Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: Stroke. 2024 Mar 25;55(4):801–811. doi: 10.1161/STROKEAHA.123.044172

Soluble biomarkers of cerebrovascular pathologies

Kate E Foley 1,2, Donna M Wilcock 1,2
PMCID: PMC10965230  NIHMSID: NIHMS1969731  PMID: 38527143

Abstract

Vascular contributions to cognitive impairment and dementia, VCID, is an all-encompassing term that describes cognitive impairment due to cerebrovascular origins. With the advancement of imaging and pathological studies, we now understand that VCID is often comorbid with Alzheimer’s disease (AD). While researchers in the AD field have been working for years to establish and test blood-based biomarkers for AD diagnosis, prognosis, clinical therapy discovery, and early detection, blood-based biomarkers for VCID are in their infancy, and also face challenges. VCID is heterogeneous, comprising many different pathologic entities (ischemic, or hemorrhagic), and spatial and temporal differences (acute or chronic). This review highlights pathways that are aiding the search for sensitive and specific blood-based cerebrovascular dysfunction markers, describes promising candidates, and explains ongoing initiatives to discover blood-based VCID biomarkers.

1. Introduction

1.1. Define VCID and various pathologies

Vascular contributions to cognitive impairment and dementia, or VCID, is an umbrella term adopted to collectively describe the various pathologic events of vascular origin that influence cognition1. VCID comprises both micro- and macro-scale cerebrovascular deficits that can be overt or subtle in clinical presentation, causing an array of cognitive and behavioral symptoms based on location and disease severity1. A prevalent component of VCID is cerebral small vessel disease (cSVD), which can be defined as a gradual reduction in efficiency and integrity of the small vessels of the brain including arterioles, capillaries, and venules. Researchers have defined six categories of cSVD: 1) arteriolosclerosis, 2) cerebral amyloid angiopathy (CAA), 3) inherited or genetic cSVD, 4) inflammatory mediated cSVD, 5) venous collagenases, and 6) other, suggesting that there are many presentations and vascular deficits in small vessels that can lead to cognitive impairment and dementia2–4.

1.2. VCID, masked by AD diagnoses, is the largest contributor to dementia

Alzheimer’s disease (AD) is certainly a leading cause of dementia; however it also often presents with cerebrovascular pathologies such as arteriolosclerosis, CAA, and superficial siderosis 5,6. A seminal study by Medina et al., showed that vascular changes are among the earliest change in biomarkers abnormality in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort, which has historically pre-screened their participants to reduce the presence of non-AD brain pathologies 7. Today, more than ever, it has become clear through many clinical-pathological studies that AD and VCID are often co-occurring, with cerebrovascular changes contributing to over 50% of dementia pathologies 8–13. A influential publication that reinforces this statistic came from the researchers of the Religious Orders Study and the Memory and Aging Project (ROS-MAP). In this clinical-pathological study, approximately 40-45% of brains with AD pathologic changes also had comorbid cerebrovascular pathologies at autopsy 14–16. Given the prevalence of cerebrovascular changes in the aging brain, it is clear that accessible, sensitive, and specific antemortem biomarkers are needed to identify VCID pathologies.

2. The need for accessible, sensitive, and specific biomarkers of VCID

Biomarkers are invaluable tools to help patients, doctors, and researchers. A good biomarker for a disease or risk can be helpful in understanding susceptibility, making accurate diagnoses, stratification of patients during a clinical trial, establishing prognosis, monitoring disease progression, or demonstrating target engagement with therapy. Existing biomarkers for VCID pathologies derive from the neuroimaging field, reflecting tissue damage like white matter hyperintensities (WMH), infarcts, and microbleeds. Fluid biomarkers afford the possibility of identifying specific molecular changes that precede the occurrence of these structural events by months or years. Sensitivity is the measure of a test accurately identifying a diseased individual as positive for the biomarker, while specificity is the measure of a test accurately identifying a non-diseased individual as negative for the biomarker17. The ideal biomarker will have both high sensitivity and specificity. Robust validation of fluid biomarkers is essential, including reproducibility across sites performing the assay, sensitivity and specificity across diverse populations, both racially, ethnically, and underlying disease pathologies including neurodegenerative and cerebrovascular changes 18.

3. Potential pathways that could be candidates for cerebrovascular pathology biomarkers.

When considering potential fluid biomarkers for VCID it is important that we consider the potential biological pathways that have been implicated in VCID, whether through preclinical studies or human autopsy studies. In section 3, we discuss the major pathways that are being considered for fluid biomarker development.

3.1. Vascular Injury - Endothelial cell dysfunction & Angiogenesis

Endothelial cells are the first barrier to the brain in the blood brain barrier system and adhesion molecules such as ICAM-1 and VCAM-1, angiogenic mediators such as members of the VEGF family, and inflammatory cytokines such as IL8, can all be measured in plasma and CSF. Endothelial cells are extremely dynamic and have been shown to ‘activate’ and participate in inflammatory signaling. In 1998, endothelial cell activation was defined as having five major components: 1) blood brain barrier breakdown at the endothelial cell level, 2) upregulation of leukocyte adhesion molecules on the endothelial surface (e.g. ICAM-1, VCAM-1 (Table 1), E-selectin (binds P-selectin on immune cells)), 3) a shift from homeostatic anti-thrombotic to prothrombotic, 4) production of inflammatory cytokines (e.g. IL6 (Table 1) and IL8), and 5) increased HLA molecule signaling 44,45.

Table 1: A subset of promising soluble biomarkers and their association with cerebrovascular pathologies.

Biomarker of Interest Association
CRP (C-reactive protein) Increased in plasma with risk for AD and VCID Dementia19
no difference in plasma between lacunar or cortical strokes20
Increased in plasma with risk of recurrent ischemic stroke21
increased in plasma risk of major vascular event21
increased in plasma with decreased fractional anisotropy in deep WM22
increased in plasma with increased mean diffusivity in deep WM22
Increased in plasma with lower fractional anisotropy in periventricular WM22
Increased in plasma with greater WM hyperintensity volume in Apoe4 carriers22
no association with cerebral infarcts22
no association with microbleeds22
C1q (Complement component 1q) increased in plasma after acute ischemic stroke23
increased in plasma with greater infarct volume post-stroke 23
increased in plasma with worsened score on NIHSS23
Increased in plasma following intracerebral hemorrhage24
Increased in plasma with worse outcomes on Glasgow Coma scale24
Increased in plasma with larger hematoma volume24
Increased in plasma with worsened 3mo outcome 24
C3/C3a (Complement component 3/3a) Plasma C3a increased early in aneurysmal subarachnoid hemorrhage25
Plasma C3 increased at 10days and 3mo post ischemic stroke26
Plasma C3a elevation limited to acute phase26
Elevated plasma C3 in LVD ischemic stroke indicated worsened 3mo outcome26
elevated plasma C3a in acute phase associated with worsened 3mo outcome26
Plasma C3a elevated in ischemic stroke at 1,3, and 5 days post-stroke27
Plasma C3a elevated at day28 post ischemic stroke27
ET-1 (Endothelin-1) Increased in plasma with increased WMH volume28
Increased in plasma with increased lacune counts28
Increased in plasma with increased microbleed counts28
Increased in plasma over course of 6 years28
elevated in plasma with ischemic stroke29
Increased in plasma 24hrs after nonhemorragic infarct stroke30
ICAM-1 (Intercellular adhesion molecule 1) Increased in plasma with increased WMH volume28
Increased in plasma with increased lacune counts28
Increased in plasma with increased microbleed counts28
increased in plasma over course of 6 years 28
not associated in plasma with increased risk for AD or VCID28
no difference in plasma between lacunar or cortical strokes19
increased in plasma with occurrence of lacunar infarction with early neurological dysfunction31
Plasma levels associated with poorer outcome at 3mo post lacunar stroke31
Increased in plasma with apparent leukoraiosis at 90 days post stroke32
increased in plasma with increased WMH volume 33
Increased in plasma with silent cerebral infarcts33
IL6 (Interleukin 6) Increase in plasma and CSF associated with risk for AD and VCID Dementia19
increased in plasma with increased risk for radiological progression of cSVD34
increased in plasma with increased risk for new lacunes (2yrs) 34
increased in VCID patient plasma over AD, MCI, and normal cognition35
no difference in plasma between lacunar or cortical strokes20
increased in plasma with lacunar infarction with early neurological dysfunction31
Plasma levels associated with poorer outcome at 3mo post lacunar stroke31
Plasma levels increased after stroke (peak at 4d) 36
Plasma levels higher in ischemic over hemorrhagic stroke at days 1 and 436
MMP9 (Matrix metalloproteinase9) Plasma levels increased in ischemic stroke37
Plasma levels associated with increased infarct volume37
Plasma levels increased in patients with slower recovery post-stroke at 48hr37
no difference in plasma of CAA-induced hemorrhagic stroke patients38
Plgf (Placental growth factor) Plasma levels associated with increased Fazekas score39
Plasma levels associated with greater functional cognitive impairment39
TNFα (Tumor necrosis factor alpha) no difference in plasma levels between lacunar or cortical strokes20
increased in plasma with lacunar infarction with early neurological dysfunction31
Plasma levels associated with poorer outcome at 3mo post lacunar stroke31
Plasma levels increased after stroke from days 1-9036
Plasma levels increased in patients with periventricular temporoparietal or juxtacortical and deep WMH without microbleeds (Fz =3) 40
VCAM-1 (Vascular cell adhesion molecule 1) Plasma levels increased with increasing WMH volume28
Plasma levels associated with increase in lacune counts28
Plasma levels associated with increase in microbleed counts28
Plasma levels increased over course of 6 years 28
Plasma levels associated with leukoraiosis at 90 days post stroke32
VEGF (Vascular endothelial growth factor) Plasma levels increased from day 0–14 in patients with ischemic stroke41
Plasma levels associated with larger infarct volume41
Plasma levels associated with worsened clinical disability41
Plasma levels associated with peripheral leukocytosis41
Plasma levels are higher in large vessel disease patients than small vessel disease patients42
Higher plasma levels acutely after stroke predict better outcome42
Plasma VEGF is associated with increased incidence of stroke/TIA43
Increased in plasma associated with increased number of lacunes32
Plasma levels increased with greater global small vessel disease burden32
Plasma levels are lower in patients with widespread periventricular WMH in all hemispheres, also with lacune, microbleeds, and atrophy (Fz=3) 40
Open in a new tab

Acronyms: AD; Alzheimer’s disease; VCID: vascular contributions to cognitive impairment and dementia; WM: white matter; Apoeε: apolipoprotein epsilon; mo: months; WMH: white matter hyperintensity; Fz: Fazekas score.

When considering candidate fluid biomarkers that may reflect endothelial cell activation, soluble vascular cell adhesion molecule 1 (VCAM-1) and intercellular adhesion molecule 1 (ICAM-1) are cleaved cell adhesion molecules that are respectively associated with worsened WMH volume and overall stroke outcomes when measured in plasma 31,46. Additionally, plasma ICAM-1 has been associated with lacunar stroke prevalence. While blood brain barrier breakdown can occur across multiple cells of the neurovascular unit (highlighted in the next section), endothelial tight junction disassembly increases permeability of blood molecules (fibrin, albumin, immunoglobulin) into the brain. These mechanisms of endothelial cell activation provide initial avenues for soluble biomarker research as markers for more widespread cerebral small vessel disease (Table 1).

Endothelin (ET-1) is a potent vasoconstrictor peptide expressed by endothelial cells, initially presenting as preproendothelin, and then cleaved during hypoxic events to its active signaling form to alter vasodilation 47,48. Plasma ET-1 is elevated in individuals after various types of strokes, and also was shown to predict worse outcomes in future WM lesions 28–30,49. ET-1 has also been associated with VCID risk with increased levels revealing worse WMHs and increased lacunes and microbleeds50. Garcia et al., highlight these proteins as well as many other endothelial based biomarkers in their extensive review50.

Angiogenesis is the term that describes the process of creating new blood vessels. This process occurs during development and continues throughout old age in response to various injuries and ischemic events. Angiogenesis is known to be significantly dysregulated in tumor formation and growth, and angiogenesis is impaired following ischemic and hemorrhagic strokes 51,52. After a CNS injury such as ischemia, or in the presence of neurodegeneration, the process of angiogenesis is essential. The VEGF family includes proteins measurable in plasma and CSF such as VEGF-A, VEGF-B, VEGF-C, VEGF-D, and PLGF (placental growth factor). These are prominent molecules in vasculogenesis and angiogenesis (See Section 4.1 and 4.3, Table 1) 53,54. When a cell is oxygen deficient, HIF-1α is expressed which then induces the release of VEGFs. VEGFs also form dimers (homodimers and heterodimers) which are necessary for receptor binding. VEGFs have several receptors including FLT1, FLK1 and KDR 55–57. Disruptions in VEGF signaling can result in abnormal vascular growth and formation, resulting in dysfunctional cerebrovascular pathogenesis (Table 1).

3.2. Blood brain barrier breakdown – pericytes and basement membrane

The blood brain barrier is formed and maintained through careful regulation and communication between neurovascular unit cells, including endothelial cells, pericytes, smooth muscle cells (SMCs), the basement membrane, and astrocytes. These cells work in concert to allow appropriate exchange of nutrients and waste between the brain and blood, including many highly dynamic processes (e.g., trans-endothelial cell migration, pericyte/SMC autoregulation, angiogenesis, etc.) 58. Platelet-derived growth factor receptor-β (PDGFRβ) is a marker specifically expressed on pericytes that appears to be shed and therefore measurable in plasma and CSF 59,60. Recently, pericyte marker PDGFRβ has been shown to shed in response to Aβ or hypoxic exposure, creating soluble PDGFRβ (sPDGFRβ) which can be measured in plasma (Table 1) 61, 62. Pericyte loss through cell death can result in reduced blood flow and vascular uncoupling63, 64. Improper vasodilation or maintenance of vascular walls can contribute to improper blood flow creating an environment primed for further VCID pathologies.

The basement membrane of the neurovascular unit acts as the structural component with multiple layers of extracellular matrices supporting the cerebrovasculature. The molecular mechanisms involved in basement membrane remodeling relate to angiogenic and inflammatory processes that can disrupt the vasculature and, therefore, present an opportunity for measurement as candidate biomarkers. Multiple collagen- and laminin-based networks (as well as other structural components) comprise this matrix and allow physical cell-to-cell communication among astrocytes, endothelial cells, and pericytes, to remain stable65. Basement membrane restructuring is a process that is delicately balanced between digesting and building; with breakdown of the basement membrane partially mediated through matrix metalloproteases (MMPs) and tissue inhibitor of metalloproteases (TIMPs). When this process becomes unbalanced, in the case of stroke, mutations, or AD there is increased risk for cerebrovascular pathologies that contribute to VCID (See Section 4.2, Table 1) 65,66. Several MMP proteins can be measured in plasma and CSF including MMP9 and MMP2 and may act as indicators of vascular remodeling and basement membrane breakdown.

3.3. Inflammation

Inflammation can be both a cause and consequence of cerebrovascular deficits like hypoperfusion, infarction, or ischemia 67. Interleukin-6 (IL6) is a pleiotropic proinflammatory molecule released during strokes to target the liver to upregulate C-reactive protein (CRP) and fibrinogen to aid in stroke repair2,19. IL6 is one of the systemic cytokines, measurable in plasma and CSF, that can be produced by monocytes or endothelial cells and that can drive a vascular response50,68. Studies show that IL6 can stimulate the cerebrovasculature to upregulate angiotensin II type 1 (AT1), enhance vasoconstriction, promote atherosclerosis, and affect blood brain barrier permeability68,69. IL6 has been consistently identified as upregulated in the plasma and brain tissue of patients with dementia, and in some studies, IL6 is able to differentiate between VCID and other dementias (Table 1) 34,35,70. Across multiple diverse populations, increased in IL-6 is associated with more lacunes, WMHs, and infarcts. Interestingly, elevated plasma IL-6 also predicted increased risk for vascular events, suggesting that IL-6 may be a blood-based biomarker for VCID in both a predictive and responsive method. A remaining question regarding IL6 is whether its expression may be contingent on existing neurodegenerative pathologies. Indeed, there is a study that suggests CSF IL6 may be neuroprotective 71.

Plasma C-reactive protein (CRP) is a frequent indicator of systemic inflammation that is assessed in routine clinical laboratory testing. More specifically, CRP increases in response to stroke, is associated with worsened WMHs, and is chronically elevated in VCID through multiple mechanisms (e.g. dampen angiogenesis, increase ET-1 release, increase atherosclerosis, decrease the expression of eNOS, etc.) 50. Plasma CRP increases platelet binding signals on endothelial cells, and also increases atherosclerosis, and angiogenesis72. One study identified that increased plasma CRP levels at midlife are associated with worsened WM integrity in both deep WM regions and periventricular WM regions 22. Further, the plasma CRP correlation to WMH volume was conditional on an APOEε4 negative genotype 22. Plasma CRP levels at midlife do not predict infarcts or hemorrhagic microbleeds later in life 22. While plasma CRP may not be predictive for later cSVD or VCID pathologies, it may serve as a biomarker of ongoing WM damage. Plasma CRP could be candidate biomarker for negative cognitive outcomes due to white matter abnormalities in cerebrovascular pathologies (Table 1) 20,21.

TNFα is elevated in the plasma of patients following acute ischemic and acute lacunar strokes (Table 1) 31,73. To date, TNFα has been linked to many cerebrovascular-relevant processes including regulation of blood brain barrier permeability, myelin dysregulation, reduced nitric oxide (NO) production in endothelial cells, and impaired vasodilation 74–76. IL-6, TNFα, and IL1β, another pleiotropic immune mediator, all signal to each other interchangeably to propagate further inflammation, and all have been found to be upregulated in the local tissue shortly after an ischemic event, especially in the white matter 70,77.

Some studies suggest that chronic increased inflammatory signaling can trigger occlusions and ischemia (nicely reviewed by Mun, Hinman78). The complement cascade is an innate inflammatory response that can provide both beneficial or harmful effects, depending on which of the three cascades that is initiated, the injury type, and acute or chronic signaling. Key components that are measurable in plasma include soluble complement receptor 1 (sCR1), complement component 3 (C3), and complement component 1q (C1q). These have been implicated in stroke response and tissue repair (Table 1) 79,80. The complement cascades have been linked to beneficial effects during stroke, with Clarke et al., arguing that timing of complement inhibition may be the most important factor, and that complement is beneficial for recovery after stroke79,81. Biomarker studies have identified that C1q increased in the plasma of intracerebral hemorrhage patients who fared worse on hematoma volume 24. Plasma C1q also has been shown to serve as a biomarker for acute ischemic stroke, differentiating stroke patients with higher C1q in the blood from non-stroke patients (Table 1) 23. Plasma C3a levels are also increased in blood after hemorrhage and can indicate worsened tissue and cognitive outcomes25,26. Plasma C3a is elevated in ischemic stroke patients on days 1,3, 7, and 28 after stroke27. C5a has also been shown to increase in the plasma on after stroke on days 7 and 14 27. Other complement proteins such as SC5b-9 and C4d were increased in the plasma of ischemic stroke patients and correlated with worsened cognitive and functional outcomes 82. Overall complement proteins are significantly associated with both ischemic and hemorrhagic strokes and their outcomes, however their specificity and sensitivity for the various types of strokes and for use as biomarkers have yet to be further evaluated. Inflammatory proteins may lack the specificity to serve as a VCID biomarker given the many peripheral sources of these inflammatory mediators including peripheral inflammatory conditions and acute infections.

4. A deep dive into three current promising biomarkers (summarized in Figure 1 and Table 1)

Figure 1: Cerebrovascular dysfunction candidate biomarkers showing major sources of each biomarker and how they are associated with one another biologically.

Figure 1:

Open in a new tab

These processes are discussed in detail in sections 3.1, 4.1 and 4.2 (ICAM1, VCAM1, VEGF, PlGF and VEGFR), 3.2, 4.2 and 4.3 (MMPs including MMP9), and 3.3 and 4.3 (CRP, TNFα, IL6, C1q and C3). Acronyms: VEGF: vascular endothelial growth factor; PlGF: Placental growth factor; ET-1; endothelin 1; CRP; c-reactive protein; MMP9; matrix metalloproteinase 9; TNFα: tumor necrosis factor alpha; C1q: complement component 1q; C3: complement component 3; IL6: interleukin 6; VCAM1: vascular cell adhesion molecule 1; ICAM1: intercellular adhesion molecule 1; VEGFR1, 2, 3: VEGF receptor 1, 2, 3. Created with BioRender.com

4.1. Placental Growth Factor - PLGF - candidate biomarker for WMH and cognitive dysfunction due to cerebral small-vessel disease.

Placental Growth Factor (PLGF) is an angiogenic regulator in the VEGF growth factor family. Although named after its primary role and highest expression in placenta vascularization, PLGF is also expressed in the thyroid, fat, lymph nodes, heart, and brain 83. PLGF isoforms can be membrane bound, supplying communication via probable autocrine signaling, or diffusible, as a possible paracrine signal84. PLGF can dimerize with itself, or with VEGF-A, to modulate the angiogenic processes such as endothelial cell survival, vascular permeability, migration, or promote non-branching angiogenesis 50,85. PLGF can bind to VEGFR1 (aka FLT-1) or neuropilin1 (NP-1), propagating pro-angiogenic responses. Interestingly, PLGF can help boost angiogenesis through binding VEGFR1, leaving more VEGF-A to bind the more potent and pleiotropic VEGFR2 (aka FLK-1 or KDR) 56,84.

Previously, PLGF has been examined as a soluble plasma biomarker identifying risk for preeclampsia (high blood pressure during pregnancy), which can lead to eclampsia and have harmful effects to mother and baby. Low PLGF levels during early pregnancy can result in hypoperfusion of the placenta and inefficient blood exchange between the two entities 85. Clinical evaluation found plasma PLGF levels to be highly efficient in sensitivity and predictability as a marker for preeclampsia 84,85. Only recently has PLGF been identified as a possible soluble biomarker for VCID. With strong linkages between impaired angiogenesis and VCID mechanisms, PLGF was added to list of plasma biomarkers to be explored by the MarkVCID NIH consortium established in 2016 86,87. A pivotal publication from the MarkVCID group provided convincing evidence for PLGF as a sensitive biomarker for both WMH pathology and cognitive decline 39. Work by the MarkVCID consortium showed increased odds ratio (OR) of 1.16 for worsened WM deficits measured by Fazekas score (0=none, 4 = severe WM burden), and an increased OR of 1.22 for functional cognitive impairment (CDR scale), with increased PLGF in the blood. Further, Hinman et al., also demonstrated that given the presence of Fazekas score 2 or above, plasma PLGF increased in diagnostic accuracy from no cognitive impairment (CDR = 0; ROC AUC = 0.66), to mild cognitive impairment (CDR = 0.5; ROC AUC 0.74), and performed best for those with dementia (CDR = 1; ROC AUC = .89) 39. Another study also showed that increased levels of plasma PLGF indicated increase white matter lesions, regardless of cognitive status or Aβ presence 88.

It is thought that PLGF could not only serve as a sensitive and specific biomarker of cerebrovascular injury but also provides us with mechanistic avenues to explore for VCID. This is being explored by several groups currently.

4.2. VEGF-A – candidate biomarkers for dysregulated angiogenesis.

VEGF-A and PLGF are both key modulators of angiogenesis, and therefore have similar, yet different, responses to VCID pathologies. VEGF-A is a rate limiting step in the angiogenesis pathways that, once produced, aims to bind VEGF receptors (e.g., VEGFR1/FLT1, VEGFR2/KDR/FLK1, VEGFR3/FLT4), which are receptor tyrosine kinases to propagate the angiogenesis signal. While there are multiple VEGF molecules (e.g., VEGFA, VEGFB, VEGFC, and VEGFD), VEGF-A is the predominant form with multiple downstream effects such as vasodilation, anti-apoptosis, hypotension, vascular permeability, monocyte chemotaxis and more54. As VEGF-A is a critical component for blood vessel growth and remodeling, it is tightly regulated through two major mechanisms; oxygen levels, and growth factors 54. Chronic hypoperfusion, ischemic, and/or hemorrhagic strokes can all affect oxygen availability. When oxygen is depleted, ‘HIF-1α – oxygen’ complexes become unbound and signals for angiogenesis through upregulation of VEGF mediated pathways. Growth factors, such as TGFα, TGFβ, IGF1, IL1α, or IL6, synthesized by multiple different cerebral cell types, can also increase VEGF expression.

VEGF-A expression increases rapidly after stroke, with elevated levels located near the stroke as early as 3 hours after insult, peaking at 24 hours, and can be detected for as long as 7 days89. In 2000, Slevin et al., took serial serum measurements from patients with ischemic strokes and identified increased VEGF-A in the blood that peaked at 7 days post-stroke, and was detected up to 14 days post stroke 41. Further experiments showed increases in VEGF-A in large vessel strokes at both an acute (24hr) and chronic (3mo) timepoint post stroke, but no elevation in serum levels in small vessel ischemic stroke 42. Additionally, multiple studies identified a reduction in plasma VEGF-A levels in AD patients 90–92. Results from the Framingham Heart Study published in 2013, showed that increased VEGF-A in serum associated with increased risk for stroke, suggesting baseline VEGF-A may be a candidate biomarker for the prediction of future strokes 43. Others showed that 90-days post-stroke, blood levels of VEGF-A associated with lacunes and overall cSVD burden32. Interestingly, increased serum VEGF levels 24 hours post stroke predicted better stroke outcomes (NIH stroke scale, NIHSS) at 90 days42. Similarly, increased levels of VEGF-A and VEGF-D were noted in the CSF of MCI individuals with WM lesions, however in cognitively normal individuals, the VEGFs showed no such association. One study evaluating serum VEGF and MRI cSVD markers found that decreased levels of serum VEGF-A associated with more WMH, lacunes, microbleeds, atrophy, and cognitive impairments in participants with advanced Fazekas scores (Fz>2) 40. Other studies have reported no association between VEGF-A and cSVD pathologies such as WMH or microbleeds, however this study’s sample was community based and only included those with low prevalence of cardiovascular disease (CVD) risk factors (i.e., excluding those with strokes and infarcts) 93. This is not surprising as CVD has been linked to serum VEGF levels in an inverted ‘U’ pattern with lower VEGF in participants with lower risk for CVD94.

Overall, it is likely that the potential of VEGF-A as a biomarker for VCID is moderate at best. VEGF-A is implicated in many disease processes. Also, as the “master regulator” of angiogenesis, being the most potent of the VEGF family and the most abundant, there are many regulators of its impact. For instance, with VEGF-A up- or down-regulated, the levels of homodimer and heterodimer with other VEGF family members are likely modulated to tune the impact of VEGF-A. Also, VEGF-A receptors can be up- and down-regulated in response to injuries, either by their internalization or insertion in the membrane, or enzymatic cleavage from the membrane. What is most likely is that VEGF-A as a ratio with either circulating receptor levels, or another member of the VEGF members, will provide more specific potential as a biomarker for VCID.

4.3. MMPs – candidate biomarkers for blood-brain barrier integrity.

Matrix metalloproteinases (MMPs) are a family of enzymes involved in the breakdown of matrices, collagens, and basement membranes, but also are key in signaling cascades for ischemia and inflammation 95. Secreted as an inactive zymogen form (proenzymes), most MMPs require cleavage activity prior to activation and signaling. In humans, there are 23 different MMPs, each with different sources, organ locations, and targets. The main groups of MMPs involved in cerebrovascular pathophysiology, are the gelatinases (MMP2, MMP9), collagenases (MMP1, MMP9, MMP13), and membrane bound MMPs (MMP14, MMP15, MMP16, MMP17, MMP24, MMP25). With cascade-like signaling transductions, there are also MMP inhibitors; tissue inhibitor of metalloproteinases (TIMP1, TIMP2, TIMP3, and TIMP4), that are required to regulate protease activity. There is significant redundancy in MMP systems, such that they activate each other, and can also be activated by many other molecules such as proinflammatory TNFα and IL1β, hypoxic responding NO or HIF-1α release, and even the Aβ peptide 95–98. Some MMPs are responsible for breaking down membranes and extracellular matrices during angiogenesis or blood brain barrier remodeling but dysregulation of the delicate balance of MMPs can result in increased tissue damage, cerebrovascular permeability, edema, and hemorrhages 99.

MMP9 is minimally expressed during homeostatic conditions in the brain but in the early phase post-stroke (>24 hours) there is a strong increase in MMP9 in the surrounding tissue, initially attributed primarily to infiltrating leukocytes 96. Later, MMP9 is elevated in the plasma patients during the late phases of stroke, and is associated with increased hemorrhagic transformation, increased infarct volume, and worsened cognitive outcomes 37,100. MMP9 expression in the brain tissue and the plasma can be concomitant with CAA and increased CAA-associated microbleeds 38,101. Preclinical studies evaluating genetic and pharmacological inhibition of MMP9 improved outcomes after stroke including attenuated tight junction protein loss, reduced tissue loss, and improved cognitive outcomes 102–104. Elevated brain MMP2 has also been associated with ischemic tight junction protein loss 105.

5. Challenges and Future Directions

Current biomarkers, primarily neuroimaging biomarkers, reflect the end result of diverse biological mechanisms including angiogenesis, inflammation, hypoxia, and stress. Therefore, these biomarkers lack the necessary specificity to inform the underlying mechanisms of an individual’s cerebrovascular pathologies. Neurofilament light (NfL) is also being evaluated in VCID, however, NfL is not reflective of a specific pathologic process but rather is a final measure of neuronal damage that most closely associates with cognitive impairment106. While the AD field utilized CSF for a number of years before plasma biomarkers appeared promising, one benefit of studying VCID is that many of the biomarkers derive from the vasculature itself, therefore it is likely that plasma biomarkers can be developed more readily. While there are some candidate biomarkers in CSF, it is clear the aging and dementia field is moving toward plasma biomarkers and this should be the target for the VCID field as well.

A major challenge facing the field for the discovery of fluid biomarkers of VCID is identifying molecules that will be specific for cerebrovascular pathologies. Many of those candidate biomarkers outlined above are likely also impacted by peripheral cardiovascular disease such as atherosclerosis, coronary artery disease and myocardial infarction. In addition, biomarkers of inflammation are also elevated in many conditions including arthritis, autoimmune disorders, acute injuries and acute infections. It is likely that specificity will be achieved through utilizing ratios, as has been successful in the AD field, or through the combination of neuroimaging and fluid biomarkers. Also, in a lesson learned from the AD field, an ideal fluid biomarker will have a broad dynamic range, allowing more precise identification of a cutoff that will provide optimal sensitivity and specificity.

One of the largest obstacles is the need for a real “ground truth” against which we can evaluate our candidate biomarkers. The fluid biomarkers for AD advanced rapidly with the arrival of amyloid PET and tau PET. These PET scans provided the AD field with “ground truth” in large populations that accelerated development of plasma biomarkers. For VCID, we have WMH and phenomenon such as microbleeds, however, we do not have antemortem specific measures of pathologies like arteriolosclerosis, microinfarcts, microbleeds or CAA. In the absence of ground truth in large population cohorts we must rely on autopsy identification of these pathologies, however, well-characterized autopsy cohorts do not provide the sample size needed to fully characterize candidate fluid biomarkers. In addition, many autopsy cohorts focused on cognitive impairment derive from memory centers / Alzheimer’s disease research centers that excluded overt cerebrovascular comorbidities for many years in their cohorts.

The biomarkers highlighted in this review are promising, yet still early. The NIH has funded several large consortia that will provide well-characterized longitudinal cohorts of individuals with cerebrovascular disease for the assessment and validation of novel biomarkers. MarkVCID 86,87, DISCOVERY 107 and DiverseVCID (see https://diversevcid.ucdavis.edu/) are key examples of these consortia. Also, longitudinal cardiovascular cohorts such as Framingham 108, ARIC (Atherosclerosis Risk In Communities) 109 and MESA (Multi-Ethnic Study of Atherosclerosis) 110 are all including neurodegenerative and cognitive measures in their studies now that the cohorts are at the age of risk for dementia. These multi-center, diverse consortia will provide well-characterized, robust populations for biomarker development.

6. Summary and vision for future

While not exhaustive, this review aimed to summarize the most promising fluid biomarkers being evaluated and validated for VCID. It is important to note that at the time of this writing there are no validated fluid biomarkers for VCID and more research is needed on the candidates discussed here, as well as discovery of new candidates. The utility of fluid biomarkers in the future cannot be overstated. It is likely that fluid biomarkers allow us to detect early changes to the cerebrovasculature that precedes the appearance of neuroimaging abnormalities such as WMH, microinfarcts or microbleeds, potentially by years. This is because the fluid biomarkers reflect early events in the pathologic cascades that initiate the sequelae leading to the ultimate vascular pathologic hallmarks. Furthermore, fluid biomarkers will likely also represent sensitive biomarkers to confirm target engagement in future clinical trials since the fluid biomarkers reflect components of biologic processes that will be targeted by therapies.

Funding Sources:

Work funded by NIH grants UF1NS115390 and R01NS116990 (DW).

Disclosures:

Dr Wilcock reports travel support from Alzheimer’s Association; compensation from Alzheimer’s Association for other services; and travel support from National Institutes of Health.

Abbreviations:

VCID

Vascular contributions to cognitive impairment and dementia

AD

Alzheimer’s disease

CAA

Cerebral amyloid angiopathy

WMH

White matter hyperintensities

ICAM-1

Intercellular adhesion molecule 1

VCAM-1

Vascular cell adhesion molecule 1

VEGF

Vasecular endothelial growth factor

PLGF

Placental growth factor

ET-1

endothelin-1

MMP

Matrix metalloproteinase

TIMP

Tissue inhibitor of metalloproteinase

IL

Interleukin

TNFα

Tumor necrosis factor alpha

References

  • 1.Zlokovic BV, Gottesman RF, Bernstein KE, Seshadri S, McKee A, Snyder H, Greenberg SM, Yaffe K, Schaffer CB, Yuan C, et al. Vascular contributions to cognitive impairment and dementia (VCID): A report from the 2018 National Heart, Lung, and Blood Institute and National Institute of Neurological Disorders and Stroke Workshop. Alzheimer’s & Dementia. 2020;16:1714–1733. doi: 10.1002/alz.12157 [DOI] [PubMed] [Google Scholar]
  • 2.Inoue Y, Shue F, Bu G, Kanekiyo T. Pathophysiology and probable etiology of cerebral small vessel disease in vascular dementia and Alzheimer’s disease. Molecular Neurodegeneration. 2023;18. doi: 10.1186/s13024-023-00640-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Skrobot OA, O’Brien J, Black S, Chen C, DeCarli C, Erkinjuntti T, Ford GA, Kalaria RN, Pantoni L, Pasquier F, et al. The Vascular Impairment of Cognition Classification Consensus Study. Alzheimers Dement. 2017;13:624–633. doi: 10.1016/j.jalz.2016.10.007 [DOI] [PubMed] [Google Scholar]
  • 4.Pantoni L Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9:689–701. doi: 10.1016/S1474-4422(10)70104-6 [DOI] [PubMed] [Google Scholar]
  • 5.Toledo JB, Arnold SE, Raible K, Brettschneider J, Xie SX, Grossman M, Monsell SE, Kukull WA, Trojanowski JQ. Contribution of cerebrovascular disease in autopsy confirmed neurodegenerative disease cases in the National Alzheimer’s Coordinating Centre. Brain. 2013;136:2697–2706. doi: 10.1093/brain/awt188 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Nelson PT, Pious NM, Jicha GA, Wilcock DM, Fardo DW, Estus S, Rebeck GW. APOE-epsilon2 and APOE-epsilon4 correlate with increased amyloid accumulation in cerebral vasculature. J Neuropathol Exp Neurol. 2013;72:708–715. doi: 10.1097/NEN.0b013e31829a25b9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Iturria-Medina Y, Sotero RC, Toussaint PJ, Mateos-Pérez JM, Evans AC, Weiner MW, Aisen P, Petersen R, Jack CR, Jagust W, et al. Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis. Nature Communications. 2016;7:11934. doi: 10.1038/ncomms11934 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Iadecola C, Gottesman RF. Cerebrovascular Alterations in Alzheimer Disease. Circ Res. 2018;123:406–408. doi: 10.1161/CIRCRESAHA.118.313400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Sweeney MD, Montagne A, Sagare AP, Nation DA, Schneider LS, Chui HC, Harrington MG, Pa J, Law M, Wang DJJ, et al. Vascular dysfunction-The disregarded partner of Alzheimer’s disease. Alzheimers Dement. 2019;15:158–167. doi: 10.1016/j.jalz.2018.07.222 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Carare RO, Kalaria R. Cerebrovascular pathology: the dark side of neurodegeneration. Acta Neuropathol. 2016;131:641–643. doi: 10.1007/s00401-016-1573-x [DOI] [PubMed] [Google Scholar]
  • 11.Iadecola C The pathobiology of vascular dementia. Neuron. 2013;80:844–866. doi: 10.1016/j.neuron.2013.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Montine TJ, Koroshetz WJ, Babcock D, Dickson DW, Galpern WR, Glymour MM, Greenberg SM, Hutton ML, Knopman DS, Kuzmichev AN, et al. Recommendations of the Alzheimer’s disease-related dementias conference. Neurology. 2014;83:851–860. doi: 10.1212/WNL.0000000000000733 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Snyder HM, Corriveau RA, Craft S, Faber JE, Greenberg SM, Knopman D, Lamb BT, Montine TJ, Nedergaard M, Schaffer CB, et al. Vascular contributions to cognitive impairment and dementia including Alzheimer’s disease. Alzheimers Dement. 2015;11:710–717. doi: 10.1016/j.jalz.2014.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Schneider JA, Wilson RS, Bienias JL, Evans DA, Bennett DA. Cerebral infarctions and the likelihood of dementia from Alzheimer disease pathology. Neurology. 2004;62:1148–1155. doi: 10.1212/01.wnl.0000118211.78503.f5 [DOI] [PubMed] [Google Scholar]
  • 15.Schneider JA, Arvanitakis Z, Leurgans SE, Bennett DA. The neuropathology of probable Alzheimer disease and mild cognitive impairment. Ann Neurol. 2009;66:200–208. doi: 10.1002/ana.21706 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Markesbery WR. Brain infarction and the clinical expression of Alzheimer disease. The Nun Study. JAMA. 1997;277:813–817. [PubMed] [Google Scholar]
  • 17.Wang H, Wang B, Zhang X, Feng C. Relations among sensitivity, specificity and predictive values of medical tests based on biomarkers. Gen Psychiatr. 2021;34:e100453. doi: 10.1136/gpsych-2020-100453 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Research CfDEa. Biomarker Qualification: Evidentiary Framework. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/biomarker-qualification-evidentiary-framework. 2018. Accessed November 10. [Google Scholar]
  • 19.Engelhart MJ, Geerlings MI, Meijer J, Kiliaan A, Ruitenberg A, van Swieten JC, Stijnen T, Hofman A, Witteman JC, Breteler MM. Inflammatory proteins in plasma and the risk of dementia: the rotterdam study. Arch Neurol. 2004;61:668–672. doi: 10.1001/archneur.61.5.668 [DOI] [PubMed] [Google Scholar]
  • 20.Wiseman SJ, Doubal FN, Chappell FM, Valdes-Hernandez MC, Wang X, Rumley A, Lowe GD, Dennis MS, Wardlaw JM. Plasma Biomarkers of Inflammation, Endothelial Function and Hemostasis in Cerebral Small Vessel Disease. Cerebrovasc Dis. 2015;40:157–164. doi: 10.1159/000438494 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Elkind MS, Luna JM, McClure LA, Zhang Y, Coffey CS, Roldan A, Del Brutto OH, Pretell EJ, Pettigrew LC, Meyer BC, et al. C-reactive protein as a prognostic marker after lacunar stroke: levels of inflammatory markers in the treatment of stroke study. Stroke. 2014;45:707–716. doi: 10.1161/STROKEAHA.113.004562 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Walker KA, Power MC, Hoogeveen RC, Folsom AR, Ballantyne CM, Knopman DS, Windham BG, Selvin E, Jack CR, Jr., Gottesman RF. Midlife Systemic Inflammation, Late-Life White Matter Integrity, and Cerebral Small Vessel Disease: The Atherosclerosis Risk in Communities Study. Stroke. 2017;48:3196–3202. doi: 10.1161/STROKEAHA.117.018675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Wang DD, Hou XH, Li HQ, Xu W, Ma YH, Dong Q, Yu JT, Tan L. Association of Serum Complement C1q Concentration with Severity of Neurological Impairment and Infarct size in Patients with Acute Ischemic Stroke. J Stroke Cerebrovasc Dis. 2020;29:105363. doi: 10.1016/j.jstrokecerebrovasdis.2020.105363 [DOI] [PubMed] [Google Scholar]
  • 24.Wang Z, Wu X, Yan T, Liu M, Yu W, Du Q, Hu W, Zheng Y, Zhang Z, Wang K, et al. Elevated Plasma Complement C1q Levels Contribute to a Poor Prognosis After Acute Primary Intracerebral Hemorrhage: A Prospective Cohort Study. Front Immunol. 2022;13:920754. doi: 10.3389/fimmu.2022.920754 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Mack WJ, Ducruet AF, Hickman ZL, Garrett MC, Albert EJ, Kellner CP, Mocco J, Connolly ES. Early plasma complement C3a levels correlate with functional outcome after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2007;61:255–260; discussion 260–251. doi: 10.1227/01.NEU.0000255518.96837.8E [DOI] [PubMed] [Google Scholar]
  • 26.Stokowska A, Olsson S, Holmegaard L, Jood K, Blomstrand C, Jern C, Pekna M. Plasma C3 and C3a levels in cryptogenic and large-vessel disease stroke: associations with outcome. Cerebrovasc Dis. 2011;32:114–122. doi: 10.1159/000328238 [DOI] [PubMed] [Google Scholar]
  • 27.Mocco J, Wilson DA, Komotar RJ, Sughrue ME, Coates K, Sacco RL, Elkind MSV, Connolly ES. ALTERATIONS IN PLASMA COMPLEMENT LEVELS FOLLOWING HUMAN ISCHEMIC STROKE. Neurosurgery. 2006;59:28–33. doi: 10.1227/01.neu.0000243280.75920.f4 [DOI] [PubMed] [Google Scholar]
  • 28.Chen Y, Yu H, Zhu J, Zhang H, Zhao Y, Dong Y, Cui Y, Gong G, Chai Q, Guo Y, et al. Low carotid endothelial shear stress associated with cerebral small vessel disease in an older population: A subgroup analysis of a population-based prospective cohort study. Atherosclerosis. 2019;288:42–50. doi: 10.1016/j.atherosclerosis.2019.07.006 [DOI] [PubMed] [Google Scholar]
  • 29.Sapira V, Cojocaru IM, Lilios G, Grigorian M, Cojocaru M. Study of endothelin-1 in acute ischemic stroke. Rom J Intern Med. 2010;48:329–332. [PubMed] [Google Scholar]
  • 30.Ziv I, Fleminger G, Djaldetti R, Achiron A, Melamed E, Sokolovsky M. Increased plasma endothelin-1 in acute ischemic stroke. Stroke. 1992;23:1014–1016. doi: 10.1161/01.str.23.7.1014 [DOI] [PubMed] [Google Scholar]
  • 31.Castellanos M, Castillo J, Garcia MM, Leira R, Serena J, Chamorro A, Davalos A. Inflammation-mediated damage in progressing lacunar infarctions: a potential therapeutic target. Stroke. 2002;33:982–987. doi: 10.1161/hs0402.105339 [DOI] [PubMed] [Google Scholar]
  • 32.Arba F, Giannini A, Piccardi B, Biagini S, Palumbo V, Giusti B, Nencini P, Maria Gori A, Nesi M, Pracucci G, et al. Small vessel disease and biomarkers of endothelial dysfunction after ischaemic stroke. European Stroke Journal. 2019;4:119–126. doi: 10.1177/2396987318805905 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Shoamanesh A, Preis SR, Beiser AS, Vasan RS, Benjamin EJ, Kase CS, Wolf PA, DeCarli C, Romero JR, Seshadri S. Inflammatory biomarkers, cerebral microbleeds, and small vessel disease: Framingham Heart Study. Neurology. 2015;84:825–832. doi: 10.1212/WNL.0000000000001279 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Staszewski J, Piusińska-Macoch R, Brodacki B, Skrobowska E, Stepien A. IL-6, PF-4, sCD40 L, and homocysteine are associated with the radiological progression of cerebral small-vessel disease: a 2-year follow-up study. Clinical Interventions in Aging. 2018;Volume 13:1135–1141. doi: 10.2147/cia.s166773 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Dukic L, Simundic AM, Martinic-Popovic I, Kackov S, Diamandis A, Begcevic I, Diamandis EP. The role of human kallikrein 6, clusterin and adiponectin as potential blood biomarkers of dementia. Clin Biochem. 2016;49:213–218. doi: 10.1016/j.clinbiochem.2015.10.014 [DOI] [PubMed] [Google Scholar]
  • 36.Ferrarese C, Mascarucci P, Zoia C, Cavarretta R, Frigo M, Begni B, Sarinella F, Frattola L, De Simoni MG. Increased cytokine release from peripheral blood cells after acute stroke. J Cereb Blood Flow Metab. 1999;19:1004–1009. doi: 10.1097/00004647-199909000-00008 [DOI] [PubMed] [Google Scholar]
  • 37.Demir R, Ulvi H, Ozel L, Ozdemir G, Guzelcik M, Aygul R. Relationship between plasma metalloproteinase-9 levels and volume and severity of infarct in patients with acute ischemic stroke. Acta Neurol Belg. 2012;112:351–356. doi: 10.1007/s13760-012-0067-4 [DOI] [PubMed] [Google Scholar]
  • 38.Hernandez-Guillamon M, Martinez-Saez E, Delgado P, Domingues-Montanari S, Boada C, Penalba A, Boada M, Pagola J, Maisterra O, Rodriguez-Luna D, et al. MMP-2/MMP-9 plasma level and brain expression in cerebral amyloid angiopathy-associated hemorrhagic stroke. Brain Pathol. 2012;22:133–141. doi: 10.1111/j.1750-3639.2011.00512.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hinman JD, Elahi F, Chong D, Radabaugh H, Ferguson A, Maillard P, Thompson JF, Rosenberg GA, Sagare A, Moghekar A, et al. Placental growth factor as a sensitive biomarker for vascular cognitive impairment. Alzheimers Dement. 2023;19:3519–3527. doi: 10.1002/alz.12974 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Dobrynina LA, Zabitova MR, Shabalina AA, Kremneva EI, Akhmetzyanov BM, Gadzhieva ZS, Berdalin AB, Kalashnikova LA, Gnedovskaya EV, Krotenkova MV. MRI Types of Cerebral Small Vessel Disease and Circulating Markers of Vascular Wall Damage. Diagnostics. 2020;10:354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Slevin M, Krupinski J, Slowik A, Kumar P, Szczudlik A, Gaffney J. Serial measurement of vascular endothelial growth factor and transforming growth factor-beta1 in serum of patients with acute ischemic stroke. Stroke. 2000;31:1863–1870. doi: 10.1161/01.str.31.8.1863 [DOI] [PubMed] [Google Scholar]
  • 42.Lee SC, Lee KY, Kim YJ, Kim SH, Koh SH, Lee YJ. Serum VEGF levels in acute ischaemic strokes are correlated with long-term prognosis. European Journal of Neurology. 2010;17:45–51. doi: 10.1111/j.1468-1331.2009.02731.x [DOI] [PubMed] [Google Scholar]
  • 43.Pikula A, Beiser AS, Chen TC, Preis SR, Vorgias D, DeCarli C, Au R, Kelly-Hayes M, Kase CS, Wolf PA, et al. Serum brain-derived neurotrophic factor and vascular endothelial growth factor levels are associated with risk of stroke and vascular brain injury: Framingham Study. Stroke. 2013;44:2768–2775. doi: 10.1161/STROKEAHA.113.001447 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Hunt BJ, Jurd KM. Endothelial cell activation. A central pathophysiological process. BMJ. 1998;316:1328–1329. doi: 10.1136/bmj.316.7141.1328 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Gross PL, Aird WC. The endothelium and thrombosis. Semin Thromb Hemost. 2000;26:463–478. doi: 10.1055/s-2000-13202 [DOI] [PubMed] [Google Scholar]
  • 46.Rouhl RP, Damoiseaux JG, Lodder J, Theunissen RO, Knottnerus IL, Staals J, Henskens LH, Kroon AA, de Leeuw PW, Tervaert JW, et al. Vascular inflammation in cerebral small vessel disease. Neurobiol Aging. 2012;33:1800–1806. doi: 10.1016/j.neurobiolaging.2011.04.008 [DOI] [PubMed] [Google Scholar]
  • 47.Yamashita K, Discher DJ, Hu J, Bishopric NH, Webster KA. Molecular regulation of the endothelin-1 gene by hypoxia. Contributions of hypoxia-inducible factor-1, activator protein-1, GATA-2, AND p300/CBP. J Biol Chem. 2001;276:12645–12653. doi: 10.1074/jbc.M011344200 [DOI] [PubMed] [Google Scholar]
  • 48.Kanazawa F, Nakanishi K, Osada H, Kanamaru Y, Ohrui N, Uenoyama M, Masaki Y, Kanatani Y, Hiroi S, Tominaga S, et al. Expression of endothelin-1 in the brain and lung of rats exposed to permanent hypobaric hypoxia. Brain Res. 2005;1036:145–154. doi: 10.1016/j.brainres.2004.12.019 [DOI] [PubMed] [Google Scholar]
  • 49.Zhang J, Narayan VM, Juedes N, Patel JM. Hypoxic upregulation of preproendothelin-1 gene expression is associated with protein tyrosine kinase-PI3K signaling in cultured lung vascular endothelial cells. Int J Clin Exp Med. 2009;2:87–94. [PMC free article] [PubMed] [Google Scholar]
  • 50.Jaime Garcia D, Chagnot A, Wardlaw JM, Montagne A. A Scoping Review on Biomarkers of Endothelial Dysfunction in Small Vessel Disease: Molecular Insights from Human Studies. International Journal of Molecular Sciences. 2023;24:13114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Petcu EB, Smith RA, Miroiu RI, Opris MM. Angiogenesis in old-aged subjects after ischemic stroke: a cautionary note for investigators. J Angiogenes Res. 2010;2:26. doi: 10.1186/2040-2384-2-26 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Szpak GM, Lechowicz W, Lewandowska E, Bertrand E, Wierzba-Bobrowicz T, Dymecki J. Border zone neovascularization in cerebral ischemic infarct. Folia Neuropathol. 1999;37:264–268. [PubMed] [Google Scholar]
  • 53.Holmes DI, Zachary I. The vascular endothelial growth factor (VEGF) family: angiogenic factors in health and disease. Genome Biol. 2005;6:209. doi: 10.1186/gb-2005-6-2-209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med. 2003;9:669–676. doi: 10.1038/nm0603-669 [DOI] [PubMed] [Google Scholar]
  • 55.Yamaguchi S, Iwata K, Shibuya M. Soluble Flt-1 (soluble VEGFR-1), a potent natural antiangiogenic molecule in mammals, is phylogenetically conserved in avians. Biochem Biophys Res Commun. 2002;291:554–559. doi: 10.1006/bbrc.2002.6478 [DOI] [PubMed] [Google Scholar]
  • 56.Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptor signalling - in control of vascular function. Nat Rev Mol Cell Biol. 2006;7:359–371. doi: 10.1038/nrm1911 [DOI] [PubMed] [Google Scholar]
  • 57.Shibuya M Vascular Endothelial Growth Factor (VEGF) and Its Receptor (VEGFR) Signaling in Angiogenesis: A Crucial Target for Anti- and Pro-Angiogenic Therapies. Genes Cancer. 2011;2:1097–1105. doi: 10.1177/1947601911423031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Uemura MT, Maki T, Ihara M, Lee VMY, Trojanowski JQ. Brain Microvascular Pericytes in Vascular Cognitive Impairment and Dementia. Front Aging Neurosci. 2020;12:80. doi: 10.3389/fnagi.2020.00080 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Smyth LCD, Rustenhoven J, Scotter EL, Schweder P, Faull RLM, Park TIH, Dragunow M. Markers for human brain pericytes and smooth muscle cells. J Chem Neuroanat. 2018;92:48–60. doi: 10.1016/j.jchemneu.2018.06.001 [DOI] [PubMed] [Google Scholar]
  • 60.Bondjers C, He L, Takemoto M, Norlin J, Asker N, Hellstrom M, Lindahl P, Betsholtz C. Microarray analysis of blood microvessels from PDGF-B and PDGF-Rbeta mutant mice identifies novel markers for brain pericytes. FASEB J. 2006;20:1703–1705. doi: 10.1096/fj.05-4944fje [DOI] [PubMed] [Google Scholar]
  • 61.Sagare AP, Sweeney MD, Makshanoff J, Zlokovic BV. Shedding of soluble platelet-derived growth factor receptor-beta from human brain pericytes. Neurosci Lett. 2015;607:97–101. doi: 10.1016/j.neulet.2015.09.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Payne LB, Abdelazim H, Hoque M, Barnes A, Mironovova Z, Willi CE, Darden J, Houk C, Sedovy MW, Johnstone SR, et al. A Soluble Platelet-Derived Growth Factor Receptor-β Originates via Pre-mRNA Splicing in the Healthy Brain and Is Upregulated during Hypoxia and Aging. Biomolecules. 2023;13:711. doi: 10.3390/biom13040711 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Kisler K, Nelson AR, Rege SV, Ramanathan A, Wang Y, Ahuja A, Lazic D, Tsai PS, Zhao Z, Zhou Y, et al. Pericyte degeneration leads to neurovascular uncoupling and limits oxygen supply to brain. Nat Neurosci. 2017;20:406–416. doi: 10.1038/nn.4489 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Hartmann DA, Berthiaume A-A, Grant RI, Harrill SA, Koski T, Tieu T, McDowell KP, Faino AV, Kelly AL, Shih AY. Brain capillary pericytes exert a substantial but slow influence on blood flow. Nature Neuroscience. 2021;24:633–645. doi: 10.1038/s41593-020-00793-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Thomsen MS, Routhe LJ, Moos T. The vascular basement membrane in the healthy and pathological brain. J Cereb Blood Flow Metab. 2017;37:3300–3317. doi: 10.1177/0271678X17722436 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Cabral-Pacheco GA, Garza-Veloz I, Castruita-De la Rosa C, Ramirez-Acuna JM, Perez-Romero BA, Guerrero-Rodriguez JF, Martinez-Avila N, Martinez-Fierro ML. The Roles of Matrix Metalloproteinases and Their Inhibitors in Human Diseases. Int J Mol Sci. 2020;21. doi: 10.3390/ijms21249739 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Tian Z, Ji X, Liu J. Neuroinflammation in Vascular Cognitive Impairment and Dementia: Current Evidence, Advances, and Prospects. Int J Mol Sci. 2022;23. doi: 10.3390/ijms23116224 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Wassmann S, Stumpf M, Strehlow K, Schmid A, Schieffer B, Bohm M, Nickenig G. Interleukin-6 induces oxidative stress and endothelial dysfunction by overexpression of the angiotensin II type 1 receptor. Circ Res. 2004;94:534–541. doi: 10.1161/01.RES.0000115557.25127.8D [DOI] [PubMed] [Google Scholar]
  • 69.Saija A, Princi P, Lanza M, Scalese M, Aramnejad E, De Sarro A. Systemic cytokine administration can affect blood-brain barrier permeability in the rat. Life Sci. 1995;56:775–784. doi: 10.1016/0024-3205(95)00008-t [DOI] [PubMed] [Google Scholar]
  • 70.Yoshizaki K, Adachi K, Kataoka S, Watanabe A, Tabira T, Takahashi K, Wakita H. Chronic cerebral hypoperfusion induced by right unilateral common carotid artery occlusion causes delayed white matter lesions and cognitive impairment in adult mice. Exp Neurol. 2008;210:585–591. doi: 10.1016/j.expneurol.2007.12.005 [DOI] [PubMed] [Google Scholar]
  • 71.Capogna E, Watne LO, Sorensen O, Guichelaar CJ, Idland AV, Halaas NB, Blennow K, Zetterberg H, Walhovd KB, Fjell AM, et al. Associations of neuroinflammatory IL-6 and IL-8 with brain atrophy, memory decline, and core AD biomarkers - in cognitively unimpaired older adults. Brain Behav Immun. 2023;113:56–65. doi: 10.1016/j.bbi.2023.06.027 [DOI] [PubMed] [Google Scholar]
  • 72.Badimon L, Pena E, Arderiu G, Padro T, Slevin M, Vilahur G, Chiva-Blanch G. C-Reactive Protein in Atherothrombosis and Angiogenesis. Front Immunol. 2018;9:430. doi: 10.3389/fimmu.2018.00430 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Sairanen T, Carpen O, Karjalainen-Lindsberg ML, Paetau A, Turpeinen U, Kaste M, Lindsberg PJ. Evolution of cerebral tumor necrosis factor-alpha production during human ischemic stroke. Stroke. 2001;32:1750–1758. doi: 10.1161/01.str.32.8.1750 [DOI] [PubMed] [Google Scholar]
  • 74.Turner RC, Millns H, Neil HA, Stratton IM, Manley SE, Matthews DR, Holman RR. Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom Prospective Diabetes Study (UKPDS: 23). BMJ. 1998;316:823–828. doi: 10.1136/bmj.316.7134.823 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Pan W, Kastin AJ. Tumor necrosis factor and stroke: role of the blood-brain barrier. Prog Neurobiol. 2007;83:363–374. doi: 10.1016/j.pneurobio.2007.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Zhang H, Park Y, Wu J, Chen X, Lee S, Yang J, Dellsperger KC, Zhang C. Role of TNF-alpha in vascular dysfunction. Clin Sci (Lond). 2009;116:219–230. doi: 10.1042/CS20080196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Hou X, Liang X, Chen JF, Zheng J. Ecto-5’-nucleotidase (CD73) is involved in chronic cerebral hypoperfusion-induced white matter lesions and cognitive impairment by regulating glial cell activation and pro-inflammatory cytokines. Neuroscience. 2015;297:118–126. doi: 10.1016/j.neuroscience.2015.03.033 [DOI] [PubMed] [Google Scholar]
  • 78.Mun KT, Hinman JD. Inflammation and the Link to Vascular Brain Health: Timing Is Brain. Stroke. 2022;53:427–436. doi: 10.1161/STROKEAHA.121.032613 [DOI] [PubMed] [Google Scholar]
  • 79.Clarke AR, Christophe BR, Khahera A, Sim JL, Connolly ES, Jr. Therapeutic Modulation of the Complement Cascade in Stroke. Front Immunol. 2019;10:1723. doi: 10.3389/fimmu.2019.01723 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Huang J, Kim LJ, Mealey R, Marsh HC, Jr., Zhang Y, Tenner AJ, Connolly ES, Jr., Pinsky DJ. Neuronal protection in stroke by an sLex-glycosylated complement inhibitory protein. Science. 1999;285:595–599. doi: 10.1126/science.285.5427.595 [DOI] [PubMed] [Google Scholar]
  • 81.Alawieh A, Elvington A, Tomlinson S. Complement in the Homeostatic and Ischemic Brain. Front Immunol. 2015;6:417. doi: 10.3389/fimmu.2015.00417 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Szeplaki G, Szegedi R, Hirschberg K, Gombos T, Varga L, Karadi I, Entz L, Szeplaki Z, Garred P, Prohaszka Z, et al. Strong complement activation after acute ischemic stroke is associated with unfavorable outcomes. Atherosclerosis. 2009;204:315–320. doi: 10.1016/j.atherosclerosis.2008.07.044 [DOI] [PubMed] [Google Scholar]
  • 83.Fagerberg L, Hallstrom BM, Oksvold P, Kampf C, Djureinovic D, Odeberg J, Habuka M, Tahmasebpoor S, Danielsson A, Edlund K, et al. Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol Cell Proteomics. 2014;13:397–406. doi: 10.1074/mcp.M113.035600 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84.Chau K, Hennessy A, Makris A. Placental growth factor and pre-eclampsia. J Hum Hypertens. 2017;31:782–786. doi: 10.1038/jhh.2017.61 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 85.Agrawal S, Shinar S, Cerdeira AS, Redman C, Vatish M. Predictive Performance of PlGF (Placental Growth Factor) for Screening Preeclampsia in Asymptomatic Women: A Systematic Review and Meta-Analysis. Hypertension. 2019;74:1124–1135. doi: 10.1161/HYPERTENSIONAHA.119.13360 [DOI] [PubMed] [Google Scholar]
  • 86.Wilcock D, Jicha G, Blacker D, Albert MS, D’Orazio LM, Elahi FM, Fornage M, Hinman JD, Knoefel J, Kramer J, et al. MarkVCID cerebral small vessel consortium: I. Enrollment, clinical, fluid protocols. Alzheimers Dement. 2021;17:704–715. doi: 10.1002/alz.12215 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87.Lu H, Kashani AH, Arfanakis K, Caprihan A, DeCarli C, Gold BT, Li Y, Maillard P, Satizabal CL, Stables L, et al. MarkVCID cerebral small vessel consortium: II. Neuroimaging protocols. Alzheimers Dement. 2021;17:716–725. doi: 10.1002/alz.12216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 88.Gertje EC, Janelidze S, van Westen D, Cullen N, Stomrud E, Palmqvist S, Hansson O, Mattsson-Carlgren N. Associations Between CSF Markers of Inflammation, White Matter Lesions, and Cognitive Decline in Individuals Without Dementia. Neurology. 2023;100:e1812–e1824. doi: 10.1212/WNL.0000000000207113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 89.Plate KH, Beck H, Danner S, Allegrini PR, Wiessner C. Cell type specific upregulation of vascular endothelial growth factor in an MCA-occlusion model of cerebral infarct. J Neuropathol Exp Neurol. 1999;58:654–666. doi: 10.1097/00005072-199906000-00010 [DOI] [PubMed] [Google Scholar]
  • 90.Tang H, Mao X, Xie L, Greenberg DA, Jin K. Expression level of vascular endothelial growth factor in hippocampus is associated with cognitive impairment in patients with Alzheimer’s disease. Neurobiology of Aging. 2013;34:1412–1415. doi: 10.1016/j.neurobiolaging.2012.10.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 91.Provias J, Jeynes B. Reduction in vascular endothelial growth factor expression in the superior temporal, hippocampal, and brainstem regions in Alzheimer’s disease. Curr Neurovasc Res. 2014;11:202–209. doi: 10.2174/1567202611666140520122316 [DOI] [PubMed] [Google Scholar]
  • 92.Mateo I, Llorca J, Infante J, Rodriguez-Rodriguez E, Fernandez-Viadero C, Pena N, Berciano J, Combarros O. Low serum VEGF levels are associated with Alzheimer’s disease. Acta Neurol Scand. 2007;116:56–58. doi: 10.1111/j.1600-0404.2006.00775.x [DOI] [PubMed] [Google Scholar]
  • 93.Dobrynina LA, Zabitova MR, Shabalina AA, Kremneva EI, Akhmetzyanov BM, Gadzhieva ZS, Berdalin AB, Kalashnikova LA, Gnedovskaya EV, Krotenkova MV. MRI Types of Cerebral Small Vessel Disease and Circulating Markers of Vascular Wall Damage. Diagnostics (Basel). 2020;10. doi: 10.3390/diagnostics10060354 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 94.Kaess BM, Preis SR, Beiser A, Sawyer DB, Chen TC, Seshadri S, Vasan RS. Circulating vascular endothelial growth factor and the risk of cardiovascular events. Heart. 2016;102:1898–1901. doi: 10.1136/heartjnl-2015-309155 [DOI] [PubMed] [Google Scholar]
  • 95.Linton AE, Weekman EM, Wilcock DM. Pathologic sequelae of vascular cognitive impairment and dementia sheds light on potential targets for intervention. Cerebral Circulation - Cognition and Behavior. 2021;2:100030. doi: 10.1016/j.cccb.2021.100030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 96.Weekman EM, Wilcock DM. Matrix Metalloproteinase in Blood-Brain Barrier Breakdown in Dementia. J Alzheimers Dis. 2016;49:893–903. doi: 10.3233/JAD-150759 [DOI] [PubMed] [Google Scholar]
  • 97.Achilli C, Ciana A, Minetti G. Amyloid-beta (25–35) peptide induces the release of pro-matrix metalloprotease 9 (pro-MMP-9) from human neutrophils. Mol Cell Biochem. 2014;397:117–123. doi: 10.1007/s11010-014-2178-0 [DOI] [PubMed] [Google Scholar]
  • 98.Ridnour LA, Dhanapal S, Hoos M, Wilson J, Lee J, Cheng RY, Brueggemann EE, Hines HB, Wilcock DM, Vitek MP, et al. Nitric oxide-mediated regulation of beta-amyloid clearance via alterations of MMP-9/TIMP-1. J Neurochem. 2012;123:736–749. doi: 10.1111/jnc.12028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 99.Rosenberg GA. Extracellular matrix inflammation in vascular cognitive impairment and dementia. Clin Sci (Lond). 2017;131:425–437. doi: 10.1042/CS20160604 [DOI] [PubMed] [Google Scholar]
  • 100.Castellanos M, Leira R, Serena J, Pumar JM, Lizasoain I, Castillo J, Davalos A. Plasma metalloproteinase-9 concentration predicts hemorrhagic transformation in acute ischemic stroke. Stroke. 2003;34:40–46. [PubMed] [Google Scholar]
  • 101.Zhao L, Arbel-Ornath M, Wang X, Betensky RA, Greenberg SM, Frosch MP, Bacskai BJ. Matrix metalloproteinase 9-mediated intracerebral hemorrhage induced by cerebral amyloid angiopathy. Neurobiol Aging. 2015;36:2963–2971. doi: 10.1016/j.neurobiolaging.2015.07.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 102.Chen W, Hartman R, Ayer R, Marcantonio S, Kamper J, Tang J, Zhang JH. Matrix metalloproteinases inhibition provides neuroprotection against hypoxia-ischemia in the developing brain. J Neurochem. 2009;111:726–736. doi: 10.1111/j.1471-4159.2009.06362.x [DOI] [PubMed] [Google Scholar]
  • 103.Ringland C, Schweig JE, Eisenbaum M, Paris D, Ait-Ghezala G, Mullan M, Crawford F, Abdullah L, Bachmeier C. MMP9 modulation improves specific neurobehavioral deficits in a mouse model of Alzheimer’s disease. BMC Neurosci. 2021;22:39. doi: 10.1186/s12868-021-00643-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 104.Tang J, Liu J, Zhou C, Alexander JS, Nanda A, Granger DN, Zhang JH. Mmp-9 deficiency enhances collagenase-induced intracerebral hemorrhage and brain injury in mutant mice. J Cereb Blood Flow Metab. 2004;24:1133–1145. doi: 10.1097/01.WCB.0000135593.05952.DE [DOI] [PubMed] [Google Scholar]
  • 105.Liu J, Jin X, Liu KJ, Liu W. Matrix metalloproteinase-2-mediated occludin degradation and caveolin-1-mediated claudin-5 redistribution contribute to blood-brain barrier damage in early ischemic stroke stage. J Neurosci. 2012;32:3044–3057. doi: 10.1523/JNEUROSCI.6409-11.2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 106.Ekenze O, Pinheiro A, Demissie S, Aparicio HJ, Charidimou A, Beiser AS, Satizabal CL, Kautz T, DeCarli C, Greenberg S, et al. Association of MRI Visible Perivascular Spaces and Neurofilament Light Chain: The Framingham Heart Study. J Alzheimers Dis. 2023;95:1133–1145. doi: 10.3233/JAD-221260 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 107.Rost NS, Meschia JF, Gottesman R, Wruck L, Helmer K, Greenberg SM, Investigators D. Cognitive Impairment and Dementia After Stroke: Design and Rationale for the DISCOVERY Study. Stroke. 2021;52:e499–e516. doi: 10.1161/STROKEAHA.120.031611 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 108.Weinstein G, Wolf PA, Beiser AS, Au R, Seshadri S. Risk estimations, risk factors, and genetic variants associated with Alzheimer’s disease in selected publications from the Framingham Heart Study. J Alzheimers Dis. 2013;33 Suppl 1:S439–445. doi: 10.3233/JAD-2012-129040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 109.Orlando A, Sharrett AR, Schneider ALC, Gottesman RF, Knopman DS, Rawlings A, Mosley TH, Jack CR, Wong D, Pike JR, et al. Brain Imaging Features Associated with 20-Year Cognitive Decline in a Community-Based Multiethnic Cohort without Dementia. Neuroepidemiology. 2022;56:183–191. doi: 10.1159/000524731 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 110.Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, Greenland P, Jacob DR, Kronmal R, Liu K, et al. Multi-Ethnic Study of Atherosclerosis: objectives and design. Am J Epidemiol. 2002;156:871–881. doi: 10.1093/aje/kwf113 [DOI] [PubMed] [Google Scholar]

RESOURCES