Six-month longitudinal tracking of arterial stiffness and blood pressure in young adults following SARS-CoV-2 infection.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) can increase arterial stiffness 3-4 wk following infection, even among young, healthy adults. However, the long-term impacts of SARS-CoV-2 infection on cardiovascular health and the duration of recovery remain unknown. The purpose of this study was to elucidate potential long-lasting effects of SARS-CoV-2 infection on markers of arterial stiffness among young adults during the 6 mo following infection. Assessments were performed at months 1, 2, 3, 4, and ∼6 following SARS-CoV-2 infection. Doppler ultrasound was used to measure carotid-femoral pulse wave velocity (cfPWV) and carotid stiffness, and arterial tonometry was used to measure central blood pressures and aortic augmentation index at a heart rate of 75 beats·min-1 (AIx@HR75). Vascular (VCAM-1) and intracellular (ICAM-1) adhesion molecules were analyzed as circulating markers of arterial stiffness. From months 1-6, a significant reduction in cfPWV was observed (month 1: 5.70 ± 0.73 m·s-1; month 6: 4.88 ± 0.65 m·s-1; P < 0.05) without any change in carotid stiffness measures. Reductions in systolic blood pressure (month 1: 123 ± 8 mmHg; month 6: 112 ± 11 mmHg) and mean arterial pressure (MAP; month 1: 97 ± 6 mmHg; month 6: 86 ± 7 mmHg) were observed (P < 0.05), although AIx@HR75 did not change over time. The month 1-6 change in cfPWV and MAP were correlated (r = 0.894; P < 0.001). A reduction in VCAM-1 was observed at month 3 compared with month 1 (month 1: 5,575 ± 2,242 pg·mL-1; month 3: 4,636 ± 1,621 pg·mL-1; P < 0.05) without a change in ICAM-1. A reduction in cfPWV was related with MAP, and some indicators of arterial stiffness remain elevated for several months following SARS-CoV-2 infection, possibly contributing to prolonged recovery and increased cardiovascular health risks.NEW & NOTEWORTHY We sought to investigate potential long-lasting effects of SARS-CoV-2 infection on markers of arterial stiffness among young adults for 6 mo following infection. Carotid femoral pulse wave velocity was significantly reduced while carotid stiffness measures remained unaltered over the 6-mo period. These findings suggest several months of recovery from infection may be necessary for young adults to improve various markers of arterial stiffness, possibly contributing to cardiovascular health and recovery among those infected with SARS-CoV-2.

INTRODUCTION

The novel coronavirus disease of 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has resulted in over 5.7 million deaths worldwide as of February 2022, with hundreds of millions surviving the disease (1). Following the initial viral infection and inflammatory response, substantial evidence demonstrates adverse consequences to several physiological systems, particularly in the systemic vasculature (2–7), which may provoke persistent symptoms for several months following infection (8, 9). Indeed, evolving evidence suggests these impairments may continue through 6 mo among patients who were hospitalized for SARS-CoV-2 (10). These lasting effects may share characteristics among other coronaviruses, including SARS-CoV-1 (11), in which symptoms may continue for years in severe cases (12). However, much less is known regarding the long-term physiological effects among those with milder symptoms, which is clinically relevant for millions of those recovering from SARS-CoV-2 infection.

Our group has previously provided evidence of vascular dysfunction (3) and elevated arterial stiffness (2, 3) among young adults 3–4 wk following infection with SARS-CoV-2. These data included significantly attenuated responses to brachial artery flow-mediated dilation, hyperemic response to passive limb movement, and brachial artery blood flow during handgrip exercise, in addition to elevated carotid-femoral pulse wave velocity (cfPWV) in young adults with SARS-CoV-2 compared with healthy controls (3, 13). Similarly, Schnaubelt et al. (14) observed elevated cfPWV 1 mo after infection with SARS-CoV-2. Further, we identified increases in carotid stiffness, aortic augmentation index (AIx) standardized for a heart rate of 75 beats per minute (AIx@HR75), and Young’s modulus, in addition to a reduction in carotid compliance and distensibility among young adults with SARS-CoV-2 compared with healthy young adults (2). Conversely, Nandadeva et al. (15) did not observe a difference in PWV or AIx@75 among young adults with SARS-CoV-2 whose recovery time ranged from 1–5 mo when cross-sectionally compared with healthy control subjects, this time range may have confounded these results. To date, no studies have longitudinally tracked the impact of SARS-CoV-2 on arterial stiffness in young adults, nor the potential mechanisms behind deleterious functional changes.

SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor, which is prevalent throughout the heart, lungs, and blood vessels (16), causing a systemic, cytokine-induced inflammatory response (17). By binding to ACE2, SARS-CoV-2 occupies a binding site for angiotensin II, leading to vasoconstriction and increasing susceptibility to inflammation. Infection with SARS-CoV-2, therefore, causes a systemic, cytokine-induced inflammatory response (17), possibly involving adhesion molecules as inflammatory mediators (18–21). Furthermore, since patients with mild and severe cases of SARS-CoV-2 exhibited elevated levels of the atherosclerotic circulating factors (22) as well as vascular cell adhesion molecule-1 (VCAM-1) and intracellular adhesion molecule-1 (ICAM-1) (23), these adhesion molecules may be linked to changes in indicators of arterial stiffness. The appearance of these circulating adhesion molecules quite possibly may accompany arterial stiffness following infection of SARS-CoV-2.

Thus, the purpose of this investigation was to longitudinally track pulse wave velocity, arterial stiffness, and carotid distensibility, as well as circulating markers of VCAM-1 and ICAM-1, during the first 6 mo following initial infection with SARS-CoV-2 in young adults. Due to the potential for lasting symptoms among a subset of those infected with SARS-CoV-2 (8, 9), we hypothesized improvement in arterial stiffness following SARS-CoV-2 infection would require several months, as suggested by decreases in cfPWV, carotid stiffness, and AIx@HR75, along with VCAM-1 and ICAM-1.

METHODS

Subjects

All procedures were approved by the Appalachian State University Institutional Review Board (IRB_20–0304). The subjects provided written informed consent, in accordance with the standards outlined by the Declaration of Helsinki before testing. Subjects were relatively healthy young adults free from chronic metabolic, cardiovascular, and pulmonary diseases. Female subjects were premenopausal, not pregnant, or trying to become pregnant. Subjects were nonsmokers and had no orthopedic limitations. Subjects tested positive for SARS-CoV-2 using a nasopharyngeal swab polymerase chain reaction (PCR) assay 3–4 wk before study inclusion. The current investigation includes subjects from previously published studies (2, 3, 7, 13) but all data presented are new to this investigation.

Study Procedures

Subjects came into the laboratory for five total study visits at approximately 1-mo intervals for 6 mo. Before the initial testing visit, subjects completed a health history questionnaire detailing their personal and family medical history, average physical activity frequency and duration per week, as well as medication usage. Subjects arrived for all testing in a >4-h fasted state, having abstained from caffeine for 12 h and from exercise and alcohol for 24 h. All study procedures were performed in a quiet, thermoneutral environment after 20 min of supine rest. For standardization purposes, study procedures were ordered similarly between subjects.

COVID-19 Symptom Severity Survey

On the day of testing, subjects ranked their COVID-19 symptoms on a scale of 0–100 with increasing severity, which can be found elsewhere (7). Measured symptoms included chest pain, chills, diarrhea, dizziness or vertigo, dry cough, dry eyes, dry mouth, fatigue, fever over 37.9°C, headache, lack of appetite, loss of smell or taste (anosmia), muscle or body aches, nasal congestion or runny nose, nausea or vomiting, shortness of breath, difficulty breathing, dyspnea, sore joints, or sore throat. The values for all symptoms were totaled and averaged for an average symptom severity score. Mild severity scores were categorized as a score of 0–33, moderate from 34 to 66, and severe from 67 to 100.

Experimental Measurements

Pulse wave velocity.

Measurements were recorded at the femoral, radial, and carotid arteries by a single operator with a Doppler ultrasound system with ECG integration (GE Logiq eR7 and L4-12T-RS transducer, GE Medical Systems, Milwaukee, WI) and an imaging frequency of ∼12 MHz, in B-mode, to assess central arterial stiffness in the form of cfPWV and peripheral arterial stiffness as carotid-radial pulse wave velocity (crPWV). Distance measurements were obtained from the sternal notch to the base of the Doppler ultrasound probe at the femoral (Df) radial (Dr), and carotid (Dc) measurement sites using a tape measure pulled tight, ensuring not to conform to the contours of the subject’s body. Distances for femoral and radial measurements were then subtracted from the carotid distance. Both cfPWV and crPWV were calculated using the foot-to-foot ECG-gated method (24), where the time differential from the R-wave on the simultaneous recording of the ECG to the initial positive deflection of the blood velocity waveform from baseline was identified visually by a single operator using the highest resolution sweep speed (1 s spatial resolution) and temporal resolution (nearest nanosecond). An average of five separate time measurements were made at the femoral (Tf), radial (Tr), and carotid (Tc) sites and differences between carotid-femoral and carotid-radial sites were used to determine cfPWV and crPWV, respectively: cfPWV (m·s−1) = (Df – Dc)/(Tf – Tc); crPWV (m·s−1) = (Dr – Dc)/(Tr – Tc). PWV was corrected for mean arterial pressure (MAP) using the equation PWVcorr (m·s−1·mmHg−1) = PWV/MAP.

Carotid stiffness.

An automated brachial artery cuff was used to determine supine blood pressures (EVOLV, Omron Healthcare, Kyoto, Japan) before carotid stiffness measurements. Subjects remained supine for assessment of carotid stiffness using the ultrasound system over the common carotid artery, ≥1–2 cm proximal to the bifurcation, with an imaging frequency of ∼12 MHz in B-mode. Between 10 and 15 cardiac cycles were recorded and analyzed by a single operator with the Carotid Studio (Cardiovascular Suite, Quipu srl, Pisa, Italy) software. Measurements of stiffness, distensibility, compliance, Young’s modulus, carotid diameters, and carotid intima-media thickness (cIMT) were calculated as previously described (2).

Carotid pulse wave analysis.

A SphygmoCor CPv (AtCor Medical, Sydney, Australia) device was used to record carotid pulse wave analysis (PWA) via applanation tonometry during all visits, except during the final visit for three subjects due to insufficient allocation of resources. During these three visits when the SphygmoCor CPv could not be used, a SphygmoCor XCEL (AtCor Medical, Sydney, Australia) device was used. The same supine blood pressures from the automated cuff were input into the SphygmoCor CPv for PWA measures. The tonometer was positioned over the common carotid artery. After 20 sequential waveforms were recorded, the intrinsic transfer function of the SphygmoCor generated estimated central aortic blood pressures (25). From this waveform, the central hemodynamic parameters were estimated, including AIx, AIx@HR75, aortic augmentation pressure (AP), and central pressures (systolic blood pressure, SBP; diastolic blood pressure, DBP; pulse pressure, PP; and MAP), heart rate (HR) period, duration from the start of the pulse to the second systolic peak (aortic T2), pressure obtained at the first peak/shoulder (P1 height), end-systolic pressure, mean systolic pressure, and mean diastolic pressure. Mean systolic pressure was defined as the average pressure between the beginning of the waveform and the end of the first peak, whereas mean diastolic pressure was defined as the average pressure from the end of the first peak to the end of the waveform. Identification of these measures was performed as previously stated (2). Rate pressure product (RPP), an indication of myocardial oxygen consumption (26), was calculated as: RPP (A.U.) = SBP × HR.

Indices of myocardial oxygen demand and perfusion.

During PWA, the Buckberg subendocardial viability ratio (SEVR) was determined as the ratio between the diastolic pressure-time index (DPTI) and the systolic pressure-time index (SPTI), defined as the areas under the pulse pressure during diastole and systole (27, 28). Ejection duration was measured as the period, in milliseconds, from the beginning of the waveform to the incisura (29).

Endothelial cell adhesion molecules.

Plasma blood samples were collected from each subject at the end of each study visit. Enzyme-linked immunosorbent assay kits were used to quantitatively determine VCAM-1 (BOSTER #EK0537) and ICAM-1 (BOSTER #EK0370), where samples were run in duplicate.

Statistical Analysis

Statistical analysis was performed using commercially available software. A repeated-measures analysis of variance was conducted to examine the symptom severity data (IBM SPSS Statistics Version 26, Armonk, NY; Armonk, NY). When statistical significance was observed, pairwise comparisons were performed to detect where differences existed between visits. A Bonferroni correction was used to adjust the error rate for the number of comparisons. All other variables were checked for normality using Kolmogorov–Smirnov tests. Normality was confirmed visually with QQ-plot observations. Log transformations were made for data not conforming to a normal distribution. Linear mixed models were used to determine the main effects of time (visit) for all outcome variables (SAS Version 9.4, Cary, NC). Studentized residuals falling outside of three standard deviations were considered outliers and removed from the analysis. Significant findings were determined by an α level of 0.05 for pairwise comparisons. Tukey–Kramer post hoc correction was used where significant effects were observed. Comparisons with reference values were made using one-sample t and z tests. Pearson correlations were performed as changes from month 1 to month 6 for cfPWV and crPWV versus MAP, as well as for VCAM-1 and ICAM-1 versus cfPWV, crPWV, stiffness, distensibility, compliance, Young’s modulus, SBP, DBP, MAP, AIx, AIx@HR75, and AP to determine potential relationships between these variables.

RESULTS

Subject characteristics of the 14 subjects (7 M/7 F) are presented in Table 1. Study attrition over the 6-mo period led to combining the subjects’ final visit (months 5 and 6), with 9 subjects (4 M/5 F) completing their final visit at month 6 and 3 male subjects at month 5, for a total of 12 subjects completing five visits (month 1: 25 ± 6 days after positive SARS-CoV-2 test; month 2: 57 ± 7 days; month 3: 87 ± 8 days; month 4: 119 ± 13 days; and month ∼6: 174 ± 15 days). Five female subjects were taking prescribed oral contraceptives and two female subjects were taking selective serotonin reuptake inhibitors throughout the entirety of the study duration. Two subjects received their first SARS-CoV-2 vaccine doses between months 2 and 3, one subject with Pfizer and one subject with Moderna, with their second doses occurring between months 3 and 4. One additional subject received their first Moderna vaccine dose between month 1 and month 2, with the second dose between months 2 and 3. A 2-wk delay between SARS-CoV-2 vaccination and study testing was allotted to mitigate any potential confounding effects of the vaccination, a design previously used in vaccination studies (30). These subjects who sought out SARS-CoV-2 vaccinations throughout the study generally responded similarly over the study as those who did not seek out vaccinations and were, therefore, grouped with the unvaccinated subjects as a complete cohort.

Table 1.

Subject characteristics

Characteristics	Month 1 (25 ± 6 Days)	Month 2 (57 ± 7 Days)	Month 3 (87 ± 8 Days)	Month 4 (119 ± 13 Days)	Month 6 (174 ± 15 Days)
Subjects, n (M/F)	14 (7 M/7 F)	14 (7 M/7 F)	12 (7 M/5 F)	13 (7 M/6 F)	12 (7 M/5 F)
Age, yr	21 ± 1	21 ± 1	22 ± 1	21 ± 1	21 ± 1
Height, cm	177 ± 10	177 ± 9	176 ± 11	176 ± 10	176 ± 10
Weight, kg	73 ± 12	74 ± 13	73 ± 7	73 ± 10	74 ± 12
Body mass index, kg·m2	23 ± 3	23 ± 3	24 ± 2	24 ± 3	24 ± 3
Physical activity
Frequency, days/wk	4 ± 1	4 ± 1	4 ± 1	4 ± 1	4 ± 1
Duration, min/day	40 ± 12	40 ± 12	39 ± 13	40 ± 13	39 ± 13
Female contraceptive use	5	5	3	4	3

Data are means ± SD.

SARS-CoV-2 Symptom Severity Survey

Average symptom severity score, average number of symptoms, and number of subjects with symptoms are presented in Table 2. One female was asymptomatic and was not included in the symptom severity survey. At month 6, one female subject reported lingering symptoms.

Table 2.

SARS-CoV-2 symptom severity survey

Symptom	Month 1(25 ± 6 Days)(7M/6F)	Month 2(57 ± 7 Days)(7M/6F)	Month 3(87 ± 8 Days)(7M/5F)	Month 4(119 ± 13 Days)(7M/5F)	Month 6(174 ± 15 Days)(7M/4F)
Chest pain	2 ± 8(1 F)	1 ± 3(1 M)	0	0	0
Chills	0	2 ± 6(1 M)	0	0	0
Diarrhea	0	0	0	0	0
Dizziness /vertigo	3 ± 6(3 F)	0	0	1 ± 3(1 F)	0
Dry cough	3 ± 6(1 M/2 F)	4 ± 11(1 M/1 F)	0	3 ± 9(1 F)	0
Dry eyes	3 ± 6(3 M/1F)	2 ± 6(1 M/1F)	2 ± 6(1 M)	2 ± 6(1 M)	0
Dry mouth	3 ± 6(4 M)	1 ± 3(1 M)	2 ± 4(2 M)	2 ± 4(2 M)	0
Fatigue	8 ± 11(3 M/3 F)	2 ± 6(1 M/2 F)	3 ± 7(1 M/2 F)	0	0 ± 1(1 F)
Fever over 100.3°F	0	0	0	0	0
Headache	3 ± 12(1 F)	2 ± 6(2 F)	5 ± 10(3 F)	2 ± 6(1 F)	0
Lack of appetite	1 ± 3(1 M)	0	0	0 ± 0(1 F)	0 ± 1(1 F)
Loss of smell/taste, anosmia	11 ± 23(1 M/3 F)	3 ± 8(1 M/2 F)	5 ± 10(1 M/3 F)	3 ± 9(2 F)	0 ± 1(1 F)
Muscle or body aches	2 ± 3(2 M/1 F)	0 ± 1(1 F)	2 ± 6(1 M)	0	0
Congestion or runny nose	7 ± 13(5 M/2 F)	3 ± 9(1 M/1 F)	1 ± 3(1 M)	3 ± 9(1 M/1 F)	0
Nausea or vomiting	4 ± 14(1 F)	2 ± 6(1 F)	0	0	0
Shortness of breath, difficulty breathing, dyspnea	5 ± 14(4 F)	1 ± 3(1 F)	0	0	0
Sore joints	0	2 ± 6(1 M)	2 ± 6(1 M)	0	0
Sore throat	3 ± 7(2 M/1 F)	0	0	0	0
Number of symptoms	3 ± 2(7 M/6 F)	2 ± 1(5 M/6 F)	1 ± 2(3 M/4 F)	1 ± 1(3 M/3 F)	0 ± 1(1 F)
Average severity score	3.32 ± 3.10	1.36 ± 1.26‡	1.18 ± 1.73*	0.81 ± 1.14*	0.03 ± 0.07*

Data are means ± SD. * vs. month 1, P < 0.05; ‡ vs. month 6, P < 0.05. Header depicts total sample sizes available during each visits, while sample sizes within specific variables depicts how many subjects had individual symptoms. One female subject was asymptomatic and excluded from the table.

Pulse Wave Velocity

There was a significant effect of time on cfPWV (P = 0.013, Fig. 1), with cfPWV being reduced at months 4 and 6 compared with month 1. There was a tendency for crPWV (P = 0.074, Fig. 1) to decrease over time, with statistical analyses being performed on the transformed data. The change in cfPWV and crPWV was correlated with the change in MAP from month 1 to month 6 (cfPWV: r = 0.894, P < 0.001; crPWV: r = −0.896, P < 0.001, Table 4). Over time, crPWVcorr (P = 0.165) and cfPWVcorr pressure (P = 0.616) remained unchanged.

Figure 1.

Carotid-femoral pulse wave velocity (A), carotid-radial pulse wave velocity (B), and central blood pressures (C) in young adults with SARS-CoV-2. Month 1: 25 ± 6 days after positive SARS-CoV-2 test, n = 14 (7 M/7 F); month 2: 57 ± 7 days, n = 14 (7 M/7 F); month 3: 87 ± 8 days, n = 12 (7 M/5 F); month 4: 119 ± 13 days, n = 13 (7 M/6 F); month 6: 174 ± 15 days, n = 12 (7 M/5 F). Linear mixed models were used to determine main effects of time. Mean arterial blood pressures are presented as closed triangles (▴), systolic blood pressure is as open circles (○), and diastolic blood pressures as closed circles (●). * versus month 1, P < 0.05. Data are means ± SD.

Carotid Stiffness

Carotid stiffness (P = 0.933, Fig. 2), distensibility (P = 0.955, Fig. 2), and compliance (P = 0.861, Fig. 2) were unchanged over time. Young’s modulus (P = 0.800, Fig. 2) remained unchanged, with statistical analyses being performed on the transformed data. Carotid diameters and cIMT were unchanged over time (P > 0.05) (Table 3).

Figure 2.

Carotid stiffness (A), distensibility (B), compliance (C), and Young’s modulus (D) in young adults with SARS-CoV-2. Month 1: 25 ± 6 days after positive SARS-CoV-2 test, n = 14 (7 M/7 F); month 2: 57 ± 7 days, n = 14 (7 M/7 F); month 3: 87 ± 8 days, n = 12 (7 M/5 F); month 4: 119 ± 13 days, n = 13 (7 M/6 F); month 6: 174 ± 15 days, n = 12 (7 M/5 F). Linear mixed models were used to determine main effects of time. Data are means ± SD.

Table 3.

Carotid diameters and cIMT

	Month 1(25 ± 6 Days)(7M/7F)	Month 2(57 ± 7 Days)(7M/7F)	Month 3(87 ± 8 Days)(7M/5F)	Month 4(119 ± 13 Days)(7M/6F)	Month 6(174 ± 15 Days)(7M/5F)
Mean diameter, mm	6.73 ± 0.50	6.55 ± 0.54	6.63 ± 0.39	6.64 ± 0.49	6.72 ± 0.63
Systolic diameter, mm	7.04 ± 0.53	6.83 ± 0.54	6.79 ± 0.48	6.93 ± 0.50	6.89 ± 0.70
Diastolic diameter, mm	6.41 ± 0.47	6.25 ± 0.55	6.47 ± 0.47	6.34 ± 0.49	6.54 ± 0.70
cIMT, mm	0.45 ± 0.09	0.46 ± 0.09	0.46 ± 0.10	0.48 ± 0.05	0.48 ± 0.05

Data are means ± SD. Linear mixed models were used to determine main effects of time. cIMT, carotid intima media thickness.

Carotid Pulse Wave Analysis

AIx, AIx@HR75, and AP were unaltered over time (Table 4). There was a significant effect of time on mean systolic pressure (P = 0.007), which tended to be lower at month 3 (P = 0.088) and was lower at month 4 (P = 0.017) and month 6 (P = 0.008) compared with month 1 (Table 4). There was a significant effect of time on central aortic MAP (P = 0.002) with month 6 being lower than month 1 and central aortic systolic blood pressure (P = 0.023) with month 4 tending to be lower compared with month 1 (P = 0.051) and month 6 being lower than month 1 (P = 0.027) (Fig. 1). Central aortic diastolic blood pressure remained unaltered over time (P = 0.129; Fig. 1).

Table 4.

Central pressures and carotid pulse wave analysis measures

Parameter	Month 1(25 ± 6 Days)(7M/7F)	Month 2(57 ± 7 Days)(7M/7F)	Month 3(87 ± 8 Days)(7M/5F)	Month 4(119 ± 13 Days)(7M/6F)	Month 6(174 ± 15 Days)(7M/5F)
Peripheral carotid pressures
Peripheral systolic pressure, mmHg	128 ± 10	124 ± 10	122 ± 8	120 ± 7	119 ± 10
Peripheral diastolic pressure, mmHg	73 ± 6	71 ± 9	68 ± 8	68 ± 8	66 ± 6
Peripheral pulse pressure, mmHg	55 ± 9	53 ± 8	54 ± 10	52 ± 9	53 ± 11
Peripheral mean arterial pressure, mmHg	97 ± 6	91 ± 9	90 ± 6	90 ± 7	86 ± 7*
Central aortic pressures
Central pulse pressure, mmHg	51 ± 7	45 ± 9	49 ± 7	47 ± 7	45 ± 11
Pulse wave analysis
Heart rate, beats/min	64 ± 11	62 ± 8	64 ± 9	64 ± 9	61 ± 9
Heart rate period, ms	958 ± 156	988 ± 138	911 ± 283	953 ± 148	1,006 ± 9
Ejection duration, ms	320 ± 14	318 ± 15	318 ± 16	320 ± 15	326 ± 18
Ejection duration, %	34 ± 5	33 ± 4	34 ± 4	34 ± 4	33 ± 4
Aortic T2	222 ± 26	231 ± 26	231 ± 23	226 ± 25	217 ± 30
P1 Height (P1-D1)	44 ± 8	41 ± 9	44 ± 9	43 ± 8	40 ± 11
Aortic augmentation pressure, mmHg	7 ± 5	6 ± 8	4 ± 7	3 ± 4	5 ± 6
Aortic AIx, AP/PP; %	13 ± 9	12 ± 17	9 ± 15	8 ± 9	10 ± 13
Aortic AIx, P2/P1; %	117 ± 12	117 ± 23	112 ± 19	109 ± 10	111 ± 14
Aortic AIx@HR75, %	8 ± 8	6 ± 16	3 ± 15	3 ± 9	3 ± 15
End systolic pressure, mmHga	106 ± 7	102 ± 12	102 ± 15	98 ± 9	96 ± 10
Mean systolic pressure, mmHg	112 ± 7	107 ± 9	105 ± 6*	96 ± 25*	101 ± 8*
Mean diastolic pressure, mmHg	89 ± 6	84 ± 9	83 ± 7	83 ± 7	79 ± 7*
Rate pressure product, A.U.	7,891 ± 1,194	7,285 ± 841	7,416 ± 1,016	7,373 ± 1,048	6,807 ± 7,995
crPWVcorr, m·s−1·mmHg-1	0.072 ± 0.010	0.082 ± 0.012	0.079 ± 0.012	0.074 ± 0.009	0.078 ± 0.010
cfPWVcorr, m·s−1·mmHg−1	0.059 ± 0.008	0.059 ± 0.005	0.058 ± 0.003	0.056 ± 0.007	0.057 ± 0.008
Myocardial oxygen demand and perfusion
Buckberg SEVR (%)	159 ± 33	166 ± 31	159 ± 28	156 ± 31	161 ± 28
SPTI, mmHg·s·min−1	2,280 ± 300	2,082 ± 243	2,106 ± 232	2,130 ± 257	1,994 ± 228
DPTI, mmHg·s·min−1	3,537 ± 391	3,411 ± 464	3,309 ± 356	3,273 ± 369	3,170 ± 358

AIx, augmentation index; A.U., arbitrary units; cfPWVcorr, carotid-femoral pulse wave velocity corrected for MAP; crPWVcorr, carotid-radial pulse wave velocity corrected for MAP; DP, diastolic pressure; DPTI, diastolic pressure-time index; HR, heart rate; P1, first systolic peak; P2, second systolic peak; SEVR, subendocardial viability ratio; STPI, systolic pressure-time index; T2, duration from the start of the pulse to the second systolic peak. Data are means ± SD. Linear mixed models were used to determine main effects of time. *versus month 1, P < 0.05; statistical analyses were performed on the transformed variable.

Indices of Myocardial Oxygen Demand and Perfusion

Indices of myocardial oxygen demand and perfusion, including Buckberg SEVR, SPTI, DPTI, and ejection duration were unaltered over time (Table 4) (P > 0.05). Statistical analyses were performed on the transformed values for Buckberg SEVR and DPTI.

Endothelial Cell Adhesion Molecules

Measures of VCAM-1 and ICAM-1 are depicted in Table 5, where the statistical analyses for ICAM were performed on the transformed variable. There was a tendency for VCAM-1 to decrease (coefficient of variation: 6.3%; P = 0.088). ICAM-1 remained unchanged over time (coefficient of variation: 5.6%; P = 0.168), with statistical analyses being performed on the transformed data.

Table 5.

Endothelial cell adhesion molecules

Circulating Biomarker	Month 1(25 ± 6 Days)	Month 2(57 ± 7 Days)	Month 3(87 ± 8 Days)	Month 4(119 ± 13 Days)	Month 6(174 ± 15 Days)
VCAM-1, pg·mL−1	5,550 ± 2,540(5 M/4 F)	5,185 ± 2,610(5 M/4 F)	4,678 ± 1,807(5 M/3 F)	5,064 ± 2,031(5 M/3 F)	4,588 ± 1,405(5 M/2 F)
ICAM-1, pg·mL−1a	4,311 ± 1,249(6 M/5 F)	4,361 ± 1,296(6 M/5 F)	4,031 ± 1,192(6 M/4 F)	4,511 ± 1,550(6 M/4 F)	4,136 ± 1,153(6 M/3 F)

Data are means ± SD. Linear mixed models were used to determine main effects of time. *versus month 1, P < 0.05. statistical analyses were performed on the transformed variable. VCAM-1, vascular adhesion molecule; ICAM-1, intracellular adhesion molecule.

The change from month 1 to month 6 in VCAM-1 (5 M/2 F) was significantly correlated with changes in SBP (r = 0.607, P = 0.047) and MAP (r = 0.610, P = 0.046), but not with changes in cfPWV (r = 0.524, P = 0.098), crPWV (r = 0.594, P = 0.054), stiffness (r = 0.586, P = 0.058), distensibility (r = 0.486, P = 0.130), compliance (r = 0.474, P = 0.141), Young’s modulus (r = 0.460, P = 0.155), DBP (r = 0.600, P = 0.051), AIx (r = −0.118, P = 0.729), AIx@HR75 (r = −0.292, P = 0.384), or AP (r = −0.004, P = 0.991). The change in ICAM-1 from month 1 to month 6 (6 M/3 F) was significantly correlated with changes in cfPWV (r = 0.782, P = 0.004), crPWV (r = 0.904, P = 0.000), stiffness (r = 0.779, P = 0.005), distensibility (r = 0.631, P = 0.037), SBP (r = 0.796, P = 0.003), DBP (r = 0.703, P = 0.016), and MAP (r = 0.783, P = 0.004), but not with changes in compliance (r = 0.549, P = 0.080), Young’s modulus (r = 0.582, P = 0.060), AIx (r = 0.234, P = 0.488), AIx@HR75 (r = 0.042, P = 0.902), or AP (r = 0.316, P = 0.344).

DISCUSSION

The purpose of this investigation was to longitudinally track the potential improvement in indices of arterial stiffness of young adults who were infected with SARS-CoV-2 for 6 mo following initial infection. Our hypotheses were partially correct as, after initial infection, cfPWV significantly declined after 4 mo of recovery, whereas carotid stiffness measures remained unchanged across the 6 mo. However, SBP and MAP were significantly reduced at 6 mo after SARS-CoV-2 infection compared with 1 mo following infection. The changes from month 1 to month 6 in cfPWV and crPWV were correlated with changes in MAP, stressing the importance of MAP in interpreting changes to PWV and arterial stiffness. Together, these results provide evidence for prolonged time to improvement in some markers of arterial stiffness and blood pressure following infection with SARS-CoV-2 in young adults.

Despite requiring only ∼18 days after symptom onset to test negative for SARS-CoV-2 by a PCR test and for viral shedding cessation (31, 32), symptoms may last for months following infection. A recent follow-up study assessing patients hospitalized with SARS-CoV-2 over 12 mo observed fatigue and muscle weakness are the most prevalent, lingering symptom (33). Fatigue was also the most common and longest-lasting symptom reported in patients with SARS-CoV-1, persisting up to 4 years after initial infection (12). In the current study, the average symptom severity, total number of subjects with symptoms, and number of symptoms per subject decreased over 6 mo of recovery. Symptoms of dry eyes, dry mouth, fatigue, anosmia, congestion, and shortness of breath were most prevalent at 1 mo after infection with SARS-CoV-2, whereas fatigue and anosmia persisted the longest. Average symptom severity score was reduced to negligible levels, with the exception of one female subject at 6 mo following infection. Notably, in previous studies of SARS, the health status of survivors was lower than the general population, both physically and mentally, 1 year (34–36), and even up to 2 years (37), after infection. The current cohort of young adults with SARS-CoV-2 is unique as they had mild symptoms and were not hospitalized, yet symptoms persisted 4 mo after a positive SARS-CoV-2 PCR test, confirming potential effects may continue for at least several months after infection.

Pulse Wave Velocity and SARS-CoV-2

In the current investigation, cfPWV started at 5.70 m·s−1 and decreased by 16% over the 6-mo investigation period. Our observation that cfPWV takes 4 mo to significantly decrease from the initial measurement suggests vascular improvement from infection with SARS-CoV-2 may take 4 to 6 mo, with most young adults tending to improve by 6 mo. At month 4, 54% of subjects (7/13 subjects) had reduced crPWV and 85% (11/13 subjects) had reduced cfPWV. Likewise, at month 6, 66% (8/12 subjects) had reduced crPWV and 83% (10/12 subjects) had reduced cfPWV. After acute infection with SARS-CoV-2, previous investigations have observed elevated cfPWV. Our group previously observed elevations of cfPWV in young adults as compared with healthy controls (3). In middle-aged adults, Lambadiari et al. (38) observed elevated cfPWV 4 mo after SARS-CoV-2 infection as compared with controls. Schnaubelt et al. (14) observed elevated cfPWV in elderly adults with SARS-CoV-2 as compared with healthy controls, with a higher cfPWV noted in subjects who passed away ∼20 days following data collection. Further, Jud et al. (39) examined hospitalized older adult patients ∼7 mo post-SARS-CoV-2 infection and observed elevated PWV as compared with an age- and sex-matched control group. However, Nandadeva et al. (15) did not observe a difference in PWV among young adults with SARS-CoV-2 whose recovery time ranged from 1 to 5 mo when cross-sectionally compared with healthy control subjects, a time range which may have masked potential observational differences. In children with Kawasaki’s disease, PWV was found to be elevated, as compared with controls, even 8 yr following disease onset (40). Subsequent to the spread of SARS-CoV-2, children diagnosed with Kawasaki disease present with features of macrophage activation syndrome, hemophagocytic lymphohistiocytosis (41). Although this is a rare, yet serious, example of systemic inflammation, it suggests similarities between the SARS-CoV-2 and Kawasaki disease states.

PWV is indicative of arterial stiffness, where cfPWV is associated with central arterial stiffness and crPWV with peripheral arterial stiffness (42). As a ∼1 m·s−1 elevation in PWV is associated with a 15% higher risk for cardiovascular events, mortality, and all-cause death (43), the nearly 2 m·s−1 decrease in cfPWV observed over time in the current study suggests a reduction in cardiovascular event risk. Moreover, the cfPWV observed in the current study at 6 mo (4.88 ± 0.65 m·s−1) is not different compared with our previous control group (5.17 ± 0.66 m·s−1) (P = 0.937) (3), which also resembles the reference value of 4.36 m·s−1 expected in young adults (44). Though changes in crPWV did not reach statistical significance, it tended to decrease, suggesting it could be changing over time. This suggests crPWV has a different recovery period than cfPWV, possibly due to the differing measurement locations. These results demonstrate arterial stiffness is likely increased following infection and remains elevated for ∼3 mo. However, by month 4 following infection, arterial stiffness appears reduced compared with the first month following infection, especially among the central arteries.

PWV is dependent on the distending pressure inside the artery, with pressures in the artery being expressed as MAP (45–47). In the current study, the PWVcorr value remained unchanged over time, suggesting changes in cfPWV may be associated with changes in MAP. Restructuring of elastic arterial walls, including alterations to the composition of elastic and collagen fibers, may be caused by chronic changes in pressures, such as with advancing age and hypertension. Yet, short-term changes in pressure result in transient alterations to elastic arteries, causing a passive change in wall stiffness, distensibility, and PWV (46). In the current study, these changes in blood pressure over a 6-mo period may impact PWV and subsequent interpretation of arterial stiffness (48). More work may be necessary to identify the long-term impact of a change in MAP and PWV, whether changes to elastic and collagen fiber composition within the elastic arterial walls coincide with these pressure changes, and if arterial health is altered years after infection, especially among those with persistent symptoms or among those reinfected with emerging SARS-CoV-2 variants who may not fully recover from one SARS-CoV-2 infection to the next.

Carotid Stiffness and SARS-CoV-2

The current study is the first to demonstrate young adults with SARS-CoV-2 have elevated carotid stiffness 6 mo after initial infection. Observed carotid stiffness values in the current investigation (5.58 m·s−1) were significantly greater than the reference value of 4.47 m·s−1 expected of young adults (P = 0.001) (49). SARS-CoV-2 infection is associated with increased oxidative stress, which is believed to be linked to increased arterial stiffness (50–55). Similarly, compliance at month 6 of this investigation [1.04 (10−6·m2·kPa−1)] tended to be lower than the reference value of 1.24 (10−6·m2·kPa−1) (P = 0.054) for adults (49). Young’s modulus uniquely incorporates wall thickness into the measurement and provides a more integrated assessment of arterial circumferential stretching which is devoid by compliance and distensibility measures (56) and is also an independent risk factor for adverse cardiovascular events (57–59). Young’s modulus was not observed in the current study by month 6 (457 kPa) to be different from the reference value of ∼400 kPa (P = 0.253) for young adults (60). Although not altered over the first 6 mo following SARS-CoV-2 infection in the current investigation, we previously identified Young’s modulus to be elevated in young adults 1 mo after testing positive for SARS-CoV-2 compared with a control group (2), which may necessitate further longitudinal tracking beyond 6 mo to identify if Young’s modulus improves further. Furthermore, the distensibility values observed in young adults with SARS-CoV-2 during month 6 of the current investigation [0.46 (10−3·kPa−1)] was lower compared with the reference value for healthy young adults [∼40.28 (10−3·kPa−1)] (P < 0.01) (61). Consistently elevated carotid stiffness measures from our acute study among young adults 1 mo after SARS-CoV-2 infection (2), along with the current investigation, suggest an increased risk of both cardiovascular events and all-cause mortality during recovery from SARS-CoV-2, independent of other risk factors (62, 63). Together, these results provide evidence for long-term alterations to carotid stiffness indices which may require more than 6 mo to improve.

Our observation of lack of improvement in carotid stiffness indices may appear contradictory to our findings of decreased cfPWV across the 6 mo of recovery, yet these differences are quite possibly due to the heterogeneity across the systemic vasculature. Vascular assessments are not equivocal across the vascular tree (64–66), as regional differences in arterial stiffness have been documented in several pathophysiological states. In systemic lupus erythematosus, arterial stiffening may progress quicker in the aorta than the carotid arteries, suggesting regional and measure-specific sensitivity of the arteries to disease progression (67). In SARS-CoV-2, Nandadeva et al. (15) noted decrements to peripheral, but not central, vascular function, albeit in young adults across a range of 1–5 mo following SARS-CoV-2 infection. Together, these seemingly contradictory results stress the importance of region-specific arterial stiffness measurements to distinguish varying levels of vessel susceptibility to infection and time to improve following infection.

Aortic AIx is derived from wave reflections and is a sensitive marker of systemic arterial aging in young adults (24, 68–71). Interestingly, Nandadeva et al. (15) did not observe a difference in AIx@HR75 among healthy young adults compared with young adults who were recovering from SARS-CoV-2 infection, albeit there was a range of subjects studied from 1 to 5 mo after infection which could explain the lack of differences observed. Our group previously noted elevated AIx and AIx@HR75 in young adults with SARS-CoV-2 1 mo after infection as compared with a control group (2). Unchanged measures of AIx and AIx@HR75 after 6 mo of recovery from initial infection with SARS-CoV-2 in the current study may imply lingering arterial dysfunction. The differences observed between AIx and PWV may appear inconsistent, as cfPWV decreased over time despite AIx remaining elevated. Although PWV assesses arterial stiffness between two recording sites, AIx is more indicative of changes in the timing and amplitude of the reflected pressure wave, possibly caused by alterations to stiffness or arterial diameter (72). Furthermore, AIx is a more complex measure as it considers HR, the cardiac cycle, PWV, and the amplitude of the reflected wave when calculated (73).

At month 6, central aortic SBP and MAP were significantly lower than month 1, with diastolic blood pressure remaining unchanged. Furthermore, the mean SBP measured between the beginning of the waveform and the end of the first peak of the pulse wave tended to be lower by month 3 and was lower by months 4 and 6. These pressure alterations may be due to alterations in sympathetic activity or endothelial function. Our group previously observed elevated basal MSNA in young adults after acute infection with SARS-CoV-2 (7), which may be reflected in increased measures of arterial stiffness (74). Likewise, blood pressure may have also decreased due to downregulation of the ACE2 receptor and therefore the renin-angiotensin-aldosterone system (RAAS). Indeed, the binding of SARS-CoV-2 to the ACE2 receptor may result in overexpression of RAAS, leading to vasoconstriction due to accumulation of angiotensin II (75–77).

Buckberg SEVR is a noninvasive estimate of myocardial workload, oxygen supply, and perfusion (28, 78, 79), where SPTI is representative of myocardial oxygen demand and DPTI of oxygen supply and perfusion (80, 81). A prior study of Middle East respiratory syndrome coronavirus indicated acute myocarditis following infection (45), and myocardial injury and inflammation have also been reported after infection with SARS-CoV-2 (82–84). One month following infection, we noted an increase in SPTI as compared with controls but no differences in DPTI or Buckberg SEVR, suggesting the myocardium may have been undergoing a stress response due to infection with SARS-CoV-2, while the oxygen supply was maintained (2). SPTI remained unchanged in the current investigation, suggesting prolonged time to improvement from a systemic cardiovascular stress response 6 mo after infection with SARS-CoV-2. Furthermore, we noted a shorter ejection duration as compared with controls 1 mo after SARS-CoV-2 infection (2), possibly due to increased arterial stiffness (85, 86). Our current observation of unchanged ejection duration in the current study supports our assessments of arterial stiffening. Ultimately, these changes can lead to cardiac hypertrophy and may alter myocardial perfusion (87).

Endothelial Cell Adhesion Molecules and Cardiovascular Disease Risk

In the current study, there was a tendency for VCAM-1 to decrease without a change to ICAM-1 over the 6 mo following SARS-CoV-2 infection in young adults. In older adults hospitalized with SARS-CoV-2, levels of both VCAM-1 and ICAM-1 were elevated as compared with controls, as patients with mild illness had 123% greater VCAM-1 levels and 76% greater ICAM-1 levels (23). As levels of cell adhesion molecules increase in disease states associated with endothelial dysfunction and vascular remodeling, the current findings of a tendency for VCAM-1 to decrease over a 6-mo period in young adults suggest a reduction in arterial stiffness several months following SARS-CoV-2 infection (2, 22, 88). The difference in ligand binding, expression duration, as well as cell and tissue specificity, dictates the varying effects of VCAM-1 and ICAM-1 (89), with ICAM-1 being associated with atherosclerotic burden in several disease states (19). Furthermore, ICAM-1 has been studied in relation to infectious disease severity and outcome in young adults, however, study results are contentious and dependent on the population and timepoint of investigation (88). Notably, in the current study, changes in both VCAM-1 and ICAM-1 were associated with changes in SBP and MAP. Additionally, changes in ICAM-1 were associated with, and may quite possibly be predictive of, changes in functional markers of arterial stiffness, including PWV.

Study Limitations

It is important to highlight the association between preexisting arterial stiffness and comorbidities with severity of SARS-CoV-2 infection, where individuals with comorbidities have more severe symptoms and a higher mortality (90). However, our cohort consisted of young, otherwise healthy adults free of cardiovascular, metabolic, and pulmonary disease, experiencing no-to-mild symptoms of the virus. Further investigation is warranted to elucidate the long-term effects of SARS-CoV-2 on aging populations and those with preexisting medical conditions.

We recognize there are several limitations of this study. The longitudinal design of this study strengthens our analysis of the observed changes in vascular measures over time. However, due to an absence of baseline data before participants contracted SARS-CoV-2, we are still somewhat limited in our interpretation of observed changes. It is possible that these changes are innate monthly fluctuations within this particular cohort and not necessarily different from pre-SARS-CoV-2 infection. Furthermore, subject attrition limited the power of our statistical analysis. Increased levels of psychological stress in college students (91, 92) and decreased levels of physical activity due to the pandemic (93) may have provoked decrements to the vasculature and overall health, as mental health conditions can impact cardiovascular parameters (94). These factors may have altered recovery after infection with SARS-CoV-2. Future longitudinal studies are needed to fully understand the implications of and recovery from SARS-CoV-2 on arterial health. However, this study indicates at least partial improvement in vascular function from SARS-CoV-2 in young adults 6 mo after initial infection. We recognize that a control group or similar disease state would have been useful to understand normal physiological variability, although comparing with disease states such as influenza is contentious (95). Furthermore, rapidly evolving SARS-CoV-2 variants, reinfection potential, SARS-CoV-2 vaccination status and viral naiveté, as well as false-negative SARS-CoV-2 testing methodologies among a control group throughout the rapidly evolving COVID-19 pandemic certainly would have complicated the possibility of comparing a healthy or similar disease state during this longitudinal study (96–98). Future investigations should consider these confounding variables when designing future studies.

Clinical significance

The acute and potential long-term effects of SARS-CoV-2 infection should not be overlooked. Young adults are currently accounting for a majority of positive SARS-CoV-2 cases worldwide, emphasizing the need for elucidating the long-term impacts on health following SARS-CoV-2 infection. The thorough analyses of arterial stiffness in several vascular regions following SARS-CoV-2 infection suggest the virus may have varying effects throughout the vascular tree. Changes observed in blood pressure even within young adults highlight the importance of monitoring individuals with SARS-CoV-2 in tandem with underlying comorbidities, such as hypertension. Perhaps vaccination from SARS-CoV-2 will mitigate these long-term effects, although variants of SARS-CoV-2 may induce different perturbations to arterial health. Although the current study focused on young adults, more work is warranted to better understand the long-term health implications and recovery among those with underlying health conditions, older populations, and those with lifestyle habits that increase risk of cardiovascular complications, and whether their recovery efforts require a prolonged duration.

Conclusions

Arterial dysfunction, along with heightened blood pressure, may be an effect of SARS-CoV-2 infection (18, 21). The current study indicates previously observed decrements to arterial health following SARS-CoV-2 infection may be prolonged and related to blood pressure, even in young healthy individuals, as recovery from some of the observed decrements did not improve until 4–6 mo after infection, which may influence cardiovascular health and wellbeing. Continued longitudinal investigations into changes in arterial stiffness and blood pressure among aging populations and those with comorbidities are needed to elucidate the possible long-term effects of SARS-CoV-2 on cardiovascular health.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

GRANTS

This study was partially supported by an internal COVID-19 Research Cluster Award at Appalachian State University.

DISCLOSURES

S. M. Ratchford and A. S. L. Stickford now work for Medtronic, a publicly traded biomedical device company. They may hold stock within the company. However, the data obtained from this manuscript does not have direct benefit for the company and is an investigation into the structure/function of young adults recovering from SARS-CoV-2. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

J.L.S., A.S.L.S., and S.M.R. conceived and designed research; R.E.S., N.L.S., V.M.P., M.A.A., J.L.S., A.S.L.S., and S.M.R. performed experiments; R.E.S., N.L.S., V.M.P., M.A.A., J.L.S., A.S. L.S., and S.M.R. analyzed data; R.E.S., N.L.S., M.A.A., J.L.S., A.S.L.S., and S.M.R. interpreted results of experiments; R.E.S. and S.M.R. prepared figures; R.E.S., V.M.P., A.S.L.S., and S.M.R. drafted manuscript; R.E.S., N.L.S., V.M.P., M.A.A., J.L.S., A.S.L.S., and S.M.R. edited and revised manuscript; R.E.S., N.L.S., V.M.P., M.A.A., J.L.S., A.S.L.S., and S.M.R. approved final version of manuscript.