Stiffening of the carotid artery or other elastic arteries is the gradual loss of elastin fibers and accumulation of stiffer collagen fibers in the media over time (4, 5). This process, which can occur independent of the development of atherosclerosis, leads to loss of the ability of vasculature to appropriately accommodate to changes in blood pressure variation (6). This loss of responsive distensibility leads to higher pulsatile pressures and eventually increased flow load experienced by cerebral microvasculature and ultimately the brain parenchyma (4).

Arterial stiffness is not routinely measured in clinical practice but has been well-studied in several epidemiologic cohort studies. It is most commonly measured via indirect methods by measurement of pulse wave velocity (PWV) (7). Pulse wave velocity is an estimation of central arterial stiffness via measurement of pressure waves in two different vascular beds, commonly the carotid and femoral arterial beds (8). This indirect measure is a surrogate for aortic stiffness and has been widely used in multiple cohort studies. In addition to these indirect measures of central arterial stiffness, there are additional methods to directly measure vascular stiffness, specifically in the carotid artery (7, 9). These are most commonly performed via ultrasound measurement techniques.

Since there are many shared risk factors for the development of arterial stiffness and other common cardiovascular diseases, it can be difficult to determine the specific effects of arterial stiffness on the downstream cerebral parenchyma. One of the major vascular risk factors contributing to increased arterial stiffness is increasing age. In addition to age, hypertension, diabetes, and smoking are additional factors that seem to accelerate the development of arterial stiffness (10). A major proposed mechanism by which arterial stiffness contributes to cognitive dysfunction is through the increased flow load experienced by the cerebral parenchyma leading to end-organ damage (11, 12). The resultant damage may manifest as cerebral small vessel disease evident on brain imaging, including white matter hyperintensities, covert brain infarctions, or cerebral microbleeds. These findings of cerebral small vessel disease are also independently associated with cognitive impairment and dementia, in addition to stroke and overall mortality (13–16).

Arterial stiffness has been associated with several imaging findings which are in turn also associated with cognitive dysfunction and dementia (Table 1). Specifically, there have been several studies showing that decreased carotid compliance (or increased carotid stiffness), is associated with decreased total brain and cortical gray matter volumes and also decreased volume in the hippocampal and parahippocampal regions (11, 17, 18). In the Age, Gene/Environment Susceptibility (AGES)-Reykjavik Study in 422 participants free from cerebrovascular disease and dementia, the authors found that carotid stiffness was associated with lower whole brain (-0.127 ± 0.037 SD/SD, p < 0.001), gray matter (−0.079 ± 0.038 SD/SD, p = 0.038), and white matter volumes (−0.128 ± 0.039 SD/SD, p = 0.028) along with lower memory scores (11). In the SMART-MR study of 526 participants, the authors found a cross-sectional association with increased carotid stiffness and lower total brain and cortical gray matter volume (B = −0.24%, 95% confidence interval [CI] −0.44 to −0.04%, and B = −0.47%, 95% CI −0.75 to −0.19%) but this association was no longer significant when evaluated prospectively after a mean of 4 years (17). When evaluating 614 participants in the Atherosclerosis Risk in the Community Study (ARIC), authors found a significant association between decreased carotid stiffness and lower parahippocampal and hippocampal volumes 20 years later [R = 0.218(0.144–0.291), p < 0.001 and R = 0.181 (0.105–0.255), P < 0.001], even after adjusting for confounders. Decreased total brain volumes and decreased volumes in the parahippocampal and hippocampal regions are independently associated with cognitive dysfunction and dementia (19).

In addition to decreased cerebral parenchymal volumes, there is evidence of the association between carotid stiffness to imaging findings of cerebral small vessel disease, including white matter hyperintensities, covert brain infarctions, and cerebral microbleeds, including in the AGES-Reykjavik, SMART-MR, and ARIC epidemiologic cohort studies (20–22). In the SMART-MR study, there was a significant association between carotid stiffness and larger volume of white matter hyperintensities (B = 0.09%, 95% CI −0.01 to 0.19%) as well as cortical and subcortical brain infarcts (RR = 1.44, 95% CI 1.14–1.81). In the ARIC study, this association was found to be significant 20 years after the carotid stiffness measures (20). These imaging markers are also independently associated with mortality, stroke, cognitive dysfunction, and dementia (14, 16, 23).

Though there is fairly robust evidence of an association between carotid stiffness to brain imaging markers that are associated with dementia, the association of stiffness to cognitive impairment and dementia are more clinically relevant. There are several strong studies demonstrating that central arterial stiffness is an independent predictor of cognitive dysfunction and dementia (24–28). These studies, many of which were performed in epidemiologic cohorts including the Framingham Heart Study, the Maastricht study, and the PARTAGE study, found strong associations between stiffness and cognitive impairment. Interestingly, in the Maastricht study, the authors found a strong association with central stiffness (−0.018 SD [95% CI, −0.036 to −0.000]), but not with carotid stiffness (24). In addition, there are multiple studies establishing a cross-sectional association of central arterial stiffness to cognitive decline (29–32). Though there is relatively strong data supporting a link between central arterial stiffening to cognitive dysfunction, there is less supporting evidence between carotid stiffness and dementia with some studies supporting this association (Table 2) (33–38), and other studies showing no association, after adjusting for other confounders (24, 67, 68). This is certainly an area worthy of future investigation to further our understanding of the role of carotid stiffness, an asymptomatic marker of vascular aging, to the development of dementia. Future longitudinal prospective cohort studies may be able to aid in clarifying this potential association.

Another marker of asymptomatic carotid atherosclerosis is thickening of the carotid intima-media. This subtle thickening of the arterial wall measured in either the distal common carotid artery or proximal internal carotid artery using ultrasound is a marker of subclinical atherosclerosis. Similar to arterial stiffness, both hypertension and smoking, along with other vascular risk factors, are known associations to CIMT. Since carotid intima-media thickening (CIMT) is a precursor for the development of atherosclerotic plaque, it is often considered an imaging marker of generalized cardiovascular risk. Though this thickening is asymptomatic, it is associated with cognitive dysfunction and dementia, along with stroke and overall increased mortality risk (47, 60, 69, 70).

There is evidence of an association of baseline increased CIMT to brain imaging markers of cerebral small vessel disease including white matter hyperintensities, covert brain infarctions, cerebral microbleeds, and also lower total brain volumes (71–76) (Figure 2). Specifically, a systematic review and meta-analysis of nine studies found that CIMT was associated with white matter hyperintensities [odds ratio (OR) 1.42, 95% CI 1.22–1.66, p < 0.0001], covert brain infarctions (OR 1.89, CI 1.46–2.45, p < 0.0001) (74). As for the association of CIMT with brain volume, there is evidence of lower total brain volume (−0.05 per SD, P < 0.05) in the Framingham Heart Study (77) and lower total brain and cortical gray matter (−0.29 per SD) in the SMART-MR study cohort (72). In addition, there is evidence that the progression of CIMT over time is also associated with lower hippocampal volumes in the Framingham Heart Study (78). As discussed previously, these imaging markers of cerebral small vessel disease and brain aging are independently associated with cognitive dysfunction and dementia.

Several studies have evaluated the association between CIMT on ultrasound and the future development of dementia and cross-sectional association with cognitive function. Several studies from epidemiologic cohorts have found a positive association between increased CIMT at baseline and future development of cognitive decline (44–46). These studies followed cohorts of participants over time and found that having higher baseline CIMT on ultrasound was correlated with poorer future performance on cognitive testing and increased rates of dementia diagnoses. Hazard ratio for development of cognitive impairment based on elevated CIMT was 1.251 (95% CI 1.006–1.555, p = 0.044) in the Korean Longitudinal Study on Health and Aging after adjusting for basic demographics and baseline cognition (44). Further, in the Northern Manhattan Study, a cohort study of stroke-free participants, found that those with greater CIMT had worse performance on cognitive testing (ß = −0.60, p = 0.04 for episodic memory) (46). There are also many studies demonstrating a cross-sectional association between increased CIMT and poorer cognitive function (39–43).

The exact pathophysiologic mechanism underpinning this association is unclear at this time, though it may be similar to the proposed mechanisms for arterial stiffness. Since CIMT is thought to be marker for systemic atherosclerosis, it may not necessarily be a direct contributor to cognitive dysfunction, but rather reflect general cardiovascular risk. Further evaluation of this association is necessary in order to identify potentially modifiable vascular contributions to cognitive dysfunction.

Flow-limiting carotid stenosis is a well-documented risk factor for stroke and is an established indication for carotid revascularization. Severe carotid stenosis is traditionally defined as 70–99% narrowing of the vessel lumen by various measuring methods, most commonly the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method in the United States (79). Moderate stenosis is defined as 50–70% stenosis. The degree of stenosis is a leading risk factor in the development of ischemic stroke with annual rates of ischemic stroke estimated to be around 1% (80). Though not nearly as well-studied, some have found that the annual rate of dementia in the presence of severe carotid stenosis to be around 1 (55).

The main mechanism by which carotid stenosis is hypothesized to contribute to cognitive impairment is via hypoperfusion (81). By limiting the flow to the brain parenchyma, some hypothesize that this may lead to end-organ damage, parenchymal atrophy, and neurodegneration. Cerebral hypoperfusion is thought to accelerate amyloid and tau deposition, which is a potential link between flow-limitation and cognitive dysfunction (82, 83). Other potential mechanistic explanations include via covert brain infarctions from embolization, which may act as an intermediate step between stenosis and cognitive dysfunction. Similar to many other features of carotid atherosclerosis, it is difficult to disentangle the complex associations between carotid stenosis, other vascular risk factors, and other findings of cerebral parenchymal damage. One potential method for clarifying these associations is by evaluating individuals with unilateral stenosis to determine if there are varying degrees of parenchymal damage downstream from the affected side. Currently, the exact mechanism by which flow-limiting stenosis contributes to cognitive dysfunction is not clearly established.

Several studies have shown an association with carotid stenosis and poorer performance on cognitive testing (47–52), while others have failed to find an association. There are several studies finding that asymptomatic individuals with severe carotid stenosis with evidence of altered perfusion are more likely to develop cognitive impairment (51, 84, 85). In one of the largest studies evaluating this association in over 4,000 participants, the authors found that high-grade carotid artery stenosis was associated with cognitive impairment (OR 6.7, 95% CI 2.4 to 18.1) and cognitive decline (OR 2.6, CI, 1.1 to 6.3) (48). Some have found that left sided carotid stenosis is more likely to result in cognitive dysfunction than right-sided carotid stenosis, indicating that carotid stenosis may be an independent risk factor for cognitive impairment (48, 53). A systematic review and meta-analysis including over 760 subjects with asymptomatic carotid stenosis found an association between the presence of carotid stenosis and cognitive impairment (54).

Several studies have attempted to isolate the effect of carotid stenosis by assessing for changes in cognitive function after carotid endarterectomy with conflicting results (56). Overall, the majority of studies have found an improvement in cognition after CEA, but there are many studies showing either no change, and even a deterioration in cognitive function. In a large study of patients with asymptomatic carotid stenosis, there was no difference in mini-mental status examination scores in those who received medical therapy compared to those who underwent CEA (57). Another study based on patients from a randomized controlled trial with severe carotid stenosis without history of stroke or known dementia (ACST-1) found that carotid endarterectomy had no significant effect on the incidence of dementia (55). There is less data regarding cognitive changes after CAS, however, there are similarly mixed results with some studies showing an improvement in cognition after undergoing CAS (56, 58, 59).

There is strong evidence that the presence of flow-limiting stenosis results in both cerebral atrophy as well as other markers of subclinical vascular injury, such as white matter hypterintensities and CBIs. There is evidence that severe carotid stenosis is associated with progression of brain atrophy, while the same link was not as pronounced in those with moderate stenosis (72). Further, there is strong evidence that carotid stenosis is associated with markers of parenchymal damage, including white matter hyperintensities and microstructural damage to both gray and white matter (86–88). In addition, there is strong evidence of an association between carotid stenosis and downstream CBI with a systematic review and meta-analysis of 11 studies reporting an OR of 2.78 (95% CI, 2.19 to 3.52, p < 0.0001) (89). Further, studies demonstrate that there are asymmetries in prevalence of CBI in cerebral hemispheres downstream from severe carotid stenosis, specifically with more cortical CBIs downstream from stenosis (90).

This role of flow-limiting stenosis in the development of dementia is an area worthy of further investigation. Though CEA and CAS procedures are primarily performed for stroke risk reduction, the potential added bonus of improved cognition may alter the risk calculus when identifying patients for carotid revascularization.

Carotid plaque is a specific marker of advanced atherosclerosis usually found in the carotid bifurcation. There are many different features of the plaque itself that carry varying degrees of associated embolic risk. There has been increased attention on individual plaque components and plaque volume in addition to the degree of stenosis when assessing stroke risk from carotid atherosclerotic disease. There is strong evidence that certain plaque components are more strongly associated with future and recurrent stroke, including intraplaque hemorrhage, lipid-rich necrotic core, and plaque ulceration (91). Many of these specific plaque elements are often more strongly associated with stroke than degree of stenosis which has led to a paradigm shift in stroke risk assessment (92). Though the major concern for carotid plaques is their leading to symptomatic strokes and/or transient ischemic attacks, many of these high-risk plaque features are also seen in asymptomatic individuals.

Though there is a strong association with high-risk plaque elements to stroke, there are fewer studies evaluating the association of plaque with the development of cognitive dysfunction and dementia. The current literature has shown mixed results with respect to the association of vulnerable plaque components and cognitive dysfunction.

Some studies have evaluated the association of increased plaque volume to performance on cognitive examinations and have found that there is worse performance on cognitive testing with higher plaque volume, even when accounting for education level and other confounding (60–66). For example, a study evaluating high risk plaque features on ultrasound found that those with more plaques were more likely to have poor performance on cognitive testing, including mini-mental status examinations (OR 1.72, 1.00–2.96) (61). Most of the studies evaluating this association used either ultrasound or CT imaging techniques to evaluate plaque. Other studies have shown no significant difference in cognitive function when accounting for other cardiovascular risk factors (93).

There is evidence that vulnerable plaque features are associated with other markers of neurodegeneration and subclinical vascular injury. For instance, there is evidence that vulnerable plaque features contribute to cortical micro-infarcts detected on MR which are in turn associated with poor cognitive function (94). There are other studies which demonstrate an association with white matter hyperintensities and CBIs as well (74), though there are few studies explicitly looking at specific plaque features (Figure 3).

The exact mechanism behind this potential association is unclear at this time. Whether high risk plaque is directly associated with dementia due to repeated microembolic phenomena is not well established. Though there is evidence that high-risk plaque lead to increased risk of stroke, it is unclear if these plaque features may also contribute to microembolic phenomenon contributing to cognitive dysfunction. Other studies have found that there are increased markers of subclinical vascular injury which are associated with high-risk plaque features, including covert brain infarctions, cerebral microbleeds, and white matter hyperintensities which are independently associated with cognitive dysfunction. At this time, this exact association is unclear and further studies would be helpful in elucidating the exact contribution of high-risk plaque features to cognitive dysfunction and dementia.