Cerebral Vasomotor Reactivity in Amnestic Mild Cognitive Impairment

Abstract

Background:

Cerebral blood flow (CBF) is sensitive to changes in arterial CO₂, referred to as cerebral vasomotor reactivity (CVMR). Whether CVMR is altered in patients with amnestic mild cognitive impairment (aMCI), a prodromal stage of Alzheimer disease (AD), is unclear.

Objective:

To determine whether CVMR is altered in aMCI and is associated with cognitive performance.

Methods:

Fifty-three aMCI patients aged 55 to 80 and 22 cognitively normal subjects (CN) of similar age, sex, and education underwent measurements of CBF velocity (CBFV) with transcranial Doppler and end-tidal CO₂ (EtCO₂) with capnography during hypocapnia (hyperventilation) and hypercapnia (rebreathing). Arterial pressure (BP) was measured to calculate cerebrovascular conductance (CVCi) to normalize the effect of changes in BP on CVMR assessment. Cognitive function was assessed with Mini-Mental State Examination (MMSE) and neuropsychological tests focused on memory (Logical Memory, California Verbal Learning Test) and executive function (Delis-Kaplan Executive Function Scale; DKEFS).

Results:

At rest, CBFV and MMSE did not differ between groups. CVMR was reduced by 13% in CBFV% and 21% in CVCi% during hypocapnia and increased by 22% in CBFV% and 20% in CVCi% during hypercapnia in aMCI when compared to CN (all p < 0.05). Logical Memory recall scores were positively correlated with hypocapnia (r = 0.283, r = 0.322, p < 0.05) and negatively correlated with hypercapnic CVMR measured in CVCi% (r = –0.347, r = –0.446, p < 0.01). Similar correlations were observed in D-KEFS Trail Making scores.

Conclusion:

Altered CVMR in aMCI and its associations with cognitive performance suggests the presence of cerebrovascular dysfunction in older adults who have high risks for AD.

Keywords

Cerebral blood flow cerebral vasomotor reactivity hypercapnia hypocapnia mild cognitive impairment transcranial Doppler

INTRODUCTION

Cerebral blood flow (CBF) is sensitive to changes in arterial partial pressure of carbon dioxide (PaCO₂). Elevated PaCO₂ (hypercapnia) increases CBF via cerebral vasodilation, whereas reduced PaCO₂ (hypocapnia) decreases CBF due to vasoconstriction [1]. These hemodynamic responses referred to as cerebral vasomotor reactivity (CVMR) have been measured in clinical and research settings to assess cerebrovascular function [2]. Recent studies suggest that CVMR could be a potential biomarker used to detect cerebrovascular dysfunction in neurodegenerative disease, including Alzheimer’s disease (AD) and mild cognitive impairment (MCI) [3 –5].

Altered cerebrovascular function, as manifested by brain hypoperfusion and impaired CBF regulation, may contribute to the onset and development of AD [3, 5]. In this regard, most population-based studies have shown that global or regional CBF, or both are reduced in patients with AD and MCI [3, 5]. However, whether CVMR is altered currently remains inconclusive, likely due to differences in the methodology used to assess CVMR as well as the heterogeneity of study populations [3, 5]. For instance, Shim et al. (2015) using transcranial Doppler (TCD) and a breath-holding method reported that CVMR was impaired in patients with AD, but not in MCI when compared with cognitively normal (CN) older adults, and that reduced CVMR was associated with cognitive performance measured with Mini-Mental State Examination (MMSE) [6]. However, Cantin et al. (2011) and Richiardi et al. (2015), using blood oxygen level-dependent (BOLD) functional magnetic resonance imaging (MRI), showed that CVMR was impaired in both AD and MCI patients [7, 8]. In addition, alterations in arterial blood pressure (BP) and autonomic neural activity induced by changes in PaCO₂ via chemoreflex may confound CVMR assessment and their effects were not accounted for CVMR assessment in most of previous studies [9 –12].

The purpose of this study was to determine whether CVMR is altered in patients with amnestic MCI, a prodromal stage of AD, when compared with CN adults. We hypothesized that 1) CVMR would be reduced in patients with amnestic MCI after accounting for changes in BP induced by changes in CO₂, and 2) alterations in CVMR would be associated with cognitive performance [4, 6]. Based on the observed asymmetrical CBF responses to decreases and increases in CO₂ [12], we assessed CVMR during both hypo- and hypercapnia.

MATERIALS AND METHODS

Study participants

Fifty-three patients with amnestic MCI and 22 CN subjects were recruited from local newspaper advertisements, senior centers, and the University of Texas Southwestern Medical Center Alzheimer’s Disease Center. The diagnosis of amnestic MCI was based on Petersen criteria [13, 14] as modified by the Alzheimer’s Disease Neuroimaging Initiative (http://adni-info.org/). Specifically, we used a global Clinical Dementia Rating scale of 0.5 with a score of 0.5 in the memory category and MMSE score between 24 and 30 [15]. Clinical evaluation and diagnosis of MCI was performed based on the recommendations from the Alzheimer’s disease Cooperative Study Diagnostic Criteria (http://adni-info.org/) via multidisciplinary consensus following a review of all available information. A subgroup of 42 patients with amnestic MCI had ¹⁸F-florbetapir positron emission tomography (PET) and 36 of these patients (85.7%) were amyloid positive using a threshold of mean cortical standardized uptake value ratio (SUVR)≥1.10 [16]. Objective memory function was assessed using the education-adjusted scores on immediate and delayed recall scores from the Logical Memory (LM) subtest of the Wechsler Memory Scale-Revised [15]. Additional neuropsychological tests included the long delayed free recall and total learning scores from the California Verbal Learning Test-second edition (CVLT-II) [17] and the Trail Making, Color-word inhibition, and Letter and Category fluency test scores from the Delis-Kaplan Executive Function System (D-KEFS) [16, 18]. The D-KEFS results are reported using the standardized scores.

Inclusion criteria were men and women aged between 55 and 80 years who were diagnosed with amnestic MCI or judged to be cognitively normal. Exclusion criteria included major neurologic, or psychiatric disorders, uncontrolled hypertension, clinically diagnosed or self-reported diabetes mellitus, sleep apnea, body mass index (BMI) ≥35 kg/m², or current or history of smoking. Individuals with a physically active lifestyle (defined as participation in moderate to vigorous intensity aerobic exercise training over the past two years for 3 times per week with each session lasting >30 min) were also excluded due to potential effect of exercise training on CVMR [19]. Detailed inclusion and exclusion criteria are presented on ClinicalTrials.gov (NCT01146717). Screening procedures included a detailed medical history and medication questionnaire, a comprehensive physical examination, 12-lead electrocardiogram (ECG), echocardiography, carotid artery ultrasound to exclude severe stenosis (>50%), and 24 h ambulatory BP monitoring.

This study was approved by the Institutional Review Board of the University of Texas Southwestern Medical Center and Texas Health Presbyterian Hospital Dallas, and was performed in accordance with the guidelines of the Declaration of Helsinki and Belmont Report. All subjects provided informed written consent prior to participation. All collected data were de-identified prior to further analysis and blinded for outcome measures.

Study protocol

All procedures were performed in an environmentally controlled laboratory with an ambient temperature of ∼22°C. Subjects refrained from high intensity exercise, caffeinated beverages, or alcohol > 24 h before experiment. After subjects rested in the supine position for > 10 min, a nose clip was placed and subjects breathed through a mouthpiece with a Y valve, with one end connected to the mouthpiece, one end open to room air, and one end connected to a 5L rebreathing bag [9 , 20].

The protocol used in this study is explained in detail in our recent study [12]. First, baseline CBFV, HR, MAP, and EtCO₂ were recorded simultaneously for 3 min. During these measurements, subjects were instructed to breathe normally and avoid body movement or Valsalva maneuvers. After baseline data collection, subjects were coached by an investigator to perform voluntary hyperventilation for 20 s (1 breath/s) which induced a brief period of hypocapnia to assess the cerebral vasoconstrictor response. Following hyperventilation, a > 5 min recovery period was provided until all hemodynamic variables returned to the baseline level [9]. Then, a modified rebreathing protocol was used to induce hypercapnia to assess cerebral vasodilatory response [10]. At the end of a deep inspiration, the Y-way valve of the mouthpiece was switched to the rebreathing bag to induce a progressive increase in EtCO₂ for 3 min [10]. During rebreathing, a small amount of oxygen was added to the rebreathing bag based on each subject’s basal metabolic rate (estimated using the Harris-Benedict formula) to maintain constant arterial blood oxygen saturation [10]. Intermittent brachial cuff BP was measured at baseline and during the rebreathing protocol to corroborate finger arterial BP measurement. The rebreathing protocol was tolerated by all subjects.

Measurements of cerebral and systemic hemodynamics

CBF velocity (CBFV) was measured from the middle cerebral artery (MCA) using a 2 MHz TCD probe (Multi-Dop X2, Compumedics/DWL, Germany). The probe was securely attached to the temporal acoustic window using either an individually created mold to fit the facial bone structure or a probe holder (Spencer Technologies, Seattle, WA, USA) to keep the position and angle of the probe unchanged during the study [21]. TCD has a high temporal resolution (>100 ms) and allows a non-invasive measurement of beat-by-beat changes in CBF. End-tidal carbon dioxide (EtCO₂), an estimate of PaCO₂ [22], and breathing frequency were monitored using capnography (Carpnograd, Novamatrix, Wallingford, CT, USA). Arterial blood oxygen saturation (SaO₂) was measured by a pulse oximeter (Biox 3700, Ohmeda Monitoring Systems, Boulder, CO, USA). Brachial BP was intermittently measured from the right upper arm using an electro-sphygmomanometer (Suntech, Morrsville, NC, USA). Beat-to-beat mean arterial pressure (MAP) was continuously monitored from the middle finger of the left hand using Finapres (Finapres Medical Systems, Amsterdam, The Netherlands). The finger pressure transducer was fixed at the heart level during the study. Heart rate (HR) was recorded via a 3-lead ECG system (Hewlett-Packard, Palo Alto, CA, USA). All data were collected with a sampling frequency of 1000 Hz and stored in a computer for off-line analysis using a data acquisition and analysis software (Acknowledge, BIOPAC systems, Goleta, CA, USA).

Data analysis

Baseline data were obtained by averaging a 3 min steady-state data segment under resting condition before hyperventilation and averaging a 1 min data segment before rebreathing protocol. Cerebrovascular conductance index (CVCi) and resistance index (CVRi) were calculated from the ratio of mean CBFV and MAP. CVCi was calculated to account for the effects of changes in MAP on CBFV during hypo- and hypercapnia [10]. The magnitude of absolute changes in CBFV, CVCi, MAP, HR, and EtCO₂ during hypo- and hypercapnia are presented as ΔCBFV, ΔCVCi, ΔMAP, ΔHR, and ΔEtCO₂, respectively. Because relative change in CBFV is proportional to change in volumetric CBF [23], percentage changes in hypo- and hypercapnic ΔCBFV% and ΔCVCi% were calculated relative to their corresponding baseline values and reflect the magnitude of cerebral vasoconstriction and vasodilation respectively.

Hypocapnic CVMR

During hyperventilation, maximal hemodynamic changes were calculated from the average of 3 breath cycles after the reduction of EtCO₂ reached nadir. Then, hypocapnic CVMR (cerebral vasoconstriction) was calculated as the ratio of ΔCFBV% to ΔEtCO₂ and ΔCVCi% to ΔEtCO₂ [10]. Maximal reductions in CBFV% and CVCi% in response to the maximal reduction in EtCO₂ were used to assess CVMR. Cardiovascular reactivity to changes in EtCO₂ was calculated as the ratio of ΔMAP to the corresponding changes in EtCO₂.

Hypercapnic CVMR

Baseline data for hypercapnia were obtained by averaging a 1-minute steady-state data before rebreathing protocol. Breath-by-breath data were extracted during rebreathing [10, 24]. Due to a deep inspiration performed while switching a Y-way valve, a brief reduction in both EtCO₂ and CBFV was observed at the beginning of rebreathing. Hence, analysis was performed on ΔEtCO₂ and ΔCBFV% above their baseline levels. Linear regression analysis of ΔCBFV% versus ΔEtCO₂ and ΔCVCi% versus ΔEtCO₂ was performed within each subject and then group averaged for statistical analysis. The slopes of these regression lines were used as the estimates of hypercapnic CVMR (cerebral vasodilation) [10]. Cardiovascular reactivity to ΔEtCO₂ was assessed by the slope of linear regression between ΔMAP and ΔEtCO_2 . The results of linear regression fitting were examined by the coefficient of determination (R²). For data visualization, group-averaged bin plots of CBFV%, CVCi%, and MAP were created based on every 4 mmHg increase in EtCO₂ from the baseline (Fig. 1).

Fig.1

Group-averaged plots showing the cerebral blood flow velocity (CBFV%), cerebrovascular conductance index (CVCi%), mean arterial pressure (MAP), and end-tidal CO₂ (EtCO₂) during rebreathing in patients with amnestic mild cognitive impairment and cognitively normal subjects. The error bars represent standard deviations. Each bin represents 4 mmHg changes in EtCO₂.

Statistical analysis

Group differences in categorical variables were examined by χ-square test and continuous variables by Student’s t-test. Pearson’s product-moment correlation was used to examine the relationship between hypo- and hypercapnic CVMRs and cognitive performances. Partial correlation was used to examine the association between CVMR and cognitive performance after adjustment for age, sex, and education. Data normality was checked by the Shapiro-Wilk test and the visual inspection of histogram and Q-Q plots. An α-level of 0.05 was set as the criterion for statistical significance. All statistical analyses were performed using SPSS 20.0 (SPSS, Chicago, IL, USA).

RESULTS

Baseline hemodynamics

Subject demographics and clinical status are presented in Table 1. Amnestic MCI and CN subjects did not differ in age, sex, education, BMI, BP, MMSE, D-KEFS Color Word inhibition or Letter fluency scores, HR, mean CBFV, CVCi, and CVRi. LM Immediate and Delayed Recall scores, CVLT-II total and long delayed free recall score, D-KEFS Trail Making Test, and Category fluency were lower in patients with amnestic MCI than CN subjects.

Table 1

Subject characteristics, neurocognitive measures, and baseline hemodynamics

Variables	CN	Amnestic MCI	p
Men/Women (n)	13/9	22/31	0.224
Age (y)	65.8±7.4	64.0±5.8	0.260
Education (y)	16.7±2.1	15.9±2.4	0.156
Height (cm)	171.9±8.4	168.1±9.7	0.121
Body mass (kg)	80.2±15.8	77.9±15.7	0.558
BMI (kg/m²)	27.1±4.5	27.4±4.3	0.760
Systolic BP (mmHg)	122.2±15.2	122.2±12.1	0.986
Diastolic BP (mmHg)	73.8±8.9	73.4±7.8	0.845
Neurocognitive measures
CDR	0	0.5	-
MMSE	29.1±0.9	29.0±1.3	0.658
LM Immediate Recall	15.4±3.0	12.0±2.6	<0.001
LM Delayed Recall	14.9±2.7	10.4±3.2	<0.001
CVLT-II total score	54.4±11.0	45.3±10.1	0.001
CVLT-II long delayed free recall	12.2±2.5	9.8±2.4	<0.001
D-KEFS Trail Making Test	12.6±1.9	11.0±2.9	0.022
D-KEFS Color-Word inhibition	11.0±2.3	10.4±3.0	0.457
D-KEFS Letter fluency	11.2±2.3	10.3±3.0	0.193
D-KEFS Category fluency	13.4±2.7	10.6±3.3	<0.001
Baseline Hemodynamics
EtCO₂ (mmHg)	37.4±3.7	37.1±3.7	0.789
SaO₂ (%)	98.1±1.1	98.1±1.4	0.990
Heart rate (bpm)	65.8±9.8	66.3±10.8	0.841
MAP (mmHg)	95.9±13.5	96.9±14.0	0.783
Mean CBFV (cm/s)	53.0±13.3	48.7±9.0	0.117
CVCi (cm/s/mmHg)	0.56±0.15	0.51±0.11	0.108
CVRi (mmHg/cm/s)	1.90±0.49	2.05±0.49	0.224

Data are mean±standard deviation. Bold values represent p < 0.05. BMI, body mass index; BP, blood pressure; CDR, clinical dementia rating; MMSE, Mini-Mental States Exam; LM, logical memory; CVLT-II, California Verbal Learning Test-second edition; D-KEFS, Delis-Kaplan Executive Function System; EtCO₂, end-tidal CO₂; SaO₂, arterial blood oxygen saturation; bpm, beats per minute; MAP, mean arterial pressure; CBFV, cerebral blood flow velocity; CVCi, cerebrovascular conductance index; CVRi, cerebrovascular resistance index.

Hypocapnic CVMR

Hemodynamic data during hyperventilation are presented in Table 2. During hyperventilation, EtCO₂, MAP, CBFV, and CVCi all decreased while HR increased from the baseline. Despite similar reductions of EtCO₂ in both groups, the magnitudes of both absolute and CBFV% and CVCi% reductions was attenuated in patients with amnestic MCI. Figure 2 presents group differences in hypocapnic CVMRs and cardiovascular reactivity. Both ΔCBFV% /ΔEtCO₂ and ΔCVCi% /ΔEtCO₂ were reduced in patients with amnestic MCI when compared with CN subjects, while ΔMAP/ΔEtCO₂ was similar between the groups.

Table 2

Cerebral and systemic hemodynamics during hypocapnia

Variables	CN	Amnestic MCI	p
ΔEtCO₂ (mmHg)	–18.8±2.8	–18.5±4.6	0.721
ΔHeart rate (bpm)	9.0±7.6	9.7±7.6	0.701
ΔMAP (mmHg)	–12.0±5.2	–11.9±5.0	0.925
ΔCBFV (cm/s)	–20.6±7.5	–16.0±6.4	0.010
ΔCBFV (%)	–38.2±6.6	–32.2±9.4	0.009
ΔCVCi (cm/s×mmHg)	–0.17±0.06	–0.12±0.06	0.002
ΔCVCi (%)	–29.3±6.0	–22.9±9.2	0.003
ΔCBFV/ΔEtCO₂ (cm/s/mmHg)	1.08±0.30	0.88±0.32	0.012
ΔCBFV/ΔEtCO₂ (% /mmHg)	2.03±0.24	1.77±0.42	0.001
ΔCVCi/ΔEtCO₂ (cm/s/mmHg/mmHg)	0.009±0.002	0.006±0.003	0.001
ΔCVCi/ΔEtCO₂ (% /mmHg)	1.56±0.20	1.23±0.37	<0.001
ΔMAP/ΔEtCO₂	0.64±0.27	0.67±0.30	0.734

Data are mean±standard deviation. Bold values represent p < 0.05. CN, cognitively normal subjects; MCI, mild cognitive impairment; EtCO₂, end-tidal CO₂; bpm, beats per minute; MAP, mean arterial pressure; CBFV, cerebral blood flow velocity; CVCi, cerebrovascular conductance index.

Fig.2

Box-and-whisker plots of cerebral vasomotor reactivity during hypo- and hypercapnia. Changes in cerebral blood flow velocity (ΔCBFV), cerebrovascular conductance index (ΔCVCi), and mean arterial pressure (ΔMAP) in response to changes in end-tidal CO₂ (ΔEtCO₂ in patients with amnestic mild cognitive impairment (MCI) and cognitively normal (CN) subjects. The boundaries of the box represent the 25th and 75th percentiles, the lines within the box are the median values. Error bars indicate the 10th and 90th percentiles.

Hypercapnic CVMR

Hemodynamic data during rebreathing are summarized in Table 3. Despite similar increases in EtCO₂ in both groups, elevations of ΔCBFV% and ΔCVCi% were greater in patients with amnestic MCI. Similarly, hypercapnic CVMRs measured from the linear regressions of ΔCBFV% versus ΔEtCO₂ and ΔCVCi% versus ΔEtCO₂ were steeper in patients with amnestic MCI compared with the CN subjects (Table 3 and Figs. 1 and 2). Both the absolute and the slope of ΔMAP versus ΔEtCO₂ was greater in patients with amnestic MCI than in CN subjects, while the HR response was similar between the groups (Table 3 and Figs. 1 and 2).

Table 3

Cerebral and systemic hemodynamics during hypercapnia

Variables	CN	Amnestic MCI	p
ΔEtCO₂ (mmHg)	19.3±3.1	18.4±2.8	0.223
ΔSaO₂ (%)	0.05±1.56	0.20±1.49	0.695
ΔHeart rate (bpm)	4.4±5.3	5.7±8.2	0.477
ΔMAP (mmHg)	18.6±3.9	20.9±4.4	0.040
ΔCBFV (cm/s)	39.8±10.7	43.2±11.6	0.234
ΔCVCi (cm/s/mmHg)	0.25±0.07	0.28±0.10	0.153
ΔCBFV (%)	76.1±18.2	91.6±26.8	0.005
ΔCVCi (%)	47.9±15.4	58.1±21.9	0.049
ΔCBFV/ΔEtCO₂
Slope (cm/s/mmHg)	2.59±0.99	2.75±0.71	0.421
Slope (% /mmHg)	4.76±1.18	5.81±1.49	0.005
R²	0.93±0.03	0.94±0.04	0.290
ΔCVCi/ΔEtCO₂
Slope (cm/s/mmHg/mmHg)	0.016±0.005	0.018±0.006	0.157
Slope (% /mmHg)	3.05±1.06	3.66±1.24	0.047
R²	0.87±0.08	0.89±0.06	0.238
ΔMAP/ΔEtCO₂
Slope (mmHg/mmHg)	1.10±0.25	1.28±0.31	0.014
R²	0.92±0.06	0.91±0.07	0.863

Data are mean±standard deviation. Bold values represent p < 0.05. CN, cognitively normal subjects; MCI, mild cognitive impairment; EtCO₂, end-tidal CO₂; SaO₂, arterial blood oxygen saturation; bpm, beats per minute; MAP, mean arterial pressure; CBFV, cerebral blood flow velocity; CVCi, cerebrovascular conductance index; R², coefficient of determination.

Relationship between hypo- and hypercapnic CVMR

The negative correlations were observed between hypo- and hypercapnic CVMR, as measured by both CBFV% and CVCi% versus ΔEtCO₂ (Fig. 3). These correlations indicate that greater reductions in CBFV% or CVCi% during hypocapnia are associated with smaller increases in CBFV% or CVCi% during hypercapnia.

Fig.3

Simple linear correlations between hypo- and hypercapnic cerebral vasomotor reactivity (CVMR) across all subjects. CVMRs were calculated from the slope of cerebral blood flow velocity (CBFV%) versus end-tidal CO₂ (mmHg) (ΔCBFV/ΔEtCO₂) (upper panel) and cerebrovascular conductance index (CVCi%) versus end-tidal CO₂ (mmHg) (ΔCVCi/ΔEtCO₂) (lower panel).

Relations between CVMR and cognitive performance

Episodic memory, as reflected by LM Immediate and Delayed Recall and CVLT-II long delayed free recall scores, were positively and significantly correlated with hypocapnic CVMR and negatively with hypercapnic CVMR (Fig. 4 and Supplementary Table 1). The correlations between LM and CVLT-II memory scores and CVMR remained significant after controlling for age, sex, and education (p < 0.05) (Supplementary Table 2). Similar correlations were observed between CVLT-II total and D-KEFS Trail Making Test scores and CVMR (Supplementary Table 1 and 2).

Fig.4

Simple linear correlations between hypo- and hypercapnic cerebral vasomotor reactivity (CVMR) and logical memory (LM) immediate and delayed recall scores. CVMR was calculated from the slope of cerebrovascular conductance index (CVCi%) versus end-tidal CO₂ (mmHg) (ΔCVCi/ΔEtCO₂).

DISCUSSION

Our main findings are as follows. First, patients with amnestic MCI had lower hypocapnic, but higher hypercapnic CVMRs than CN older adults. Second, hypo- and hypercapnic CVMRs were inversely correlated to each other across all subjects. Third, BP responses to hypercapnia was enhanced in patients with amnesic MCI when compared to CN older adults. Finally, lower hypocapnic CVMR and higher hypercapnic CVMR were associated with lower scores on measures of episodic memory and aspects of executive function. Collectively, these findings suggest the presence of cerebrovascular dysfunction in older adults who are at high risks for AD.

Resting CBF and CVMR

There is a growing interest to understand the vascular contributions to AD [3 –5]. Both regional and global CBF reductions at rest have been reported in AD [25]. Findings of CBF at rest in MCI are inconclusive, likely due to the heterogeneity of the study population and relatively small sample size in previous studies [3]. In the present study, we found that mean CBFV measured in the MCA at rest was similar between amnestic MCI and CN subjects, consistent with some [26, 27], but not other studies [28 –30]. This inconsistency in CBFV at rest may be related to the severity of cognitive impairment of study participants. For example, Sun et al. (2007) and Roher et al. (2011) reported a lower CBFV in MCI patients with a group mean MMSE scores < 27 [29, 30]. In contrast, similar CBFV was observed between MCI patients and normal subjects with group mean MMSE scores≥29 [26]. However, it must be acknowledged that measurement of CBFV does not equal to volumetric CBF [2]. In this regard, global volumetric CBF measured with phase-contrast MRI or color-corded duplex ultrasonography, normalized by brain size, was reduced in MCI patients [31, 32], which may be related to brain hypometabolism or cerebral vasoconstriction, or both [33].

In the present study, hypocapnic CVMR was lower and hypercapnic CVMR was higher in amnestic MCI than CN subjects. Previous studies using TCD also reported lower hypocapnic CVMR in patients with AD or vascular dementia [34], but not in MCI [35]. Differences in the methods used to quantify CVMR and relatively small sample size in these studies may explain this discrepancy.

In contrast to previous studies which showed either a lower or unchanged hypercapnic CVMR in MCI when compared with normal subjects [5–7 , 36], we observed that hypercapnic CVMR actually was higher in patients with amnestic MCI. As discussed above, differences in the methods used to measure CVMR as well as the heterogeneity of MCI population may have led to these inconsistent findings [37, 38]. Further studies are needed to better understand these confounding factors in the assessment of CVMR for its usefulness in the study of AD.

The vascular mechanisms for the observed lower hypocapnic CVMR (vasoconstriction) and higher hypercapnic CVMR (vasodilation) in patients with amnestic MCI can only be speculated. First, in patients with amnestic MCI, the presence of AD pathology (i.e., amyloid-β and hyperphosphorylated tau) may lead to cerebral endothelial dysfunction, pericyte-mediated capillary vasoconstriction, blood vessel wall smooth muscle degeneration, and increases in basal cerebrovascular tone [39]. Cerebral vasoconstriction associated with AD pathology at rest, thus, may reduce further vasoconstriction during hypocapnia, but increase the extent of cerebral vasodilation during hypercapnia due to an increase in the vasodilatory reserve associated with the changes of basal cerebrovascular tone. The observed inverse relationship between hypo- and hypercapnic CVMR is consistent the hypothesis that cerebral vasoconstriction and increases in basal cerebrovascular tone may reduce hypocapnic, but augment hypercapnic CVMR [20]. Second, potential differences in the extent of cerebral small vessel disease as assessed by MRI white matter hyperintensities may lead to differences in CVMR (39). Third, the greater increases in arterial pressure during hypercapnia in amnestic MCI observed in this study may lead to a larger increase in CBF than that observed in normal subjects, particularly under conditions of impaired cerebral autoregulation in patients with MCI or AD [39 –41]. Finally, altered autonomic neural regulation of CBF during hypo- or hypercapnia may contribute the differences in CVMR between MCI and CN subjects [10 , 43].

CVMR and cognitive performance

Altered CVMR has been associated with cognitive performance in some, but not other studies [4, 6]. For instance, Richiardi et al. (2015) [8] and Cantin et al. (2011) [7] measured CVMR in AD, amnestic MCI, and healthy control subjects using BOLD MRI and inhalation of CO₂. They observed that altered CVMR in multiple brain regions was correlated with MMSE scores. In addition, Shim et al. using TCD and a breath-hold method, observed that reduced CVMR was associated with lower MMSE scores [6]. However, Glodzik et al. (2011), using arterial spin labeling MRI and a rebreathing method to induce hypercapnia, did not find associations between CVMR and MMSE in MCI and health control subjects [44]. In the present study, we did not observe significant correlations between CVMR and MMSE scores. However, both lower hypo- and higher hypercapnic CVMR were significantly correlated with lower memory scores, and similar correlations were also observed between CVMR and D-KEFS Trail Making scores. These findings suggest that altered CVMR may play a role in the development of cognitive impairment in older adults.

Study Strengths and limitations

The strengths of the present study are that our participants were carefully screened and demographically similar between the amnestic MCI and CN groups. This allowed us to examine their group differences by balancing out the effects of potential cofounding factors (e.g., hypertension) on CVMR. In addition, both hypo- and hypercapnic stimuli were used with simultaneous BP measurements to reveal asymmetrical CBF responses to changes in CO₂ after accounting for changes in BP [12]. Finally, we used MMSE and standard neuropsychological measures of memory and executive function to assess both global and domain- specific cognitive performance. The findings that altered CVMR was not correlated with MMSE, but correlated with sensitive memory and executive function tests suggest a link between cerebrovascular dysfunction and domain specific cognitive impairment (particularly episodic memory) in older adults.

The major limitations are the cross-sectional design, which does not allow for inferences regarding causal relationships between MCI and CVMR. Second, changes in CBFV reflect changes in CBF only if the insonated MCA diameter remained relatively constant under hypo- and hypercapnic conditions. The direct measurement of the MCA diameter during moderate changes in arterial pressure and EtCO₂ during craniotomy did not show significant changes [45]. In contrast, recent studies using MR angiographic studies showed that MCA may dilate during moderate hypercapnia (∼15 mmHg) which would suggest a potential underestimation of CVMR assessed by TCD in this study [46]. However, the magnitude of changes in EtCO₂ during both hyperventilation and rebreathing were similar in both groups; therefore, the hypo- and hypercapnic effects on the MCA, if they did exist, were likely to be similar in both groups, which should reduce potential effects of changes in the MCA diameter during CO₂ stimuli on the conclusion of this study. Third, it should be noted that MCI is a heterogeneous disease entity and the vascular mechanisms underlying AD pathology (amyloid-β and hyperphosphorylated tau), cerebral small vessel disease, and CVMR need to be elucidated in future studies.

Conclusions

We found that patients with amnestic MCI had lower hypocapnic, but higher hypercapnic CVMR than that observed in CN subjects, and that lower hypocapnic and higher CVMR were associated with lower episodic memory and aspects of executive function performance. These findings suggest the presence of cerebrovascular dysfunction and its potential impact on cognitive impairment in older adults who have high risks for AD.

Footnotes

ACKNOWLEDGMENTS

We thank all our study participants for their willingness, time, and effort devoted to this study. The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by NIH R01AG033106, R01HL102457, and NIH P30AG012300-19.

Authors’ disclosures available online ().

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

References

Kety

, Schmidt

(1948) The effects of altered arterial tensions of carbon dioxide and oxygen on cerebral blood flow and cerebral oxygen consumption of normal young men. J Clin Invest 27, 484–492.

Willie

, Colino

, Bailey

, Tzeng

, Binsted

, Jones

, Haykowsky

, Bellapart

, Ogoh

, Smith

, Smirl

, Day

, Lucas

, Eller

, Ainslie

(2011) Utility of transcranial Doppler ultrasound for the integrative assessment of cerebrovascular function. J Neurosci Methods 196, 221–237.

Beishon

, Haunton

, Panerai

, Robinson

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

Catchlove

, Pipingas

, Hughes

, Macpherson

(2018) Magnetic resonance imaging for assessment of cerebrovascular reactivity and its relationship to cognition: A systematic review. BMC Neurosci 19, 21.

Glodzik

, Randall

, Rusinek

, de Leon

(2013) Cerebrovascular reactivity to carbon dioxide in Alzheimer’s disease. J Alzheimers Dis 35, 427–440.

Shim

, Yoon

, Shim

, Kim

, An

, Yang

(2015) Cognitive correlates of cerebral vasoreactivity on transcranial Doppler in older adults. J Stroke Cerebrovasc Dis 24, 1262–1269.

Cantin

, Villien

, Moreaud

, Tropres

, Keignart

, Chipon

, Le Bas

, Warnking

, Krainik

(2011) Impaired cerebral vasoreactivity to CO2 in Alzheimer’s disease using BOLD fMRI. Neuroimage 58, 579–587.

Richiardi

, Monsch

, Haas

, Barkhof

, Van de Ville

, Radu

, Kressig

, Haller

(2015) Altered cerebrovascular reactivity velocity in mild cognitive impairment and Alzheimer’s disease. Neurobiol Aging 36, 33–41.

Brothers

, Lucas

, Zhu

, Crandall

, Zhang

(2014) Cerebral vasomotor reactivity: Steady-state versus transient changes in carbon dioxide tension. Exp Physiol 99, 1499–1510.

10.

Claassen

, Zhang

, Fu

, Witkowski

, Levine

(2007) Transcranial Doppler estimation of cerebral blood flow and cerebrovascular conductance during modified rebreathing. J Appl Physiol (1985) 102, 870–877.

11.

Cullen

, Eger

2nd (1974) Cardiovascular effects of carbon dioxide in man. Anesthesiology 41, 345–349.

12.

Tomoto

, Riley

, Turner

, Zhang

, Tarumi

(2020) Cerebral vasomotor reactivity during hypo- and hypercapnia across the adult lifespan. J Cereb Blood Flow Metab 40, 600–610.

13.

Petersen

, Doody

, Kurz

, Mohs

, Morris

, Rabins

, Ritchie

, Rossor

, Thal

, Winblad

(2001) Current concepts in mild cognitive impairment. Arch Neurol 58, 1985–1992.

14.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–308.

15.

Tarumi

, Harris

, Hill

, German

, Riley

, Turner

, Womack

, Kerwin

, Monson

, Stowe

, Mathews

, Cullum

, Zhang

(2015) Amyloid burden and sleep blood pressure in amnestic mild cognitive impairment. Neurology 85, 1922–1929.

16.

Tarumi

, Rossetti

, Thomas

, Harris

, Tseng

, Turner

, Wang

, German

, Martin-Cook

, Stowe

, Womack

, Mathews

, Kerwin

, Hynan

, Diaz-Arrastia

, Lu

, Cullum

, Zhang

(2019) Exercise training in amnestic mild cognitive impairment: A one-year randomized controlled trial. J Alzheimers Dis 71, 421–433.

17.

Woods

, Delis

, Scott

, Kramer

, Holdnack

(2006) The California Verbal Learning Test–second edition: Test-retest reliability, practice effects, and reliable change indices for the standard and alternate forms. Arch Clin Neuropsychol 21, 413–420.

18.

Delis

, Kramer

, Kaplan

, Holdnack

(2004) Reliability and validity of the Delis-Kaplan Executive Function System: An update. J Int Neuropsychol Soc 10, 301–303.

19.

Tarumi

, Zhang

(2015) The role of exercise-induced cardiovascular adaptation in brain health. Exerc Sport Sci Rev 43, 181–189.

20.

Zhu

, Tarumi

, Tseng

, Palmer

, Levine

, Zhang

(2013) Cerebral vasomotor reactivity during hypo- and hypercapnia in sedentary elderly and Masters athletes. J Cereb Blood Flow Metab 33, 1190–1196.

21.

Giller

, Giller

(1997) A new method for fixation of probes for transcranial Doppler ultrasound. J Neuroimaging 7, 103–105.

22.

Razi

, Moosavi

, Omidi

, Khakpour Saebi

, Razi

(2012) Correlation of end-tidal carbon dioxide with arterial carbon dioxide in mechanically ventilated patients. Arch Trauma Res 1, 58–62.

23.

Bishop

, Powell

, Rutt

, Browse

(1986) Transcranial Doppler measurement of middle cerebral artery blood flow velocity: A validation study. Stroke 17, 913–915.

24.

Hancock

, Mahajan

, Athanassiou

(2003) Noninvasive estimation of cerebral perfusion pressure and zero flow pressure in healthy volunteers: The effects of changes in end-tidal carbon dioxide. Anesth Analg 96, 847–851.

25.

Austin

, Nair

, Meier

, Xu

, Rowley

, Carlsson

, Johnson

, Prabhakaran

(2011) Effects of hypoperfusion in Alzheimer’s disease. J Alzheimers Dis 26(Suppl 3), 123–133.

26.

Tarumi

, Dunsky

, Khan

, Liu

, Hill

, Armstrong

, Martin-Cook

, Cullum

, Zhang

(2014) Dynamic cerebral autoregulation and tissue oxygenation in amnestic mild cognitive impairment. J Alzheimers Dis 41, 765–778.

27.

Viola

, Viola

, Buongarzone

, Fiorelli

, Litterio

(2013) Tissue oxygen saturation and pulsatility index as markers for amnestic mild cognitive impairment: NIRS and TCD study. Clin Neurophysiol 124, 851–856.

28.

Gommer

, Martens

, Aalten

, Shijaku

, Verhey

, Mess

, Ramakers

, Reulen

(2012) Dynamic cerebral autoregulation in subjects with Alzheimer’s disease, mild cognitive impairment, and controls: Evidence for increased peripheral vascular resistance with possible predictive value. J Alzheimers Dis 30, 805–813.

29.

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.

30.

Sun

, Zhu

, Liu

, Zhu

, Zhou

, Liu

(2007) Decreased cerebral blood flow velocity in apolipoprotein E epsilon4 allele carriers with mild cognitive impairment. Eur J Neurol 14, 150–155.

31.

de Eulate

, Goni

, Galiano

, Vidorreta

, Recio

, Riverol

, Zubieta

, Fernandez-Seara

(2017) Reduced cerebral blood flow in mild cognitive impairment assessed using phase-contrast MRI. J Alzheimers Dis 58, 585–595.

32.

Liu

, Zhu

, Khan

, Brunk

, Martin-Cook

, Weiner

, Cullum

, Lu

, Levine

, Diaz-Arrastia

, Zhang

(2014) Global brain hypoperfusion and oxygenation in amnestic mild cognitive impairment. Alzheimers Dement 10, 162–170.

33.

Dai

, Lopez

, Carmichael

, Becker

, Kuller

, Gach

(2009) Mild cognitive impairment and alzheimer disease: Patterns of altered cerebral blood flow at MR imaging. Radiology 250, 856–866.

34.

Provinciali

, Minciotti

, Ceravolo

, Angeleri

, Sanguinetti

(1990) Transcranial Doppler sonography as a diagnostic tool in vascular dementia. Eur Neurol 30, 98–103.

35.

Anzola

, Galluzzi

, Mazzucco

, Frisoni

(2011) Autonomic dysfunction in mild cognitive impairment: A transcranial Doppler study. Acta Neurol Scand 124, 403–409.

36.

Thomas

, Sheng

, Tseng

, Tarumi

, Martin-Cook

, Womack

, Cullum

, Levine

, Zhang

, Lu

(2017) Reduced global brain metabolism but maintained vascular function in amnestic mild cognitive impairment. J Cereb Blood Flow Metab 37, 1508–1516.

37.

Fierstra

, Sobczyk

, Battisti-Charbonney

, Mandell

, Poublanc

, Crawley

, Mikulis

, Duffin

, Fisher

(2013) Measuring cerebrovascular reactivity: What stimulus to use? J Physiol 591, 5809–5821.

38.

van Beek

, de Wit

, Olde Rikkert

, Claassen

(2011) Incorrect performance of the breath hold method in the old underestimates cerebrovascular reactivity and goes unnoticed without concomitant blood pressure and end-tidal CO(2) registration. J Neuroimaging 21, 340–347.

39.

Iadecola

(2004) Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat Rev Neurosci 5, 347–360.

40.

Claassen

, Zhang

(2011) Cerebral autoregulation in Alzheimer’s disease. J Cereb Blood Flow Metab 31, 1572–1577.

41.

Liu

, Tseng

, Khan

, Tarumi

, Hill

, Mirshams

, Hodics

, Hynan

, Zhang

(2016) Individual variability of cerebral autoregulation, posterior cerebral circulation and white matter hyperintensity. J Physiol 594, 3141–3155.

42.

Mohan

, Amara

, Cunningham

, Duffin

(1999) Measuring central-chemoreflex sensitivity in man: Rebreathing and steady-state methods compared. Respir Physiol 115, 23–33.

43.

Shoemaker

, Vovk

, Cunningham

(2002) Peripheral chemoreceptor contributions to sympathetic and cardiovascular responses during hypercapnia. Can J Physiol Pharmacol 80, 1136–1144.

44.

Glodzik

, Rusinek

, Brys

, Tsui

, Switalski

, Mosconi

, Mistur

, Pirraglia

, de Santi

, Li

, Goldowsky

, de Leon

(2011) Framingham cardiovascular risk profile correlates with impaired hippocampal and cortical vasoreactivity to hypercapnia. J Cereb Blood Flow Metab 31, 671–679.

45.

Giller

, Bowman

, Dyer

, Mootz

, Krippner

(1993) Cerebral arterial diameters during changes in blood pressure and carbon dioxide during craniotomy. Neurosurgery 32, 741–732.

46.

Verbree

, Bronzwaer

, Ghariq

, Versluis

, Daemen

, van Buchem

, Dahan

, van Lieshout

, van Osch

(2014) Assessment of middle cerebral artery diameter during hypocapnia and hypercapnia in humans using ultra-high-field MRI. J Appl Physiol 117, 1084–1089.