A Cross-Sectional Study on the Impact of Arterial Stiffness on the Corpus Callosum,a Key White Matter Tract Implicated in Alzheimer’s Disease

Abstract

Background:

Vascular risk factors such as arterial stiffness play an important role in the etiology of Alzheimer’s disease (AD), presumably due to the emergence of white matter lesions. However, the impact of arterial stiffness to white matter structure involved in the etiology of AD, including the corpus callosum remains poorly understood.

Objective:

The aims of the study are to better understand the relationship between arterial stiffness, white matter microstructure, and perfusion of the corpus callosum in older adults.

Methods:

Arterial stiffness was estimated using the gold standard measure of carotid-femoral pulse wave velocity (cfPWV). Cognitive performance was evaluated with the Trail Making Test part B-A. Neurite orientation dispersion and density imaging was used to obtain microstructural information such as neurite density and extracellular water diffusion. The cerebral blood flow was estimated using arterial spin labelling.

Results:

cfPWV better predicts the microstructural integrity of the corpus callosum when compared with other index of vascular aging (the augmentation index, the systolic blood pressure, and the pulse pressure). In particular, significant associations were found between the cfPWV, an alteration of the extracellular water diffusion, and a neuronal density increase in the body of the corpus callosum which was also correlated with the performance in cognitive flexibility.

Conclusion:

Our results suggest that arterial stiffness is associated with an alteration of brain integrity which impacts cognitive function in older adults.

Keywords

Aging Alzheimer’s disease arterial spin labelling arterial stiffness corpus callosum diffusion weighted imaging

INTRODUCTION

Vascular risk factors play an important role in the etiology of Alzheimer’s disease (AD), presumably due to the emergence of white matter lesions [1]. AD and dementia in general are developing into one of the major public health problems of the 21st century since the proportion of older individuals is on the rise [2, 3]. In 2010, approximately 35.6 million people suffered from dementia around the world and this number is expected to double every 20 years [4]. These alarming predictions call for a better understanding of the mechanisms underpinning cognitive decline in the elderly. Currently, there is no cure nor disease-modifying treatment for AD. Thus, reducing the risk of developing dementia through controlling modifiable risk factors takes on added importance [5].

The key risk factors for dementia are old age, low education, APOE ɛ4 allele, and cardiovascular risk factors [2 , 6–8], among which the downstream effect of arterial stiffness on cerebral integrity has gained considerable research interest over the past few years [9 –11]. Furthermore, arterial stiffness which is associated with hypertension and aging, has attracted a lot of attention for its impact on white matter [12, 13]. However, the impact of arterial stiffness on white matter structure involved in the etiology of AD, including the corpus callosum remains poorly understood.

Arterial stiffness of large vessels refers to their reduced capacity for dampening blood pulsatility arising from the heart during each contraction [2]. As a result, arterial pulse waves arrive at the periphery of the arterial system with higher magnitudes causing arterial wave reflections to arrive back at the aorta during the systole rather than the diastole phase of the cardiac cycle thereby escalating systolic blood pressure (SBP) [14, 15]. During aging, diastolic blood pressure (DBP) is also known to decrease. These two factors lead arterial stiffness to be associated with increased pulse pressure (PP) (the difference between SBP and DBP) [16].

Given its potential impact on the microcirculation, non-invasive methods have been developed to measure arterial stiffness. Among these methods, carotid-femoral pulse wave velocity (cfPWV) is considered to be the gold standard [17]. One can also examine the arterial waveform at the radial site in order to calculate the Augmentation Index (AIx), a ratio derived from the augmentation of the PP, which considered to be an indicator of arterial stiffness [18]. It is worth noting that given their close relationship, the cfPWV, AIx, SBP, and PP fall under the same umbrella and can be referred as index of vascular aging [19].

Arterial stiffness has also been shown to be associated with ischemic white matter hyperintensities (WMHs) manifesting on T2-weighted and fluid-attenuated inversion recovery (FLAIR) images [12, 13]. These white matter lesions are often associated with mild cognitive impairment and dementia [20 –23]. Whether these WMHs are reversible or not is still unclear, however, they may be preceded by microstructural changes that can be detected with diffusion tensor imaging (DTI) [13, 24].

One of the earliest reports looking at the association of arterial stiffness and DTI metrics has showed that arterial stiffness is associated with altered fractional anisotropy (FA) and radial diffusivity (RD) in four vulnerable regions [9] among which is the corpus callosum. These findings have been confirmed by similar voxel- and tract-wise analysis [10 , 25]. Authors of these works speculated that arterial stiffness leads to axonal demyelination. However, diffusion MRI is although sensitive to myelin, far from specific. Other MRI contrasts such as magnetization transfer (MT) imaging sensitize the MR signal to hydrogen bound to macromolecules and thus provide a means to estimate myelin volume fraction [26, 27]. By using DTI metric and MT to assess axonal integrity and myelination of nerve fibers respectively, we previously showed that arterial stiffness as assessed by cfPWV was associated with axonal degeneration rather than demyelination [10]. However, DTI metrics lack specificity, which makes our interpretation of the effect of arterial stiffness on the microstructure difficult [28]. Fortunately, recent advances in diffusion imaging have provided more sophisticated models [28] from which one popular example is the neurite orientation dispersion and density imaging (NODDI) [29]. NODDI enables the estimation of separate water compartments: intracellular, extracellular, and isotropic (i.e., cerebrospinal fluid) that can be combined to estimate the neuronal density (intracellular volume fraction, ICVF) and extracellular water diffusion (isotropic volume fraction, ISOVF). As such, these metrics can be used to brings us one step closer to a better understanding of the association between arterial stiffness and the underlying white matter microstructure.

Moreover, the integrity of the brain has a critical dependence on a nearly impeccable cardiovascular supply [30]. This means that even slight hemodynamic changes due to arterial stiffness may reduce cerebral blood flow (CBF) and consequently brain integrity, in particular in vulnerable white matter fibers. However, no prior study has looked specifically at the association between arterial stiffness and blood flow (BF) in white matter regions previously denoted vulnerable to arterial stiffness. Nevertheless, such changes in CBF could be quantified using the magnetic tagging of the inflowing blood before it reaches the capillary bed. Based on the subtraction of tagged and untagged cerebral images, arterial spin labeling (ASL) can be used to derive several perfusion metrics, such as a quantitative CBF map [31]. This map can be used to better understand the impact of peripheral vascular aging on the CBF of vulnerable regions.

Finally, there is accumulating evidence suggesting that determinants of vascular aging (such as arterial stiffness) may impact the trajectory of cognition later in life [12], in particular in domains such as processing speed, executive skills, and cognitive flexibility [12, 18]. However, the association between vascular aging, white matter microstructural integrity, CBF, and cognitive decline in the elderly is still unclear.

The present study aims to investigate the relationship between arterial stiffness, cognitive performance, NODDI metrics (ISOVF and ICVF), and perfusion (BF) measured in the corpus callosum. The corpus callosum has been chosen as a region of interest not only because it is one of the most vulnerable regions to arterial stiffness but also because changes in the corpus callosum may impact the trajectory of cognition later in life [32, 33].

Firstly, we propose that vascular aging is associated with an alteration of the microstructural integrity (ISOVF, ICVF) and BF of the corpus callosum. In particular, we hypothesized that vascular aging is associated with an increase in the extracellular water diffusion (ISOVF) and a decrease in neuronal density (ICVF) and BF of the corpus callosum. Secondly, considering the close relationship between the gold standard measure of arterial stiffness (cfPWV) and other markers of vascular aging (SBP, PP, and AIx), we sought to identify which index of vascular aging best predicts blood flow and microstructural integrity in the corpus callosum. We hypothesized that the ISOVF, ICVF, and BF of the corpus callosum are best predicted by cfPWV as compared to other indexes of vascular aging (24h SBP, 24h PP, and AIx). Finally, as a recent literature review [13] supports the idea that white matter regions vulnerable to arterial stiffness contribute mainly to poorer performance in executive skills and cognitive flexibility, we wanted to investigate the relationship between our MRI metrics and the performance in executive function. We hypothesized that increased neuronal density and BF have a positive association with performance in cognitive flexibility.

MATERIAL AND METHODS

Study participants

Seventy-three healthy French speakers between 65–75 years of age were recruited from the bank of participants of the Centre de Recherche de l’Institut Universitaire de Gériatrie de Montréal (CRIUGM). The ethics approval was obtained from the ethical review board on the CRIUGM and the Montréal Clinical Research Institute (IRCM). An informed consent approved by the local committee was signed by each participant. Candidate eligibility was assessed during a telephone-based screening interview. Exclusion criteria included: major psychiatric or neurological disorders, malignant hypertension (i.e., >180/120 mmHg), diabetes mellitus, heart failure (NYHA III-IV), myocardial infarction (in the previous 3 months), cardiac arrhythmia, rheumatic mitral valve disease, diabetes, liver failure, renal failure (creatinine clearance of less than 30 mL/min), stroke, non-compensated thyroid disorder, respiratory problems (i.e., asthma, emphysema), metallic implants and/or pacemaker, claustrophobia, a current or history of alcohol or drug abuse and a Mini-Mental State Examination (MMSE) score below 26. From all recruited participants, 12 participants voluntarily withdrew themselves from the study for personal reasons, 2 were excluded, and 1 did not complete the study. Among the 58 participants left from which we obtained the MRI images, 1 was excluded due to incidental findings and 4 were excluded from this study because of missing data. Therefore, for this study, only the remaining 53 participants were included. The rule of thumb for our analyses is that a minimum of 10 observation is required per variable to produce a reliable parameter estimate [34 –36]. Among all 53 participants included in this study, 14 were on antihypertensive medications, 7 participants were on angiotensin-converting-enzyme inhibitor, 3 were on calcium channel blockers, 2 were on angiotensin-converting-enzyme inhibitor and calcium channel blockers, 1 was on angiotensin-converting-enzyme inhibitor and hydrochlorothiazide, and 1 was on calcium channel blockers and hydrochlorothiazide.

The characteristics of all participants have been previously published [10] and include a comparison of demographics, cardiovascular, neuropsychological, and MRI measures between drug-naive participants and participants on antihypertensive medication. No significant difference was found between MRI metrics, thus all 53 participants have been included in our study.

Cardiovascular measurements

Noninvasive hemodynamic data acquisition was performed by a trained expert (Adrián Noriega, MD) using an established reproducible protocol between 9 am–1 pm. All physiological measurements were performed in an environmentally controlled laboratory using an established and reproducible protocol. All subjects abstained from alcohol, caffeinated beverages, and intense physical activity for at least 3 h prior to data collection as required by the Van Bortel protocol [17].

The cfPWV values were calculated from tonometry waveforms using the SphygmoCor system (AtCor Medical Pty Ltd) as described previously [10]. In brief, the SphygmoCor system used the foot of the waveform as the onset point for calculating the time difference between the R wave of the ECG and the pulse waveforms at each site: the carotid and the femorale sites. The cfPWV was then automatically calculated by the SphygmoCor system by dividing the arterial pulse traveling distance by the transit time and expressed in meters per second. The arterial pulse traveling distance was measured as the straight distance between the right common carotid and common femoral arteries measurement sites using a tape ruler. The transit time was determined from the time delay between proximal and distal “foot” waveforms”. The cfPWV measures were done twice according to the Van Bortel protocol. If the difference between the two measures was larger than 0.5 m/s, a third measure was taken [17]. Recently, a risk treshold of cfPWV of 10 m/s has been suggested by a study group on behalf of the Artery Society, the European Society of Hypertension Working Group on Vascular Structure and Function, and the European Network for Noninvasive Investigation of Large Arteries [37]. This cut-off is actually of 9.6 m/s but has been set it up to 10 m/s for an easier implementation in clinical practice [17].

Prior to the measure of cfPWV, a pulse wave analysis was performed at the radial artery to calculate the AIx as described elsewhere [38]. Two pressure peaks characterized the arterial waveform. The first peak resulted from the left cardiac ventricle ejection whereas the second one resulted from the wave reflections from the periphery. The difference between these two pressure peaks reflected the degree of the arterial pressure augmentation due to wave reflection. The AIx is defined as the ratio of the increase of the second (wave reflection) to the first peak (arterial pressure waveform) of the pressure wave and is expressed as a percentage.

In addition, a 24 h Ambulatory Blood Pressure Monitoring 90207-3Q model (Spacelabs Healthcare^®) was installed in the patient’s non-dominant arm following the recommendation of Hypertension Canada [39], in order to get a 24 h measures of SBP and a 24 h measure of PP.

To calibrate the Ambulatory Blood Pressure Monitoring device, conventional blood pressure was measured manually three times by sphygmomanometer (Korotkoff phase I and IV) after subjects rested in a seated position for at least 15 minutes.

Neurocognitive assessments

The Trail Making Test (TMT) was administered by a trained expert (Adrián Noriega, MD) and consists of two parts (A and B). In the first part of the test (TMTA), the participants had to connect a series of numbers in correct order as fast as possible (i.e., 1–2–3, etc.) whereas in the second part of the test (TMTB), the participants had to connect numbers and letters by matching them in ascending order (i.e., 1–A–2–B, etc.) as fast as possible. In this study, we considered the difference between the time spent in completing the TMTB and TMTA in seconds as a measure of cognitive flexibility [40, 41]. More precisely, the cost of switching task was measured by subtracting the reaction time TMTA from TMTB, allowing us to isolate the switching component of the TMT [42]. This means that a lower score in the TMTB-A implies a better cognitive flexibility performance.

Brain MRI analysis

MRI acquisition

MRI data were acquired at 3T (MAGNETOM Prisma Fit, Siemens Healthineers, Erlangen, Germany) using a 32-channel head receive coil.

Diffusion-weighted images were acquired using echo planar imaging with factor = 3 simultaneous multislice acceleration [43]. Axial slices with 2.0 mm isotropic resolution were acquired parallel to the posterior-anterior commissure line, encoded with 3 b-value shells: 300/1k/2k s/mm², along with 7/29/64 directions, 9 b = 0 images, and a scan duration of 4.37 min. Other acquisition parameters were: imaging Echo Time (TE) = 63 ms, Repetition Time (TR) = 2200 ms, and FOV = 220 × 220 mm².

A pCASL acquisition was acquired at rest according to Wu et al. [44] using single-echo readout with the following parameters: TR = 3850 ms, TE = 8.2 ms, voxel size = 5.0 mm isotropic, number of slices = 20, post-labeling delay = 1.55 s, tag duration = 1.5 s, flip angle = 90°, FOV = 320 mm × 320 mm². Scan duration was 4.03 min.

A T1 weighted image was acquired using an MPRAGE sequence. Sagittal slices of 0.8 mm isotropic were acquired with a TR = 2400.0 ms, TE = 2.33 ms, T1 = 979 ms, and flip angle = 9° and FOV = 230 mm × 230 mm². Scan duration was 7.02 min.

MRI data processing

Computation of NODDI metrics and cerebral BF. Processing of diffusion data was done using the Toolkit for Analysis in Diffusion MRI (TOAD) (https://github.com/UNFmontreal/toad) and included motion and eddy current distortion correction as well as denoising using an MP-PCA algorithm [45]. After pre-processing, the diffusion data was fed into NODDI modeling, using the AMICO (Accelerated Microstructure Imaging via Convex Optimization) tool [46]. The aim was to generate meaningful voxelwise microstructural parameters, including ISOVF and ICVF. To minimize partial volume effect, we conservatively discarded voxels with FA < 0.2 since these voxels are considered to be potentially dominated by gray matter or CSF partial voluming.

Resting CBF was quantified using an open-source pipeline available on Github (https://github.com/SaraDupont/ASL_processing). Processing of pCASL data included i) motion correction, ii) denoising of each frame, iii) intensity normalization of each individual frame based on the intensity of the 1st control frame, iv) computation of the average of control frames and tag frames (separately) to increase SNR, and v) computation of the difference Control-Tag to get a perfusion weighted image [47]. Finally, the resting cerebral BF map has been calculated according to Alsop et al. [48] using the perfusion weighted image and proton density image (also called M0) assuming a blood-brain partition coefficient = 0.90, a label efficiency = 0.90, and blood T1 = 1.65 s. Considering the fact that the range of CBF in the white matter is between 20–40 mL/100 g/min [49], only voxels with a CBF > 0 and <40 were included in our analysis to avoid the majority of voxels at the interface of the corpus callosum.

Registration of the corpus callosum on NODDI and BF metrics

The analysis of the corpus callosum was performed in the MNI space using the deep WM atlas (ICBM-DTI-81 WM atlas) [50], which is composed of 48 WM tract labels, including the corpus callosum.

To perform the registration, the FA maps computed for each participant were first aligned to the Johns Hopkins International Consortium for Brain Mapping (ICBM) FA template using a multi-step and multi-resolution registration by means of the ANTs non-linear registration tools [51]. Then, a Laplacian transform was applied on the 3D volumes (kernel size: 2 mm) to help the registration by enhancing the edge at the interface between white and grey matter. The resulting volumes were registered to the ICBM FA template by means of a BsplineSyN transformation. All steps have been thoroughly described previously [10]. The same transformations were used to align the ISOVF and ICVF to the template.

The ICBM-DTI-81 WM atlas was then registered to the raw pCASL data. Firstly, brain extraction was applied to the pCASL data using an atlas-based method 2 . Then, the MNI152 T1 template was registered to the pCASL space using a multi-step registration process: i) linear registration steps using a block-matching algorithm [52], with rigid, then affine transformations, ii) a non-linear anatomical registration step as described in [53]. The same transformation was then applied to the ICBM-DTI-81 WM atlas.

Statistical analysis

To test the first hypothesis that arterial stiffness is associated not only with the microstructural integrity (ISOVF, ICVF) but also with the BF of the corpus callosum, a partial correlation between cfPWV and either MRI metrics independently (ISOVF, ICVF, and BF) was performed with age and sex as covariates. To investigate which subregions of the corpus callosum have the strongest associations with cfPWV, a partial correlation analysis was performed between cfPWV and either ISOVF or ICVF in the genu, the body, and the splenium of the corpus callosum respectively adjusting for age and sex. This was motivated by the fact that several studies assessing age-related microtrustructral alterations have been previously shown that changes in FA appear sooner in the genu and body of the corpus callosum than in any other white matter regions of the brains [54 –56]. The ASL global analysis of the corpus callosum instead of subregional was performed due to the known limitations of ASL discussed in the manuscript.

To test the second hypothesis that the ISOVF, ICVF and BF of the corpus callosum are best predicted by cfPWV compared to other index of vascular aging (24 h SBP, 24 h PP, and AIx), a stepwise linear model was performed. For this analysis, each MRI metrics was treated as a dependent variable and all the following variables were considered as independent variables: cfPWV, 24 h SBP, 24 h PP, and AIx. Age and sex were forced into the model to control for biological factors.

To test the third hypothesis that the microstructural integrity (ISOVF, ICVF) and the BF of the corpus callosum are associated with cognitive flexibility, a partial correlation analysis was performed with the MRI metrics defined as independent variables and the TMTB-A defined as the dependent variable. In this partial correlation analysis, age, sex, and education levels were considered as covariates.

We tested the presence of outliers in each variable and removed three outliers using the 1.5 interquartile rule in the cfPWV and TMTB-A data prior to data analysis. All statistical tests were implemented in R version 3.5.2. and SPSS (IBM SPSS 25 Statistics, Chicago, IL). Correction for multiple comparisons was performed using the false discovery rate (FDR) procedure as described in Benjamini-Hochberg et al. [57]. Since the p-threshold adjustment for significance under the Benjamini-Hochberg procedure not only depends on the number of tests but also on the calculated p-value for each test, original p-values were reported and only the significant ones after correction were highlighted. In addition, the adjusted threshold under which a p-value is declared significant after the FDR correction is shown for each case in the figure legend.

To allow for good reproducibility of this study, the data and code to perform all statistical analyses and figures is publicly available at https://github.com/atefbadji/cfPWV_Microstructure-Perfusion_CorpusCallosum.

RESULTS

Arterial stiffness-associated parameters, white matter neuronal integrity, and BF in the corpus callosum

Table 1 shows participants MRI characteristics including a comparison among normotensive drug naive participants and those on antihypertensive treatments.

Table 1

Sample characteristics of all participants including a comparisons of drug naive participants and those on Anti-HT Treatment

	All	On Anti-HT Treatment	Drug naïve	p
Women/men (N)	39/14 (53)	7/7(14)	30/9 (39)	0.190
Age (y)	69.900±3.341	69.961±3.138	69.878±3.454	0.936
Education (y)	16.260±3.451	17.769±3.370	15.729±3.363	0.074
Neurocognitive measures
TMTB-A (s)	36.360±17.348	39.000±16.206	35.432±17.852	0.512
Cardiovascular measures
24 h SBP (mmHg)	126.160±11.741	127.846±11.985	125.567±11.762	0.559
24 h DBP (mmHg)	73.880±7.283	75.384±6.278	73.351±7.612	0.352
24 h PP (mmHg)	52.400±8.315	52.461±7.287	52.378±8.741	0.973
24 h AIx (mmHg)	32.380±8.852	30.357±7.333	33.076±9.412	0.280
cfPWV (m/s)	9.248±2.298	9.384±2.627	9.200±2.209	0.823
MRI measures
ISOVF_cc	0.208±0.025	0.211±0.027	0.206±0.025	0.861
ICVF_cc	0.672±0.033	0.671±0.019	0.672±0.037	0.827
BF_cc	18.172±2.997	18.190±3.907	18.172±2.672	0.793

Values are mean±standard deviation. ISOVF_cc, isotropic volume fraction of the corpus callosum; ICVF_cc, intracellular volume fraction of the corpus callosum; BF_cc, blood flow of the corpus callosum.

To answer our first hypothesis, that arterial stiffness is associated with the microstructural integrity (ISOVF, ICVF) and with the BF of the corpus callosum, Fig. 1 shows a significant correlation between cfPWV and both the ISOVF and ICVF measured in the corpus callosum (r = 0.478, p = 0.00059 and r = 0.339, p = 0.018 respectively). This means that an increase in arterial stiffness was found to be associated with a higher neuronal density and extracellular water diffusion of the white matter fibers of the corpus callosum. However, no significant correlation was found between cfPWV and the BF of the corpus callosum (p = 0.065). We decided to provide scatter plots (Figs. 1 –4) with data points according to the cfPWV cut-off value of 9.6 m/s according to European scientific societies [51] to help the reader better grasp the results.

Fig. 1

Scatter plot for partial correlation analysis of carotid-femoral pulse wave velocity (cfPWV) with either the isotropic volume fraction (ISOVF), the intracellular volume fraction (ICVF) or the blood flow (BF) along the corpus callosum. Significant p-values are in bold font (adjusted threshold with FDR was 0.018). Covariates included age and sex.

Fig. 2

Scatter plot for partial correlation analysis of carotid-femoral pulse wave velocity (cfPWV) with either the isotropic volume fraction (ISOVF) and the intracellular volume fraction (ICVF) along the anterior part, the genu and the splenium of the corpus callosum (CC). Significant p-values are in bold font (adjusted threshold with FDR was 0.025 and 0.05 and 0.025 for the genu, the body and the splenium of the CC, respectively). Covariates included are age and sex.

Fig. 3

Scatter plot for the partial correlation analysis of TMTB-A with either the isotropic volume fraction (ISOVF), the intracellular volume fraction (ICVF) or the blood flow (BF) along the anterior part, the genu and the splenium of the corpus callosum (CC). Significant p-values are in bold font (adjusted threshold with FDR was 0.016). Covariates included are age and sex.

Fig. 4

Scatter plot for the partial correlation analysis of TMTB-A with either the isotropic volume fraction (ISOVF) or the intracellular volume fraction (ICVF) along the body of the corpus callosum. Significant p-values are in bold font (adjusted threshold with FDR was 0.025). Covariates included age, sex, and years of schooling.

A closer look at the association between cfPWV and MRI metrics reveals that cfPWV was found to be significantly associated with both the ISOVF (r = 0.459, p = 0.001) and ICVF (r = 0.459, p = 0.001) of the body of the corpus callosum (Fig. 2). However, cfPWV was found to be significantly associated only with the ISOVF measured in both the genu and splenium of the corpus callosum (p = 0.008 and p = 0.009, respectively) (Fig. 2).

To test the second hypothesis that the ISOVF, ICVF, and BF of the corpus callosum are best predicted by cfPW among cardiovascular predictor, we performed three stepwise linear regression (Table 2, Supplementary Table 2). In Table 2, model ISOVF1 revealed that age and sex predict ISOVF independently (p = 0.001). This analysis indicated an R² of 0.249, which means that age and sex accounted for about 25% of the total variance in ISOVF in the corpus callosum. Interestingly, model ISOVF2 also highlights that cfPWV is found to be the best predictor of ISOVF among the arterial stiffness-associated parameters included in this study. Indeed, the contribution of arterial stiffness as assessed by cfPWV on top of age and sex was found to account for significantly more variance than the previous model (F change at p < 0.012). The overall model of ISOVF2 (age, sex, and cfPWV) was also found to be a significant predictor of ISOVF (p < 0.001). This analysis indicated an R² of 0.341 which means that age, sex, and cfPWV accounted for 34.1% of the total variance of ISOVF in the corpus callosum. None of the other arterial stiffness-associated parameters included in this study were found to be a significant predictor of ISOVF once cfPWV was taken into account.

Table 2

Result of the stepwise entry linear model based on a priori hypotheses on the influence of age, sex, and cardiovascular risk factors on the integrity of the corpus callosum

Name	Dependent variable	Independent variables	R²	p	F change p
1	ISOVF	Age, sex	0.249	0.001	N/A
2		Age + sex + cfPWV	0.341	<0.001	0.024

Five cardiovascular risk factors were fed to the stepwise linear model: SBP, PP, cfPWV and AIx. Model ISOVF, aimed to find the arterial stiffness-associated parameters that best predict the isotropic volume fraction (ISOVF) of the corpus callosum. Fixed covariates were age and sex. R² of the significant models, the significant overall p-value of the model and the significant F changes when comparing with the previous model are in bold.

In Supplementary Table 2, model ICVF shows that age and sex are not significantly associated with ICVF (p = 0.079). In addition, none of the arterial stiffness-associated parameters of this study were found to improve the prediction of ICVF. Likewise, model BF shows that age and sex were not significantly associated with the level of resting BF in the corpus callosum (p = 0.410) and none of the arterial stiffness-associated parameters improved the prediction.

Linking corpus callosum microstructure, BF, and cognition

Figure 3 shows the results of the partial correlation analysis with cognitive flexibility (TMTB-A) and the microstructure (ISOVF, ICVF) and perfusion (BF) of the corpus callosum, respectively. A significant negative correlation was found between the time to complete the TMTB-A and the neuronal density (ICVF) of the white matter fibers of the corpus callosum (p < 0.0001, r = –0.477). This means that a better performance in cognitive flexibility characterized by a shorter time to complete the TMTB-A, was associated with an increase in neuronal density of the corpus callosum.

A closer look at the association between the performance in TMTB-A and our MRI metrics (ISOVF, ICVF) measured in sub-regions of the corpus callosums (genu, body, and splenium) revealed that the neuronal density (ICVF) of the body of the corpus callosum is associated with the performance of cognitive flexibility (p < 0.001).

DISCUSSION

In this study, we investigated the relationship between arterial stiffness (measured with cfPWV) and the integrity of the corpus callosum in the elderly using 1) NODDI metrics (ISOVF, ICVF) reflecting the white matter microstructure and 2) a CBF metric to account for brain perfusion. An important finding is that arterial stiffness as measured by cfPWV better predicts the microstructural integrity of the corpus callosum when compared with other indices of vascular aging (24 h SBP, 24 h PP, and AIx). In particular, a significant association was found between the cfPWV and the neuronal density and extracellular water diffusion of the white matter fibers of the body of the corpus callosum. In addition, a significant association was observed between the microstructure of the body of the corpus callosum and the performance in cognitive flexibility in our healthy elderly population between 65–75 years old. These results suggest that arterial stiffness may influence brain integrity which, then, may impact cognition in older adults.

Central artery stiffness, a significant predictor variable for extracellular water diffusion and neurite density in the corpus callosum

In the emerging literature where the downstream effects of arterial stiffness on cerebral health are considered, several studies have shown that increased arterial stiffness as assessed by cfPWV is associated with a decrease in FA and increase in RD in the corpus callosum [9 , 25]. However, the cfPWV was not found to be associated with the myelin volume fraction in the corpus callosum [10], suggesting that arterial stiffness is more related to axonal degeneration than it is to demyelination [10]. Nevertheless, dissecting the microstructural reasons behind the changes in DTI metrics in the presence of arterial stiffness remained highly challenging. For example, both decompaction and a decrease in the number of fibers can account for lower FA values [58]. To disentangle this complexity, the present work introduced the use of more specific diffusion metrics to the white matter microstructure: the NODDI metrics (ISOVF, ICVF). Our findings showed that the cfPWV is positively associated with both the extracellular water diffusion (ISOVF) and the neuronal density (ICVF) of the corpus callosum, constituting the first evidence of an association between cfPWV and NODDI metrics in the literature. In particular, our results suggest that the body of the corpus callosum is particularly vulnerable to an increase in arterial stiffness.

A large study with a related aim exists on the comparison of NODDI metrics between 4,659 non-hypertensive, pre-hypertensive, and hypertensive individuals [59]. Suzuki et al. reported in this particular study that the ISOVF in pre-hypertensive individuals is higher than that of non-hypertensive individuals [59]. This finding suggested that an increase in white matter ISOVF may be the earliest microstructural alterations following vascular aging prior to the appearance of established hypertension [59]. However, their study did not control for arterial stiffness, which may be the underlying vascular parameter altering the white matter microstructure [10]. A supporting basis for this hypothesis was provided by a study looking at the effect of cfPWV on white matter microstructure using the free water metric, which quantifies the fraction of water molecules that diffuse equally in all directions such as the ISOVF [11]. The authors reported that an increase in free water in the corpus callosum is associated with increased arterial stiffness and blood pressure, agreeing well with the results of the present work [11]. Interestingly, their analysis further suggested that the increase in free water associated to arterial stiffness precede the changes seen in FA and the appearance of white matter lesions [11]. Altogether, these results suggest that arterial stiffness may result in axonal dispersion, lessening the constraint of water directionality along axons before the appearance of hypertension and irreversible white matter damage [11, 59]. Another potential hypothesis is that the increased in free water following arterial stiffness reflects a vasogenic edema due to an alteration of the blood brain barrier and capillary fluid leakage [60, 61]. Additional studies with a validation of these MRI metrics in an animal model of arterial stiffness could help us better understand the underlying mechanism associating arterial stiffness and the white matter microstructure.

In addition, Suzuki et al. also found that the neuronal density as assessed with the ICVF was lower in hypertensive related to non-hypertensive individuals in several white matter tracts (e.g., the right cingulate gyrus part of the cingulum, the forceps minor, and the bilateral superior and inferior longitudinal fasciculus) [59]. Although Suzuki et al. did not include the corpus callosum as a region of interest, the authors suggested that hypertension is associated with a decrease in axonal density in some white matter tracts. In contrast, our results show that arterial stiffness is associated with an increase in the ICVF of the corpus callosum. Our finding may suggest that individuals at higher risk for cognitive decline demonstrate early compensatory mechanisms before the appearance of atrophy as well as signs of cognitive decline. A recent study conducted by Benitez et al. supports this claim. They compared the axonal density between 69 cognitively intact older adults and 39 matching healthy controls using a Diffusional Kurtosis Estimator to derive metrics of axonal density (axonal water fraction - AWFI) and found that axonal density was higher in the corpus callosum of the older adults, agreeing well with the results of the present study [62]. The authors suggested that the aging brain is capable of compensatory mechanisms to maintain brain homeostasis following an insult. However, as seen in multiple sclerosis patients using connectomics [63, 64], as well as in a transgenic Alzheimer rat lines (TgF344-AD) using the NODDI model [65], compensatory mechanisms may reach a plateau after a certain period of time and then exhibit an inverse trend. This remains to be verified in a longitudinal study aiming to investigate the impact of cardiovascular risk factors on the white matter microstructure and cognitive performance in the elderly.

Central artery stiffness and BF

Vascular integrity is of central importance in the process of autoregulation, stabilizing the global CBF despite changes in blood pressure [66]. Therefore, an increase in the aortic stiffness may alter cerebral perfusion through damage to the microvessels [66]. However, studies examining this relationship have yielded contradicting findings. While more recent reports point to lower regional BF in the gray matter of individuals with increased arterial stiffness [67, 68], earlier studies did not find such association [25 , 70]. This may be due to different study designs (cross-sectional versus longitudinal) and/or heterogeneity of target populations (young adult vs older adults) and/or differences in means to measure arterial stiffness as well as brain perfusion [13].

The impact of arterial stiffness in the perfusion of white matter has been examined less often than that of gray matter [69]. Tarumi et al. evaluated regional BF in frontal and parietal white matter as well as in the centrum semiovale in 37 adults (40–60 years old) with no cardiovascular or neurological disease and found that increased cfPWV was significantly associated with lower BF in frontal and parietal white matter [69]. Their findings constituted a first step towards the understanding of the mechanisms involved in white matter damage following vascular aging [66]. In the present study, we did not find a significant correlation between cfPWV and the BF of the corpus callosum. However, considering the known limitations of ASL to accurately measure the BF in the white matter discussed in the limitation section, additional studies with more advanced ASL technique assessing the BF in the white matter in the presence of increased arterial stiffness are needed for advancing the literature.

Considering the fact that arterial stiffening leads to an increase in arterial pulsatility in small vessels (due to a reduced dampening of the pulsatility in large vessels), increased research focus looking at the impact of arterial pulsatility on white matter BF would further help to better understand the association between vascular aging and brain perfusion.

The association between neuronal density and cognitive function

In general, white matter fibers can be seen as efficient pathways for the transfer of information within a brain network [71]. With around 200 million interhemispheric fibers [72], the microstructural integrity of the corpus callosum pathway appears to be critical for a preserved cognitive performance in the elderly. For instance, loss of microstructural integrity of the corpus callosum in patients with either mild cognitive impairment or AD has been shown using DTI metrics [73, 74]. In addition, Madden et al. showed that an altered white matter microstructure in the corpus callosum measured by FA is associated with a slower response in a visual task in older adults [75]. Other DTI studies showed that the microstructural organization of the corpus callosum is associated with individual performance in cognitive flexibility as assessed by the TMTB-A in the elderly [9, 10].

Bringing more specificity to the microstructural analysis, our work presents the first evidence concerning an association between cognitive flexibility and NODDI metrics in the corpus callosum. However, one previous study used the same NODDI model in order to investigate regional patterns of grey matter microstructure in healthy individuals across the adult lifespan (21–84 years) [76]. Interestingly, the authors reported that the ISOVF measured in the frontal pole mediates a negative relationship between age and executive function [76]. Our results add to the literature by showing that the neuronal density (ICVF) of the corpus callosum, and in particular the body of the corpus callosum, is significantly associated with the performance on the TMTB-A. More precisely, our results suggest that a better performance in cognitive flexibility, characterized by a shorter time to complete the TMTB-A, was associated with an increase in neuronal density of the body of the corpus callosum.

One must note that several studies assessing age-related microstructural alterations have previously shown that changes in FA appear sooner in the genu and body of the corpus callosum than in any other white matter regions of the brain [54 –56]. Using metrics more specifically related to the physiological changes in the microstructure (ISOVF and ICVF), our results suggest that following changes in the neuronal density of the body of the corpus callosum may offer a potential tool for early detection of cognitive dysfunction in the elderly.

Strengths and limitations

The strength of this study lies in the inclusion of quantitative MRI metrics (NODDI) that are sensitive and specific to the white matter microstructure in order to investigate the association between arterial and white matter integrity. Previous studies have used DTI metrics to estimate changes in white matter microstructure following arterial stiffness which are physiologically unspecific. In contrast, by using the NODDI model, this work was able to disentangle changes in neuronal density and changes in the extracellular water diffusion following hemodynamic changes. However, the NODDI model also suffers from some limitations. For example, NODDI requires fixing the intracellular and isotropic diffusivity which could influence results. Nonetheless, due to their good reproducibility [77, 78], NODDI metrics are valuable for studying the neuroaxonal pathology in a broad range of conditions including the impact of cardiovascular risk factors on white matter microstructure. Relating T2 hyperintensities with NODDI metrics would have been interesting, however T2-weighted scans were not acquired in the study.

While another strength of the current study is the use of non-invasive quantitative perfusion imaging to calculate white matter BF, one must note that the ASL technique suffers from several limitations. In particular, the SNR of ASL data in the white matter is known to be relatively poor due to small perfusion fraction and long transit times [79]. Moreover, although experimental data suggest that it may be possible to measure white matter perfusion by using an appropriate tagging duration and post-labelling delay in healthy individuals [80], our post-labelling delay might be too low to accurately detect white matter perfusion. The range of temporal SNR (tSNR) has been reported to be around 1.81±0.13 in older adults for different gray matter regions such as the frontal or parietal lobe [81]. In the present study, we calculated the tSNR for each voxel as the ratio of the mean and std of the temporal signal in the raw perfusion weighted image (difference between control-tag). We found a tSNR of 1.04±0.62 in the whole gray matter for randomly selected subjects, agreeing with the literature. We then calculated the tSNR in the corpus callosum by averaging the tSNR values in the voxels that belong to the corpus callosum. We found a tSNR of 0.23±0.41 in the corpus callosum. In comparison with the gray matter, the BF in the white matter has been reported to be 1.6 to 4.6 times lower [82 –84] and it takes a longer transit time for the tags to travel from the labelling region to capillaries, which negatively impacts the tSNR [80]. Nevertheless, our tSNR in the corpus callosum may not be optimal. Additional studies with an increased pulse labelling delay following the recommendation of Alsop et al. may be necessary when it comes to ASL acquisition in the elderly since a prolonged transit time is expected [48]. Future studies could also use multiple pulse labeling, multiple band approaches and/or additional recommendations from the ISMRM perfusion study group and the European consortium for ASL in dementia [48] to optimize parameter selection for the population studied. Future studies could also apply robust method for correcting partial volume effect in ASL [85] in order to optimize parameter selection for the population of interest.

The evidence linking directly arterial stiffness and cognitive decline is highly inconsistent due to different study designs, heterogeneity of target populations and cognitive screening tools. Therefore, we did not investigate the direct effect of arterial stiffness on executive function. A much larger population is needed to control for many cofactors. Instead, since the corpus callosum is known to be vulnerable to arterial stiffness and has been previously shown to be implicated in AD, we wanted to assess if its microstructural information and perfusion is associated with executive function. Finally, although this study provides valuable knowledge on the impact of arterial stiffness in the aging brain, a follow-up study with better statistical power is needed to confirm our findings. A longitudinal study would also help us better understand the axonal pathology throughout the lifespan before the appearance of irreversible white matter damage and cognitive impairment. Such a longitudinal study will also help answering the question on the presence of potential compensatory mechanisms in the aging brain following vascular aging.

Conclusion

This study is the first to investigate the association of arterial stiffness on the perfusion, neuronal density and extracellular water diffusion of the corpus callosum in older adults. Using MRI metrics specific to the white matter microstructure (ISOVF, ICVF), this work provides evidence that the ISOVF and ICVF are sensitive measures of subtle brain injury associated with increased cfPWV. In addition, a lower microstructural integrity of the corpus callosum was related to a lower cognitive performance, which may explain the important role of the corpus callosum and its atrophy in AD.

However, additional studies are needed to better understand the underlying mechanism behind the increase in extracellular water diffusion (ISOVF) and neuronal density (ICVF) seen in the corpus callosum in relation to increased arterial stiffness. A validation of these MRI metrics (ISOVF, ICVF) in an animal model of arterial stiffness would greatly help in that direction.

In addition, a decrease in neuronal density of the body of the corpus callosum as assessed with the ICVF has been shown to translate into a decline in cognitive flexibility performance (longer time to complete the TMTB-A). Considering the strong interplay between arterial stiffness and a decline in both brain structure and function, the present work emphasizes 1) the possibility of using the cfPWV measure to identify individuals at higher risk for altered white matter microstructure, 2) the need for novel interventions aiming to preserve the white matter microstructure, particularly in vulnerable individuals. This may in turn help preserve brain function in the aging brain before irreversible structural damage occurs. However, we must keep in mind that these observations were made in a restricted age group and that these associations could vary in subjects of another age group. A follow up study using a large biobank data with a broader age range will help us verify whether these results generalize across different age range.

Footnotes

ACKNOWLEDGMENTS

The authors would like to acknowledge Carollyn Hurst and André Cyr from the Functional Neuroimaging Unit (CRIUGM, Université de Montréal) for helping with data acquisitions. Arnaud Boré and Tanguy Duval are acknowledged for helping with the analysis of diffusion data. Finally, Ernesto L. Schiffrin, Pierre Paradis and Julio Fraulob-Aquino (Lady Davis Institute for Medical Research, McGill University) are acknowledged for their help in the acquisition of pulse wave velocity measurements.

This study was supported by the Centre de Recherche de l’Institut Universitaire de Gériatrie de Montréal (CRIUGM) and the MerckSharp & Dohme Corp Program of the Faculty of Medicine of the Université de Montréal, the Canada Research Chair in Quantitative Magnetic Resonance Imaging [950-230815], the Canadian Institute of Health Research [CIHR FDN-143263], the Canada Foundation for Innovation [32454, 34824], the Fonds de Recherche du Québec - Santé, the Fonds de Recherche du Québec - Nature et Technologies [2015-PR-182754], the Natural Sciences and Engineering Research Council of Canada [435897-2013], the Canada First Research Excellence Fund (IVADO and TransMedTech), the Quebec BioImaging Network [5886]. Hélène Girouard was the holder of a new investigator award from the Fonds de Recherche du Québec-Santé (FRSQ) and the Heart and Stroke Foundation of Canada (HSFC). Atef Badji and Agah Karakuzu were supported by a TransMedTech excellence scholarship. Adrián Noriega de la Colina was supported by a Doctoral Fellowship from the Société Québécoise d’Hypertension Artérielle (SQHA). Charley Gros was supported by an IVADO excellence scholarship [EX-2018-4].

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Sweeney

, Montagne

, Sagare

, Nation

, Schneider

, Chui

, Harrington

, Pa

, Law

, Wang

DJJ

, Jacobs

, Doubal

, Ramirez

, Black

, Nedergaard

, Benveniste

, Dichgans

, Iadecola

, Love

, Bath

, Markus

, Salman

, Allan

, Quinn

, Kalaria

, Werring

, Carare

, Touyz

, Williams

SCR

, Moskowitz

, Katusic

, Lutz

, Lazarov

, Minshall

, Rehman

, Davis

, Wellington

, González

, Yuan

, Lockhart

, Hughes

, Chen

CLH

, Sachdev

, O’Brien

, Skoog

, Pantoni

, Gustafson

, Biessels

, Wallin

, Smith

, Mok

, Wong

, Passmore

, Barkof

, Muller

, Breteler

MMB

, Román

, Hamel

, Seshadri

, Gottesman

, van Buchem

, Arvanitakis

, Schneider

, Drewes

, Hachinski

, Finch

, Toga

, Wardlaw

, Zlokovic

(2019) Vascular dysfunction—the disregarded partner of Alzheimer’s disease. Alzheimers Dement 15, 158–167.

Wiesmann

, Kiliaan

, Claassen

JAHR

(2013) Vascular aspects of cognitive impairment and dementia. J Cereb Blood Flow Metab 33, 1696–1706.

World Health Organization (2015) The World Health Report 2008 - Primary Health Care (Now More Than Ever). World Health Organization, Geneva.

Prince

, Bryce

, Albanese

, Wimo

, Ribeiro

, Ferri

(2013) The global prevalence of dementia: A systematic review and metaanalysis. Alzheimers Dement 9, 63–75.e2.

Baumgart

, Snyder

, Carrillo

, Fazio

, Kim

, Johns

(2015) Summary of the evidence on modifiable risk factors for cognitive decline and dementia: A population-based perspective. Alzheimers Dement 11, 718–726.

Roberts

, Knopman

(2013) Classification and epidemiology of MCI. Clin Geriatr Med 29, 753–772.

Villeneuve

, Jagust

(2015) Imaging vascular disease and amyloid in the aging brain: Implications for treatment. J Prev Alzheimers Dis 2, 64–70.

Kalaria

, Akinyemi

, Ihara

(2012) Does vascular pathology contribute to Alzheimer changes? J Neurol Sci 322, 141–147.

Tarumi

, de Jong

DLK

, Zhu

, Tseng

, Liu

, Hill

, Riley

, Womack

, Kerwin

, Lu

, Munro Cullum

, Zhang

(2015) Central artery stiffness, baroreflex sensitivity, and brain white matter neuronal fiber integrity in older adults. Neuroimage 110, 162–170.

10.

Badji

, Noriega de la Colina

, Karakuzu

, Duval

, Desjardins-Crépeau

, Joubert

, Bherer

, Lamarre-Cliche

, Stikov

, Girouard

, Cohen-Adad

(2019) Arterial stiffness and white matter integrity in the elderly: A diffusion tensor and magnetization transfer imaging study. Neuroimage 186, 577–585.

11.

Maillard

, Mitchell

, Himali

, Beiser

, Fletcher

, Tsao

, Pase

, Satizabal

, Vasan

, Seshadri

, DeCarli

(2017) Aortic stiffness, increased white matter free water, and altered microstructural integrity: A continuum of injury. Stroke 48, 1567–1573.

12.

Singer

, Trollor

, Baune

, Sachdev

, Smith

(2014) Arterial stiffness, the brain and cognition: A systematic review. Ageing Res Rev 15, 16–27.

13.

Badji

, Sabra

, Bherer

, Cohen-Adad

, Girouard

, Gauthier

(2019) Arterial stiffness and brain integrity: A review of MRI findings. Ageing Res Rev 53, 100907.

14.

Laurent

, Boutouyrie

, Lacolley

(2005) Structural and genetic bases of arterial stiffness. Hypertension 45, 1050–1055.

15.

Pase

(2012) Modifiable vascular markers for cognitive decline and dementia: The importance of arterial aging and hemodynamic factors. J Alzheimers Dis 32, 653–663.

16.

Vlachopoulos

, O’Rourke

, Nichols

(2011) McDonald’s Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, CRC Press.

17.

Van Bortel

, Laurent

, Boutouyrie

, Chowienczyk

, Cruickshank

, De Backer

, Filipovsky

, Huybrechts

, Mattace-Raso

FUS

, Protogerou

, Schillaci

, Segers

, Vermeersch

, Weber

; Artery Society; European Society of Hypertension Working Group on Vascular Structure and Function; European Network for Noninvasive Investigation of Large Arteries (2012) Expert consensus document on the measurement of aortic stiffness in daily practice using carotid-femoral pulse wave velocity. J Hypertens 30, 445–448.

18.

Iulita

, Noriega de la Colina

, Girouard

(2018) Arterial stiffness, cognitive impairment and dementia: Confounding factor or real risk? J Neurochem 144, 527–548.

19.

Mikael

L de R

, Paiva

AMG de

, Gomes

, Sousa

ALL

, Jardim

PCBV

, Vitorino

PV de O

, Euzébio

, Sousa

W de M

, Barroso

WKS

(2017) Vascular aging and arterial stiffness. Arq Bras Cardiol 109, 253–258.

20.

Bombois

, Debette

, Delbeuck

, Bruandet

, Lepoittevin

, Delmaire

, Leys

, Pasquier

(2007) Prevalence of subcortical vascular lesions and association with executive function in mild cognitive impairment subtypes. Stroke 38, 2595–2597.

21.

Yoshita

, Fletcher

, Harvey

, Ortega

, Martinez

, Mungas

, Reed

, DeCarli

(2006) Extent and distribution of white matter hyperintensities in normal aging, MCI, and AD. Neurology 67, 2192–2198.

22.

Vermeer

, Prins

, den Heijer

, Hofman

, Koudstaal

, Breteler

MMB

(2003) Silent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med 348, 1215–1222.

23.

Skoog

(1998) A review on blood pressure and ischaemic white matter lesions. Dement Geriatr Cogn Disord 9 Suppl 1, 13–19.

24.

Maillard

, Fletcher

, Lockhart

, Roach

, Reed

, Mungas

, DeCarli

, Carmichael

(2014) White matter hyperintensities and their penumbra lie along a continuum of injury in the aging brain. Stroke 45, 1721–1726.

25.

Maillard

, Mitchell

, Himali

, Beiser

, Tsao

, Pase

, Satizabal

, Vasan

, Seshadri

, DeCarli

(2016) Effects of arterial stiffness on brain integrity in young adults from the Framingham Heart Study. Stroke 47, 1030–1036.

26.

Edzes

, Samulski

(1977) Cross relaxation and spin diffusion in the proton NMR or hydrated collagen. Nature 265, 521–523.

27.

Wolff

, Balaban

(1989) Magnetization transfer contrast (MTC) and tissue water proton relaxation in vivo. Magn Reson Med 10, 135–144.

28.

Duval

, Stikov

, Cohen-Adad

(2016) Modeling white matter microstructure. Funct Neurol 31, 217–228.

29.

Zhang

, Schneider

, Wheeler-Kingshott

, Alexander

(2012) NODDI: Practical in vivo neurite orientation dispersion and density imaging of the human brain. Neuroimage 61, 1000–1016.

30.

Tarumi

, Zhang

(2017) Cerebral blood flow in normal aging adults: Cardiovascular determinants, clinical implications, and aerobic fitness. J Neurochem 144, 595–608.

31.

Golay

, Petersen

(2006) Arterial spin labeling: Benefits and pitfalls of high magnetic field. Neuroimaging Clin N Am 16, 259–268.

32.

Kennedy

, Raz

(2009) Pattern of normal age-related regional differences in white matter microstructure is modified by vascular risk. Brain Res 1297, 41–56.

33.

Frederiksen

, Garde

, Skimminge

, Ryberg

, Rostrup

, Baaré

WFC

, Siebner

, Hejl

, Leffers

, Waldemar

(2011) Corpus callosum atrophy in patients with mild Alzheimer’s disease. Neurodegener Dis 8, 476–482.

34.

Harrell

Jr , Lee

, Califf

, Pryor

, Rosati

(1984) Regression modelling strategies for improved prognostic prediction. Stat Med 3, 143–152.

35.

Harrell

Jr , Lee

, Mark

(1996) Multivariable prognostic models: Issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 15, 361–387.

36.

Peduzzi

, Concato

, Kemper

, Holford

, Feinstein

(1996) A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 49, 1373–1379.

37.

Butlin

, Qasem

(2017) Large artery stiffness assessment using sphygmocor technology. Pulse 4, 180–192.

38.

O’Rourke

(1999) Wave travel and reflection in the arterial system. J Hypertens Suppl 17, S45–S47.

39.

Nerenberg

, Zarnke

, Leung

, Dasgupta

, Butalia

, McBrien

, Harris

, Nakhla

, Cloutier

, Gelfer

, Lamarre-Cliche

, Milot

, Bolli

, Tremblay

, McLean

, Padwal

, Tran

, Grover

, Rabkin

, Moe

, Howlett

, Lindsay

, Hill

, Sharma

, Field

, Wein

, Shoamanesh

, Dresser

, Hamet

, Herman

, Burgess

, Gryn

, Grégoire

, Lewanczuk

, Poirier

, Campbell

, Feldman

, Lavoie

, Tsuyuki

, Honos

, Prebtani

APH

, Kline

, Schiffrin

, Don-Wauchope

, Tobe

, Gilbert

, Leiter

, Jones

, Woo

, Hegele

, Selby

, Pipe

, McFarlane

, Oh

, Gupta

, Bacon

, Kaczorowski

, Trudeau

, Campbell

NRC

, Hiremath

, Roerecke

, Arcand

, Ruzicka

, Prasad

GVR

, Vallée

, Edwards

, Sivapalan

, Penner

, Fournier

, Benoit

, Feber

, Dionne

, Magee

, Logan

, Côté

A-M

, Rey

, Firoz

, Kuyper

, Gabor

, Townsend

, Rabi

, Daskalopoulou

, Hypertension Canada (2018) Hypertension Canada’s 2018 guidelines for diagnosis, risk assessment, prevention, and treatment of hypertension in adults and children. Can J Cardiol 34, 506–525.

40.

Corrigan

, Hinkeldey

(1987) Relationships between parts A and B of the Trail Making Test. J Clin Psychol 43, 402–409.

41.

Reitan

(1958) Validity of the Trail Making Test as an indicator of organic brain damage. Percept Mot Skills 8, 271–276.

42.

Arbuthnott

, Frank

(2000) Executive control in set switching: Residual switch cost and task-set inhibition. Can J Exp Psychol 54, 33–41.

43.

Setsompop

, Cohen-Adad

, Gagoski

, Raij

, Yendiki

, Keil

, Wedeen

, Wald

(2012) Improving diffusion MRI using simultaneous multi-slice echo planar imaging. Neuroimage 63, 569–580.

44.

, Fernández-Seara

, Detre

, Wehrli

, Wang

(2007) A theoretical and experimental investigation of the tagging efficiency of pseudocontinuous arterial spin labeling. Magn Reson Med 58, 1020–1027.

45.

Manjón

, Coupé

, Concha

, Buades

, Collins

, Robles

(2013) Diffusion weighted image denoising using overcomplete local PCA. PLoS One 8, e73021.

46.

Daducci

, Canales-Rodríguez

, Zhang

, Dyrby

, Alexander

, Thiran

(2015) Accelerated microstructure imaging via convex optimization (AMICO) from diffusion MRI data. Neuroimage 105, 32–44.

47.

Liu

, Wong

(2005) A signal processing model for arterial spin labeling functional MRI. Neuroimage 24, 207–215.

48.

Alsop

, Detre

, Golay

, Günther

, Hendrikse

, Hernandez-Garcia

, Lu

, MacIntosh

, Parkes

, Smits

, van Osch

MJP

, Wang

DJJ

, Wong

, Zaharchuk

(2015) Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia: Recommended Implementation of ASL for Clinical Applications. Magn Reson Med 73, 102–116.

49.

Watanabe

, Liao

, Jara

, Sakai

(2013) Multispectral quantitative MR imaging of the human brain: Lifetime age-related effects. Radiographics 33, 1305–1319.

50.

Mori

, Oishi

, Jiang

, Li

, Akhter

, Hua

, Faria

, Mahmood

, Woods

, Toga

, Pike

, Neto

, Evans

, Zhang

, Huang

, Miller

, van Zijl

, Mazziotta

(2008) Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. Neuroimage 40, 570–582.

51.

Avants

, Tustison

, Song

, Cook

, Klein

, Gee

(2011) A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 54, 2033–2044.

52.

Commowick

, Wiest-Daessle

, Prima

(2012) Block-matching strategies for rigid registration of multimodal medical images. 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI).

53.

Suarez

, Commowick

, Prabhu

, Warfield

(2012) Automated delineation of white matter fiber tracts with a multiple region-of-interest approach. Neuroimage 59, 3690–3700.

54.

Kochunov

, Williamson

, Lancaster

, Fox

, Cornell

, Blangero

, Glahn

(2012) Fractional anisotropy of water diffusion in cerebral white matter across the lifespan. Neurobiol Aging 33, 9–20.

55.

Pfefferbaum

, Sullivan

, Hedehus

, Lim

, Adalsteinsson

, Moseley

(2000) Age-related decline in brain white matter anisotropy measured with spatially corrected echo-planar diffusion tensor imaging. Magn Reson Med 44, 259–268.

56.

Sullivan

, Adalsteinsson

, Hedehus

, Ju

, Moseley

, Lim

, Pfefferbaum

(2001) Equivalent disruption of regional white matter microstructure in ageing healthy men and women. Neuroreport 12, 99–104.

57.

Benjamini

, Hochberg

(1995) Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 57, 289–300.

58.

Beaulieu

(2009) The biological basis of diffusion anisotropy. In Diffusion MRI, Johansen-Berg H, Behrens TEJ, eds. Academic Press, San Diego, pp. 105-126.

59.

Suzuki

, Gao

, Bai

, Evangelou

, Glocker

, O’Regan

, Elliott

, Matthews

(2017) Abnormal brain white matter microstructure is associated with both pre-hypertension and hypertension. PLoS One 12, e0187600.

60.

Iulita

, Muhire

, Vallerand

, Youwakim

, Petry

, Gratuze

, Ferland

, Planel

, Girouard

(2017) Arterial stiffness due to carotid calcification disrupts cerebral blood flow regulation before cognitive deficits manifest. J Cereb Blod Flow Metab 37(1 Supp), 368–369.

61.

Muhire

, Iulita

, Vallerand

, Youwakim

, Gratuze

, Petry

, Planel

, Ferland

, Girouard

(2019) Arterial stiffness due to carotid calcification disrupts cerebral blood flow regulation and leads to cognitive deficits. J Am Heart Assoc 8, e011630.

62.

Benitez

, Jensen

, Helpern

(2017) Axonal density and myelin integrity in cognitive decline: A diffusional kurtosis imaging study. Alzheimers Dement 13, P774–P775.

63.

Fleischer

, Gröger

, Koirala

, Droby

, Muthuraman

, Kolber

, Reuter

, Meuth

, Zipp

, Groppa

(2017) Increased structural white and grey matter network connectivity compensates for functional decline in early multiple sclerosis. Mult Scler 23, 432–441.

64.

Mangeat

, Badji

, Ouellette

, Treaba

, Herranz

, Granberg

, Louapre

, Stikov

, Sloane

, Bellec

, Mainero

, Cohen-Adad

(2018) Changes in structural network are associated with cortical demyelination in early multiple sclerosis. Hum Brain Mapp 39, 2133–2146.

65.

Fick

RHJ

, Daianu

, Pizzolato

, Wassermann

, Jacobs

, Thompson

, Town

, Deriche

(2017) Comparison of biomarkers in transgenic alzheimer rats using multi-shell diffusion MRI. In Computational Diffusion MRI, Springer International Publishing, pp. 187–199.

66.

Mitchell

(2011) Aortic stiffness and cerebral blood flow. Am J Hypertens 24, 1056.

67.

Jefferson

, Cambronero

, Liu

, Moore

, Neal

, Terry

, Nair

, Pechman

, Rane

, Davis

, Gifford

, Hohman

, Bell

, Wang

, Beckman

, Carr

(2018) Higher aortic stiffness is related to lower cerebral blood flow and preserved cerebrovascular reactivity in older adults. Circulation 138, 1951–1962.

68.

Tarumi

, Thomas

, Wang

, Zhang

, Liu

, Turner

, Riley

, Tangella

, Womack

, Kerwin

, Cullum

, Lu

, Vongpatanasin

, Zhu

, Zhang

(2017) Ambulatory pulse pressure, brain neuronal fiber integrity, and cerebral blood flow in older adults. J Cereb Blood Flow Metab 39, 926–936.

69.

Tarumi

, Shah

, Tanaka

, Haley

(2011) Association between central elastic artery stiffness and cerebral perfusion in deep subcortical gray and white matter. Am J Hypertens 24, 1108–1113.

70.

Tarumi

, Ayaz Khan

, Liu

, Tseng

, Parker

, Riley

, Tinajero

, Zhang

(2014) Cerebral hemodynamics in normal aging: Central artery stiffness, wave reflection, and pressure pulsatility. J Cereb Blood Flow Metab 34, 971–978.

71.

Filley

(2012) The Behavioral Neurology of White Matter, Oxford University Press.

72.

Biegon

, Eberling

, Richardson

, Roos

, Wong

, Reed

, Jagust

(1994) Human corpus callosum in aging and Alzheimer’s disease: A magnetic resonance imaging study. Neurobiol Aging 15, 393–397.

73.

Paola

, Di Paola

, Di Iulio

, Cherubini

, Blundo

, Casini

, Sancesario

, Passafiume

, Caltagirone

, Spalletta

(2010) When, where, and how the corpus callosum changes in MCI and AD: A multimodal MRI study. Neurology 74, 1136–1142.

74.

Paola

, Di Paola

, Spalletta

, Caltagirone

(2010) In vivo structural neuroanatomy of corpus callosum in Alzheimer’s disease and mild cognitive impairment using different MRI techniques: A review. J Alzheimers Dis 20, 67–95.

75.

Madden

, Whiting

, Huettel

, White

, MacFall

, Provenzale

(2004) Diffusion tensor imaging of adult age differences in cerebral white matter: Relation to response time. Neuroimage 21, 1174–1181.

76.

Nazeri

, Chakravarty

, Rotenberg

, Rajji

, Rathi

, Michailovich

, Voineskos

(2015) Functional consequences of neurite orientation dispersion and density in humans across the adult lifespan. J Neurosci 35, 1753–1762.

77.

Tariq

, Schneider

, Alexander

, Wheeler-Kingshot

, Zhang

(2012) Scan-rescan reproducibility of neurite microstructure estimates using NODDI. In Medical Image Understanding and Analysis 2012: Proceedings of the 16th Conference on Medical Image Understanding and Analysis, Xie X, ed. The British Machine Vision Association and Society for Pattern Recognition, Swansea, UK, pp. 255-261.

78.

Parvathaneni

, Nath

, Blaber

, Schilling

, Hainline

, Mojahed

, Anderson

, Landman

(2018) Empirical reproducibility, sensitivity, and optimization of acquisition protocol, for neurite orientation dispersion and density imaging using AMICO. Magn Reson Imaging 50, 96–109.

79.

Van Gelderen

, De Zwart

, Duyn

(2008) Pittfalls of MRI measurement of white matter perfusion based on arterial spin labeling. Magn Reson Med 59, 788–795.

80.

, Lin

, Wang

, Chen

, Li

(2013) Measurement of cerebral white matter perfusion using pseudocontinuous arterial spin labeling 3T magnetic resonance imaging–an experimental and theoretical investigation of feasibility. PLoS One 8, e82679.

81.

Gauthier

, Madjar

, Desjardins-Crépeau

, Bellec

, Bherer

, Hoge

(2013) Age dependence of hemodynamic response characteristics in human functional magnetic resonance imaging. Neurobiol Aging 34, 1469–1485.

82.

van Osch

MJP

, Teeuwisse

, van Walderveen

MAA

, Hendrikse

, Kies

, van Buchem

(2009) Can arterial spin labeling detect white matter perfusion signal? Magn Reson Med 62, 165–173.

83.

Asllani

, Borogovac

, Brown

(2008) Regression algorithm correcting for partial volume effects in arterial spin labeling MRI. Magn Reson Med 60, 1362–1371.

84.

Roberts

, Detre

, Bolinger

, Insko

, Leigh

(1994) Quantitative magnetic resonance imaging of human brain perfusion at 1.5 T using steady-state inversion of arterial water. Proc Natl Acad Sci U S A 91, 33–37.

85.

Liang

, Connelly

, Calamante

(2013) Improved partial volume correction for single inversion time arterial spin labeling data. Magn Reson Med 69, 531–537.

86.

Badji

, de la Colina

, Karakuzu

, Duval

, Desjardins-Crépeau

, Parizet

, Joubert

, Bherer

, Lamarre-Cliche

, Stikov

, Cohen-Adad

, Girouard

(2019) Arterial stiffness cut-off value and white matter integrity in the elderly. Neuroimage Clin 26, 102007.