Dementia is a major health issue with considerable health, social and economic costs for individuals, their families, and societies. It presents significant challenges to healthcare systems, affecting approximately 57 million people worldwide, a number which is predicted to rise to 153 million by 2050. As no cure exists, the identification of risk factors and the implementation of prevention strategies are a crucial step towards decreasing dementia incidence (Nichols et al., 2022). Cardiovascular health is believed to play an important role in this pathogenesis, and arterial stiffening has been identified as a key risk factor that increases dementia risk (Scuteri, 2007).

The deleterious effects of elevated arterial stiffness (AS) on cerebral health are well-established. Increased arterial stiffening has been associated with lower total brain volumes and a higher incidence of cerebral artefacts and white matter lesions in healthy older adults and in patient cohorts in both cross-sectional and longitudinal studies (Henskens et al., 2008; Matsumoto et al., 2007; Ohmine et al., 2008). The literature also supports a link between elevated arterial stiffness and cognitive impairment with progression to dementia (Henskens et al., 2008; Pase et al., 2016; Rouch et al., 2018; Benetos et al., 2012; Scuteri et al., 2007; Hanon et al., 2005). Negative associations between AS and total hippocampal volumes (a hallmark of AD) have been observed in many cross-sectional and longitudinal studies (Baradaran et al., 2021; Bown et al., 2021).

It has been proposed that increased arterial stiffness leads to cerebral hypoperfusion (Ajamu et al., 2021; Suri et al., 2020; Sadekova et al., 2018; Jefferson et al., 2018; Muhire et al., 2019; Tarumi et al., 2011). This may, in turn, lead to neuronal death with subsequent cerebral atrophy, as is seen selectively in the hippocampus in memory decline (Jack et al., 2000; Jack et al., 1998; Jobst et al., 1994). Several studies have reported an association between lower cerebral blood flow with smaller hippocampal volumes and progression to cognitive impairment (Knoops et al., 2009; Mak et al., 2014; Chao et al., 2010). The link between increasing arterial stiffness and reduced cerebral blood flow is also well established (Ajamu et al., 2021; Suri et al., 2020; Sadekova et al., 2018; Jefferson et al., 2018; Muhire et al., 2019; Tarumi et al., 2011). However, the causal relationships and interactions between these pathways, as well as associated compensatory mechanisms are less well understood. For example, to date, no studies have together described the independent and moderating effects of increased arterial stiffness and reduced CBF on the hippocampal volume in a healthy older population, but it could be assumed that raised arterial stiffness either alone or in combination with a reduction in cerebral blood flow may lead to selective atrophy of the hippocampus through this proposed mechanism of hypoperfusion and neuronal death. Additionally, no studies have addressed the issue of specific hippocampal subfields, and some subfields appear to be more vulnerable to hypoxia and ischemia than others, like the Cornu Ammonis (CA) 1, CA4, and subiculum subfields (Schmidt-Kastner and Freund, 1991; Bonnekoh et al., 1990; Kim et al., 2015). For example, a recent study by Pin et al. (2021), examined the rate of volume loss in the hippocampal subfields CA1-CA3, CA4 and the subiculum over four-years in 249 non-demented older adults and examined their interaction with white matter hyperintensities, believed to be a surrogate marker for vascular damage. They identified the subiculum to be particularly vulnerable to ischemia with a significant association between WMH’s and subiculum volume loss, independent of vascular risk factors (p = 0.03) (Pin et al., 2021).

In this study, we investigate the independent and moderating effects of AS and CBF on total hippocampal volume in a large MRI sample of community-dwelling older adults from a nationally representative population-based study, the Irish Longitudinal Study on Ageing (TILDA). In supplementary analyses, we also examine if these effects are specific to certain hippocampal subfields which are known to be vulnerable to AS and CBF alterations: the CA1, CA3, CA4, subiculum, and dentate gyrus (DG). Specifically, we test the hypotheses that (i) higher AS is associated with lower total hippocampal volumes and that (ii) reduced CBF may moderate this relationship. We also test the hypotheses that (iii) persistently higher AS over a four-year follow-up period and (iv) the co-occurrence of higher AS and low CBF may have more significant effects on hippocampal volumes. Finally we hypothesize that these effects may be exerted maximally in the CA-1, CA-4 and subiculum subfields.

We examined the independent and moderating effects of elevated arterial stiffness and reduced cerebral blood flow on hippocampal volumes and subfield volumes, in a relatively large MRI sample of community-dwelling older adults from ‘The Irish Longitudinal Study on Ageing’.

Our cross-sectional analyses showed that there were no independent associations between raised AS or reduced CBF with total hippocampal volumes. Longitudinal analyses of arterial stiffness suggested that higher AS in the preceding 4 years was independently associated with lower total hippocampal volumes when adjusted for AS at wave 3 in an additive model, however this effect was small. Significant associations between carotid compliance (a surrogate marker for arterial stiffness) and hippocampal volume have been observed in a study of 614 subjects at 20-year follow-up MRI as part of the Atherosclerosis Risk in the Community (ARIC) Study (Baradaran et al., 2021). Carotid compliance was calculated by continuous measurement of the variation in carotid artery diameter during the cardiac cycle, using ultrasound. It is defined as the ‘absolute volume increase within the carotid artery segment during the cardiac cycle, divided by the arterial pulse pressure’. Therefore, the lower the carotid compliance value, the greater the degree of arterial stiffness. Whilst cf-PWV has been consistently cited as the ‘gold standard’ measure of arterial stiffness, conversion tools have been utilised to convert carotid compliance values to local-PWV values, which show significant correlation to the patients cf-PWV measurements (Alhalimi et al., 2021).

Similarly, Bown’s longitudinal study of 278 adults over the age of 60 who had serial MRI brain studies at baseline, 18 months, 3, 5 and 7 years as part of the Vanderbilt Memory and Ageing Project, showed that an elevated baseline pulse wave velocity was associated with a greater decrease in hippocampal volume over time (Bown et al., 2021). As per the SMART-MR Study, the association between cerebral blood flow measurements and hippocampal volumes were lost after adjustment for age and sex (Knoops et al., 2009).

Our analyses examining the combined effects of raised AS and reduced CBF on total hippocampal volume (HV) and the prolonged effect of raised AS over a four-year period showed that (i) prolonged elevated AS (at wave 1 and 3), (ii) the interaction between higher AS at Wave 1 (baseline) and lower CBF at Wave 3 and (iii) the interaction between prolonged elevated AS (at wave 1 and wave 3) and reduced CBF at wave 3 were associated with smaller hippocampal volumes. Our findings suggest a lag effect in the arterial stiffness and hippocampal volume relationship. We propose that the subsequent reduction in cerebral blood flow observed with elevated arterial stiffness may be the missing link in the pathway linking arterial stiffness to hippocampal atrophy. This association between raised arterial stiffness and reduced cerebral blood flow is already reported in the literature in both animal and human subjects (Ajamu et al., 2021; Suri et al., 2020; Sadekova et al., 2018; Jefferson et al., 2018; Muhire et al., 2019). However, contrary to our expectations, reduced CBF at wave 3 in isolation was not associated with hippocampal atrophy, nor was elevated arterial stiffness at Wave 1. Only the combination of both in a model including an interaction term was associated with reduced HV, suggesting CBF plays a buffering effect in the presence of elevated AS. Total hippocampal volume was smaller only in the presence of both higher AS and lower CBF whereas the association between AS and HV was positive in the presence of preserved CBF. The literature does not fully explain why this may be the case. Perhaps when arterial stiffness starts to increase, there is an initial compensatory mechanism that transiently increases flow, but with prolonged exposure, this relationship reverses. A number of authors have proposed that this is due to the fact that in healthy individuals, a compensatory reduction in vascular bed resistance occurs in response to cerebral hypoperfusion, and this maintains adequate cerebral perfusion (Marstrand et al., 2002; Mitchell et al., 2011). Perhaps in the setting of progressive cerebral small vessel disease (i.e increased arterial stiffness), this autoregulation is impaired and thus increased AS leads to an increase in the transmission of harmful pulsatility into the thin-walled cerebral microcirculation, which is unequipped to deal with this increased pressure. This could, over time, ultimately result in vascular remodelling, and resultant reduction in cerebral blood flow and perfusion (Marstrand et al., 2002; Mitchell et al., 2011; Mitchell, 2021). The structure of the hippocampus makes it especially vulnerable to ischemia. The blood supply to the hippocampus is derived primarily from branches of the posterior cerebral artery (PCA) and contributory vessels from the anterior choroidal artery. Animal studies have examined the microvasculature of the hippocampus and unanimously report that the hippocampus has one of the lowest microvascular densities, relative to the surrounding cerebral cortices (Shaw et al., 2021; Nair et al., 1960; Zhang et al., 2019; Kirst et al., 2020).

The hippocampus also has a relatively long distance between its micro vessels and parent supplying artery. On top of this, the hippocampus is highly sensitive to hypoxia and ischemic insult with its high metabolic demand, and this therefore creates a perfect storm scenario whereby the hippocampus is highly vulnerable to vascular insult whilst simultaneously being ill-equipped to deal with states of hypoperfusion (Muhire et al., 2019; Wang and Michaelis, 2010). Therefore, maintaining adequate hippocampal perfusion is essential to preserve cognitive and memory function.

We hypothesised that the CA-1 subfield of the hippocampus would demonstrate preferential vulnerability to ischemia, compared to the other subfields examined; CA2, CA3, CA4, and DG. We hypothesized from our literature search that the CA4 and subiculum may also show regional vulnerability to raised arterial stiffness. However, this hypothesis was only partially confirmed by our results. The supplementary analyses demonstrated that the subiculum and the DG may in fact be most sensitive to higher AS and lower CBF, but these results were not consistent across our analyses. Also, a positive cross-sectional association was observed between AS and the CA-1, CA-3 and CA-4 subfields. The analyses examining these interaction effects suggest a similar effect of raised AS and reduced CBF across all subfields, with smaller subfield volumes observed across the board, as per total HV.

Whilst no prior study, to the best of our knowledge, has specifically examined the effects of increased arterial stiffness and reduced cerebral blood flow on hippocampal subfield volumes, several authors have examined the link between vascular risk factors on specific hippocampal subfield volumes as well as animal models which have investigated the effects of global brain ischemia on the hippocampus and its subfields (Wilde et al., 1997; Sugawara et al., 2000; Wu et al., 2008; Shing et al., 2011).

CA-1 is widely reported in the literature (Sugawara et al., 2000; Wu et al., 2008; Shing et al., 2011) to be the subfield that harbours the most sensitivity to ischemia and hypoxia and thus we tested the independent and moderating effects of AS and CBF on the subfields CA3, CA4, DG and subiculum, with the CA-1 as a reference level. However, this effect of hippocampal atrophy with raised AS and reduced CBF, contrary to our hypothesis, was exerted equally across all subfields tested. We had hypothesised that the CA-1 subfield would be most vulnerable owing to the vascularisation of the hippocampus and the fragility of its vascular network (Spallazzi et al., 2019; Duvernoy et al., 2013; Marinkovic et al., 1999), but this was not borne out in our study. However, along with several authors we observed small independent associations between elevated AS and reduced CBF with smaller volumes of the subiculum (Kim et al., 2015; Pin et al., 2021; Wu et al., 2008; Li et al., 2016).

Pin et al. investigated the effects of neurovascular damage on hippocampal subfields in their study of 249 non-demented adults who underwent two 1.5 Tesla MRI scans, 4 years apart. They examined the association between the atrophy of the CA1-3, CA4 and subiculum with the presence of white matter hyperintensities (WMH’s), which are thought to result from chronic hypoxia (Pin et al., 2021; Tubi et al., 2020). Interestingly, they reported that the subiculum was the only tested subfield in which volume loss correlated with white matter hyperintensities, independent of other vascular risk factors. Whilst these authors did not examine the impact of raised arterial stiffness or reduced cerebral blood flow on hippocampal subfield volumes, they used white matter hyperintensities as a marker of cerebral small vessel disease and linked the increasing quantities of WMH’s to the increasing rates of subiculum atrophy in their longitudinal analyses of their study cohort (Pin et al., 2021). The findings from our study of healthy ageing adults supports the findings of Pin (Pin et al., 2021) and other authors (Kim et al., 2015; Li et al., 2016) and highlight the subiculum as a hippocampal subfield of interest that may demonstrate preferential sensitivity to hypoxia, ischemia, and cerebrovascular insult. There is an obvious paucity of recent research in the literature examining this selective vulnerability of specific subfields to ischemia and in particular the effects of a combination of raised arterial stiffness and reduced cerebral blood flow on the subfield volumes. The hippocampus is the principal site of neurogenesis (specifically within the dentate gryus) and the rate of hippocampal neurogenesis has been shown to decline in patients with Alzheimer’s dementia compared to healthy subjects (Moreno-Jimenez et al., 2019; Lempriere, 2019). This is important in the identification of possible modifiable risk factors that can be used to treat patients with hippocampal vascular insults and thus impaired neurogenesis. Medications such as angiotensin-converting enzymes inhibitors, angiotensin II receptor blockers and calcium channel blockers have been demonstrated to reduce arterial stiffness, promote vascular remodelling and improve endothelial function (Dudenbostel and Glasser, 2012). Further research is required to establish whether these benefits may be extended to the improvement of hippocampal perfusion and neurogenesis.

It is important to better understand this long-accepted link between vascular disease and cognitive decline in order to develop potential therapeutic targets. Better understanding of the pattern of hippocampal atrophy within distinct subfields will help clinicians identify patients that are at risk of cognitive decline. Thus, by mitigating a patient’s vascular risk profile we may reduce their risk of embarking on this pathway of elevated arterial stiffness, reduced cerebral blood flow, subsequent hippocampal atrophy and utimately cognitive decline.

Our study shows that increased arterial stiffness and reduced CBF are not independently associated with whole hippocampal or hippocampal subfield volumes. However, when combined, increased arterial stiffness for a longer duration (spanning the 4 year follow-up period) in combination with a reduction in cerebral blood flow is associated with lower hippocampal volumes. These effects were equally exerted across all hippocampal subfields tested in this study.

Improving our knowledge surrounding the mechanisms that lead to the hippocampal atrophy has clinical implications. Understanding the hippocampal vasculature and its response to hypoperfusion is critical in the development of potential therapeutic targets in patients with an elevated vascular risk profile. Targeting these vascular risk factors, such as arterial stiffness, may help lower the rate of hippocampal atrophy and thus slow the onset of cognitive impairment in our ageing population.