Background
Hypertension is related to a higher risk of dementia and more rapid cognitive decline in older adults [
1‐
5]. More specifically, hypertension is associated not only with the development of vascular dementia and vascular cognitive impairment [
6‐
8], but also with that of Alzheimer’s disease (AD) dementia [
9‐
11]. However, the underlying pathological changes that link hypertension to AD remain poorly understood.
While one postmortem study showed that midlife hypertension was associated with more senile plaque and neurofibrillary tangle pathologies at death [
12], other studies reported that late-life high blood pressure (BP) was related only to neurofibrillary tangles or even to neither neurofibrillary tangles nor senile plaques [
13,
14]. In vivo AD biomarker studies based on positron emission tomography (PET) imaging or cerebrospinal fluid (CSF) analysis yielded similarly conflicting findings. Although a couple of studies demonstrated a significant association between high BP and cerebral β-amyloid (Aβ) deposition in older adults [
15,
16], many other studies did not find such direct relationship [
17‐
24]. In terms of tau pathology, while a recent CSF study revealed that current BP had a significant positive association with tau level [
17] and a PET study showed that systolic BP (SBP) synergistically interacted with Aβ on tau deposition [
25], other studies reported that a history of hypertension was not related to CSF tau level or tau deposition on PET [
19,
24]. As the current BP level reflects the appropriateness of hypertension management and a large proportion of hypertensive individuals exhibit poorly controlled BP, current BP status may explain the adverse effects of hypertension better than a history thereof [
26‐
28]. However, few studies have evaluated the effects of both a history of hypertension and current BP on AD pathologies in the same population.
In this context, we aimed to examine the relationships of a history of hypertension and current BP with in vivo AD pathologies (cerebral Aβ and tau deposition on PET) in cognitively normal (CN) older adults. We also investigated the modulatory effects of hypertension and current BP on the association between Aβ and tau deposition.
Discussion
In the current study, neither a history of hypertension nor the current BP was directly associated with Aβ or tau deposition (Tables
2 and
3; Fig.
1). However, the synergistic interaction effect of current BP with Aβ on tau deposition was significant, whereas the interaction effect between a history of hypertension and Aβ was not (Table
5 and Fig.
2). Our finding of no direct association of hypertension history or current BP with Aβ deposition is in line with many previous in vivo AD biomarker studies [
17‐
24], although one postmortem study reported a positive association between BP measured in midlife and senile plaque pathology at death [
12] and a biomarker study reported a positive relationship between BP and Aβ deposition in late middle-aged adults [
15]. Taken together, although we cannot exclude the possibility that high BP in midlife may influence cerebral Aβ accumulation, high current BP or history of hypertension in late-life does not seem to be directly related to an increased Aβ burden.
We found no significant direct association of current BP or hypertension history with cortical tau deposition as well, which is comparable to the results in previous CSF and PET studies [
18,
19,
24]. In contrast to these findings, a recent Chinese study reported a significant association of hypertension history and higher SBP with greater CSF tau level in individuals aged between 40 and 90 years [
17]. However, its subgroup analyses revealed that association was significant only for the younger subgroup (< 65 years of age), but not for the older subgroup (≥ 65 years of age), of which the age distribution and result are similar to those of our study.
While we found no direct association between current BP and tau deposition, there was a significant synergistic interaction effect of current SBP and DBP with Aβ on tau deposition. The relationship between Aβ and tau deposition was greater in individuals with high SBP (or DBP) than in those with low SBP (or DBP), suggesting that current (late-life) BP moderates the relationship between cerebral Aβ and tau. Similarly, a synergistic interaction of SBP and cerebral Aβ burden on cortical tau deposition was recently reported, although the effect of DBP was not investigated [
25]. It is difficult to determine the exact mechanism by which high BP acts in AD pathobiology. However, decreased cerebral blood flow and damaged blood-brain barrier, both of which can be caused by high BP, might be the contributing factors since both of them interact with Aβ to increase tau deposition [
39]. Tau pathology may be exacerbated by increased tau hyperphosphorylation associated with chronic cerebral hypoperfusion and reduced tau clearance related with impairment of blood-brain barrier and glymphatic system [
40‐
42]. Additionally, given that tau pathology is related to the missorting of axonal tau into somatodendritic compartment by Aβ deposition and other neuronal insults [
43], we assessed the potential mediation of cerebrovascular insult, measured by white matter hyperintensities (WMHs), for the synergistic interaction effect of current BP with Aβ on tau deposition. We first analyzed the interaction effect of global WMH volume (instead of a history of hypertension or current BP) with global Aβ on tau deposition. As shown in Additional file
1: Supplementary Table 1, the interaction effect was not significant. Furthermore, the results presented in Table
5 were largely unchanged even when we controlled global WMH volume as an additional covariate (See Additional file
1: Supplementary Table 2). These findings indicate that cerebrovascular injury measured by WMHs does not mediate the synergistic effect of high BP on tau pathology.
In contrast to high current BP, the interaction effect of a history of hypertension with Aβ on tau deposition was not significant. This may support the importance of strict BP monitoring and control in individuals with hypertension in order to prevent AD-related cognitive decline. Multiple studies reported that individuals with untreated hypertension or less intensive BP control were at higher risk of cognitive decline [
44‐
46]. In the present study, most (32/34, 94%) participants with a history of hypertension were on antihypertensive medication. However, 14 (44%) of them still had high current SBP or DBP, i.e., poorly controlled hypertension.
A novel point of this study is that we specifically identified that high SBP and DBP in late-life, but not a history of hypertension, synergistically interact with in vivo cerebral amyloid deposition on tau accumulation. However, several limitations should be considered. First, as the study was cross-sectional, we could not infer causality in the relationship between high BP and AD pathology. In addition, long-term BP variability may be a more important factor than BP at single point [
47]. Further longitudinal studies with larger sample size and sufficient long-term BP variability data are required. Second, the participants had relatively narrow ranges of BP (SBP, 80 mmHg to 160 mmHg; DBP, 50 mmHg to 100 mmHg). Even in the group with a history of hypertension, most individuals (32/34, 94%) were on anti-hypertensive medication and only a few had very high BP (three with SBP ≥ 160 mmHg and three with DBP ≥ 100 mmHg). This might have reduced the likelihood of detecting any direct association between very high BP and in vivo AD pathology. Third, we did not consider the effect of specific details for antihypertensive medication history, such as onset, duration, and dosage of the medication, and specific drug(s) administered, although such details may affect the association between hypertension history and AD biomarkers. Finally, the impact of cerebral blood flow alteration, caused by hypertension, on PET radioligand binding dynamics needs to be considered when interpreting the results. High BP reduces cerebral blood flow, which can potentially decrease estimates of PET ligand binding (i.e., estimates of Aβ or tau deposition) in the brain [
48,
49]. However, given that high BP, synergistically with Aβ, increased cerebral tau deposition rather than decreased it, it is not likely that the influence of cerebral blood flow on estimates of PET ligand binding was a major determinant of our results.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.