVR, FDG SUVR, and structural grey matter density were extracted from the global cortex (PET) and bilateral hippocampus (MRI). Florbetapir PET scans were also compared between groups on a voxel-wise level in SPM8. Results: A significant difference in amyloid deposition was observed between groups on both voxel-wise and ROI analyses. HC-risk, SMC, and SMC-plus demonstrated more global and regional amyloid deposition than HC without risk, including in the global cortex (p¼0.004) and precuneus (p<0.001). Differences in glucose metabolism and MTL atrophy were more variable with some regions showing a trend towards hypometabolism and MTL atrophy in HC-risk and SMC groups. Conclusions: Participants with SMC and HC participants at risk for progression to AD due to genetic background show increased amyloid deposition relative to HC without risk. This suggests that older adults with SMC, especially those with informant corroboration, are at increased risk for future cognitive decline and therefore may be a good target population for enrichment of clinical trials.[1] Risacher et al. (2013) Frontiers in Aging Neuroscience.
MONDAY, JULY 14, 2014 ORAL SESSIONS O2-03 NEUROIMAGING: IMAGING IN MILD COGNITIVE IMPAIRMENT AND SUBJECTIVE MEMORY COMPLAINT O2-03-01
INCREASED AMYLOID DEPOSITION IN OLDER ADULTS AT RISK FOR PROGRESSION TO ALZHEIMER’S DISEASE DUE TO GENETIC BACKGROUND AND/OR THE PRESENCE OF SIGNIFICANT MEMORY CONCERNS
Shannon Leigh Risacher1, Sungeun Kim1, Kwangsik T. Nho1, John West1, Yang Wang1, Ronald Carl Petersen2, Paul S. Aisen3, Clifford R. Jack4, William J. Jagust5, Robert Koeppe6, Michael Walter Weiner7, Andrew J. Saykin8, 1Indiana University School of Medicine, Indianapolis, Indiana, United States; 2Mayo Clinic Rochester, Rochester, Minnesota, United States; 3UCSD, La Jolla, California, United States; 4Mayo Clinic, Rochester, Minnesota, United States; 5University of California, Berkeley, Berkeley, California, United States; 6University of Michigan, Ann Arbor, Michigan, United States; 7Center for Imaging of Neurodegenerative Diseases; VA Medical Center and UCSF, San Francisco, California, United States; 8Indiana University School of Medicine, Indianapolis, Indiana, United States. Contact e-mail: [email protected] Background: Older adults with significant memory concerns (SMC) and/or genetic risk for AD are key groups of interest due to risk of progression.
Figure 1. Increased Amyloid Deposition in Older Adults at Risk for AD O2-03-02
REGIONAL BRAIN METABOLISM AND CORTICAL THICKNESS IN F18-FLUTEMETAMOL AMYLOID-POSITIVE VERSUS -NEGATIVE MILD COGNITIVE IMPAIRMENT PATIENTS
Bernard Hanseeuw1, Laurence Dricot2, Cecile Grandin3, Renaud Lhommel1, Lisa Quenon4, Adrian Ivanoiu5, 1Saint-Luc University Hospital, Brussels, Belgium; 2Universite Catholique de Louvain, Bruxelles, Belgium; 3Saint-Luc University Hospital, Bruxelles, Belgium; 4Universite Catholique de Louvain, Brussels, Belgium; 5Universite Catholique de Louvain, Saint Luc Hospital, Brussels, Belgium. Contact e-mail: bernard. [email protected] Background: New criteria for Alzheimer’s disease (AD) define prodromal AD as patients suffering from mild cognitive impairment (MCI) and presenting both an amyloid and a neurodegenerative marker. However, recent studies proved that some MCI do not carry brain amyloid. These patients have been named ’neurodegenerative only’ MCI (NO-MCI). We aimed at comparing the regional pattern of cortical thinning and hypometabolism
P168
Oral Sessions: O2-03: Neuroimaging: Imaging in Mild Cognitive Impairment and Subjective Memory Complaint
in NO-MCI versus amyloid-positive MCI (AP-MCI). Methods: Sixty non-demented patients attending our Memory Clinic were included. All participants underwent neuropsychological examination, 3T-brain MRI, FDG-PET and F18-flutemetamol. Brain metabolism was assessed using P-mod. Cortical thickness and hippocampus volume were assessed using FreeSurfer. Patients were classified in three groups:- AP-MCI (n¼28) had amyloid deposits on the F18-flutemetamol scan (>pc90 of an independent group of 31 elderly controls).- NO-MCI (n¼19) had no amyloid but cognitive impairment or hippocampus atrophy or else hypometabolism (
O2-03-03
BIN1 AND CR1 VARIANTS AFFECT COGNITIVE PERFORMANCE, NEURODEGENERATION, AND BRAIN AMYLOIDOSIS IN ADNI SUBJECTS
Anna Emilia Blanken1, Daniel H. Silverman1, Nare Torosyan2, Manogna Manne1, Beata Durcanova1, Andrew J. Saykin1, Clifford R. Jack3, Liana G. Apostolova1, 1UCLA, Los Angeles, California, United States; 2 University of California, Los Angeles, Los Angeles, California, United States; 3Mayo Clinic, Rochester, Minnesota, United States. Contact e-mail: [email protected] Background: Genome-wide association studies (GWAS) have identified many risk genes for Alzheimer’s disease (AD). The precise mechanism through which many of these genes exert their effect on AD remains unknown. Methods: We downloaded the top 10 AD gene single nucleotide polymorphism, baseline clinical and MR volumetric data (hippocampal, ventricular, fusiform and entorhinal volumes) of 33 ADNI-2/GO cognitively normal and 126 mild cognitive impairment subjects. The corresponding AV45 images were downloaded and processed with the FDA-approved
AmyQ software. For each individual AmyQ derived the mean SUVR in 47 sVOIs with whole brain as a reference, and 1 SUVR combining multiple regions optimized for quantifying florbetapir retention with whole cerebellum as a reference. We employed ANOVA with Bonferroni correction for multiple comparisons to investigate the effect of minor allele dosage on cognitive and functional performance, neurodegeneration and brain amyloidosis. Results: CR1 rs6691117 and BIN1 rs749008 had significant effects on multiple measures. The minor allele of CR1 rs6691117 showed significant association with functional decline at 6, 12 and 24 months as measured with the Functional Activity Questionnaire (FAQ). The minor allele of CR1 rs6691117 also associated with lower entorhinal and fusiform volumes as well as with greater brain amyloidosis in the lateral temporal, parietal and occipital association cortices at baseline. The minor allele of BIN1 showed a negative association with ventricular size and amyloid load in the frontal association, sensorimotor, posterior cingulate, medial temporal cortices, the pons, the midbrain and the lentiform nuclei. Conclusions: Both CR1 rs6691117 and BIN1 rs749008 demonstrate significant associations with brain amyloidosis and neurodegeneration providing further support for the relevance of these two genes to AD pathophysiology. The minor allele of CR1 rs6691117 appears to have a disease-promoting effect while the minor allele of BIN1 rs749008 seems to convey a disease-protective effect. The associations differed significantly in terms of localization in the brain.
O2-03-04
MR DIFFUSION IMAGING AND PHYSICAL ACTIVITY IN PEOPLE AT RISK OF ALZHEIMER’S DISEASE
Bernd Merkel1, Nicola T. Lautenschlager2, Kay Lorraine Cox3, Charles Malpas4, Sila Genc4, Kathryn A. Ellis5, Elizabeth Cyarto6, David Ames7, Patricia Desmond8, 1Royal Melbourne Hospital, Parkville, Australia; 2University of Melbourne, Kew, Australia; 3University of Western Australia, Perth, Australia; 4The University of Melbourne, Parkville, Australia; 5St. Georges Hospital, Parkville, Australia; 6 National Ageing Research Institute, Parkville, Australia; 7National Ageing Research Institute Inc. (NARI), Parkville, Australia; 8University of Melbourne and Royal Melbourne Hospital, Melbourne, Australia. Contact e-mail: [email protected] Background: Magnetic Resonance Imaging (MRI) has been proven to be a useful biomarker for early diagnosis in Alzheimer’s disease. Both structural (T1, T2, FLAIR, diffusion) as well as functional imaging (fMRI, perfusion) are able to show early signs of the disease. We investigated the association of physical activity (PA) with white matter microstructure using diffusion tensor MRI (DTI) in people at risk of developing Alzheimer’s disease with subjective memory complaints (SMC) or mild cognitive impairment (MCI). Methods: MRI data was acquired in a cohort of 100 people (79 SMC, 21 MCI, mean age 73.26 5.7, 44 male, 56 female) on a 3T scanner (Tim TrioÒ, Siemens) at the Royal Melbourne Hospital. The protocol included a DTI sequence. Using the FSL software, we created images of fractional anisotropy (FA), as well as mean (MD), axial (AD), and radial (RD) diffusivity. This data was taken as an input for TBSS (tract based spatial statistics) and the four mean parameter values for each patient were calculated, based on the white matter skeleton (Fig.). Level of PA was measured at baseline using the CHAMPS (Community Healthy Activities Model Program for Seniors) questionnaire. This questionnaire, designed for older adults, collects information on various physical activities undertaken during a typical week in the previous 4 weeks. Statistical analysis was performed using SPSS 22 (SPSS, Chicago, IL). Results: Diffusion parameters did not vary between SMC and MCI groups (see attached Table and Fig). In the whole group controlling for age, partial correlation was significant between overall PA (freq/week) and FA (r ¼ .30, p ¼ .003) (Fig.), MD (r ¼ -.27, p ¼ .008), RD (r ¼ -.29, p ¼ .004), but not between PA and AD (r ¼ -.17, p ¼ .10). The correlation coefficients were not significantly different between the SMC and MCI group Conclusions: To our knowledge, this is the first study reporting an association between PA and DTI in a cohort of people with SMC and MCI. This suggests that increased levels of PA may promote better preservation of white matter integrity in this group. Future studies addressing the longitudinal effects of PA in this cohort are already underway.
control groups during the first year, and the proportion of energy from fat increased (paired t-test for groups separately, p<0.01). Intakes of vitamin E, vitamin D, and dietary fiber increased in the intervention group (p<0.05), and decreased in the control group (table 1). Furthermore, quality of fat improved in the intervention group: intake of saturated fatty acids decreased while the intake of unsaturated fatty acids increased. Participants were categorized into 3 groups based on their level of cognitive function, and dietary changes were compared across the tertiles. In the control group, those with highest cognitive scores had less decrease in vitamin D (p for trend across tertiles 0.032) and in vitamin E (p¼ 0.012) during the 1 st year. In the intervention group the positive dietary changes were observed irrespective of cognitive function. The FINGER dietary intervention shows that beneficial dietary changes can be achieved in a group of older, high-risk individuals. The lack of association between baseline cognition and dietary changes in the intervention group suggests that participants with different levels of baseline cognitive function benefit from a dietary intervention. The future results of the FINGER study will show, if these positive changes in the diet will also prevent cognitive decline and dementia. Table 1 Nutrient intakes at baseline and the change during the first year of intervention in the FINGER study. Data are presented as mean (sd) and stars indicate statistically significant difference between the groups (***p<0.001)
Energy intake (MJ) Fat (E%) Saturated fat (E%) Unsaturated fat (mono- and polycombined, E%) Protein (E%) Carbohydrate (E%) Fiber (g/MJ) Vitamin E (mg/MJ) Vitamin D (mg/MJ) Folate (mg/MJ)
Intake at baseline
Change during the 1st year of intervention
Intervention group
Control group
Intervention group
7.6 (2.3) 33.7 (6.5) 12.4 (3.5) 16.6 (3.9)
7.6 (2.2) 33.6 (6.6) 12.3 (3.6) 16.7 (4.1)
-0.2 (1.8) 1.0 (6.8) -0.6 (3.4) 1.4 (4.7)
-0.2 (1.8) 1.0 (6.9) 0.6 (3.4) *** 0.4 (4.5) ***
17.4 (3.2) 46.0 (7.3) 2.87 (0.90) 1.45 (0.92) 1.85 (1.52) 36.0 (18.4)
17.3 (3.4) 46.2 (7.2) 2.91 (0.91) 1.60 (2.70) 1.88 (1.62) 36.4 (19.4)
0.1 (3.7) -0.5 (6.8) 0.14 (0.93) 0.10 (0.96) 0.34 (1.89) 1.6 (20.8)
-0.1 (3.6) -0.7 (6.8) -0.13 (0.84) *** -0.14 (1.42) *** -0.09 (1.68) *** -0.5 (25.9)
Control group
P167
Our goal was to evaluate amyloid deposition, glucose metabolism, and medial temporal lobe (MTL) atrophy in SMC participants from ADNI. Methods: 569 participants were selected from the ADNI cohort, including 177 healthy controls (HC), 93 participants with SMC, and 299 patients with early mild cognitive impairment (EMCI). The HC participants were further divided into those with genetic risk (APOE ε4 positive and/or family history of AD (HC-risk)) and those without genetic risk (HC). The SMC participants were also further divided by the presence or absence of informant complaints about the participant’s memory. SMC participants were considered to have self-only concerns (SMC) if the informant did not endorse >1.5 SD above the HC mean on the ECog Memory domain (w62% of items) and to have self-plus-informant concerns if the informant endorsed > 62% (SMC-plus). Florbetapir and structural MRI scans were downloaded from the ADNI site for the baseline visit and processed as previously described [1]. FDG PET scans were also downloaded from this timepoint and processed using standard techniques. Average Florbetapir SUVR, FDG SUVR, and structural grey matter density were extracted from the global cortex (PET) and bilateral hippocampus (MRI). Florbetapir PET scans were also compared between groups on a voxel-wise level in SPM8. Results: A significant difference in amyloid deposition was observed between groups on both voxel-wise and ROI analyses. HC-risk, SMC, and SMC-plus demonstrated more global and regional amyloid deposition than HC without risk, including in the global cortex (p¼0.004) and precuneus (p<0.001). Differences in glucose metabolism and MTL atrophy were more variable with some regions showing a trend towards hypometabolism and MTL atrophy in HC-risk and SMC groups. Conclusions: Participants with SMC and HC participants at risk for progression to AD due to genetic background show increased amyloid deposition relative to HC without risk. This suggests that older adults with SMC, especially those with informant corroboration, are at increased risk for future cognitive decline and therefore may be a good target population for enrichment of clinical trials.[1] Risacher et al. (2013) Frontiers in Aging Neuroscience.
MONDAY, JULY 14, 2014 ORAL SESSIONS O2-03 NEUROIMAGING: IMAGING IN MILD COGNITIVE IMPAIRMENT AND SUBJECTIVE MEMORY COMPLAINT O2-03-01
INCREASED AMYLOID DEPOSITION IN OLDER ADULTS AT RISK FOR PROGRESSION TO ALZHEIMER’S DISEASE DUE TO GENETIC BACKGROUND AND/OR THE PRESENCE OF SIGNIFICANT MEMORY CONCERNS
Shannon Leigh Risacher1, Sungeun Kim1, Kwangsik T. Nho1, John West1, Yang Wang1, Ronald Carl Petersen2, Paul S. Aisen3, Clifford R. Jack4, William J. Jagust5, Robert Koeppe6, Michael Walter Weiner7, Andrew J. Saykin8, 1Indiana University School of Medicine, Indianapolis, Indiana, United States; 2Mayo Clinic Rochester, Rochester, Minnesota, United States; 3UCSD, La Jolla, California, United States; 4Mayo Clinic, Rochester, Minnesota, United States; 5University of California, Berkeley, Berkeley, California, United States; 6University of Michigan, Ann Arbor, Michigan, United States; 7Center for Imaging of Neurodegenerative Diseases; VA Medical Center and UCSF, San Francisco, California, United States; 8Indiana University School of Medicine, Indianapolis, Indiana, United States. Contact e-mail: [email protected] Background: Older adults with significant memory concerns (SMC) and/or genetic risk for AD are key groups of interest due to risk of progression.
Figure 1. Increased Amyloid Deposition in Older Adults at Risk for AD O2-03-02
REGIONAL BRAIN METABOLISM AND CORTICAL THICKNESS IN F18-FLUTEMETAMOL AMYLOID-POSITIVE VERSUS -NEGATIVE MILD COGNITIVE IMPAIRMENT PATIENTS
Bernard Hanseeuw1, Laurence Dricot2, Cecile Grandin3, Renaud Lhommel1, Lisa Quenon4, Adrian Ivanoiu5, 1Saint-Luc University Hospital, Brussels, Belgium; 2Universite Catholique de Louvain, Bruxelles, Belgium; 3Saint-Luc University Hospital, Bruxelles, Belgium; 4Universite Catholique de Louvain, Brussels, Belgium; 5Universite Catholique de Louvain, Saint Luc Hospital, Brussels, Belgium. Contact e-mail: bernard. [email protected] Background: New criteria for Alzheimer’s disease (AD) define prodromal AD as patients suffering from mild cognitive impairment (MCI) and presenting both an amyloid and a neurodegenerative marker. However, recent studies proved that some MCI do not carry brain amyloid. These patients have been named ’neurodegenerative only’ MCI (NO-MCI). We aimed at comparing the regional pattern of cortical thinning and hypometabolism
P168
Oral Sessions: O2-03: Neuroimaging: Imaging in Mild Cognitive Impairment and Subjective Memory Complaint
in NO-MCI versus amyloid-positive MCI (AP-MCI). Methods: Sixty non-demented patients attending our Memory Clinic were included. All participants underwent neuropsychological examination, 3T-brain MRI, FDG-PET and F18-flutemetamol. Brain metabolism was assessed using P-mod. Cortical thickness and hippocampus volume were assessed using FreeSurfer. Patients were classified in three groups:- AP-MCI (n¼28) had amyloid deposits on the F18-flutemetamol scan (>pc90 of an independent group of 31 elderly controls).- NO-MCI (n¼19) had no amyloid but cognitive impairment or hippocampus atrophy or else hypometabolism (
O2-03-03
BIN1 AND CR1 VARIANTS AFFECT COGNITIVE PERFORMANCE, NEURODEGENERATION, AND BRAIN AMYLOIDOSIS IN ADNI SUBJECTS
Anna Emilia Blanken1, Daniel H. Silverman1, Nare Torosyan2, Manogna Manne1, Beata Durcanova1, Andrew J. Saykin1, Clifford R. Jack3, Liana G. Apostolova1, 1UCLA, Los Angeles, California, United States; 2 University of California, Los Angeles, Los Angeles, California, United States; 3Mayo Clinic, Rochester, Minnesota, United States. Contact e-mail: [email protected] Background: Genome-wide association studies (GWAS) have identified many risk genes for Alzheimer’s disease (AD). The precise mechanism through which many of these genes exert their effect on AD remains unknown. Methods: We downloaded the top 10 AD gene single nucleotide polymorphism, baseline clinical and MR volumetric data (hippocampal, ventricular, fusiform and entorhinal volumes) of 33 ADNI-2/GO cognitively normal and 126 mild cognitive impairment subjects. The corresponding AV45 images were downloaded and processed with the FDA-approved
AmyQ software. For each individual AmyQ derived the mean SUVR in 47 sVOIs with whole brain as a reference, and 1 SUVR combining multiple regions optimized for quantifying florbetapir retention with whole cerebellum as a reference. We employed ANOVA with Bonferroni correction for multiple comparisons to investigate the effect of minor allele dosage on cognitive and functional performance, neurodegeneration and brain amyloidosis. Results: CR1 rs6691117 and BIN1 rs749008 had significant effects on multiple measures. The minor allele of CR1 rs6691117 showed significant association with functional decline at 6, 12 and 24 months as measured with the Functional Activity Questionnaire (FAQ). The minor allele of CR1 rs6691117 also associated with lower entorhinal and fusiform volumes as well as with greater brain amyloidosis in the lateral temporal, parietal and occipital association cortices at baseline. The minor allele of BIN1 showed a negative association with ventricular size and amyloid load in the frontal association, sensorimotor, posterior cingulate, medial temporal cortices, the pons, the midbrain and the lentiform nuclei. Conclusions: Both CR1 rs6691117 and BIN1 rs749008 demonstrate significant associations with brain amyloidosis and neurodegeneration providing further support for the relevance of these two genes to AD pathophysiology. The minor allele of CR1 rs6691117 appears to have a disease-promoting effect while the minor allele of BIN1 rs749008 seems to convey a disease-protective effect. The associations differed significantly in terms of localization in the brain.
O2-03-04
MR DIFFUSION IMAGING AND PHYSICAL ACTIVITY IN PEOPLE AT RISK OF ALZHEIMER’S DISEASE
Bernd Merkel1, Nicola T. Lautenschlager2, Kay Lorraine Cox3, Charles Malpas4, Sila Genc4, Kathryn A. Ellis5, Elizabeth Cyarto6, David Ames7, Patricia Desmond8, 1Royal Melbourne Hospital, Parkville, Australia; 2University of Melbourne, Kew, Australia; 3University of Western Australia, Perth, Australia; 4The University of Melbourne, Parkville, Australia; 5St. Georges Hospital, Parkville, Australia; 6 National Ageing Research Institute, Parkville, Australia; 7National Ageing Research Institute Inc. (NARI), Parkville, Australia; 8University of Melbourne and Royal Melbourne Hospital, Melbourne, Australia. Contact e-mail: [email protected] Background: Magnetic Resonance Imaging (MRI) has been proven to be a useful biomarker for early diagnosis in Alzheimer’s disease. Both structural (T1, T2, FLAIR, diffusion) as well as functional imaging (fMRI, perfusion) are able to show early signs of the disease. We investigated the association of physical activity (PA) with white matter microstructure using diffusion tensor MRI (DTI) in people at risk of developing Alzheimer’s disease with subjective memory complaints (SMC) or mild cognitive impairment (MCI). Methods: MRI data was acquired in a cohort of 100 people (79 SMC, 21 MCI, mean age 73.26 5.7, 44 male, 56 female) on a 3T scanner (Tim TrioÒ, Siemens) at the Royal Melbourne Hospital. The protocol included a DTI sequence. Using the FSL software, we created images of fractional anisotropy (FA), as well as mean (MD), axial (AD), and radial (RD) diffusivity. This data was taken as an input for TBSS (tract based spatial statistics) and the four mean parameter values for each patient were calculated, based on the white matter skeleton (Fig.). Level of PA was measured at baseline using the CHAMPS (Community Healthy Activities Model Program for Seniors) questionnaire. This questionnaire, designed for older adults, collects information on various physical activities undertaken during a typical week in the previous 4 weeks. Statistical analysis was performed using SPSS 22 (SPSS, Chicago, IL). Results: Diffusion parameters did not vary between SMC and MCI groups (see attached Table and Fig). In the whole group controlling for age, partial correlation was significant between overall PA (freq/week) and FA (r ¼ .30, p ¼ .003) (Fig.), MD (r ¼ -.27, p ¼ .008), RD (r ¼ -.29, p ¼ .004), but not between PA and AD (r ¼ -.17, p ¼ .10). The correlation coefficients were not significantly different between the SMC and MCI group Conclusions: To our knowledge, this is the first study reporting an association between PA and DTI in a cohort of people with SMC and MCI. This suggests that increased levels of PA may promote better preservation of white matter integrity in this group. Future studies addressing the longitudinal effects of PA in this cohort are already underway.