University of Southern California Dissertations and Theses

Indexing cerebrovascular resistance in cognitive decline and Alzheimer's disease

(USC Thesis Other)

Indexing cerebrovascular resistance in cognitive decline and Alzheimer's disease

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Running head: CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 1

Indexing Cerebrovascular Resistance in Cognitive Decline
and Alzheimer’s Disease

Belinda Yew

Master of Arts - Psychology
University of Southern California

August 9, 2016

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 2

Table of Contents

Abstract............................................................................................................................................3
Introduction......................................................................................................................................4
Overview and hypotheses ..............................................................................................................15
Methods..........................................................................................................................................17
Participants.........................................................................................................................17
Measures ............................................................................................................................18
Analyses.............................................................................................................................22
Results............................................................................................................................................24
Cross-sectional...................................................................................................................24
Longitudinal.......................................................................................................................26
Discussion......................................................................................................................................29
References......................................................................................................................................35
Tables.............................................................................................................................................45
Table 1: Baseline demographics ...............................................................................................45
Table 2: Group sample sizes .....................................................................................................46
Table 3: Physiological values ...................................................................................................47
Table 4: Baseline group cerebral blood flow differences .........................................................48
Table 5: Baseline group cerebrovascular resistance index (CVRi
MAP
) differences ..................49
Table 6: Baseline CVRi
MAP
x amyloid effects on cognitive performance ................................50
Table 7: Mixed model results ...................................................................................................51
Figures............................................................................................................................................56
Figure 1: Baseline group cerebral blood flow ...........................................................................56
Figure 2: Baseline group CVRi
MAP
..........................................................................................57
Figure 3: Baseline CVRi
MAP
x amyloid interaction effects on cognitive performance ............58
Figure 4: Longitudinal mOFC CVRi
MAP
x amyloid .................................................................59
Figure 5: Baseline CVRi
MAP
x amyloid prediction of longitudinal global cognition ...............60
Figure 6: Baseline CVRi
MAP
x amyloid prediction of longitudinal executive function ...........61
Figure 7: Baseline CVRi
MAP
x amyloid prediction of recognition ...........................................64
Appendix 1: ADNI 2 sample composition.....................................................................................65

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 3

Abstract

Given the proposed role of vascular factors in normal and pathological aging, we investigated
whether elevated cerebrovascular resistance in brain regions typically affected by cognitive
aging and Alzheimer’s disease (AD) was associated with amyloid-β accumulation and cognitive
decline. Cerebral blood flow (CBF) was measured in a sample of older North American adults (N
= 232) using arterial spin labeling magnetic resonance imaging. An estimate of cerebrovascular
resistance (CVRi
MAP
) was calculated as the ratio of mean arterial blood pressure to CBF in
regions of interest. Amyloid-β retention and positivity were indexed using positron emission
tomography with a florbetapir-fluorine-18 (
18
F) tracer, while cognitive performance was
evaluated via assessments of global cognition, memory, and executive function. Cross-sectional
analyses indicated highest CVRi
MAP
for AD (n = 33), intermediate values for non-demented
amyloid-positive (n = 87), and lowest estimates for amyloid-negative (n = 112) participants.
Group differences were more pronounced, and present in more regions, for CVRi
MAP
relative to
CBF. Furthermore, CVRi
MAP
but not CBF differentiated non-demented amyloid-positive from
amyloid-negative individuals. Further analyses detected interactive effects of CVRi
MAP
and
amyloid-β deposition on cognition such that cognitive deficits were moderated by, if not
dependent upon, concurrent elevation in cerebrovascular resistance. Longitudinal analyses
revealed similar interactive effects for cognitive decline across 3 time points. Declines in
cognitive performance over time were accelerated by elevated baseline CVRi
MAP
, particularly for
amyloid-positive individuals. Earlier elevations in CVRi
MAP
for frontal regions predicted later
whole-brain amyloid increases. Our results thus suggest that increases in cerebrovascular
resistance may represent a crucial component of AD pathogenesis, working synergistically with,
and in some areas pre-dating, amyloidosis to produce cognitive decline.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 4

Introduction
Alzheimer’s disease (AD) is a neurodegenerative disorder classically characterized by
marked deficits in episodic memory early in the disease process, followed by impairments to
other cognitive and behavioral domains as the disease progresses. These declines are thought to
result from accumulation in the brain of amyloid-β plaques and neurofibrillary tangles, both of
which have been associated with a host of genetic and environmental risk factors (Ittner & Götz,
2011). Affecting predominantly older (i.e. over 60 years of age) adults, AD represents the most
common form of dementia with 1 in 9 older Americans currently carrying a diagnosis. This
group is forecasted to increase in size as a growing number of Americans approach typical age of
onset for AD, escalating demands for diagnostic tools and treatment options (Ballard et al., 2011;
Herbert, Scherr, Bienias, Bennett, & Evans, 2003). Satisfaction of such demands will depend
greatly on continued development of existing models, as well as exploration of genetic and
environmental risk factors and the mechanisms by which they produce AD neuropathology.
As abovementioned, early manifestations of AD are typically characterized by memory
difficulties, with the most prominent declines observed in the ability to form new memories and
recall more recent memories. These deficits are attributed to neuronal and synapse loss in regions
associated with memory formation and recollection (i.e. medial temporal lobes) even prior to
presentation of clinical symptoms (McDonald et al., 2009). While complaints of memory deficits
are also common in healthy older adults, progression to mild cognitive impairment (MCI, a
condition often viewed as a transitional phase preceding AD), and AD is evident in worsening
memory impairment (i.e. increased severity and declines in other memory types). This is often
accompanied by deficits in other cognitive domains (e.g. executive function) as well.
Furthermore, memory complaints from the affected individual may decrease with disease

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 5

progression as insight is lost due to pathology of the frontal lobes (Peña-Casanova, Sánchez-
Benavides, de Sola, Manero-Borrás, & Casals-Coll, 2012). This “spread” of cognitive deficits to
non-memory functions (e.g. attention, word-finding, and praxis) is reflective of increased
atrophy in bilateral frontal and parietal regions along with continued deterioration of temporal
structures (Frisoni, Fox, Jack, Scheltens, & Thompson, 2010).
Neuronal and synapse loss underlying cognitive deficits in AD have been explored using
a number of neuroimaging techniques. Unsurprisingly, structural magnetic resonance imaging
(MRI) has revealed brain atrophy consistent with observed cognitive deficits. Earlier stages of
disease are associated with greatest atrophy in medial temporal regions, while more advanced
stages see marked increases in atrophy of prefrontal, parietal, posterior temporal, and cingulate
cortex regions (McDonald et al., 2009). Reduced hippocampal volume has consistently been
observed for MCI and AD samples, with greatest reduction seen in the latter. Furthermore,
baseline hippocampal volume predicts conversion to AD and disease progression (Wang, 2014).
Devanand et al. (2007) reported analogous findings as well as decreased entorhinal cortex
volume in MCI and AD, again with larger reductions for AD. Entorhinal cortex volume was also
found to predict conversion to AD and disease progression (Devanand et al., 2007). Vemuri et al.
(2009) have similarly examined atrophy increases over time, identifying degree and location of
atrophy as predictors of conversion to AD (i.e. in healthy and MCI individuals 2 years after
MRI). Atrophy in medial temporal regions such as the hippocampus and entorhinal cortex as
well as enlargement of ventricles, are thus structural differences found to reliably differentiate
those afflicted with AD pathology (including early stage AD and MCI groups), from healthy
controls (Frisoni et al., 2010).

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 6

Imaging investigations of brain activity (i.e. functional MRI) have yielded concordant,
albeit less clear, findings. The bulk of these studies have employed blood-oxygen-level-
dependent (BOLD) functional MRI (fMRI), which indexes brain activation via mapping of
variations in magnetization that follow changes in blood oxygenation associated with increased
neural activity. Essentially, elevations in blood flow that follow increased neural activity result in
increased blood oxygenation relative to other neural regions, which in turn produces detectable
changes in magnetic resonance signal (i.e. BOLD signal change). Time course of such change
can be correlated with presentation of a stimulus to infer neuronal activation associated with a
particular task, or examined during “rest” state where no explicit activity is performed (Buxton,
Uludag, Dubowitz, & Liu, 2004).
Consistent with observed cognitive deficits, task-based BOLD fMRI studies have
identified diminished activation of left parietal and hippocampal regions in AD relative to
healthy individuals during episodic memory retrieval. In contrast, regions typically unaffected by
AD show normal or greater than normal activation, with the latter potentially reflecting
compensatory processes. Cognitive impairments characteristic of AD often render completion of
relevant (e.g. memory-based) fMRI tasks difficult, if not impossible, for affected individuals.
Resting-state fMRI research has thus been of increasing interest and importance (Small et al.,
2008). These investigations have focused largely on AD-related abnormalities in neural regions
typically active during wakeful rest (i.e. the default-mode network). Specifically, brain areas
comprising the default-mode network (e.g. posterior cingulate and hippocampus) show reduced
activity and decreased coordination during resting state (Greicius, Srivastava, Reiss, & Menon,
2004). Healthy adults carrying a genetic risk factor for AD have also demonstrated resting state
activation predictive of later declines in memory, an association not observed for non-carriers

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 7

(Bookheimer et al., 2000). The BOLD fMRI literature for AD is thus suggestive of atypical
activation, particularly in regions typically disrupted by AD pathology.
The BOLD approach to fMRI is often criticized for employing indirect measurement of
activation (i.e. obtained values reflect cerebral metabolic rate of oxygen [CMRO
2
] rather than
activation per se). Furthermore, obtained values are sensitive to cerebral blood volume in
addition to CBF, and variations in baseline physiological state can exacerbate difficulties in
BOLD signal interpretation (Buxton et al., 2004). Arterial spin labeling (ASL) fMRI offers a
more direct gauge of CBF through measurement of magnetized arterial blood water. By applying
a radiofrequency pulse to arterial blood (e.g. at the common carotid artery), net magnetization of
the blood water is inverted and water molecules “labeled”. When labeled blood enters brain
regions of interest (ROI), it is exchanged with tissue water, altering overall tissue magnetization
and resulting in reduced MR signal, at which point a “tag” image is taken. This image is
compared to a control image wherein blood is not labeled. The resulting perfusion image
therefore reflects the amount of arterial blood delivered to each voxel of interest within a
specified time period (Wang, 2014).
In recent years ASL has been increasingly applied to AD research, with reduced blood
flow (i.e. hypoperfusion) and elevated blood flow (i.e. hyperperfusion) interpreted as decreased
and increased activation, respectively. Several studies have identified abnormalities in AD and
MCI samples relative to healthy controls (Wang, 2014). Johnson et al. (2005), for example,
identified bilateral inferior parietal, precuneus/posterior cingulate, and prefrontal hypoperfusion
in AD relative to control participants. While methodological limitations prevented detection of
perfusion in anatomically lower areas, including the medial temporal regions typically affected
by AD, a later study identified hyperperfusion in the hippocampus and other regions disrupted

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 8

during early stages of AD (e.g. parahippocampal gyrus, anterior cingulate, and superior temporal
pole) when corrected for atrophy (i.e. calculation of perfusion per voxel). Consistent with
Johnson et al.’s (2005) findings, the study also observed hypoperfusion in the bilateral precuneus
and parietal association cortex, as well as the inferior temporal lobe (Alsop, Casement, de
Bazelaire, Fong, & Press, 2008). Dai et al. (2009) replicated these findings in another AD
sample, with hypoperfusion also observed in left prefrontal regions. Finally, perfusion
abnormalities have also been identified in MCI individuals, with findings of low posterior
cingulate CBF in representative samples (Chao et al., 2009). Notably, AD abnormalities detected
by ASL are largely anatomically distinct from those evident in structural MRI, suggesting unique
contributions from both imaging modalities in diagnosis and disease prediction (Wang, 2014)
Structural and functional abnormalities observed in AD-afflicted individuals have been
attributed to a number of neuropathological changes. Accumulation of abnormal protein tau
tangles (i.e. microtubule-associated protein tau, MAPT) in neuronal cell bodies and dendrites
have been correlated with clinical progression in AD patients, and mutation of the tau gene
associated with a familial AD subtype. AD brain changes have also been linked to amyloid-β
plaques, with a separate subset of inherited AD connected to mutated production of an amyloid
precursor protein (APP) (Bertram & Tanzi, 2005). Despite substantial progress in genetic
research concerning familial forms, the majority of AD cases constitute sporadic, later-onset
variants in which causes of amyloid and tau accumulation are unknown. Polymorphisms of the
apolipoprotein gene (APOE) have been consistently associated with increased risk of AD but are
not essential nor do they guarantee development of AD. The nature of APOE-related changes
responsible for AD risk is unclear, although dysfunction in amyloid-β clearance and/or lipid
metabolism is probable (Bertram & Tanzi, 2005). As Cuyvers and Sleegers (2016) concede,

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 9

there are likely many as yet undiscovered AD candidate genes and the relative contributions of
known mutations may therefore be minor.
Elevated AD risk has also been attributed to a number of vascular factors based on
associations between vascular variables and cognitive and neurodegenerative features of AD.
Raz and Rodrigue (2006), for example, cite consistent findings of neurostructural differences in
individuals with cardiovascular risk factors such as hypertension and history of ischemia (i.e.
restricted blood supply to tissues, causing insufficient oxygenation and glucose supply that can
lead to tissue death). Such individuals experience increased lesions to brain white matter (i.e.
white matter hyperintensities, WMH), accelerated atrophy to medial temporal structures such as
the hippocampus, and longitudinal reductions in neural regions typically associated with stable
volume in healthy aging (Raz & Rodrigue, 2006). This is consistent with other findings of
greater ventricular volume, reduced gray matter, and reduced overall brain volume in subjects
suffering cardiovascular risk factors (Gianaros, Greer, Ryan, & Jennings, 2006; Strassburger et
al., 1997; Wiseman et al., 2004). Unsurprisingly, these structural deficits are also associated with
cognitive impairments and increased risk of AD. Further, where hypertension is treated with
certain medications, structural and cognitive declines appear to be attenuated (Qiu, Winblad, &
Fratiglioni, 2005).
These results have been extended to neurodegenerative disease contexts, with several
studies identifying links between vascular factors like blood pressure and aforementioned
structural and functional deficits in AD-afflicted brains. Beason-Held et al. (2007), for example,
report reduced cerebral glucose metabolism for hypertensive versus healthy control participants
in neural regions associated with AD and general aging. This suggests that regions typically
susceptible to AD neuropathology are similarly vulnerable to hypertension-related dysfunction

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 10

(Beason-Held et al., 2007). Analogous findings were reported by Langbaum et al. (2012), with
systolic and diastolic blood pressure, as well as pulse pressure (PP, the difference between
systolic and diastolic blood pressure, which indexes fluctuation in pressure with each heart
contraction), inversely related to glucose metabolism in frontal and temporal brain regions, as
well as amyloid-β deposition (i.e. as indexed by higher Pittsburgh Compound-B [PiB] retention)
in cognitively healthy participants. This replicates correlations published by Petrovitch et al.
(2000) between mid-life blood pressure and late-life brain amyloid-β and atrophy, and is
consistent with reports of increased MAP-estimated cerebrovascular resistance (CVRi
MAP
, the
ratio of mean arterial pressure [MAP] to CBF) in AD relative to control participants (Nation et
al., 2013). Bangen et al. (2014) have also identified links between age and reduced CBF in
individuals with multiple vascular risk factors but not those with low vascular risk. There is thus
extensive evidence of connection between vascular variables, particularly those reflecting
vascular resistance, and AD neuropathology.
In synthesis of these vascular findings with known AD pathology, various models have
proposed vascular mechanisms through which amyloid-β accumulation may occur (Kress et al.,
2014; Weller, Subash, Preston, Mazanti, & Carare, 2008; Zlokovic, 2011). Reductions in CBF
alongside disintegration of the blood brain barrier (BBB), for example, have been posited as
precursors to dementia (Zlokovic, 2011). Certainly, reduced CBF has been associated with both
normal and pathological aging across a number of brain regions (Iadecola, 2004), with decreases
in older relative to younger adults (Chen, Rosas, & Salat, 2011), and accelerated or more intense
declines observed in cases of preclinical and clinical AD (Alsop, Dai, Grossman, & Detre, 2010;
Dai et al., 2009). Moreover, CBF dysfunction seems to predate neurodegeneration, amyloid-β
accumulation, and cognitive decline in those at genetic risk for AD (e.g., amyloid-β precursor

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 11

protein [APP] gene mutation or apolipoprotein E ε4 [ApoE4] genotype), suggesting CBF
reductions may contribute to early disease stages (Bookheimer et al., 2000; Iadecola, 2004;
Sheline et al., 2010).
Even relatively modest declines in CBF (i.e., oligemia or mild hypoperfusion) can disrupt
protein synthesis, impeding synaptic plasticity necessary for memory formation (Mies, Ishimaru,
Xie, Seo, & Hossmann, 1991). With more severe CBF declines, hypoxia may arise, hampering
enzymatic activity crucial for initiation of action potentials and disturbing brain pH, electrolyte,
and water gradient homeostasis. Prolonged disruption may then instigate white matter lesion
formation and neurotoxicity, including the pathophysiological hallmarks of AD (i.e.
accumulation of amyloid-β and hyperphosphorylated tau) (Zlokovic, 2011). Ischemic hypoxia,
for example, promotes activity of amyloid-β producing enzymes (Guglielmotto et al., 2009),
while decreasing amyloid-β degrading enzyme neprilysin (Wang et al., 2011). These enzymatic
changes may be exacerbated by hypoxia-driven alterations to genes expressed by the
cerebrovascular endothelium, causing deterioration of cerebral blood vessels and CBF
deregulation (Chow et al., 2007; Wu et al., 2005).
Stiffening cerebrovasculature may drive and/or operate parallel to these CBF
abnormalities. Weller et al. (2008), for example, have proposed that blood vessel pulsations
provide the mobility necessary for drainage. Stiffening of vessels (i.e. reduced compliance of
vessels in accommodation of blood pressure changes) therefore reduces pulsation amplitude and
consequently capacity for amyloid-β drainage. This is consistent with amyloid-β deposition
along drainage-facilitating basement membranes of cerebrovascular vessel walls in AD, leading
to changes like smooth muscle loss that eventually cause vessel sclerosis (Weller et al., 2008).
Impaired clearance ability consequently causes accumulation of amyloid-β, leading to plaque

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 12

formation and/or increased risk of white-matter lesions that further promote AD symptomatology
(Bell et al., 2009; Qiu et al., 2005; Weller, Boche, & Nicoll, 2009). Amyloid-β accumulation can
in turn produce further obstruction of blood/fluid flow via effects of cerebral amyloid angiopathy
(i.e. amyloid-β accumulation in brain blood vessels impedes drainage) and neurotoxicity of
amyloid-β oligomers (Weller et al., 2008). Amyloid-associated vasotoxicity may also produce
vasoconstriction itself (Han et al., 2015).
Vessel stiffening can be gauged using cerebrovascular resistance, the ratio of cerebral
perfusion pressure (P
α
) to CBF, where P
α
is calculated as the difference between MAP and
intracranial pressure (ICP; see Equation 1 below). Under typical conditions wherein ICP is
normal and considerably lower than MAP, cerebrovascular resistance may be estimated as the
ratio of MAP to CBF (see Equation 2). Homeostasis of ICP and CBF is maintained by
autoregulatory processes amid even significant changes in MAP (e.g. fluctuations of 60 to 150
mmHg) (Buxton, 2009; Paulson, Strandgaard, & Edvinsson, 1990). These highly efficient
processes modulate CBF by adjusting arteriolar diameter in response to changes in blood
pressure (BP), while also integrating responses to carbon dioxide (CO
2
) tension, neural activity
and metabolism, autonomic activity, and other local and systemic stimuli. Put simply, MAP
increase elicits reduction in blood vessel diameter while MAP decrease results in vessel dilation.
Chronic MAP elevation is thus associated with persistent constriction of blood vessels. The
relationship between blood vessel diameter (or vessel radius, r) and CBF is also influenced by
vessel length (θ) and blood viscosity (η), as described by Poiseuille’s law (see Equation 3).
Given the relatively stable nature of θ and η, however, CBF changes are largely determined by
vessel radius, which powerfully modulates CBF by virtue of its fourth power proportion. For
example, a 5% change in vessel radius leads to a 6.25 fold change in CBF (Buxton, 2009).

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 13

Changes in vessel diameter thus represent the primary proximal factor determining dynamic CBF
changes that constitute cerebral autoregulation.

Experimental animal studies have demonstrated that autoregulation of cerebral vessel
diameter is predominantly exerted by the arteriolar compartment, as invasively measured by
vascular resistance changes. Changes are minimal in larger upstream arteries but precipitous
drops occur prior to the capillary bed (Nichols, O' Rourke, & Vlachopoulos, 2011). Although
beat-to-beat cerebral arteriolar diameter is modulated by the convergence of many homeostatic
processes, blood pressure is the primary determinant of steady state changes on the macroscale.
According to our model, chronic CBF reductions in response to blood pressure elevation are
therefore indicative of a chronic state of increased cerebrovascular resistance. At steady state,
this increase in baseline resistance could be due to chronic vasoconstriction (r) or reduced
arteriolar density (θ). Cerebral arterioles typically adopt a partially constricted baseline state,
allowing for either dilation or further constriction in response to chemical or hemodynamic
signalling. This tonic baseline state of dynamic vasoconstriction may underlie the chronic
vasoconstriction (i.e. inability to dilate despite intact capacity for constriction) that characterizes
cerebrovascular stiffening.
Support for the role of cerebrovascular stiffening in AD pathogenesis (e.g. as proposed
by amyloid clearance models like those abovementioned) is evident in findings of elevated CVRi
[1] Cerebral perfusion pressure (P
α
) = MAP – ICP
[2] Cerebrovascular resistance = P
α
→ MAP
CBF CBF
[3] Blood flow = ∆ BP x π x r
4

8 x η x θ

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 14

throughout subcortical, medial-temporal, posterior cingulate, precuneus, inferior parietal, and
superior temporal regions of AD relative to control brains. Fittingly, intermediate CVRi
magnitudes were observed in MCI participants, and across AD and MCI cases thalamus and
caudate CVRi values were inversely related to global cognition (Nation et al., 2013). Consistent
findings have been obtained using a cerebrovascular resistance measure derived from total CBF
(i.e. averaged across brain regions) in amnestic MCI (aMCI) and control participants (Liu et al.,
2014). Relative to controls, the aMCI group had cerebrovascular resistance estimates almost 13%
higher (Liu et al., 2014).
More broadly, these findings implicate deficits in cerebrovascular autoregulation (i.e.
maintenance of stable CBF) in the greater cascade through which AD arises (Meel-van den
Abeelen, Lagro, van Beek, & Claassen, 2014). Significant dysfunction of cerebrovascular
resistance regulation in response to induced blood pressure changes (i.e. through repeated cycles
of sitting and standing), for example, was observed in AD patients compared to controls. CBF is
also influenced by vasomotor reactivity (i.e. in response to CO
2
tension) and neuronal glucose
metabolism, both of which are abnormal in AD (Buckner et al., 2005; Meel-van den Abeelen et
al., 2014). Furthermore, recent work has linked vasomotor reactivity to loss of vascular
compliance (V. Marmarelis, personal communication, 5/10/16), suggesting that these processes
are highly enmeshed. Consequently, while clear connections have been identified between CBF
and amyloid-β deposition, the relative contributions of cerebrovascular resistance, vasomotor
reactivity, and neurovascular coupling (i.e. the relationship between neural activity/metabolism
and corresponding CBF changes) will need to be disentangled (Vlassenko et al., 2010).
Exploration of BP, CO
2
tension, and neuronal metabolism (which index cerebrovascular
resistance, vasomotor reactivity, and neurovascular coupling, respectively) in relation to

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 15

amyloid-β deposition is thus imperative to more thorough understanding of the mechanisms
involved in amyloid-β accumulation and subsequent progression to AD.
Overview and Hypotheses
To our knowledge, no studies have investigated estimates of cerebrovascular resistance in
relation to amyloid-β retention and longitudinal decline in neuropsychological function. Further,
the majority of studies exploring CBF in Alzheimer’s samples have employed BOLD fMRI
measures of CBF, which may offer insufficient accuracy for evaluation of vascular properties
like cerebrovascular resistance (Bookheimer et al., 2000; Greicius et al., 2004). In the present
study we therefore sought to address these concerns through comparisons of estimated
cerebrovascular resistance using ASL-derived CBF in AD patients, amyloid-positive individuals
without diagnosis of AD (heretofore referred to as “non-demented amyloid-positive”), and
amyloid-negative participants recruited through the Alzheimer’s Disease Neuroimaging Initiative
(ADNI). Specifically, we sought to identify group differences for CVRi
MAP
in ROIs both cross-
sectionally and longitudinally, and using CBF values both adjusted and un-adjusted for neuronal
metabolism (i.e. as indexed by fludeoxyglucose [FDG]-PET). In addition, we examined cross-
sectional and longitudinal CVRi
MAP
and amyloid-β retention differences, and their relation to
cognitive decline.
Our overarching hypothesis was that vascular mechanisms driving amyloid-β
accumulation would be evident in association of cerebrovascular resistance estimates with
amyloid-β retention and cognitive decline. Neural regions generally associated with abnormality
in AD were therefore hypothesized to show higher cerebrovascular resistance, which we
predicted would correlate with elevated amyloid load and cognitive decline. More explicitly, we
hypothesized that:

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 16

1. Baseline cerebrovascular resistance (i.e. CVRi
MAP
) would be highest for AD,
intermediate for non-demented amyloid-positive, and lowest for amyloid-negative
participants, in regions typically affected by AD (i.e. medial orbitofrontal cortex
[mOFC], rostral middle frontal gyrus [rMFG], hippocampus, inferior temporal cortex
[ITC], inferior parietal cortex [IPC], and precuneus). Regions showing CVRi differences
were predicted to be more widespread, emerge earlier (i.e. show more differences
between non-demented amyloid-positive and amyloid-negative groups), and reflect
greater statistical significance and effect sizes than CBF differences.
2. Baseline CVRi
MAP
increases would correspond to declines in global cognition (i.e. Mini-
Mental State Examination [MMSE] score), memory (i.e. as measured by the Rey
Auditory Verbal Learning Test [RAVLT]), and executive function (i.e. as measured by
the Trail Making Test [TMT]). Performance on the MMSE, RAVLT delayed recall and
recognition, and TMT Trials A and B, were thus predicted to inversely correlate with
CVRi
MAP
and amyloid load (i.e. amyloid-negative > amyloid-positive) at baseline.
3. Higher baseline CVRi
MAP
would predict greater declines in cognitive performance across
the three time points. Furthermore, longitudinal cerebrovascular resistance changes
would predict cognitive decline (i.e. those with larger CVRi
MAP
increases would show
poorest cognitive performance)
4. Higher CVRi
MAP
would predict greater amyloid load and larger increases in amyloid at
future time points. More specifically, we predicted that higher baseline CVRi
MAP
would
be associated with higher amyloid load at times 2 and 3, and larger amyloid increases
from baseline to time 2, and time 2 to time 3. Likewise, we predicted that higher time 2

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 17

CVRi
MAP
would be associated with greater amyloid load at time 3, and larger amyloid
increases from time 2 to time 3.
Methods
Participants
Participants were selected from the ADNI database, a repository of data obtained from
volunteer adults aged 54 and above at over 50 sites across North America. All participants
completed the MMSE and Clinical Dementia Rating scale (CDR). The former is a 30-item
measure of cognitive function across a variety of domains such that scores below 25 indicate
mild (19-24 points), moderate (10-18 points) or severe (≤9 points) impairment (Folstein,
Folstein, & McHugh, 1975). The CDR is a structured interview assessing severity of cognitive
and functional dementia symptoms. Scores range from 0 (no symptoms) to 3 (severe
symptoms)(Morris, 1993). AD participants met National Institute of Neurological and
Communicative Disorders and Stroke, and Alzheimer’s Disease and Related Disorders
Association (NINCDS-ADRDA) criteria for probable Alzheimer’s disease and had MMSE
scores between 20 and 26, as well as CDR of 0.5 or 1 (ADNI, 2011) at the time of their first ASL
scan. Remaining participants were classified based on amyloid-β load at initial ASL scan. Those
with summary uptake value ratio (SUVR) greater than 1.11 were categorized as amyloid-
positive, while those with SUVR ≤ 1.11 were classified as amyloid-negative. Included AD
participants were all amyloid-positive.
ADNI data collection officially began in 2005 and the first (ADNI 1) and second (ADNI
GO) waves of the study have since concluded. A third wave (ADNI 2) commenced in 2011 and
is scheduled for completion this year. The latter iterations, ADNI GO and ADNI 2, represented
continued efforts to determine biomarkers of AD development while extending the study

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 18

protocol to include testing for blood, DNA, and CSF markers of AD; notably, ASL fMRI and
florbetapir PET were added as routine procedures. Each study wave comprised both newly
enrolled participants as well as eligible MCI and control participants from prior phases. In
addition to new recruits, ADNI GO thus included a number of participants originally enrolled in
ADNI 1, and ADNI 2 included a number of participants originally enrolled in ADNI 1 and/or
ADNI GO. ADNI 2 also included 200 newly recruited AD patients. Analyses in the present
study will be restricted to ADNI GO and ADNI 2 participants as ASL fMRI and florbetapir PET
were not administered during the ADNI 1 wave. Participants comprising each stage of ADNI 2
are described in Appendix 1. Only participants who underwent ASL fMRI and florbetapir PET
scanning were included in the present study. Baseline (i.e. at first ASL scan) demographic and
vital signs information for participants included in the present study are summarized in Table 1.
Group sample sizes for each year included in longitudinal analyses are presented in Table 2.
Measures
Blood pressure (BP) was measured using a calibrated mercury sphygmomanometer and
blood pressure cuff while the participant was seated with forearm positioned horizontally at
approximately the same level as their heart. Where possible, pressure was measured using the
participant’s dominant arm.
Cerebral blood flow (CBF) was determined from resting ASL fMRI imaging undergone
by a subset of participants. Scanning was conducted on 3.0 T scanners using a pulsed ASL
method (QUIPS II with thin-slice TI1 periodic saturation) with echo-planar imaging (EPI). Scan
parameters were as follows: field of view = 256mm, repetition time (RT) = 3400ms, echo time =
12ms, inversion time for arterial spins (TI1) = 700ms, total transit time of spins (TI2) = 1900ms,
matrix = 64 x 64, tag thickness = 100mm, gap from tag to proximal slice = 25.4mm, slice

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 19

thickness = 4mm, number of axial slices = 24, and time between slices = 22.5ms. CBF was
calculated by normalizing scaled, distortion-corrected, co-registered, and partial volume
corrected perfusion weighted images to a reference image estimating blood water magnetization.
This yielded an estimate of CBF based on physical units of arterial water density (ml/100g/min).
More detailed descriptions of ASL acquisition and processing procedures are provided by ADNI
(ADNI, 2011). CBF values will be analyzed both with and without adjustment for neuronal
metabolism. Adjustment will be accomplished by residualizing regional FDG-PET values in
regression and mixed models analyses.
Mean arterial pressure (MAP), the average blood pressure within an individual’s
arteries during a single cardiac cycle, was calculated by multiplying diastolic blood pressure
(DBP) by 2 then adding systolic blood pressure (SBP) and dividing by 3. This can also be written
as: SBP + 2 (DBP)
3
Cerebrovascular resistance was measured using a cerebrovascular resistance index
derived by dividing regional CBF by MAP (i.e. CVRi
MAP
). This was computed for regions
typically affected in AD (i.e. rMFG, mOFC, hippocampus, ITC, IPC, and precuneus).
Amyloid-β retention was indexed using florbetapir-fluorine-18 (
18
F). Due its tendency to
bind to amyloid-β, this compound is used as a tracer in PET scanning to identify amyloid-β
deposits. Each participant was therefore administered a single intravenous bolus of florbetapir-
fluorine-18 prior to AV-45 PET imaging. PET imaging commenced approximately 50-70
minutes following injection and ran for 20 minutes. Image reconstruction was conducted
immediately after scanning and, where motion artefact was identified, scanning repeated.
Florbetapir was quantified using 3.0 T 3D MPRAGE structural scans collected earlier. Average
florbetapir uptake (i.e. SUVR) was computed for whole brain grey matter relative to uptake

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 20

within the entire cerebellum (i.e. grey and white matter). Non-AD participants with SUVR
greater than 1.11 were classified amyloid-positive, while participants with SUVR ≤ 1.11 were
classified amyloid-negative. This cutoff equates to the upper limit of a 95% confidence limit for
mean SUVR obtained in a sample of normal controls (Joshi et al., 2012). More detailed
information regarding procedures is available through ADNI (ADNI, 2011).
Glucose uptake was indexed using fludeoxyglucose (FDG) PET imaging. FDG is a
glucose analog with radionuclide fluorine-18 (
18
F), which indexes glucose uptake by tissue and is
thus thought to reflect brain metabolism. Each participant was injected with a single intravenous
bolus of FDG 30 minutes prior to scanning. Scan duration was approximately 30 minutes. Image
reconstruction was conducted immediately after scanning and, where motion artefact was
identified, scanning repeated. Images were spatially normalized to a Montreal Neurological
Institute (MNI) PET template before mean intensity values were extracted for ROIs defined
based on prior studies of metabolic changes in pathological aging. A composite region based on
combined values for all ROIs was also calculated. Individual and composite ROIs were intensity
normalized by dividing by the mean for a pons/cerebellar vermis reference region (Landau et al.,
2011). The composite value was utilized in the present study as an index of global brain
metabolism.
Global cognitive decline was measured using the aforementioned MMSE. This scale
comprises items evaluating orientation to time and place, registration (i.e. repeating 3 named
objects), recall of registered items, attention and calculation, language, repetition, and following
complex commands. Administration takes approximately 5 to 10 minutes. Each correct answer is
awarded 1 point; higher scores thus reflect better global cognitive ability.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 21

Memory was assessed using the delayed recall and recognition subtests of the RAVLT.
Briefly, this test involves orally presenting participants with a list of 15 words that they are
instructed to remember. The list of words is given 5 times, with participants asked to recall as
many words as possible after each presentation. After these 5 recall trials, a second list of 15
unrelated words is presented and the participant must, again, recall as many words as possible.
This is followed by another recall trial for the original list (i.e. participants recall as many of the
words from the first list as they can). After a 20-30 minute delay, another recall trial is conducted
to test for longer-term retention. The number of correctly recalled words is summed for a delayed
recall score. A recognition test is then administered wherein participants are read a list of 30
words and asked to identify which were on the original list; the number of correctly recognized
words is summed for a recognition score (Rey, 1941).
Executive function was evaluated using the Trails A and B trials of the TMT. In Trails A
participants are presented with a random spread of 25 circles numbered from 1 to 25. Beginning
at circle 1, the participant must draw lines connecting circles in ascending order as quickly as
possible (e.g. from 1 to 2; 2 to 3; 3 to 4, and so on). A maximum of 150 seconds is allowed
before the trial is ended. Trails B also comprises 25 circles, with 13 numbered from 1 to 13, and
13 labelled with letters A through L. Participants must connect circles by alternating between
numbers and letters in ascending and alphabetical orders respectively. Beginning at circle A,
they would thus draw a line to circle 1, then from 1 to B; B to 2; 2 to C, and so forth. A
maximum of 500 seconds for completion is allowed. For both trials, scoring is based on number
of seconds taken to complete each task (Spreen & Strauss, 1998).

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 22

Analyses
Cross-sectional
To evaluate differences in cerebrovascular resistance across degrees of pathology, we
conducted a one-way analysis of variance (ANOVA) for baseline CVRi
MAP
by group (i.e. AD vs.
amyloid-positive vs. amyloid-negative), controlling for age, APOE genotype, body mass index
(BMI), sex, and FDG uptake. Where significant main effects were identified, post-hoc least
significant difference (LSD) pairwise analyses of AD vs. amyloid-positive, AD vs. amyloid-
negative, and amyloid-positive vs. amyloid-negative comparisons were conducted to clarify
group differences.
Effects of cerebrovascular resistance and amyloid-β status on cognitive performance were
evaluated via multivariate analysis of variance (MANOVA). Cognitive test scores were entered
as dependent variables, with regional CVRi
MAP
, amyloid group (positive or negative), and
CVRi
MAP
x amyloid group interaction terms entered as independent variables. Age, sex, BMI,
FDG uptake, and APOE genotype were included as covariates. We employed post-hoc LSD
pairwise analyses to more precisely identify group differences where significant omnibus results
were obtained.
Baseline group differences across demographic variables (i.e. age, sex, education, BMI
and APOE genotype) were analysed via ANOVA. Baseline group differences across
physiological variables (i.e. SBP, DBP, MAP, and PP) were also evaluated using ANOVA, with
age, sex, BMI, and APOE genotype included as covariates.
Longitudinal
Longitudinal changes in CVRi
MAP
were assessed using repeated measures ANOVAs.
Regional CVRi
MAP
values from 2 time points were entered as within-subjects variables, with

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 23

amyloid-status (i.e. negative or positive) entered as a between-subjects factor. APOE genotype,
sex, and mean-centered age, FDG uptake and BMI were entered as covariates. Where significant
omnibus effects were detected, post-hoc LSD pairwise comparisons were performed to identify
differences between amyloid-negative and positive groups (i.e. in longitudinal CVRi
MAP

changes).
Effects of baseline CVRi
MAP
on longitudinal changes in cognitive performance were
evaluated using linear mixed models employing maximum likelihood estimation. Cognitive
measures (i.e. MMSE, RAVLT, and Trails) were entered as dependent variables, with fixed
effects of amyloid-status, baseline CVRi
MAP
, time, time x amyloid-status, time x CVRi
MAP
, and
time x amyloid-status x CVRi
MAP
. Covariates of APOE genotype, sex, and mean-centered age,
BMI, education, and FDG uptake were also included. Time, time x amyloid-status, time x
CVRi
MAP
, and time x amyloid-status x CVRi
MAP
, were entered as random effects with an
autoregressive covariance structure.
Effects of CVRi
MAP
changes over time on cognitive performance were analysed using
linear regression. Specifically, cognitive test scores at time 2 and 3 were entered as dependent
variables, while amyloid-status, CVRi
MAP
change scores (i.e. time 1 vs. 2, and time 2 vs. 3), and
amyloid x CVRi
MAP
interaction terms were entered as independent variables. Covariates of
APOE genotype, sex, and mean-centered age, BMI, education and FDG uptake were also
included. Significant interaction effects were deconstructed via stratified analyses (e.g. for
amyloid-negative and amyloid-positive groups).
Finally, prediction of amyloid load at time 3 by CVRi
MAP
change was assessed using
linear regression. Amyloid load was entered as the dependent variable, with CVRi
MAP
change

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 24

scores (i.e. time 1 vs. 2, and 2 vs. 3) entered as independent variables. Covariates of APOE
genotype, sex, and mean-centered age, BMI, and FDG uptake were also included.
Results
Cross-sectional Analyses
Group comparisons of demographic and physiological variables
Baseline demographic and vital signs information for included participants is the same as
that presented earlier in Table 1. The AD and amyloid-positive groups were significantly older
than amyloid-negative group (both p < .001). The AD participants also had significantly lower
BMIs than their amyloid-negative counterparts (p = .015). There were significantly more carriers
of a single ApoE4 allele in the AD and amyloid-positive groups relative to the amyloid-negative
group (p < .05). There were also significantly more carriers of two ApoE4 alleles in the AD
relative to amyloid-positive and amyloid-negative groups (p < .05).
Group means for additional physiological variables used in cerebrovascular resistance
analyses are presented in Table 3. No significant group differences were detected (all p > .10).
Group comparisons of CBF
Group means for rCBF in target regions are presented in Figure 1. MANCOVA indicated
a significant effect of group on rCBF in three regions (see Table 4). Post-hoc LSD tests showed
that CBF was reduced for AD relative to amyloid-negative participants in the left hippocampus
(p = .031) and left ITC (p = .013). The AD group also exhibited reduced right IPC CBF
compared to the amyloid-positive (p < .001) and amyloid-negative (p < .001) groups. No
differences between amyloid-negative and -positive groups were detected for CBF (all p > .10).

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 25

Group comparisons of CVRi
MAP

Mean group CVRi
MAP
values are presented in Figure 2. In regions for which MANCOVA
omnibus tests were significant, post-hoc LSD tests identified a number of significant group
differences (see Table 5). Specifically, elevated CVRi
MAP
was found in AD compared to
amyloid-positive participants for the right IPC (p = .018), and left rMFG (p = .038) and
hippocampus (p = .017). CVRi
MAP
for the AD group was also higher than the amyloid-negative
group in the left mOFC (p = .006), rMFG (p < .001), hippocampus (p = .001) and ITC (p = .001),
as well as the right mOFC (p = .040), rMFG (p = .003), hippocampus (p = .001 ), ITC (p = .017),
and IPC (p < .001). Amyloid-positive participants exhibited elevated CVRi
MAP
relative to
amyloid-negative participants in the left mOFC (p = .050) and rMFG (p = .020), as well as the
right rMFG (p = .017), hippocampus (p = .026), ITC (p = .001), and IPC (p = .029).
Amyloid Status x CVRi
MAP
effects on cognitive performance
Analyses of CVRi
MAP
and amyloid (i.e. group) prediction of cognitive performance
yielded several significant interaction and main effects. Resulting statistics are presented in
Table 5. As depicted in Figure 3, differences in cognitive performance for amyloid-positive
versus negative participants varied as a product of regional CVRi
MAP
. More specifically,
recognition memory was worse for amyloid-positive versus negative participants only when
combined with elevated CVRi
MAP
in the left IPC. Recognition deficits in the amyloid-positive
relative to amyloid-negative group were also exacerbated when combined with elevated
CVRi
MAP
. Trails B performance was worse in amyloid-positive relative to amyloid-negative
participants only when concurrent with elevated CVRi
MAP
in the left IPC, right IPC, or right
mOFC.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 26

Longitudinal Analyses
Longitudinal CVRi
MAP
changes x amyloid status
Across amyloid groups, CVRi
MAP
increases between times 1 and 2 were detected in the
left mOFC (F (1,135) = 4.042, p = .046), as depicted in Figure 5.
Baseline CVRi
MAP
x amyloid status prediction of cognitive decline
i. Global cognition
For both amyloid groups, averaged across time points, higher baseline right rMFG
and right IPC CVRi
MAP
were associated with lower MMSE score. These associations
were more pronounced at later time points (i.e. more rapid declines for cases with higher
CVRi
MAP
between time 2 and 3, relative to between time 1 and 2), and stronger for
amyloid-positive versus amyloid-negative participants. Fixed effects parameter estimates
for the mixed model analyses are presented in Tables 7A and 7B. Longitudinal MMSE
values for median split (i.e. high and low) CVRi
MAP
in amyloid-positive and -negative
groups are presented in Figure 5.
ii. Executive function
Higher baseline mOFC CVRi
MAP
was associated with longer time taken to
complete Trails A, across time points. Higher baseline right ITC CVRi
MAP
was associated
with accelerated declines in Trails A performance over time, with more marked
acceleration seen for amyloid-positive versus amyloid-negative individuals.
Elevated baseline right hippocampal CVRi
MAP
predicted worse deterioration of
Trails B performance, with greater acceleration seen for the amyloid-positive compared
to the amyloid-negative group. This was also the case for baseline left IPC CVRi
MAP
,
with larger increases in Trails B time observed as CVRi
MAP
increased, and more

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 27

pronounced acceleration of decline seen for amyloid-positive relative to -negative
participants. Similarly, associations were detected for left mOFC CVRi
MAP
and Trails B
time. Fixed effects parameter estimates for the mixed model analyses are presented in
Tables 7C-G. Longitudinal Trails A and Trails B times for median split (i.e. high and
low) CVRi
MAP
in amyloid-positive and -negative groups are presented in Figures 6A and
6B, respectively.
iii. Memory
Across amyloid groups, RAVLT recognition performance decreased with time.
Such declines were worse in cases with higher right hippocampus CVRi
MAP,
with more
pronounced effects seen for amyloid-positive individuals. Fixed effects parameter
estimates for the mixed model analyses are presented in Table 7H. Longitudinal RAVLT
recognition scores for median split (i.e. high and low) CVRi
MAP
in amyloid-positive and -
negative groups are presented in Figure 7.
Prediction of cognitive performance by CVRi
MAP
change
i. Global cognition
In amyloid-positive but not -negative groups, larger late (i.e. between times 2 and
3) increases in left (ß = -.390, p = .010) and right (ß = -.284, p = .06) mOFC predicted
poorer MMSE performance at final assessment. In amyloid-positive but not amyloid-
negative participants, greater early (i.e. between times 1 and 2) increases in right ITC
CVRi
MAP
predicted higher MMSE scores at time 2 (ß = .221, p = .049).
ii. Executive function
For amyloid-positive but not -negative individuals, larger early increases in left
ITC CVRi
MAP
predicted longer Trails A time at time 2 (ß = .256, p = .048). Similarly, for

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 28

amyloid-positive participants only, greater late increases in left IPC CVRi
MAP
, predicted
worse Trails A performance at final assessment (ß = -.345, p = .052). Later increases in
left ITC (ß = .321, p = .022) and right ITC (ß = .265, p = .014) CVRi
MAP
, were associated
with larger increases in final assessment Trails A time for amyloid-positive relative to
amyloid-negative participants. Similarly, later increases in left IPC (ß = .242, p = .026)
and precuneus (ß = .243, p = .023), CVRi
MAP
were associated with greater increases in
final assessment Trails B time for amyloid-positive versus -negative participants. Later
increases in right hippocampus (ß = .238, p = .044), ITC (ß=.295, p = .004), and IPC (ß
=.290, p = .007) CVRi
MAP
, were associated with greater increases in final assessment
Trails B time for amyloid-positive relative to amyloid-negative participants.
iii. Memory
In amyloid-positive but not -negative participants, greater early increases in left
precuneus CVRi
MAP
were associated with fewer RAVLT words recalled at time 2 (ß = -
.276, p = .037), and final assessment (ß = -.349, p = .011). For the amyloid-negative but
not amyloid-positive group, larger later increases in left hippocampus CVRi
MAP
were
associated with poorer RAVLT recall.
For amyloid-positive but not amyloid-negative individuals, larger early increases
in right rMFG CVRi
MAP
predicted poorer time 2 RAVLT recognition (ß = -.294, p =
.022). Greater early increases in left rMFG CVRi
MAP
predicted worse time 2 RAVLT
recognition for amyloid-positive versus -negative participants (ß = .311, p = .008). Larger
early increases in right IPC CVRi
MAP
predicted poorer late RAVLT recognition (ß = .214,
p = .039) for the amyloid-positive relative to amyloid-negative group.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 29

Prediction of amyloid load by CVRi
MAP

Amyloid load at time 3 was predicted by time 2 CVRi
MAP
in the right rMFG (ß = .329,
t(25) = 2.575, p = .017). Amyloid load at time 2 was predicted by baseline CVRi
MAP
in the left
ITC (ß = .194, t (128) = 2.522, p = .013), and the right rMFG (ß = .159, t(137) = 2.119, p = .036)
and ITC (ß = .285, t(128) = 3.750, p < .001).
Prediction of amyloid load by CVRi
MAP
change
Higher final (i.e. time 3) amyloid load was predicted by larger early (i.e. between time 1
and 2) increases in right rMFG CVRi
MAP
(ß = .277, p = .044). Greater final amyloid load,
however, was predicted by smaller late changes in right rMFG CVRi
MAP
(ß = -.399, p = .016).
Prediction of amyloid changes by CVRi
MAP

Higher right mOFC at time 2 predicted later (i.e. between time 2 and 3) increases in
amyloid load (ß = .217, p = .026). Higher right (ß = -.485, p = .038) and left (ß = -.553, p = .006)
ITC CVRi
MAP
at time 2 predicted later decreases in amyloid load.
Attrition
Chi-square tests were performed to determine whether attrition rates differed among
amyloid status groups, APOE genotypes, and older versus younger (defined via age median split)
participants. Attrition from baseline to time 2 did not differ for any of these variables (all p >
.05). Attrition from time 2 to time 3 was proportionately higher in amyloid-positive relative to
amyloid-negative participants (χ
2
(1) = 4.331, p = .037), and older relative to younger participants
(χ
2
(1) = 6.858, p = .009), but did not differ among APOE genotypes (p > .05).
Discussion
Results of the present study suggest that elevations in cerebrovascular resistance occur
earlier and across a broader range of brain regions than previously identified CBF changes

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 30

(Mattson et al., 2014). This was reflected in CVRi
MAP
increases for prodromal (i.e. non-
demented amyloid-positive) relative to amyloid-negative older adults, over various cortical areas
associated with AD. In contrast, CBF decreases were detected only in cases already diagnosed
with dementia, and for fewer cortical regions. These findings link elevations in cerebrovascular
resistance to Alzheimer’s pathology (i.e. increased amyloid load), and suggest that changes in
cerebrovascular resistance may be more sensitive to early stages of AD than CBF. This is
consistent with prior reports of cerebrovascular resistance increases in cases identified as MCI, a
condition often equated to prodromal AD (Liu et al., 2014; Nation et al., 2013). Brain areas in
which CVRi
MAP
elevations were detected could therefore represent neural regions particularly
susceptible to arterial stiffening and other processes through which Alzheimer’s pathophysiology
arises (Weller et al., 2008).
Early increases in cerebrovascular resistance were also associated with later cognitive
decline, further implicating cerebrovascular dysfunction in clinical progression to AD.
Longitudinal analyses revealed synergistic effects of cerebrovascular resistance alongside
amyloidosis in erosion of cognitive faculties. Specifically, non-demented amyloid-positive cases
demonstrated minimal, if any, cognitive decline unless concurrent elevations in CVRi
MAP
were
also observed. Furthermore, in amyloid-positive individuals, increased CVRi
MAP
was associated
with accelerated cognitive decline. Notably, CVRi
MAP
effects were independent of fluctuations in
neuronal metabolism indexed by FDG-PET. CVRi
MAP
changes thus represent more than mere
products of blood flow abnormality driven by metabolic dysfunction or neurodegeneration.
Taken together, these data support hypothesized contributions of cerebrovascular dysfunction to
initiation and evolution of Alzheimer’s pathophysiology, including cerebral amyloidosis and
resulting cognitive decline.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 31

These findings are in line with models recognizing amyloid-β as a necessary but
insufficient element in the broader cascade through which clinical features of AD arise
(Drachman, 2014; Musiek & Holtzman, 2015). More specifically, our results suggest that
cerebrovascular stiffening and resulting disruption to CBF may be necessary for the development
of dementia in addition to, or in some cases regardless of, amyloid-β status. This is consistent
with vascular models such as the two-hit hypothesis (Iadecola, 2004), which posits that even
moderate declines in capillary perfusion can trigger neurodegeneration and APP-driven increases
in amyloid-β production, both of which contribute to cognitive decline characteristic of AD.
Cerebrovascular resistance and associated or parallel CBF irregularities may thus modify, or
operate in conjunction with, amyloid-β accumulation in progression to dementia (Zlokovic,
2011). This is consistent with our findings of cognitive repercussions for the interactive effects
of elevated cerebrovascular resistance and amyloid deposition, and further supported by
associations of early CVRi
MAP
with later amyloid load.
CVRi
MAP
exacerbation of amyloid-driven executive function deficits was particularly
pronounced for inferior parietal and frontal regions. This is neuroanataomically consistent with
prior reports of a frontoparietal basis for executive function performance (Moll, Oliveira-Souza,
Moll, Bramati, & Andreiuolo, 2002; Zakzanis, Mraz, & Graham, 2005). While the role of frontal
regions in executive function has been well established (Alvarez & Emory, 2006), more recent
research has linked Trails B performance to activation of parietal nodes within this broader
frontoparietal executive network (Seeley et al., 2007). Executive function deficits commonly
seen in preclinical and clinical AD may therefore reflect disruption of inferior parietal and frontal
cerebrovasculature in conjunction with regional amyloid deposition. Frontal CVRi
MAP

exacerbation of cognitive decline in prodromal cases (i.e. amyloid-positive individuals without

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 32

an AD diagnosis), paired with frontal CVRi
MAP
prediction of later amyloid load, implicates
anterior cerebrovascular dysfunction in early AD pathogenesis.
Frontal cerebrovascular abnormalities were also evident in longitudinal relationships
between CVRi
MAP
and global cognition. Both baseline and later increases in frontal
cerebrovascular resistance were predictive of future declines in global cognition for amyloid-
positive cases. While higher baseline CVRi
MAP
in these regions also predicted greater executive
function declines, these effects were not borne out in CVRi
MAP
change scores (i.e. increases in
frontal CVRi
MAP
did not significantly predict degree or rate of executive function decline). This
may reflect “peaking” of frontal CVRi
MAP
-driven declines in executive function earlier on, with
later erosion of executive function motivated by cerebrovascular dysfunction across a broader
and more posterior range of regions (as evidenced by prediction of later Trails performance
declines by later changes in inferior temporal and inferior parietal areas).
Given that our amyloid-positive participants were older than their amyloid-negative
counterparts, questions regarding later prognosis of amyloid-negative individuals were raised.
Specifically, would amyloid-negative cases “catch up” to the amyloid-positive individuals as
they aged, experiencing increases in amyloid load along with comparable vascular and cognitive
changes? Examination of amyloid negative cases 4 years (roughly the average age difference
between amyloid-negative and -positive groups) from baseline indicated only minor increases in
amyloid load (i.e. insufficient to render the majority amyloid-positive). Furthermore,
comparisons of amyloid-negative participants at time 3 and amyloid-positive participants at
baseline, indicated that higher CVRi
MAP
was only associated with poorer cognitive performance
in the latter. When original analyses were stratified by age (e.g. amyloid group comparisons for
only the “young old” or only the “older old” participants), differences in CVRi
MAP
between

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 33

amyloid-negative and -positive groups were more pronounced for older old than young old cases.
In addition, elevated CVRi
MAP
for these individuals was also associated with poorer cognitive
performance, generally irrespective of amyloid status. There is thus strong support for
cerebrovascular resistance as a crucial contributor to AD-related changes in amyloid load and
cognition, with more marked pathology evident in older age.
Across brain regions, higher baseline CVRi
MAP
and larger changes in CVRi
MAP
between
time points seemed to predict worse cognitive performance and often, faster declines. These
effects were most pronounced in individuals who were also amyloid-positive, suggesting
synergistic roles for cerebrovascular stiffening and cerebral amyloidosis in AD. Overall, our
results are therefore consistent with growing evidence for associations between cerebrovascular
variables and amyloid-β load. Increased cerebrovascular resistance and blood pressure (Liu et al.,
2014; Nation et al., 2013), and decreased CBF (Dumas et al., 2012; Gietl et al., 2015; Mattson et
al., 2014), for example, have been repeatedly identified in clinical and preclinical Alzheimer’s
cases. Similarly, CBF and amyloid-β imaging were previously found to be equivalent in
prediction of amyloid-β positivity (Tosun, Schuff, Jagust, & Weiner, 2016), and age-related CBF
declines appear to be present only where amyloid-β load is pathologically elevated (Gietl et al.,
2015). Moreover, disruption of CBF in APP knock-in mice produces both cerebral amyloid
angiopathy (CAA) and elevated parenchymal amyloid-β deposition (Li et al., 2014). Our results
extend these findings by identifying processes (i.e. cerebrovascular stiffening in key neural
regions) that operate in tandem with, and/or potentially precede, hypoperfusion associated with
cognitive decline.
Amyloid-β has shown vasoconstrictive and vasotoxic properties in animal and cell model
studies (Niwa et al., 2002; Sole, Minano-Molina, & Unzeta, 2014), which may underlie

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 34

cerebrovascular abnormalities. Cerebrovascular stiffening and associated cerebral hypoperfusion
could conversely exacerbate amyloidosis (Zlokovic, 2011). Our findings of elevated
cerebrovascular resistance, and to a lesser extent reduced CBF, in clinical and preclinical AD
implicate blood pressure changes in Alzheimer’s disease processes (e.g. amyloid-β deposition).
This is consistent with reports of reduced dementia-associated structural and cognitive declines
(Qiu et al., 2005), and attenuation of age-related cerebral amyloid-β deposition and dementia
(Nation, Ho, & Yew, 2015), in individuals taking antihypertensive medications. Further research
into the mechanisms responsible for cerebrovascular resistance increases in preclinical AD may
thus improve our understanding of AD pathogenesis but also reveal treatment targets.
While the present study’s large sample, assessment of cerebrovascular resistance
adjusting for brain glucose metabolism, and longitudinal nature represent valuable strengths, a
number of limitations must also be acknowledged. High attrition rates across follow-up visits
proved problematic (e.g. preventing employment of more sophisticated statistical techniques)
and may have reduced power to detect longitudinal effects. Furthermore, higher attrition rates
between time 2 and 3 were observed for amyloid-positive and older participants, potentially
limiting our understanding of later stage cerebrovascular changes in higher risk individuals. The
heterogeneous nature of the ADNI sample, which included participants from over 50 sites,
recruited through varied sources, studied at varying follow-up intervals, may also have limited
generalization of findings to community samples. In addition, selection criteria employed by
ADNI exclude individuals diagnosed with certain cardiovascular conditions, resulting in a
sample which likely under-represents vascular risk factors relevant to the models investigated in
the present study. Replication of study findings in a larger, more representative community-
based sample with more consistent and extensive follow-up may therefore be warranted.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 35

References
ADNI. (2011). PET Technical Procedures Manual: AV-45 (Florbetapir F18) & FDG. Retrieved
15th March, 2015, from http://adni.loni.usc.edu/wp-
content/uploads/2010/05/ADNI2_PET_Tech_Manual_0142011.pdf
Alsop, D. C., Casement, M., de Bazelaire, C., Fong, T., & Press, D. Z. (2008). Hippocampal
hyperperfusion in Alzheimer's disease. NeuroImage, 42(4), 1267-1274.
Alsop, D. C., Dai, W., Grossman, M., & Detre, J. A. (2010). Arterial Spin Labeling Blood Flow
MRI: Its Role in the Early Characterization of Alzheimer's Disease. Journal of
Alzheimer's Disease, 20(3), 871-880.
Alvarez, J. A., & Emory, E. (2006). Executive function and the frontal lobes: A meta-analytic
review. Neuropsychology Review, 16(1), 17-42.
Ballard, C., Gauthier, S., Corbett, A., Brayne, C., Aarsland, D., & Jones, E. (2011). Alzheimer's
disease. The Lancet, 377(9770), 1019-1031.
Bangen, K. J., Nation, D. A., Clark, L. R., Harmell, A. L., Wierenga, C. E., Dev, S. I., et al.
(2014). Interactive effects of vascular risk burden and advanced age on cerebral blood
flow. Frontiers in Aging Neuroscience, 6, 159.
Beason-Held, L. L., Moghekar, A., Zonderman, A. B., Kraut, M. A., & Resnick, S. M. (2007).
Longitudinal changes in cerebral blood flow in the older hypertensive brain. Stroke,
38(6), 1766-1773.
Bell, R. D., Deane, R., Chow, N., Long, X., Sagare, A., Singh, I., et al. (2009). SRF and
myocardin regulate LRP-mediated amyloid-beta clearance in brain vascular cells. Nature
Cell Biology, 11(2), 143.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 36

Bertram, L., & Tanzi, R. E. (2005). The genetic epidemiology of neurodegenerative disease.
Journal of Clinical Investigation, 115(6), 1449-1457.
Bookheimer, S. Y. P., Strojwas, M. H. B. S., Cohen, M. S. P., Saunders, A. M. P., Pericak-
Vance, M. A. P., Mazziotta, J. C. M. D. P., et al. (2000). Patterns of brain activation in
people at risk for Alzheimer's disease. The New England Journal of Medicine, 343(7),
450-456.
Buckner, R. L., Snyder, A. Z., Shannon, B. J., LaRossa, G., Sachs, R., Fotenos, A. F., et al.
(2005). Molecular, Structural, and Functional Characterization of Alzheimer's Disease:
Evidence for a Relationship between Default Activity, Amyloid, and Memory. The
Journal of Neuroscience, 25(34), 7709-7717.
Buxton, R. (2009). An introduction to functional magnetic resonance imaging: Principles and
techniques (2nd ed.). New York, NY: Cambridge University Press.
Buxton, R., Uludag, K., Dubowitz, D. J., & Liu, T. T. (2004). Modeling the hemodynamic
response to brain activation. NeuroImage, 23, Supplement 1, S220-S233.
Chao, L. L., Pa, J., Duarte, A., Schuff, N., Weiner, M. W., Kramer, J. H., et al. (2009). Patterns
of cerebral hypoperfusion in amnestic and dysexecutive MCI. Alzheimer's Disease and
Associated Disorders, 23, 245-252.
Chen, J. J., Rosas, H. D., & Salat, D. H. (2011). Age-associated reductions in cerebral blood flow
are independent from regional atrophy. NeuroImage, 55(2), 468-478.
Chow, N., Bell, R. D., Deane, R., Streb, J. W., Chen, J., Brooks, A., et al. (2007). Serum
response factor and myocardin mediate arterial hypercontractility and cerebral blood flow
dysregulation in Alzheimer's phenotype. Proceedings of the National Academy of
Sciences of the United States of America, 104(3), 823-828.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 37

Dai, W. Y., Lopez, O. L., Carmichael, O. T., Becker, J. T., Kuller, L. H., & Gachm, H. M.
(2009). Mild cognitive impairment and Alzheimer disease: Patterns of altered cerebral
blood flow at MR imaging. Radiology, 250, 856-866.
Devanand, D. P., Pradhaban, G., Liu, X., Khandji, A., De Santi, S., Segal, S., et al. (2007).
Hippocampal and entorhinal atrophy in mild cognitive impairment: Prediction of
Alzheimer disease. Neurology, 68, 828-836.
Drachman, D. A. (2014). The amyloid hypothesis, time to move on: Amyloid is the downstream
result, not cause, of Alzheimer's disease. Alzheimer's & Dementia: The Journal of the
Alzheimer's Association, 10(3), 372-380.
Dumas, A., Dierksen, G. A., Gurol, M. E., Halpin, A., Martinez‚ÄêRamirez, S., Schwab, K., et
al. (2012). Functional magnetic resonance imaging detection of vascular reactivity in
cerebral amyloid angiopathy. Annals of Neurology, 72(1), 76-81.
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Mini-mental state: A practical method
for grading the cognitive state of patients for the clinician. Journal of Psychiatric
Research, 12, 189-198.
Frisoni, G. B., Fox, N. C., Jack, C. R., Scheltens, P., & Thompson, P. M. (2010). The clinical use
of structural MRI in Alzheimer disease. Nature Reviews Neurology, 6, 67-77.
Gianaros, P. J., Greer, P. J., Ryan, C. M., & Jennings, J. R. (2006). Higher blood pressure
predicts lower regional grey matter volume: Consequences on short-term information
processing. NeuroImage, 31(2), 754-765.
Gietl, A. F., Warnock, G., Riese, F., K√§lin, A. M., Saake, A., Gruber, E., et al. (2015). Regional
cerebral blood flow estimated by early PiB uptake is reduced in mild cognitive

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 38

impairment and associated with age in an amyloid-dependent manner. Neurobiology of
Aging, 36(4), 1619-1628.
Greicius, M. D., Srivastava, G., Reiss, A. L., & Menon, V. (2004). Default-mode network
activity distinguishes Alzheimer's disease from healthy aging: Evidence from functional
MRI. Proceedings of the National Academy of Sciences of the United States of America,
101(13), 4637-4642.
Guglielmotto, M., Aragno, M., Autelli, R., Giliberto, L., Novo, E., Colombatto, S., et al. (2009).
The up-regulation of BACE1 mediated by hypoxia and ischemic injury: role of oxidative
stress and HIF1α. Journal of Neurochemistry, 108(4), 1045-1056.
Han, B. H., Zhou, M.-l., Johnson, A. W., Singh, I., Liao, F., Vellimana, A. K., et al. (2015).
Contribution of reactive oxygen species to cerebral amyloid angiopathy, vasomotor
dysfunction, and microhemorrhage in aged Tg2576 mice. [10.1073/pnas.1414930112].
Proceedings of the National Academy of Sciences, 112(8), E881-E890.
Herbert, L. S., Scherr, P. A., Bienias, J. L., Bennett, D. A., & Evans, D. A. (2003). Alzheimer's
disease in the U.S. population: Prevalence estimates using the 2000 census. Archives of
Neurology, 60, 1119-1122.
Iadecola, C. (2004). Neurovascular regulation in the normal brain and in Alzheimer's disease.
[10.1038/nrn1387]. Nat Rev Neurosci, 5(5), 347-360.
Ittner, L. M., & Götz, J. (2011). Amyloid-β and tau - a toxic pas de deux in Alzheimer's disease.
Nature Reviews Neuroscience, 12(2), 67-72.
Johnson, N. A., Jahng, G.-H., Weiner, M. W., Miller, B. L., Chui, H. C., Jagust, W. J., et al.
(2005). Pattern of cerebral hypoperfusion in Alzheimer disease and Mild Cognitive

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 39

Impairment measured with arterial spin-labeling MR imaging: Initial experience.
Radiology, 234(3), 851-859.
Joshi, A. D., Pontecorvo, M., Clark, C. M., Carpenter, A. P., Jennings, D. L., Sadowsky, C. H., et
al. (2012). Performance characteristics of amyloid PET and florbetapir F 18 in patients
with Alzheimer's disease and cognitively normal subjects. The Journal of Nuclear
Medicine, 53(3), 378-384.
Kress, B. T., Iliff, J. J., Xia, M., Wang, M., Wei, H. S., Zeppenfeld, D., et al. (2014). Impairment
of paravascular clearance pathways in the aging brain. Annals of Neurology, 76, 845-861.
Landau, S. M., Harvey, D., Madison, C. M., Koeppe, R. A., Reiman, E. M., Foster, N. L., et al.
(2011). Associations between cognitive, functional, and FDG-PET measures of decline in
AD and MCI. Neurobiology of Aging, 32(7), 1207-1218.
Langbaum, J. B. S., Chen, K., Launer, L. J., Fleisher, A. S., Lee, W., Liu, X., et al. (2012). Blood
pressure is associated with higher brain amyloid burden and lower glucose metabolism in
healthy late middle-age persons. Neurobiology of Aging, 33(4), 827.e811-827.e819.
Li, H., Guo, Q., Inoue, T., Polito, V. A., Tabuchi, K., Hammer, R. E., et al. (2014). Vascular and
parenchymal amyloid pathology in an Alzheimer disease knock-in mouse model:
Interplay with cerebral blood flow. Molecular Neurodegeneration, 9.
Liu, J., Zhu, Y.-S., Khan, M. A., Brunk, E., Martin-Cook, K., Weiner, M. F., et al. (2014).
Global brain hypoperfusion and oxygenation in amnestic mild cognitive impairment.
Alzheimer's & Dementia, 10(2), 162-170.
Mattson, R. H., Tosun, D., Insel, P. S., Simonson, A., Jack, C. R., Jr., Beckett, L. A., et al.
(2014). Association of brain amyloid-ß with cerebral perfusion and structure in
Alzheimer's disease and mild cognitive impairment. Brain, 137, 1550-1561.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 40

McDonald, C. R., McEvoy, L. K., Gharapetian, L., Fennema-Notestine, C., Hagler, D. J., Jr.,
Holland, D., et al. (2009). Regional rates of neocortical atrophy from normal aging to
early Alzheimer disease. Neurology, 73, 457-465.
Meel-van den Abeelen, A. S. S., Lagro, J., van Beek, A. H. E. A., & Claassen, J. A. H. R. (2014).
Impaired cerebral autoregulation and vasomotor reactiity in sporadic Alzheimer's disease.
Current Alzheimer Research, 11, 11-17.
Mies, G., Ishimaru, S., Xie, Y., Seo, K., & Hossmann, K. A. (1991). Ischemic Thresholds of
Cerebral Protein Synthesis and Energy State Following Middle Cerebral Artery
Occlusion in Rat. J Cereb Blood Flow Metab, 11(5), 753-761.
Moll, J., Oliveira-Souza, R. d., Moll, F. T., Bramati, I. E., & Andreiuolo, P. A. (2002). The
cerebral correlates of set-shifting: an fMRI study of the trail making test. Arquivos de
Neuro-Psiquiatria, 60, 900-905.
Morris, J. C. (1993). The Clinical Dementia Rating (CDR): Current version and scoring rules.
Neurology, 43, 2412-2414.
Musiek, E. S., & Holtzman, D. M. (2015). Three dimensions of the amyloid hypothesis: Time,
space and 'wingmen'. Nature Neuroscience, 18, 800-806.
Nation, D. A., Ho, J., & Yew, B. (2015). Older adults taking AT1-receptor blockers exhibit
reduced cerebral amyloid retention. Journal of Alzheimer's Disease, 50(3).
Nation, D. A., Wierenga, C. E., Clark, L. R., Dev, S. I., Stricker, N. H., Jak, A. J., et al. (2013).
Cortical and subcortical cerebrovascular resistance index in Mild Cognitive Impairment
and Alzheimer's disease. Journal of Alzheimer's Disease, 36, 689-698.
Nichols, W. M., O' Rourke, M. F., & Vlachopoulos, C. (2011). McDonald's Blood Flow in
Arteries. Boca Raton, FL: CRC Press.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 41

Niwa, K., Kazama, K., Younkin, L., Younkin, S. G., Carlson, G. A., & Iadecola, C. (2002).
Cerebrovascular autoregulation is profoundly impaired in mice overexpressing amyloid
precursor protein. American Journal of Physiology - Heart and Circulatory Physiology,
283, H315-H323.
Paulson, O. B., Strandgaard, S., & Edvinsson, L. (1990). Cerebral autoregulation.
Cerebrovascular and Brain Metabolism Reviews, 2, 161-192.
Peña-Casanova, J., Sánchez-Benavides, G., de Sola, S., Manero-Borrás, R. M., & Casals-Coll,
M. (2012). Neuropsychology of Alzheimer's disease. Archives of Medical Research, 43,
686-693.
Petrovitch, H., White, L. R., Izmirilian, G., Ross, G. W., Havlik, R. J., Markesbery, W., et al.
(2000). Midlife blood pressure and neuritic plaques, neurofibrillary tangles, and brain
weight at death: The HAAS. Neurobiology of Aging, 21(1), 57-62.
Qiu, C., Winblad, B., & Fratiglioni, L. (2005). The age-dependent relation of blood pressure to
cognitive function and dementia. The Lancet Neurology, 4(8), 487-499.
Raz, N., & Rodrigue, K. M. (2006). Differential aging of the brain: Patterns, cognitive correlates
and modifiers. Neuroscience & Biobehavioral Reviews, 30(6), 730-748.
Rey, A. (1941). L'examen psychologique dans les cas d'encéphalopathie traumatique. Archives
de Psychologie, 28, 215-285.
Seeley, W. W., Menon, V., Schatzberg, A. F., Keller, J., Glover, G. H., Kenna, H., et al. (2007).
Dissociable intrinsic connectivity networks for saliene processing and executive control.
Journal of Neuroscience 27(9), 2349-2356.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 42

Sheline, Y. I., Morris, J. C., Snyder, A. Z., Price, J. L., Yan, Z., D'Angelo, G., et al. (2010).
APOE4 allele disrupts resting state fMRI connectivity in the absence of amyloid plaques
or decreased CSF Aβ42. Journal of Neuroscience, 30(50), 17035-17040.
Small, G. W., Bookheimer, S. Y., Thompson, P. M., Cole, G. M., Huang, S. C., Kepe, V., et al.
(2008). Current and future uses of neuroimaging for cognitively impaired patients. The
Lancet Neurology, 7(2), 161-172.
Sole, M., Minano-Molina, A. J., & Unzeta, M. (2014). A cross-talk between Abeta and
endothelial SSAO/VAP-1 accelerates vascular damage and Abeta aggregation related to
CAA-AD. Neurobiology of Aging.
Spreen, O., & Strauss, E. (1998). A compendium of neuropsychological tests (2nd ed.). New
York: Oxford University Press.
Strassburger, T. L., Lee, H.-C., Daly, E. M., Szczepanik, J., Krasuski, J. S., Mentis, M. J., et al.
(1997). Interactive effects of age and hypertension on volumes of brain structures. Stroke,
28(7), 1410-1417.
Tosun, D., Schuff, N., Jagust, W., & Weiner, M. W. (2016). Discriminative Power of Arterial
Spin Labeling Magnetic Resonance Imaging and 18F-Fluorodeoxyglucose Positron
Emission Tomography Changes for Amyloid-β-Positive Subjects in the Alzheimer's
Disease Continuum. Neurodegenerative Diseases, 16, 87-94.
Vemuri, P., Wiste, H. J., Weigand, S. D., Shaw, L. M., Trojanowski, J. Q., Weiner, M. W., et al.
(2009). MRI and CSF biomarkers in normal, MCI, and AD subjects: Predicting future
clinical change. Neurology, 73(4), 294-301.
Vlassenko, A. G., Vaishnavi, N., Couture, L., Sacco, D., Shannon, B. J., Mach, R. H., et al.
(2010). Spatial correlation between brain aerobic glycolysis and amyloid-ß (Aß)

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 43

deposition. Proceedings of the National Academy of Sciences of the United States of
America, 107, 17763-17767.
Wang, Z. (2014). Characterizing early Alzheimer's disease and disease progression using
hippocampal volume and arterial spin labeling perfusion MRI. Journal of Alzheimer's
Disease, 42, S495-S502.
Wang, Z., Yang, D., Zhang, X., Li, T., Li, J., Tang, Y., et al. (2011). Hypoxia-Induced Down-
Regulation of Neprilysin by Histone Modification in Mouse Primary Cortical and
Hippocampal Neurons. PLoS ONE, 6(4), e19229.
Weller, R. O., Boche, D., & Nicoll, J. A. R. (2009). Microvasculature changes and cerebral
amyloid angiopathy in Alzheimer's disease and their potential impact on therapy. Acta
Neuropathologica, 118, 87-102.
Weller, R. O., Subash, M., Preston, S. D., Mazanti, I., & Carare, R. O. (2008). Perivascular
drainage of amyloid-ß peptides from the brain and its failure in cerebral amyloid
angiopathy and Alzheimer's disease. Brain Pathology, 18, 253-266.
Wiseman, R. M., Saxby, B. K., Burton, E. J., Barber, R., Ford, G. A., & O'Brien, J. T. (2004).
Hippocampal atrophy, whole brain volume, and white matter lesions in older
hypertensive subjects. Neurology, 63, 1892-1897.
Wu, Z., Guo, H., Chow, N., Sallstrom, J., Bell, R. D., Deane, R., et al. (2005). Role of the
MEOX2 homeobox gene in neurovascular dysfunction in Alzheimer disease.
[10.1038/nm1287]. Nat Med, 11(9), 959-965.
Zakzanis, K. K., Mraz, R., & Graham, S. J. (2005). An fMRI study of the Trail Making Test.
Neuropsychologia, 43(13), 1878-1886.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 44

Zlokovic, B. V. (2011). Neurovascular pathways to neurodegeneration in Alzheimer's disease
and other disorders. Nature Reviews Neuroscience, 12(12), 723-738.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 45

Tables
Table 1.
Mean (SD) baseline demographics for included participants

Amyloid-
negative
Amyloid-
positive
AD Total F or χ
2
p-value
N 112 87 33 232

Age 69.22 (6.58) 73.73 (6.81) 73.16 (6.68) 71.47 (7.00) 12.38 <.001
Sex (M/F) 53/59 47/40 18/15 118/114 1.09 .58
Education 16.74 (2.37) 16.22 (3.13) 16.42 (2.39) 16.50 (2.68) .95 .39
BMI 28.01 (4.83) 26.67 (4.15) 25.15 (4.00) 26.90 (4.43) 3.08 .05
Proportion
APOE4 carriers
(none/1/2)
#

89/20/3 34/44/9 9/13/11 132/77/23 60.40 .00
#
percentage of participants with no, 1, or 2 ApoE4 alleles

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 46

Table 2.
Group sample sizes for longitudinal analyses
Time 1 Time 2 Time 3
Amyloid-negative 112 85 65
Amyloid-positive 87 61 41
AD 33 18 15
Total 232 164 121

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 47

Table 3.
Mean (SE) group values for physiological variables

Amyloid-
negative
Amyloid-
positive
AD F p-value
SBP 130.00 (1.77) 133.29 (1.90) 133.58 (3.15) .805 .45
DBP 73.22 (1.02) 75.24 (1.09) 73.97 (1.81) .85 .43
MAP 92.15 (1.11) 94.59 (1.20) 93.84 (1.98) .99 .37
PP 56.78 (1.49) 58.06 (1.60) 59.61 (2.65) .40 .67

Note. SBP = systolic blood pressure; DBP = diastolic blood pressure; MAP = mean arterial
pressure; PP = pulse pressure. Reported statistics reflect values controlling for age, sex, BMI,
and APOE genotype.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 48

Table 4.
Baseline group differences in regional cerebral blood flow (rCBF)
df F p-value ƞ
P
2

Left
mOFC 2, 224 1.587 .207 .015
rMFG 2, 224 2.566 .079 .025
Hippocampus 2, 224 3.198 .043 .030
ITC 2, 224 3.269 .040 .031
IPC 2, 224 2.641 .074 .025
Precuneus
2, 224
2.412 .092 .023
Right
mOFC 2, 224 1.303 .274 .013
rMFG 2, 224 1.619 .201 .005
Hippocampus 2, 224 2.559 .080 .024
ITC 2, 224 2.062 .130 .020
IPC 2, 224 8.026 < .001 .073
Precuneus
2, 224
1.297 .276 .013

Results of MANCOVA comparing regional cerebral blood flow in Alzheimer’s disease, non-
demented amyloid-positive, and amyloid-negative groups. Age, APOE genotype, body mass
index, sex, and composite FDG PET values were included as covariates.
Note. mOFC = medial orbitofrontal cortex; rMFG = rostral middle frontal gyrus; ITC = inferior
temporal cortex; IPC = inferior parietal cortex.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 49

Table 5.
Baseline group differences in regional CVRi
MAP

df F p-value ƞ
P
2

Left
mOFC
2, 224
4.156 .017 .040
rMFG
2, 224
6.796 .001 .065
Hippocampus
2, 224
5.941 .003 .057
ITC
2, 224
7.108 .001 .067
IPC
2, 224
1.931 .148 .019
Precuneus
2, 224
2.527 .082 .025
Right

mOFC
2, 224
2.346 .098 .023
rMFG
2, 224
5.320 .006 .051
Hippocampus
2, 224
5.860 .003 .056
ITC
2, 224
6.115 .003 .058
IPC
2, 224
7.319 .001 .069
Precuneus
2, 224
1.908 .151 .019

Results of MANCOVA comparing regional cerebrovascular resistance index (CVRi
MAP
) in
Alzheimer’s disease, non-demented amyloid-positive, and amyloid-negative groups at baseline.
Age, APOE genotype, body mass index, sex, and composite FDG PET values were included as
covariates.
Note. mOFC = medial orbitofrontal cortex; rMFG = rostral middle frontal gyrus; ITC = inferior
temporal cortex; IPC = inferior parietal cortex.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 50

Table 6.
Baseline CVRi
MAP
x amyloid effects on cognitive performance

Results of MANCOVA comparing non-demented amyloid-positive, and amyloid-negative
groups on regional CVRi
MAP
prediction of cognitive performance at baseline. Age, APOE
genotype, body mass index (BMI), sex, composite FDG PET values, and education were
included as covariates.
Note. CVRi
MAP
= cerebrovascular resistance index derived from mean arterial pressure; mOFC =
medial orbitofrontal cortex; rMFG = rostral middle frontal gyrus; ITC = inferior temporal cortex;
IPC = inferior parietal cortex.

Cognitive variable CVRi
MAP
region df F p-value ƞ
P
2

L ITC 1, 191 5.401 .021 .033
L IPC 1, 191 7.936 .005 .044
Recognition memory
(RAVLT recognition)
R precuneus 1, 191 5.675 .018 .032
Delayed recall
(RAVLT delayed recall)
L rMFG 1, 191 4.406 .037 .025
L IPC 1, 191 11.205 .001 .061
R mOFC 1, 191 5.484 .020 .031
Executive function
(Trails B time)
R IPC 1, 191 8.387 .004 .047

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 51

Table 7.
A. Baseline right rMFG CVRi
MAP
x amyloid prediction of longitudinal MMSE performance
Parameter Estimate SE df t p-value
Intercept
29.297 0.506 421.715 57.931 0.000
Age
0.000 0.018 396.678 0.005 0.996
Education
0.112 0.045 406.899 2.476 0.014
BMI
0.017 0.028 389.655 0.604 0.546
Sex
0.287 0.240 400.171 1.195 0.233
APOE4
0.066 0.222 391.060 0.297 0.767
FDG-PET
0.422 0.191 415.269 2.210 0.028
Amyloid (0= Aβ+; 1=Aβ-)
-0.679 0.411 364.726 -1.651 0.100
Right precentral CBF
-0.003 0.013 397.416 -0.261 0.794
Right rMF CVRi
-0.032 0.161 306.789 -0.200 0.842
Aβ- * Right rMF CVRi
-0.006 0.221 326.101 -0.026 0.979
Time
-1.113 0.206 377.237 -5.399 0.000
Right rMF CVRi * Time
0.507 0.185 186.368 2.744 0.007
Aβ- * Right rMG CVRi *
Time -0.471 0.219 225.961 -2.155 0.032
Aβ- * time
1.027 0.264 364.915 3.888 0.000

B. Baseline right IPC CVRi
MAP
x amyloid prediction of longitudinal MMSE performance
Parameter Estimate SE df t p-value
Intercept
29.502 0.533 428.723 55.392 0.000
Age
0.014 0.020 387.676 0.739 0.461
Education
0.107 0.048 389.775 2.243 0.025
BMI
0.011 0.031 383.640 0.369 0.712
Sex
0.190 0.257 391.502 0.739 0.460
APOE4
-0.024 0.235 398.558 -0.104 0.918
FDG-PET
0.554 0.215 399.042 2.572 0.010
Amyloid (0= Aβ+; 1=Aβ-)
-0.717 0.434 358.892 -1.652 0.099
R precentral CBF
0.001 0.014 399.317 0.070 0.944
R IPC CVRi
0.834 0.489 344.936 1.708 0.089
Aβ- * R rMF CVRi
-0.817 0.556 345.687 -1.469 0.143
Time
-1.475 0.216 380.765 -6.830 0.000
R IPC CVRi * Time
-1.013 0.393 355.340 -2.579 0.010
Aβ- * R IPC CVRi * Time
1.100 0.440 349.646 2.498 0.013
Aβ- * time
1.403 0.281 359.890 4.990 0.000

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 52

C. Baseline right ITC CVRi
MAP
x amyloid prediction of longitudinal Trails A performance
Parameter Estimate SE df t p-value
Intercept
33.254 3.285 380.301 10.123 0.000
Age
0.436 0.113 250.494 3.847 0.000
Education
0.309 0.283 299.048 1.093 0.275
BMI
0.146 0.169 254.958 0.865 0.388
Sex
-0.477 1.413 238.965 -0.338 0.736
APOE4
0.799 1.332 272.739 0.600 0.549
FDG-PET
-4.370 1.216 287.426 -3.594 0.000
Amyloid (0= Aβ+; 1=Aβ-)
2.523 3.367 285.757 0.749 0.454
R precentral CBF
-0.116 0.093 389.322 -1.252 0.211
R ITC CVRi
-0.252 1.200 257.264 -0.210 0.834
Aβ- * R ITC CVRi
2.072 2.045 279.194 1.014 0.312
Time
2.282 1.352 319.504 1.689 0.092
R ITC CVRi * Time
2.290 0.957 155.142 2.393 0.018
Aβ- * R ITC CVRi * Time
-3.265 1.360 233.413 -2.402 0.017
Aβ- * time
-4.332 1.926 364.930 -2.249 0.025

D. Baseline left mOFC CVRi
MAP
x amyloid prediction of longitudinal Trails A performance
Parameter Estimate SE df t p-value
Intercept
33.123 3.200 412.347 10.352 0.000
Age
0.473 0.110 326.628 4.283 0.000
Education
0.515 0.267 348.584 1.927 0.055
BMI
0.141 0.170 300.417 0.833 0.405
Sex
0.240 1.407 334.007 0.170 0.865
APOE4
0.208 1.316 338.110 0.158 0.874
FDG-PET
-5.004 1.174 350.735 -4.263 0.000
Amyloid (0= Aβ+; 1=Aβ-)
-0.412 2.870 334.046 -0.144 0.886
L precentral CBF
-0.051 0.081 403.296 -0.631 0.529
L mOFC CVRi
3.836 1.782 347.412 2.152 0.032
Aβ- * R rMF CVRi
-3.197 2.281 339.018 -1.402 0.162
Time
1.987 1.295 390.886 1.534 0.126
L mOFC CVRi * Time
-1.343 1.324 244.567 -1.014 0.312
Aβ- * L mOFC CVRi *
Time 1.155 1.582 279.431 0.730 0.466
Aβ- * time
-2.722 1.653 396.347 -1.646 0.101

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 53

E. Baseline left mOFC CVRi
MAP
x amyloid prediction of longitudinal Trails B performance
Parameter Estimate SE df t p-value
Intercept
70.141 10.986 382.403 6.385 0.000
Age
2.016 0.415 194.826 4.854 0.000
Education
0.362 0.987 244.196 0.366 0.714
BMI
0.736 0.655 237.033 1.124 0.262
Sex
11.227 5.289 220.738 2.122 0.035
APOE4
7.021 4.873 206.556 1.441 0.151
FDG-PET
-24.396 4.382 238.690 -5.567 0.000
Amyloid (0= Aβ+; 1=Aβ-)
0.854 9.305 209.394 0.092 0.927
L precentral CBF
0.242 0.289 319.722 0.839 0.402
L mOFC CVRi
-0.403 5.592 193.568 -0.072 0.943
Aβ- * R rMF CVRi
3.801 7.268 183.513 0.523 0.602
Time
9.873 4.578 368.922 2.156 0.032
L mOFC CVRi * Time
8.322 4.370 248.100 1.904 0.058
Aβ- * L mOFC CVRi *
Time -12.312 5.558 228.722 -2.215 0.028
Aβ- * time
-14.706 5.893 368.151 -2.496 0.013

F. Baseline right hippocampus CVRi
MAP
x amyloid prediction of longitudinal Trails B
performance
Parameter Estimate SE df t p-value
Intercept
77.628 10.991 297.653 7.063 0.000
Age
1.648 0.401 135.928 4.106 0.000
Education
0.083 1.015 222.445 0.082 0.935
BMI
0.376 0.643 192.677 0.585 0.559
Sex
9.533 5.195 143.844 1.835 0.069
APOE4
3.048 5.022 159.489 0.607 0.545
FDG-PET
-26.809 4.227 156.827 -6.342 0.000
Amyloid (0= Aβ+; 1=Aβ-)
-5.985 9.169 207.903 -0.653 0.515
R precentral CBF
0.016 0.297 297.098 0.054 0.957
R hippocampus CVRi
-2.428 6.435 203.161 -0.377 0.706
Aβ- * R hippocampus CVRi
2.128 8.435 188.291 0.252 0.801
Time
8.399 4.540 344.004 1.850 0.065
R hippocampus CVRi *
Time 13.208 5.123 238.346 2.578 0.011
Aβ- * R hippocampus CVRi
* Time -15.005 6.529 199.980 -2.298 0.023
Aβ- * time
-11.303 5.854 322.735 -1.931 0.054

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 54

G. Baseline left IPC CVRi
MAP
x amyloid prediction of longitudinal Trails B performance
Parameter Estimate SE df t p-value
Intercept
70.008 10.329 327.130 6.778 0.000
Age
1.684 0.384 157.060 4.382 0.000
Education
-0.005 0.939 229.162 -0.005 0.996
BMI
0.175 0.632 215.313 0.278 0.782
Sex
8.610 5.020 188.754 1.715 0.088
APOE4
5.744 4.706 221.278 1.220 0.224
FDG-PET
-20.206 4.411 209.609 -4.581 0.000
Amyloid (0= Aβ+; 1=Aβ-)
2.484 8.595 219.200 0.289 0.773
L precentral CBF
0.298 0.282 289.123 1.054 0.293
L IPC CVRi
1.421 6.074 225.032 0.234 0.815
Aβ- * L IPC CVRi
2.040 7.201 212.139 0.283 0.777
Time
15.492 4.501 351.169 3.442 0.001
L IPC CVRi * Time
13.204 4.775 324.808 2.765 0.006
Aβ- * L IPC CVRi * Time
-16.963 5.778 286.206 -2.936 0.004
Aβ- * time
-19.287 5.850 371.333 -3.297 0.001

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 55

H. Baseline right hippocampus CVRi
MAP
x amyloid prediction of longitudinal RAVLT
recognition performance

Parameter Estimate SE df t p-value
Intercept
11.215 0.714 409.147 15.696 0.000
Age
-0.018 0.026 296.431 -0.684 0.495
Education
0.040 0.063 318.174 0.642 0.521
BMI
-0.003 0.040 293.425 -0.065 0.948
Sex
0.503 0.331 296.516 1.521 0.129
APOE4
-0.177 0.308 302.249 -0.574 0.566
FDG-PET
1.216 0.274 314.282 4.444 0.000
Amyloid (0= Aβ+; 1=Aβ-)
0.900 0.614 319.128 1.465 0.144
R precentral CBF
0.009 0.020 390.417 0.473 0.636
R hippocampus CVRi
0.715 0.432 348.255 1.657 0.098
Aβ- * R hippocampus CVRi
-0.860 0.569 325.506 -1.512 0.131
Time
-0.584 0.276 411.249 -2.120 0.035
R hippocampus CVRi *
Time -0.525 0.311 323.385 -1.687 0.093
Aβ- * R hippocampus CVRi
* Time 0.850 0.404 322.790 2.106 0.036
Aβ- * time
0.627 0.355 405.647 1.763 0.079

Fixed effects estimates for linear mixed models examining baseline CVRi
MAP
and amyloid status
prediction of cognitive performance across three time points. Time, regional CVRi
MAP
, amyloid
group, CVRi
MAP
x amyloid, age, APOE genotype, BMI, sex, FDG PET values, and education
were entered as fixed effects. Time, time x amyloid, time CVRi
MAP
, and time x CVRi
MAP
x
amyloid were entered as random effects with an autoregressive covariance structure.
Note. CVRi = cerebrovascular resistance index; mOFC = medial orbitofrontal cortex; rMFG =
rostral middle frontal gyrus; ITC = inferior temporal cortex; IPC = inferior parietal cortex.
Modeled CVRi values are residuals from regression of original CVRi values on precentral CBF.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 56

Figures
Figure 1.
Regional cerebral blood flow (rCBF) for amyloid-negative, amyloid-positive, and AD groups.

Participants underwent arterial spin labeling (ASL) functional magnetic resonance imaging
(fMRI) to measure blood flow in regions typically affected by Alzheimer’s disease.
Note. Error bars represent standard error. L = left; R= right; mOFC = medial orbitofrontal cortex;
rMFG = rostral middle frontal gyrus; hipp = hippocampus; ITC = inferior temporal cortex; IPC =
inferior parietal cortex. * p < .05; ** p < .01.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 57

Figure 2.
Regional CVRi
MAP
for amyloid-negative, amyloid-positive, and AD groups.

Cerebrovascular resistance was estimated by dividing baseline mean arterial pressure (MAP) by
regional cerebral blood flow (rCBF) for each region of interest. Note. *p < .05; **p < .01

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 58

Figure 3.
CVRi
MAP
x amyloid status interaction effects on cognitive performance

Cerebrovascular resistance x amyloid effects on time taken to complete Trails B. Analyses were
performed using continuous CVRi
MAP
variables upon which median splits (yielding high and low
CVRi
MAP
groupings) were later performed for graphing purposes.

10
11
12
13
14
15
AB- AB+
RA VLT recognition
L IPC CVRi x Aβ
Low CVRi High CVRi
40
60
80
100
120
140
AB- AB+
Trails B time (seconds)
R mOFC CVRi x Aβ
40
60
80
100
120
140
AB- AB+
Trails B time (seconds)
L IPC CVRi x Aβ
40
60
80
100
120
140
AB- AB+
Trails B time (seconds)
R IPC CVRi x Aβ

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 59

Figure 4.
Longitudinal CVRi
MAP
for amyloid groups

Left mOFC CVRi
MAP
at baseline and time 2, for amyloid negative and amyloid positive
participants.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 60

Figure 5.
Baseline CVRi
MAP
x amyloid prediction of longitudinal global cognition

Cerebrovascular resistance x amyloid effects on time taken to complete Trails B. Analyses were
performed using continuous CVRi
MAP
variables upon which median splits (yielding high and low
CVRi
MAP
groupings) were later performed for graphing purposes.

22
23
24
25
26
27
28
29
30
Time 1 Time 2 Time 3
MMSE score
Aβ+ Right IPC CVRi
Low CVRi High CVRi
22
23
24
25
26
27
28
29
30
Time 1 Time 2 Time 3
MMSE score
Aβ+ Right rMFG CVRi
22
23
24
25
26
27
28
29
30
Time 1 Time 2 Time 3
MMSE score
Aβ- Right rMFG CVRi
22
23
24
25
26
27
28
29
30
Time 1 Time 2 Time 3
MMSE score
Aβ- Right IPC CVRi

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 61

Figure 6.
Baseline CVRi
MAP
x amyloid prediction of longitudinal executive function

A. Trail Making Test - Part A

20
30
40
50
60
Time 1 Time 2 Time 3
Trails A time (seconds)
Aβ+ Left mOFC CVRi
Low CVRi High CVRi
20
30
40
50
60
Time 1 Time 2 Time 3
Trails A time (seconds)
Aβ+ Right ITC CVRi
20
30
40
50
60
Time 1 Time 2 Time 3
Trails A time (seconds)
Aβ- Right ITC CVRi
20
30
40
50
60
Time 1 Time 2 Time 3
Trails A time (seconds)
Aβ- Left mOFC CVRi

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 62

B. Trail Making Test - Part B
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ+ Left IPC CVRi
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ+ Right Hippocampus CVRi
Low CVRi High CVRi
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ- Right Hippocampus CVRi
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ- Left IPC CVRi
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ- Left mOFC CVRi
50
70
90
110
130
150
Time 1 Time 2 Time 3
Trails B time (seconds)
Aβ+ Left mOFC CVRi

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 63

Cerebrovascular resistance x amyloid effects on time taken to complete Trails A and Trails B.
Analyses were performed using continuous CVRi
MAP
variables upon which median splits
(yielding high and low CVRi
MAP
groupings) were later performed for graphing purposes.

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 64

Figure 7.
Baseline CVRi
MAP
x amyloid prediction of longitudinal recognition memory

Cerebrovascular resistance x amyloid effects on Rey Auditory Verbal Learning Test (RAVLT)
recognition score. Analyses were performed using continuous CVRi
MAP
variables upon which
median splits (yielding high and low CVRi
MAP
groupings) were later performed for graphing
purposes.
8
9
10
11
12
13
14
Time 1 Time 2 Time 3
RA VLT recognition
Aβ+ Right Hippocampus CVRi
Low CVRi High CVRi
8
9
10
11
12
13
14
Time 1 Time 2 Time 3
RA VLT recognition
Aβ- Right Hippocampus CVRi

CEREBROVASCULAR RESISTANCE IN ALZHEIMER’S DISEASE 65

Appendix 1.
ADNI 2 Sample Composition
Year 1 Year 2 Year 3 Year 4 Year 5
Control
Existing
Newly enrolled
352
202 (57%)
150 (43%)
331
190 (57%)
141 (43%)
312
179 (57%)
133 (43%)
293
168 (57%)
125 (43%)
276
158 (57%)
118 (43 %)
MCI
Existing
Newly enrolled
724
474 (65%)
250 (35%)
681
446 (65%)
235 (35%)
641
420 (66%)
221 (34%)
602
394 (65%)
208 (35%)
566
370 (65%)
196 (35%)
AD 150 141 133 † †
Total 1226 1153 1086 895 842
† AD participants were only followed for 3 years

Number of participants in the control, mild cognitive impairment (MCI), and Alzheimer’s
disease (AD) groups for each year of the Alzheimer’s Disease Neuroimaging Initiative – 2
(ADNI 2) study. Proportions of participants that were existing (i.e. continuing from ADNI I
and/or ADNI GO study phases) versus newly enrolled are also indicated.

Abstract (if available)

Abstract Given the proposed role of vascular factors in normal and pathological aging, we investigated whether elevated cerebrovascular resistance in brain regions typically affected by cognitive aging and Alzheimer’s disease (AD) was associated with amyloid-β accumulation and cognitive decline. Cerebral blood flow (CBF) was measured in a sample of older North American adults (N = 232) using arterial spin labeling magnetic resonance imaging. An estimate of cerebrovascular resistance (CVRiMAP) was calculated as the ratio of mean arterial blood pressure to CBF in regions of interest. Amyloid-β retention and positivity were indexed using positron emission tomography with a florbetapir-fluorine-18 (¹⁸F) tracer, while cognitive performance was evaluated via assessments of global cognition, memory, and executive function. Cross-sectional analyses indicated highest CVRiMAP for AD (n = 33), intermediate values for non-demented amyloid-positive (n = 87), and lowest estimates for amyloid-negative (n = 112) participants. Group differences were more pronounced, and present in more regions, for CVRiMAP relative to CBF. Furthermore, CVRiMAP but not CBF differentiated non-demented amyloid-positive from amyloid-negative individuals. Further analyses detected interactive effects of CVRiMAP and amyloid-β deposition on cognition such that cognitive deficits were moderated by, if not dependent upon, concurrent elevation in cerebrovascular resistance. Longitudinal analyses revealed similar interactive effects for cognitive decline across 3 time points. Declines in cognitive performance over time were accelerated by elevated baseline CVRiMAP, particularly for amyloid-positive individuals. Earlier elevations in CVRiMAP for frontal regions predicted later whole-brain amyloid increases. Our results thus suggest that increases in cerebrovascular resistance may represent a crucial component of AD pathogenesis, working synergistically with, and in some areas pre-dating, amyloidosis to produce cognitive decline.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Contributions of dynamic cerebrovascular function to cognitive decline and dementia: development and validation of a novel neuroimaging approach

PDF

The role of blood pressure variability in cognitive decline, cerebrovascular disease and Alzheimer’s disease

PDF

Long-term blood pressure variability across the clinical and biomarker spectrum of Alzheimer’s disease

PDF

Effects of AT-1 receptor blockers on cognitive decline and Alzheimer's disease

PDF

Longitudinal neurocognitive profiles of empirically-derived Alzheimer’s disease variants

PDF

The role of the locus coeruleus in Alzheimer’s disease and cerebrovascular function: insights from neuroimaging, neuropsychology, and biofluid markers

PDF

Brainstem structural integrity in the progression of Alzheimer's disease

PDF

A virtual reality exergaming system to enhance brain health in older adults at risk for Alzheimer’s disease

PDF

Blood-brain barrier pathophysiology in cognitive impairment and injury

PDF

Insulin sensitivity in cognition, Alzheimer's disease and brain aging

PDF

Role of oxidative stress in age-associated mild cognitive impairment and Alzheimer's disease

PDF

Affective neuropsychiatric symptoms and neural connectivity in the early stages of Alzheimer’s disease

PDF

Cross-sectional association of blood pressure, antihypertensive medications, MRI volumetric measures and cognitive function scores in an aging population

PDF

Untreated hyperglycemia associated with tau pathology and worse cognitive performance in older adults

PDF

Accuray of subjective cognitive complaints in a longitudinal context: the effect of depression and dementia status

PDF

TREM2 and C1q signaling regulates immunoproteostasis in Alzheimer's disease

PDF

Cerebrovascular disease of white matter in patients with chronic anemia syndrome

PDF

Memory abnormalities in Alzheimer's disease and anxiety models

PDF

Using neuroinformatics to identify genomic and proteomic markers of suboptimal aging and Alzheimer's disease

PDF

Vascular contributions to brain aging along the Alzheimer's disease continuum

Asset Metadata

Creator Yew, Belinda (author)

Core Title Indexing cerebrovascular resistance in cognitive decline and Alzheimer's disease

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 07/22/2016

Defense Date 05/24/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aging,Alzheimer's disease,amyloid-β,cerebral blood flow,cerebrovascular resistance,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Nation, Daniel A. (committee chair), Bechara, Antoine (committee member), Gatz, Margaret (committee member)

Creator Email belinday@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-274464

Unique identifier UC11280538

Identifier etd-YewBelinda-4590.pdf (filename),usctheses-c40-274464 (legacy record id)

Legacy Identifier etd-YewBelinda-4590.pdf

Dmrecord 274464

Document Type Thesis

Format application/pdf (imt)

Rights Yew, Belinda

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA