Skip to main content
NCBI home page
Brain logoLink to Brain
. 2017 Jun 1;140(7):1987–2001. doi: 10.1093/brain/awx112

Cerebrovascular resistance: effects on cognitive decline, cortical atrophy, and progression to dementia

Belinda Yew 1, Daniel A Nation 1,✉; for the Alzheimer’s Disease Neuroimaging Initiative*
PMCID: PMC6059092  PMID: 28575149

See Markus (doi:10.1093/awx161) for a scientific commentary on this article.

Abnormalities in cerebral blood pressure and flow have been implicated in Alzheimer’s disease. Yew and Nation show that elevated cerebrovascular resistance, defined as the ratio of blood pressure to cerebral blood flow, is associated with greater cognitive decline, brain atrophy, and likelihood of progression to dementia over two years in older adults.

Keywords: Alzheimer’s disease, cerebral blood flow, blood pressure, beta-amyloid, dementia biomarkers

Abstract

See Markus (doi:10.1093/awx161) for a scientific commentary on this article.

Evidence for vascular contributions to Alzheimer’s disease has been increasingly identified, with increased blood pressure and decreased cerebral blood flow both linked to in vivo biomarkers and clinical progression of Alzheimer’s disease. We therefore hypothesized that an elevated ratio of blood pressure to cerebral blood flow, indicative of cerebrovascular resistance, would exhibit earlier and more widespread associations with Alzheimer’s disease than cerebral blood flow alone. Further, we predicted that increased cerebrovascular resistance and amyloid retention would synergistically influence cognitive performance trajectories, independent of neuronal metabolism. Lastly, we anticipated associations between cerebrovascular resistance and later brain atrophy, prior to amyloid accumulation. To evaluate these hypotheses, we investigated associations between cerebrovascular resistance and amyloid retention, cognitive decline, and brain atrophy, controlling for neuronal metabolism. North American older adults (n = 232) underwent arterial spin labelling magnetic resonance imaging to measure regional cerebral blood flow in brain regions susceptible to ageing and Alzheimer’s disease. An estimated cerebrovascular resistance index was then calculated as the ratio of mean arterial pressure to regional cerebral blood flow. Positron emission tomography with 18F-florbetapir and fludeoxyglucose was used to quantify amyloid retention and neuronal metabolism, respectively. Cognitive performance was evaluated via annual assessments of global cognition, memory, and executive function. Results indicated diminished inferior parietal and temporal cerebral blood flow for patients with Alzheimer’s disease (n = 33) relative to both non-demented groups, but no cerebral blood flow differences between non-demented amyloid-positive (n = 87) and amyloid-negative (n = 112) cases. In contrast, the cerebrovascular resistance index was significantly elevated in amyloid-positive versus amyloid-negative cases, with additional elevation in patients with Alzheimer’s disease. Furthermore, cerebrovascular resistance index group differences were of greater statistical effect size and encompassed a greater number of brain regions than those for cerebral blood flow alone. Cognitive decline over 2-year follow-up was accelerated by elevated baseline cerebrovascular resistance index, particularly for amyloid-positive individuals. Increased baseline cerebrovascular resistance index also predicted greater progression to dementia, beyond that attributable to amyloid-positivity. Finally, increased cerebrovascular resistance index predicted greater regional atrophy among non-demented older adults who were amyloid-negative. Findings suggest that increased cerebrovascular resistance may represent a previously unrecognized contributor to Alzheimer’s disease that is independent of neuronal hypometabolism, predates changes in brain perfusion, exacerbates and works synergistically with amyloidosis to produce cognitive decline, and drives amyloid-independent brain atrophy during the earliest stage of disease.

Introduction

Alzheimer’s disease is a neurodegenerative disorder classically characterized by accumulation of cerebral amyloid plaques and neurofibrillary tangles, resulting in marked deficits to episodic memory, and progressively broader cognitive impairment with disease evolution. Beyond these classic neuropathological hallmarks, vascular contributions to Alzheimer’s disease have also been increasingly recognized (Snyder et al., 2015). For example, vascular risk factors have been implicated in Alzheimer’s pathogenesis (Breteler, 2000; Viswanathan et al., 2009), with hypertension, heart disease, diabetes, and a host of other vascular vulnerabilities associated with cognitive decline, brain atrophy, white matter lesions, brain hypometabolism, and amyloid deposition (Luchsinger et al., 2005; Raz and Rodrigue, 2006; Langbaum et al., 2012). More recently, a series of studies have identified predictive relationships of elevated blood pressure and aortic stiffening with in vivo Alzheimer’s biomarkers (Langbaum et al., 2012; Nation et al., 2013a, 2015a; Rodrigue et al., 2013; Hughes et al., 2014), further confirming the pre-eminence of vascular markers during the earliest stages of disease (Iturria-Medina et al., 2016).

A number of models have been proposed to account for the synergistic interaction of the amyloid-induced pathophysiological cascade with vascular compromise. Many models emphasize the impact of reduced cerebral blood flow (CBF) (Weller et al., 2008; Zlokovic, 2011; Kress et al., 2014), which has been associated with both normal and pathological ageing across a number of brain regions (Iadecola, 2004). Moreover, accelerated declines are observed in cases of prodromal and clinical Alzheimer’s dementia (Dai et al., 2009; Alsop et al., 2010), with changes in CBF seemingly predating neurodegeneration, amyloid accumulation, and cognitive decline in those at genetic risk for Alzheimer’s disease. CBF reductions have thus been heavily implicated in early disease stages (Bookheimer et al., 2000; Iadecola, 2004; Sheline et al., 2010).

Increased cerebrovascular stiffening, endothelial dysfunction and/or smooth muscle contractility may drive or operate in parallel to these CBF abnormalities. It has been hypothesized, for example, that blood vessel pulsations provide the mobility necessary for perivascular drainage of amyloid (Weller et al., 2008). Stiffening of vessels can lead to reduced vascular compliance in accommodation of blood pressure changes, which may in turn decrease pulsation amplitude and consequently capacity for amyloid drainage (Weller et al., 2008). Cerebrovascular dysfunction may similarly disrupt brain lymphatic vessels, further impairing clearance capacity (Nedergaard, 2013; Iliff et al., 2015). Reduced clearance ability consequently instigates accumulation of amyloid, leading to plaque formation and/or increased risk of white matter lesions that can further promote Alzheimer’s symptomatology (Qiu et al., 2005; Bell et al., 2009; Weller et al., 2009). Amyloid accumulation can in turn generate further obstruction of blood and perivascular fluid flow via effects of cerebral amyloid angiopathy and neurotoxicity of amyloid oligomers (Weller et al., 2008). Amyloid-associated vasotoxicity may also produce vasoconstriction itself (Han et al., 2015).

Cerebrovascular stiffening and amyloid-induced vasoconstriction can lead to increased cerebrovascular resistance, the ratio of cerebral perfusion pressure (Pα) to CBF, where Pα is the difference between mean arterial pressure and intracranial pressure. Previous studies estimating cerebrovascular resistance in Alzheimer’s disease have identified increases within several brain regions (Nation et al., 2013b; Liu et al., 2014; den Abeelen et al., 2014). Although cerebrovascular autoregulation ensures steady cerebral perfusion across a broad range of mean arterial pressure values (∼60–150 mmHg), this is accomplished through prodigious changes in cerebrovascular resistance. Examining tissue perfusion without regard to blood pressure levels may thus mask changes in cerebrovascular resistance. Consistent with this, our prior work has identified more widespread cerebrovascular resistance changes than those of cerebral perfusion alone (Nation et al., 2013b). Furthermore, high blood pressure and low CBF have been independently linked to cerebral amyloidosis, suggesting that their ratio (i.e. cerebrovascular resistance) may increase during the earlier stages of Alzheimer’s disease.

To date, no studies have investigated cerebrovascular resistance in relation to cerebral amyloid, or evaluated whether these two pathological events synergistically drive future cognitive decline. The present study sought to examine this and establish the contribution of cerebrovascular dysfunction to Alzheimer’s disease through longitudinal study of an estimate of cerebrovascular resistance in Alzheimer’s patients, amyloid-positive individuals without diagnosis of dementia (heretofore referred to as ‘non-demented amyloid-positive’), and amyloid-negative participants. Specifically, we evaluated whether estimated cerebrovascular resistance would show earlier and more widespread changes than CBF alone, independent of cerebral hypometabolism measured by fluorodeoxyglucose (FDG)-PET. We also explored whether these resistance changes interacted with cerebral amyloid status to predict future regional brain atrophy and cognitive decline in older adults initially free of dementia. Finally, we sought to determine whether cerebrovascular resistance is predictive of early atrophy within corresponding brain regions prior to amyloid accumulation (i.e. among non-demented, amyloid-negative older adults).

Materials and methods

Participants

Participants were selected from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database, a repository of data obtained from volunteer adults aged 55–90, recruited at over 50 North American sites. Accepted participants had Geriatric Depression Scale scores < 6, Hachinski Ischemic Scale scores ≤ 4, and no significant neurological disease (other than suspected Alzheimer’s disease). Only participants who underwent arterial spin labelling (ASL) functional MRI and florbetapir PET scanning (i.e. individuals sourced from sites with Siemens scanners) were included in the present analyses. Participants were divided into those with and without a diagnosis of Alzheimer’s dementia. The Alzheimer’s dementia group 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, with Mini-Mental State Examination (MMSE) (Folstein et al., 1975) scores of 20 or above, and Clinical Dementia Rating scale (CDR) (Morris, 1993) scores of 0.5 or 1. Remaining participants had MMSE scores ≥24 and CDR scores of 0.5 or 0. These individuals were further classified based on amyloid load, determined by initial florbetapir PET scanning. Specifically, participants with SUVR >1.11 were categorized as non-demented amyloid-positive, while those with SUVR ≤ 1.11 were classified as amyloid-negative. A subset of participants received follow-up MRI and cognitive testing at ∼12 months and 24 months after baseline. Follow-up florbetapir PET scans were collected at 24 and 48 months. Sample sizes are shown in Supplementary Table 1.

Procedures and measures

Cerebrovascular resistance index: blood pressure to blood flow ratio

Blood pressure was measured using a calibrated mercury sphygmomanometer and stethoscope. Measurements were taken while the participant was seated with forearm positioned horizontally and at heart level. Where possible, measurements were taken from the dominant arm.

The average blood pressure within an individual’s arteries during a single cardiac cycle, mean arterial pressure, was estimated by multiplying diastolic blood pressure (DBP) by 2 then adding systolic blood pressure (SBP) and dividing by 3 (Equation 1).

CBF was determined using ASL functional MRI 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 = 256 mm, repetition time = 3400 ms, echo time = 12 ms, inversion time for arterial spins (TI1) = 700 ms, total transit time of spins (TI2) = 1900 ms, matrix = 64 × 64, tag thickness = 100 mm, gap from tag to proximal slice = 25.4 mm, slice thickness = 4 mm, number of axial slices = 24, and time between slices = 22.5 ms. 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/100 g/min). More detailed descriptions of ASL acquisition and processing procedures are provided by ADNI (ADNI, 2011). Left and right medial orbitofrontal cortex (mOFC), rostral middle frontal gyrus (rMFG), hippocampus, inferior temporal cortex (ITC), and inferior parietal cortex (IPC) CBF were examined. These regions were selected based on previously identified associations with cerebrovascular dysfunction (Du et al., 2006; Nation et al., 2013b; Mattson et al., 2014) in Alzheimer’s disease. Regional CBF values were residualized by precentral gyrus CBF to adjust for individual variation in flow. This reference region was selected due to its relative sparing from damage in Alzheimer’s disease (Thompson et al., 2003) and use in prior study of ADNI CBF data (Mattson et al., 2014). CBF values were analysed both with and without adjustment for neuronal metabolism. Adjustment was accomplished by covarying for composite FDG-PET values. ASL MRI was performed at baseline, 12, and 24 months. Sample sizes are presented in Supplementary Table 1.

Structural MRI was performed during the same session as ASL. A T1-weighted 3D MPRAGE sequence was used, with field of view = 256 mm, repetition time = 2300 ms, echo time = 2.98 ms, flip angle = 9°, and resolution = 1.1 × 1.1 ± 1.2 mm3. Structural scans were skull-stripped, segmented, and parcellated using FreeSurfer version 4.5.0 (surfer.nmr.mgh.harvard.edu).

Cerebrovascular resistance is the ratio of cerebral perfusion pressure (Pα, the difference between mean arterial pressure and intracranial pressure) to CBF (Equation 2). Under typical conditions wherein intracranial pressure is normal and considerably lower than mean arterial pressure, however, cerebrovascular resistance may be estimated by the ratio of mean arterial pressure to CBF (Equation 3). This estimate, cerebrovascular resistance index (CVRi), was used to index cerebrovascular resistance. CVRi values were analysed for the aforementioned regions (i.e. mOFC, rMFG, hippocampus, ITC, and IPC).

MAP=SBP+2(DBP)3 (1)
Cerebrovascular resistance=PαCBF=MAP − ICPCBF (2)
CVRi=MAPCBF (3)

Cerebral amyloidosis

Amyloid retention was indexed using PET scanning with an 18F-florbetapir tracer. Each participant was administered a single intravenous bolus of 18F-florbetapir prior to AV-45 PET imaging. PET imaging commenced ∼50–70 min following injection and ran for 20 min. Image reconstruction was conducted immediately after scanning and, where motion artefact was identified, scanning repeated. Florbetapir was quantified using 3 T 3D MPRAGE structural scans collected earlier. Florbetapir uptake was calculated for the frontal, temporal, and parietal lobes. A cortical summary standardized reuptake value ratio (SUVR) was calculated by dividing the average florbetapir uptake across frontal, cingulate, lateral parietal, and lateral temporal cortices, by mean cerebellar florbetapir uptake. In non-demented participants, amyloid status (i.e. positive or negative) was then identified using a 1.11 threshold shown to reliably differentiate Alzheimer’s patients from cognitively normal individuals, and strongly correlate with classifications made by another amyloid indexing radiotracer (i.e. Pittsburgh compound B). This cut-off equates to the upper limit of a 95% confidence limit for mean SUVR obtained in a sample of normal controls (Landau et al., 2011; Joshi et al., 2012). Florbetapir PET scanning was conducted at baseline, and ∼24 and 48 months after baseline. Sample sizes are presented in Supplementary Table 1.

Cerebral metabolism

Glucose uptake was indexed using FDG PET imaging, 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 min prior to scanning. Scan duration was ∼30 min. 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 regions of interest defined based on prior studies of metabolic changes in pathological ageing. A composite region based on combined values for all regions of interest was also calculated. Individual and composite regions of interest were intensity normalized by dividing by the mean for a pons/cerebellar vermis reference region (Landau et al., 2011). The composite value was used in the present study as an index of global brain metabolism.

Cerebral atrophy

For amyloid-negative participants, regional cortical thickness values were obtained using the aforementioned T1-weighted 3D MPRAGE scans, collected at baseline, and ∼12 and 24 months after baseline. FreeSurfer segmentation of subcortical white and grey matter volumes was performed as described by Fischl et al. (2004). Sample sizes for cortical thickness analyses are presented in Supplementary Table 1.

White matter hyperintensity volume

White matter hyperintensity (WMH) volume was calculated using an automated Bayesian Markov-Random Field (MRF) approach generated from T1-, T2, and proton density-weighted brain MRI scans. The resulting model incorporated probability priors derived from manually assessed example fluid-attenuated (FLAIR) images. Specifically, the MRF framework combined scan intensity distributions with prior probabilities of WMH in a given voxel, and contextual probability of WMH given the status (WMH present or absent) of neighbouring voxels (Schwarz et al., 2009). Schwarz et al. (2009) describe these methods in greater detail. Scans for WMH analyses were collected at baseline, 12, and 24 months. Sample sizes are presented in Supplementary Table 1.

Neuropsychological decline

Global cognitive decline was measured using the MMSE (Folstein et al., 1975). Given evidence of both amnestic and dysexecutive preclinical phenotypes (Dickerson et al., 2011), we also explored performance on memory and executive function tasks. Verbal memory encoding was assessed using the Rey Auditory Verbal Learning Test (RAVLT) recognition subtest (Rey, 1941), which has been shown to identify memory decline due to Alzheimer’s disease (Andersson et al., 2006; Schoenberg et al., 2006; Jedynak et al., 2012). Executive function was evaluated using Trails B of the Trail Making Test, which is also sensitive to clinical progression in Alzheimer’s disease (Ashendorf et al., 2008). Trails B scores reflect time taken to complete the trail. Higher times/scores thus suggest poorer executive function (Spreen and Strauss, 1998). Cognitive performance was collected at baseline, and ∼12 and 24 months after baseline. Sample sizes are presented in Supplementary Table 1.

Data used in the preparation of this article were obtained from the ADNI database (adni.loni.usc.edu). The ADNI was launched in 2003 as a public-private partnership, led by Principal Investigator Michael W. Weiner, MD. The primary goal of ADNI has been to test whether serial MRI, PET, other biological markers, and clinical and neuropsychological assessment can be combined to measure the progression of mild cognitive impairment (MCI) and early Alzheimer’s disease. For up-to-date information, see www.adni-info.org.

Statistical analyses

All analyses were performed using IBM SPSS Statistics (Version 22). Baseline group differences across demographic and physiological [i.e. age, sex, education, body mass index (BMI), APOE ɛ4 carrier status, systolic and diastolic blood pressure, and mean arterial pressure] variables were evaluated using MANOVA and chi-square tests of independence. MANOVAs were used to evaluate CBF and CVRi differences in all participants (i.e. demented and non-demented). Each MANOVA compared CBF or CVRi in the five selected regions among Alzheimer’s, non-demented amyloid-positive and amyloid-negative groups, controlling for age, APOE ɛ4 carrier status, BMI, sex, and FDG uptake. To control for multiple comparisons, all five regions were entered into MANOVA simultaneously, and follow-up analyses restricted to those regions for which MANOVA was significant. Where significant group MANOVA effects were detected, least significant difference (LSD) pairwise analyses (e.g. Alzheimer’s versus amyloid-positive, amyloid-positive versus amyloid-negative) were employed to clarify group differences.

The effect of APOE genotype on regional CVRi was also explored, with MANOVA investigating differences in CVRi between APOE ɛ4 carriers and non-carriers. Potential APOE × amyloid-status (i.e. amyloid-positive versus amyloid-negative) effects were also investigated using linear regression. All APOE analyses controlled for age, BMI, sex, and FDG uptake.

Longitudinal analyses were restricted to non-demented participants, and investigated CVRi and cognitive change across baseline, 12 months and 24 months. Effects of baseline CVRi on longitudinal changes in cognitive performance were evaluated using linear mixed models employing maximum likelihood estimation. Cognitive measures (i.e. MMSE, RAVLT recognition, and Trails B scores) were entered as dependent variables, with fixed effects of amyloid-status, baseline CVRi, and time. Covariates of age, sex, APOE ɛ4 carrier status genotype, BMI, years of education, and FDG uptake were included. To minimize comparisons, analyses of RAVLT and Trails B scores were restricted to a priori selected brain regions known to underpin examined cognitive abilities (i.e. rMFG, mOFC, and IPC for Trails B; hippocampus and ITC for RAVLT). Time was entered as a random effect with an autoregressive covariance structure.

To evaluate the relationship between CVRi and cerebral amyloid load, linear regression investigated baseline CVRi as a predictor of regional amyloid load at baseline, 24 months and 48 months. Results were adjusted for baseline severity through inclusion of baseline MMSE score as a covariate in analyses of 24- and 48-month amyloid load. Regional amyloid SUVR was entered as the dependent variable, with baseline regional CVRi entered as an independent variable. As analyses sought to investigate contributions of CVRi to cognition, and were therefore exploratory in nature, we did not correct for multiple comparisons. However, to restrict the total number of analyses, only associations between CVRi and amyloid load in the corresponding region (e.g. frontal amyloid predicted by mOFC CVRi, or temporal amyloid predicted by ITC CVRi) were explored.

To evaluate the relationship between CVRi and cerebral atrophy, independent of cerebral amyloid, linear regression was used to analyse baseline CVRi as a predictor of cortical thickness among those who were amyloid-negative. For each region (excluding hippocampus), cortical thickness was regressed on baseline CVRi in the same region (e.g. mOFC CVRi prediction of mOFC cortical thickness). Separate regressions were conducted for each brain region at each time point (i.e. baseline, 12 months, and 24 months). Results were adjusted for baseline severity through inclusion of baseline MMSE score as a covariate in analyses of 12- and 24-month cortical thickness.

The relationship between CVRi and WMH volume, independent of cerebral amyloid, was assessed using linear regression. WMH volume at baseline, 12, and 24 months was regressed on baseline CVRi. Separate regressions were conducted for each CVRi brain region at each WMH volume time point. Results were adjusted for baseline severity through inclusion of baseline MMSE score as a covariate in analyses of 12- and 24-month WMH volume.

Cox regressions were also used to determine relationships between CVRi, amyloid load, and progression to dementia. Specifically, median splits were performed on regional CVRi values in amyloid-positive cases, yielding amyloid-positive high CVRi, and amyloid-positive low CVRi groups. These groups were then compared to amyloid-negative participants in prediction of progression to Alzheimer’s disease over follow-up of 24 months. As some participants had follow-up beyond these 24 months, Cox regressions were also run on an extended dataset with variable follow-up from 3 to 120 months in order to confirm findings from the fixed interval follow-up analysis.

For all Cox and linear regression analyses, age, sex, BMI, APOE ɛ4 carrier status, and FDG uptake were entered as covariates.

To address potential inflation of type I error due to multiple comparisons, false discovery rate (FDR) was controlled using the Benjamini-Hochberg procedure (Benjamini and Hochberg, 1995). Results were assessed when FDR was controlled at both 0.05 and 0.10.

Experimental design

CDR raters were blinded to participants’ cognitive test performance. MRI and PET processing and analyses were performed using automated procedures. Group sample sizes met previously reported requirements for power to detect associations between diagnosis (i.e. Alzheimer’s disease versus MCI versus cognitively normal), cognitive performance, and brain volume (Morra et al., 2009).

Results

Baseline differences in cerebrovascular resistance index versus cerebral blood flow at each Alzheimer’s disease stage

Participant baseline demographic and clinical data are presented in Table 1. Alzheimer’s and amyloid-positive groups were significantly older than the amyloid-negative group (both P < 0.001). The Alzheimer’s participants also had significantly lower BMI than their amyloid-negative counterparts (P = 0.02). There were significantly more APOE ɛ4 allele carriers in the Alzheimer’s (P < 0.001) and amyloid-positive groups (P < 0.001) relative to the amyloid-negative group. There were also significantly more carriers of two APOE ɛ4 alleles in the Alzheimer’s relative to amyloid-positive and amyloid-negative groups (P < 0.05). There were significantly more MCI cases in the amyloid-positive relative to amyloid-negative group (P < 0.001). There were no significant group differences in blood pressure (all P > 0.10). Expected group differences in baseline cognitive performance were detected. Amyloid-negative participants outperformed non-demented amyloid-positive participants, who in turn outperformed the Alzheimer’s group (Table 1).

Table 1.

Baseline demographic and clinical data

Amyloid- negative Amyloid- positive Alzheimer’s disease F or χ2 P
n 112 87 33
Age 69.2 (0.6) 73.7 (0.7) 73.2 (1.2) 12.38 <0.01
Sex (M/F) 53/59 47/40 18/15 1.09 0.58
Education 16.7 (0.3) 16.2 (0.3) 16.4 (0.5) 0.95 0.39
BMI 27.3 (0.4) 26.6 (0.5) 25.3 (0.7) 3.08 0.05
MCI 58 (52%) 67 (77%) - 13.34 <0.01
Proportion APOE ɛ4 carriers (none/1/2)a 79/18/3 39/51/1 27/39/33 60.40 <0.01
SBPb 130.0 (1.1) 133.3 (1.9) 133.6 (3.2) 0.8 0.45
DBPb 73.2 (1.0) 75.2 (1.1) 74.0 (1.8) 0.85 0.43
MAPb 92.2 (1.1) 94.6 (1.2) 93.8 (2.0) 0.99 0.37
PP* 56.8 (1.5) 58.1 (1.6) 59.6 (2.7) 0.40 0.67
MMSEc,d,f,g 29.0 (0.2) 28.0 (0.2) 23.6 (0.3) 122.0 <0.01
Trails B (s)c,e,f,g 84.6 (6.4) 109.6 (7.1) 147.8 (11.7) 10.0 <0.01
Recognitionc,e,f,g 12.7 (0.3) 11.4 (0.4) 7.8 (0.6) 24.4 <0.01
Open in a new tab

Data are mean (SD).

aPercentage of participants with none, one, or two APOE ɛ4 alleles.

bReported statistics are adjusted for age, sex, BMI, and APOE genotype.

c Reported statistics are adjusted for age, sex, BMI, APOE genotype, and education.

dAβ− > Aβ+, P < 0.01; e Aβ− > Aβ+, P < 0.05; f Aβ− > AD, P < 0.01; g Aβ+ > AD, P < 0.01.

BMI = body mass index; SBP = systolic blood pressure; DBP = diastolic blood pressure; MAP = mean arterial pressure; PP = pulse pressure.

Group means for CBF in target regions are presented in Fig. 1A. Multivariate ANCOVA (MANCOVA) indicated a significant effect of group on regional CBF in three brain areas (Supplementary Table 2). Post hoc LSD tests showed that CBF was reduced for Alzheimer’s relative to amyloid-negative participants in the left hippocampus (P = 0.03) and left ITC (P = 0.01). The Alzheimer’s group also exhibited reduced right IPC CBF compared to the amyloid-positive (P < 0.001) and amyloid-negative (P < 0.001) groups. No differences between amyloid-negative and -positive groups were detected for CBF (all P > 0.10).

Figure 1.

Figure 1

Open in a new tab

Comparisons of regional CBF and cerebrovascular resistance index (CVRi). (A) Mean regional CBF and (B) mean CVRi, for Alzheimer’s disease (AD), non-demented amyloid-positive (Aβ+), and amyloid-negative (Aβ−) groups. Reduced regional CBF was observed for AD and Aβ+ participants relative to Aβ− participants. Highest CVRi was observed in AD, intermediate CVRi in Aβ+, and lowest CVRi in Aβ− participants. L = left; R = right; hipp = hippocampus. *P < 0.05; **P < 0.01. Error bars represent ± 1 standard error.

Mean group CVRi values are presented in Fig. 1B. MANCOVA indicated a significant effect of group on regional CVRi in eight brain regions (Supplementary Table 2). Post hoc LSD tests revealed elevated CVRi in Alzheimer’s compared to amyloid-positive participants for the right IPC (P = 0.02), and left rMFG (P = 0.04) and hippocampus (P = 0.02). CVRi for the Alzheimer’s group was also higher than the amyloid-negative group in the left mOFC (P = 0.01), rMFG (P < 0.001), hippocampus (P < 0.01) and ITC (P < 0.01), as well as the right mOFC (P = 0.04), rMFG (P < 0.01), hippocampus (P < 0.01), ITC (P = 0.02), and IPC (P < 0.001). Amyloid-positive participants exhibited elevated CVRi relative to amyloid-negative participants in the left mOFC (P = 0.05) and rMFG (P = 0.02), as well as the right rMFG (P = 0.02), hippocampus (P = 0.03), ITC (P < 0.01), and IPC (P = 0.03). CVRi did not differ significantly between APOE genotype groups (all P > 0.05). With FDR limited to 0.05, the amyloid-positive versus Alzheimer’s group difference in left rMFG CVRi, amyloid-negative versus Alzheimer’s difference in right mOFC, and amyloid-positive versus -negative differences in left mOFC, right hippocampus, and right IPC CVRi remained significant.

No main effects of APOE genotype on CVRi were detected (all P > 0.05). A significant APOE × amyloid status effect was found for left hippocampal CVRi (P = 0.04). However, when analyses were stratified by amyloid status, no significant effect of APOE on left hippocampal CVRi was detected for amyloid-positive or amyloid-negative participants. No significant APOE × amyloid interactions were detected for any other brain region (all P > 0.05).

Longitudinal prediction of amyloid load by baseline cerebrovascular resistance index

Higher baseline left and right ITC CVRi were associated with greater temporal amyloid load at baseline. Higher right ITC also predicted higher temporal amyloid load at 24, and 48 months. Higher baseline right rMFG CVRi was also associated with greater baseline frontal amyloid load (Table 2). CVRi prediction of amyloid load was not identified for any other brain regions (all P > 0.05).

Table 2.

Baseline CVRi prediction of longitudinal regional amyloid load

Brain region Amyloid-β time point ß t df P
Right rMFG Baseline 0.12 2.05 188 0.04
Left ITC Baseline 0.14 2.04 175 0.04
Right ITC Baseline 0.17 2.40 175 0.01Ψ
24 months 0.20 2.43 130 0.02Ψ
48 months 0.36 3.05 64 0.01Ψ
Open in a new tab

Results from linear regression of baseline and follow-up regional amyloid load on baseline CVRi in the same region. Covariates of age, body mass index, sex, APOE genotype, baseline MMSE score, and composite FDG PET values were included. All findings remain significant under false discovery rate (FDR) of 0.10.

ΨSignificant under FDR of 0.05.

Baseline cerebrovascular resistance index × amyloid status prediction of longitudinal cognitive decline

Global cognition

There were significant CVRi × amyloid interactions predicting decline in MMSE score for the right rMFG [F(1, 224) = 4.47, P = 0.04] and IPC [F(1, 347) = 6.27, P = 0.01]. Specifically, higher baseline right rMFG and IPC CVRi were associated with larger declines in MMSE for amyloid-positive, but not amyloid-negative, participants (Fig. 2). CVRi prediction of MMSE scores was not detected for the left rMFG and IPC, or the left and right mOFC and ITC (all P > 0.05). Differences associated with the right IPC but not the right rMFG remained significant when FDR was limited to 0.05.

Figure 2.

Figure 2

Open in a new tab

Baseline CVRi × amyloid effects on MMSE total score. In amyloid-positive cases, higher baseline CVRi predicted poorer and larger declines in MMSE score at later time points. Aβ+ = non-demented amyloid-positive; Aβ− = amyloid-negative. Error bars represent±1 SE. Analyses were performed using continuous CVRi variables. Median splits (yielding high and low CVRi groupings) are provided here for visual comparison.

Executive function

As shown in Fig. 3, there were significant CVRi × amyloid interactions predicting longer Trails B completion time, for the left mOFC [F(1, 224) = 4.87, P = 0.03] and IPC [F(1, 283) = 8.74, P < 0.01]. Higher baseline left mOFC CVRi predicted faster increases in Trails B for amyloid-positive relative to amyloid-negative cases. Higher baseline left IPC CVRi predicted increases in Trails B time for amyloid-positive but not amyloid-negative participants. CVRi prediction of Trails B times was not detected for the right mOFC or IPC (both P > 0.05). Differences associated with the left IPC but not the left mOFC remained significant when FDR was limited to 0.05.

Figure 3.

Figure 3

Open in a new tab

Baseline CVRi × amyloid effects on time taken to complete Trails B. In amyloid-positive cases, higher baseline CVRi predicted poorer and larger declines in Trails B time at later time points. Aβ+ = non-demented amyloid-positive; Aβ− = amyloid-negative; Trails B = Trail Making Test Part B. Error bars represent±1 SE. Analyses were performed using continuous CVRi variables. Median splits (yielding high and low CVRi groupings) are provided here for visual comparison.

Memory

As shown in Fig. 4, higher baseline right hippocampal CVRi was associated with greater declines in RAVLT recognition for amyloid-positive compared to amyloid-negative individuals [F(1, 200) = 5.30, P = 0.022]. This effect remained significant after FDR was limited to 0.05. CVRi prediction of recognition scores was not detected for the left hippocampus, or the right and left ITC (all P > 0.05).

Figure 4.

Figure 4

Open in a new tab

CVRi × amyloid effects on RAVLT recognition score. In amyloid-positive cases, higher baseline CVRi predicted poorer and larger declines in recognition score at later time points. Aβ+ = non-demented amyloid-positive; Aβ− = amyloid-negative. Error bars represent±1 SE. Analyses were performed using continuous CVRi variables. Median splits (yielding high and low CVRi groupings) are provided here for visual comparison.

Longitudinal prediction of regional cortical thickness by baseline cerebrovascular resistance index

In amyloid-negative participants, higher regional CVRi at baseline was associated with lower cortical thickness within the same brain region, particularly in temporal areas (Table 3).

Table 3.

Baseline CVRi prediction of regional cortical thickness

Brain region Cortical thickness time point ß t df P
Left ITC 24 months −0.29 −2.64 64 0.01Ψ
12 months −0.32 −3.28 91 <0.01Ψ
Baseline −0.30 −3.31 100 <0.01Ψ
Right ITC 12 months −0.21 −2.27 91 0.03
Left IPC Baseline −0.20 −2.10 106 0.04
Open in a new tab

Results from linear regression of baseline and follow-up regional cortical thickness on baseline CVRi in the same region. Only amyloid-negative cases were analysed. Covariates of age, APOE genotype, body mass index, sex, baseline MMSE score, and composite FDG PET values were included. All findings remain significant under false discovery rate (FDR) of 0.10.

ΨSignificant under FDR of 0.05.

Baseline cerebrovascular resistance index × amyloid status prediction of longitudinal WMH volume

WMH volume was not predicted by CVRi in any region (all Ps > 0.05).

Prediction of progression to dementia by baseline cerebrovascular resistance index

Elevated risk of progression to dementia, above and beyond the risk conveyed by amyloid positivity, was associated with high CVRi in the left mOFC [odds ratio (OR) = 7.94, P = 0.03], hippocampus (OR = 6.19, P = 0.05), ITC (OR = 4.08, P = 0.02), and IPC (OR = 6.91, P = 0.03), as well as the right mOFC (OR = 8.01, P = 0.02), rMFG (OR = 8.04, P = 0.03), and hippocampus (OR = 6.09, P = 0.05) (Fig. 5). For the left rMFG, there was a non-significant trend in the same direction (OR = 5.82, P = 0.06). Elevated dementia risk associated with higher left ITC, right mOFC, and right rMFG CVRi remained significant when FDR was limited to 0.05.

Figure 5.

Figure 5

Open in a new tab

Risk of progression to dementia associated with high and low CVRi. Survival plots derived from Cox regressions comparing progression to dementia in amyloid-negative, amyloid-positive with low CVRi, and amyloid-positive with high CVRi, cases. Elevated CVRi was associated with increased risk of progression to Alzheimer’s disease, above and beyond risk associated with amyloid positivity. Aβ+ = non-demented amyloid-positive; Aβ− = amyloid-negative. In amyloid-positive participants, CVRi scores were subjected to a median split yielding ‘low’ and ‘high’ CVRi groupings.

Attrition

Chi-square tests were performed to determine whether attrition rates differed among amyloid status groups, APOE ɛ4 carrier status, and older versus younger (defined via age median split) participants. Attrition from baseline to 12 months did not differ for any of these variables (all P > 0.05). Attrition from 12 months to 24 months was proportionately higher in amyloid-positive relative to amyloid-negative participants [χ2(1) = 4.33, P = 0.04], and older relative to younger participants [χ2(1) = 6.86, P = 0.01], but did not differ among APOE ɛ4 carrier status (P > 0.05).

False discovery rate

Statistical significance of all reported findings was retained under a 0.10 FDR. Results for which statistical significance was not maintained under 0.05 FDR are indicated as such above.

Discussion

Results of the present study suggest that elevations in cerebrovascular resistance are larger, occur earlier, and impact a broader range of brain regions than previously identified CBF changes (Alsop et al., 2010). This was reflected in CVRi increases that were larger in statistical effect size, differentiated prodromal Alzheimer’s cases (i.e. non-demented amyloid-positive versus -negative older adults), and manifested in frontal, hippocampal, temporal and parietal regions. In contrast, CBF decreases were of smaller effect size, detected only for cases already diagnosed with dementia, and observed in 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 Alzheimer’s disease than CBF alone. This is consistent with previous reports of increased blood pressure and decreased blood flow in association with Alzheimer’s pathophysiology (Langbaum et al., 2012; Nation et al., 2012, 2013a; Mattson et al., 2014), and other studies indicating increased cerebrovascular resistance in MCI and Alzheimer’s dementia (Nation et al., 2013b; Liu et al., 2014).

Prior studies using ASL MRI for measurement of arteriolar and capillary blood flow have suggested that brain tissue perfusion is attenuated in some regions in Alzheimer’s disease. These studies have not taken into account the role of blood pressure, despite its important role in both CBF and Alzheimer’s disease. Although cerebral autoregulation enables maintenance of steady CBF across a broad range of arterial pressures, this is accomplished through substantial changes in cerebrovascular resistance. Therefore, failure to consider concomitant arterial pressure levels may mask vascular resistance changes, underestimating the role of early vascular changes in Alzheimer’s disease. If CBF declines observed in Alzheimer’s disease are caused by vascular pathology, upstream changes in vascular resistance, which is chiefly generated by larger arteries and arterioles, might be expected prior to capillary-level changes in tissue perfusion. Our findings are consistent with this notion since cerebrovascular resistance increases appear to predate those of absolute CBF, and encompass a wider range of brain regions.

Early increases in cerebrovascular resistance were also associated with future amyloid accumulation within corresponding brain areas for frontal and temporal regions. Amyloid has shown vasoconstrictive and vasotoxic properties in animal and cell model studies (Niwa et al., 2002; Sole et al., 2015), which may contribute to cerebrovascular abnormalities. Consistent with our findings, cerebrovascular stiffening could conversely exacerbate amyloidosis, potentially through impairment of amyloid clearance mechanisms (Weller et al., 2008; Zlokovic, 2011; Iliff et al., 2015). This is consistent with reports of reduced dementia-associated structural and cognitive declines (Qiu et al., 2005), and attenuation of age-related cerebral amyloid deposition and dementia (Nation et al., 2015b), in individuals taking antihypertensive medications. Further research into the mechanisms responsible for cerebrovascular resistance increases in preclinical Alzheimer’s disease may thus improve our understanding of Alzheimer’s disease pathogenesis but also reveal treatment targets.

In addition to predicting amyloid accumulation, cerebrovascular resistance and cerebral amyloidosis showed synergistic effects on the erosion of cognitive faculties in a longitudinal analysis, further implicating cerebrovascular dysfunction in clinical progression of Alzheimer’s disease. Specifically, non-demented amyloid-positive cases demonstrated limited cognitive decline unless concurrent elevations in cerebrovascular resistance were also observed. Furthermore, increased cerebrovascular resistance in amyloid-positive individuals was associated with accelerated decline in global cognition, memory and executive function. Cerebrovascular resistance-based exacerbation of amyloid-driven executive function deficits was particularly marked in inferior parietal and frontal regions. This is neuroanatomically consistent with prior reports of a frontoparietal basis for executive function performance (Moll et al., 2002; Zakzanis et al., 2005). While the role of frontal regions in executive function has been well established (Alvarez and 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 Alzheimer’s may therefore reflect disruption of inferior parietal and frontal cerebrovasculature in conjunction with regional amyloid deposition. Hippocampal cerebrovascular resistance effects on amyloid-driven memory declines in amyloid-positive cases, are consistent with known vulnerability of the hippocampus and adjacent regions to the early stages of Alzheimer’s pathology (Braak et al., 1993; Thal et al., 2002) as well as vascular disease (Wu et al., 2008).

The observed synergistic effects of cerebrovascular resistance and cerebral amyloidosis on cognitive decline are all consistent with the notion that vascular and amyloid pathologies interactively drive the progression of Alzheimer’s disease. Our findings are also in line with models recognizing amyloid as a necessary but insufficient element in the broader cascade through which the clinical features of Alzheimer’s disease arise (Drachman, 2014; Musiek and Holtzman, 2015). More specifically, our results suggest that cerebrovascular dysfunction and resulting disruption to CBF may be necessary for the development of dementia in addition to amyloid status. This is in line 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 production, both of which contribute to cognitive decline characteristic of Alzheimer’s disease. Cerebrovascular resistance and associated or parallel CBF irregularities may thus modify, or operate in conjunction with, amyloid 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 cerebrovascular resistance increases with later amyloid load.

Although cerebral amyloidosis is thought to be pathognomonic of Alzheimer’s disease, it has been hypothesized that vascular changes may actually predate amyloid accumulation, a notion that has received some empirical support (Iturria-Medina et al., 2016). To address the question of whether increased cerebrovascular resistance may occur prior to cerebral amyloidosis, and whether it may independently contribute to early neurodegenerative change, we examined cerebrovascular resistance as a predictor of cortical atrophy among amyloid-negative participants. Findings indicated that in addition to its interaction with cerebral amyloid, cerebrovascular resistance also appears to exert effects above and beyond those produced by increased amyloid load. In amyloid-negative cases, elevations in inferior temporal cerebrovascular resistance at baseline were associated with later cortical thinning within these same regions. These findings suggest that age-related cerebrovascular changes within the inferior temporal regions could predate amyloid accumulation, which has been identified in early disease stages (Thal et al., 2002; Cho et al., 2016) and associated with preclinical memory declines (Chételat et al., 2011). Increased baseline cerebrovascular resistance within temporal, but also frontal and parietal, brain areas also predicted greater risk of progression to Alzheimer’s dementia beyond that attributed to cerebral amyloidosis, further supporting the notion that cerebrovascular resistance may impact clinical progression through both amyloid-dependent and independent pathways.

Importantly, all observed cerebrovascular resistance effects were independent of fluctuations in FDG-indexed neuronal metabolism, to which hypoperfusion is often attributed or linked (Du et al., 2006; Musiek et al., 2012). We conclude that cerebrovascular resistance changes are distinct from, and more than the mere products of, blood flow abnormality driven by metabolic dysfunction. This is underscored by the fact that CVRi, as the ratio of pressure to flow, may not be as directly linked to cerebral metabolism as CBF alone. We also noted that controlling for FDG-PET had little to no impact on observed group differences in either CBF or CVRi, further suggesting that previously described relationships between reduced CBF and Alzheimer’s disease are not driven by cerebral hypometabolism. Taken together, these data support hypothesized contributions of cerebrovascular dysfunction to initiation and evolution of Alzheimer’s pathophysiology. Study findings were also independent of white matter lesion burden, as cortical cerebrovascular resistance showed no relationships with white matter lesions, suggesting that the effects of cerebrovascular resistance on both amyloid and cognitive decline are not merely due to concomitant vascular brain injuries. This is also consistent with our prior work indicating that subcortical cerebrovascular resistance, but not cortical cerebrovascular resistance, was related to white matter lesion burden (Nation et al., 2013b). Whereas the present study focused only on cortical cerebrovascular resistance, future studies will further evaluate the role of subcortical cerebrovascular resistance in both vascular and Alzheimer’s pathologies.

Contrary to prior reports (Hajjar et al., 2015; Suri et al., 2015; Wolters et al., 2016), no significant differences in cerebrovascular resistance were identified between APOE genotypes.

This absence of APOE effects on cerebrovascular resistance did not vary with amyloid-status (i.e. amyloid-positive versus -negative), suggesting that cerebrovascular resistance-driven effects may be distinct from those attributable to APOE ɛ4. Given that APOE ɛ4 likely contributes to observed outcomes via amyloid retention (Kanekiyo, 2014), it is also possible that our use of groups defined by amyloid load expended variance attributable to APOE. APOE effects may therefore be evident within alternatively defined groups (e.g. MCI versus no MCI). Furthermore, APOE × diagnostic group effects could be examined in relation to a number of outcomes beyond baseline CVRi (e.g. APOE × amyloid × CVRi effects on cognition and brain volume). While more thorough examination of APOE interactions with CVRi is beyond the scope of the present study, it is nonetheless an important focus for future research.

The strengths of the present study include large sample size, assessment of cerebrovascular resistance adjusting for brain glucose metabolism, utilization of both cortical atrophy and neuropsychological decline as outcomes, and longitudinal design. We also chose to evaluate CVRi differences and predictive utility in relation to amyloid status rather than cognitive status in these non-demented older adults. This approach allows for investigation of non-biased relationships with Alzheimer’s disease pathophysiology and neuropsychological outcomes. A number of limitations, however, must also be acknowledged. Our estimate of cerebrovascular resistance relied on simple seated blood pressure measures from the brachial artery, ignoring potential differences in intracranial pressure. Calculation of actual cerebrovascular resistance would require simultaneous measurement of cerebral perfusion pressure and cerebral blood flow. Support for the validity of our measure is, however, provided by more dynamic and acute indexes of cerebrovascular resistance. Neuroimaging during gas challenge indicates impaired cerebrovascular reactivity to carbon dioxide in hypertensive relative to normotensive individuals (Hajjar et al., 2010). This is consistent with the increases in cerebrovascular resistance we observed in individuals with higher blood pressure (i.e. we would expect higher vascular resistance in the brain to reduce capacity for cerebrovascular response to blood pressure and CO2 changes), although it is possible that autoregulatory mechanisms responsible for hypercapnic hyperaemia differ from those underlying more customary maintenance of basal CBF (White et al., 1998), particularly in older adults (Kamper et al., 2004). Our index should thus be regarded as a relative estimate that may improve understanding of blood pressure contributions to regional CBF changes observed in Alzheimer’s disease.

The study entailed a large number of analyses, necessitating a correction for multiple comparisons. While statistical significance of all results was retained when FDR was controlled at 0.10, a number of findings were no longer significant under 0.05 FDR. This did not change the overall pattern of results but merely reduced the number of brain regions involved. Furthermore, 80% of results remained significant or trends even at 0.05 FDR (P < 0.06). High attrition rates across follow-up visits also 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 12 and 24 months 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, and studied at varying follow-up intervals, may also have limited generalization of findings to community samples. In addition, the exact methodology employed for blood pressure measurement across ADNI sites is unclear. Optimal acquisition would use a standardized procedure (e.g. three seated measurements, the first of which is discarded). Whether such methods were used and/or whether multiple readings were collected is unknown. Comparisons with other studies of blood pressure should thus be approached with caution. Lastly, selection criteria employed by the 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.

Overall, our results are consistent with growing evidence implicating cerebrovascular dysfunction in Alzheimer’s disease. Recent studies have shown that CBF and amyloid imaging are equivalent in prediction of amyloid retention (Tosun et al., 2016), and age-related CBF declines appear to be present only where amyloid load is pathologically elevated (Gietl et al., 2015). Moreover, disruption of CBF in APP knock-in mice produces both cerebral amyloid angiopathy and elevated parenchymal amyloid deposition (Li et al., 2014). Our results extend these findings by identifying both interactive and independent effects of cerebrovascular dysfunction and cerebral amyloid burden on cortical atrophy and cognitive decline, and suggest that cerebrovascular dysfunction may occur earlier than previously thought in the Alzheimer’s cascade.

Web resources

ADNI. PET Technical Procedures Manual: AV-45 (Florbetapir F18) & FDG. 2011 [cited 2015 15th March]; Available from: http://adni.loni.usc.edu/wp-content/uploads/2010/05/ADNI2_PET_Tech_Manual_0142011.pdf

Funding

The study data analysis was supported by NIH grants (P50 AG005142 and P01 AG052350). Data collection and sharing for this project was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Biogen; Bristol-Myers Squibb Company; CereSpir, Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Therapeutic Research Institute at the University of Southern California. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

Supplementary material

Supplementary material is available at Brain online.

Supplementary Material

Supplementary Data
Click here for additional data file. (85.1KB, docx)

Glossary

Abbreviations

ADNI

Alzheimer’s Disease Neuroimaging Initiative

CBF

cerebral blood flow

CVRi

cerebrovascular resistance index

IPC

inferior parietal cortex

ITC

inferior temporal cortex

mOFC

medial orbitofrontal cortex

MCI

mild cognitive impairment

MMSE

Mini-Mental Status Examination

RAVLT

Rey Auditory Verbal Learning Test

rMFG

rostral middle frontal gyrus

WMH

white matter hyperintensity

References

  1. Alsop DC, Dai W, Grossman M, Detre JA. Arterial spin labeling blood flow MRI: its role in the early characterization of Alzheimer’s disease. J Alzheimers Dis 2010; 20: 871–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alvarez JA, Emory E. Executive function and the frontal lobes: a meta-analytic review. Neuropsychol Rev 2006; 16: 17–42. [DOI] [PubMed] [Google Scholar]
  3. Andersson C, Lindau M, Almkvist O, Engfeldt P, Johansson SE, Eriksdotter Jönhagen M. Identifying patients at high and low risk of cognitive decline using Rey Auditory Verbal Learning Test among middle-aged memory clinic outpatients. Dement Geriatr Cogn Disord 2006; 21: 251–9. [DOI] [PubMed] [Google Scholar]
  4. Ashendorf L, Jefferson AL, ÓÄôConnor MK, Chaisson C, Green RC, Stern RA. Trail making test errors in normal aging, mild cognitive impairment, and dementia. Arch Clin Neuropsychol 2008; 23: 129–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bell RD, Deane R, Chow N, Long X, Sagare A, Singh I. et al. SRF and myocardin regulate LRP-mediated amyloid-beta clearance in brain vascular cells. Nat Cell Biol 2009; 11: 143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Methodol 1995; 57: 289–300. [Google Scholar]
  7. Bookheimer SY, Strojwas MH, Cohen MS, Saunders AM, Pericak-Vance MA, Mazziotta JC. et al. Patterns of brain activation in people at risk for Alzheimer’s disease. N Engl J Med 2000; 343: 450–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Braak H, Braak E, Bohl J. Staging of Alzheimer-related cortical destruction. Eur Neurol 1993; 33: 403–8. [DOI] [PubMed] [Google Scholar]
  9. Breteler MM. Vascular risk factors for Alzheimer’s disease: an epidemiologic perspective. Neurobiol Aging 2000; 21: 153–60. [DOI] [PubMed] [Google Scholar]
  10. Chételat G, Villemagne VL, Pike KE, Ellis KA, Bourgeat P, Jones G. et al. Independent contribution of temporal β-amyloid deposition to memory decline in the pre-dementia phase of Alzheimer’s disease. Brain 2011; 134: 798. [DOI] [PubMed] [Google Scholar]
  11. Cho H, Choi JY, Hwang MS, Kim YJ, Lee HM, Lee HS. et al. In vivo cortical spreading pattern of tau and amyloid in the Alzheimer disease spectrum. Ann Neurol 2016; 80: 247–58. [DOI] [PubMed] [Google Scholar]
  12. Dai WY, Lopez OL, Carmichael OT, Becker JT, Kuller LH, Gachm HM. Mild cognitive impairment and Alzheimer disease: patterns of altered cerebral blood flow at MR imaging. Radiology 2009; 250: 856–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. den Abeelen AS, Lagro J, van Beek AH, Claassen JA. Impaired cerebral autoregulation and vasomotor reactiity in sporadic Alzheimer’s disease. Curr Alzheimer Res 2014; 11: 11–17. [DOI] [PubMed] [Google Scholar]
  14. Dickerson BC, Wolk DA; Alzheimer’s Disease Neuroimaging Initiative. Dysexecutive versus amnesic phenotypes of very mild Alzheimer’s disease are associated with distinct clinical, genetic and cortical thinning characteristics. J Neurol Neurosurg Psychiatry 2011; 82: 45–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Drachman DA. The amyloid hypothesis, time to move on: amyloid is the downstream result, not cause, of Alzheimer’s disease. Alzheimers Dement 2014; 10: 372–80. [DOI] [PubMed] [Google Scholar]
  16. Du AT, Jahng GH, Hayasaka S, Kramer JH, Rosen HJ, Gorno-Tempini ML. et al. Hypoperfusion in frontotemporal dementia and Alzheimer disease by arterial spin labeling MRI. Neurology 2006; 67: 1215–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fischl B, Salat DH, van der Kouwe AJ, Makris N, Segonne F, Quinn BT. et al. Sequence-independent segmentation of magnetic resonance images. Neuroimage 2004; 23 (Suppl 1): S69–84. [DOI] [PubMed] [Google Scholar]
  18. Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–98. [DOI] [PubMed] [Google Scholar]
  19. Gietl AF, Warnock G, Riese F, Kälin AM, Saake A, Gruber E. et al. Regional cerebral blood flow estimated by early PiB uptake is reduced in mild cognitive impairment and associated with age in an amyloid-dependent manner. Neurobiol Aging 2015; 36: 1619–28. [DOI] [PubMed] [Google Scholar]
  20. Hajjar I, Sorond F, Lipsitz LA. Apolipoprotein E, Carbon Dioxide vasoreactivity, and cognition in older adults: effect of hypertension. J Am Geriatr Soc 2015; 63: 276–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hajjar I, Zhao P, Alsop D, Novak V. Hypertension and cerebral vasoreactivity: a continuous arterial spin labeling magnetic resonance imaging study. Hypertension 2010; 56: 859–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Han BH, Zhou ML, Johnson AW, Singh I, Liao F, Vellimana AK. et al. Contribution of reactive oxygen species to cerebral amyloid angiopathy, vasomotor dysfunction, and microhemorrhage in aged Tg2576 mice. Proc Natl Acad Sci USA 2015; 112: E881–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hughes TM, Kuller LH, Barinas-Mitchell EM, McDade EM, Klunk WE, Cohen AD. et al. Arterial stiffness and β-amyloid progression in nondemented elderly adults. JAMA Neurol 2014; 71: 562–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Iadecola C. Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat Rev Neurosci 2004; 5: 347–60. [DOI] [PubMed] [Google Scholar]
  25. Iliff JJ, Goldman SA, Nedergaard M. Implications of the discovery of brain lymphatic pathways. Lancet Neurol 2015; 14: 977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Iturria-Medina Y, Sotero RC, Toussaint PJ, Mateos-Pérez JM, Evans AC. Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis. Nat Commun 2016; 7: 11934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jedynak BM, Lang A, Liu B, Katz E, Zhang Y, Wyman BT. et al. A computational neurodegenerative disease progression score: method and results with the Alzheimer‚Äôs disease neuroimaging initiative cohort. Neuroimage 2012; 63: 1478–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Joshi AD, Pontecorvo MJ, Clark CM, Carpenter AP, Jennings DL, Sadowsky CH. et al. Performance characteristics of amyloid PET and florbetapir F 18 in patients with Alzheimer’s disease and cognitively normal subjects. J Nucl Med 2012; 53: 378–84. [DOI] [PubMed] [Google Scholar]
  29. Kamper AM, Spilt A, de Craen AJM, van Buchem MA, Westendorp RGJ, Blauw GJ. Basal cerebral blood flow is dependent on the nitric oxide pathway in elderly but not in young healthy men. Exp Gerontol 2004; 39: 1245–8. [DOI] [PubMed] [Google Scholar]
  30. Kanekiyo T, Xu H, Bu G. ApoE and Aβ in Alzheimer’s disease: accidental encounters or partners? Neuron 2014; 81: 740–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kress BT, Iliff JJ, Xia M, Wang M, Wei HS, Zeppenfeld D. et al. Impairment of paravascular clearance pathways in the aging brain. Ann Neurol 2014; 76: 845–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Landau SM, Harvey D, Madison CM, Koeppe RA, Reiman EM, Foster NL. et al. Associations between cognitive, functional, and FDG-PET measures of decline in AD and MCI. Neurobiol Aging 2011; 32: 1207–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Langbaum JB, Chen K, Launer LJ, Fleisher AS, Lee W, Liu X. et al. Blood pressure is associated with higher brain amyloid burden and lower glucose metabolism in healthy late middle-age persons. Neurobiol Aging 2012; 33: 827.e11–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Li H, Guo Q, Inoue T, Polito VA, Tabuchi K, Hammer RE. et al. Vascular and parenchymal amyloid pathology in an Alzheimer disease knock-in mouse model: interplay with cerebral blood flow. Mol Neurodegener 2014; 9: 28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Liu J, Zhu YS, Khan MA, Brunk E, Martin-Cook K, Weiner MF. et al. Global brain hypoperfusion and oxygenation in amnestic mild cognitive impairment. Alzheimers Dement 2014; 10: 162–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Luchsinger JA, Reitz C, Honig LS, Tang MX, Shea S, Mayeux R. Aggregation of vascular risk factors and risk of incident Alzheimer disease. Neurology 2005; 65: 545–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mattson RH, Tosun D, Insel PS, Simonson A, Jack CR Jr, Beckett LA. et al. Association of brain amyloid-ß with cerebral perfusion and structure in Alzheimer’s disease and mild cognitive impairment. Brain 2014; 137: 1550–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Moll J, Oliveira-Souza Rd, Moll FT, Bramati IE, Andreiuolo PA. The cerebral correlates of set-shifting: an fMRI study of the trail making test. Arq Neuropsiquiatr 2002; 60: 900–5. [DOI] [PubMed] [Google Scholar]
  39. Morra JH, Tu Z, Apostolova LG, Green AE, Avedissian C, Madsen SK. et al. Automated 3D mapping of hippocampal atrophy and its clinical correlates in 400 subjects with Alzheimer’s disease, mild cognitive impairment, and elderly controls. Hum Brain Mapp 2009; 30: 2766–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 1993; 43: 2412–14. [DOI] [PubMed] [Google Scholar]
  41. Musiek ES, Chen Y, Korczykowski M, Saboury B, Martinez PM, Reddin JS. et al. Direct comparison of fluorodeoxyglucose positron emission tomography and arterial spin labeling magnetic resonance imaging in Alzheimer’s disease. Alzheimers Dement 2012; 8: 51–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Musiek ES, Holtzman DM. Three dimensions of the amyloid hypothesis: time, space and ‘wingmen’. Nat Neurosci 2015; 18: 800–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nation DA, Delano-Wood L, Bangen KJ, Wierenga CE, Jak AJ, Hansen LA. et al. Antemortem pulse pressure elevation predicts cerebrovascular disease in autopsy-confirmed Alzheimer’s disease. J Alzheimers Dis 2012; 30: 595–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Nation DA, Edland SD, Bondi MW, Salmon DP, Delano-Wood L, Peskind ER. et al. Pulse pressure is associated with Alzheimer biomarkers in cognitively normal older adults. Neurology 2013a; 81: 2024–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Nation DA, Edmonds EC, Bangen KJ, Delano-Wood L, Scanlon BK, Han SD. et al. Pulse pressure in relation to tau-mediated neurodegeneration, cerebral amyloidosis, and progression to dementia in very old adults. JAMA Neurol 2015a; 72: 546–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Nation DA, Ho J, Yew B. Older adults taking AT1-receptor blockers exhibit reduced cerebral amyloid retention. J Alzheimers Dis 2015b; 50: 779–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Nation DA, Wierenga CE, Clark LR, Dev SI, Stricker NH, Jak AJ. et al. Cortical and subcortical cerebrovascular resistance index in mild cognitive impairment and Alzheimer’s disease. J Alzheimers Dis 2013b; 36: 689–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Nedergaard M. Garbage truck of the brain. Science 2013; 340: 1529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Niwa K, Kazama K, Younkin L, Younkin SG, Carlson GA, Iadecola C. Cerebrovascular autoregulation is profoundly impaired in mice overexpressing amyloid precursor protein. Am J Physiol Heart Circ Physiol 2002; 283: H315–23. [DOI] [PubMed] [Google Scholar]
  50. Qiu C, Winblad B, Fratiglioni L. The age-dependent relation of blood pressure to cognitive function and dementia. Lancet Neurol 2005; 4: 487–99. [DOI] [PubMed] [Google Scholar]
  51. Raz N, Rodrigue KM. Differential aging of the brain: patterns, cognitive correlates and modifiers. Neurosci Biobehav Rev 2006; 30: 730–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Rey A. L’examen psychologique dans les cas d’encéphalopathie traumatique. Arch Psychol 1941; 28: 215–85. [Google Scholar]
  53. Rodrigue KM, Rieck JR, Kennedy KM, Devous MD, Diaz-Arrastia R, Park DC. Risk factors for β-Amyloid deposition in healthy aging: vascular and genetic effects. JAMA Neurol 2013; 70: 600–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Schoenberg MR, Dawson KA, Duff K, Patton D, Scott JG, Adams RL. Test performance and classification statistics for the Rey Auditory Verbal Learning Test in selected clinical samples. Arch Clin Neuropsychol 2006; 21: 693–703. [DOI] [PubMed] [Google Scholar]
  55. Schwarz C, Fletcher E, DeCarli C, Carmichael O. Fully-automated white matter hyperintensity detection with anatomical prior knowledge and without FLAIR. Inf Process Med Imaging 2009; 21: 239–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Seeley WW, Menon V, Schatzberg AF, Keller J, Glover GH, Kenna H. et al. Dissociable intrinsic connectivity networks for saliene processing and executive control. J Neurosci 2007; 27: 2349–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sheline YI, Morris JC, Snyder AZ, Price JL, Yan Z, D’Angelo G. et al. APOE4 allele disrupts resting state fMRI connectivity in the absence of amyloid plaques or decreased CSF Aβ42. J Neurosci 2010; 30: 17035–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Snyder HM, Corriveau RA, Craft S, Faber JE, Greenberg SM, Knopman D. et al. Vascular contributions to cognitive impairment and dementia including Alzheimer’s disease. Alzheimers Dement 2015; 11: 710–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Sole M, Minano-Molina AJ, Unzeta M. A cross-talk between Abeta and endothelial SSAO/VAP-1 accelerates vascular damage and Abeta aggregation related to CAA-AD. Neurobiol Aging 2015; 36: 762–75. [DOI] [PubMed] [Google Scholar]
  60. Spreen O, Strauss E. A compendium of neuropsychological tests. 2nd edn New York, NY: Oxford University Press; 1998. [Google Scholar]
  61. Suri S, Mackay CE, Kelly ME, Germuska M, Tunbridge EM, Frisoni GB. et al. Reduced cerebrovascular reactivity in young adults carrying the APOE ɛ4 allele. Alzheimers Dement 2015; 11: 648–57.e1. [DOI] [PubMed] [Google Scholar]
  62. Thal DR, Rüb U, Orantes M, Braak H. Phases of abeta-deposition in the human brain and its relevance for the development of AD. Neurology 2002; 58: 1791–800. [DOI] [PubMed] [Google Scholar]
  63. Thompson PM, Hayashi KM, de Zubicaray G, Janke AL, Rose SE, Semple J. et al. Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci 2003; 23: 994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Tosun D, Schuff N, Jagust W, Weiner MW. 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. Neurodegener Dis 2016; 16: 87–94. [DOI] [PubMed] [Google Scholar]
  65. Viswanathan A, Rocca WA, Tzourio C. Vascular risk factors and dementia: how to move forward? Neurology 2009; 72: 368–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Weller RO, Boche D, Nicoll JA. Microvasculature changes and cerebral amyloid angiopathy in Alzheimer’s disease and their potential impact on therapy. Acta Neuropathol 2009; 118: 87–102. [DOI] [PubMed] [Google Scholar]
  67. Weller RO, Subash M, Preston SD, Mazanti I, Carare RO. Perivascular drainage of amyloid-ß peptides from the brain and its failure in cerebral amyloid angiopathy and Alzheimer’s disease. Brain Pathol 2008; 18: 253–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. White RP, Deane C, Vallance P, Markus HS. Nitric oxide synthase inhibition in humans reduces cerebral blood flow but not the hyperemic response to hypercapnia. Stroke 1998; 29: 467. [DOI] [PubMed] [Google Scholar]
  69. Wolters FJ, de Bruijn RF, Hofman A, Koudstaal PJ, Ikram MA. cerebral vasoreactivity, apolipoprotein E, and the risk of dementia: a population-based study. Arterioscler Thromb Vasc Biol 2016; 36: 204–10. [DOI] [PubMed] [Google Scholar]
  70. Wu W, Brickman AM, Luchsinger J, Ferrazzano P, Pichiule P, Yoshita M. et al. The brain in the age of old: the hippocampal formation is targeted differentially by diseases of late life. Ann Neurol 2008; 64: 698–706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Zakzanis KK, Mraz R, Graham SJ. An fMRI study of the trail making test. Neuropsychologia 2005; 43: 1878–86. [DOI] [PubMed] [Google Scholar]
  72. Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci 2011; 12: 723–38. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
Click here for additional data file. (85.1KB, docx)

Articles from Brain are provided here courtesy of Oxford University Press

RESOURCES

OSZAR »