Aarhus University Seal / Aarhus Universitets segl

Association between memory and amyloid deposition–synaptic dysfunction in people with EMCI, LMCI and Alzheimer’s disease

Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

  • Marina Dauar, McGill University, Canada
  • Jared Rowley, McGill University, Canada
  • Sara Mohades, McGill University, Canada
  • Monica Shin, McGill University, Canada
  • Antoine Leuzy, McGill University, Canada
  • Liyong Wu, Capital Medical University, China
  • Laksanun Cheewakriengkrai, McGill University, Canada
  • Simon Fristed Eskildsen
  • Vladimir Fonov, McGill University, Canada
  • Serge Gauthier, McGill University, Canada
  • Pedro Rosa-Neto, McGill University, Canada
Background: Deficits in list of words or history recall performance is associated
with regional declines in brain metabolism, neurodegeneration and
amyloid deposition. However, little is known regarding the neurocorrelates
of these tests in early stages of AD. Here we aimed to investigate the association
between these tests and synaptic depletion and amyloid deposition
measured by [18 F]FDG and [18 F]AV45 respectively. Methods: We analyzed
a subsample of participants from ADNIGO & ADNI2 who had
clinical, neuropsychological, and [18F]AV45/[18F]FDG data collected
in a single visit. Diagnosis of cognitively normal (CN), early mild cognitive
impairment (EMCI), late mild cognitive impairment (LMCI) and AD
was adjusted using ADNI2 guidelines. Scores on the Rey Auditory Verbal
Learning Test 30 min delay (RAVLT30), total recall (RAVLTT), sum of
trials (RAVLTST), Logical Memory 30 min delay (LM) were obtained
from the ADNI database. RAVLTST was calculated by adding the 5 initial
trials of the Rey Auditory Verbal Learning Test. T1 MRIs underwent
non-uniformity correction, were skull-stripped and nonlinearly registered
to MNI152 space. After registration to MRI, PET uptake ratios were calculated
dividing [18F]AV45 and [18F]FDG scans by the median counts of
cerebellar-gm and pons, respectively. PET images were subsequently registered
to MNI152 space. Voxel-based (age corrected) regression between
PET images and LM, RAVLT30, RAVLTT and RAVLST were calculated
with PET-UR resampled to MNI space and blurred with a 6mm Gaussian filter. Results: Demographics are summarized in table 1,. [18 F]FDG uptake
correlated positively (Figure 1,) with RAVLT30 and RAVLTST in the hippocampus
of EMCI. Hypometabolism correlated with LM (left hippocampus)
in LMCI and with RAVLTST (temporal lobe) in AD. [18 F]AV45 uptake
(Figure 2) correlated negatively with LM (temporal and parietal) and
RAVLTT (parietal) in controls and with RAVLT30, RAVLTTand RAVLTST
in the frontal and parietal lobes of EMCI. For all groups collapsed, all tests
correlated with brain hypometabolism and [18 F]AV45 uptake in temporal,
parietal and frontal areas. Conclusions: RAVLT30 and RAVLTST convey
neurodegeneration in EMCI while LM reflects neurodegeneration in
LMCI. LM30 correlates with amyloid deposition only in controls, while
RAVLT30, RAVLTT and RAVLTST reflect amyloidosis in EMCI.
Original languageEnglish
Publication yearJul 2013
Number of pages1
Publication statusPublished - Jul 2013
EventAlzheimer's Association International Conference - Boston, United States
Duration: 13 Jul 201318 Jul 2013


ConferenceAlzheimer's Association International Conference
CountryUnited States

    Research areas

  • Alzheimer, MCI, memory, amyloid

See relations at Aarhus University Citationformats

ID: 69482726