Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Mar 2;89(5):971-982.
doi: 10.1016/j.neuron.2016.01.028.

PET Imaging of Tau Deposition in the Aging Human Brain

Free PMC article

PET Imaging of Tau Deposition in the Aging Human Brain

Michael Schöll et al. Neuron. .
Free PMC article


Tau pathology is a hallmark of Alzheimer's disease (AD) but also occurs in normal cognitive aging. Using the tau PET agent (18)F-AV-1451, we examined retention patterns in cognitively normal older people in relation to young controls and AD patients. Age and β-amyloid (measured using PiB PET) were differentially associated with tau tracer retention in healthy aging. Older age was related to increased tracer retention in regions of the medial temporal lobe, which predicted worse episodic memory performance. PET detection of tau in other isocortical regions required the presence of cortical β-amyloid and was associated with decline in global cognition. Furthermore, patterns of tracer retention corresponded well with Braak staging of neurofibrillary tau pathology. The present study defined patterns of tau tracer retention in normal aging in relation to age, cognition, and β-amyloid deposition.

Keywords: PET; aging; amyloid; memory; tau.


Figure 1
Figure 1. AV-1451 mean images of participant groups
Surface projections and coronal slices are presented for young healthy adults (YA), older adults (OA) split into PiB- and PiB+ subgroups for illustration, and AD patients. Differences between cognitively healthy participants were primarily localized in inferior and lateral temporal subregions, while patient vs. control differences extended into other temporal as well as parietal and frontal cortical regions.
Figure 2
Figure 2. Braak stage ROI AV-1451 uptake by group
(A) Results from Mann-Whitney U Test, uncorrected for partial volume effects. (B) Results from Mann-Whitney U Test, corrected for partial volume effects. * p < .05, ** p < .01, *** p < .001. See also Tables S1-S2, Figure S1, and Movies S1-S3.
Figure 3
Figure 3. In vivo Braak staging using conditional inference trees
Braak staging of healthy young adults (YA, n = 5), healthy older adults (OA, n = 33) and AD patients (n = 15) based on AV-1451 Braak ROI uptake, with participant characterization by in vivo assigned Braak stage. ICV = Intracranial volume. Lines indicate means and first and third quartiles.
Figure 4
Figure 4. AV-1451 in Braak ROIs vs. cognition in healthy elderly
(A) Significant relationships of AV-1451 in Braak ROI I/II with cross-sectional performance and retrospective change (slopes) in episodic memory and (B) of AV-1451 in all Braak stage ROIs with retrospective change (slopes) in global cognition in healthy OA participants, grouped into PiB+ and PiB-. See also Figure S3.
Figure 5
Figure 5. Cross-sectional and longitudinal episodic memory vs. AV-1451 uptake in OA
Nonparametric voxelwise regression of AV-1451 on (A) cross-sectional performance and (B) retrospective longitudinal change in episodic memory. See also Table S3.
Figure 6
Figure 6. Age, Aβ, and tau associations among OA
(A) Positive associations between chronological age and AV-1451 accumulation in OA (n = 33), controlling for global PiB DVR, exist in medial temporal and ventral frontal subregions. Note all Age FWE results (left MTL, ventral frontal) are overlapped and only liberal and overlap colors are visible. (B) Positive associations between PiB DVR and AV-1451 accumulation, controlling for chronological age, appear in additional temporal regions, as well as in parietal and frontal cortex. All results from voxelwise nonparametric regression, with mean-centered covariates of age (years) and PiB (global cortical DVR). Multiple statistical thresholding approaches were considered: a combined uncorrected cluster (k > 100 voxels) plus height (p < .005) threshold (“liberal”), and an FWE p < .05 correction (“conservative”). See also Table S4, Figure S4, and Movie S4.
Figure 7
Figure 7. Temporal lobe regional patterns of tau accumulation in OA
Relations of bilateral temporal ROI AV-1451 SUVR values with age (top row) and with PiB DVR (bottom row). Left to right within each row represents a medial-to-lateral path across the inferior aspect of temporal cortex. Also shown are regression results for models with one or two independent variables. ROIs: entorhinal cortex (ERC), parahippocampal gyrus (PHG), fusiform gyrus (FG), inferior temporal cortex (ITC). In bottom row, participants are median-split by age for descriptive purposes (young-old: <79y, old-old: ≥79y).

Similar articles

See all similar articles

Cited by 262 articles

See all "Cited by" articles

Publication types

MeSH terms

LinkOut - more resources