Associations Between Cerebral Vasoreactivity and Cognitive Function in the Middle-Aged Non-Demented Population

Abstract

Background:

Increasing evidence shows early vascular dysregulation in the pathophysiology of Alzheimer’s disease (AD) in elderly population.

Objective:

We wondered about the relationship between vascular health and cognitive performance in middle-aged adults. The present study aims to evaluate whether and which brain vascular hemodynamic parameters are associated with cognitive functions in a middle-aged, non-demented population.

Methods:

We recruited 490 middle-aged community-based participants (30–60 years). Transcranial color-coded sonography was used to measure cerebral vascular hemodynamics, including mean flow velocity, pulsatility index, and breath-holding index (BHI) in the middle cerebral arteries (MCAs). Cognitive functions were assessed using the Montreal Cognitive Assessment (MoCA). A multivariate linear regression model was used to determine the association between the MoCA scores and each intracranial hemodynamic parameter.

Results:

In 369 participants (median age 52 years [IQR 47–56], 48.8% men) with robust acoustic windows, the factors related to poorer MoCA scores were older age, less education extent, and the habitats of cigarette smoking or alcohol consumption. Multivariate analyses did not show a significant association between any intracranial hemodynamic parameters in both MCAs and MoCA scores in the total study population. Left MCA BHI was found to be significantly and independently correlated with the MoCA scores only in people aged 55–60 years (n = 111, B = 0.70, 95% confidence interval, 0.13–1.26, p = 0.017), however, not in people younger than 55 years.

Conclusion:

Our results emphasize the role of neurovascular abnormalities in the early pathophysiology of cognitive impairment and suggest cerebral vasoreactivity as the earliest detectable cognition-associated hemodynamic parameter.

Keywords

Alzheimer’s disease cerebrovascular reactivity cognitive impairment vasomotor system vasoreactivity

INTRODUCTION

Brain microvessels, namely cerebral small vessels, are responsible for regulating cerebral blood flow (CBF) to maintain adequate brain perfusion and energy supply in response to altered systemic blood pressure or several metabolic stimuli [1, 2]. Transcranial Doppler (TCD) is a non-invasive technique that can measure intracranial hemodynamics such as cerebral arterial flow velocity, pulsatility index (PI), and flow velocity changes in response to PaCO2 increment. These hemodynamic parameters could indicate the brain microvascular functions. Lower mean flow velocity (MFV) and higher PI compared with norms indicate decreased global or regional CBF and increased microvascular resistance, respectively, while fewer MFV changes in response to breath-holding represent impaired vasoreactivity [3 –5].

Several studies have shown that people with dementia, either vascular or Alzheimer’s disease (AD), have abnormal intracranial hemodynamics including lower intracranial arterial MFV, higher PI, and impaired vasoreactivity [6] compared with non-demented age-matched subjects [7]. Our previous study also showed that demented, abnormal intracranial hemodynamics, for example, elevated PI, are correlated with poorer neuropsychological scores in old-aged individuals with subjective cognitive decline or mild cognitive impairment (MCI) [8]. Additionally, higher PI in the left middle cerebral artery (MCA) predicts a higher risk of conversion to dementia in these participants [8]. These results suggest that brain microvascular abnormalities could be detected preceding significant cognitive impairment in the disease process of dementia, and certain intracranial hemodynamic parameters could be used to detect cognitive-associated vascular dysfunction at an early stage of the disease.

The present study used transcranial color-coded sonography (TCCS) to measure intracranial hemodynamics in a community-based 30–60 years non-demented population. We wondered whether the association between intracranial hemodynamics and cognitive function could be detected even at a younger age. Additionally, the results might reveal the earliest-detectable hemodynamic parameters that are associated with cognitive function.

METHODS

Study population

The Cardiovascular Disease Risk FACtors Two-Township Study (CVDFACTS) is an ongoing longitudinal cardiovascular cohort study, community-based in two towns, Chu-Dong and Pu-Tzu, of Taiwan [9 –11]. CVDFACTS, which began in 1989, had gone through six recruiting cycles and included over 10,000 participants. The study population of the present study was the sixth cycle of CVDFACTS, which recruited middle-aged participants (30–60 years) from August 2017.

The participants had undergone detailed demographic and medical history recording, neuropsychological and depression questionnaires including the Taiwan version of the Montreal Cognitive Assessment (MoCA), basic laboratory serum testing, and body weight/height measurements. Those with malignancy, severe cardiac or pulmonary disease, major depression, or organic brain diseases such as traumatic brain injury, brain tumor, stroke, or dementia were not included in the present study. Those with global cognitive impairment (MoCA scores < 24) and/or absent temporal acoustic windows on both sides were also excluded for the present analyses. A flowchart of participant recruitment is depicted in Fig. 1. All participants provided written informed consent. This study was approved by the Institutional Review Board of National Yang-Ming University, Taipei, Taiwan.

Fig. 1

Study population recruitment algorithm. L + R- indicates participants with robust left temporal acoustic window and absent right temporal acoustic window; L– R+, robust right temporal acoustic window and absent left temporal acoustic window; L + R+, robust temporal acoustic windows on both right and left sides.

Cognitive assessment: MoCA

The MoCA scale, with high interrater reliability and internal consistency for detecting cognitive impairment, is a cognitive screening test designed to address the limitations of the Mini-Mental State Examination (MMSE) [12]. The major advantages of the MoCA include its sensitivity at an earlier stage of dementia, 90% sensitivity for MCI, and 100% sensitivity for mild dementia. The optimal cutoff point of the MoCA score to define cognitive impairment has been documented [13]. It can be administered in approximately 10 min. The total MoCA score is 30. A higher score indicates better cognition.

Intracranial hemodynamics: MFV and PI

To minimize angle-corrected bias, we used TCCS instead of TCD to evaluate the intracranial hemodynamics. TCCS was conducted by the same examiner using the Phillips ultrasound system (iU22; Philips, New York, NY, USA). All individuals who performed the TCCS were laid for rest for at least 5 min before the TCCS examination. We used a center transmitted frequency of 2 MHz in the color mode and a Doppler gate at 5–10 mm. Angle-corrected velocity was determined whenever the angle correction was less than 60° in a straight arterial segment of at least 2 mm in length in all other cases. The hemodynamics of both sides of the MCAs were assessed and measured. The M1 segment of the MCA was measured and determined through the temporal skull window from depths of 50–75 mm as unidirectional signals toward the probe at the mesencephalic plane and their anatomical association with the other intracranial arterial segments. The measured parameters included peak systolic and end-diastolic flow velocities (PSV and EDV). MFV and PI were determined as (1/3PSV + 2/3EDV) and (PSV –EDV)/MFV, respectively.

Intracranial hemodynamics: CO2 vasoreactivity (Breath-Holding Index)

First, the time-averaged mean velocity (TAMV) of M1 of the MCA at baseline (V_baseline) was obtained from three randomized cardiac cycles on the Doppler spectrum. Individuals were then instructed to hold their breath after normal expiration for 20 s. The TAMV of the last cardiac cycle on the Doppler spectrum at the end of breath-holding was recorded (V_CO2). TAMV acquisition was continuously from baseline to the end of breath-holding to ensure that the Doppler cursor was at the same position. The Breath-Holding Index (BHI) was calculated from these data as the percent increase in TAMV after breath-holding divided by seconds of breath-holding ([V_CO2- V_baseline/ V_baseline] ^*100/20) [5, 14]. CO2 vasoreactivity was also assessed and measured by the same examiner. BHI in 20 randomly sampled subjects’ images was evaluated again separately, and the intra-rater was calculated (k = 0.81; 95% confidence interval: 0.79–0.90). To avoid inter-rater variability, all sonographic investigations and measurements were performed by one single examiner who was blind to participant’s clinical characteristics.

Statistical analyses

Data are presented as mean values (±standard deviation [SD]) for continuous variables if normally distributed, median values (±interquartile range [IQR]) if not normally distributed, or number with percentage for categorical data. Data normality was examined using the Kolmogorov-Smirnov test. Student’s T test or Mann-Whitney U test were used for comparisons. Spearman’s rank correlation was used to analyze the associations between cognitive function and each clinical characteristic or intracranial hemodynamic parameter. Multivariate analyses were then performed to adjust potential confounding factors and validate the significant associations. Statistical significance was set as a two-sided p value < 0.05. SPSS software (version 22.0; IBM Corp. SPSS Statistics for Windows, Armonk, NY, USA) was used for statistical analyses.

RESULTS

We included 369 eligible participants (median age 52 years [IQR 47–56], 48.8% men) in the final analysis (Fig. 1). Those with absent temporal acoustic windows (n = 93) were older, predominantly female, had lower education years, and included fewer current smokers than those with robust temporal acoustic windows (n = 369, 44 with left temporal acoustic window only, 43 with right temporal acoustic window only, and 282 with both temporal acoustic windows) (Supplementary Table 1). Among these 369 participants, 326 had a robust temporal acoustic window on the left side (LMCA group) and 325 on the right side (RMCA group) (Fig. 1). We then analyzed the associations between cognitive function and intracranial hemodynamics in the LMCA and RMCA separately.

The clinical characteristics and hemodynamic parameters of the study population are shown in Table 1. There were no significant differences in the clinical characteristics and hemodynamic parameters between the LMCA and RMCA groups (Table 1).

Table 1

Clinical characteristics and intracranial hemodynamics of the total study population and stratified by the present acoustic window

		Present acoustic window
Variables	Total study population (n = 369)	LMCA (n = 326)	RMCA (n = 325)	p
MoCA scores	28.0±1.6	28.0±1.6	28.0±1.6	0.941
Age, y (IQR)	52 (47–56)	51 (46–56)	51 (46–56)	0.929
30–54	240 (65.0%)	215 (66.0%)	217 (66.8%)
55–60	129 (35.0%)	111 (34.0%)	108 (33.2%)
Sex, Male, n (%)	180 (48.8%)	165 (50.6%)	165 (50.8%)	0.938
Education, y (IQR)	12 (12–16)	16 (12–16)	16 (9–16)	0.402
≤6	8 (2.2%)	6 (1.8%)	7 (2.2%)
6–9	43 (11.7%)	34 (10.4%)	39 (12.0%)
9–12	136 (36.9%)	117 (35.9%)	113 (34.8%)
> 12	182 (49.3%)	169 (51.8%)	166 (51.1%)
BMI, kg/m²	24.4±3.7	24.3±3.8	24.4±3.7	0.161
SBP, mmHg	121.0±16.4	121.0±16.6	121.1±16.4	0.884
DBP, mmHg	75.5±10.5	75.4±10.5	75.6±10.4	0.692
PP, mmHg	45.5±9.6	45.6±9.8	45.5±9.4	0.841
Vascular risk factors
Hypertension, n (%)	58 (15.7%)	47 (14.4%)	52 (16.0%)	0.174
Diabetes, n (%)	28 (7.6%)	25 (7.7%)	21 (6.5%)	0.202
Hyperlipidemia, n (%)	30 (8.1%)	26 (8.0%)	25 (7.7%)	0.766
Smoking, n (%)	111 (30.1%)	97 (29.8%)	102 (31.4%)	0.214
Drinking, n (%)	91 (24.7%)	82 (25.2%)	80 (24.6%)	0.668
Hemodynamics
TAMV, cm/s	48.9±13.0	47.9±13.6	49.8±14.4	0.742
PSV, cm/s	128.9±33.4	125.7±33.9	132.9±37.8	0.587
EDV, cm/s	52.9±15.7	52.1±16.2	54.5±17.3	0.532
PI_b	1.0±0.2	1.0±0.2	1.0±0.2	0.781
PI_h	0.9±0.2	0.9±0.2	0.9±0.2	0.550
BHI	0.9±0.5	0.9±0.6	1.0±0.6	0.603

p value is derived from comparisons between those with robust left but absent right temporal acoustic window versus robust right but absent left temporal acoustic window. MoCA, Montreal Cognitive Assessment; LMCA, participants with robust temporal acoustic window for the left middle cerebral artery; RMCA, participants with robust temporal acoustic window for the right middle cerebral artery; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; TAMV, time-averaged mean flow velocity; PSV, peak systolic flow velocity; EDV, end-diastolic flow velocity; PI_b, pulsatility index before breath-holding; PI_h, pulsatility index after breath-holding; BHI, breath-holding index.

Factors associated with cognitive functions in the total study population

The results showed that older age (coefficient =–0.118, p = 0.033; coefficient = –0.113, p = 0.042), less education extent (p = 0.032; p = 0.033), the habit of cigarette smoking (p = 0.041; p = 0.033), and alcohol consumption (p = 0.013; p = 0.044) were significantly associated with poorer MoCA scores in both LMCA and RMCA groups (Table 2). Additionally, there was no significant association between each intracranial hemodynamic parameter (MFV, PI, and BHI) and MoCA scores in both groups (Table 2).

Table 2

Associations with MoCA scores in the study population stratified by the present acoustic window

	LMCA (n = 326)		RMCA (n = 325)
Variables	Coefficient	p	Coefficient	p
Age	– 0.118	0.033^*	– 0.113	0.042^*
Sex		0.950		0.134
Education	0.119	0.032^*	0.120	0.033^*
BMI	0.007	0.906	0.018	0.754
SBP	– 0.074	0.186	– 0.039	0.488
DBP	– 0.084	0.132	– 0.027	0.630
PP	– 0.042	0.447	– 0.059	0.294
Vascular risk factors
Hypertension		0.185		0.698
Diabetes		0.150		0.364
Hyperlipidemia		0.511		0.203
Smoking		0.041^*		0.033^*
Drinking		0.013^*		0.044^*
Hemodynamics
TAMV	0.009	0.872	– 0.01	0.918
PSV	– 0.004	0.949	0.02	0.727
EDV	– 0.016	0.779	0.03	0.626
PI_b	– 0.031	0.571	– 0.04	0.469
PI_h	– 0.029	0.624	– 0.03	0.588
BHI	0.113	0.052	0.01	0.845

^*p < 0.05. LMCA, participants with robust temporal acoustic window for the left middle cerebral artery; RMCA, participants with robust temporal acoustic window for the right middle cerebral artery; BMI indicates body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; TAMV, time-averaged mean flow velocity; PSV, peak systolic flow velocity; EDV, end-diastolic flow velocity; PI_b, pulsatility index before breath holding; PI_h, pulsatility index after breath holding; BHI, breath-holding index.

Left intracranial vasoreactivity significantly associated with cognitive function only in people aged 55–60 years

Since age was a significant factor associated with the MoCA scores, we also evaluated whether age was associated with intracranial hemodynamics in our population. Table 3 shows the results of associations analyses between each hemodynamic parameter and factors including age. The results showed that age was also a significant factor with certain intracranial hemodynamics, the TAMV and PI but not BHI. Therefore, we assessed the associations between intracranial hemodynamic parameters and cognitive functions would differ in different age subgroups. We divided our study population of LMCA and RMCA into younger (30–54 years) and older (55–60 years) subgroups. Compared with the younger subgroup, the elder subgroup significantly comprised more men, had higher systolic blood pressure, higher pulse pressure, less education years, lower MoCA scores, and more prevalent vascular risk factors such as hypertension, diabetes, cigarette smoking, and alcohol drinking in both LMCA and RMCA groups (Supplementary Table 2). As to the hemodynamic parameters, the elderly subgroup had significantly lower TAMV, lower PSV, lower EDV, and higher PI after breath-holding compared with the younger subgroup in both LMCA and RMCA groups (Supplementary Table 2). In the LMCA group, the PI at baseline was significantly higher in the elderly group (1.02±0.20) than in the younger group (0.95±0.18).

Table 3

Associations between the intracranial hemodynamic parameters and clinical demographics of the study population stratified by the present acoustic window

	LMCA (n = 326)						RMCA (n = 325)
	TAMV		PI_b		BHI		TAMV		PI_b		BHI
Variables	Coefficient	p	Coefficient	p	Coefficient	p	Coefficient	p	Coefficient	p	Coefficient	p
Age	– 0.142	0.010^*	0.173	0.002^*	– 0.031	0.593	– 0.156	0.005^*	0.120	0.030^*	– 0.099	0.089
Hypertension		0.358		0.510		0.891		0.116		0.106		0.722
Diabetes		0.635		0.157		0.202		0.165		0.414		0.429
Hyperlipidemia		0.852		0.855		0.539		0.375		0.729		0.802
Smoking		0.075		0.138		0.189		0.033		0.269		0.152
Drinking		0.708		0.728		0.190		0.080		0.415		0.029^*

^*p < 0.05. LMCA, the participants with robust temporal acoustic window for left middle cerebral artery; RMCA, participants with robust temporal acoustic window for right middle cerebral artery; TAMV, time-averaged mean flow velocity; PI_b, pulsatility index before breath holding; BHI, breath-holding index.

The relationships between intracranial hemodynamics and cognitive function in the different age subgroups are shown in Table 4 (Figs. 2 3). We found that LMCA BHI was significantly associated with MoCA score in the elderly group (55–60 years) (coefficient 0.24, p = 0.016) but not in the younger group (30–54 years). Additionally, no associations between MoCA and hemodynamic parameters in RMCA was found (Table 4).

Table 4

Correlation between MoCA scores and intracranial hemodynamics in different age subgroups

	LMCA				RMCA
	Elderly (n = 111)		Young (n = 215)		Elderly (n = 108)		Young (n = 217)
Hemodynamics	Coefficient	p	Coefficient	p	Coefficient	p	Coefficient	p
TAMV	0.09	0.361	–0.07	0.285	–0.05	0.610	0.01	0.881
PSV	0.07	0.465	–0.09	0.195	–0.01	0.900	–0.06	0.399
EDV	0.04	0.660	–0.10	0.155	–0.05	0.611	–0.06	0.402
PI_b	0.02	0.863	–0.03	0.687	0.08	0.400	0.02	0.736
PI_h	–0.09	0.358	0.03	0.642	0.13	0.218	–0.001	0.993
BHI	0.24	0.016^*	0.04	0.563	–0.05	0.627	–0.01	0.906

^*p < 0.05. LMCA, participants with robust temporal acoustic window for the left middle cerebral artery; RMCA, participants with robust temporal acoustic window for the right middle cerebral artery; TAMV, time-averaged mean flow velocity; PSV, peak systolic flow velocity; EDV, end-diastolic flow velocity; PI_b, pulsatility index before breath-holding; PI_h, pulsatility index after breath-holding; BHI, breath-holding index.

Fig. 2

Scatter plot and linear regression between hemodynamics and MoCA scores in the older group. MoCA, Montreal cognitive assessment; LMCA, left middle cerebral artery; RMCA, right middle cerebral artery; TAMV, time-averaged mean flow velocity; PI_b, pulsatility index before breath-holding; BHI, breath-holding index.

Fig. 3

Scatter plot and linear regression between hemodynamics and MoCA scores in the younger group. MoCA, Montreal cognitive assessment; LMCA, left middle cerebral artery; RMCA, right middle cerebral artery; TAMV, time-averaged mean flow velocity; PI_b, pulsatility index before breath-holding; BHI, breath-holding index.

Associations between the intracranial hemodynamic parameters and the clinical demographics of the study population are provided in Table 3. There were no significantly-associated vascular factors with each hemodynamic parameter including BHI in both LMCA and RMCA. Multivariate analyses adjusting for age, education, cigarette smoking, and alcohol drinking (these were potential confounders since they were associated with poorer MoCA shown in Table 2) remained statistically significant in the association between LMCA BHI and MoCA scores in the elderly (B = 0.70, 95% confidence interval, 0.13–1.26, p = 0.017) (Table 5; Fig. 4). We did not find any other significant factor associated with LMCA BHI (Table 6).

Table 5

Multivariate linear regression analyses of the relationship between LMCA BHI and MoCA score in the elder subgroup

	Unstandardized		Standardized
n = 111	B	95% CI	B	p
Age	–0.03	–0.26–0.20	–0.03	0.800
Education	–0.12	–0.53–0.28	–0.06	0.550
Smoking	–0.13	–0.95–0.69	–0.04	0.761
Drinking	–0.25	–0.59–1.10	0.07	0.553
LMCA BHI	0.70	0.13–1.26	0.24	0.017

LMCA BHI, breath-holding index of the left middle cerebral artery.

Fig. 4

Scatter plot and linear regression between LMCA BHI and MoCA scores in the elder subgroup. MoCA, Montreal cognitive assessment; LMCA BHI, breath-holding index of the left middle cerebral artery.

Table 6

Association between BHI and clinical demographics in the LMCA elderly subgroup

Variables	Coefficient	p
Age	– 0.036	0.723
Sex		0.911
Education	– 0.005	0.957
BMI	0.095	0.353
SBP	0.030	0.764
DBP	0.018	0.856
PP	0.012	0.908
Vascular risk factors
Hypertension		0.636
Diabetes		0.265
Hyperlipidemia		0.795
Smoking		0.429
Drinking		0.729
Hemodynamics
TAMV	0.027	0.792
PSV	– 0.021	0.835
EDV	– 0.054	0.589
PI_b	0.060	0.549
PI_h	– 0.233	0.019^*

^*p < 0.05. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; TAMV, time-averaged mean flow velocity; PSV, peak systolic flow velocity; EDV, end-diastolic flow velocity; PI_b, pulsatility index before breath holding; PI_h, pulsatility index after breath holding.

DISCUSSION

The present study revealed an association between cerebral hemodynamic parameters and cognitive function in middle-aged individuals (30–60 years) who were free of cognitive impairment or stroke. Among the MFV, PI, and BHI of LMCAs and RMCAs, only LMCA BHI was found to be correlated with cognitive function; less vasoreactivity of LMCA was associated with poorer global cognitive performance. Additionally, the association was significant when participants were older than 55 years, but not at younger ages.

Many pieces of evidence support a link between abnormal brain vasculatures and cognitive impairment [15 –17]. Neuropathological studies have revealed prevalent neurovascular pathologies such as arteriolosclerosis, amyloid angiopathy, atherosclerosis, and lipohyalinosis in those with cognitive impairment [18]. These neurovascular pathologies have been shown as major risk factors for AD with clinical evident dementia even after taking apolipoprotein E (APOE) ɛ4 into account and lowering the threshold for dementia for a given AD pathology burden [18, 19]. Several pre-mortem studies using ultrasound or brain magnetic resonance imaging (MRI) have also shown that impaired neurovascular functions and white matter hyperintensities, a presentation of cerebral small vessel disease on brain parenchyma, are correlated with cognitive impairments in vascular dementia and AD, respectively [6 , 21]. Besides, several studies have suggested that neurovascular abnormalities are involved quite early in the pathophysiology of dementia [15 , 22–24]. Large comprehensive studies collecting multiple biomarkers revealed the early role of vascular dysregulation in memory decline in AD before cerebral amyloidosis and tau-mediated neurodegeneration [23], along with the hemodynamic alternations observed before amyloid fibril deposition and neurodegeneration in rodent models [25, 26]. Our previous study also provides evidence supporting an early role of brain microvascular dysfunction in the disease course of dementia; elevated left MCA PI occurred before dementia and independently predicted AD conversion in individuals with subjective cognitive decline or MCI [8]. The present study further showed that even in middle-aged population who had no symptoms, cerebral vascular dysfunction was associated with cognitive impairment. Our results emphasize the importance of vascular abnormalities in the development of dementia.

In the present study, MFV and PI were not associated with cognitive impairment in the middle-aged, non-demented population. Age- or dementia-related MFV decrement and PI elevation generally indicate microangiopathic changes in decreased vascular density and increased vascular stiffness [27, 28]. In contrast, neurovascular functional impairment such as diminished vasodilation in response to CO2 accumulation may occur before significant visible structural lesions on conventional imaging [29, 30]. BHI is a convenient screening method to assess the ability of hypercapnia-induced cerebral arteriolar dilation. We used a relatively short breath-holding time, but it seemed that there is no difference between the BHI with a short (< 27 s) and long (> 27 s) measurement times [5]. The complex pathogenesis of cognition-related neurovascular functions and structural impairments, the timing, and the order in which they occur remains unclear. Our results suggest that hemodynamic parameters that reflect microvascular function (BHI) are the optimal tools to detect the earliest microvascular abnormalities in the disease process of cognitive decline. This postulation is consistent with the current acknowledged concept that vascular dysfunction occurs in the early stages of dementia before vascular remodeling and notable structural change [15–17 , 22].

We did not find any associations between vascular risk factors and cerebral hemodynamics or cognition in the present study population. This could be owing to the delayed effects of vascular risk factors, as previous studies have illustrated that uncontrolled vascular factor at mid-life increase the risk of poorer cognition in later life [31]. Since we only recruited middle-aged adults and there was a lower prevalence of vascular risk factors among our participants, the potential effects of vascular risk factors on cognitive function might not be detected at this stage.

Age is the most important common contributor to brain microvascular disorders and cognitive impairment [28]. In the present study, the association between BHI and cognitive function in the middle-aged non-demented population was shown until participants were more than 55 years old. Consistent with our results, another small-sample-sized study [32] in healthy adults also demonstrated a significant association between memory test scores and cerebral vasoreactivity in the hippocampus in older adults (55–60 years), but not in younger adults (21–45 years). There were more men, lower education, and more prevalence vascular risk factors in the elderly group compared with the younger group in the present study. These factors known to be related to neurovascular damages and lower cognitive function might explain the age difference in the relationship between BHI and cognition. However, the relationship between lower LMCA BHI and poorer cognitive function in the elderly group remained significant after adjusting for these potential confounders.

Our previous longitudinal study revealed that higher MCA pulsation (PI≥1.1) on the left side but not on the right side independently predicted the conversion from MCI to AD [8]. We have concluded that abnormalities in left MCA which supplies larger brain regions, i.e., the dominant hemisphere rather than the other cerebral arteries might cause a more widespread and severe cognitive impairment and a higher risk of conversion to dementia. The present study analyzed the left and right sides’ vascular hemodynamic parameters separately and still found the sides’ preference for the association between microvascular dysfunction and cognitive impairment; MCA BHIs on the left side but not the right side were associated with cognitive impairment in middle-aged non-demented participants. Most studies evaluating the relationship between neurovascular and cognitive impairment used the mean value of the hemodynamic parameters of bilateral sides or data from only one side [6 , 24]. Therefore, the side difference of cognition-related neurovascular abnormalities might not be revealed in these studies.

As per our findings, a study found that only the BHI in left MCA but not right MCA was significantly lower in early patients with AD compared with controls [33], and another study reported that the cerebral amyloid positron emission tomography (PET) positivity-associated endothelial autoregulation reduction in non-demented subjects was also detected in only LMCAs but not RMCAs [34]. Besides ultrasound, side asymmetry was detected in PET, single-photon emission tomography (SPECT), and MRI studies as well [35, 36]. A lower cerebral perfusion detected in patients with AD and the elderly was found predominantly in the left hemisphere [35, 36]. A recent published study also provides evidence for increased left hemisphere vulnerability in AD [37]. These results indicate that the interactions between brain microvascular dysfunction and cognitive impairment might be different in left and right cerebral hemispheres; and left cerebral hemisphere might be more vulnerable to brain microvascular or/and neurodegenerative pathological insults. Future relevant studies should record both sides’ hemodynamics and analyze them separately.

There were limitations to the present study. Due to the technical limitations of ultrasonography, approximately 20% of the participants had an absent temporal acoustic window. To overcome this, we recruited a relatively large population compared to previous studies [3 , 33]. For breath-holding maneuver in demented or stroke patients, one would measure the end-PaCO2 to make sure the correct breath-holding performance in study participants. We did not obtain this measurement since our participants were relatively young with normal mentality. All sonographic evaluations were performed by one skilled technician who was blind to participant’s clinical characteristics and the intra-rater of BHI measurement was good (k = 0.81; 95% confidence interval: 0.79–0.90). Neuropsychological tests were not performed for the different cognitive domains. Consequently, we could not assess associations with specific cognitive domains in the present study. The cross-sectional study design precludes our ability to infer causal relationships. Future studies integrating APOE ɛ4 or other AD biomarkers are needed to provide mechanistic information regarding the association between impaired cerebral vasoreactivity and cognitive function.

Conclusions

The present study found a significant association between cerebral microvascular dysfunction and cognitive function in a middle-aged, non-demented population. Our results again emphasize the important role of neurovascular abnormalities in the early pathophysiology of cognitive impairment and suggest cerebral vasoreactivity as the earliest detectable cognition-associated hemodynamic parameter.

Footnotes

ACKNOWLEDGMENTS

This work was supported by the Ministry of Science and Technology, Taiwan (MOST 109-2314-B-075-048-MY2, MOST 109-2314-B-075-084-, MOST 109-2321-B-009-007- to Chung; MOST 106-2314-B-010-044-, MOST 110-2321-B-010-007 to Wang), and Taipei Veterans General Hospital (V110C-044 to Chung).

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Caplan

(2015) Lacunar infarction and small vessel disease: Pathology and pathophysiology. J Stroke 17, 2–6.

Faraci

, Heistad

(1998) Regulation of the cerebral circulation: Role of endothelium and potassium channels. Physiol Rev 78, 53–97.

Aries

, Elting

, Keyser

, Kremer

, Vroomen

(2010) Cerebral autoregulation in stroke: A review of transcranial Doppler studies. Stroke 41, 2697–2704.

Aaslid

, Markwalder

, Nornes

(1982) Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg 57, 769–774.

Markus

, Harrison

(1992) Estimation of cerebrovascular reactivity using transcranial Doppler, including the use of breath-holding as the vasodilatory stimulus. Stroke 23, 668–673.

Sabayan

, Jansen

, Oleksik

, van Osch

, van Buchem

, van Vliet

, de Craen

, Westendorp

(2012) Cerebrovascular hemodynamics in Alzheimer’s disease and vascular dementia: A meta-analysis of transcranial Doppler studies. Ageing Res Rev 11, 271–277.

Tsivgoulis

, Katsanos

, Papageorgiou

, Dardiotis

, Voumvourakis

, Giannopoulos

(2014) The role of neurosonology in the diagnosis of vascular dementia. J Alzheimers Dis 42(Suppl 3), S251–257.

Chung

, Lee

, Lin

, Wang

(2017) Cerebral artery pulsatility is associated with cognitive impairment and predicts dementia in individuals with subjective memory decline or mild cognitive impairment. J Alzheimers Dis 60, 625–632.

Chuang

, Cheng

, Bai

, Yeh

, Chen

, Pan

(2016) Blood pressure, carotid flow pulsatility, and the risk of stroke: A community-based study. Stroke 47, 2262–2268.

10.

Chuang

, Bai

, Chen

, Lien

, Pan

(2009) Fibrinogen independently predicts the development of ischemic stroke in a Taiwanese population: CVDFACTS study. Stroke 40, 1578–1584.

11.

Chuang

, Bai

, Cheng

, Chen

, Yeh

, Hsu

, Liu

, Pan

(2016) Common carotid artery end-diastolic velocity is independently associated with future cardiovascular events. Eur J Prev Cardiol 23, 116–124.

12.

Nasreddine

, Phillips

, Bédirian

, Charbonneau

, Whitehead

, Collin

, Cummings

, Chertkow

(2005) The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53, 695–699.

13.

Tsai

, Chen

, Chu

, Yang

, Chung

, Liao

, Chou

(2016) Comparing the sensitivity, specificity, and predictive values of the Montreal Cognitive Assessment and Mini-Mental State Examination when screening people for mild cognitive impairment and dementia in Chinese population. Arch Psychiatr Nurs 30, 486–491.

14.

Reinhard

, Schwarzer

, Briel

, Altamura

, Palazzo

, King

, Bornstein

, Petersen

, Motschall

, Hetzel

, Marshall

, Klijn

, Silvestrini

, Markus

, Vernieri

(2014) Cerebrovascular reactivity predicts stroke in high-grade carotid artery disease. Neurology 83, 1424–1431.

15.

Kisler

, Nelson

, Montagne

, Zlokovic

(2017) Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer disease. Nat Rev Neurosci 18, 419–434.

16.

Sweeney

, Montagne

, Sagare

, Nation

, Schneider

, Chui

, Harrington

, Pa

, Law

, Wang

DJJ

, Jacobs

, Doubal

, Ramirez

, Black

, Nedergaard

, Benveniste

, Dichgans

, Iadecola

, Love

, Bath

, Markus

, Salman

, Allan

, Quinn

, Kalaria

, Werring

, Carare

, Touyz

, Williams

SCR

, Moskowitz

, Katusic

, Lutz

, Lazarov

, Minshall

, Rehman

, Davis

, Wellington

, González

, Yuan

, Lockhart

, Hughes

, Chen

CLH

, Sachdev

, O’Brien

, Skoog

, Pantoni

, Gustafson

, Biessels

, Wallin

, Smith

, Mok

, Wong

, Passmore

, Barkof

, Muller

, Breteler

MMB

, Román

, Hamel

, Seshadri

, Gottesman

, van Buchem

, Arvanitakis

, Schneider

, Drewes

, Hachinski

, Finch

, Toga

, Wardlaw

, Zlokovic

(2019) Vascular dysfunction-The disregarded partner of Alzheimer’s disease. Alzheimers Dement 15, 158–167.

17.

Wardlaw

, Smith

, Dichgans

(2019) Small vessel disease: Mechanisms and clinical implications. Lancet Neurol 18, 684–696.

18.

Kapasi

, DeCarli

, Schneider

(2017) Impact of multiple pathologies on the threshold for clinically overt dementia. Acta Neuropathol 134, 171–186.

19.

Arvanitakis

, Capuano

, Leurgans

, Bennett

, Schneider

(2016) Relation of cerebral vessel disease to Alzheimer’s disease dementia and cognitive function in elderly people: A cross-sectional study. Lancet Neurol 15, 934–943.

20.

Bos

, Wolters

, Darweesh

SKL

, Vernooij

, de Wolf

, Ikram

, Hofman

(2018) Cerebral small vessel disease and the risk of dementia: A systematic review and meta-analysis of population-based evidence. Alzheimers Dement 14, 1482–1492.

21.

Malojcic

, Giannakopoulos

, Sorond

, Azevedo

, Diomedi

, Oblak

, Carraro

, Boban

, Olah

, Schreiber

, Pavlovic

, Garami

, Bornstein

, Rosengarten

(2017) Ultrasound and dynamic functional imaging in vascular cognitive impairment and Alzheimer’s disease. BMC Med 15, 27.

22.

van de Haar

, Jansen

JFA

, van Osch

MJP

, van Buchem

, Muller

, Wong

, Hofman

PAM

, Burgmans

, Verhey

FRJ

, Backes

(2016) Neurovascular unit impairment in early Alzheimer’s disease measured with magnetic resonance imaging. Neurobiol Aging 45, 190–196.

23.

Iturria-Medina

, Sotero

, Toussaint

, Mateos-Pérez

, Evans

(2016) Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis. Nat Commun 7, 11934.

24.

Beishon

, Haunton

, Panerai

, Robinson

(2017) Cerebral hemodynamics in mild cognitive impairment: A systematic review. J Alzheimers Dis 59, 369–385.

25.

Niwa

, Kazama

, Younkin

, Carlson

, Iadecola

(2002) Alterations in cerebral blood flow and glucose utilization in mice overexpressing the amyloid precursor protein. Neurobiol Dis 9, 61–68.

26.

Garcia-Alloza

, Gregory

, Kuchibhotla

, Fine

, Wei

, Ayata

, Frosch

, Greenberg

, Bacskai

(2011) Cerebrovascular lesions induce transient β-amyloid deposition. Brain 134, 3697–3707.

27.

Kalback

, Esh

, Castaño

, Rahman

, Kokjohn

, Luehrs

, Sue

, Cisneros

, Gerber

, Richardson

, Bohrmann

, Walker

, Beach

, Roher

(2004) Atherosclerosis, vascular amyloidosis and brain hypoperfusion in the pathogenesis of sporadic Alzheimer’s disease. Neurol Res 26, 525–539.

28.

, Wang

, Ren

, Hu

, Greenberg

, Chen

, Xie

, Jin

(2017) Age-related impairment of vascular structure and functions. Aging Dis 8, 590–610.

29.

Lavi

, Gaitini

, Milloul

, Jacob

(2006) Impaired cerebral CO2 vasoreactivity: Association with endothelial dysfunction. Am J Physiol Heart Circ Physiol 291, H1856–1861.

30.

Sam

, Crawley

, Conklin

, Poublanc

, Sobczyk

, Mandell

, Venkatraghavan

, Duffin

, Fisher

, Black

, Mikulis

(2016) Development of white matter hyperintensity is preceded by reduced cerebrovascular reactivity. Ann Neurol 80, 277–285.

31.

Mukadam

, Sommerlad

, Huntley

, Livingston

(2019) Population attributable fractions for risk factors for dementia in low-income and middle-income countries: An analysis using cross-sectional survey data. Lancet Global Health 7, e596–e603.

32.

Catchlove

, Parrish

, Chen

, Macpherson

, Hughes

, Pipingas

(2018) Regional cerebrovascular reactivity and cognitive performance in healthy aging. J Exp Neurosci 12, 1179069518785151.

33.

Urbanova

, Schwabova

, Magerova

, Jansky

, Markova

, Vyhnalek

, Laczo

, Hort

, Tomek

(2018) Reduced cerebrovascular reserve capacity as a biomarker of microangiopathy in Alzheimer’s disease and mild cognitive impairment. J Alzheimers Dis 63, 465–477.

34.

Brickman

, Guzman

, Gonzalez-Castellon

, Razlighi

, Gu

, Narkhede

, Janicki

, Ichise

, Stern

, Manly

, Schupf

, Marshall

(2015) Cerebral autoregulation, beta amyloid, and white matter hyperintensities are interrelated. Neurosci Lett 592, 54–58.

35.

Roher

, Garami

, Tyas

, Maarouf

, Kokjohn

, Belohlavek

, Vedders

, Connor

, Sabbagh

, Beach

, Emmerling

(2011) Transcranial doppler ultrasound blood flow velocity and pulsatility index as systemic indicators for Alzheimer’s disease. Alzheimers Dement 7, 445–455.

36.

Kogure

, Matsuda

, Ohnishi

, Asada

, Uno

, Kunihiro

, Nakano

, Takasaki

(2000) Longitudinal evaluation of early Alzheimer’s disease using brain perfusion SPECT. J Nucl Med 41, 1155–1162.

37.

Roe

, Vidal-Piñeiro

, Sørensen

, Brandmaier

, Düzel

, Gonzalez

, Kievit

, Knights

, Kühn

, Lindenberger

, Mowinckel

, Nyberg

, Park

, Pudas

, Rundle

, Walhovd

, Fjell

, Westerhausen

; Australian Imaging Biomarkers and Lifestyle Flagship Study of Ageing (2021) Asymmetric thinning of the cerebral cortex across the adult lifespan is accelerated in Alzheimer’s disease. Nat Commun 12, 721.