As the population ages, dementia has become the focus of global attention, affecting more than 35 million people globally. Among them, dementia caused by vascular diseases ranks second in the prevalence of cognitive impairment. However, unlike other types of dementia, the pathologic mechanism of vascular dementia (VaD) is complex, and diagnosis and evaluation of VaD are expected to be more objective. The pathologic and clinical features of VaD depend on the mechanism of stroke and the extent of tissue loss, which may affect the connectivity of neural pathways. The underlying cellular molecular mechanism is mainly related to the insufficient perfusion of brain tissues caused by vascular stenosis, which activates inflammatory pathways and leads to oxidative stress response, chronic hypoxia, and neuroendocrine disorders, causing decline of cognitive function. However, the damage of brain white matter fiber structure plays a crucial role in the pathogenesis of VaD. And Diffusion tensor imaging (DTI) is particularly valuable as a new imaging technique for white matter integrity assessment including connectivity. Moreover, DTI can monitor the structural development process of VaD over time. This may not only provide prognostic information, but also objectively assess the response to the intervention.1 Currently, the only method that can noninvasively track white matter fiber tracts is DTI, which can provide information on the microstructure of the organization. The DTI is better at detecting the organizational structure and functional characteristics compared with any other diagnosis method. Fractional anisotropy (FA) and mean diffusivity (MD) are the 2 main parameters calculated based on DTI. FA is defined as the coefficient of variation of eigenvalues, whereas MD is consistent with the magnitude of water self-diffusion.2
Therefore, DTI is valuable in determining severity and prognosis of VaD. In this paper, the authors investigated potential utility of DTI in differential diagnosis of VaD by searching and analyzing the related literature in recent years.
METHODS
Literature Inclusion and Exclusion Criteria
The inclusion criteria were as follows: the study type was case-control study and the language was limited to English.
Exclusion criteria: duplicate publications; research papers without full text, incomplete information or inability to conduct data extraction; animal experiments; reviews and systematic reviews.
Search Strategy
In this meta-analysis, the authors searched Pubmed, Embase, and Cochrane Library from establishment of the database to Mar 2021. The search terms mainly included “diffusion tensor imaging,” “DTI,” “Vascular Dementia,” “Arteriosclerotic Dementia,” “Cognition,” and “Cognitive.”
Literature Screening and Data Extraction
The literature search, screening, and information extraction were all independently completed by 2 researchers. During any doubts or disagreements, the decision was made after discussion or consultation with a third person. The data extraction included the author, year, country, study design, sample size, mini-mental status examination, FA including corpus callosum (CC), Genu, Splenium, Hippocampus, Temporal lobe, anterior periventricular white (Ant-PVWM), posterior periventricular white (Post-PVWM), inferior fronto-occipital fasciculus (IFO), superior longitudinal fasciculus (SLF) and inferior longitudinal fasciculus (ILF) and MD including Genu, Splenium, Ant-PVWM , Post-PVWM, SLF, and ILF.
Literature Quality Assessment
Two researchers independently conducted literature quality evaluations using the Newcastle-Ottawa Scale (NOS) for cohort study.3 NOS included 4 items (4 points) for “Research Subject Selection,” 1 item (2 points) for “Comparability between Groups,” and 3 items (3 points) for “Result Measurement,” with a full score of 9 points and ≥7 is regarded as High-quality literature, whereas <7 is classified into lower-quality literature. When the opinions were inconsistent, the decision was made through discussion or consultation with a third investigator. The meta-analysis was performed based on the related items of the Preferred Reporting Items for Systematic Reviews and Meta-analysis statement (PRISMA statement).4
Data Synthesis and Statistical Analysis
STATA 15.1 was used to analyze the data. Weighted mean difference (WMD) [95% confidence interval (CI)] was used to evaluate the difference in FA and MD between patients with VaD and healthy individuals. The heterogeneity test value of P≥0.1 and I2≤50% indicated that there was homogeneity among studies, and the fixed effects model was used for combined analysis; whereas the corresponding values of P<0.1 and I2>50% indicated that the studies were heterogeneous and sensitivity analysis was used to find the source of heterogeneity. Even after this the heterogeneity was still large, the random effects model was used or the set of results were abandoned and descriptive analysis was used. Funnel plot and Egger test were used to analyze publication bias.
RESULTS
The Results of Literature Search
In this study, a total of 124 studies were retrieved from the databases. After eliminating duplicate studies, 91 unique studies were obtained. A total of 77 studies were obtained after browsing titles and abstracts. Finally, 8 articles were included in the meta-analysis through full-text reading (Fig. 1).
;)
FIGURE 1: Flow diagram for selection of studies.
Baseline Characteristics and Quality Assessment of the Included Studies
A total of 8 case-control studies were included in this meta-analysis. The sample size of patients ranged from 35 to 60, including 166 patients in the VaD group and 177 healthy controls. The age of the VaD group was 65.3(5.9) to 78.2(7.7) years, whereas the mean age of the control group was 63.9(5.3) to 74.3(7.5) years. The VaD group had a male-to-female ratio of 1.20, whereas the control group had a male-to-female ratio of 0.91. Patients in 6 included studies were from Asia, and in the other 2 studies the patients were from Europe. The NOS scores used for quality assessment were all above 7 and met the requirements (Table 1).
TABLE 1 -
Baseline Characteristics and Quality Assessment of the Included Studies
|
|
|
|
Sample Size |
Sex |
Age |
MMSE |
|
| Author |
Year |
Study Design |
Country |
VaD |
Control |
VaD |
Control |
VaD |
Control |
VaD |
Control |
NOS Score |
| D’Souza5
|
2017 |
Case-control |
India |
30 |
30 |
18/12 |
14/16 |
66.9±8.4 |
65.6±3.9 |
— |
— |
7 |
| Chen6
|
2009 |
Case-control |
China |
18 |
20 |
7/11 |
9/11 |
78.2±87.7 |
70.1±7.1 |
— |
— |
7 |
| Fu7
|
2012 |
Case-control |
China |
20 |
20 |
9/11 |
10/10 |
73.6±5.9 |
74.3±7.5 |
17.3±4.2 |
29.6±0.5 |
8 |
| Kim8
|
2011 |
Case-control |
Korea |
10 |
20 |
4/6 |
8/12 |
69.3±5.3 |
69.8±3.5 |
21.5±2.2 |
28.5±1.6 |
8 |
| Ostojic9
|
2015 |
Case-control |
Serbia |
25 |
25 |
16/9 |
15/10 |
69.5±8.8 |
71.6±7.5 |
— |
— |
7 |
| Sugihara10
|
2004 |
Case-control |
Japan |
20 |
10 |
4/6 |
8/12 |
65.4 |
72.4 |
— |
— |
7 |
| Wu11
|
2014 |
Case-control |
China |
30 |
30 |
17/13 |
16/14 |
65.3±5.9 |
63.9±5.3 |
>24 |
7 |
|
| Zarei12
|
2009 |
Case-control |
Netherlands |
13 |
22 |
10/3 |
9/13 |
74.3±7.0 |
70.7±6.0 |
22.1±3.3 |
28.7±1.4 |
8 |
MMSE indicates mini-mental status examination; NOS, Newcastle-Ottawa Scale; VaD, vascular dementia.
Results of Meta-analysis
FA
There were 2 studies, in which the FA values in CC of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=0.0%, P=0.543), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the FA values in the CC of VaD patients were significantly lower than those of the healthy controls (WMD=−0.05, 95% CI: −0.06 to −0.04, P=0.000; Fig. 2a).
FIGURE 2: Comparison of the fractional anisotropy values in corpus callosum (A), Genu (B), Splenium (C), Hippocampus (D), Temporal lobe (E), anterior periventricular white (F), posterior periventricular white (G), inferior fronto-occipital fasciculus (H), superior longitudinal fasciculus (I) and inferior longitudinal fasciculus (J) between patients with vascular dementia and healthy controls. CI indicates confidence interval; WMD, weighted mean difference.
In 4 of the included studies, the FA values in Genu of patients with VaD and healthy controls were compared. As there was significant heterogeneity (I2=95.5%, P=0.000), a meta-analysis was conducted through a random effects model. The pooled results showed that the FA values in the Genu of VaD patients were significantly lower than those of the healthy people (WMD=−0.10, 95% CI: −0.18 to −0.02, P=0.007; Fig. 2b).
There were 4 studies, in which the FA values in Splenium of patients with VaD and healthy controls were compared. As there was significant heterogeneity (I2=85.4%, P=0.000), a meta-analysis was conducted through a random effects model. The pooled results showed that the FA values in the Splenium of VaD patients were significantly lower than those of the healthy controls (WMD=−0.10, 95% CI: −0.15 to −0.05, P=0.000; Fig. 2c).
In 2 of the included studies, the FA values in hippocampus between patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=0.0%, P=0.459), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the difference between the FA values in the Hippocampus of VaD patients and the healthy people was not statistically significant (WMD=−0.01, 95% CI: −0.02 to 0.01, P=0.282; Fig. 2d).
There were three studies, in which the FA values in temporal lobe of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=0.0%, P=0.476), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the FA values in the temporal lobe of VaD patients were significantly lower than those of healthy controls (WMD=−0.03, 95% CI: −0.04 to −0.02, P=0.000; Fig. 2e).
In 2 of the included studies, the FA values in Ant-PVWM of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=27.8%, P=0.239), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the FA values in the Ant-PVWM of VaD patients were significantly lower than those of healthy controls (WMD=−0.04, 95% CI: −0.06 to −0.02, P=0.000; Fig. 2f).
There were 2 studies, in which the FA values in Post-PVWM of patients with VaD and healthy controls were compared. As there was significant heterogeneity (I2=92.6%, P=0.000), a meta-analysis was conducted through a random effects model. The pooled results showed that the difference between the FA values in the Post-PVWM of VaD patients and healthy controls was not statistically significant (WMD=−0.09, 95% CI: −0.18 to 0.01, P=0.082; Fig. 2g).
In 2 of the included studies, the FA values in IFO of patients with VaD and healthy controls were compared. As there was significant heterogeneity (I2=98.9%, P=0.000), a meta-analysis was conducted through a random effects model. The pooled results showed that the FA values in the IFO of VaD patients were significantly lower than those of healthy controls (WMD=−0.08, 95% CI: −0.15 to −0.02, P=0.012; Fig. 2h).
There were 2 studies, in which the FA values in SLF of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=0.0%, P=1.000), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the FA values in the SLF of VaD patients were significantly lower than those of the healthy controls (WMD=−0.03, 95% CI: −0.04 to −0.02, P=0.000; Fig. 2i).
There were 3 studies, in which the FA values in ILF of patients with VaD and healthy controls were compared. As there was significant heterogeneity (I2=80.2%, P=0.006), a meta-analysis was conducted through a random effects model. The pooled results showed that the FA values in the ILF of VaD patients were significantly lower than those of the healthy controls (WMD=−0.02, 95% CI: −0.04 to −0.01, P=0.002; Fig. 2j).
MD
The MD values in Genu of patients with VaD and healthy controls were compared in 2 of the included studies. As there was no significant heterogeneity (I2=0.0%, P=0.509), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the MD values in Genu of VaD patients were significantly higher than those of healthy controls (WMD=0.09, 95% CI: 0.05 to 0.14, P=0.000; Fig. 3a).
FIGURE 3: Comparison of the mean diffusivity values in Genu (A), Splenium (B), anterior periventricular white (C), posterior periventricular white (D), superior longitudinal fasciculus (E) and inferior longitudinal fasciculus (F) of patients with vascular dementia and healthy controls. CI indicates confidence interval; WMD, weighted mean difference.
There were 2 studies, in which the MD values in Splenium of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=0.0%, P=0.754), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the MD values in Splenium of VaD patients were significantly higher than those of healthy controls (WMD=0.17, 95% CI: 0.07 to 0.26, P=0.000; Fig. 3b).
There were 2 studies, in which the MD values in Ant-PVWM of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=89.7%, P=0.000), a meta-analysis was conducted through a random effects model. The pooled results showed that the MD values in Ant-PVWM of VaD patients were significantly higher than those of healthy controls (WMD=0.16, 95% CI: 0.06 to 0.25, P=0.002; Fig. 3c).
The MD values in Post-PVWM of patients with VaD and healthy controls were compared in 2 of the included studies. As there was no significant heterogeneity (I2=79.0%, P=0.029), a meta-analysis was conducted through a random effects model. The pooled results showed that the MD values in Post-PVWM of VaD patients were significantly higher than those of healthy controls (WMD=0.13, 95% CI: 0.01 to 0.26, P=0.038; Fig. 3e).
There were 2 studies, in which the MD values in SLF of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=70.8%, P=0.064), a meta-analysis was conducted through a random effects model. The pooled results showed that the MD values in SLF of VaD patients were significantly higher than those of healthy controls (WMD=0.07, 95% CI: 0.05 to 0.09, P=0.000; Fig. 3f).
There were 2 studies, in which the MD values in ILF of patients with VaD and healthy controls were compared. As there was no significant heterogeneity (I2=8.7%, P=0.295), a meta-analysis was conducted through a fixed effects model. The pooled results showed that the MD values in ILF of VaD patients were significantly higher than those of healthy controls (WMD=0.05, 95% CI: 0.02 to 0.08, P=0.000; Fig. 3g).
Sensitivity Analysis
To assess impact of any single included study on the results of the current meta-analysis, each included study was eliminated one by one using sensitivity analysis and a summary analysis was performed on the remaining studies. The result of the sensitivity analysis is shown in Figs. (S1–4), Supplemental Digital Content 1, https://links.lww.com/NRL/A103. The results showed that none of the studies had an excessive impact on the results of the meta-analysis, indicating that the results of the remaining studies were stable and reliable.
Publication Bias
The funnel plot of the comparison of the FA values in Genu of patients with VaD and healthy controls seemed basically symmetric (Fig. 4a), and the P value of Egger test was 0.986, indicating that there was no obvious publication bias in this study. In addition, the funnel plot of the comparison of the FA values in Splenium of patients with VaD and healthy controls seemed basically symmetric (Fig. 4b), and the P value of Egger test was 0.935, indicating that there was no obvious publication bias in this study.
FIGURE 4: Funnel plot of the comparison of the fractional anisotropy values in Genu (A), and Splenium (B) of patients with vascular dementia and healthy controls. WMD indicates weighted mean difference.
DISCUSSION
In 1672, Willis reported a case of VaD after stroke. On the basis of extensive research and exploration, currently, VaD is classified into 4 major subtypes: (1) poststroke dementia, defined as dementia onset within 6 months after a stroke; (2) subcortical ischemic VaD; (3) multi-infarct (cortical) dementia; and (4) mixed dementia.13
Even though all the 4 types of pathology are named VaD, their neuropathology are still very complex, involving both large and small blood vessels, which may also co-exist with degenerative diseases. According to the different neuropathologic mechanisms, it can be divided into small vessel disease (SVD is an umbrella term that includes various pathologic findings, including small infarcts, microscopic infarcts, microbleeds, arteriolosclerosis, intracranial atherosclerosis, and cerebral amyloid angiopath), small and microscopic infarcts (microinfarction is defined by imaging, especially infarct lesions less than 3 mm in diameter), large infarcts (VaD caused by large infarction is most closely related to the location of the infarction), microbleeds and other hemorrhages and shares characteristics with neurodegenerative pathology.14
Because of the complex pathologic basis of VaD, it is also difficult to diagnose and evaluate this disease, and it is even more difficult to predict and make an objective evaluation of this disease. In addition to the classic neuropsychiatric scale tests and blood and cerebrospinal fluid tests for the etiology of the disease, neuroimaging examination plays a very important role in the diagnostic assessment of VaD. Among them, magnetic resonance imaging (MRI) examination is particularly excellent because it provides more information in evaluating dementia lesions, and it is considered as the “gold standard” for evaluation based on neuroimaging.15
Among them, DTI is particularly valuable for white matter integrity, including connectivity. DTI was first proposed by Basser et al.16 It is a quantitative MRI technology developed and optimized by diffusion weighted imaging, which aims to detect ultrastructure damage through water diffusion measurement. DTI technology can observe small damages and divide the cortex into multiple areas. It also analyzes the correlation between the areas of brain and continuously measures changes in those areas. Hence, DTI can improve the detection of the lesion to the cognitive function.17 There are 3 types of VaD, like subcortical ischemic VaD, multi-infarct (cortical) dementia, and mixed dementia. The onset is insidious, so the prevention is difficult. Although DTI for VaD assessment may be earlier than clinical symptoms and conventional imaging diagnosis. In a DTI study of microstructural changes in VaD hippocampus, connectivity impairment was observed despite conventional MRI which indicated normal gray matter.
The main parameters of DTI include apparent diffusion coefficient, FA, MD, axial diffusivity, and radial diffusivity. The FA value reflects the anisotropy of water molecule diffusion and is sensitive to axonal integrity. It is the most used anisotropy indicator whose value ranges from 0 to 1. The higher FA values indicate more intact axon fibers. MD reflects the overall level of molecular diffusion and diffusion resistance.18 The MD coefficient varies according to the progressive stage of the injury, leading to an overall measurement of water molecule movements. The primary diffusion metrics (FA and MD) reflect overall white matter health, organization, and maturation.19 As a longitudinal diffusion metric along axons, axial diffusivity is indicative of axonal damage and is negatively correlated with axonal injury and positively correlated with axonal repair.20 Radial diffusivity reflects diffusivity perpendicular to the axon and higher values were previously associated with demyelination processes of white matter tracts both in humans and in animal models.21Application of combination of multiple indices can fully reflect the complex tissue structures of the brain and the changes in the microstructure of the tissue under pathologic conditions.
Currently, an increasing number of clinical and experimental studies use DTI as an evaluation index to quantify cognitive impairment that involves vascular and neurodegenerative diseases. White matter hyperintensity (WMHs) caused by SVD are taken as an example, and DTI can provide white matter microstructure and WMH penumbra information, and can monitor the changes in FA and mean diffusion coefficient in the DTI of these lesions over time that can indicate axonal damage.22 Although longitudinal studies in different populations consistently show that increased WMH volume predicts cognitive decline, mild cognitive impairment.23 Another group using DTI in a case-control study of lacunar stroke patients and controls showed that patients had less network efficiency than controls and those with WMH severity.24 DTI is more closely associated with cognitive dysfunction compared with conventional measures. Subsequent studies not only independently replicated these findings, but also showed that reduced network efficiency can predict conversion to dementia in SVD patients.25,26 Patients with cerebral amyloid angiopath also exhibit reduced network efficiency over time and are associated with cognitive dysfunction and cognitive decline.27,28
Because of its objective, quantitative, microscopic, and sensitive characteristics, DTI offers an advantage over conventional imaging markers in the study of related clinical features and disease involving dementia. In addition, owning to its advantages, it will be possible to distinguish the types, and evaluate the progress and prognosis of cognitive disorders.
In this study, DTI was used to diagnose vascular cognitive impairment, and it was indicated that it has advantages in the diagnosis of VaD. This study has some certain limitations. First, the included studies were all observational studies, and the quality of the evidence was relatively low. More randomized controlled studies need to be further included. There was heterogeneity in some studies; however, sensitivity analyses did not reveal sources of heterogeneity, and we were unable to perform subgroup analyses due to incomplete descriptions of the underlying characteristics of each study.
CONCLUSION
The results of this study showed that DTI imaging results of brain tissues in VaD patients, including FA and MD values at various sites, were significantly different from healthy individuals. Therefore, it may be feasible to use DTI imaging as a diagnostic method for VaD. It is noteworthy that the quality score of the literature included in this study is high, and there is no potential publication bias, indicating that the overall quality of this study is acceptable.
REFERENCES
1. Bir SC, Khan MW, Javalkar V, et al. Emerging concepts in
vascular dementia: a review.
J Stroke Cerebrovasc Dis. 2021;30:105864.
2. Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. 1996. J Magn Reson. 2011;213:560–570.
3. Cook DA, Reed DA. Appraising the quality of medical education research methods: the Medical Education Research Study Quality Instrument and the Newcastle-Ottawa Scale-Education. Acad Med. 2015;90:1067–1076.
4. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097.
5. D’Souza MM, Gorthi SP, Vadwala K, et al. Diffusion tensor tractography in cerebral small vessel disease: correlation with
cognitive function. Neuroradiol J. 2018;31:83–89.
6. Chen TF, Lin CC, Chen YF, et al. Diffusion tensor changes in patients with amnesic mild cognitive impairment and various dementias. Psychiatry Res. 2009;173:15–21.
7. Fu JL, Zhang T, Chang C, et al. The value of
diffusion tensor imaging in the differential diagnosis of subcortical ischemic
vascular dementia and Alzheimer’s disease in patients with only mild white matter alterations on T2-weighted images. Acta Radiol. 2012;53:312–317.
8. Kim MJ, Seo SW, Lee KM, et al. Differential diagnosis of idiopathic normal pressure hydrocephalus from other dementias using
diffusion tensor imaging. AJNR Am J Neuroradiol. 2011;32:1496–1503.
9. Jelena O, Dusko K, Aleksandra P, et al. Hippocampal
diffusion tensor imaging microstructural changes in
vascular dementia. Acta Neurol Belg. 2015;115:557–62..
10. Sugihara S, Kinoshita T, Matsusue E, et al. Usefulness of
diffusion tensor imaging of white matter in Alzheimer disease and
vascular dementia. Acta Radiol. 2004;45:658–663.
11. Wu X, Gao Y, Yang J, et al. Quantitative measurement to evaluate morphological changes of the corpus callosum in patients with subcortical ischemic
vascular dementia. Acta Radiol. 2015;56:214–218.
12. Zarei M, Damoiseaux JS, Morgese C, et al. Regional white matter integrity differentiates between
vascular dementia and Alzheimer disease. Stroke. 2009;40:773–779.
13. Skrobot OA, Black SE, Chen C, et al. Progress toward standardized diagnosis of vascular cognitive impairment: Guidelines from the Vascular Impairment of Cognition Classification Consensus Study. Alzheimers Dement. 2018;14:280–292.
14. Lobo A, Launer LJ, Fratiglioni L, et al. Prevalence of dementia and major subtypes in Europe: A collaborative study of population-based cohorts. Neurologic Diseases in the Elderly Research Group. Neurology. 2000;54(suppl 5):1–9.
15. Iadecola C, Duering M, Hachinski V, et al. Vascular cognitive impairment and dementia: JACC Scientific Expert Panel. J Am Coll Cardiol. 2019;73:3326–3344.
16. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophys J. 1994;66:259–267.
17. Smith EE, Beaudin AE. New insights into cerebral small vessel disease and vascular cognitive impairment from MRI. Curr Opin Neurol. 2018;31:36–43.
18. Xiao J, He X, Tian J, et al. Diffusion kurtosis imaging and pathological comparison of early hypoxic-ischemic brain damage in newborn piglets. Sci Rep. 2020;10:17242.
19. Liu J, Zhu J, Yuan F, et al. Abnormal brain white matter in patients with right trigeminal neuralgia: a
diffusion tensor imaging study. J Headache Pain. 2018;19:46.
20. Chen Q, Wang Z, Lv H, et al. Reorganization of brain white matter in persistent idiopathic tinnitus patients without hearing loss: evidence from baseline data. Front Neurosci. 2020;14:591.
21. Leonard M, Renard F, Harsan L, et al.
Diffusion tensor imaging reveals diffuse white matter injuries in locked-in syndrome patients. PLoS One. 2019;14:e0213528.
22. Maniega SM, Valdés Hernández MC, Clayden JD, et al. White matter hyperintensities and normal-appearing white matter integrity in the aging brain. Neurobiol Aging. 2015;36:909–918.
23. Alber J, Alladi S, Bae HJ, et al. White matter hyperintensities in vascular contributions to cognitive impairment and dementia (VCID): knowledge gaps and opportunities. Alzheimers Dement (NY). 2019;5:107–117.
24. Lawrence AJ, Chung AW, Morris RG, et al. Structural network efficiency is associated with cognitive impairment in small-vessel disease. Neurology. 2014;83:304–311.
25. Tuladhar AM, van Uden IW, Rutten-Jacobs LC, et al. Structural network efficiency predicts conversion to dementia. Neurology. 2016;86:1112–1119.
26. Tuladhar AM, van Dijk E, Zwiers MP, et al. Structural network connectivity and cognition in cerebral small vessel disease. Hum Brain Mapp. 2016;37:300–310.
27. Reijmer YD, Fotiadis P, Martinez-Ramirez S, et al. Structural network alterations and neurological dysfunction in cerebral amyloid angiopathy. Brain. 2015;138(Pt 1):179–188.
28. Reijmer YD, Fotiadis P, Riley GA, et al. Progression of brain network alterations in cerebral amyloid angiopathy. Stroke. 2016;47:2470–2475.
