Carotid Artery Stiffness and Hemodynamic Pulsatility During Cognitive Engagement in Healthy Adults: A Pilot Investigation

Abstract

BACKGROUND

The matching of vascular supply to neuronal metabolic demand during cognitive engagement is known as neurovascular coupling (NVC). Arterial stiffness is a prominent determinant of pulsatility in the systemic circulation and may thus indirectly impact NVC. In this pilot investigation, we explored changes in carotid artery stiffness and cerebrovascular hemodynamic pulsatiltiy during cognitive engagement in healthy adults.

METHODS

Twenty-seven adults (age 39±3 years, BMI 24±1kg/m²) underwent Doppler ultrasonography of the common carotid artery (CCA) combined with applanation tonometry to derive (i) CCA elastic modulus (Ep) and β-stiffness index; (ii) CCA flow pulsatility index (PI); (iii) CCA pulse pressure, (iv) CCA augmentation index (AIx). Cerebral PI was assessed using transcranial Doppler at the middle cerebral artery (MCA). All measures were made at rest and during an incongruent Stroop task.

RESULTS

CCA PI was reduced (1.75±0.06 to 1.57±0.06, P < 0.05) while MCA PI was unchanged (0.75±0.02 to 0.75±0.02, P > 0.05) during Stroop. Brachial pulse pressure increased during Stroop (43±1 to 46±1mm Hg, P < 0.05) while CCA pulse pressure was unchanged (36±1 to 35±1mm Hg, P > 0.05). Similarly, CCA Ep (54.5±5.5 to 53.8 ± 4.9 kPa, P > 0.05) and β-stiffness index (4.4±0.4 to 4.2±0.3 aU, P > 0.05) were unchanged. CCA AIx increased (1±4 to 13±4%, P < 0.05).

CONCLUSION

Carotid pressure pulsatility is unaltered while carotid flow pulsatility is reduced during cognitive engagement. Carotid artery stiffness does not change suggesting that factors other than the dynamic elastic properties of the CCA buffer cerebrovascular hemodynamic pulsatility during cognitive engagement.

The matching of cerebral oxygen supply to neuronal metabolic demand, also known as neurovascular coupling (NVC), is a significant determinant of cognitive performance.¹ The brain is an obligate high flow, low impedance organ and is therefore particularly susceptible to hemodynamic pulsatility.² Excessive flow pulsatility may damage or impair function of small vessels in the brain, predisposing to microvascular hypoperfusion, ischemia, and focal infarcts.^3–8 Central artery stiffness (i.e., aorta or carotid) directly moderates transmission of pulsatile pressure and flow into the cerebral circulation^5,9 and may thus play an important role in affecting NVC.^9,10 Indeed recent cross-sectional studies note strong associations between central artery stiffness, hemodynamic pulsatility, cerebral perfusion, cerebrovascular damage (i.e., white matter hyperintensities, β-amyloid plaque deposition), and cognitive dysfunction.^{2–9,11–19} Ultimately, this hemodynamic sequela may have implications for future risk of dementia and Alzheimer’s disease.²⁰

Arterial stiffness can be altered acutely by any factor that affects vascular function (i.e., vascular smooth muscle tone). Mental stress with concomitant cognitive engagement and sympathoexcitation is known to increase blood pressure.²¹ In theory, the pressor response would affect vessel wall load bearing by transferring stress from elastin fibers to stiffer collagen fibers. Sympathoexcitation may also increase vascular smooth muscle tone, functionally stiffening the vessel.²² As such, an inappropriate hemodynamic and vascular response to cognitive challenge may have a detrimental effect on NVC; acute increases in arterial stiffness may increase cerebrovascular pulsatility and reduce cognitive performance. Cross-sectional studies note that higher hemodynamic reactivity to cognitive challenge is associated with carotid intima-media thickness,²³ cerebrovascular damage, and cognitive decline.^24–26

Changes in arterial stiffness during cognitive engagement have not been fully elucidated, in part due to inherent constraints of research methodology. Many cognitive tasks include a verbal component. Acoustic interference and movement artifact associated with talking severely compromises measures of blood pressure, blood flow, and arterial stiffness taken at the carotid artery. Our lab recently developed a technique that enables the simultaneous assessment of carotid blood pressure, carotid and cerebral blood flow, and carotid artery stiffness during cognitive engagement. Participants remain in the supine position and there is no verbal component. Using this highly novel approach, the purpose of this pilot study was to examine carotid artery stiffness and cerebrovascular hemodynamic pulsatility during cognitive engagement in healthy adults. A secondary purpose was to examine the reproducibility of this approach to elicit consistent changes in central hemodynamics during cognitive perturbation.

METHODS

Twenty-seven healthy adults between the ages of 18 and 58 were recruited to participate in this study. Exclusion criteria were history of smoking, stroke, hypertension, diabetes mellitus, hyperlipidemia, obesity (body mass index (BMI) ≥ 35kg/m²), pulmonary disease, renal disease, neurological disease, recent head trauma resulting in concussion or loss of consciousness, depressive symptom score > 18 signifying high depressive symptomology, Montreal Cognitive Assessment score < 24 signifying cognitive impairment, severe arrhythmia, peripheral artery disease, use of vasoactive medications of any kind, and color blindness. This study was approved by the Institutional Review Board of Syracuse University. All participants provided written informed consent.

Study design

All participants reported to the Human Performance Laboratory on two separate occasions for screening (visit 1) and vascular hemodynamic assessment (visit 2). A subset of participants (n = 10) returned a third day for additional vascular and hemodynamic assessment that included a time control condition (visit 3) described below. The initial visit 1 was carried out in the early morning (06:00–09:00) following an overnight fast. Following consent, participants completed a health history questionnaire, International Physical Activity Questionnaire (IPAQ), visual acuity test, Ishihara color-blindness test, body composition assessment (air displacement plethysmography, BodPod; COSMED, Concord, CA), urinalysis (Clinitek Status+ Analyzer, Siemans, IL), fasting glucose and lipid assessment via finger stick (Cholestech LDX, Alere, Waltham, MA), depressive symptomology appraisal via the Center for Epidemiologic Studies Depression Scale (CES-D) and a global cognitive appraisal using the Montreal Cognitive Assessment (MOCA), trails A, and trails B. The difference between trials B performance time and trails A performance time (trails B–trails A) was taken as a measure of executive function and vascular brain senescence.³

Immediately following the screening visit 1, participants were outfitted with an accelerometer (Actigraph Model 7164 Manufacturing Technology, FL) on the right hip and instructed to wear the device during all waking hours (at least 10 hours) for 5 consecutive days (including at least 1 weekend day), except during swimming or bathing. The monitors were set at 60-second epochs and total movement counts were calculated for each participant based on data from 3 valid days (defined as >10 hours of wear with <1 hours of no accelerometer data (i.e., continuous zeros).

Participants reported back to the Human Performance Lab approximately 1 week after screening for vascular and hemodynamic data acquisition (visit 2). Participants returned the accelerometer during this visit. Vascular hemodynamic testing was conducted at a consistent time of day (13:00–15:00) in a quiet, dimly lit, temperature-controlled laboratory. Participants were instructed to fast for a minimum of 4 hours and avoid vigorous exercise and consumption of caffeine or alcohol for a minimum of 12 hours prior to testing. Baseline measures of blood pressure (brachial and carotid), blood flow (carotid and cerebral), arterial stiffness, and a salivary cortisol were obtained following a 15-minute supine rest. All hemodynamic variables were then simultaneously measured (in triplicate) during two separate 4 minutes blocks of a cognitive perturbation (incongruent Stroop task); 4 minutes of recovery time was afforded between incongruent Stroop trials and no vascular measures occurred during this recovery epoch (quiet rest).

A third visit was conducted in a subset of participants (n = 10) and occurred within 1 week of visit 2. Visit 3 was designed to recapitulate visit 2 with the exception that one incongruent Stroop task was replaced with a neutral time control condition. Participants reported to the lab in a similar fasting state as described above. Following a 15-minute supine rest period and baseline vascular hemodynamic measures (collected in the same fashion as during visit 2), participants underwent a “time control” condition which consisted of viewing a black screen for 4 minutes. This was then followed by a 4-minute recovery period and finally a 4-minute incongruent Stroop task. All hemodynamic variables were simultaneously measured (in triplicate) during the black screen condition and the incongruent Stroop task. No vascular measures occurred during the recovery epoch (quiet rest).

Brachial blood pressure

Systolic blood pressure (SBP), diastolic brachial blood pressure (DBP), and mean arterial pressure (MAP) was measured on the left arm using a validated, automated oscillometric cuff (EW3109, Panasonic Electric Works, Secaucus, NJ).

Carotid blood pressure

Pressure waveforms were obtained in the right carotid artery using applanation tonometry from a 10-second epoch (SphygmoCor, AtCor Medical, Sydney, Australia). Carotid pressure waveforms were ensemble averaged to a single waveform and calibrated to brachial MAP and DBP. Pulse pressure (PP) was calculated as SBP minus DBP.

Carotid blood flow and arterial stiffness

Images of the left common carotid artery (CCA) were obtained using Doppler ultrasound (ProSound α7, Aloka, Tokyo, Japan) and 7.5–10.0 MHz linear-array probe. Images were acquired 5–10mm below the carotid bulb. The distance from the near wall to far wall lumen–intima interface was continuously traced using eTracking software and used to determine CCA diameters. Flow velocity waveforms were measured using range gated spectral Doppler signals averaged along the Doppler beam. An insonation angle ≤ 60° was maintained for all measures and sample volume was manually adjusted to encompass the entire vessel. CCA PI was calculated as: (V _s − V _d)/V _m, where V _s is the peak systolic velocity, V _d diastolic velocity, and V _m the mean velocity. CCA volume flow rate per minute was calculated as π × (mean radius)² × V _m × 60. CCA mean shear rate was calculated as 4 × (V _m/mean diameter). Arterial stiffness measures included beta stiffness index (β) and Peterson’s pressure–strain elastic modulus (Ep):

$$\begin{array}{l}\beta =\text{ln}(P_{\text{Max}}/P_{\text{Min}})/[(D_{\text{Max}}-D_{\text{Min}})/D_{\text{Min}}]\\ \text{Ep}=(P_{\text{Max}}-P_{\text{Min}})/[({\text{D}}_{\text{Max}}-D_{\text{Min}})/D_{\text{Min}}]\end{array}$$

where P and D correspond to pressure and diameter, respectively, and Max and Min refer to maximum (systolic) and minimum (diastolic) values during the cardiac cycle. Augmentation index (AIx) was calculated as augmented pressure (i.e., the difference between the early [P₁] and late [P₂] systolic peaks of both the carotid pressure and distension waveforms) relative to the total PP expressed as a percentage (augmented pressure/PP × 100) and taken as a proxy of global wave reflections.

Cerebral blood flow velocity

Middle cerebral artery (MCA) blood flow velocity was assessed using a 2-MHz transcranial Doppler (TCD) ultrasound probe (DWL Doppler Box-X, Compumedics, Germany) applied to the left temporal window. Mean cerebral blood flow velocity (CBFv) and PI were measured at insonation depths of 45–60mm. MCA PI was calculated via an automated waveform tracking function using the same equation described for CCA PI.

Cognitive challenge protocol

A modified incongruent Stroop color-word interference task was used to elicit changes in central hemodynamics. The Stroop test is one of the most widely utilized neurocognitive challenges in the field of psychophysiology used to elicit reproducible and reliable changes in neural substrates of executive function.^27–30 Participants remained in the supine position. A specialized wall mount was used to horizontally suspend a 42-inch flat screen television over the participant. The television was approximately 40-inches above the participant as per manufacturer suggestions. Font was focally displayed on a black background at a font size of 3-cm which is comparable to the size of font displayed on a standard 17-inch computer monitor (1.5cm) with participants sitting 2.5 feet from the monitor. The television was interfaced with a laptop (Dell) and remote response clicker that ran a customized incongruent Stroop protocol (E-Prime, Psychology Software Tools, Sharpsburg, PA).

In single trials, participants were prompted to identify the color of a target word displayed centrally by selecting one of four identifier words appearing in one row on the bottom of the screen using the remote response clicker. The identifier words (1–4) corresponded to separate buttons (1–4) on the response clicker located on the exam table at the participant’s side. Buttons 1–4 corresponded to individual digits (index finger–pinky finger). For this incongruent Stroop paradigm, the target word appeared in a color incongruent with the target word (i.e., the word “Red” presented in blue font). Moreover, all identifier words also appeared in colors incongruent with the colors that the identifier words identified. Two 4-minute blocks were presented to participants with a 4-minute rest period afforded between trials. This time block was selected based on previous studies that note prominent changes in BP and HR to occur within this allotted time.³⁰ Longer time blocks may affect participant valence and cause mental fatigue, resulting in progressive tapering of cardiovascular reactivity. This time block was also sufficient in length to afford ample time to collect all hemodynamic variables in triplicate (determined during pilot testing).

In an attempt to elicit comparable hemodynamic reactivity between trials while controlling for potential individual differences in task performance, each participant’s accuracy at target word identification was titrated to ~50% by manipulating presentation times in 400ms steps (range 300 ms–5 s). For example, three consecutive correct responses prompted shorter response time windows, increasing task difficulty; less accurate performance lengthened response time windows, reducing task difficulty. This protocol has previously been used successfully to assess cardiovascular reactivity and neural activation during functional magnetic resonance imaging.³⁰ Salivary cortisol was analyzed from 25 μl of sample using a highly sensitive salivary enzyme immunoassay (Salimetrics, State College, PA) at baseline and 30-minute following the Stroop task as a marker of generalized stress responsiveness.

STATISTICAL ANALYSES

All data is reported as mean ± standard error of the mean and statistical significance was established a priori as P < 0.05. Normality of distribution for variables was assessed qualitatively using histograms and Q–Q plots as well as quantitatively using the Shapiro–Wilk test. Degree of consistency across testing trials between individuals was evaluated by the intraclass correlation coefficient (ICC). To further examine reliability of measures obtained during Stroop task, we computed the coefficient of variation (CV, standard deviation/mean) for all hemodynamic variables. Given high ICCs and low CVs, values were collapsed (averaged) across trials and used for subsequent analyses. We examined differences in central hemodynamic parameters using a priori contrasts. Homogeneity of variance was assessed with the Levene statistic, and appropriate comparisons were made (equal variance assumed or not assumed) based on results. Absolute change (Δ) scores for each hemodynamic parameter were tabulated as the average value obtained during Stroop—Baseline. A one sample t-test was used to compare the change score to a value of zero (i.e., was the change significantly different than zero). We examined associations of interest between reactivity scores using Pearson’s correlation coefficients.

Exploratory analyses examined the effect of sex and age on the hemodynamic response to Stroop. A 2×2 analysis of variance (ANOVA) with repeated measures was used to assess sex differences in the response to Stroop: 2 groups (men vs. women) by 2 time points (baseline vs. Stroop). K-means cluster analysis was used to divide participants into 2 groups, stratified by age: younger and older. A 2×2 ANOVA with repeated measures was then used to explore age-group differences in the response to Stroop: 2 groups (younger vs. older) by 2 time points (baseline vs. Stroop). Partial correlations were used to examine associations of interest adjusting for age. All statistical analyses were executed using Statistical Package for the Social Sciences (SPSS, version 21, IBM, Chicago, IL).

RESULTS

Participant characteristics are presented in Table 1. Based on aforementioned inclusion and exclusion criteria, two participants were excluded (CES-D score 34 and MOCA score 21, n = 1; BMI 38kg/m², n = 1). Brachial SBP, DBP, PP, and MAP increased during Stroop (Table 2, P < 0.05). While carotid SBP increased during Stroop (P < 0.05), carotid PP was unchanged (P > 0.05). The increase in carotid SBP was smaller than the increase in brachial SBP (Δ5mm Hg vs. Δ 9mm Hg, P < 0.05). HR increased during Stroop (P < 0.05). AIx, AP, CCA mean diameter, and CCA flow volume increased during Stroop (Table 3, P < 0.05). CCA PI and P₁ were reduced during Stroop (P < 0.05). Arterial stiffness (β-stiffness and Ep) and CCA shear rate were unchanged (P > 0.05). MCA Vmean increased during Stroop (Table 3, P < 0.05). MCA PI remain unchanged (P > 0.05). Salivary cortisol was unchanged (0.165±0.016 vs. 0.147±0.015 μg/dl, P > 0.05).

Table 1.

Participant descriptive characteristics

Variable	Mean ± SEM
Age, years	39±3
Women, n	13
Body mass index, kg/m2	24±1
Body fat, %	23±2
Total cholesterol, mg/dl	186±7
HDL cholesterol, mg/dl	69±4
Glucose, mg/dl	88±2
Urinary creatinine, mg/dl	110±18
MOCA, score	27±1
Trails A, s	20±1
Trails B, s	42±3
CES-D, score	9±1
Family history CVD, n	3
IPAQ, MET min/week	4,424±560
Accelerometer counts (n = 21)	1.37×105 ± 1.38×104

Abbreviations: CES-D, Center for Epidemiologic Studies Depression scale; CVD, cardiovascular disease; HDL, high-density lipoprotein; IPAQ, International Physical Activity Questionnaire; MOCA, Montreal Cognitive Assessment.

Table 2.

Blood pressure and heart rate at rest and during Stroop task

Variable	Baseline	Stroop	Δ Score	t (df)	P value
Brachial SBP, mm Hg	117±2	126±2a	9±1b	6.1 (24)	<0.001
Brachial DBP, mm Hg	74±1	79±1a	5±1b	6.5 (24)	<0.001
Brachial PP, mm Hg	43±1	46±1a	3±1b	3.5 (24)	0.002
Brachial MAP, mm Hg	88±1	95±1a	7±1b	6.7 (24)	<0.001
Carotid SBP, mm Hg	110±2	114±2a	4±1b	3.9 (24)	0.001
Carotid PP, mm Hg	36±1	35±1	−1±1	1.0 (24)	0.324
Heart Rate, bpm	60±2	72±3a	12±2b	6.4 (24)	<0.001

Abbreviations: DBP, diastolic blood pressure; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure.

^aSignificantly different from baseline (P < 0.05).

^bSignificantly different from zero (P < 0.05).

Table 3.

Carotid and cerebral hemodynamics during Stroop task

Variable	Baseline	Stroop	Δ Score	t (df)	P value
CCA β-Stiffness, aU	4.4±0.4	4.2±0.3	−0.2±0.2	−1.3 (23)	0.224
CCA Ep, kPa	54.5±5.5	53.8±4.9	−1.1±2.5	−0.4 (23)	0.670
CCA pressure AIx, %	1±4	13±4a	12±3b	4.6 (24)	<0.001
CCA distension AIx, %	4±2	8±2a	4±1b	4.0 (23)	0.001
CCA P1, mm Hg	34±8	29±7a	−5±1b	5.5 (24)	0.001
CCA AP, ms	1±6	4±6a	3±5b	3.4 (24)	0.102
CCA mean diameter, mm	5.62±0.13	5.74±0.13a	0.09±0.03b	2.5 (22)	0.010
CCA Vmean, cm/s	38.9±1.2	39.6±1.4	0.8±0.9	0.8 (24)	0.423
CCA shear rate, /s	28.6±1.3	28.2±1.6	0.4±0.9	0.4 (22)	0.680
CCA flow, ml/min	575.5±21.1	607.4±19.6a	31.8±17.1	1.9 (22)	0.039
CCA PI	1.75±0.06	1.57±0.06a	−0.18±0.08b	2.3 (24)	0.016
MCA Vmean, cm/sec	49.4±2.3	52.5±2.5a	3.1±1.5b	1.9 (24)	0.035
MCA PI	0.75±0.02	0.75±0.01	0.003±0.01	0.2 (24)	0.841

Abbreviations: AIx, augmentation index; AP, augmented pressure; CCA, common carotid artery; Ep, elastic modulus;P₁, pressure at the inflection point; PI, pulsatility index MCA, middle cerebral artery; Vmean, mean velocity.

^aSignificantly different from baseline (P < 0.05).

^bSignificantly different than zero (P < 0.05).

A correlation matrix is presented in Table 4. Adjusting for age had negligible effects on the associations between Stroop % Correct and: ∆ CCA PI (r = −0.38, P = 0.04), ∆ CCA PP (r = −0.36, P = 0.04) or ∆ CCA diameter (r = 0.35, P = 0.06). Additional correlations of interest are as follows. Trails B–trails A was associated with ∆ MCA PI (r = 0.41, P = 0.03), ∆ CCA PI (r = 0.49, P = 0.01), ∆ CCA PP (r = 0.32, P = 0.061), and associated negatively with ∆ CCA diameter (r = −0.56, P = 0.03). Finally, trails B–trails A was associated positively with mean reaction time for correct choices during Stroop tasks (r = 0.34, P = 0.04) and associated negatively with the mean % correct for Stroop performances (r = −0.48, P = 0.02). After adjusting for age with partial correlation, the association between trails B–trails A performance and ∆ MCA PI was attenuated (r = 0.29, P = 0.09) as was ∆ CCA PP (r = 0.27, P = 0.10). Adjusting for age had no significant effect on the association between trails B–trails A performance and ∆ CCA PI (r = 0.48, P = 0.01) or ∆ CCA diameter (r = −0.40, P = 0.04).

Table 4.

Correlation matrix of hemodynamic reactivity scores with Stroop performance (n = 25 participants)

	Stroop %Correct	Stroop RT	Δ CCA PI	Δ MCA PI	Δ CCA PP	Δ CCA P1	Δ CCA AP
Stroop RT	−0.72 *
Δ CCA PI	−0.34 *	0.35 *
Δ MCA PI	−0.21	−0.03	0.70 *
Δ CCA PP	−0.45 *	0.23	0.34 *	0.19
Δ CCA P1	−0.24	0.24	0.39 *	0.11	0.64 *
Δ CCA AP	−0.07	0.08	−0.15	−0.07	0.17	−0.26
Δ CCA diameter	0.35 *	−0.48 *	−0.57 *	−0.42 *	−0.41 *	−0.55 *	0.04

Abbreviations: AP, augmented pressure; CCA, common carotid artery; MCA, middle cerebral artery; PP, pulse pressure; P₁, pressure at the inflection point;PI, pulsatility index; RT, reaction time.

Significant univariate association (^* P < 0.05).

Overall ICCs were high (0.90–0.99) and CVs low (0.88–8.53) suggesting good reproducibility for all hemodynamic measures (See Supplementary Table S1). There was a slight but significant increase in brachial SBP, carotid SBP and MAP during the time control condition (P < 0.05; See Supplementary Table S2). No other parameter changed during the time control (See Supplementary Table S2, P > 0.05).

Demographic data stratified by age and sex are presented in Supplementary Table S3. There were no sex differences in the vascular hemodynamic response to Stroop (See Supplementary Table S4, P > 0.05 for all). There were no differences in the vascular hemodynamic response to Stroop when comparing younger adults to older adults within our group of participants (See Supplementary Table S5, P > 0.05 for all).

DISCUSSION

Changes in cerebral blood flow are tightly coupled to level of cognitive perturbation³¹ such that cerebral flow increases with increasing neuronal metabolic demand.^32,33 In this pilot study, during increased cognitive demand associated with an incongruent color-word interference task, there were increases in both CCA mean flow and MCA flow velocity (i.e., a surrogate for increased inflow volume) consistent with the premise of NVC. The novel findings of the present pilot study were as follows: (i) increases in flow volume occurred concomitant with reductions in CCA flow pulsatility and no change in MCA flow pulsatility; (ii) although there was an increase in peripheral pressure pulsatility, there was no change in CCA pressure pulsatility; (iii) there was no change in CCA stiffness during cognitive engagement. Our preliminary findings indicate that transmission of both pressure and flow pulsatility into the brain is moderated during cognitive engagement despite an increase in mean flow volume. Changes in CCA pulsatility occur without changes in CCA stiffness per se suggesting that mechanisms other than the dynamic elastic properties of the CCA buffer central hemodynamic pulsatility during cognitive challenge.

Overall, our preliminary findings did not support our hypothesis as cognitive perturbation (Stroop task) did not alter carotid artery stiffness. The response to cognitive perturbation in healthy adults may be vascular bed specific. Muscular arteries such as the radial artery may increase in stiffness during mental stress²² while elastic arteries such as the CCA do not change. This may ensure optimal transfer of flow into the cerebral circulation. CCA PI was reduced with no change in MCA PI during cognitive engagement. Eighty percent of common carotid blood flow feeds the internal carotid artery with subsequent branching giving rise to the MCA which supplies 80% of the blood supply to the brain.³⁴ We recently reported an association between resting carotid PI and MCA PI in healthy adults.³⁵ In the current pilot study, we noted that change in carotid PI during Stroop was associated with MCA PI suggesting that for ideal NVC and preservation of the cerebral hemodynamic milieu, a reduction in CCA flow pulsatility may be necessary.

We noted significant increases in CCA diameter during cognitive engagement and this is consistent with previous findings.³⁶ Increases in common carotid vasodilatory capacity during cognitive challenge contribute to increases in carotid and MCA flow.³⁶ Changes in CCA diameter during cognitive engagement reflect carotid endothelial function³⁶ and may be due to β-adrenergically mediated vasodilation, which is dependent on nitric oxide mediated smooth muscle relaxation.^37,38 Endothelial dysfunction affects blood flow into the cerebral vascular bed,³⁹ preventing efficient oxygen extraction at the capillary level⁴⁰ and ultimately reducing neurocognitive capacity. Inhibition of nitric oxide has been shown to reduce cerebral blood flow⁴¹ without affecting carotid artery stiffness.⁴² There were negative associations between carotid dilation and cerebrovascular pressure and flow pulsatility in this study. Therefore, carotid vasodilation during cognitive challenge may help to favorably buffer cerebrovascular pulsatility independent of modulation of carotid stiffness, contributing to optimal NVC.

There was a divergent effect of cognitive perturbation on peripheral (i.e., brachial) vs. central (i.e., carotid) pressure pulsatility. While pulse pressure increased in the peripheral vascular bed, there was no change in the central vascular bed. This occurred concomitant with a paradoxical increase in carotid AIx. The AIx quantifies the interaction between forward and reflected travelling pressure waves in the systemic and regional circulation. An increase in AIx is often attributed to an increase in pressure from global wave reflections but may also occur with a fall in the forward travelling pressure wave. In this pilot study, there was a significant increase in augmented pressure as well as a reduction in pressure at the inflection point (P₁, a proxy of forward wave pressure) during cognitive engagement. Increases in augmented pressure from wave reflections may be due to increases in peripheral vascular resistance and stiffness that are known to occur during mental stress.²² The fall in forward wave pressure was likely due to the significant carotid dilation. It is also interesting to note that reductions in pressure at the inflection point were also associated with favorable changes in carotid pressure and flow pulsatility. These preliminary findings allude to the importance of impedance properties of the extra-cranial vascular bed as an important moderator of hemodynamic pulsatility entry into the cerebral circulation during cognitive engagement and suggest that the CCA may not simply be a passive conduit during NVC.⁴³

High cerebrovascular hemodynamic pulsatility is associated with reduced cognitive function.² Moreover, interventions that reduce cerebrovascular pulsatility have been shown to improve cognitive function.⁴⁴ In the present pilot study, there was an association between change in CCA PI and PP and performance during Stroop task (percentage of correct answers and the reaction time for correct answers), a measure of executive function.^27–30 Change in CCA diameter, CCA PI and MCA PI assessed during Stroop task were also each inversely associated with trails B–trails A performance, an additional measure of executive function and vascular brain aging.³ These preliminary findings support the possibility that CCA dilation and reductions in cerebrovascular pulsatility during cognitive challenge are associated with optimal NVC and improved cognitive performance.⁴⁵

We did not find a significant effect of age on our overall findings. Age is one of the strongest, most consistent determinants of central artery stiffness⁴⁶ which in turn plays an important role in age-related differences in cerebral hemodynamics.¹² The lack of effect of age on changes in central hemodynamics during Stroop and subsequent associations with cognitive performance noted herein is likely due to small sample size (i.e., inability to detect an effect) coupled with lack of inclusion of older adults over 60 years of age (when changes in large artery stiffness and impedance matching may more profoundly impact hemodynamic pulsatility transmission into the brain). Our population consisted of fairly healthy adults. Indeed when statistically separating our participants into a younger group and an older group, results revealed no group differences in resting BP, lipids, fasting glucose, or physical activity levels, established arbiters of arterial stiffness, cerebral reactivity and cognitive performance.^47–49 In the absence of CVD risk factors, several studies have shown maintenance of NVC with advancing age.^36,50–54 Our findings also noted no sex differences in NVC. Carotid stiffness and carotid flow may change across the menstrual cycle,^55,56 as does mental stress mediated hemodynamic reactivity⁵⁷ and cognitive function⁵⁸ arguing for possible hormonal modulation of NVC. Interestingly, acute administration of estrogen has been shown to improve arterial compliance and endothelial function,⁵⁹ attenuate hemodynamic reactivity to cognitive perturbation,⁶⁰ and improve cerebral vasoreactivity in vitro.⁶¹ Additional studies are needed to specifically explore sex- and age-associated changes in large artery stiffness and cerebrovascular hemodynamic pulsatility during cognitive perturbation.

Limitations to this pilot study should be acknowledged. TCD does not measure CBF directly. It is often argued that the diameter of the MCA does not change during various perturbations such as cognitive engagement suggesting that changes in CBF velocity may be used as a suitable proxy of cerebral volume flow changes.⁶² It is possible that noted hemodynamic changes are not due to cognitive engagement per se but may be a generalized mental stress response. Indeed there was a slight but significant increase in SBP during the time control condition. It should be noted that the increases in SBP were greater with Stroop than time control suggesting that hemodynamic change was likely due to a combination of mental stress and cognitive engagement. And the lack of change in salivary cortisol in our participants with perturbation would suggest that this protocol was not particularly stressful. This pilot study used healthy participants across a wide age range, but was not specifically designed to examine the impact of sex/gender or aging per se on hemodynamic pulsatility during cognitive engagement. Given that “typical” vascular and hemodynamic responses during cognitive engagement (as assessed using this protocol) are largely unknown, we designed this pilot study to first establish the “typical” response to cognitive perturbation in the absence of disease. We also sought to establish reproducibility of our novel method in eliciting consistent vascular and hemodynamic responses.

In conclusion, carotid pressure and flow pulsatility are buffered during cognitive engagement in healthy adults and this may be an important aspect of optimal NVC. Carotid artery stiffness does not change during cognitive engagement suggesting that factors other than the dynamic elastic properties of the CCA moderate changes in cerebrovascular hemodynamic pulsatility during NVC. Future research is needed to explore potential factors affecting central hemodynamic pulsatility during cognitive engagement such as carotid vasoreactivity and pressure from wave reflections and whether favorably modifying these factors improves NVC and cognitive function.

SUPPLEMENTARY MATERIAL

Supplementary materials are available at American Journal of Hypertension (http://ajh.oxfordjournals.org).

DISCLOSURE

We have no conflicts of interest to disclose.