If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Clinical Sciences in Malmö, Division of Geriatric Medicine, Skåne University Hospital, Lund University, Jan Waldenströms gata 35, pl13, Malmö SE 205 02, Sweden
Department of Clinical Sciences in Malmö, Division of Geriatric Medicine, Skåne University Hospital, Lund University, Jan Waldenströms gata 35, pl13, Malmö SE 205 02, Sweden
Department of Clinical Sciences in Malmö, Division of Geriatric Medicine, Skåne University Hospital, Lund University, Jan Waldenströms gata 35, pl13, Malmö SE 205 02, Sweden
Magnetic Resonance Images of the brain reveals changes in >80% of 70–87 year olds.
•
Distribution of small vessel disease and brain atrophy in general cohort is diverse.
•
Microbleeds were present in 27% of subjects, lacunar infarctions in 8%.
•
Carotid End Diastolic Velocity was inversely associated with white matter changes.
•
Cortical atrophy was associated with Carotid flow, -resistivity and -pulsatility.
Abstract
Objectives
A growing body of evidence links age related brain pathologies to systemic vascular processes. We aimed to study the prevalence and interrelations between magnetic resonance imaging (MRI) markers of cerebral small vessel disease and patterns of brain atrophy, and their association to carotid duplex ultrasound flow parameters.
Materials and methods
We investigated a population based randomised cohort of older adults (n=391) aged 70-87, part of the Swedish Good Aging in Skåne Study. Peak systolic and end diastolic velocities of the carotid arteries were measured by ultrasound, and resistivity- and pulsatility indexes were calculated. Subjects with increased peak systolic velocity indicating carotid stenosis were excluded from analysis. Nine MRI findings were rated by visual scales: white matter changes, pontine white matter changes, microbleeds, lacunar infarctions, medial temporal lobe atrophy, global cortical atrophy, parietal atrophy, precuneus atrophy and central atrophy.
Results
MRI pathologies were found in 80% of subjects. Mean end diastolic velocity in common carotid arteries was inversely associated with white matter hyperintensities (OR=0.92; p=0.004), parietal lobe atrophy (OR=0.94; p=0.039), global cortical atrophy (OR=0.90; p=0.013), precuneus atrophy (OR=0.94; p=0.022), “number of CSV pathologies” (β=-0.07; p<0.001) and “MRI-burden score” (β=-0.11; p<0.001), after adjustment for age and sex. The latter three were also associated with pulsatility and resistivity indexes.
Conclusions
Low carotid end diastolic velocity, as well as increased carotid resistivity and pulsatility, were associated with signs of cerebral small vessel disease and patterns of brain atrophy, indicating a vascular component in the process of brain aging.
Cerebral small vessel disease (CSVD) affects small arteries, arterioles, veins and capillaries of the brain. Although a common and often silent condition occurring among healthy elderly people, it is associated with stroke, dementia, gait disturbances, cognitive decline and depression.
The LADIS Study Group 2001-2011: A decade of the LADIS (Leukoaraiosis And DISability) study: what have we learned about white matter changes and small-vessel disease?.
Prevalence of CSVD differs among studies and is highly dependent on the cohort's age, as well as investigative techniques, such as: magnetic field strength, or on visual or automated rating scales. WMC were present in 23-98% of elderly subjects in different population studies.
Pathophysiological pathways are still debated and differ among the different manifestations of CSVD. Common risk factors: age, sex, hypertension, diabetes, body mass index, dyslipidaemia, markers of athero- and arteriosclerosis, amyloidosis and genetic predisposition, all predispose to CSVD in different grade.
Brain atrophy is associated with normal aging but accelerates in both CSVD and neurodegenerative disorders such as Alzheimer's Disease (AD). Focal patterns of atrophy can be attributed to specific brain pathologies. Medial temporal lobe atrophy (MTA), Precuneus atrophy (PA) and Parietal lobe atrophy (KPA) are highly prevalent, although not exclusive, in AD and MCI patients.
Central atrophy (CA) reflects white matter degeneration and widening of ventricles, whereas global cortical atrophy (GCA) reflects grey matter degeneration.
The correlation between CSVD and altered hemodynamics in the large arteries of the head and neck is still not fully comprehended. Vascular flow can reflect both upstream and downstream pathology: reflected by resistivity and pulsatility indexes. To our knowledge, no previous study has investigated the association between carotid flow parameters and neuroimaging of CSVD in older adults with no carotid stenosis. Furthermore, impact of hemodynamic changes on brain atrophy is still largely unresearched. We hypothesized that changes in carotid flow parameters indicating increased arterial stiffness could be associated with a greater incidence of classical signs of CSVD or focal brain atrophy.
Materials and methods
Study cohort
The Good Aging in Skåne (GÅS) study is a longitudinal, general population based aging study, and a part of the Swedish National study on Aging and Care (SNAC).
The study, still ongoing, was initialized in 2001 and initially comprised 2931 subjects in nine different age cohorts ranging from 60 to 93 years in the baseline examination, with new participants recruited every sixth year. Subjects are evaluated by medical examinations, bloodworks, questionnaires, functionality tests and cognitive tests; details described elsewhere.
In conjunction with the fourth survey, all participants aged 70-87 from the Malmö-based test center were offered an extended examination protocol. 647 participants accepted and underwent ultrasonography of carotid arteries, and 408 of them subsequently accepted an invitation for a supplementary MRI scan during 2015 – 2017, which was performed within a year from ultrasound examination.
Background characteristics were acquired from baseline- and fourth survey protocols. General cognitive performance was assessed by the Mini-Mental State Examination (MMSE) in conjunction with the fourth survey, except 45 subjects (11.5%), whose MMSE results or background characteristics, due to missing data, were chosen from the previous or subsequent visit, whichever was closest in time to the MRI scan. Written informed consent was obtained from each patient. Ethical approval was obtained from the Regional Ethical Review Board, Lund, 2015/859.
Carotid flow measurements
Carotid Duplex Ultrasonography was performed by one and the same laboratory technician in a controlled environment. Examination included: peak systolic velocity (PSV) and end diastolic velocity (EDV) of: left and right common carotid artery (CCA), and left and right internal carotid artery at both proximal (ICAp) and distal (ICAd) sites.
Magnetic resonance image acquisition and processing
Subjects were examined with a 3 Tesla MRI (General Electric, discovery MR 750w), protocol described elsewhere.
MR images were visually assessed by an experienced neuroradiologist and evaluated with regard to occurrence of the following: (1) white matter changes (WMC) classified according to Fazekas´ scale,
and pontine white matter changes (WMP), (2) lacunar infarcts (LAC) defined as less than 1 cm in diameter, (3) cerebral microbleeds (MB) defined as small (2-5 mm) hypointense lesions
(7) central atrophy(CA) and (8) precuneus atrophy (PA).
408 subjects underwent the Carotid Flow Measurement and MRI examination. Two cases were excluded due to missing data/unable to complete MRI examination. Cases with carotid stenosis (n=15), defined as PSV values >120 cm/s
in any location, were excluded. Subsequently, 391 subjects remained for statistical analysis.
Statistical analysis
Carotid flow parameters and MRI-findings were assessed in the total cohort and stratified according to age and sex. Based upon clinical significance, mild/sporadic WMC (Fazekas score 1, n=204) and mild MTA (Scheltens’ scale 1, n=213) were coded as “no pathologies”. “Number of MRI pathologies” was calculated as the sum of how many different types of pathologies found (where Fazekas score 1 (WMC) and Scheltens score 1 (MTA) were coded as “no pathology”). The “MRI-burden score” was assessed by calculating the sum of the original, non-dichotomized severity scores for each of the MRI-findings as follows: the maximum score for the individual variables was three (no (0) - mild (1) – moderate (2) – severe (3)) for scale variables (WMC, MB, MTA, KPA, GCA). The maximum score for dichotomous variables (LAC, PA, WMP, CA) was one. Number of microbleeds was divided into a three-scale severity code where one MB was coded as 1, two through nine MB coded as 2 and ten or more Mb coded as 3. A “MRI-severity index” was defined as “MRI burden score” divided by number of original lesions found, including WMC Fazeka score 1 and MTA Schelten score 1.
For each of the carotid flow locations (CCA left and right, ICAp and ICAd left and right respectively) resistivity and pulsatility indexes (RI and PI) were calculated using the Pourcelot formula (RI = (PSV-EDV)/PSV) and Goslin's index (PI=(PSV-EDV)/V (mean), where V(mean)=(PSV – EDV) /3+EDV). For regression analysis, index scores were multiplied by ten.
Pairwise correlation was established by Yule's Q formula for colligation, and the dissimilarity between conditions was defined as 1-Yule's Q. According to Yule's Q, covariance between MRI pathologies were investigated by hierarchical cluster analysis to determine clusters of variables with a higher degree of covariance.
Co-occurrences between MRI pathologies were also investigated in a cross tabulation using Pearson's chi2 to check for associations.
Regression analyses were performed with each of the MRI-findings as well as “MRI-burden score” and “number of MRI pathologies” as dependent variables, and carotid flow parameters (peak systolic volume, end diastolic volume, pulsatility index and resistivity index for CCA, ICAp and ICAd, left and right side respectively) as independent variables. Multivariable models were adjusted for age and sex. To test the robustness of the findings, analyses were repeated in sub cohorts stratified according to age and sex. Primary analyses were performed on all variables separately (left-right, CCA, ICAp and ICAd, respectively). For brevity, subsequent analyses were performed on mean values of left and right side for CCA only.
IBM SPSS Statistics v25 was used for all statistical analyses.
Results
Baseline characteristics
The characteristics of the study population is presented in Table 1. Out of 391 subjects, 57% (n=224) were female. The age ranged between 70 and 87, median age being 76 years.
Carotid flow values are reported in Table 1 and displayed in Fig. 1. PSV and EDV decreased significantly with age and increased significantly distally from CCA to ICAp to ICAd (not shown). EDV was significantly lower, and PI and RI significantly higher in males. There was no statistical difference in PSV between the sexes.
Fig. 1Carotid flow parameters in relation to sex and age group. PSV was associated with age but not with sex. EDV, RI and PI was associated with age and sex in bivariable analysis models. Abbreviations: CCA: common carotid artery; PSV: Peak Systolic Velocity; EDV: End Diastolic Velocity; RI: Resistivity index; PI: Pulsatility Index.
Prevalence and severity of MRI-pathology are displayed in Table 1 and in supplementary Table 2. Pathologies, as we defined them, existed in 80% (n=312) of the subjects, and 25% of them had only one kind of MRI sign, 23% presented with two different MRI signs, while 31% had three types of MRI signs or more (Table 2). WMC ≥2 was associated with many of the manifestations, except cortical atrophies (GCA, KPA, PA) (Fig. 2).
Table 2The distribution of combined MRI pathologies.
n (%)
MRI-severity index (median (range) 25th;75th percentile)
0 pathologies
73 (18.7%)
1(0-1) 1;1
1 kind of pathology
99 (25.3%)
1(1-3) 1;1.5
MB
24 (6.1%)
KPA
22 (5.6%)
WMC ≥2
18 (4.6%)
GCA
11 (2.8%)
Other
24 (6.1%)
2 combinations
91 (23.3%)
1.33 (1-2.5) 1;1.5
KPA + GCA
19 (4.9%)
KPA + MB
7 (1,8%)
WMP + WMC ≥2
7 (1.8%)
MB + WMC ≥2
7 (1.8%)
Other
51 (13.0%)
3 combinations
60 (15.3%)
1.25 (1-2) 1.25;1.67
PA+CA+GCA
6 (1.5%)
WMP+MB+ WMC ≥2
5 (1.3%)
PA+KPA+ WMC ≥2
5 (1.3%)
PA + KPA + MB
5 (1.3%)
CA + PA + MTA ≥2
5 (1.3%)
Other
34 (8.7%)
4 combinations
38 (9.7%)
1.5 (1-2.5) 1.20;1.75
5 -7 combinations
24 (6.1%)
1.6 (1.14 – 2.20) 1.33;1.78
Data missing
6 (1.5%)
Note: WMC Fazekas grade 1 and MTA Scheltens grade 1 are defined as “no pathologies”. MRI severity index is defined as “MRI burden score” / number of lesions found, including WMC Fazekas grade 1 and MTA Scheltens grade 1.
Abbreviations: WMC: white matter changes; MB: microbleeds; WMP: pontine white matter changes; LAC: lacunar infarctions; KPA: Koedam parietal atrophy; PA: precuneus atrophy; MTA: medial temporal lobe atrophy; CA: central atrophy; GCA: global cortical atrophy
Fig. 2Co-occurrence of MRI-pathologies expressed as number of mutual cases and percentage within the variable expressed in the vertical rows. Individuals with data missing are excluded test wise, percentage is based on valid percentage. Bold font express significant association between variables according to Pearson's chi squared. Abbreviations: WMC: white matter changes; MB: microbleeds; WMP: pontine white matter changes; LAC: lacunar infarctions; KPA: parietal atrophy; PA: precuneus atrophy; MTA: medial temporal lobe atrophy; CA: central atrophy; GCA: global cortical atrophy * WMC Fazekas grade 1 and MTA Scheltens grade 1 are coded as “no lesions”.
Based on the results of cluster analysis (Fig. 3), cut-off value for distance between clusters was arbitrarily set to 0.9 –since it entailed the identification of four sets of definable clusters: cluster 1 included WMC≥2 + GCA, cluster 2 included WMP + KPA + MTA≥2, cluster 3 included MB + PA, cluster 4 included LAC + CA.
Fig. 3Clustering of MRI-findings displayed as a dendrogram, where a higher degree of co-occurrence is illustrated by a sooner point of convergence. Pairwise coefficient of colligation was established using Yule´s Q formula, and distance between manifestations was defined as “1-Yule´s Q” Abbreviations: WMC: white matter changes; GCA: global cortical atrophy; WMP: pontine white matter changes; KPA: Koedam parietal atrophy; MTA: medial temporal lobe atrophy; MB: microbleeds; PA: precuneus atrophy; LAC: lacunar infarctions; CA: central atrophy.
Association between carotid flow and cerebral small vessel disease
Table 3 shows associations between mean carotid flow parameters and MRI-findings. For brevity, only results for whole cohort CCA are presented. In multivariable linear or logistic regression models, mean EDV in CCA was negatively associated with “MRI-burden score” (β:-0.11; p<0.001), “number of MRI pathologies” (β:-0.07; p<0.001), “ WMH ≥2” (OR: 0.92; p=0.004), GCA (OR: 0.90; p=0.013), KPA (OR: 0.94; p=0.039), and PA (OR:0.94; p=0.022), adjusted for age and sex. Greater pulsatility- and resistivity indexes were associated with “number of MRI pathologies” (β: 0.12; p=0.001 and β: 0.47; p<0.001), “MRI-burden score” (β: 0.22; p<0.001 and β: 0.85; p<0.001), and PA (OR: 1.13; p=0.029 and OR: 1.67; p= 0.027), after adjusting for age and sex. There were no associations between carotid flow parameters and presence of MB, WMP, LAC, MTA, or CA in analysis of the whole, unstratified cohort.
Table 3Association between CSVD and carotid flow in multivariable regression models adjusted for age and sex.
OR (95% CI)
p
OR (95% CI)
p
WMC ≥2
n=112
MTA ≥2
n=75
PSV mean
0.99 (0,97-1.00)
0.086
PSV mean
1.00 (0.98 – 1.03)
0.682
EDV mean
0.92 (0.87 – 0.97)
0.004
EDV mean
0.98 (0.91 – 1.05)
0.537
Pulsatility index
1.09 (0.98 – 1.22)
0.132
Pulsatility index
1.07 (0.94 – 1.22)
0.334
Resistivity index
1.40 (0.91 – 2.18)
0.129
Resistivity index
1.30 (0.77 – 2.18)
0.331
MB
n=105
CA
n=65
PSV mean
1.01 (0.99 – 1.03)
0.443
PSV mean
1.02 (0.99 – 1.04)
0.149
EDV mean
0.97 (0.92 – 1.03)
0.381
EDV mean
0.98 (0.91 – 1.05)
0.520
Pulsatility index
1.12 (0.99 – 1.26)
0.055
Pulsatility index
1.14 (0.99-1.31)
0.059
Resistivity index
1.55 (0.98 – 2.45)
0.061
Resistivity index
1.71 (0.98 – 2.97)
0.059
WMP
n=63
GCA
n=48
PSV mean
1.00 (0.98 – 1.02)
0.957
PSV mean
0.98 (0.96 – 1.01)
0.143
EDV mean
0.98 (0.91 – 1.05)
0.477
EDV mean
0.90 (0.83 – 0.98)
0.013
Pulsatility index
1.08 (0.94 – 1.24)
0.265
Pulsatility index
1.10 (0.95 – 1.28)
0.210
Resistivity index
1.34 (0.78 – 2.32)
0.301
Resistivity index
1.44 (0.78 – 2.65)
0.244
LAC
n=32
β (95% CI)
p
PSV mean
0.960 (0.93 – 1.00)
0,050
“number of MRI
EDV mean
0.94 (0.86 – 1.04)
0.221
pathologies”
Pulsatility index
0.86 (0.72 – 1.04)
0.111
PSV mean
-0.01 (-0.02 – 0.01)
0.409
Resistivity index
0.55 (0.27 – 1.09)
0.087
EDV mean
-0.07 (-0.11 - -0.03)
<0.001
Pulsatility index
0.12 (0.05 – 0.20)
0.001
KPA
n=126
Resistivity index
0.47 (0.18 – 0.76)
<0.001
PSV mean
1.0 (0.99 -1.01)
0.694
EDV mean
0.94 (0.89 – 0.99)
0.039
“MRI burden score”
Pulsatility index
1.11 (1.0 – 1.24)
0.062
PSV mean
-0.01 (-0.02 – 0.01)
0.483
Resistivity index
1.47 (0.95 – 2.26)
0.081
EDV mean
-0.11 (-0.17 - -0.06)
<0.001
Pulsatility index
0.22 (0.11 – 0.33)
<0.001
PA
n=115
Resistivity index
0.85 (0.42 – 1.29)
<0.001
PSV mean
1.0 (0.98 – 1.01)
0.806
EDV mean
0.94 (0.88 – 0.98)
0.022
Pulsatility index
1.13 (1.01 – 1.27)
0.029
Resistivity index
1.67 (1.06 – 2.61)
0.027
Abbreviations: PSV: peak systolic velocity (cm/s); EDV: end diastolic velocity (cm/s); WMC: white matter changes; MB: microbleeds; WMP: pontine white matter changes; LAC: lacunar infarctions; KPA: Koedam parietal atrophy; PA: precuneus atrophy; MTA: medial temporal lobe atrophy; CA: central atrophy; GCA: global cortical atrophy
In this study we investigated the prevalence and co-occurrence of CSVD markers and patterns of brain atrophy in a general population-based cohort of older adults without carotid stenosis. We also investigated the association between mentioned MRI pathologies and carotid flow parameters. MRI pathologies were present in 80% of subjects and 54% had multiple pathologies, indicating that brain changes are common and a part of the aging process. Low EDV in the common carotid artery was associated with cortical atrophies such as GCA, KPA, PA, with moderate/severe WMC, and with pooled MRI-burden variables.
Comparison of prevalences
As noted previously, the prevalence of MRI-findings differs greatly among studies, and is dependent on cohort age and health, MRI magnetic field strength, visual or computerized assessment of pathologies, and neuroradiologist experience. This study, representative for a community dwelling elderly population, somewhat healthier than the average population of the same age due to voluntary and involuntary selection bias, reports a lower degree of total WMC than several other large studies. This is especially true compared to the ones that are machine assessed. In our study, microbleeds were present in 27% of study participants, similarly to a population from the Rotterdam Scan study, where microbleeds were present in one out of five persons over the age of 60, and in one out of three persons over 80 (4). In PIVUS, a Swedish population-based study of 75-year-old individuals, MTA≥2 was present in 54% of subjects compared to 20% of our study participants in age interval 75-79 yrs. (n=132).
Additionally, in the PIVUS study, 6 out of 390 subjects presented with MTA grade 4, compared to none in our study. PIVUS study had also a higher proportion of low educated individuals, a factor that the study noted as a confounder. In SNAC-K study, with a similar, swedish, randomized, non-demented population based cohort, MTA≥2 was found in 19% of the 72 year old's.
Few studies have investigated the covariance between different aspects of CSVD and patterns of atrophy. The best described association between different CSVD manifestations is between WMC and MB, as well as between WMC and general brain atrophy.
The heterogenic manifestation of brain aging is shown by the even dispersion of different combinations of pathologies in our study (Fig. 2) and the relatively high average distance between clusters in cluster analysis (Fig. 3), giving further support to the “whole brain disease” angle. The cluster analysis is based upon differences between lesions, and as every lesion can be in only one cluster, the average distance between the clusters tends to be high. As a result, manifestations that tend to co-occur together, but less commonly with other manifestations, would be clustered closer. Put in another way, large distance between clusters do not necessarily imply that the variables do not co-occur, but that they co-occur with other variables as well. On the contrary, the Pearson's Chi2 test is based on similarities, and the similarity between two variables is unaffected by similarities to other lesions. In choosing to set the cut of value between clusters to 0.9, a few interesting groupings emerged. Cluster 1 included WMC≥2 and GCA – both reflecting global morphological brain changes, and both associated to lower EDV velocities in our study. Cluster 2 included KPA and MTA, both associated to Alzheimer´s dementia. Cluster 4 included LAC and CA, both reflecting pathologies in deeper brain structures. Interestingly, none of these combinations showed a significant association according to Pearson's Chi2 analysis. Furthermore, as could be seen in Table 2, individuals with higher number of MRI-pathologies also had higher “MRI burden score”, indicating that number of lesions and lesion severity go hand in hand.
Arterial stiffness and vessel flow physiology
In the attempt to investigate underlying pathophysiological mechanisms behind cerebral pathologies, the distinction between arterial stiffness and atherosclerosis is of interest. Atherosclerosis, buildup of lesions containing among other things cholesterol and cell remnants in the intima of the vessel walls,
is often associated with but distinct from arterial stiffening/arteriosclerosis, defined as a thickening and hardening of the vessel wall, mainly in the media. Classical vascular risk factors associated with atherosclerosis including high blood glucose or lipids, inflammatory activity, genetic predisposition, life-style factors, such as physical exercise, smoking, alcohol consumption and body composition, are inconclusively associated with arterial stiffness.
Carotid-femoral pulse wave velocity (cf-PWV) has been proposed as a gold standard of arterial stiffness and several previous studies have seen associations to CSVD, cerebral insults, cognitive impairment or dementia.
It is believed that a disproportionate stiffening of the aorta augments the excess pulsatile force wave energy into the arteries supplying the brain. This is confirmed by previous studies showing that stiffness of the aorta is associated with an increased pulsatility in CCA and MCA, and excessive pulsatile energy in the carotid arteries.
In one study, investigating relationship between pulsatility in the greater vessels of the neck and skull and CSVD, the association is stronger the more distally it is measured (CCA, ICA or MCA). It is not clear if this is rather a mirror of increased vascular resistance in the downstream vascular beds, and in fact a reversed causality.
Associations between MRI pathologies and carotid flow parameters
The determinants of arterial blood flow velocity and pulsatility are many, complex, and only partly understood. Cardiac output, great vessel diameter and distensibility, pulse wave augmentation, muscular artery lumen, endothelial and humoral responses and downstream vascular resistance are a few known factors that influence arterial flow. In healthy cerebral vasculature, the vessel resistance is low, allowing continuous arterial blood flow in both systole and diastole. With increasing distal resistance and flow obstruction, the EDV flow is affected prior to PSV, as it is delivered at a lower arterial blood pressure to overcome this resistance, and could be an early marker of mild to moderate cerebrovascular disease. In studies estimating cerebral hypoperfusion with resting-state magnetoencephalography, low EDV in CCA was associated with impaired brain activity and cognitive status via hypoperfusion,
and could be responsible for white matter degeneration, probably due to the poor collateral bed.
In our material, WMC and the cortical atrophies were associated with decreased EDV, while MB, CA and LAC, although associated to WMC, were not. The connection between WMC and decreased EDV could be that the cerebral white matter has an end artery supply which could render it vulnerable to ischemic damage. Another possible implication for these findings might be that MB, CA and LAC might be fully or partly explained by other causal pathways also affecting the development of WMC. Except for MTA, all the investigated cortical atrophies results were stringent and might be a reflection of cortical vulnerability to alterations in blood flow. The pulsatility and resistivity indexes were in our material only associated with the compound MRI variables and with Precuneus Atrophy. The location of precuneus area on the medial surface of the hemispheres could make it especially vulnerable to flow alterations. The main blood supply of the precuneus area is provided from the posterior cerebral artery, not from the carotids. Associations between atrophies and flow alterations in the carotid arteries indicate that this is an upstream or systemic phenomenon, rather than a local flow alteration. In previous observations, patients with early onset Alzheimer´s Disease were presented with lower grey matter density in the precuneus.
Hypoperfusion in the posterior cingulate and precuneal areas was associated with cognitive function but could not be statistically separated from that of global atrophy.
Our findings could be in further support of emerging data that suggests a vascular vulnerability also in the progression of AD.
Notably, for the composite measures “number of MRI-pathologies” and “MRI-burden score”, the inverse association to EDV and positive association to PI and RI was stronger than was the case with the individual parameters. Individual brain ageing process is diverging, starting and progressing in different areas, depending on large or small vessel disease, constellation of vascular risk and genetic factors. Therefore, “number of MRI pathologies” and “MRI burden score” better reflect severity of brain ageing, and as a concept have been used previously.
The stronger association could also be partly explained by the fact that the individual pathologies were not compared to healthy controls, but to “the rest” – certainly including other forms of MRI-pathologies. As shown in Table 3, the statistical method with the composite variables were another; the linear regression model can make use of more information regarding the degree of carotid flow and cerebral pathological changes. For a comprehensive investigation, analyses were made with all carotid flow values, including PSV. However, PSV-results should be interpreted with precaution since cases with PSV values >120 cm/s were excluded before analysis.
Previous studies looking at associations between carotid flow and CSVD have somewhat discordant results, although, not necessarily contradictive. Several studies have linked CSVD manifestations to high PSV values or other manifestations of atherosclerotic plaques.
Our findings do not contradict these results, on the contrary, in our study individuals with suspected carotid stenosis were excluded as we were curious to investigate flow parameters undisguised by lumen restriction or embolic related pathologies. A few previous studies have linked CSVD manifestations to low EDV or increased pulsatility index in CCA, ICA or MCA, which is in line with our results.
associated lower diastolic blood pressure (BP) and higher pulse pressure (PP) to higher ICA-pulsatility, which in turn could be associated to WMH in 73-year-old subjects. Recently, Chuang et al.
showed that EDV in the common carotid arteries was negatively associated with a presence of CSVD and positively associated with white matter volume. Their interpretation of the results was that the low EDV was attributable to a generalized decreased flow because of an increased downstream resistance. To our knowledge, no previous study has assessed the association between carotid flow parameters and patterns of brain atrophy.
Strengths and weaknesses
A major strength of this study is the high participant age and the relatively high prevalence of brain lesions that follows, applicable to a community dwelling healthy elderly population. Another major strength of the study is the high magnetic field strength, and manual evaluation of MR-images by an experienced neuroradiologist. Visual assessment is, compared to computerized assessment, a lower risk for artefacts and more applicable to a normal clinical setting. All cases are assessed by the same neuroradiologist, and by the same ultrasound technician, which could be disputed as a weakness since internal validation is lacking, but also a strength since inter-rater variability would be low.
Although the variables “MRI-burden score” and “number of MRI-pathologies” are constructed based upon our specific variables, rendering them hard to compare to other studies, the construction of pooled variables is not uncommon. Shi & Wardlaw has suggested a whole-brain approach to the concept of CSVD,
as heterogenic manifestations of vascular burden, where number or severity of lesions might be a better evaluator of cognitive risk than the individual manifestations. In many studies, including this one, indeed pooled variables of MRI-burden show stronger associations to hemodynamic risk factors or, in other cases, cognitive decline. Possible explanations to the difference between our different MRI pathologies might be partly due to power.
After analysing the MRI lesions compared to each of the flow parameters of CCA, ICAp and ICAd, we chose for brevity reasons only to display CCA figures. Similarly, we chose to proceed with a mean of the left and right value for each of the carotid flow variables, to reduce statistical type one error. The rationales for this procedure were, firstly, the MRI-lesions are not reported according to side, why side specific analysis of flow might be misleading, Secondly, since stenotic subjects with the possibility of unilateral highly effected flow velocities have been excluded by study design, side to side variation might be smaller. Thirdly, comparisons were made to assure that results would not be affected.
Considering that our study is a part of a large population study, several risk factors were available for adjusting the models. We choose however not to adjust for vascular risk factors, contrary to many other similar studies, since exact causal pathogenic pathways are not known and our aim with this cross-sectional study was to test for associations, not prove causation.
Future studies
Combined results of this and previous studies suggest future intervention studies might focus on markers of arteriosclerosis rather than hypertension.
Conclusions
Manifestations of CSVD and brain atrophy are common in the ageing brain. The extensive distribution of different cerebral manifestations paints a heterogenic picture of brain aging. Carotid flow parameters, whether a manifestation of upstream or downstream vascular changes, can be linked to both focal and global manifestations of cerebral changes. Low EDV, a proposed marker of systemic arterial stiffness, is associated with increased CSVD burden and cortical atrophy, indicating a vascular component in the process of brain aging.
Author contribution
S.E is study PI and designed the initial study and sub-study, made an important intellectual contribution and critical draft revision. A.S-L. participated in sub-study design and acquisition, including expertise and important intellectual content, and critical draft revision. K.A-K. is responsible for neuroimage acquisition and interpretation. K.E has performed the statistical analysis and drafted the article. All authors have contributed to the intellectual content and revision of the article and has given final approval of the version to be published.
Supplementary material
The supplementary material is available in the electronic version of this article.
Declaration of Competing Interest
The authors declare no conflict of interest.
Acknowledgements
We greatly thank ultrasound technician Nour Moussa for careful examination of all participants, Zinka Tucek and Lena Berggren for administrative support, and statistician Mats Pihlsgård for aid in cluster analysis.
This study was funded by the Swedish Research Council (grants 521-2013-8604; 2017-01613), Gustaf V and Victoria Foundation and the Medical Faculty, Lund University.
WMC Fazekas grade 1 and MTA Scheltens grade 1 are coded as “no pathologies”
00005-8&id=gr1.jpg){kind=link}
00005-8&id=gr2.jpg){kind=link}
00005-8&id=gr3.jpg){kind=link}