Past antihypertensive medication use is associated with lower levels of small vessel disease and lower Aβ plaque stage in the brains of older individuals

Abstract

Aims: This study assesses the association of antihypertensive medication use on the severities of neuropathological cerebrovascular disease (CVD excluding lobar infarction) in older individuals.

Methods: Clinical and neuropathological data were retrieved for 149 autopsy cases >75 years old with or without CVD or Alzheimer's disease and no other neuropathological diagnoses. Clinical data included hypertension status, hypertension diagnosis, antihypertensive medication use, antihypertensive medication dose (where available) and clinical dementia rating (CDR). Neuropathological CVD severity was evaluated for differences with anti-hypertensive medication usage.

Results: Antihypertensive medication use was associated with less severe white matter small vessel disease (SVD, mainly perivascular dilatation and rarefaction), with a 5.6-14.4 times greater likelihood of less severe SVD if medicated. No significant relationship was detected between infarction (presence, type, number and size), lacunes or cerebral amyloid angiopathy and antihypertensive medication use. Only increased white matter rarefaction/oedema and not perivascular dilation was associated with Alzheimer's pathology, with a 4.3 times greater likelihood of reduced Aβ progression through the brain if white matter rarefaction severity was none or mild. Antihypertensive medication use was associated with reduced Aβ progression but only in those with moderate to severe white matter SVD.

Conclusions: This histopathological study provides further evidence that antihypertensive medication use in older individuals is associated with white matter SVD and not with other CVD pathologies. This is mainly due to a reduction in white matter perivascular dilation and rarefaction/oedema. Even in those with moderate to severe white matter SVD, antihypertensive medication use reduced rarefaction and Aβ propagation through the brain.

Keywords: Alzheimer's disease; antihypertensive medication; cerebrovascular disease; neuropathology; small vessel disease.

FIGURE 1

Antihypertensive medication use and cerebrovascular disease pathologies. Stacked bar charts representing the percentage distributions of various cerebrovascular disease pathologies, namely, infarction (A), lacunes (B) and cerebral amyloid angiopathy (C) comparing between antihypertensive medication use (medicated = orchid colour, not medicated = cantaloupe colour). (Ai) Presence or absence of infarct(s). (Aii) Infarct category; presence of none, single or multiple infarcts. (Aiii) Of those that had an infarct(s), proportions of the number of brain regions involved. (Aiv) Of those that had an infarct(s), approximate average volume of infarct(s) was measured in millilitres. Error bars = 95% confidence interval. (Av) Representative macroscopic image of a portion of a brain slice depicting a haemorrhagic infarct (red arrow) seen in the superior frontal gyrus. (Bi) Absence or presence (few small vs many small and large) of lacunes. (Bii) Representative macroscopic image of a portion of a brain slice illustrating lacunes (red arrows) in white matter underlying the insular cortex. (Ci) Presence or absence of cerebral amyloid angiopathy. (Cii) Representative micrograph depicting a blood vessel with cerebral amyloid angiopathy (red arrow).

FIGURE 2

Antihypertensive medication use and small vessel disease severity. (A) Schematic illustrating the neuropathological characteristics and staging of small vessel disease (SVD) used and operationalised in this study from the staging scheme described by Esiri et al [10]. None/limited SVD includes infrequent small perivascular spaces (PVS) in at least one cortical region. Mild SVD includes occasional small PVS and common medium‐sized rarefaction (RR) in at least one cortical region. Moderate SVD includes common medium‐sized PVS and common medium‐sized RR. Severe SVD includes abundant large‐sized PVS and abundant large‐sized RR. Frequency of pathology was assessed by the percentage of vessels that were affected within a sampled region, <20% of vessels affected were rated as infrequent, 20%–30% as occasional, 30%–60% as common and >60% of vessels affected as abundant. The severity of pathology was judged by the size of the lesion (PVS or RR) compared to the thickness of the vessel wall, with a lesion size <3x the thickness of the associated vessel wall rated as small, ≈3x as medium and >3x the vessel wall thickness as large. Large‐rated examples of each type of pathology are shown (inset). Scale bar = 200 μm. *indicates other related pathologies such as spongiosis, cellular infiltrate, congophilic angiopathy and abnormal vessel walls that may or may not be present. (B) Stacked bar charts representing the percentage distribution of the severity of SVD as either none, mild, moderate or severe as outlined in Figure 2A. (C) Pertinent results of the multinomial logistic regression carried out to assess the effect of variables of interest on the likelihood of SVD severity membership using severe SVD as the reference group. For full details of these results, see Table S3.

FIGURE 3

Investigation of the main components of small vessel disease pathology (perivascular space and rarefaction) with antihypertensive medication use, cortical Aβ density, A and B scores for AD neuropathologic change (progression of Aβ and neuritic pathologies through the brain respectively). (A) Column graph depicting percentage area of perivascular spaces averaged across antihypertensive medication groups (medicated = orchid colour, not medicated = cantaloupe colour). Error bars = 95% confidence interval. (B) Stacked bar chart comparing the percentage distributions of rarefaction severity ratings (none, mild, moderate and severe) across antihypertensive medication groups (medicated = orchid colour, not medicated = cantaloupe colour). (C) Scatter plot illustrating the percentage area of cortical Aβ plaques density and perivascular space. (D) Column graph depicting percentage area of cortical Aβ plaques density averaged across rarefaction severity ratings (none, mild, moderate and severe). Error bars = 95% confidence interval. (E) Column graph depicting percentage area of perivascular spaces averaged across the dichotomised A component (A0 or A1 vs A2 or A3) from the ABC score for AD neuropathologic change paradigm. Error bars = 95% confidence interval. (F) Stacked bar chart comparing the percentage distributions of rarefaction severity ratings (none, mild, moderate and severe) across the dichotomised A component (A0 or A1 vs A2 or A3) from the ABC score for AD neuropathologic change paradigm. (G) Column graph depicting percentage area of perivascular spaces averaged across the dichotomised B component (B0 or B1 vs B2 or B3) from the ABC score for AD neuropathologic change paradigm. Error bars = 95% confidence interval. (H) Stacked bar chart comparing the percentage distributions of rarefaction severity ratings (none, mild, moderate and severe) across the dichotomised B component (B0 or B1 vs B2 or B3) from the ABC score for AD neuropathologic change paradigm. Correlation coefficients between severity scores are given in Table S4.

FIGURE 4

Associations between ACE levels, pathologies and antihypertensive medication use. (A) Bar graph showing no difference in frontal cortex ACE levels and the presence or absence of pathological small vessel disease. (B) Bar graph showing no difference in frontal cortex ACE levels in those medicated with antihypertensive medications during life or not. (C) Bar graph showing no difference in frontal cortex ACE levels in those with or without cerebral amyloid angiopathy. (D) Scatter plot illustrating frontal ACE levels vs Aβ levels. (E) Bar graph showing levels of ACE detected in the frontal cortex of cases who were (orchid colour) and were not (cantaloupe colour) medicated with antihypertensive medications during life further dichotomised into a not/low or intermediate/high level of AD neuropathologic change as measured in arbitrary intensity units (AIU) from western immunoblotting. Case types and number of cases are illustrated in key underneath bar graph. (F) Representative western immunoblots of the data presented in (A). Lanes denoted as AD change “−” indicate not/low level of AD change, while those denoted as “+” indicate an intermediate/high level of AD change. Lanes denoted as AH medicated “−” indicated not antihypertensive medicated, while those denoted as AH medicated “+” indicate antihypertensive medicated cases. (G) Bar graphs illustrating levels of Aβ detected in the frontal cortex via western immunoblotting (i) and percentage of Aβ plaque area as measured via immunohistochemistry (ii) in those who had and had not taken ACEIs. Cases types and number of cases are illustrated in key underneath bar graph. Error bars = standard error of the mean. *p < 0.05 and **p < 0.01.