Association Between Cerebral Microbleeds and Circulating Levels of Mid-Regional Pro-Adrenomedullin

Abstract

Background:

Mid-regional pro-adrenomedullin (MR-proADM) is a novel biomarker for cognitive decline based on its association with cerebral small vessel disease (SVD). Cerebral microbleeds (MBs) are characteristic of SVD; however, a direct association between MR-proADM and MBs has not been explored.

Objective:

We aimed to examine whether circulating levels of MR-proADM are associated with the identification of MBs by brain magnetic resonance imaging (MRI) and whether this association could be linked with cognitive impairment.

Methods:

In total, 214 participants (mean age: 75.9 years) without history of cerebral infarction or dementia were prospectively enrolled. All participants underwent brain MRI, higher cognitive function testing, blood biochemistry evaluation, lifestyle examination, and blood MR-proADM measurement using a time-resolved amplified cryptate emission technology assay. For between-group comparisons, the participants were divided into two groups according to whether their levels of MR-proADM were normal (< 0.65 nmol/L) or high (≥0.65 nmol/L).

Results:

The mean MR-proADM level was 0.515±0.127 nmol/L. There were significant between-group differences in age, hypertension, and HbA1c levels (p < 0.05). In the high MR-proADM group, the MR-proADM level was associated with the identification of MBs on brain MR images and indications of mild cognitive impairment (MCI). In participants with ≥3 MBs and MCI, high MR-proADM levels remained a risk factor after multivariate adjustment (OR: 2.94; p < 0.05).

Conclusion:

High levels of MR-proADM may be a surrogate marker for the early detection of cognitive decline associated with the formation of cerebral MBs. This marker would be valuable during routine clinical examinations of geriatric patients.

Keywords

Adrenomedullin biomarkers cerebral small vessel diseases cognitive dysfunction magnetic resonance imaging peptides

INTRODUCTION

In a super-aging society, dementia is an important health concern for which no curative treatments exist. There is an increasing need for risk factor identification that could allow early detection and intervention of cognitive decline. Cerebral arteriosclerosis is related to vascular dementia, which constitutes approximately one-third of dementia diagnoses. In humans, subcortical vascular dementia is typically characterized by cognitive dysfunction, the presence of cerebral microbleeds (MBs), and the detection of ischemic cerebral white matter lesions [1]. However, it is clinically challenging to diagnose vascular dementia in the early stage, resulting in inconsistencies in dementia-related epidemiological data. Thus, a novel and highly sensitive surrogate marker of vascular dementia is required to achieve early clinical detection and for the development of effective management strategies, and several studies have aimed at identifying related risk factors [2, 3].

MBs are associated with cerebral small vessel diseases (SVD) caused by arteriolosclerosis and cerebral amyloid angiopathy (CAA) [4]. The occurrence of MBs originating from microaneurysms and arteriolosclerosis of the medullary artery perfusing the white matter is reported to increase with age. Importantly, these MBs play a role in cognitive decline characterized by a reduced cerebral metabolism and circulation [5 –7], and their effects on cognitive function vary according to their numbers and locations [8 –12]. Owing to technical advancements in brain magnetic resonance imaging (MRI) techniques, MBs have been detected with high sensitivity in asymptomatic patients [13 –16]. Thus, cerebral MBs reflect a patient’s degree of microvascular vulnerability and are considered direct risk factors for vascular dementia [17, 18]. However, the underlying pathology of cerebral MBs has not been fully elucidated.

Adrenomedullin (ADM) is an angiogenic peptide that is released from vascular endothelium as a secondary response to cerebrovascular insufficiency [19]. Preclinical studies in animal models have demonstrated that ADM prevents tissue damage induced by inflammation and suppresses neuronal apoptosis [19, 20]. As the blood-brain barrier deteriorates with advancing age, it becomes more permeable to ADM, which subsequently enters the bloodstream as its precursor, mid-regional pro-adrenomedullin (MR-proADM). Previously, we have reported that MR-proADM is a physiologically active substance involved in the cerebral microvascular vulnerability associated with cognitive decline [7]. However, to our knowledge, no studies have examined the clinical significance of MR-proADM levels as markers for MBs in humans.

In this study, we aimed to examine whether cir-culating levels of MR-proADM are associated with brain MRI findings identifying the presence of MBs. Therefore, we sought to examine whether MR-proADM levels could be an indication of patient-specific clinical characteristics linked with cognitive decline. To achieve this, we recruited a cohort of older adult participants and categorized them into two groups according to whether their circulating levels of MR-pro ADM were normal or high.

MATERIALS AND METHODS

Participants

Participants from Kyoto, Japan were prospectively recruited by random mailing. Those who consented to undergo a brain MRI as a medical check-up for dementia between December 2019 and February 2020 were enrolled in the study. Those who did not complete the study questionnaire or did not undergo brain MRI were excluded. Participants were also excluded if they had a history of dementia or stroke, or had a history of treatment with anti-coagulant, anti-thrombotic, or anti-dementia therapies. The study was approved by the research ethics board of the Kyoto Prefectural University of Medicine (approval number: G-144), and written consent was obtained from all participants. The study was conducted in accordance with the ethical standards of the Helsinki Declaration of 1975.

The following examinations were performed for all participants: completion of a lifestyle questionnaire and performance of a routine clinical check-up, laboratory evaluations of blood biochemistry and genotyping, brain MRI and the measurement of blood MR-proADM levels, and higher cognitive function testing. Then, the participants were categorized into two groups according to whether their blood MR-proADM levels were normal (< 0.65 nmol/L) or high (≥0.65 nmol/L). Data from the examinations were compared between the two groups.

Lifestyle questionnaire and clinical check-up

Lifestyle was assessed using a self-administered questionnaire distributed by trained staff during the initial clinical check-up. The participants were required to answer questions concerning their medical history, alcohol intake, and smoking behavior. Participants were categorized as hypertensive if they were undergoing related treatments, or if they had a systolic or diastolic blood pressure ≥140 mmHg and ≥90 mmHg, respectively, at the time of check-up. Participants were categorized as hyperlipidemic if they were undergoing related treatments or met the following criteria based on the blood test findings at the time of check-up: triglyceride ≥150 mg/dL, high-density lipoprotein < 40 mg/dL, and low-density lipoprotein ≥140 mg/dL [21]. Participants were categorized as diabetic if they were undergoing related treatments or had hemoglobin A1c levels ≥6.5% based on the blood test performed at the time of check-up. Participants were categorized as alcohol consumers if they reported daily alcohol consumption and were categorized as smokers if they reported daily smoking. The participants’ education levels were calculated as the total number of years spent in the educational system. Participants’ body mass indices (kg/m²) were measured based on their heights and weights at the time of check-up.

As is widely applied in clinical settings, we used the modified Rankin Scale for functional assessment in this study. The modified Rankin scale is a six-scale measurement of the degree of independence while performing daily activities, with grade 0 indicating no symptoms of functional impairment and grade 5 indicating severe disability [22].

Laboratory evaluations and genotyping

Measurements of high-sensitivity C-reactive protein levels were included in addition to other, standard biochemical measurements. Genotyping for the evaluation of apolipoprotein E (APOE) allele distribution was performed using a commercially available polymerase chain reaction assay (Funakoshi Co., Ltd. Tokyo, Japan). Based on the genotyping results, participants were further categorized as being APOE ɛ4 allele-carriers or not. As is commonly applied in clinical settings, systemic arteriosclerosis was evaluated be measuring brachial-to-ankle pulse wave velocity using an Automatic Waveform Analyzer (form PWV/ABI; Omron Healthcare Co. Ltd., Kyoto, Japan) [23].

Evaluations by brain MRI

Brain MRI was performed using a 1.5-T scanner (Achieva 1.5T, Philips N.V., Eindhoven, The Netherlands). A certified neurologist (N.K.) and radiologist (K.A.) assessed the images independently. Both were blinded to each other’s findings and data, and only in cases in which their findings differed they discussed them together to reach a consensus. MBs were examined using susceptibility-weighted imaging (SWI; 3D T1-fast field echo (FFE), 4 mm THK/Gapless). The acquisition parameters for axial thin-section 3D-T1-FFE-single point imaging (SPI; 1.5-T, 1.3 mm section thickness, SWI) were as follows: in-plane resolution = 0.65 mm×0.68 mm; slice thickness =1.5 mm; field of view (FOV) = 240×240 mm; echo time (TE) = 38 ms; repetition time (TR) = 104 ms; flip angle (FA) = 20°; and bandwidth (BW) = 17.0 Hz/pixel.

Using SWI, cerebral MBs were identified based on the presence of hemosiderin in MB foci. MBs appear as dotted and round low-intensity areas, without edema in the surrounding regions [24, 25]. We counted the number of MBs and examined their locations within the brain, which was classified as lobar, deep, or mixed [26].

In addition to identifying cerebral MBs, we examined the presence of deep white matter lesions (DWMLs) on fluid-attenuated inversion recovery (FLAIR) or T2-weighted images to identify incidences of cerebral SVD. Thus, image analysis was performed using T1-weighted images (TR = 611 ms; TE = 13 ms), T2-weighted images (TR = 4,431 ms; TE = 100 ms), and FLAIR images (delay time [TI] =2,200 ms; TR = 8,000 ms; TE = 100 ms). For each image, transverse images were generated with 5-mm section thicknesses.

DWMLs were categorized semi-quantitatively from grade 0 (none) to grade 4 (severe) based on the Fazekas classification [27]. Specifically, DWMLs were categorized as follows: grade 0 (absence), grade 1 (solitary, punctate foci), grade 2 (the beginning of foci aggregation), grade 3 (large confluent areas), and grade 4 (severe largely confluent areas) [28]. Similarly, periventricular hyper-intensity (PVH) on brain MRI was categorized using the de Groot classification [29] as follows: grade 0 (absence), grade 1 (“caps” or pencil-thin linings), grade 2 (halos), grade 3 (irregular PVH extending into deep white matter), and grade 4 (large PVH extending into deep white matter). These methods for brain MRI are based on our previously described protocol [6 , 31].

Evaluation of covariates and cognitive function

As is commonly conducted in clinical settings, higher cognitive function was assessed using the Mini-Mental State Examination (MMSE), word fluency test, letter fluency test, and symbol digit modalities test (SDMT). MMSE is used worldwide as a screening tool for higher cognitive function; it consists of 11 questions that add up to a total score of 30 and includes the assessment of memory, calculation, language, and construct ability [32]. Accordingly, the MMSE score was recorded on a scale from 0 to 30 points, with higher scores indicating better cognition. Based on the commonly used definition, mild cognitive decline was defined as an MMSE score of ≥27 points [33].

The word fluency [34] and letter fluency tests [35] are primarily used to assess verbal fluency and are commonly used to measure cognitive function among different age groups. In the word fluency test, participants were asked to verbally list items that fit within a particular category, such as vegetables and animals, for 1 min. According to previously published protocols, we counted the number of words that were correctly mentioned for the purpose of our analysis. According to previously published protocols for the letter fluency test, the participants were asked to verbally list words that start with “ta” and “ka” for 1 min. We counted the number of words correctly mentioned within the time limit for the purpose of our analysis.

The SDMT was used as an assessment of attention and processing speed; it measures the ability to pay attention to varying information, which is expressed as a percentage achievement rate. SDMT was selected as the tool to evaluate early cognitive decline as it can be performed in a relatively short timeframe and is considered to be a highly effective screening tool for mild cognitive impairment (MCI) [36]. During the SDMT, participants were asked to write down numbers that correspond with given geometric figures over a 90-s period per task.

The assessments of higher cognitive function were performed by trained neurologists and neuropsychologists on the same day as the brain MRI.

Measurement of blood MR-proADM levels and routine laboratory evaluations of peripheral blood

Blood tests to measure the MR-proADM levels were performed in all participants on the same day as the brain MRI. Blood samples were collected and stored in tubes containing EDTA. The samples were mixed by gentle rocking and then centrifuged at 1,600×g for 15 min at 4°C to recover the serum. One milliliter of extracted serum was stored at –80°C until the time of MR-proADM measurement. MR-proADM plasma concentrations were measured with a time-resolved amplified cryptate emission technology assay using an automated KRYPTOR analyzer (Thermo Fisher Diagnostics Inc., Waltham, MA, USA). The protocol was followed according to the manufacturer’s instructions [37, 38].

Statistical analysis

Continuous variables and variables that satisfy the assumption of normality are described as means±standard deviations. Between-group comparisons were conducted using the Chi-squared test or one-way analysis of variance (ANOVA), and a paired t-test was used to compare MR-proADM measurements.

Multivariate analysis was performed to identify the association between the MBs and high MR-proADM levels. In the high MR-proADM group, participants were further divided based on the number of identified MBs. Logistic regression analysis was performed to calculate the odds ratio (OR) and 95% confidence interval (CI) using the number of MBs (0, 1–2, ≥3) as an independent variable after adjusting for sex, age, and other clinical factors.

A Spearman’s rank correlation test was used to determine the correlation between the blood MR-proADM level and each covariate, including brain MRI findings and higher cognitive function. All analyses were performed using SPSS 21.0J for Windows (IBM Corp., Armonk, NY, USA), and p < 0.05 was considered statistically significant.

RESULTS

Participants

In total, 225 participants were initially enrolled; five participants were excluded because they did not complete the lifestyle questionnaire, and six participants were excluded because they did not undergo brain MRI. As shown in Table 1, 126 male and 88 female participants were finally included in the study. The mean age of the participants was 75.9±5.8 years. Among them, 43.0%, 30.8%, and 9.3% had hypertension, hyperlipidemia, and diabetes, respectively. These comorbidities are considered risk factors for arteriosclerosis associated with cognitive decline [28, 29]. The mean blood MR-proADM level was 0.515±0.127 nmol/L. Participants who were in the 90^th percentile (≥0.65 nmol/L) for blood MR-proADM level were placed in the high MR-proADM group; the remainder of the participants were placed in the normal MR-proADM group. There were significant between-group differences in age and the proportion of participants with hypertension (p < 0.05). The HbA1c level was also significantly higher in the high MR-proADM group (p = 0.049).

Table 1

Comparison of basic clinical characteristics between participants in the normal and high MR-proADM groups

Variables	Total	Normal MR-proADM group (< 0.65 nmol/L)	High MR-proADM group (≥0.65 nmol/L)	p
	n = 214	n = 186	n = 28
Clinical evaluations
Age (y)^a	75.9±5.8	75.4±5.8	78.2±5.0^*	0.016^*
Sex (male/female)	126/88	104/82	21/7	0.056
Education (y)^a	13.3±2.8	13.2±2.5	13.7±2.8	0.333
BMI^a	22.6±2.5	22.4±2.7	22.9±2.2	0.351
Smoking n (%)	18 (8.4)	6 (3.2)	2 (7.1)	0.302
Alcohol drinking n (%)	76 (35.5)	114 (61.3)	14 (50.0)	0.256
Systolic blood pressure (mmHg)^a	139.8±19.6	131.7±24.6	140.2±27.9	0.096
Diastolic blood pressure (mmHg)^a	78.7±13.2	72.2±19.6	79.7±18.4	0.059
HT n (%)	92 (43.0)	72 (38.7)	20 (71.4)^*	0.011^*
HL n (%)	66 (30.8)	61 (32.7)	5 (17.8)	0.116
DM n (%)	20 (9.3)	15 (8.1)	5 (17.9)	0.097
IHD n (%)	13 (6.1)	9 (4.8)	4 (14.3)	0.051
mRS score^a	0.61±0.49	0.59±0.29	0.68±0.45	0.312
Laboratory evaluations
HbA1c (%)^a	5.9±0.6	5.8±0.7	6.2±1.0	0.049^*
PWV (cm/s)^a	1857.4±386.0	1736.2±357.2	1873.6±394.0	0.063
EGFR (mL/min/1.73m²)^a	64.1±13.4	63.9±9.5	61.7±6.2	0.054
APOE ɛ4 allele carrier n (%)	44 (20.5)	37 (19.8)	7 (25.0)	0.533
hsCRP (mg/dL)^a	0.07±0.02	0.06±0.05	0.08±0.06	0.056
MR-proADM (nmol/L)^a	0.515±0.127	0.478±0.084	0.754±0.117	< 0.001^**

^aValues are presented as means±standard deviations. ^*p < 0.05 by chi-square tests and analysis of variance (ANOVA); ^**p < 0.01 by chi-square tests and ANOVA. MR-proADM, mid-regional pro-adrenomedullin; BMI, body mass index; HT, hypertension; HL, hyperlipidemia; DM, diabetes mellitus; IHD, ischemic heart disease; mRS, modified Rankin Scale; HbA1c, hemoglobin A1c; PWV, pulse wave velocity; eGFR, estimated glomerular filtration rate; APOE, apolipoprotein E; hsCRP, high-sensitivity C-reactive protein.

Higher cognitive function assessments

As shown in Table 2, the mean higher cognitive function score measured by MMSE for the entire cohort (n = 126) was 28.0±1.2. Compared with participants in the normal MR-proADM group, subjects in the high MR-proADM group had significantly lower scores on the MMSE, word fluency (animals) task, letter fluency task (“ta” and “ka”), and SDMT.

Table 2

Between-group comparisons of higher cognitive function assessments and brain MRI findings

Variables	Total	Normal MR-proADM group (< 0.65 nmol/L)	High MR-proADM group (≥0.65 nmol/L)	p
	n = 214	n = 186	n = 28
Cognitive function evaluations
MMSE score^a	28.0±1.2	28.1±1.9	26.5±2.2	< 0.001^**
Animal word fluency score^a	16.7±4.5	17.0±3.5	14.4±3.9	< 0.001^**
Vegetable word fluency score^a	13.1±4.6	13.2±4.7	12.4±3.2	0.256
Letter fluency tasks “Ta” score^a	8.8±3.3	9.1±3.3	7.3±2.9	0.007^**
Letter fluency tasks “Ka” score^a	9.8±3.5	10.1±3.5	8.5±3.1	0.023^*
SDMT (%)^a	43.8±9.7	44.7±10.1	39.4±8.8	0.009^**
Brain MRI findings: DWML n (%)
DWML Grade 0	49 (22.9)	47	2
DWML Grade 1	94 (43.9)	86	8
DWML Grade 2	43 (20.1)	32	11	0.030^*
DWML Grade 3	22 (10.3)	17	5
DWML Grade 4	6 (2.8)	4	2
DWML≥Grade 2	71 (33.1)	53 (28.5)	18 (64.3)^*	< 0.001^**
Brain MRI findings: PVH n (%)
PVH Grade 0	92 (42.9)	84	8
PVH Grade 1	83 (38.8)	71	12	0.124
PVH Grade 2	33 (15.4)	28	5
PVH Grade 3	5 (2.3)	2	3
PVH Grade 4	1 (0.5)	1	0
PVH≥Grade 2	39 (18.2)	31 (16.7)	8 (28.6)	0.128
Brain MRI findings: MBs n (%)
Lobar type	19 (13.5)	10	9^*	0. 009^**
Deep type	38 (22.4)	35	3
Mixed type	12 (5.6)	8	4

^aValues are presented as means±standard deviations. ^*p < 0.05 by chi-square tests and analysis of variance (ANOVA); ^**p < 0.01 by chi-square tests and ANOVA. MRI, magnetic resonance imaging; MR-proADM, mid-regional pro-adrenomedullin; MMSE, Mini-Mental State Examination; SDMT, symbol digit modality test; DWML, deep white matter lesions; PVH, periventricular hyper-intensity; MBs, cerebral microbleeds.

Brain MRI findings

The brain MRI findings in each group are summarized in Table 2. DWML grades ≥2 and cortical MBs were more common in subjects in the high MR-proADM group (p < 0.001); there was no significant difference in PVH of any grade. In our cohort, we found no evidence of cerebral hemorrhage, pathological brain atrophy, infectious diseases, or any other neurodegenerative diseases.

To compare the association of blood MR-proADM levels with the occurrence of MBs, participants were further categorized into three groups based on the number of MBs (0, 1–2, and ≥3). As shown in Fig. 1, the differences in blood MR-proADM levels among the three MBs subgroups were significant. Although more MBs were identified in the occipital and parietal lobes, there was no association between the location of MBs and the level of MR-proADM (Table 3). Thus, high MR-proADM was significantly associated with the occurrence of cortical MBs and the number of MBs (p < 0.01).

Fig. 1

The differences in blood MR-proADM levels among the three MBs subgroups (0, 1–2, and ≥3) were significant. MR-proADM, mid-regional pro-adrenomedullin.

Table 3

Evaluation of the association between MR-proADM levels and the number of cerebral microbleeds (MBs) as well as their locations within the cortex

Number of MBs on brain MR image	0 (n = 142)	1–2 (n = 71)	≥3 (n = 24)	p
MR-proADM (nmol/L)	0.411±0.102	0.519±0.137	0.692±0.105	0.002
Regional distribution
Frontal lobe	–	10	3	0.931
Occipital lobe	–	23	6	0.124
Parietal lobe	–	22	9	0.978
Temporal lobe	–	16	6	0.760

MR-proADM, mid-regional pro-adrenomedullin; MBs, cerebral microbleeds.

To further examine whether high MR-proADM levels were associated with the number of MBs, the odds of having a high MR-proADM level was examined in MCI patients with ≤2 or ≥3 MBs (Table 4). Specifically, a logistic regression analysis was performed on participants with 1–2 and ≥3 MBs against participants with no MBs as a reference. In the high MR-proADM group, the odds of having ≥3 MBs significantly increased after adjusting for age (OR: 5.07, 95% CI: 2.05–9.51). Furthermore, after adjusting for age, sex, and other factors, the odds of having ≥3 MBs if the MR-proADM level was high remained significant (OR: 2.94, 95% CI: 1.41–7.82). Thus, a high MR-proADM level was an independent factor associated with the development of three or more MBs in patients with MCI.

Table 4

Results of logistic regression analyses aimed at identifying whether high MR-proADM levels were independent factors associated with the number of microbleeds (MBs) identified in participants with mild cognitive impairment

	No MBs w/o MCI n = 142 (control)	MBs (1–2) with MCI n = 19 OR (95% CI)	p	MBs (≥3) with MCI n = 12 OR (95% CI)	p
Age (y)	Ref	1.21 (1.05–1.88)	0.035^*	1.17 (1.16–2.28)	0.020^*
Sex (male)	Ref	0.92 (0.47–2.07)	0.725	0.89 (0.51–2.21)	0.627
MR pro-ADM (< 0.65 nmol/L)^#	Ref	2.89 (1.38–5.29)	0.041^*	5.07 (2.05–9.51)	0.024^*
MR pro-ADM (< 0.65 nmol/L)^# #	Ref	2.08 (0.79–6.15)	0.078	2.94 (1.41–7.82)	0.039^*

^*p < 0.05 compared with control. ^#Model 1: adjusted by age and sex. ^# #Model 2: adjusted by age, sex, hypertension, hyperlipidemia, diabetes mellitus, ischemic heart disease, alcohol consumption, smoking, hsCRP levels, education levels, pulse wave velocity, apolipoprotein E ɛ4 allele carrier, estimated glomerular filtration rate, deep white matter lesions, and periventricular hyper-intensity. MR-proADM, mid-regional pro-adrenomedullin; MBs, cerebral microbleeds; MCI, mild cognitive impairment; OR, odds ratio, CI, confidence interval.

Finally, the correlation between the blood MR-proADM level and the number of MBs was examined. As shown in Fig. 2, there was a positive linear correlation between the blood MR-proADM level and the number of MBs (p < 0.01). Furthermore, MR-proADM and higher cognitive function were negatively correlated. Therefore, the blood MR-proADM level is higher in participants with a greater number of MBs and symptoms of cognitive decline.

Fig. 2

Correlation between the blood MR-proADM levels and numbers of MBs identified on brain MR images as well as participants’ cognitive function scores. A significant positive linear correlation was observed between the MR-proADM levels and the number of identified MBs. A significant negative linear correlation was observed between MR-proADM levels and MMSE, word (vegetable) and letter fluency (“Ta”) test, and SDMT scores. ^*p < 0.01 by Spearman’s rank correlation test. MBs, microbleeds; MMSE, Mini-Mental State Examination; SDMT, Symbol Digit Modality Test; MR-proADM, mid-regional pro-adrenomedullin.

DISCUSSION

In this study, we examined whether the level of MR-proADM in the blood is associated with the occurrence of MBs and cognitive decline. Patients with a high MR-proADM level (≥0.65 nmol/L) were significantly older and more likely to have hypertension and elevated HbA1c levels than those with a normal MR-proADM level (< 0.65 nmol/L). Additionally, their MMSE, word fluency, and SDMT scores for attention and processing speed were significantly lower. Brain MRI further demonstrated that the MR-proADM level is significantly associated with the identification of MBs and the number of MBs found in the cortex. The odds of having ≥3 MBs and symptoms of cognitive decline were greater in patients with high MR-proADM levels (OR: 2.94).

There are two possible mechanisms underlying our findings. First, the association between a high MR-proADM level and cognitive decline in individuals with cortical MBs may be attributed to the presence of cerebral SVD in cortical arteries, which is typically undetectable on brain MR images. MBs are predictive markers of stroke [39]. Similar with lacunar infarction and white matter hyper-intensities, MBs are also markers of cerebral SVD caused by arteriosclerosis [25]. Cerebral small vessels maintain connectivity within the network of white matter fibers and basal ganglia, which are highly metabolic neurologic structures. Disruption of these networks by SVD leads to cognitive decline [40]. Indeed, several studies have demonstrated an association between MBs and dementia [41, 42].

Arteriolosclerosis in an affected area may result in an increase in ADM, which subsequently leads to the formation of cortical MBs and a high MR-proADM level. Therefore, high MR-proADM levels may underlie cerebral arteriolosclerosis, microinfarctions, and CAA, which are involved in MB development [43]. MBs are considered deposits of hemosiderin in capillary pericytes. These capillaries actively secrete amyloid-β (Aβ), which is involved in cognitive function [8, 9]. Indeed, Fazekas et al. identified Aβ deposition in cortical MBs while performing autopsies and noted that the patients with amyloid angiopathy were more likely to have symptoms of cognitive decline [44]. Previous studies have demonstrated that Aβ deposition in the cortex and subcortical small vessel walls reduces cerebral blood flow, which subsequently results in hippocampal and cerebral functional impairment [44, 45]. Thus, MBs are indicated in conditions that cause vascular dementia, such as vascular endothelial dysfunction and CAA.

Furthermore, the presence of MBs and the increase in the number of MBs up to a certain threshold has been associated with a low MMSE score, suggesting that MBs themselves are directly associated with cognitive decline [15]. The Rotterdam Scan Study in the Netherlands demonstrated that lobar MBs are markers of cognitive decline, specifically for processing tasks [42]. Another study using diffusion tensor imaging on MRI found that a single cerebral MB can damage the cerebral white matter irrespective of its location, and that the damage becomes greater as the number of MBs increases [46]. These previous findings are consistent with the observations made in the present study.

The second possible explanation for our findings is that the level of MR-proADM, which is a marker of vascular endothelial dysfunction [47], may increase as a secondary response to vascular endothelial damage caused by SVD with a pathological level of amyloid deposition. Specifically, amyloid deposits may gradually form in the vessels of brain parenchyma and the cortical branch such that the vessel walls became fragile and CAA develops. We hypothesize that the consequent vascular damage and formation of MBs would result in an increase in detectable MR-proADM. Although autopsy examination is required to confirm the occurrence of CAA in the brain parenchyma and carotid artery, we believe that this is a plausible mechanism underlying our findings [48].

We have previously reported that the vasoactive peptide, MR-proADM, is a novel biomarker for cerebral arteriosclerosis [7]. ADM is secreted from vascular endothelial cells and smooth muscle cells and is known to have immunomodulatory and antiapoptotic effects [19 , 49]. ADM penetrates the blood-brain barrier and enters the bloodstream as MR-proADM, its precursor. With advancing age, the blood-brain barrier becomes more permeable to ADM, which may explain why participants in the high MR-proADM group were significantly older than participants in the normal MR-proADM group. Maki et al. used an animal model of chronic cerebral circulation insufficiency to demonstrate that MR-proADM is functionally secreted from neurons and cerebral vascular endothelial cells as SVD develops. In turn, they confirmed that the therapeutic administration of ADM reduced SVD lesions and improved cognitive function [19]. This suggests that ADM plays an important role in the development and regulation of SVD lesions, including MBs.

MBs often develop in the cortical white matter boundary and are generally found in the occipital lobe. Indeed, previous studies suggest that lobar MBs are involved in early cognitive decline and are markers of amyloid deposition [10, 50]. In this study, the occurrence of MBs was more common in the occipital and parietal lobes, but the effect of their location on the study outcomes was non-significant. Further, the presence of the APOE ɛ4 allele is reported to correlate with the occurrence of MBs [51]. However, we could not confirm the apparent relationship between MBs and APOE ɛ4 in this study. Additional data are needed to determine the characteristics of cortical MBs according to their location and association with being a carrier for APOE ɛ4.

There are several limitations to our study. First, we acknowledge that there are many factors that may play a role in the development of MBs. Although we included hypertension and other known factors of arteriosclerosis as possible confounding variables, this list may not be exhaustive. Second, our analysis may have been insufficient in addressing the possible effects systemic or local inflammation may have on MBs formation and MR-proADM levels. We used the hs-CRP level as an indicator of inflammation and found the association to be non-significant. In the future, it may be interesting to test the association identified in the present study while including additional inflammatory markers in the analyses. As MBs are known to increase the inflammatory cytokine and Aβ production [52, 53], additional studies are needed to specifically address this. Third, we did not directly examine the association between high MR-proADM and CAA in participants with MCI and MBs. CAA is typically diagnosed using brain amyloid positron emission tomography and by measuring the levels of Aβ in the cerebrospinal fluid. In our cohort of local volunteers, the performance of more invasive procedures was not feasible. Instead, we have planned a small pilot study to determine whether amyloid metabolism is associated with high MR-proADM levels in participants with MCI and MBs. A fourth limitation is the characterization of the participant’s impairment that we derived from the analysis of MRI images. In particular we did not record the occurrence of cerebral superficial siderosis (SS), which is often associated with cerebrovascular disease, tumor, trauma, and amyloid angiopathy (CAA) [53]. Although cerebrovascular disease, tumor, and trauma could be excluded based on the clinical information obtained from the participants, the presence of SS and potentially underlying CAA cannot be completely ruled out. Cerebral SS is a disease characterized by cerebellar ataxia, sensorineural hearing loss, and myelopathy, and in which hemosiderin tends to be deposited under the soft membrane of the cerebellum, brainstem, and other structures. Although In the present study, none of the participants exhibited, such findings, cases of SS associated with CAA may be nearly asymptomatic or have minimal clinical findings, and only microbleeds may be evident on brain MRI. Therefore, our cohort might inadvertently have included one or more of such cases. Future studies should address the potential relationship between SS associated with CAA, and MR-proADM.

In conclusion, we have demonstrated that the level of MR-proADM in the blood is a surrogate marker for the early detection of cognitive decline associated with MBs formation. Measurement of MR-proADM as a biomarker for SVD will contribute to the differential diagnosis of SVD both in future research studies and in clinical settings. From the perspective of geriatric medicine, the measurement of blood MR-proADM levels may be effectively incorporated into routine health check-up procedures for older adults with MCI and MBs. Future clinical and population-based cohort studies are required to validate our findings.

Footnotes

ACKNOWLEDGMENTS

This study was supported by a Grant-in Aid for Scientific Research (B) (18H03052) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. The study was in part supported by a Grant-in-Aid for Scientific Research on Priority Areas of Cancer (No. 17015018) and Innovative Areas (No. 221S0001) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and by the Platform of Supporting Cohort Study and Biospecimen Analysis (CoBiA, JSPS KAKENHI Grant Number 16H06277) from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.

Authors’ disclosures available online ().

References

Erkinjuntti

, Inzitari

, Pantoni

, Wallin

, Scheltens

, Rockwood

, Roman

, Chui

, Desmond

(2000) Research criteria for subcortical vascular dementia in clinical trials. J Neural Transm Suppl 59, 23–30.

Matsui

, Tanizaki

, Arima

, Yonemoto

, Doi

, Ninomiya

, Sasaki

, Iida

, Iwaki

, Kanba

, Kiyohara

(2009) Incidence and survival of dementia in a general population of Japanese elderly: The Hisayama study. J Neurol Neurosurg Psychiatry 80, 366–370.

Ihara

, Yamamoto

(2016) Emerging evidence for pathogenesis of sporadic cerebral small vessel disease. Stroke 47, 554–560.

Altmann-Schneider

, Trompet

, de Craen

, van Es

, Jukema

, Stott

, Sattar

, Westendorp

, van Buchem

, van der Grond

(2011) Cerebral microbleeds are predictive of mortality in the elderly. Stroke 42, 638–644.

Gregg

, Kim

, Gurol

, Lopez

, Aizenstein

, Price

, Mathis

, James

, Snitz

, Cohen

, Kamboh

, Minhas

, Weissfeld

, Tamburo

, Klunk

(2015) Incidental cerebral microbleeds and cerebral blood flow in elderly individuals. JAMA Neurol 72, 1021–1028.

Watanabe

, Kuriyama

, Miyatani

, Nomura

, Naka

, Nakano

, Ihara

, Iwai

, Matsui

, Ozaki

, Koyama

, Nishigaki

, Yamamoto

, Tamura

, Mizuno

, Akazawa

, Takada

, Takeda

, Yamada

, Nakagawa

, Tanaka

, Kanamura

, Friedland

, Watanabe

(2016) Oral Cnm-positive Streptococcus mutans expressing collagen binding activity is a risk factor for cerebral microbleeds and cognitive impairment. Sci Rep 6, 38561.

Kuriyama

, Ihara

, Mizuno

, Ozaki

, Matsui

, Watanabe

, Koyama

, Kondo

, Tokuda

, Tamura

, Yamada

, Akasawa

, Takeda

, Takada

, Mizuno

, Nakagawa

, Watanabe

(2017) Association between mid-regional proadrenomedullin levels and progression of deep white matter lesions in the brain accompanying cognitive decline. J Alzheimers Dis 56, 1253–1262.

Zhang-Nunes

, Maat-Schieman

, van Duinen

, Roos

, Frosch

, Greenberg

(2006) The cerebral beta-amyloid angiopathies: Hereditary and sporadic. Brain Pathol 16, 30–39.

Fisher

, French

, Ji

, Kim

(2010) Cerebral microbleeds in the elderly: A pathological analysis. Stroke 41, 2782–2785.

10.

Yakushiji

, Werring

(2016) Cerebrovascular disease: Lobar cerebral microbleeds signal early cognitive impairment. Nat Rev Neurol 12, 680–682.

11.

Van der Flier

, Cordonnier

(2012) Microbleeds in vascular dementia: Clinical aspects. Exp Gerontol 47, 853–857.

12.

Akoudad

, Wolters

, Viswanathan

, de Bruijn

, van der Lugt

, Hofman

, Koudstaal

, Ikram

, Vernooij

(2016) Association of cerebral microbleeds with cognitive decline and dementia. JAMA Neurol 73, 934–943.

13.

Greenberg

, Vernooij

, Cordonnier

, Viswanathan

, Al-Shahi Salman

, Warach

, Launer

, Van Buchem

, Breteler

, Microbleed Study Group (2009) Cerebral microbleeds: A guide to detection and interpretation. Lancet Neurol 8, 165–174.

14.

Nandigam

, Viswanathan

, Delgado

, Skehan

, Smith

, Rosand

, Greenberg

, Dickerson

(2009) MR imaging detection of cerebral microbleeds: Effect of susceptibility-weighted imaging, section thickness, and field strength. AJNR Am J Neuroradiol 30, 338–343.

15.

Yakushiji

, Noguchi

, Hara

, Nishihara

, Eriguchi

, Nanri

, Nishiyama

, Hirotsu

, Nakajima

, Kuroda

, Hara

(2012) Distributional impact of brain microbleeds on global cognitive function in adults without neurological disorder. Stroke 43, 1800–1805.

16.

Liu

, Buch

, Chen

, Choi

, Dai

, Habib

, Hu

, Jung

, Luo

, Utriainen

, Wang

, Wu

, Xia

, Haacke

(2017) Susceptibility-weighted imaging: Current status and future directions. NMR Biomed 30, 10.1002/nbm.3552.

17.

Cannistraro

, Badi

, Eidelman

, Dickson

, Middlebrooks

, Meschia

(2019) CNS small vessel disease: A clinical review. Neurology 92, 1146–1156.

18.

Wallin

, Roman

, Esiri

, Kettunen

, Svensson

, Paraskevas

, Kapaki

(2018) Update on vascular cognitive impairment associated with subcortical small-vessel disease. J Alzheimers Dis 62, 1417–1441.

19.

Maki

, Ihara

, Fujita

, Nambu

, Miyashita

, Yamada

, Washida

, Nishio

, Ito

, Harada

, Yokoi

, Arai

, Itoh

, Nakao

, Takahashi

, Tomimoto

(2011) Angiogenic and vasoprotective effects of adrenomedullin on prevention of cognitive decline after chronic cerebral hypoperfusion in mice. Stroke 42, 1122–1128.

20.

Buerger

, Ernst

, Ewers

, Uspenskaya

, Omerovic

, Morgenthaler

, Knauer

, Bergmann

, Hampel

(2009) Blood-based microcirculation markers in Alzheimer’s disease-diagnostic value of midregional pro-atrial natriuretic peptide/C-terminal endothelin-1 precursor fragment ratio. Biol Psychiatry 65, 979–984.

21.

Kinoshita

, Yokote

, Arai

, Iida

, Ishigaki

, Ishibashi

, Umemoto

, Egusa

, Ohmura

, Okamura

, Kihara

, Koba

, Saito

, Shoji

, Daida

, Tsukamoto

, Deguchi

, Dohi

, Dobashi

, Hamaguchi

, Hara

, Hiro

, Biro

, Fujioka

, Maruyama

, Miyamoto

, Murakami

, Yokode

, Yoshida

, Rakugi

, Wakatsuki

, Yamashita

(2018) Japan Atherosclerosis Society (JAS) guidelines for prevention of atherosclerotic cardiovascular diseases 2017. J Atheroscler Thromb 25, 846–984.

22.

Quinn

, Dawson

, Walters

, Lees

(2009) Reliability of the modified Rankin Scale: A systematic review. Stroke 40, 3393–3395.

23.

Laurent

, Boutouyrie

, Asmar

, Gautier

, Laloux

, Guize

, Ducimetiere

, Benetos

(2001) Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 37, 1236–1241.

24.

Buch

, Cheng

, Hu

, Liu

, Beaver

, Rajagovindan

, Haacke

(2017) Determination of detection sensitivity for cerebral microbleeds using susceptibility-weighted imaging. NMR Biomed 30, 10.1002/nbm.3551.

25.

Wardlaw

, Smith

, Biessels

, Cordonnier

, Fazekas

, Frayne

, Lindley

, O’Brien

, Barkhof

, Benavente

, Black

, Brayne

, Breteler

, Chabriat

, Decarli

, de Leeuw

, Doubal

, Duering

, Fox

, Greenberg

, Hachinski

, Kilimann

, Mok

, Oostenbrugge

, Pantoni

, Speck

, Stephan

, Teipel

, Viswanathan

, Werring

, Chen

, Smith

, van Buchem

, Norrving

, Gorelick

, Dichgans

(2013) Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 12, 822–838.

26.

Romero

, Preis

, Beiser

, DeCarli

, Viswanathan

, Martinez-Ramirez

, Kase

, Wolf

, Seshadri

(2014) Risk factors, stroke prevention treatments, and prevalence of cerebral microbleeds in the Framingham Heart Study. Stroke 45, 1492–1494.

27.

Fazekas

, Chawluk

, Alavi

, Hurtig

, Zimmerman

(1987) MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol 149, 351–356.

28.

Gordon

, Najmi

, Hsu

, Roe

, Morris

, Benzinger

(2015) The effects of white matter hyperintensities and amyloid deposition on Alzheimer dementia. Neuroimage Clin 8, 246–252.

29.

de Groot

, de Leeuw

, Oudkerk

, van Gijn

, Hofman

, Jolles

, Breteler

(2000) Cerebral white matter lesions and cognitive function: The Rotterdam Scan Study. Ann Neurol 47, 145–151.

30.

Kuriyama

, Mizuno

, Ohshima

, Yamada

, Ozaki

, Shigeta

, Mitani

, Kondo

, Matsumoto

, Takeda

, Nakagawa

, Watanabe

(2013) Intracranial deep white matter lesions (DWLs) are associated with chronic kidney disease (CKD) and cognitive impairment: A 5-year follow-up magnetic resonance imaging (MRI) study. Arch Gerontol Geriatr 56, 55–60.

31.

Ohshima

, Mizuno

, Yamada

, Matsumoto

, Nagakane

, Kondo

, Kuriyama

, Miyazaki

, Takeda

, Nishimura

, Nakagawa

, Ozasa

, Watanabe

(2013) Low vitamin and carotenoid levels are related to cerebral white matter lesions. J Nutr Health Aging 17, 456–460.

32.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

33.

Kaufer

, Williams

, Braaten

, Gill

, Zimmerman

, Sloane

(2008) Cognitive screening for dementia and mild cognitive impairment in assisted living: Comparison of 3 tests. J Am Med Dir Assoc 9, 586–593.

34.

Pasquier

, Lebert

, Grymonprez

, Petit

(1995) Verbal fluency in dementia of frontal lobe type and dementia of Alzheimer type. J Neurol Neurosurg Psychiatry 58, 81–84.

35.

Cerhan

, Ivnik

, Smith

, Tangalos

, Petersen

, Boeve

(2002) Diagnostic utility of letter fluency, category fluency, and fluency difference scores in Alzheimer’s disease. Clin Neuropsychol 16, 35–42.

36.

Fellows

, Schmitter-Edgecombe

(2019) Symbol Digit Modalities Test: Regression-based normative data and clinical utility. Arch Clin Neuropsychol 35, 105–115.

37.

Koyama

, Kuriyama

, Suzuki

, Saito

, Tanaka

, Iwao

, Tanaka

, Maki

, Itoh

, Ihara

, Shindo

, Uehara

(2021) Mid-regional pro-adrenomedullin is a novel biomarker for arterial stiffness as the criterion for vascular failure in a cross-sectional study. Sci Rep 11, 305.

38.

Caruhel

, Mazier

, Kunde

, Morgenthaler

, Darbouret

(2009) Homogeneous time-resolved fluoroimmunoassay for the measurement of midregional proadrenomedullin in plasma on the fully automated system B.R.A.H.M.S KRYPTOR. Clin Biochem 42, 725–728.

39.

Bokura

, Saika

, Yamaguchi

, Nagai

, Oguro

, Kobayashi

, Yamaguchi

(2011) Microbleeds are associated with subsequent hemorrhagic and ischemic stroke in healthy elderly individuals. Stroke 42, 1867–1871.

40.

Bullmore

, Sporns

(2012) The economy of brain network organization. Nat Rev Neurosci 13, 336–349.

41.

Werring

, Frazer

, Coward

, Losseff

, Watt

, Cipolotti

, Brown

, Jager

(2004) Cognitive dysfunction in patients with cerebral microbleeds on T2^*-weighted gradient-echo MRI. Brain 127, 2265–2275.

42.

Poels

, Ikram

, van der Lugt

, Hofman

, Niessen

, Krestin

, Breteler

, Vernooij

(2012) Cerebral microbleeds are associated with worse cognitive function: The Rotterdam Scan Study. Neurology 78, 326–333.

43.

Roob

, Schmidt

, Kapeller

, Lechner

, Hartung

, Fazekas

(1999) MRI evidence of past cerebral microbleeds in a healthy elderly population. Neurology 52, 991–994.

44.

Fazekas

, Kleinert

, Roob

, Kleinert

, Kapeller

, Schmidt

, Hartung

(1999) Histopathologic analysis of foci of signal loss on gradient-echo T2^*-weighted MR images in patients with spontaneous intracerebral hemorrhage: Evidence of microangiopathy-related microbleeds. AJNR Am J Neuroradiol 20, 637–642.

45.

Herzig

, Van Nostrand

, Jucker

(2006) Mechanism of cerebral beta-amyloid angiopathy: Murine and cellular models. Brain Pathol 16, 40–54.

46.

Akoudad

, de Groot

, Koudstaal

, van der Lugt

, Niessen

, Hofman

, Ikram

, Vernooij

(2013) Cerebral microbleeds are related to loss of white matter structural integrity. Neurology 81, 1930–1937.

47.

Koyama

, Sakurai

, Kamiyoshi

, Ichikawa-Shindo

, Kawate

, Shindo

(2015) Adrenomedullin-RAMP2 system in vascular endothelial cells. J Atheroscler Thromb 22, 647–653.

48.

Martinez-Lizana

, Carmona-Iragui

, Alcolea

, Gomez-Choco

, Vilaplana

, Sanchez-Saudinos

, Clarimon

, Hernandez-Guillamon

, Munuera

, Gelpi

, Gómez-Anson

, de Juan-Delago

, Delgado-Mederos

, Montaner

, Ois

, Amaro

, Blesa

, Martí-Fàbregas

, Lleó

, Fortea

(2015) Cerebral amyloid angiopathy-related atraumatic convexal subarachnoid hemorrhage: An ARIA before the tsunami. J Cereb Blood Flow Metab 35, 710–717.

49.

Artunc

, Nowak

, Mueller

, Breidthardt

, Twerenbold

, Wagner

, Peter

, Haering

, Ebmeyer

, Friedrich

(2014) Plasma concentrations of the vasoactive peptide fragments mid-regional pro-adrenomedullin, C-terminal pro-endothelin 1 and copeptin in hemodialysis patients: Associated factors and prediction of mortality. PLoS One 9, e86148.

50.

Yates

, Desmond

, Phal

, Steward

, Szoeke

, Salvado

, Ellis

, Martins

, Masters

, Ames

, Villemagne

, Rowe

, AIBL Research Group (2014) Incidence of cerebral microbleeds in preclinical Alzheimer disease. Neurology 82, 1266–1273.

51.

Thal

, Ghebremedhin

, Rüb

, Yamaguchi

, Del Tredici

, Braak

(2002) Two types of sporadic cerebral amyloid angiopathy. J Neuropathol Exp Neurol 61, 282–293.

52.

Takeda

, Sato

, Morishita

(2014) Systemic inflammation, blood-brain barrier vulnerability and cognitive/non-cognitive symptoms in Alzheimer disease: Relevance to pathogenesis and therapy. Front Aging Neurosci 6, 171.

53.

Cai

, Wang

, He

, Tu

, Tang

, Xiao

, Yan

(2015) Cerebral small vessel disease and Alzheimer’s disease. Clin Interv Aging 10, 1695–1704.

54.

Chen

, Tsai

, Tang

, Lee

, Liu

, Yen

, Jeng

(2019) Advances in cerebral amyloid angiopathy imaging. Ther Adv Neurol Disord 12, 1756286419844113.