The Neurovascular Unit in Dementia: An Opinion on Current Research and Future Directions

Dementia is a burgeoning public health crisis, with 50 million people currently affected worldwide (Prince et al., 2015). As the population ages, this figure is set to rise dramatically by 40% over the next 12 years (Prince et al., 2015). Dementia is an umbrella term for several disorders which result in the progressive loss of memory or other cognitive functions (Scott and Barrett, 2007). It remains an incurable disease, and current therapeutics have limited efficacy at slowing disease progression for one third of patients (Rockwood et al., 2008). Of the dementia sub-types, Alzheimer's disease (AD) remains the most prevalent, accounting for ~60–70% cases (Alzheimer's-Society, 2016). Vascular dementia (VaD) is the second most common form and is responsible for ~20% of cases, with a further 10% being a combination of these two diseases (Alzheimer's-Society, 2016). However, in practise these distinctions are somewhat arbitrary given the significant overlap in altered vascular structure and function in both of these major sub-types (Kalaria and Ballard, 1999). At least 30% of patients with AD have evidence of cerebrovascular disease on post-mortem examination, and almost all have evidence of cerebral amyloid angiopathy, microvascular degeneration, and white matter lesions (Kalaria and Ballard, 1999). Similarly, one-third of patients with VaD exhibit pathology consistent with AD (e.g., hippocampal or temporal lobe atrophy) (Kalaria and Ballard, 1999). Longitudinal studies have demonstrated that vascular risk factors (e.g., hypertension), significantly increase the risk of both AD and VaD (Rius-Pérez et al., 2018). In genetically at-risk individuals positive for apolipoprotein E4 (APOE4), atherosclerosis can increase the risk of AD by three-fold (Hoffmann et al., 2010). This article provides an opinion on the current evidence on the role of the neurovascular unit in dementia, for further information, several recent reviews are available on this topic (Nelson et al., 2016; Kisler et al., 2017)