*Review* **The Importance of Arterial Stiffness Assessment in Patients with Familial Hypercholesterolemia**

**Beáta Kovács 1, Orsolya Cseprekál 2, Ágnes Diószegi 1, Szabolcs Lengyel 1, László Maroda 3, György Paragh 1, Mariann Harangi 1,\* and Dénes Páll 1,3**


**Abstract:** Cardiovascular diseases are still the leading cause of mortality due to increased atherosclerosis worldwide. In the background of accelerated atherosclerosis, the most important risk factors include hypertension, age, male gender, hereditary predisposition, diabetes, obesity, smoking and lipid metabolism disorder. Arterial stiffness is a firmly established, independent predictor of cardiovascular risk. Patients with familial hypercholesterolemia are at very high cardiovascular risk. Non-invasive measurement of arterial stiffness is suitable for screening vascular dysfunction at subclinical stage in this severe inherited disorder. Some former studies found stiffer arteries in patients with familial hypercholesterolemia compared to healthy controls, while statin treatment has a beneficial effect on it. If conventional drug therapy fails in patients with severe familial hypercholesterolemia, PCSK9 inhibitor therapy should be administered; if these agents are not available, performing selective LDL apheresis could be considered. The impact of recent therapeutic approaches on vascular stiffness is not widely studied yet, even though the degree of accelerated athero and arteriosclerosis correlates with cardiovascular risk. The authors provide an overview of the diagnosis of familial hypercholesterolemia and the findings of studies on arterial dysfunction in patients with familial hypercholesterolemia, in addition to presenting the latest therapeutic options and their effects on arterial elasticity parameters.

**Keywords:** familial hypercholesterolemia; selective LDL apheresis; PCSK9 inhibitor monoclonal antibody; arterial stiffness

#### **1. Introduction**

Cardiovascular diseases are still the leading cause of mortality worldwide, which is primarily due to increased atherosclerosis. In the background, the most important risk factors include hypertension, age, male gender, hereditary predisposition, diabetes, obesity, smoking and lipid metabolism disorders [1]. The proportion of people with high cholesterol levels is the highest in Europe in a worldwide comparison: it affects one in every two people [2,3]. Lipid metabolism is a complex process, with several known diseases that can significantly influence the lipid parameters of the serum, including cholesterol levels (secondary hypercholesterolemia). In two thirds of the cases, genetic factors are responsible for a pathological functioning of lipid metabolism (primary hypercholesterolemia). Polygenic forms are the most common, stemming from a cumulative effect of minor genetic deviances and gene variants, but more rarely, severe lipid metabolism disorders can be attributed to the mutations of individual genes [4].

**Citation:** Kovács, B.; Cseprekál, O.; Diószegi, Á.; Lengyel, S.; Maroda, L.; Paragh, G.; Harangi, M.; Páll, D. The Importance of Arterial Stiffness Assessment in Patients with Familial Hypercholesterolemia. *J. Clin. Med.* **2022**, *11*, 2872. https://doi.org/ 10.3390/jcm11102872

Academic Editors: Andrea Grillo and Paolo Salvi

Received: 25 April 2022 Accepted: 17 May 2022 Published: 19 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **2. Genotype and Phenotype of Familial Hypercholesterolemia**

Diagnosing primary hypercholesterolemia is based on hereditary features as well as the lack of secondary factors. The group of patients suffering from familial hypercholesterolemia (FH) is prominent due to the prevalence of the disease and the severity of related cardiovascular complications. The disease, which was formerly described as a result of an inactivating mutation of the low-density lipoprotein (LDL) receptor (LDLR), is today accounted for as familial hypercholesterolemia syndrome, including both the classic form of the disease as well as a form caused by the loss-of-function mutation of apolipoprotein B100 (ApoB100), formerly termed familial defective apoB syndrome, in addition to other severe types of hypercholesterolemia similarly exhibiting autosomal dominant inheritance, some of which have been found to be related to gain-of-function mutations of the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene [5]. In 2014, STAP1 (signal transducing adaptor family member 1) was reported as a novel FH candidate gene [6]. However, functional validation studies have not been reported, and possible mechanisms by which STAP1 could influence plasma lipid levels have not been explored. Indeed, no marked changes have been demonstrated in plasma lipid profiles of carriers of STAP1 variants compare to controls as well as in a mouse model [7]. Moreover, global loss of Stap1 in mice did not result in an abnormal lipid phenotype [8]. Accordingly, following these negative findings, the combined studies exclude STAP1 as an FH gene [9]. The rather rare autosomal recessive form is caused by the mutation of LDL receptor adaptor protein 1 (LDLRAP1) [10] (Table 1).



ApoB100: apolipoprotein B100; FH: familial hypercholesterolemia; FDB: familiar defective ApoB; LDLR: lowdensity lipoprotein receptor; LDLRAP1: LDLR adaptor protein 1; PCSK9: proprotein convertase subtilisin/kexin type 9.

It must be noted that in clinically diagnosed FH patients without mutations in the classical genes, elevated LDL-C levels might have a polygenic cause. Such patients often carry a cluster of common polymorphisms affecting several loci associated with markedly raised LDL-C levels, comparable to those observed in patients carrying FH-causative mutations. Even in patients with monogenic FH, a polygenic contribution may subsist, contributing to the variable phenotypic expression [11]. Both monogenic FH and polygenic hypercholesterolemia are found to be associated with greater risk of cardiovascular disease (CVD) than hypercholesterolemia without a known genetic cause, with monogenic FH associated with the greatest risk [12].

#### *2.1. Heterozygous FH*

Genetically, the less severe heterozygous form of FH is more common than it has been previously believed. Its prevalence in the European population is currently estimated at approximately 1:300, but it may reach 1:200 [13]. According to the findings of a Hungarian research project conducted a few years ago on a large patient population, the prevalence of FH in Hungary, similarly to several European countries, is around 1:340 [14]. The study found that in addition to a substantial rise in total cholesterol and LDL-cholesterol (LDL-C) levels, the level of triglyceride did not increase in general, and the level of highdensity lipoprotein-cholesterol (HDL-C) approximated the upper normal domain. Lipid abnormalities are acquired already in childhood, so screening the entire population for cholesterol levels no later than at age 8–11 has been proposed [13–15].

The LDL-C level is also significantly high in heterozygous cases between 4.9 and 11.6 mmol/L. As a result of the early and extremely high total cholesterol, atherosclerosis is already pronounced in childhood, and without treatment it may well lead to coronary artery disease before the patient is 35, but other vascular diseases—cerebrovascular disease or lower extremity arterial disease—have a greater risk. There is also a risk of developing tendonous xanthomata above extensor tendons, xanthelasmata on the eyelids, and corneal arcus on the iris. These are conspicuous deviations, which may help in diagnosing the disease.

Diagnosing FH is assisted by the Dutch Lipid Clinic Network criteria, which rests on three pillars: clinical symptoms, deviations in laboratory findings, and—in the case of ambiguous laboratory and clinical symptoms—a genetic test [16] (Table 2). Treatment relies on the efficient administering of statins in high doses, which in most cases is complemented by ezetimibe. As the patients are at high or extremely high cardiovascular risk, in order to reach lipid targets, it may become necessary to introduce PCSK9-inhibitor monoclonal antibodies, too [17]. Inclisiran, a small interfering RNA (siRNA) therapy, is another nonstatin medication available for cholesterol management [18]. Cardiovascular screening for patients and their family members is obligatory in the case of FH [19].


**Table 2.** Dutch Lipid Clinic Network diagnostic criteria [16].

CAD: coronary artery disease; HDL: high-density lipoprotein; LDL: low-density lipoprotein; PAD: peripheral artery disease; TG: triglyceride.

#### *2.2. Homozygous FH*

The homozygous form of FH is a rare but rather severe disease, characterized by extreme increased LDL-C (above 12.9 mmol/L) from birth. Progressive atherosclerosis causes myocardial infarction and vascular complications in other parts of the body even in childhood, but cases of aortic valve stenosis as well as subvalvular aortic stenosis have also been recorded. Prevalence is 1:1,000,000, but due to premature mortality it is supposedly higher [20].

To examine the interconnections between the genotype and the phenotype, we need to pose the question whether the severe clinical phenotype is unequivocally tied to the homozygous mutation of a single gene. True homozygous cases, in which the same pathogenic mutation occurs in identical candidate genes in both DNA strands, result in a severe clinical disease, but similarly severe clinical diseases can be caused by compound and double heterozygous forms, too, in which a pathogenic mutation develops in different sections of the given gene, or in two different candidate genes. Therefore, when a homozygous form is suspected, it is reasonable to perform a genetic test.

#### **3. Treatments of Familial Hypercholesterolemia**

#### *3.1. Pharmacological Treatment*

The basis of the treatment is administering high-dose, intensive statin, then, if that should prove insufficient, it is combined with ezetimibe. If the LDL-C target is not accomplished, PCSK9-inhibitor treatment should be provided as a third stage [19]. In the case of homozygous as well as severe homozygous FH, the impact of statin, ezetimibe and PCSK9-inhibitors is for the most part inefficient. In the case of homozygous FH, the administering of lomitapid, a microsomal transfer protein inhibitor [21], or mipomersen, an ApoB100 synthesis inhibitor antisense oligonucleotide [22] is recommended.

In addition to these, in the case of severe heterozygous and homozygous cases, it may become necessary to provide selective LDL apheresis treatment.

#### *3.2. Selective LDL Apheresis Treatment*

LDL apheresis treatment involves the removal of atherogenic lipid fractions by a selective extracorporeal procedure. During apheresis, the particles containing ApoB100 are removed selectively, which can acutely reduce total and LDL-C levels by a further 50–75% beyond pharmacological treatment. In addition to that, it reduces the level of lipoprotein (a) (Lp(a)) [23], known as an individual cardiovascular risk factor also in FH patients [24,25], as well as very low-density lipoprotein levels (VLDL), also containing ApoB, by 70% [26]. During apheresis, the level of protein-like components causing cardiovascular diseases is decreased in the serum; thus, the treatment has an evidently beneficial anti-atherosclerotic impact [27]. The serum levels of inflammatory cytokines and oxidative stress are reduced [28]. Vasodilation increases, and beneficial hemorheological changes are affected [29,30].

In accordance with international recommendations (by the Food and Drug Administration), in Europe it is recommended to perform LDL apheresis in the following patient groups: *functional FH homozygote, LDL-C > 13 mmol/L, functional FH heterozygote, LDL-C > 7.8 mmol/L, functional FH heterozygote with documented ischemic heart disease and LDL-C > 5.2 mmol/L.* Alongside the recommended diet and lipid lowering pharmacological treatment at the maximum tolerated dose, LDL-C levels must exceed the specified target throughout the course of 6 months. Alternative indication is provided by findings that show *below 40% LDL-C decrease in heterozygous FH patients* alongside lipid lowering pharmacological treatment at the maximum tolerated dose. Further indication is provided by an over 60 mg/dL Lp(a) level at documented ischemic heart disease, with over 4 mmol/L LDL-C level despite the administered pharmacological treatment. The above treatments have yielded favorable results, but due to limited funding the number of patients systematically treated in Europe is unfortunately still low [31].

#### **4. Significance of Cardiovascular Risk Assessment**

Cardiovascular morbidity and mortality are significantly enhanced in patients with different types of FH; therefore; they all would require individual cardiovascular risk assessment at the earliest possible stage of the disease course [32]. Not all FH patients are at the same cardiovascular risk [33]. The main predictors of morbidity and mortality in this patient population are known to be the age HDL-s, gender, hypertension, and smoking, which all together make up the Montreal FH risk prediction score [34]. Those individual biomarkers do not provide direct information about the hard outcomes and moreover, their cumulative effect could only be declared only if hard endpoints have already occurred. Biomarkers and comorbidities result in an intermediate state of decreased arterial elasticity as a surrogate endpoint of cardiovascular hard outcome in this patient population [35]. Signs of accelerated athero- and arteriosclerosis may occur in early childhood in special types of FH, which refer to the hazard of later functional and structural arterial damage. AHA for instance classified homozygous FH as a Tier I risk group from early childhood [36].

Early risk assessment and patient education are crucial factors to prevent later fatal outcomes. Arterial stiffness measurement contributes to understanding cardiovascular morbidity and mortality risk beyond traditional risk factors or blood pressure measurement [37]. There are several methods to assess arterial stiffness. Due to the fact that there are plenty of measurement tools to assess central and peripheral vascular elasticity, none of them has been approved as a standard method to routinely measure stiffness as a surrogate marker [38]. Nonetheless, ESH suggests the non-invasive measurement of central PWV as the gold standard method to assess preclinical organ damage in patients at high cardiovascular risk [39]. Some studies proved that beyond traditional risk assessment tools, and it may offer additional value and refinement of strategies applied thus far.

Measuring arterial stiffness as a surrogate marker is of paramount importance and will further help lengthen the survival of FH patients.

#### **5. Arterial Stiffness**

Suitable non-invasive methods in early stages of atherosclerosis and related artery wall disorders include measuring arterial stiffness. Increased arterial wall stiffness is a result of complex structural changes in the tunica media of the great arteries and of their increased and progressive calcification. Severe arterial stiffness precedes the development of atherosclerosis, which then causes the symptoms. Functional parameters that are independent predictors of cardiovascular diseases include applied pulse wave velocity (PWV) and augmentation index (Aix), which are suitable for both screening and monitoring the efficiency of the treatment [40]. In the past years, the 24-h monitoring of stiffness has become increasingly widespread. Internationally these measurements are most often performed by devices using the oscillometric method, such as Vasotens® (BPLab GmBH Schwalbach am Taunus, Hessen, Germany) and Mobil-O-Graph® (IEM, Stolberg, Germany). Arterial stiffness parameters—pulse wave velocity, augmentation index, central blood pressure are recorded throughout 24 h with the help of an upper arm blood pressure monitor. The development of early-onset artery wall dysfunction with increased cardiovascular risk was first diagnosed in diseases with chronic inflammation and lipid metabolism disorder. Artery wall dysfunction and related chronic inflammation involve a change in the marker levels of several inflammatory proteins and other serum markers [41].

#### **6. Relationship between Cholesterol and Arterial Stiffness**

The link between serum cholesterol level and arterial stiffness may be explained by several potential mechanisms [42]. The more obvious and probably the most important is the development of atherosclerosis, which has been consistently associated with increased arterial stiffness in subjects with and without severe hypercholesterolemia. However, cholesterol and especially oxidatively modified LDL (oxLDL) have further, nonatheromatous effects on the arterial wall, leading to arterial stiffening. OxLDL promotes peroxynitrite formation and increased oxidative stress, which may lead to the direct damage of elastin, the main elastic element of the arterial wall [43]. Furthermore, oxLDL has pro-inflammatory effects characterized by increased serum levels of C-reactive protein (CRP), which was associated with arterial stiffness in apparently healthy individuals [44]. Inflammatory cytokines enhance the expression of some inducible enzymes, such as matrix metalloproteinase-9 (MMP-9), that may damage the structural components of the arterial wall. MMP-9 is a gelatinase secreted by immigrating inflammatory cells of the vascular wall, capable of digesting elastin leading to the remodeling of the arterial wall [45]. Local inflammation and inflammatory lipids may also promote calcium deposition in the medial elastic fibers, resulting in arterial calcification [46]. In addition to structural changes, hypercholesterolemia induces functional dysregulation of the vascular endothelium found to be associated with arterial stiffness. High serum levels of cholesterol are significantly associated with reduced bioavailability of nitric oxide, impaired L-arginine/nitric oxide pathway and increased asymmetric dimethyl arginine production, leading to impaired endothelial vasodilatation [47]. On the other hand, vascular dysfunction may also be caused by the overproduction of vasoconstrictor agents, including endothelin-1 [48].

The relationship between serum lipid levels and arterial stiffness have been examined in several former studies. Most of them demonstrated a significant positive relationship between large artery stiffness and total or LDL-cholesterol [49–52]. It must be noted that many of these studies included a relatively low number of patients with other cardiovascular risk factors. Therefore, careful interpretation of the data is essential. FH represents an extreme form of total and LDL cholesterol elevation; therefore, it may serve as an excellent model to prove the link between LDL cholesterol and arterial stiffness.

#### **7. Assessing Arterial Stiffness in Familial Hypercholesterolemia**

Patients with familial hypercholesterolemia have an extremely high risk of atherosclerosis and early-onset vascular ageing, which can be measured by the increase of arterial stiffness. This is due to high cholesterol levels, including the presence of high serum LDL-C values as well as high Lp(a), more prevalent than in the general population, and higher levels of oxidized LDL and chronic artery wall inflammation due to increased oxidative stress [53]. In the case of FH, a low-fat diet and a conventional lipid lowering treatment have limited efficiency and due to the already existing arterial complication, often it is not possible to do physical exercise to the desired effect. All of these may be exacerbated by conventional cardiovascular risk factors, such as age, excess weight, diabetes, hypertension, and smoking. These may be accompanied by other unfavorable genetic factors as well [54] (Figure 1).

**Figure 1.** Factors leading to increased vascular stiffness in familial hypercholesterolemia. LDL-C: low density lipoprotein-cholesterol, Lp(a): lipoprotein (a); oxLDL: oxidized low-density lipoprotein.

As a result of accelerated atherosclerosis, sooner or later all patients develop coronary stenosis; however, its extent and the severity of the resulting clinical symptoms have a broader spectrum. Even though the injurious effects of arterial stiffness on the population of non-FH patients have been registered by several studies, the available data on the clinical impact of arterial stiffness on FH patients is insufficient.

In a small cross-sectional comparative study, brachial-ankle pulse wave velocity (baPWV) was measured in 35 heterozygous FH subjects and 17 healthy control subjects. Although baPWV disi not differ significantly between FH patients and controls (12.5 ± 2.9 vs. 11.9 ± 2.3 m/s), among FH patients, the baPWV and carotid IMT were higher in cases with high cholesterol burden than those without. Similarly, the baPWV and carotid IMT were also higher in cases with elevated hs-CRP than those without [55].

In a former case control study of 22 patients with FH and matched healthy controls, PWV values were compared before and after lipoprotein apheresis (LA) treatment. Baseline PWV was similar between the two groups (controls 8.2 ± 0.9 m/s vs. FH 7.7 ± 0.8 m/s, *p* = 0.12). Moreover, baseline PWV did not change following LA (pre 8.8 ± 1.2 m/s vs. post 9.2 ± 1.2 m/s, *p* = 0.19) [56].

Another study involved 60 patients without documented cardiovascular events and clinical symptoms of cardiovascular diseases: 21 patients with elevated plasma LDL-C levels and genetically confirmed FH, 19 patients with elevated LDL-C levels and without FH mutations and 20 healthy controls. In each patient, echo-tracking and photoplethysmography were used to assess the parameters of arterial stiffness. They found that arterial stiffness parameters were similar between the groups [57].

A study conducted on the population of 125 FH patients as per the guidelines displayed significantly higher Aix values in comparison to those of a control group of identical age and gender (9.6 ± 17.2 vs. 2.6 ± 10.3%; *p* = 0.011), based on which the measuring of Aix value is recommended in patient tracking [58].

In a study conducted on 66 untreated FH patients and 57 first-degree non-FH relatives, when measuring carotid β-stiffness index and carotid-femoral PWV it was found that while FH patients' β-index (6.3 (4.8–8.2) vs. 5.2 (4.2–6.4); *p =* 0.005) and local PWV values (5.4 (4.5–6.4) vs. 4.7 (4.2–5.4) m/s; *p =* 0.005) were significantly higher than in the case of their non-FH relatives, there was no substantial deviation in carotid-femoral PWV values (6.76 (7.0–7.92 vs. 6.48 (6.16–7.12) m/s; *p =* 0.138). Based on all the above, the measurement of carotid arterial stiffness, especially in the case of younger patients, may indicate the extent of calcification sooner than does arterial stiffness of the aorta [59].

A Japanese group of researchers recorded changes in brachial-tibial pulse wave velocity (baPWV) as well as the development of coronary artery disease in 245 medicated FH individuals. The patients were selected on the basis of clinical criteria for FH specified by the Japan Atherosclerosis Society. According to these, two out of three clinical criteria need to be met for a diagnosis of FH, namely, LDL-C ≥ 180 mg/dL, the presence of tendonous xanthoma or xanthoma tuberosum, as well as an FH-positive family history or early-onset CAD diagnosed in second-degree relatives. Cardiovascular risk factors (age, male gender, hypertension, diabetes, smoking) have been assessed as well as deviances in lipid parameters (total cholesterol, triglycerides, HDL) and the presence of CAD. In the latter case, the diagnosis was established on the basis of coronary CT angiography by taking into account only over 50% stenosis of the main coronary arteries. Measurement of brachial-tibial pulse wave velocity was performed with a Colin VP-1000, Omron® device. The goal of the study was to establish a connection between arterial stiffness and the risk of CAD in the given population. The findings proved that in the case of FH arterial stiffness, including baPWV as a biomarker indicating high cardiovascular risk, showed correlation with CAD [60].

Parameters of arterial stiffness as well as aortic root thickness by cardiac MRI have also been tested on heterozygous FH children. Testing 33 children aged 7–18, it was found that in comparison to a non-FH group of identical age, the PWV values of FH children were significantly higher (4.5 ± 0.8 vs. 3.5 ± 0.3 m/s; *p* < 0.001), and the wall thickness of the ascending aorta was higher (1.37 ± 0.18 vs. 1.3 ± 0.02 mm; *p* < 0.05), which suggests the importance of early statin treatment [61].

However, some further studies in children and young patients with FH with low patient numbers could not demonstrated significant differences compared to control subjects [35,62,63].

Indeed, a recent meta-analysis of 8 studies involving 317 patients with FH and 244 non-FH individuals did not suggest a significantly altered PWV in FH patients versus controls, although the authors admit that different scores for FH diagnosis as well as different methods for PWV estimation were used in different studies included and there was a lack of information about the duration and type of lipid-lowering therapy [64].

Taken together, larger studies evaluating PWV in FH patients compared with controls in order to elucidate the impact of FH on arterial stiffness as measured by PWV are definitely needed. A very recent position paper of the Associations of Preventive Paediatrics of Serbia, Mighty Medic and International Lipid Expert Panel focusing on risk assessment and clinical management of children and adolescents with heterozygous FH stated that depending on the availability of noninvasive equipment for PWV measurement and staff experience, it would be clinically meaningful to perform PWV measurements in all children with FH and evaluate their changes over time. PWV values above 97th could be a possible guide for treatment initiation in ambiguous clinical cases (Level B evidence). The paper prefers oscillometric devices; it suggests the usage of Mobile-O-Graph device due to the simplicity of measurement [65].

Whether the parameters for establishing arterial stiffness can be replaced by measuring inflammatory parameters of atherosclerosis is an important question. According to the findings of a study conducted on 89 FH patients and a control group of 31, PWV values were higher in FH patients (*p* < 0.05), but no significant connections to serum inflammatory parameters have been found (C-reactive proteins and white blood count) [66].

#### **8. Effect of Traditional Oral Lipid-Lowering Treatment on Arterial Stiffness**

Several human studies have investigated the effect of HMG-CoA reductase inhibitor statins on arterial stiffness. It is well documented that in addition to cholesterol reduction, a number of other pleiotropic effects have been described with statins that may improve atherogenesis independently of cholesterol reduction, including their anti-inflammatory and antioxidant effects, leading to improved endothelial function [67,68]. With the exception of one short-term, cross-over study, which could not find any impact of pravastatin on carotid, brachial and femoral artery stiffness [69], most of these former studies reported significant reduction in stiffness parameters after higher doses of atorvastatin and cerivastatin treatment [70–73]. Even a long-term, but low-dose pravastatin treatment could improve the aortic pulse wave velocity, especially in patients with the greatest lowering of cholesterol [74]. It must be noted that most statins had significant adjuvant effects on peripheral systolic blood pressure [72,75]. On the contrary, 20-week treatment with statins (rosuvastatin or atorvastatin) combined with regular exercise significantly improved exercise capacity and brachial artery PWV, but had no effect on blood pressure [76]. In patients with coronary artery disease compared with simvastatin/ezetimibe, rosuvastatin was found to more effectively improve arterial wall stiffness [77]. Although the beneficial effect of statin on stiffness parameters in various patient populations are well established, to date, their effects on stiffness parameters in FH patients has not been investigated.

Recently, the impact of PCSK9 plasma levels on mechanical vascular impairment was verified [78]. The exact mechanism is not fully clarified, but PCSK9 might promote atherogenesis by stimulating oxidative stress and the production of proinflammatory cytokine production of the atherosclerotic lesions [79]. It must be highlighted that statins, especially the lipophilic agents significantly increase the circulating level of PCSK9. Furthermore, statin-induced PCSK9 increase may limit the absolute magnitude of statin LDL-C lowering effect, by limiting the statin-driven LDLR upregulation [80].

#### **9. Changes in Arterial Stiffness and PCSK9-Inhibitor Monoclonal Antibody Treatment**

In the past few years, new product groups have appeared in the range of FH therapeutic medicines. Breakthrough products included the aforementioned ApoB synthesis inhibitors, microsomal transfer protein inhibitors and especially PCSK9-inhibitors. Currently, the efficiency, side effect profiles and effect of PCSK9-inhibitor monoclonal antibodies on the cardiovascular endpoints are very similar. In the RUTHERFORD study, 168 heterozygous FH patients were treated with 350 mg and 420 mg evolocumab every 4 weeks alongside statin treatment. In the case of 350 mg, LDL-C decreased by 43%, in the case of 420 mg by 55%. After 12 weeks, 44% and 65% of the patients, respectively, reached the desired 1.8 mmol/L target value. Triglyceride levels decreased by 15% and 20%, respectively, while HDL-C increased by 7%, and Lp(a) fell by 23% and 32%, respectively, as a result of evolocumab treatment. In the case of administering 140 mg every 2 weeks, LDL-C decreased by 66% [81]. Therefore, the impact on total cholesterol and LDL-C is conclusive; however, the impact on Lp(a) level lags that of LDL apheresis.

Changes in lipid parameters and pulse wave velocity due to complementary treatment using PCSK9-inhibitors or ezetimibe were studied in a 6-week tracking study. The research project involved ninety-eight certified FH patients who had previously undergone other cardiovascular risk assessment. All patients had genetically certified FH. The patients had received high-dose statin (atorvastatin 40–80 mg, rosuvastatin 20–40 mg) and/or ezetimibe treatment at least 6 months prior to the commencement of the study, but still they had not reached the desired LDL-C targets. In the study, 53 patients were administered statin+ezetimibe+PCSK9-inhibitors (alirocumab 75 mg/150 mg or evolocumab 140 mg). Forty-five patients were administered ezetimibe alongside the statin already administered. PWV measurements had been taken prior to the complementary pharmacological treatment and 6 months after the optimized treatment. Measurements were conducted with the SphygmoCor CVMS® device. In the case of patients in the PCSK9 group, a more significant decrease was recorded not only in LDL-C values, but also in pulse wave velocity (−51% vs. −22.8%, *p* < 0.001 and −15% vs. −8.5%, *p* < 0.01) [82].

The effect of six-month add-on PCSK9 inhibitor monoclonal antibodies on circulating PCSK9 and PWV was detected in a cohort of FH subjects. The PCSK9 plasma level was correlated with PWV at baseline. Furthermore, reduction of PCSK9 plasma level seems to be associated with a significant mechanical vascular improvement after PCSK9 inhibitor monoclonal antibody therapy. Therefore, PCSK9 could be a novel cardiovascular biomarker of the mechanical vascular homeostasis through lipid and non-lipid pathways, and it could identify subjects at high CVD risk with a limited LDL-C lowering benefit after high-intensity statin therapy in FH [78].

#### **10. The Impact of LDL Apheresis Treatment on Vascular Parameters**

The impact of LDL apheresis treatment on arterial stiffness is the least documented in spite of the fact that in acute cases this extracorporeal procedure yields the most substantial metabolic and hemodynamic changes. Professional literature on the subject is also rather scarce. A German research team examined the impact of lipoprotein apheresis treatment on the parameters of endothelial function (circulating endothelial cells, circulating endothelial progenitor cells, flow-mediated vasodilation, microalbuminuria) as well as left ventricular ejection fraction and changes in homocystein levels. Heterozygous FH patients were examined: 21 patients were administered statin at the maximum tolerated dose, while 8 patients proved to be statin intolerant. Direct adsorption of lipoproteins (DALI) was provided on a weekly basis. Primarily in the case of the statin intolerant patient's immediate improvement was recorded in vascular contractility even after a single treatment. Regular apheresis throughout a course of 6 months clearly had a favorable effect on the metabolic parameters under survey and improved endothelial function, which is one of the key causes of clinical improvement [83].

#### **11. Conclusions**

In familial hypercholesterolemia, complex molecular and hemodynamic changes are involved in the development of cardiovascular complications. Even though an increasing amount of clinical data is available, the exact role of serum cholesterol in the changing elasticity of different arterial sections is still not clear. Several combined pharmacological treatments are available for remedying metabolic deviations, but in clinically severe cases, complex pharmacological and non-medicinal treatments, such as selective LDL apheresis, can be used jointly. Similarly, little is known about changes in arterial stiffness induced by new generations of medicines, such as PCSK9-inhibitor monoclonal antibodies, siRNAs and selective LDL apheresis treatments. Atherosclerosis and related artery wall dysfunction can be screened for by non-invasive arterial stiffness measurements. Today oscillometric ABPM devices are available for performing a 24-h measurement of arterial stiffness; these are used primarily in scientific research but not widespread in clinical practice. Since these devices are suitable for measuring biomarkers indicating high cardiovascular risk, such measurements may contribute to screening especially high-risk patients with familial hypercholesterolemia and to improving the efficiency of their treatment.

**Author Contributions:** B.K. contributed with the processing of data from professional literature, O.C., S.L., L.M. and Á.D. edited the manuscript, M.H. generated the figures and compiled the manuscript and G.P. and D.P. proofread the final version. All authors have read and agreed to the published version of the manuscript.

**Funding:** The project is co-financed by the European Union under the European Regional Development Fund. The funding number is the GINOP-2.3.2-15-2016-00062.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**


#### **References**


## *Review* **Arterial Stiffness in Thyroid and Parathyroid Disease: A Review of Clinical Studies**

**Andrea Grillo 1,2,\*, Vincenzo Barbato 1, Roberta Maria Antonello 1, Marco Fabio Cola 1, Gianfranco Parati 3,4, Paolo Salvi 3, Bruno Fabris 1,2 and Stella Bernardi 1,2**


**Abstract:** Growing evidence shows that arterial stiffness measurement provides important prognostic information and improves clinical stratification of cardiovascular risk. Thyroid and parathyroid diseases are endocrine diseases with a relevant cardiovascular burden. The objective of this review was to consider the relationship between arterial stiffness and thyroid and parathyroid diseases in human clinical studies. We performed a systematic literature review of articles published in PubMed/MEDLINE from inception to December 2021, restricted to English languages and to human adults. We selected relevant articles about the relationship between arterial stiffness and thyroid and parathyroid diseases. For each selected article, data on arterial stiffness were extracted and factors that may have an impact on arterial stiffness were identified. We considered 24 papers concerning hypothyroidism, 9 hyperthyroidism and 16 primary hyperparathyroidism and hypoparathyroidism. Most studies evidenced an increase in arterial stiffness biomarkers in hypothyroidism, hyperthyroidism and primary hyperparathyroidism, even in subclinical and mild forms, although heterogeneity of measurement methods and of study designs prevented a definitive conclusion, suggesting that the assessment of arterial stiffness may be considered in the clinical evaluation of cardiovascular risk in these diseases.

**Keywords:** arterial stiffness; thyroid; parathyroid; cardiovascular disease

#### **1. Introduction**

Growing evidence shows that arterial stiffness measurement provides important prognostic information and improves clinical stratification of cardiovascular disease [1,2]. The stiffening of arteries is notably linked to aging, but a number of risk factors and diseases may affect this process, which is usually referred as arteriosclerosis [3]. The biological mechanisms underlying arterial stiffening involve the degradation of elastin layers in the tunica media of the large arteries and the proportional increase in collagen fiber content, along with smooth muscle cell hyperplasia, fibrosis and calcification of the media [4]. The process of stiffening interacts with atheromatous plaque formation and inflammation in the development and progression of cardiovascular disease [5]. The use of arterial stiffness markers is considered of clinical interest to assess cardiovascular risk in particular among patients not presenting the classical risk factors (e.g., smoking, obesity, diabetes), in which evaluation of arterial damage may improve risk stratification [6].

Thyroid and parathyroid diseases should be considered as endocrine diseases with relevant implications in the cardiovascular system. Patients with overt hyperthyroidism

**Citation:** Grillo, A.; Barbato, V.; Antonello, R.M.; Cola, M.F.; Parati, G.; Salvi, P.; Fabris, B.; Bernardi, S. Arterial Stiffness in Thyroid and Parathyroid Disease: A Review of Clinical Studies. *J. Clin. Med.* **2022**, *11*, 3146. https://doi.org/10.3390/ jcm11113146

Academic Editor: Vanessa Bianconi

Received: 25 March 2022 Accepted: 30 May 2022 Published: 1 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

or hypothyroidism show several alterations, caused either by effects of thyroid hormones in the heart and in the vasculature or by cardiovascular risk factors (including blood pressure, dyslipidemia and inflammation), which may increase cardiovascular risk and lead to cardiovascular morbidity and mortality [7,8]. Similarly, patients with primary hyperparathyroidism often present cardiovascular abnormalities and an increased cardiovascular risk [9]. Less evidence supports the cardiovascular significance of subclinical hyper- or hypothyroidism or mild hyperparathyroidism, although detrimental cardiovascular consequences may affect these conditions as well [10–12].

In recent years, numerous clinical studies have investigated the relationship between thyroid and parathyroid diseases and arterial stiffness markers. New evidence explains the molecular and physiological pathways leading to vascular disease in conditions of hormonal excess or deficiency. The use of arterial stiffness measures has been proposed to improve clinical management of these conditions in several clinical studies.

In this review, we aimed to perform a literature search for the effects of thyroid and parathyroid hormone excess or deficiency on the structural and functional alterations of the arterial wall which leads to arterial stiffening, focusing on clinical studies conducted in humans.

#### **2. Material and methods**

A PubMed/MEDLINE search was performed to select peer-reviewed articles published from inception to 31 December 2021. The complete search string included inclusive keywords regarding arterial stiffness (e.g., "arterial stiffness", "arterial compliance", "pulse wave velocity", "PWV", "augmentation index", "AIx") and inclusive keywords regarding the evaluated hormones and their imbalance ("hypothyroidism" or "thyroid hormones" or "thyrotoxicosis" or "levothyroxine" or "Hashimoto" or "Graves" or "Basedow" or "dysthyroidism" "hyperparathyroidism" or "PTH"). Papers written in languages other than English, not pertinent with the present review, or whose full text was not available were excluded. The pertinent references were evaluated and eventually included in the final manuscript. We considered all papers in open-access and non-open access journals. The included papers were summarized and their results discussed in the text, offering an overview of current literature. Clinical studies in adult humans were organized in Tables (Table 1: hypothyroidism, Table 2: hyperthyroidism, Table 3: primary hyperparathyroidism). The manuscript was organized in the following major chapters: (1) Thyroid: hypo- and hyper-thyroidism; (2) Parathyroid.

#### **3. Results**

#### *3.1. Thyroid*

Our literature search identified 33 papers regarding dysthyroidism and its impact on arterial stiffness. Twenty-five studies evaluated arterial stiffness in patients with overt and/or subclinical hypothyroidism, while nine studies evaluated patients with overt and/or subclinical hyperthyroidism (one study evaluated patients with Hashimoto thyroiditis including both hyperthyroid and hypothyroid patients). From an overall evaluation, variable results about the impact of thyroid hormones on arterial stiffness are reported. In addition, arterial stiffness was assessed using different measurement methods, varying from ultrasounds to tonometry. Considered measures of arterial stiffness included pulse wave velocity (PWV), augmentation index and other surrogate measures. PWV is the velocity at which the pressure pulsation propagates through the circulatory system, and is defined as the distance travelled by the pulse wave divided by the time [2], and the carotid-femoral PWV is the gold standard for measuring stiffness of the aorta [3]. Variability in methodologies, together with the differences in study populations and in inclusion/exclusion criteria among studies, may explain the heterogeneity of results.

#### 3.1.1. Overt and Subclinical Hypothyroidism

The prevalence of subclinical hypothyroidism, defined as a serum thyroid stimulating hormone (TSH) level above the upper normal limit despite normal levels of serum free thyroxine (fT4), in the general population is 5–15% [13]. Of these, 2–5% progress to overt hypothyroidism [14]. Clinical and subclinical hypothyroidism are related to cardiovascular diseases, in particular atherosclerosis and ischemic heart disease [15]. Over the past years, the hypothesis that even in euthyroid subjects thyroid function may affect cardiovascular health has been supported by different authors [16–18]. A large Chinese population-based, cross-sectional study, including 812 euthyroid subjects without major cardiovascular risk factors, showed a significant and independent association of fT4 with baPWV in euthyroid subjects [19].

Regarding subclinical hypothyroidism, evidence observing an increase in arterial stiffness is conflicting and heterogeneous, both in methods used to evaluate arterial stiffness and in results. A number of small studies were conducted to evaluate arterial stiffness in subclinical hypothyroidism, in small cohorts of patients and with different methodologies. Earlier studies mainly considered the evaluation of brachial-ankle PWV, which was found to be increased [20] but not correlated to TSH [21]. Further studies confirmed the increase in aortic stiffness, measured either with brachial-ankle or heart-femoral PWV [21,22] and observed a decrease in stiffness after l-thyroxine replacement [23,24]. The increase in stiffness was confirmed by other methodologies, as the evaluation of β-stiffness index in the carotid artery [25–27], or the augmentation index [28,29], while the evaluation of global aortic distensibility gave contrasting results [30,31]. Considering the response to treatment, the augmentation index tended to reduce with levothyroxine replacement therapy [32].

More recent studies explored the relationship between the PWV in aorta and subclinical hypothyroidism. Three studies found an increase in aortic PWV in this condition [33–35], although one study conducted in a large sample and with gold-standard measurement (carotid-femoral PWV) [36] did not find an association between subclinical hypothyroidism and aortic stiffness. The CAVI, a blood pressure-independent stiffness index of the aorta, which is related to central and peripheral arterial stiffness and to 24 h blood pressure [37], was found to be increased in subclinical hypothyroidism [38], with a possible benefit after acute aerobic exercise [39].

Similarly to subclinical, most studies conducted in overt hypothyroidism found an increase in arterial stiffness, which has been evaluated across studies with a variety of biomarkers. Earlier studies evaluated markers related to reflected waves, as augmentation index [40,41]. Timing and magnitude of reflected waves was determined by hypothyroidism and improved after replacement therapy, indicating a positive effect on arterial stiffness [40]. Other studies focused on biomarkers of structural arterial stiffening, as β index [25,42,43], pulsatility index in carotid arteries [44] or brachial-ankle PWV [21]. One large study conducted in a large population, with brachial-ankle PWV, did not find a significant difference between hypothyroid and euthyroid subjects, which adds heterogeneity in the results [45]. More recently, studies that evaluated the gold-standard measure of aortic stiffness, the carotid-femoral PWV [33,44] or the heart-femoral PWV [35], found an increase in arterial stiffness markers in patients with overt hypothyroidism.

One study conducted in thyroidectomized patients on long-term replacement therapy, did not find an increase in arterial stiffness, suggesting that targeting TSH in the reference range does not seem to cause adverse cardiovascular effects [46].



**Table 1.** *Cont.*


**Table 1.** *Cont.*


*J. Clin. Med.* **2022**, *11*, 3146

**Table 1.** *Cont.* arterial stiffness index; cIMT: carotid intima-media

 thickness; CBP: central blood pressure.

#### 3.1.2. Overt and Subclinical Hyperthyroidism

Studies conducted in patients with overt hyperthyroidism suggest an increase in structural arterial stiffness biomarkers. An increase in β index in carotid arteries was found, which was reduced by antithyroid drug or radioiodine treatment [47,48]. Additionally, a marker of total arterial compliance (Pulse pressure/stroke volume) was found to be reduced in hyperthyroidism and normalized after beta-blockers therapy [49]. Considering a surrogate biomarker of central arterial stiffness derived from 24 h blood pressure monitoring, the ambulatory arterial stiffness index (AASI), one study did not find a significant difference between patients with overt or subclinical hyperthyroidism [50]. Data on subclinical hyperthyroidism (defined as a low or undetectable TSH, with normal fT3 and fT4) are scarce, considering that only one other work [51] has evaluated patients with this condition after thyroidectomy and l-thyroxine suppressive therapy, finding an increase in the β aortic stiffness by echocardiography.

Two studies evaluated the algorithm-derived PWV calculated from Mobil-O-Graph device in patients with hyperthyroidism, finding an increase in PWV in the office setting [52], which is not surprising considering that the PWV estimated by this method is strictly dependent on actual blood pressure values [53]. Interestingly, in another study the Mobil-O-Graph-derived PWV was not different in hyperthyroidism patients compared to blood-pressure matched controls [54], although the circadian profile of PWV was altered.

The evaluation of markers of reflected waves magnitude and the analysis of central pressure waves gave variable results among studies. In thyrotoxicosis, Obuobie et al. found a decrease in AIx despite an increase in central pulse pressure [40], thus suggesting a lowered central arterial stiffness. Conversely, Bodlaj et al. [55] found an increase in aortic AIx in hyperthyroidism patients with Graves' disease. A negative correlation between AIx and TSH, and a positive correlation between AIx and free thyroid hormones (fT3, fT4) was also described by Yildiz et al. [52]. Interpretation of these contrasting results probably resides in the close dependence of reflected waves timing with heart rate, which is acutely affected by thyroid hormones [56]. Alterations in functional markers of arterial stiffness can acutely affect the cardiovascular system, producing possible adverse organ damage, but the long-term effects of these transient changes remain to be elucidated.



**Table 2.** *Cont.* ASI: arterial stiffness index; PWV: pulse wave velocity; IMT: intima-media

 thickness.

#### *3.2. Parathyroid*

Primary hyperparathyroidism (pHPT) is another very common endocrine disorder, affecting between 0.4% and 11% of the population [57]. The highest rates are due to patients—mostly post-menopausal women—with mild pHPT, who exhibit inappropriately high levels of parathyroid hormone (PTH) and normal or only mildly elevated calcium levels [58]. Parathyroidectomy remains the treatment of choice for this condition, and it is recommended in patients with mild pHPT and normal calcium levels in case of young age, or signs of kidney and/or bone damage [59]. Although the kidney and the bone are the main targets of PTH actions, PTH exerts direct actions on the cardiovascular system too, including not only cardiac myocytes but also endothelial and vascular smooth muscle cells, leading to arterial remodeling [60–62].

Based on this background, several works have evaluated the effect of mild pHPT and parathyroidectomy on arterial stiffness. PTH has been found associated with arterial stiffness in the general population [63,64], and mild pHPT has been found associated with an increase of arterial stiffness, as assessed by AIx [65,66] as well as PWV [67–71], additionally to detrimental effects in other forms of vascular organ involvement as aortic intima-media thickness [72] and central blood pressure [73]. By contrast, the studies evaluating the effects of parathyroidectomy upon arterial stiffness have reported conflicting data regarding effects on arterial structure and function [74–78]. Nevertheless, in a recent meta-analysis we found that mild pHPT was associated with an increase of arterial stiffness, which was significantly reduced by parathyroidectomy [79]. These data are in line with the results of another meta-analysis analyzing the effects of parathyroidectomy upon left ventricular mass, where surgery was able to reduce it significantly [80]. These data suggest that surgery could improve arterial stiffness, as well as other signs of cardiovascular organ damage such as left ventricular hypertrophy, and reduce the cardiovascular risk profile in patients with mild pHPT.

Interestingly, hypoparathyroidism has also been associated with an increase of arterial stiffness [81,82], consistent with the higher risk of cardiovascular disease of patients affected by this condition.


 humans.


**Table 3.** *Cont.*


#### *J. Clin. Med.* **2022**, *11*, 3146

**Table 3.** *Cont.*


**Table 3.** *Cont.*

 pressure.

#### **4. Discussion**

#### *4.1. Thyroid*

Over recent years, the role of thyroid hormones in cardiovascular health has been widely studied. Several mechanisms of action of fT3, fT4 and TSH on arterial stiffness have been hypothesized. Thyroid hormones may affect the cardiovascular system by direct effects on arterial vessels (by regulating smooth muscle cells tone and endothelial function), on the heart (by influencing heart rate, rhythm, myocardial contraction and perfusion) or indirectly by influencing cardiovascular risk factors [7].

Focusing on effects on arterial vessels, thyroid hormones have both genomic and non-genomic mechanisms affecting vascular tone, by ion channel activation and regulation of specific signal transduction pathways.

Firstly, they have a cellular action on endothelium, causing production of nitric oxide via the phosphatidylinositol 3-kinase and serine/threonine protein kinase pathways [83]. Vasodilating effect on vascular smooth muscle cells and on resistance arteries results in widened pulse pressure and decreased systemic vascular resistance [84,85] in conditions of excess of thyroid hormones. Thyroid hormones thus lead to an increase in tissue oxygen consumption and in distending pressures. Higher mechanical stretch and altered perfusion patterns may lead to arterial vascular remodeling and to an increase in arterial, and in particular of aortic, stiffness [84,86].

Secondly, due to its positive inotropic and chronotropic effect, fT3 is directly responsible for acute hemodynamic changes [85], which affect the functional determinants of aortic stiffness [87]. The increase in heart rate and the increase in cardiac output driven by an excess of thyroid hormones lead to an altered hemodynamic adaptation. The resulting increase in mean arterial pressure and the shortening of left ventricular ejection may represent the major hemodynamic determinants of a functional increase in aortic stiffening. Additionally, an increase in heart rate worsens the pressure supply–demand balance, which may cause ischemia of the heart and of tissues with high blood flow supply. These conditions have been widely associated with aortic stiffening [88].

Thirdly, by acting on cardiovascular risk factors, thyroid hormones may indirectly lead to an increase in arterial stiffness. Notably the main determinant of aortic stiffness are blood pressure levels, which may be influenced by thyroid hormones. In hyperthyroidism, caused by inotropic effect on the heart and reduction in systemic vascular resistance, an increase in systolic and pulse pressure is often seen. Hypothyroidism is conversely associated with diastolic hypertension, induced by increased peripheral vascular resistance and by changes in circulating volume [89]. Variations in blood pressure components may acutely and chronically damage the arterial wall thus leading to arterial stiffness. Thyroid hormones are involved in lipid metabolism regulation. Hypothyroidism leads to an increased total cholesterol and LDL cholesterol [90], thus influencing atherosclerotic plaque burden. Non-HDL cholesterol is closely correlated with residual cardiovascular risk and arterial stiffness markers [91], although lipid levels are more linked with atherosclerotic vascular phenotypes, rather than to arteriosclerosis which is the hallmark of arterial stiffness. Thyroid hormones may be also linked to arteriosclerosis through different metabolic pathways: chronic inflammation, oxidative stress, insulin resistance [7].

An additional cause that may affect the promotion of cardiovascular disease in the presence of thyroid disease is the autoimmune process of disease per se [92,93]. Regardless of thyroid hormones level, autoimmune processes in the thyroid gland may cause a lowgrade inflammation which plays a relevant role in the atherosclerotic process and in the stiffening of arteries [94]. Subclinical inflammation has been associated with arterial stiffness in the general population [95] with a causative role played by an increase in cytokine levels (Interleukin 1 and 6, tumor necrosis factor-β) and reactive oxygen species in the degradation of elastin, migration of smooth muscle cells and increase in collagen in the arterial wall.

Regarding the effect of treatment of thyroid and parathyroid diseases in relation to thyroid disease, most studies observed an improvement in vascular markers after correction of hormone levels. In hypothyroidism, an improvement in AIx (which is mainly a functional parameter) followed correction with hormone replacement therapy. In hyperthyroidism, treatment with antithyroid drugs may restore both AIx and PWV levels, although the evidence is scarce.

We should consider an important limitation in our work, which is not having considered articles not in the English language. Considering the important heterogeneity of studies considering effects of thyroid hormones on stiffness and the complex interactions of factors determining vascular dysfunction, a further contribution is given by a recent meta-analysis, which found that both hypothyroidism and thyrotoxicosis are associated with an increase of aortic stiffness [96]. The scheme represented in Figure 1 represents the impact of thyroid disease on arterial stiffness and the effects and mutual interactions of all the previously discussed factors.

**Figure 1.** Pathophysiology of arterial stiffness in thyroid and parathyroid disease. RAAS: renin-angiotensin-aldosterone system. NO: nitric oxide. ROS: reactive oxygen species.

#### *4.2. Parathyroid*

In pHPT, a prolonged exposure to high levels of PTH may affect the vascular system in different ways, leading to vascular functional and structural changes and to arterial stiffness. Arterial remodeling is induced by PTH with a few mechanisms. A direct effect of PTH on endothelial and vascular smooth muscle cells was hypothesized [97] and confirmed by evidence of a stimulatory effect on nitric oxide synthase, which may contribute to vascular injury and arteriosclerotic progression through reactive oxygen species [60]. PTH may induce expression of other mediators associated with adverse vascular remodeling, as interleukin-6, receptor for advanced glycation end-products and vascular endothelial growth factor [61,98]. An effect mediated by increased intracellular calcium influx was also suggested [99], due to altered calcium metabolism in vascular smooth muscle cells, which may lead to increased arterial resistances [62]. Furthermore, PTH may also have systemic actions, by inducing the activity of renin-angiotensin-aldosterone system as well as the sympathetic nervous system [100,101], which have been notably associated with adverse effects on arterial stiffness [4]. Chronically elevated levels of circulating calcium may also mediate the association between elevated PTH and arterial stiffness. The association of calcium levels with stiffness was also demonstrated at a population level [102] and most likely mediated by the induction of vascular calcifications [103].

An elevation of blood pressure levels, which is a hallmark of pHPT, is associated with functional changes in arterial viscoelastic properties of the large arteries and an increase in the blood pressure-dependent component of arterial stiffness [104]. In experimental conditions, the infusion of physiologic doses of PTH in otherwise healthy adults, is known to produce an increase in blood pressure [105]. The induction of a chronic hypertensive state may also lead to the adverse arterial remodeling typical of hypertension and to arterial stiffness [106].

Regarding possible therapeutic approaches, surgical treatment of pHPT may improve arterial stiffness [79], and thus produce a favorable effect on cardiovascular risk profile in these patients.

The factors and the mechanisms influencing arterial stiffness in pPHT are schematically represented in Figure 1.

#### **5. Conclusions**

Our review of clinical studies prompts that thyroid and parathyroid diseases are able to affect arterial stiffness biomarkers, thus leading to possible adverse cardiovascular outcomes. In hypothyroidism, most studies agree that an increase in arterial stiffness is present in both subclinical and overt hypothyroidism, although there is no complete agreement among methodologies evaluating the different aspects of vascular stiffening in the arterial vasculature. Levothyroxine replacement therapy in most studies has shown to lead to an improvement in arterial stiffness markers. Regarding hyperthyroidism, few studies observed an increase in biomarkers of structural arterial stiffening, although a relevant methodological heterogeneity prevents a definite conclusion. The increased heart rate typical of hyperthyroidism significantly alters the blood pressure profile, affecting biomarkers of functional arterial stiffness. In primary hyperparathyroidism, an increase in arterial stiffness is evident also in the mild forms. Arterial stiffening is reversed by parathyroidectomy, suggesting a role of surgery in reducing cardiovascular risk. Regarding methods used to quantify arterial stiffness, most studies focused on PWV, which represents the gold standard method to evaluate this parameter and may thus represent the preferred method in the vascular evaluation of thyroid and parathyroid diseases.

Thyroid and parathyroid diseases are systemic diseases, characterized by an increase of cardiovascular risk, which present an increase in stiffness of the large arteries. Arterial stiffness measurement can be effectively used in the clinical evaluation of these conditions in order to quantify and possibly reduce the risk of cardiovascular events.

**Author Contributions:** Conceptualization, A.G. and S.B.; methodology, A.G.; data curation, V.B., R.M.A. and M.F.C.; writing—original draft preparation, A.G.; writing—review and editing, B.F., P.S. and G.P.; supervision, S.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** Research partially supported by the Italian Ministry of Health.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**


#### **References**


**Michele Colaci 1,2,\*, Luca Zanoli 2,3, Alberto Lo Gullo 4, Domenico Sambataro 2, Gianluca Sambataro 2, Maria Letizia Aprile 1, Pietro Castellino 2,3 and Lorenzo Malatino 1,2**


**Abstract:** (1) Background: Systemic sclerosis (SSc) is an autoimmune disease characterized by endothelial dysfunction and fibrosis of skin and visceral organs. In the last decade, attention has been focused on the macrovascular involvement of the disease. In particular, the observation of increased arterial stiffness represented an interesting aspect of the disease, as predictor of cardiovascular risk. (2) Methods: We recruited 60 SSc patients (52 ± 12 years old, 90% females) and 150 age/sex-matched healthy controls in order to evaluate both intima-media thickness of the right common carotid artery and arterial stiffness using the B-mode echography and the SphygmoCor system® tonometer. (3) Results: The carotid-femoral pulse wave velocity (PWV) was higher in SSc patients than in controls (8.6 ± 1.7 vs. 7.8 ± 1.5 m/s; *p* < 0.001), as was the carotid-radial PWV (7.8 ± 1.1 vs. 6.7 ± 1.4 m/s; *p* < 0.001). The intima-media thickness was higher in SSc than in controls (654 ± 108 vs. 602 ± 118 μm; *p* = 0.004). The other parameters measured at carotid (radial strain, Young's modulus, compliance and distensibility) all indicated that arterial stiffness in tension was more pronounced in SSc. Of interest, the direct correlation between PWV and age corresponded closely in SSc. Moreover, a significant difference between SSc and controls as regards the carotid parameters was evident in younger subjects. (4) Conclusions: SSc patients showed an increased arterial stiffness compared to healthy controls. In particular, an SSc-related pathologic effect was suggested by the more pronounced increase in PWV with age and lower values of carotid elasticity in younger SSc patients than in age-matched controls.

**Keywords:** pulse wave velocity; carotid distensibility; carotid strain; Young's elastic modulus; carotid compliance; systemic sclerosis; scleroderma; arterial stiffness

#### **1. Introduction**

Systemic sclerosis (SSc) is characterized by endothelial dysfunction and diffuse vasculopathy. Traditionally, the vascular involvement of SSc was considered mainly microvascular, leading to tissue ischemia. Raynaud's phenomenon, digital ulcers, as well as pulmonary arterial hypertension are typical SSc features that pathogenetically identify functional and structural alterations of microvasculature [1,2]. However, in the last decade, increasing evidence that large arteries are also affected was accumulated [3–5].

Arterial stiffness is a typical sign of arterial dysfunction, even in the absence of overt vascular abnormalities. The main pathogenetic feature is damage to elastin fibers of the aorta and its main branches, which are replaced by collagen overproduction [6]. Furthermore, the increase in vascular smooth-muscle cells and the infiltration of lymphocytes that secrete metalloproteinases and cytokines, such as transforming growth factor-ß, also contribute to arterial wall stiffening [6].

**Citation:** Colaci, M.; Zanoli, L.; Lo Gullo, A.; Sambataro, D.; Sambataro, G.; Aprile, M.L.; Castellino, P.; Malatino, L. The Impaired Elasticity of Large Arteries in Systemic Sclerosis Patients. *J. Clin. Med.* **2022**, *11*, 3256. https://doi.org/10.3390/ jcm11123256

Academic Editor: Paolo Salvi

Received: 30 April 2022 Accepted: 6 June 2022 Published: 7 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Arterial stiffness is well recognized as an independent cardiovascular risk factor; therefore, its assessment is clinically relevant [7–9]. The measurement of arterial stiffness may be easily performed by a simple, non-invasive evaluation of the pulse wave velocity (PWV). Several studies in the literature showed that SSc patients presented increased PWV if compared with healthy age/sex-matched controls, suggesting that the autoimmune disease is also responsible for macrovascular alterations [3,5,10–22]. Nonetheless, arterial stiffness is a well-recognized crossroad of different pathophysiological states, such as arterial hypertension, diabetes mellitus, smoking-related complications, and dyslipidemia [7–9,23]. In this study, we aimed to investigate the macrovascular involvement in SSc by measuring aortic stiffness together with the behavior of the so far unknown carotid elastic parameters, in order to identify additional alterations, if any, of carotid wall elasticity associated with SSc.

#### **2. Materials and Methods**

#### *2.1. Patients*

We recruited consecutive SSc patients classified according to the 2013 ACR/EULAR criteria [24] and referred to the Rheumatology Clinic of the Cannizzaro hospital or to the Rheumatology Unit of the ARNAS Garibaldi hospital, both in Catania, Italy. The SSc series was paired with a control group of the same ethnicity, matched for age and sex, randomly selected from a database including healthy outpatients referred to the clinic for cardiovascular diseases prevention, Internal Medicine Unit, Policlinico Rodolico—S.Marco, Catania, Italy.

SSc patients and controls affected by diseases associated with arterial stiffening, such as arterial hypertension, diabetes mellitus, chronic kidney disease, dyslipidemia, atherosclerotic diseases (i.e., myocardial infarction, stroke and peripheral arterial diseases), chronic heart failure and current or former smoking habits, permanent or paroxysmal atrial fibrillation, were excluded from this study.

Chronic vasoactive treatments were not interrupted for the study. However, in the case of prostanoid infusion therapy, the measurements of vascular parameters were performed at least 3 weeks after the last prostanoid administration.

All demographic, clinical, laboratory and instrumental characteristics of SSc patients had already been collected in their clinical records. These records included: clinical history, data from the physical examination, C-reactive protein, erythrocyte sedimentation rate, complete blood counts, indices of liver and renal function, autoantibody profiles, plasma NT-pro-BNP, spirometry, DLCO measurement, chest high-resolution computed tomography, ECG and echocardiography.

All patients gave their informed consent to the study, which was carried out in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments and approved by the Ethics Committee of Catania 1 (protocol n. 403, 13 March 2017).

#### *2.2. Methods*

All participants were studied in a vascular clinic by the same trained operator (L.Z.) blinded to patients' clinical histories. Vascular evaluations were performed in a quiet room with a controlled temperature of 22 ± 1 ◦C between 9:00 and 11:00 a.m. The patients were fasting and refrained from caffeine, alcohol and exercise before the study for at least 12 h.

After 15 min of rest in a supine position, brachial blood pressure was measured three times, 2 min apart, using a validated oscillometric device (Spacelabs 90217 ambulatory blood pressure monitor: Issaquah, WA, USA) [25]. The mean value of three measurements was used in this study.

The study of right common carotid artery was performed as previously reported [26]. Longitudinal B-mode (60 Hz, 128 radiofrequency lines) and fast B-mode (600 Hz, 14 radiofrequency lines) images of the artery 2 cm below the carotid bulb were obtained using a high-precision echo tracking device (MyLab One; Esaote, Maastricht, The Netherlands) equipped with a high-resolution (13 MHz) linear-array transducer. The systolic and diastolic internal diameters (Ds and Dd) and the intima-media thickness (IMT) were measured

at the right common carotid artery, according to Laurent et al. [27]. It is not excluded that IMT may be slightly different for measurements at the left side.

The right arm radial pulse wave profile was recorded by applanation tonometry (SphygmoCor system®, AtCor Medical, Sydney, Australia) after recalibration with brachial mean blood pressure (MBP) and diastolic blood pressure (DBP) in the contralateral arm and was used to calculate carotid pulse pressure (PP). Brachial MBP was calculated as brachial DBP + 1/3 × brachial PP.

The carotid PP was used for the calculation of carotid stiffness indexes [28].

The carotid-femoral and the carotid-radial PWV (cfPWV and crPWV, respectively) were measured with the SphygmoCor device, as previously reported [27], using the foot-tofoot velocity method, the intersecting tangent algorithm and the direct distance between the measurement sites [29]: cfPWV (m/s) = 0.8 × [carotid-femoral direct distance (m)/Δt]; crPWV (m/s) = 0.8 × [carotid-radial direct distance (m)/Δt]. The mean value of two consecutive recordings was used for this analysis. When the difference between the two measurements was ≥0.5 m/s, a third recording was performed, and the median value was used. The augmentation index (AIx%), an indirect measure of arterial stiffness, was calculated as the difference between the late systolic peak and the early systolic peak pressure, divided by the former.

The carotid distensibility, defined as the relative change in luminal area (ΔA) during systole for a given pressure change, was calculated as previously described [26], assuming the lumen to be circular and using the following equation: carotid distensibility = ΔA/A × carotid PP.

The carotid strain, defined as the relative change in the vessel diameter during systole, was calculated using the following equation: Strain = (Ds − Dd)/Dd, where Ds is the systolic internal diameter and Dd is the diastolic internal diameter.

The circumferential wall stress (CWS), which represents the tangential force that enlarges the vessel, was calculated as follows: CWS (kPa) = (mean BP × Dm)/2IMTm (where Dm and IMTm were the mean values of the internal diameter and the wall thickness during the cardiac cycle). The cross-sectional compliance coefficient (CC) represents the absolute change in lumen area during systole for a given pressure change and was calculated as follows: CC = stroke change in lumen area/local pulse pressure.

The incremental young elastic modulus (Einc), which represents the elastic properties of the material of the arterial wall (assuming that the vessel wall consists of a homogeneous material), was calculated as previously described [26], using the following equation: Einc = [3(1 + A/WCSA)]/DC, given that WCSA is the mean intima-media cross-sectional area.

#### *2.3. Statistical Analysis*

All continuous variables are presented as mean ± standard deviation (SD), after confirming their normal distribution by means of the Kolmogorov–Smirnov test; categorical variables are presented as a percentage value.

Clinical and hemodynamic variables were compared using analysis of variance (ANOVA) for continuous variables. Spearman linear regression analysis was performed to verify the existence of any significant correlation between two quantitative variables. *p* values < 0.05 were considered statistically significant.

The statistical analysis was performed using NCSS 2007 and PASS 11 software (Gerry Hintze, Kaysville, UT, USA).

#### **3. Results**

This cross-sectional study included 60 SSc patients and 150 age/sex-matched healthy controls. Table 1 shows demographic characteristics and findings obtained in this study.


**Table 1.** Data obtained in SSc patients and age/sex-matched healthy controls.

Data are shown as mean (standard deviation). SSc = systemic sclerosis; BMI = body mass index; SBP = systolic blood pressure; DBP = diastolic blood pressure; MBP = mean blood pressure; HR = heart rate; cfPWV = carotidfemoral pulse wave velocity; crPWV = carotid-radial pulse wave velocity; AIx% = augmentation index; CC = crosssectional compliance coefficient. The internal diameter was measured 2 cm below the bulb of the right common carotid artery. Significant *p* values have been reported in bold.

Fifty-six out of 60 (93.3%) SSc patients showed the limited skin subset, 27 patients (45%) presented digital ulcers in their clinical history, 32 (53.3%) interstitial lung disease and 3 (5%) pulmonary arterial hypertension (PAH). Furthermore, anti-centromere or anti-Scl70 autoantibodies were found in 35 (58.3%) and 22 (36.7%) SSc patients, respectively.

All patients were treated with calcium channel blockers for Raynaud's phenomenon. Moreover, 13 patients used endothelin-1 receptor antagonists for digital ulcer prevention or PAH, while 35 were treated with monthly infusion of prostanoids.

No significant difference between SSc patients and controls as regards the body mass index (BMI) was noted. The values for blood pressure confirmed that no patients with arterial hypertension were included.

The evaluation of the right common carotid artery showed an increase in IMT in SSc patients compared with controls, but a similar internal diameter. Moreover, stiffness of the arterial wall was significantly increased in SSc patients compared with controls (Table 1). In particular, PWV (measured both as carotid-femoral PWV and radial-femoral PWV) was clearly higher in SSc patients than in controls, indicating a diffuse vascular stiffening involving both elastic and muscular arteries in SSc patients.

Carotid-femoral PWV > 10 m/s was found in 13 out of 60 SSc patients and 11 out of 150 controls (21.7% vs. 7.3%; *p* = 0.003). These 13 SSc patients had a mean age of 63.9 (range 52–73) years, whereas the mean age of the entire SSc group was 52 ± 12 years (*p* = 0.0014). No other SSc characteristics distinguished this patient subgroup from the others.

Interestingly, PWV was directly and more closely correlated with age in SSc patients compared with healthy controls, with a steeper regression line, suggesting that SSc is responsible for an accelerated loss of elastic properties of large vessels (Figure 1).

**Figure 1.** Correlations between carotid femoral PWV and age in SSc patients versus healthy controls.

Macrovascular stiffness was consistent with structural changes of the arterial wall, as shown by the higher IMT of the common carotid artery in SSc than in controls (Table 1, Figure 2), although a similar pattern of direct association between PWV and IMT was demonstrated in both SSc patients and healthy subjects (Figure 3).

**Figure 3.** Correlations between carotid femoral PWV and IMT in SSc patients versus healthy controls.

An increased carotid elastic modulus index, along with reduced radial strain, compliance and distensibility, were also observed in SSc patients, compared with controls (Table 1). As shown in Figure 4, in younger SSc patients, carotid wall presented a reduction in its elasticity in comparison with healthy individuals. These findings suggest a direct effect of SSc on arterial wall, leading to a precocious arterial aging. In fact, carotid elastic changes

**Figure 4.** Carotid wall elasticity evaluation in SSc patients versus healthy controls.

Vasoactive therapies did not seem to influence our findings. Finally, except for SSc patients' age, no significant correlations between SSc characteristics and vascular parameters were found.

#### **4. Discussion**

In this study, we evaluated the arterial stiffness of large vessels in a group of SSc patients compared with healthy controls. PWV, an index of aortic stiffness, was higher in SSc patients than in controls, in accordance with the literature [3–6,10–22]. In particular, cfPWV was directly and more closely correlated with age in patients than in controls, and was positively associated with IMT. We also demonstrated that vascular stiffening involved both muscular and elastic arteries of SSc patients, as shown by the higher values for both crPWV and cfPWV (Table 1).

In a previous study [18], a cutoff of 9 m/s for cfPWV was considered in order to identify patients with a significant increase. At variance, in the present study, a cutoff of 10 m/s was considered, according to the ESC guidelines for arterial hypertension [30]. However, it would be more correct to consider the absolute values of PWV without choosing a cutoff, because no data are so far available on the clinical and prognostic significance of individual PWV values in rheumatic diseases.

Of note, the increase in carotid Young's elastic modulus and the reduction in radial strain, compliance and distensibility, provided for the first time in SSc evidence further corroborating the concept that a complex dysfunction of macrovascular arterial bed occurs in SSc patients. The carotid stiffening was particularly evident in younger SSc patients, thus emphasizing its association with SSc.

Vascular dysfunction of SSc involved both elastic (i.e., carotid artery and aorta) and muscular (i.e., brachial artery) arteries. Moreover, in SSc patients we found the coexistence of stiffening and thickening processes. The increase in the Young's elastic modulus suggested that alterations of the bioelastic material occur in the arterial wall in SSc. As a consequence of the stiffening process, the radial strain and the distensibility of the common carotid artery were reduced. Of interest, despite the increased common carotid artery IMT in SSc compared with control subjects, the internal diameter was comparable, thus suggesting that, in SSc patients, the thickening of large elastic arteries may proceed in parallel with the enlargement of the arterial wall. Further studies may be needed to confirm this hypothesis.

The increase in arterial stiffness in SSc is a crucial point in the assessment of the cardiovascular risk in SSc patients [5,10,12], providing additional predictive value above IMT measurement for the association with high risk of cardiovascular disease [31]. In our previous study [32] including female patients affected by SSc or diabetes mellitus, the incidence of established coronary artery disease was lower in SSc patients than in diabetics, but similar between the two groups when subjects older than 65 years were considered. This would mean that, when SSc patients grow older than 65 years, the SSc phenotype may become equivalent to that of diabetes.

Rheumatic diseases, such as rheumatoid arthritis (RA), were widely studied as independent risk factors for the increase in cardiovascular risk, due to the chronic inflammatory state [33–35]. In the literature, increased incidence of major adverse cardiovascular events was largely demonstrated for RA patients, in association with long disease duration and scarce disease control [33–35]. Moreover, it is well known that the chronic inflammatory state leads to endothelial activation/lesion and surface expression of adhesion molecules for migration of leukocytes [36]. The inflammatory infiltration of atherosclerotic plaques and progression of vascular injury through the inflammatory pathway may be considered as the *primum movens* of premature atherosclerosis in RA patients [33–36].

Differently from other rheumatic diseases, SSc is characterized by endothelial dysfunction even from its early phase [1,2]. Raynaud's phenomenon, or the onset of digital ulcers, may be considered a direct consequence of the imbalance between endothelium-derived relaxation and vasoconstrictive factors. Endothelial dysfunction is considered an early marker

of atherosclerosis [37], so, therefore, we may assume that SSc could contribute *per se* to the development of accelerated atherosclerosis. Consistently, our findings (Figures 1 and 4) suggested that SSc could be involved in the pathogenetic chain of arterial stiffening. In this respect, in a previous study [38], we raised a suggestive hypothesis regarding aortic wall damage in SSc patients. In particular, we found a significant association between nailfold videocapillaroscopy abnormalities and aortic root dilation. These findings could lead us to hypothesize that a microvascular dysfunction of the aortic *vasa vasorum* may contribute to the early damage of the aortic wall. Then, fibrosis, due to collagen overproduction replacing elastin fibers, completes the aortic remodeling. The absence of histological studies on aortic wall from SSc patients, unfortunately, so far does not allow confirmation of this hypothesis.

Overall, the presence of a macrovascular involvement in SSc is widely accepted, even though the large heterogeneity of the methodologies used in the literature has so far raised some inconsistencies [3]. However, the presence of SSc-specific endothelial dysfunction makes plausible the development of an SSc-related macrovascular alteration.

In clinical practice, SSc patients may be affected also by systemic hypertension, diabetes, and dyslipidemia or could be smokers. Therefore, the coexistence of several cardiovascular risk factors makes the identification of the intrinsic role of SSc more difficult. For this reason, in this study, we focused on SSc patients without other conditions known to be traditional cardiovascular risk factors.

In our study, we did not find correlations between arterial stiffness parameters and SSc patients' features, besides patients' age. It is probable that the relatively low number of cases included in our study (i.e., few cases with diffuse skin subset or PAH) did not allow us to find specific correlations found in previous studies [12,39]. Further larger cohort studies are needed to clarify this issue.

In conclusion, the evaluation of carotid elasticity, carried out for the first time in our study, could facilitate a better understanding of the specific characteristics of macrovascular abnormalities in patients with SSc.

This study, however, has limitations. In fact, some unknown or underestimated factors might have influenced our findings. For instance, many drugs used in SSc are vasoactive substances that could counteract the long-term evolution of arterial stiffness. Furthermore, immunosuppressive agents may inhibit leukocyte activities also in the arterial wall, thus influencing the natural progression of disease. Therefore, considering the high number of confounders influencing endothelial function and arterial alterations, protocols for the evaluation of cardiovascular risk in SSc should be designed within very large, multicenter, prospective studies.

**Author Contributions:** Conceptualization, M.C., L.Z., P.C. and L.M.; methodology, M.C. and L.Z.; formal analysis, L.Z.; investigation L.Z.; resources, A.L.G. and M.L.A.; data curation, D.S. and G.S.; writing—original draft preparation, M.C.; writing—review and editing, L.Z., D.S., G.S., P.C. and L.M.; supervision, P.C. and L.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of Catania 1 (protocol n. 403, 13 March 2017).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.

**Acknowledgments:** This study is part of the University Research plan 2016–2018, project #1A "Molecular and clinical-early instrumental markers in metabolic and chronic-degenerative pathologies" by Dept. of Clinical and Experimental Medicine.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Davide Agnoletti 1,2,\*, Federica Piani 1,2, Arrigo F. G. Cicero 1,2 and Claudio Borghi 1,2**


**Abstract:** The gut microbiota is a critical regulator of human physiology, deleterious changes to its composition and function (dysbiosis) have been linked to the development and progression of cardiovascular diseases. Vascular ageing (VA) is a process of progressive stiffening of the arterial tree associated with arterial wall remodeling, which can precede hypertension and organ damage, and is associated with cardiovascular risk. Arterial stiffness has become the preferred marker of VA. In our systematic review, we found an association between gut microbiota composition and arterial stiffness, with two patterns, in most animal and human studies: a direct correlation between arterial stiffness and abundances of bacteria associated with altered gut permeability and inflammation; an inverse relationship between arterial stiffness, microbiota diversity, and abundances of bacteria associated with most fit microbiota composition. Interventional studies were able to show a stable link between microbiota modification and arterial stiffness only in animals. None of the human interventional trials was able to demonstrate this relationship, and very few adjusted the analyses for determinants of arterial stiffness. We observed a lack of large randomized interventional trials in humans that test the role of gut microbiota modifications on arterial stiffness, and take into account BP and hemodynamic alterations.

**Keywords:** vascular ageing; arterial stiffness; central hemodynamics; pulse wave velocity; gut microbiota; gut microbiome; inflammation; oxidative stress

#### **1. Introduction**

For several decades, cardiovascular disease (CVD) has been the leading cause of death worldwide. CVD is mainly driven by high blood pressure (BP), which causes damage several target organs. However, there is some evidence that cardiovascular risk due to hypertension is not fully restored by antihypertensive treatment, leading to the concept of residual cardiovascular risk [1]. This is in line with the hypothesis that underlying factors drive both CVD and hypertension, and precede the clinical evidence of the disease, even before hypertension is established. Indeed, subclinical local and systemic inflammation could be one of the main drivers of target organ damage. One of the underlying factors contributing to increased cardiovascular risk is arteriosclerosis, a process of progressive stiffening of the arterial tree associated with arterial wall remodeling, which can precede hypertension and organ damage during the life course. This process has recently been mentioned as "vascular ageing" [2].

The main function of the arterial system is to dampen the pulsatility induced by the stroke volume during the systole.

This is of pivotal importance as organs with high flow and low resistance (e.g., the heart, brain, kidney) are prone to the side effects of increased pulsatility [3]. The large arteries, mainly the aorta, due to their elastic properties, contribute to preserving continuous

**Citation:** Agnoletti, D.; Piani, F.; Cicero, A.F.G.; Borghi, C. The Gut Microbiota and Vascular Aging: A State-of-the-Art and Systematic Review of the Literature. *J. Clin. Med.* **2022**, *11*, 3557. https://doi.org/ 10.3390/jcm11123557

Academic Editors: Paolo Salvi, Andrea Grillo and Emmanuel Androulakis

Received: 25 April 2022 Accepted: 18 June 2022 Published: 20 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and low pulsatile flow into target organs. The progressive stiffening of the arterial tree from the center to the periphery, together with the phenomenon of pressure wave reflection, maintains and even amplifies blood pressure in order to guarantee correct organ perfusion. Eventually, the peripheral resistances preserve the target organ microcirculation from pulsatility. During the physiological process of ageing, the arterial tree becomes more and more stiff, due to a change in the elastin–collagen ratio and to the deterioration of the arterial extracellular matrix. This leads to increased pulsatility at the peripheral level, where target organs may be injured. Early vascular ageing (EVA) is an attempt to describe early vascular modifications, leading to a stiffer arterial tree as compared with the normal aging process, and conferring higher cardiovascular risk. Arterial stiffness has become the preferred marker of EVA and is easily estimated by the measurement of pulse wave velocity (PWV). The stiffness of an arterial segment can be estimated either "locally" (e.g., in the carotid artery by the doppler ultrasound technique, or in the ascending aorta by magnetic resonance) or "globally" (e.g., the entire aorta by the carotid-femoral PWV [cfPWV] by arterial tonometry). Aortic PWV has become the gold standard for the estimation of aortic stiffness and is an established marker of cardiovascular morbidity and mortality [4,5].

The gut microbiota has emerged as a critical regulator of human physiology, and deleterious changes of its composition and function, commonly referred to as dysbiosis, have been linked to the development and progression of numerous disorders, including cardiovascular diseases. In particular, both direct and indirect roles of gut microbiota have been described on blood pressure regulation and vascular inflammation and stiffening.

The human gastrointestinal tract harbors a vast array of microorganisms that significantly affect host nutrition, metabolic function, gut development, and maturation of the immune system and intestinal epithelial cells [6–8]. In the present review, we refer to "microbiota" as the composition of the whole gut bacterial microorganism. Overall, the microbiota comprises 5 major phyla and approximately more than 1000 species in the large intestine. The gut microbiota promotes digestion and food absorption for host energy production, whereas in the colon, complex carbohydrates are digested and subsequently fermented into short chain fatty acids (SCFAs) such as n-butyrate, acetate, and propionate. The resulting SCFAs seem to regulate neutrophil function and migration, reduce colonic mucosal permeability, inhibit inflammatory cytokines, and control the redox environment in the cell. From a physiological point of view, the main producers of SCFAs belong to the Firmicutes phylum, the single largest grouping of gut bacteria. The Clostridia class in the Firmicutes phylum includes diverse bacteria of medical, environmental, and biotechnological importance. In particular, butyrate and butyrate-producing microbes have been associated with gastrointestinal health in humans and various animal species, and in the human gut are predominately members of clostridial clusters IV (phylum Firmicutes, class Clostridia, genera: *Faecalibacterium*, *Oscillibacter*, *Ruminococcus*, ... ) and XIVa (phylum Firmicutes, class Clostridia, genera: *Coprococcus*, *Roseburia*, *Clostridum\_g24*, ... ) [9]. *Clostridium* clusters XIVa and IV represent the gut's predominant bacteria, accounting for 10–40% of the total bacteria [10]. *Akkermansia muciniphila*, the only member of the Verrucomicrobia phylum, is one of the SCFA-producing bacteria and represents 3–5% of total faecal microbes. A large number of studies has shown that the abundance of *Akkermansia* in the gut is correlated with several health benefits in humans [11]. These beneficial effects are related to the ability of the bacterium of maintaining the mucus thickness and the integrity of the intestinal barrier, providing energy sources (SCFAs) for mucin-producing goblet cells [11]. Studies have shown a relationship between low *A. muciniphila* abundance and increased occurrence of inflammatory metabolic diseases, such as diabetes, obesity, and inflammatory bowel disease [12], which are associated with epithelial gut damage and high permeability.

In investigating gut microbiota characteristics, it is important to explore (i) α diversity: the microbial diversity at the smallest spatial scale (intra individual), assessed by the Shannon or Simplon index; (ii) ß diversity: the microbial diversity at the landscape scale (inter-individual diversity within the same population); (iii) Richness: the total number of species in the unit of study, measured as operational taxonomic units (Chao1 index); (iv)

Firmicutes:Bacteroidetes (F:B) ratio: an index of balance between the two most relevant bacterial families. Furthermore, a high-resolution analysis of the bacterial community down to the species level, and a functional profiling for the assessment of the most represented genes/metabolic pathways and their relative abundance is obtained by a metagenomics analysis, whereas metatranscriptomics is exploited to elucidate which microbial genes annotated in metagenomes are actually transcribed and to what extent. The abundance of health-promoting or detrimental microbiome-derived metabolites (e.g., lipidome and metabolome profile) is assessed by metabolomics analysis.

The mechanisms underlying the association between vascular ageing and the gut microbiota are presented in Figure 1.

**Figure 1.** Schemes of the intercorrelation between environmental and biological mechanisms, gut microbiota, and vascular ageing.

#### **2. State-of-the-Art Review**

#### *2.1. Microbiota and Hypertension*

Data from literature indicate that fruit and vegetable intake is associated with both lower BP values and reduced cardiovascular mortality [13,14], despite a high fat intake [15]. Several micronutrients have been investigated as potential proactive means to achieve such results. Fibers interact with gut microbiota, stimulating growth of specific bacterial phyla, and fiber intake has been associated with lower cardiovascular and all-cause mortality [16].

Among the metabolites produced by the gut microbiota, alanine, n-methylnicotinate, hippurate, and formate have been associated with BP levels. While formate and hippurate are negatively correlated with BP, alanine (produced mainly under a carnivore diet) is associated with higher BP [17]. 4-hydroxyhippurate production by polyphenols microbial metabolism was associated with higher risk of developing hypertension at 10 years (1.17 [95%CL 1.08–1.28]) in a black normotensive population [18]. Subjects with prehypertension or stage 1 hypertension underwent a randomized crossover trial with three diet interventions: carbohydrate-rich, protein-rich, or mono-unsaturated fat-rich diet for 6 weeks each. Urinary metabolites were associated with blood pressure, in particular: proline-betaine (derived from carbohydrate, protein and fat-rich diets), 4-cresyl sulfate and phenylacetylglutamine (derived from the fat-rich diet), N-methyl-2-pyridone-5-carboxamide (derived from the carbohydrate-rich diet) were inversely associated with BP, while carnitine (derived from the protein-rich diet) and hippurate (derived from the

carbohydrate-rich diet) were positively associated with BP levels [19]. These results show that different diet interventions, addressing macronutrient content, are associated with distinct hemodynamic effects.

From the available data, hypertension is associated with gut microbiota dysbiosis, characterized by an increased F:B ratio, as well as a drastic decrease in acetate-, butyrate-, and an accumulation of lactate-producing microbial populations. Treatment with an oral minocycline dose, which interferes with microbial growth, has been found to attenuate hypertension and produce beneficial effects on dysbiosis in a rat model [20]. Otherwise, primary alterations in the gut microbiota may elicit hypertension, as was highlighted by an animal study where hypertensive, stroke-prone rats (SHRSP) presented a dysbiotic gut. The main result was that, after transplantation of the SHRSP microbiota in normotensive rats, the authors observed a significant increase in systolic blood pressure (SBP) [21].

If the relationship between gut microbiota and hypertension is so far well described, it is not easy to understand by which pathophysiological mechanisms do the microbial environment and its metabolites regulate BP levels. One of the main characters in the scene is the group of SCFAs.

#### *2.2. The Role of SCFAs*

For more than two decades, SCFAs, mainly acetate, propionate and butyrate, have been found to be involved in dilatation in rat tail arteries and human colonic resistance arteries in a concentration-dependent way [22–24]. More recent findings suggest that propionate enhances renin release from juxta glomerular cells, and reduces BP levels in hypertensive mice, as well as in wild mice [25]. In a model of deoxycorticosterone acetate (DOCA)-salt mice, chronic acetate intake was associated with lower BP levels, together with reduced myocardial fibrosis and hypertrophy, and better cardiac function [26]. Interestingly, in the same study, the authors describe that fiber intake could modify the gut microbiota, increasing acetate-producing bacteria, and that fiber and acetate supplementation improved dysbiosis. The positive biological effects of fibers and acetate were ascribed to: (i) the downregulation of cardiac and renal genes for early-growth-response-protein-1 (Egr1), involved in myocardial hypertrophy, fibrosis, and inflammation; (ii) the downregulation of the renin–angiotensin system in the kidney [26].

Several metabolite-sensing G-protein-coupled receptors (GPCRs) have been found to bind with SCFAs, and are important for gut health and immune response regulation [27]. Impaired signaling of these receptors may occur due to excess fat or sugar intake, and could be involved in the deterioration of the intestinal barrier, with lipopolysaccharide (LPS) translocation and subsequent local and systemic inflammation (leaky gut syndrome, see paragraph 2.6) [27,28]. In addition, SCFAs could influence cellular gene expression, binding to diverse histone deacetylases. One of the most important GPCRs is the olfactory receptor 51E2 (named Olfr78 or OR51E2), which is found mainly in arterial smooth muscle cells, autonomic nerves, and in juxtaglomerular apparatus, and binds acetate and propionate, resulting in increased renin release. In Olfr78-/- mice, both BP and plasma renin levels were found to be lower than in wild mice. Indeed, a challenge of propionate in wild mice resulted in a dose-dependent BP lowering, while in Olfr78-/- mice, an accentuated BP-lowering effect was observed with a very low propionate dose, indicating that Olfr78 activation antagonizes the acute hypotensive effect of propionate. This suggests that propionate receptors other than Olfr78 regulate BP levels [25]. GPR41 is found in the vascular endothelium and mediates vasodilation through SCFA stimulation. GPR41-/ mice are not prone to the BP-lowering effect of propionate, and present both higher BP levels and arterial stiffness [29]. This realizes a complex schema where the same metabolite (propionate) can increase renin release, but can also exert a hypotensive effect, depending on the receptor it activates.

According to the data presented, it appears that a biological link exists between microbiota composition, the production of SCFAs, and blood pressure regulation. In particular, beside the effect of blood pressure, it is worth highlighting that SCFA production is almost constantly associated with a beneficial microbiota composition, and that higher butyrate levels are mostly associated with positive local and systemic effects. They include: improvement of colonocytes health; reduction of neutrophils migration; increased tight junction protein, and an anti-inflammatory effect.

#### *2.3. Maternal Heritage and Genetic/Epigenetic Regulation*

A study on mice fed with a high-fiber diet or acetate during pregnancy showed that acetate inhibited the histone deacetylaese-9, which resulted in the downregulation of atrial natriuretic peptide (ANP) in the offspring [30]. In another mouse model, mice fed with fiber or acetate presented an improved heart and kidney function, in particular through genetic regulation of fibrosis, fluid absorption, the renin angiotensin system, and inflammation pathways (e.g., by downregulation of transcription factor Egr1, IL-1, Rasal1, Cyp4a14, Cck) [26].

In offspring exposed in utero to maternal obesity, a mouse study found a specific methylation profile in the resistance of mesenteric arteries, which was associated with vascular remodeling and impaired vasodilation [31].

Overall, from the scarce information available, it seems that maternal characteristics and behavior influence offspring in terms of both inflammation pathways and hemodynamics.

#### *2.4. Inflammation and Immune System in Hypertension*

Hypertension is associated with immune activation. Increased numbers of central memory CD8+ T cells, activated CD8+ T cells producing Interferon-gamma (IFNγ) and tumor necrosis factor (TNF), and TH17 cells have been reported in patients with hypertension [32]. Monocytes from patients with essential hypertension are preactivated, producing greater amounts of IL-1β, TNF and IL-6 following ex vivo stimulation with angiotensin II or LPS than monocytes from healthy controls [33,34]. A major goal of hypertension treatment is the prevention of end-organ damage. A substantial portion of the vascular, renal, cardiac, and brain damage and dysfunction that accompanies hypertension is mediated by inflammation within these target organs. The innate and adaptive immune responses are critical to the development of hypertension and its consequences [35]. Adaptive immunity activation of both T-cells and B-cells is initiated early in the course of the disease and greatly contributes to important pathogenetic changes, through release of pro-inflammatory cytokines and antibodies [34].

While hypertension and aging are established factors contributing to arterial stiffness, the role of inflammation in the stiffening of the arteries is less well understood.

Arterial stiffness is associated with increased production of reactive oxygen species [36] and proinflammatory cytokines [37]. Furthermore, C-reactive protein itself may play an active role in mediating arterial stiffening, by inducing endothelial dysfunction. The increased vascular inflammation increases vascular fibrosis, smooth muscle cell proliferation, and impair endothelial-mediated vasodilation, which subsequently leads to increased arterial stiffness [38]. Oxidative stress appears to play a role in the pathogenesis of arterial stiffness, as oxidative injury may result in increased vascular inflammation and increased cellular proliferation, which may subsequently lead to impaired arterial elasticity [39].

Multiple studies have shown elevated indices of arterial stiffness in subjects with primary inflammatory disorders, and prospective studies (including 2 RCTs) have demonstrated a reduction in arterial stiffness following treatment with anti-TNF and other antiinflammatory agents [40–44].

The gut microbiota is thought to modulate immune and inflammatory responses. Germ-free mice present lower levels of TH7 and Treg, and a higher TH2/TH1 ratio than wild mice, which is associated with hypertension development. Furthermore, GPCR are also localized in immunity cells, so that SCFAs are able to interact with and activate them [45]. From the evidence available so far, it is clear that: (i) Inflammatory dysregulation occurs during human hypertension, together with immune system activation; (ii) Vascular ageing, as estimated by arterial stiffness, is strictly influenced by inflammation and oxidative stress, and could precede the development of hypertension; (iii) The gut microbiota can regulate immunity cells and the inflammation response.

#### *2.5. The Role of Trimethylamine-N-Oxide*

The pivotal role of the gut microbiota in cardiovascular disease is highlighted by the data concerning trimethylamine-N-oxide (TMAO) and its link to both microbiota and atherosclerosis. TMAO is a molecule transformed from the metabolism of choline by the gut microbiota, starting from dietary phosphatidylcholine. The main sources of phosphatidylcholine are meat, eggs, and foods with a high cholesterol content. The final TMAO plasma concentration depends on diet, gut microbial composition, drugs, and the activity of the liver flavin monooxygenase [46,47].

Plasma TMAO concentration correlates with incidents of major adverse cardiovascular events in patients with acute coronary syndrome [48]. TMAO levels were also associated with ageing, systolic blood pressure, and cfPWV, independent of cardiovascular risk factors [49]. TMAO dietary supplementation increased arterial stiffness in both young and old mice, impaired aortic wall intrinsic mechanical stiffness, and increased aortic wall concentration of advanced glycation end-products [49]. Interestingly, in mice, TMAO infusion amplified the angiotensin II effect in increasing BP [17,50], but did not affect BP levels in normotensive rats, so that it is unclear whether TMAO is proatherogenic or a marker of atherosclerosis [50]. In any case, even if the results of an experimental study show an obligate role for intestinal microbiota in the generation of TMAO from the dietary lipid phosphatidylcholine [51], it is still not clear whether specific patterns of microbiota composition would be associated with different levels of TMAO production. This issue is highlighted by the results of a recent meta-analysis showing that supplementation with probiotic *Lactobacillus rhamnosus GG* was the most efficient in reducing the plasma TMAO level in both humans and animals [52].

#### *2.6. The Role of Lipopolysaccharides*

LPSs are found mainly in the gut lumen, as they form the outer membrane of gramnegative bacteria. In situations of increased permeability of the epithelial gut barrier, LPS transmigrates into the blood stream, and it binds to the toll-like receptor-4 by means of CD14 complex. This stimulation induces the release of several proinflammatory cytokines by the NF-kB pathway, with activation of the immune and inflammatory response [53].

Even if LPS enhances the atherosclerotic process and boosts the formation of unstable plaques, its link with hypertension is still debated.

Gut microbiota dysbiosis easily induces both an increase of lumen LPS and a deterioration of the gut epithelial barrier with amplification of tight junction permeability, resulting in the transmigration of LPS; this phenomenon is called the "leaky gut syndrome", and is one of the promoters of systemic inflammation driven by the gut microbiota.

#### *2.7. The Role of Salt*

As shown above, different dietary patterns, including a diverse composition of fiber, fructose, and fat, modulate the gut microbiota with various effects on inflammation, the immune system, and BP levels. In this domain, dietary sodium intake has emerged as an important player for its interaction not only with BP levels, but also with gut microbiota, inflammation, and the immune system [54,55].

Sodium and water absorption are regulated by the sodium–proton exchanger 3 (NHE3), which is found both in the gastrointestinal tract and the renal proximal tubule [56]. In a model of spontaneously hypertensive rats, the inhibition of NHE3 resulted in increased fecal content of sodium and water, decreased urinary sodium excretion, and lower BP levels [57]. NHE3-ko mice presented altered gut microbiota, with a beneficial decrease of the F:B ratio [58], but NHE3 deficiency was also found to induce irritable bowel syndrome, with gut dysbiosis [59], making the results difficult to interpret. High salt intake alters gut microbiota, inducing low microbial diversity [60] and the depletion of *Lactobacillus*

spp., which is restored after normalization of the sodium dietary content [61]. Moreover, supplementing *Lactobacillus* reduced BP levels and TH-17 cell activation in mice fed with a high-salt diet [61]. Salt intake also affects the Clostridial order, with a reduction of several genera, and an increase in Christensenellaceae, Corynebacteriaceae, Lachnospiraceae, Ruminococcaceae and *Oscillospira*, with exacerbation of colitis [60,62]. It is worth noting that the biological result of the modifications of the genera abundances in the gut is not always predictable, as it depends also on bacterial species-to-species interaction and on activation of specific genes. Indeed, mice fed with a high-salt diet show a higher abundance of *Roseburia*, a butyrate-producing species [60], but lower butyrate production, perhaps due to the loss of interaction with *Lactobacillus* spp., which is depleted [62].

High salt intake can increase several proinflammatory cytokines, such as interleukin (IL)-6 and IL-23 [63], and may activate TH-17 cells with production of IL-17 and IL-22 [55,61], which are associated with the development of hypertension [64]. Interestingly, as shown before, these activation mechanisms are likely mediated by the gut microbiota [45,65].

From the data presented here, it emerges that salt intake is associated with several biological mechanisms related to disbiosis and inflammatory pathways.

#### *2.8. Microbiota and Exercise*

A relatively recent observational study comparing the fecal bacterial profile of elite male rugby players with non-athlete healthy subjects [66] showed significant differences between the two groups; in particular, athletes had lower levels of Bacteroidetes and greater amounts of Firmicutes than the controls. After analyzing the gut microbiota composition of the participants of the American Gut Project, it was concluded that increasing exercise frequency from never to daily causes greater diversity among the Firmicutes phylum (including *Faecalibacterium prausnitzii* and species from the genus *Oscillospira*, *Lachnospira*, and *Coprococcus*), which contributes to a healthier gut environment. In the limited studies available in animal models, exercise in rats was associated with higher Bacteroidetes and lower Firmicutes in fecal matter, whereas the cecal microbiota following 6 weeks of exercise activity presented a greater abundance of selected Firmicutes species and a lower abundance of *Bacteroides/Prevotella* genera. Similarly, at the phyla level, exercise reduced Bacteroidetes, while it increased Firmicutes, Proteobacteria, and Actinobacteria in mice. Even if data on bacterial genera are lacking, this microbiota composition could represent a benefical adaptation to exercise. Rats that participated in voluntary running exercise had increased colonic butyrate concentrations compared to sedentary rats, due to higher levels of butyrate-producing bacteria from the Firmicutes phylum (SM7/11 and T2-87) in their cecum. Hsu et al. investigated the influence that intestinal microbiota has on endurance swimming time in specific pathogen-free (SPF), germ-free (GF), and *Bacteroides fragilis* (BF) gnotobiotic mice. They found that the antioxidant capacity was deeply different in the three mice models, as serum levels of glutathione peroxidase (GPx) and catalase (CAT), two major antioxidants able to convert hydrogen peroxyde into water, were greater in SPF than GF mice. Additionally, serum superoxide dismutase (SOD) activity, pivotal for the clearance of superoxyde radicals, was lower in BF than SPF and GF mice. The authors found that endurance swimming time was longer for SPF and BF mice than GF mice, suggesting that the gut microbiota composition is crucial for exercise performance, and could also be linked to the activity of antioxidant enzyme systems. The types and amount of SCFAs produced by gut microorganisms are determined by the composition of the gut microbiota and the metabolic interactions between microbial species, but also by the amount, type, and balance of the main dietary macro- and micronutrients [67].

Exercise training seems to have a role in gut microbiota composition and function, and the bacterial patterns may evolve during exercise, potentially providing beneficial adaptation to physical stress. At the same time, the gut microbiota composition itself may influence the exercise performance.

#### *2.9. Nutrition and Stiffness*

The relationship between dietary components and arterial stiffness has been investigated in limited and heterogeneous studies that seem to indicate a beneficial effect of certain nutrients on vascular ageing.

Higher anthocyanin and flavone, cocoa intake, as well as phytoestrogens such as isoflavones and lignans, are associated with lower arterial stiffness [68].

Dietary polyphenols have been investigated in several small heterogeneous studies. Cocoa and chocolate, rich in flavonoids and proanthocyanidins, seem to reduce BP levels and cardiovascular risk, with an improvement in measures of vascular health (arterial stiffness and endothelial function), possibly due to the activation of nitric oxide (NO) synthase, and to other antioxidant/anti-inflammatory properties [69,70]. The European Food Safety Authority approved a health claim about the effectiveness of cocoa polyphenols on arterial elasticity, indicating an ideal assumption of 200 mg of cocoa flavanols daily, consumed as 2.5 g high-flavanol cocoa powder, or 10 gr high-flavanol dark chocolate [71]. Although anti-inflammatory and antioxidant effects have been associated with berry and grape juice consumption, there are no sufficient data to establish their relationship with arterial stiffness. On the other hand, the importance of isoflavone (a soy metabolite) in reducing arterial stiffness and BP levels has been highlighted [72].

Curcumin capsule supplementation has shown to reduce PWV in diabetic patients in a randomized trial [73].

#### **3. Systematic Review**

#### *3.1. Aim*

This systematic review aims to investigate (i) the interdependence between gut microbiota composition and central hemodynamics, and (ii) whether modifications to gut microbiota translate into different vascular aging profiles.

#### *3.2. Methods*

#### 3.2.1. Eligibility Criteria

This systematic review is based on population, intervention, comparator, outcome, and setting criteria. Participants: humans or animals included in both observational and interventional studies. Interventions: we considered every kind of intervention (dietary, antibiotics, fecal transplant, dietary supplements, etc.). Comparators: we included any kind of comparator. Outcomes: primary outcomes: (i) modification in PWV; (ii) modification in gut microbiota composition (alpha- and beta-diversity, genera abundances); secondary outcome: relationship between changes in PWV and gut microbiota composition. Study designs: observational, experimental, and interventional trials in humans and animals are included. No restrictions were imposed on language or date of publication. Exclusion criteria: studies without information about either microbiota or arterial stiffness were excluded; editorials, study protocols, reviews, commentary, and letters were also excluded.

#### 3.2.2. Information Sources and Search

The following databases from inception to February 2022 were searched: PubMed/ MEDLINE, Scopus, Web of Science. The main electronic search strategy was designed for PubMed/MEDLINE and was adapted as appropriate for each of the other databases.

D.A. and F.P. screened titles, abstracts, and full texts of articles identified in this search, and extracted the data for eligible studies; discrepancies were resolved by consensus.

#### *3.3. Results*

The systematic search led to the identification of 24 articles from three databases, of which 12 were based on animal studies and 12 on humans. A flowchart of the final selection of items is shown in Figure 2. Main characteristics of the selected articles are summarized in Table 1.

**Figure 2.** Flow diagram of the Systematic Review.

#### *3.4. Discussion*

#### 3.4.1. Animal Studies

Among twelve studies based on animal models, four studies tested diet supplementation with soy [74], dapaglifozin [75], indole-3-propionic acid [76], and hesperidin [77]. Three studies focused on fecal transplantation [78–80], and two on antibiotic treatment [81,82]. Other studies investigated the SCFA receptor [29] and germ-free mice [83]. Only three studies reported data on BP levels [29,49,77]. Nine studies analyzed arterial stiffness by recording PWV by aortic doppler (see Table 1).

**Table 1.** Characteristics of animal studies.



#### **Table 1.** *Cont.*


#### **Table 1.** *Cont.*

VA stands for Vascular ageing; SBP, systolic blood pressure; PWV, pulse wave velocity; LPS, lipopolysaccharides; Gpr, G-protein coupled receptor; SCFA, short-chain fatty acid; GF, germ-free; F:B, Firmicutes/Bacteroidetes ratio; TMAO, trimethylamine-N-oxide.

Supplementation studies. In all studies, independent of the type of supplement, the modifications of gut microbiota were associated with parallel modification to arterial stiffness. In particular, it seems that the changes of the gut microbiota linked to a better configuration are correlated with lower arterial stiffness. (i) In rats selectively bred for low running capacity, soy supplementation significantly improved their blood lipid profile, adipose tissue inflammation, and aortic stiffness; it shifted the cecal microbiota toward a lower F:B ratio. Soy-fed rats had lower mRNA expression of CD11c (inflammatory macrophage marker) and of the proinflammatory cytokine IL-6 [74]. (ii) Diabetic mice presented higher cfPWV than controls. Supplementation with dapaglifozin was associated with reduced microbiota diversity and richness, but failed to improve arterial stiffness in the study by Lee DM. Interestingly, arterial stiffness was negatively associated with *Akkermansia* abundance and positively with F:B ratio [75]. (iii) The supplementation with indole-3-propionic acid (a microbial metabolite of the essential aromatic amino acid, tryptophan) did not improve cfPWV in mice fed a western diet, and even worsened cfPWV in control mice, which also presented a reduced abundance of *Bifidobacterium* [76]. (iv) Supplementation with hesperidin resulted in higher urinary excretions of hippurate and other polyphenols metabolites. As most polyphenols are metabolized by gut microbiota before being absorbed, in this study, urinary metabolites of hesperidin were positively correlated with a microbial family, the Bacteroidaceae (phylum Bacteroidetes). From the vascular point of view, hesperidin supplementation was able to reduce both the circulating levels of neuraminidase (a biological marker of arterial stiffness [84]) and the systolic BP [77]. These results show a biological effect of polyphenols on vascular ageing, and indicate gut microbiota modification as the potential mechanism of the effect.

Fecal transplantation. Arterial stiffness was associated with gut microbiota modifications induced by fecal transplantation. In particular, PWV was positively associated with Clostridium genus, which contains most of the deleterious *Clostridium* species (e.g., *C. botulinum*, *C. perfringens*, *C. difficile*), and with gut permeability and obese microbiota, whereas it was negatively associated with Akkermansia abundance. (i) Local carotid stiffness was investigated in mice fed with a high-fat diet and gavaged with gut microbiota of either healthy donors ("Con" group) or patients with myocardial infarction ("CAD" group). The characteristics of the gut microbiota from CAD patients were transmissible and associated with low fermentation and high inflammation, and with increased abundance of *Clostridium symbiosum* (Clostridium genus) and Eggerthella genus. CAD mice also presented a higher carotid stiffness than the Con group of about 1 m/s [78]. (ii) An interesting study from Battson M et al. investigated fecal transplantation in four groups: control mice fed with either normal microbiota (Con + Con) or microbiota from obese mice (Con + Ob); and obese mice fed with either normal microbiota (Ob + Con) or obese (Ob + Ob) microbiota. Higher PWV was observed after obese microbiota gavage in control mice, together with altered gut permeability and SCFA content. Importantly, Akkermansia abundance was strongly inversely related to PWV with r = −0.8 (*p* < 0.0001) [79]. (iii) In another experimental study, the microbiota from either lean (LM) or obese (OBM) patients was used to feed two cohorts of mice: one from male and one from female patients. Aortic stiffness was higher in OBM than germ-free mice, and in cohort 2, it was also higher in OBM than LM mice. Mouse microbiota profiles clustered according to their transplant donor groups, possibly explaining the difference in arterial stiffness [80].

Antibiotic treatment. In both studies that we found, antibiotic treatment induced deep changes in gut microbiota composition, which were associated with parallel changes in arterial stiffness. (i) The role of antibiotic treatment was investigated in mice fed with a western diet for 5 months. During the study, mice presented a progressive increase in aortic PWV, which was completely reversed by 2-month antibiotic supplementation. The western diet increased the F:B ratio and Ruminococcus abundance; it reduced the abundance of Bifidobacterium, and increased inflammatory markers like LPS-binding protein, IL-6, plasminogen activator inhibitor-1, which were normalized after the antibiotic treatment. Four groups were analyzed: old control mice (OC), old mice fed with antibiotic supplementation (OA), young control mice (YC); young mice with antibiotics (YA). Old mice presented several markers of dysbiosis and inflammation, with higher levels of TMAO and PWV. During the intervention, PWV increased in young mice without antibiotic supplementation only. In old mice, antibiotic treatment was associated with: (i) partial PWV improvement (*p* = 0.047 vs. OC); (ii) increased aortic elastin expression; (iii) suppressed TMAO levels [81]. (ii) In one study from Brunt V et al., 35 young and 38 old mice were treated with antibiotic supplement for 3 to 4 weeks. After the supplementation, most of major phyla were suppressed. The intervention restored arterial stiffness in old mice to normal levels, and normalized oxidative stress and inflammation [82].

The role of SCFA receptor Gpr41 in vascular function was investigated in Gpr41-KO mice. Gpr41 was found in the vascular endothelium, which was necessary for SCFAmediated vasodilation. At baseline, Gpr41-/- mice presented isolated systolic hypertension, but with no differences in plasma renin concentration between WT and KO mice. At 6mo, Gpr41-/- mice showed accelerated vascular ageing with higher PWV than wild mice. Of note, ex vivo analysis at 6mo showed no difference in tensile vessel properties. The disparity between higher PWV and unchanged structural vessel properties indicates that functional alterations may occur before structural modifications exist [29].

The gut microbiota influences the resistance properties of arteries, with increased stiffness in male germ-free mice with respect to either wild or female germ-free mice [83].

All these experimental animal studies consistently show a strict correlation between modification of gut microbiota and arterial stiffness. Moreover, some of them highlight the role of inflammation as a mediator between the gut microbiota and arterial stiffness.

#### 3.4.2. Human Studies

Among studies on humans (see Table 2), we found five cross-sectional and seven intervention studies. Ten studies present information on BP, and six employed cfPWV as a measure of arterial stiffness. Three studies did not directly evaluate modification in microbiota.

Cross-sectional studies. In this study setting, a correlation between gut microbiota modifications and arterial stiffness parameters was consistently found across the included studies. In particular, the abundance of beneficial bacteria (mainly butyrate producers) is constantly associated with lower arterial stiffness. (i) Menni et al. investigated more than 600 women from the TwinsUK registry, and measured tonometric carotid-femoral PWV and microbiota composition. They found that gut microbiome diversity was significantly inversely associated with arterial stiffness. PWV was also negatively associated with the abundance of Ruminococcaceae family bacteria, which are beneficial butyrate-producing bacteria linked to lower endotoxemia [85]. (ii) In 10 hemodialysis patients, Firmicutes and Bacteroidetes phyla were the most abundant. Faecalibacterium spp. (butyrate producer from the Oscillospiraceae family, class Clostridia, phylum Firmicutes) were positively associated with total carbohydrate intake (ρ = 0.636; *p* = 0.048) and negatively associated with cfPWV (ρ = −0.867, *p* = 0.001). Lipopolysaccharide-Binding Protein, a marker of bacterial translocation through the intestinal barrier and endotoxemia, was negatively associated with butyrate-producing bacteria [86]. This result supports the association between the favorable microbiota composition and vascular ageing, through a reduction of the systemic inflammation derived from leaky gut syndrome. (iii) In sixty-nine subjects not treated for hypertension who underwent an ambulatory BP monitoring, an ambulatory arterial stiffness index (AASI) was obtained. AASI is calculated from the regression line between 24 h systolic and diastolic BP values, and is believed to be a marker of arterial stiffness. Although no definite evidence exists on its accuracy, at least it seems able to predict cardiovascular events. No association was found between microbiota diversity indexes and AASI. AASI was associated with lower abundance of Lactobacillus spp. and higher abundance of several deleterious species from the genus Clostridium [87]. (iv) In children with chronic kidney disease with different categories of estimated glomerular filtration rate (G1: eGFR ≥ 90 mL/min/1.73 m2, 9.5 years-old; G2-G3: eGFR 30–89, 13.7 ys), carotid-PWV was correlated to the severity of the disease. Although various beneficial bacteria (Lactobacillus, Bifidobacterium, Akkermansia) were not influenced by the severity of the disease, genus Lactobacillus abundance was negatively correlated with PWV [88].

Supplement intervention. Three studies investigated the effect of supplementation on gut microbiota and arterial stiffness, without any significant result. Only one study found a significant effect on arterial stiffness reduction associated with a specific microbiota pattern, but the effect was only marginal. (i) A total of sixty-six healthy men were enrolled in a randomized, double-blind placebo-controlled trial where two forms of aronia supplementation were compared to a placebo after 12-week treatment: a (poly)phenol-rich aronia extract, and an aronia fruit powder. No effect on aortic stiffness was registered. Gut microbial diversity was very high and did not show significant variation among the treatment groups after aronia intake; the aronia extract group presented an increased abundance of genus Anaerostipes (butyrate producer, family Lachnospiraceae, class Clostridia) [89]. (ii) A relatively young uncontrolled hypertensive sample was randomized to receive garlic (*n* 23) supplementation or a placebo (*n* 26) for 12 weeks. Tonometric cfPWV presented a trend for reduction in the garlic group (from 12.8 to 12.1 m/s), but with no statistical difference versus the placebo. Of note, SBP was reduced in garlic versus the placebo arm, with a mean difference of 10 mmHg. The garlic group presented an increase in Lactobacillus and Clostridium spp, without deeper characterization of the bacterial species [90]. (iii) A

total of twelve young men were studied after consumption of 2 eggs/day for 2 weeks in a non-randomized trial. Brachial-ankle PWV and endothelial function improved, with no effect on BP, inflammation, oxidative stress, or TMAO. Microbiota was not modified, but a reduction of tryptophan degradation was observed [91].

Exercise intervention. Both the following studies found a significant relationship between the gut microbiota and arterial stiffness surrogates. In particular, the second study, focused on a very intense exercise training program, found a significant microbiota modification after the training, associated with improvement of the augmentation index (an indirect index of arterial stiffness, which is associated both with vascular peripheral resistances and with the phenomenon of the pulse wave reflection). Unfortunately, the trial was not randomized and the augmentation index is not only related with arterial stiffness, but also to the peripheral resistances and heart rate. This mines the interpretation of the results. (i) A crossover trial with 5 wks of exercise training and 5 wks of washout period found a significant positive correlation between Clostridium difficile and arterial stiffness in 33 men, measured by the cardio-ankle vascular index (CAVI), (r = 0.306, *p* = 0.016) and SBP, but with no exercise effect. Microbiota diversity was not affected by exercise, but the relative abundances of Oscillospira and Clostridium difficile were increased and decreased by exercise, respectively [92]. Considering that Oscillospira species (family Oscillospiraceae, class Clostridia) are associated with leanness and may have anti-inflammatory properties [93], the results of this study support the beneficial role of exercise training in microbiota composition. (ii) In a non-randomized interventional trial, 24 obese adolescents underwent a 6-wk program of endurance/strength training for 5 h/day and 6 d/wk, together with caloric restriction. The subendocardial viability ratio, an index of the workload of the left ventricle depending on ventricle-vascular coupling, and the augmentation index were improved by the program, with no significant change in BP. Microbiota diversity increased, together with the abundance of the Christensenellaceae family, which is inversely related to the host body mass index [94].

Studies without direct microbiota measures. Three studies present results on arterial stiffness modification in relation to either comorbidities or supplementation design, assuming an indirect role of gut microbiota. (i) In the paper by Ponziani et al., 39 patients with suspected small intestinal bacterial overgrowth (SIBO) were included. Vitamin-K2 status was measured, and Vitamin-K2 intake and carotid PWV were obtained. In patients with confirmed SIBO (*n* = 12), despite similar dietary vitamin-K2 intake, measured vitamin-K2 status was markedly reduced versus the no-SIBO group (*p* = 0.02), suggesting an altered vitamin-K2 production by intestinal bacteria. Median PWV was significantly higher in the SIBO group than the no-SIBO group (10.25 m/s vs. 7.68 m/s; *p* = 0.002). Furthermore, in both groups, vitamin-K2 status was significantly correlated with carotid PWV (R2 = 0.29, *p* < 0.001) [95]. This study supports the role of gut dysbiosis in vascular ageing. (ii) In a randomized trial, 15 patients received 1-month cocoa extract with 130 mg epicatechin and 560 mg procyanidins (group D1-10), 15 patients received 20 mg epicatechin and 540 procyanidins (group D2-10), and were compared with the placebo (*n* = 15). Of note, both epicatechin and procyanidins are catabolized by gut microbes in the colon. CfPWV was reduced in the D1-10 group by about 1 m/s versus the placebo and by 0.8 m/s versus D2-10 group. SBP was also reduced in D1-10 versus the control and D2-10 groups. In this study, the reduction of SBP could explain the observed variation in PWV. The impact of cocoa flavonols on vascular health was mainly linked to epicatechin, and is mediated by epicatechin metabolites, which in turn depends on gut microbiota metabolism [96]. (iii) One study focused on the effect of equol, a microbial-derived metabolite of the isoflavone daidzein. Equol is produced by gut microbiota after soy intake in almost one-third of the Western population. Acute soy supplementation slightly reduced cfPWV only in equol producers (−0.2 m/s) at 24 h, with no change in BP [97].

Even if a direct association with microbiota composition and function has not been addressed, the three studies presented here show a significant correlation between vascular ageing and the metabolites of gut microbiota. This confirms the role of gut microbiota in modulating human systemic biological pathways.

Antibiotics in humans. In contrast to what has been shown in animal studies, this paragraph aims to warn the reader not to consider antibiotic treatment beneficial for cardiovascular health in humans. Indeed, the role of antibiotics in humans is not fully established. Studies investigating the role of antibiotics in microbiota and the effect on cardiovascular risk are lacking. According to the literature, it appears that antibiotics impact the gut microbiota, reducing bacterial diversity and changing relative abundances [98]. They were also found to enhance pathways linked to increased atherosclerosis [99]. Long-term use of antibiotics in late adulthood has been associated with all-cause and cardiovascular mortality [100]. Macrolide antibiotic consumption is associated with increased risk for sudden cardiac death or ventricular tachyarrhythmias and cardiovascular death, but not increased all-cause mortality [101]. Furthermore, no association with long-term cardiovascular risk (ranging from >30 days to >3 years) was noted in observational studies or randomized controlled trials on treatment with macrolides [102]. A significant association was found between fluoroquinolone use and an increased risk for arrhythmia and cardiovascular mortality [103]. Antibiotic exposure in infancy was associated with a slightly increased risk of childhood overweight and obesity [104]. The pooled colorectal cancer risk was increased among individuals who ever used antibiotics, particularly for broad-spectrum antibiotics [105]. The pooled breast cancer risk was modestly increased among individuals who ever used antibiotics [106].

Limitations. This study presents several limitations. Firstly, due to the small number of studies and great heterogenicity among them, it was not possible to make a meta-analysis of the results. Second, the quality of most of the studies was questionable, due either to sample size, or to the assessment of microbiota or arterial stiffness. Third, for the reason just mentioned, it was not possible to entirely follow the PRISMA statement for systematic review.

#### *3.5. Conclusions*

From the available literature, we found an association between gut microbiota composition and arterial stiffness. We identified two association patterns, consistently present in most animal and human studies: (i) a direct correlation between arterial stiffness and abundances of bacteria associated with altered gut permeability and inflammation (mainly from the *Clostridium* genus), as well as with biological markers of inflammation; (ii) an inverse relationship between arterial stiffness, microbiota diversity, and abundances of bacteria associated with most fit microbiota composition (butyrate producers, *Akkermansia*, *Bifidobacterium*, *Ruminococcaceae*, *Faecalibacterium*, *Lactobacillus*).

While in animal studies most of the interventions were able to show a stable link between microbiota modification and arterial stiffness, in humans that was not the case. In particular, none of the identified interventional trials was able to demonstrate this relationship. However, most strikingly, nearly half of human studies measured BP, and very few adjusted the vascular analyses for BP variation, which is a major determinant of arterial stiffness.

The main finding of this review is the lack of large randomized interventional trials in humans that test the role of gut microbiota modifications on arterial stiffness, and take into account BP and hemodynamic alterations.


#### **Table 2.** Characteristics of human studies.


#### **Table 2.** *Cont.*

VA stands for Vascular ageing; SBP, systolic blood pressure; cfPWV, carotid-femoral pulse wave velocity; CAVI, cardio-ankle vascular index; MAP, mean arterial pressure; AIx75, augmentation index corrected for heart rate at 75 bpm; SEVR, sub-endocardial viability ratio; AASI, ambulatory arterial stiffness index; baPWV, brachial-ankle pulse wave velocity; FMD, flow-mediated dilation; eGFR, estimated glomerular filtration rate; eGFR categories: G1 ≥ 90 mL/min/1.73 m2, G2 50–89, G3 30–59; ABPM, ambulatory blood pressure measurement; TMAO, trimethylamine-N-oxide.

**Author Contributions:** Conceptualization, D.A., F.P., A.F.G.C., C.B.; methodology, D.A., F.P., A.F.G.C.; validation, D.A., F.P., A.F.G.C., C.B.; writing—original draft preparation, D.A.; writing—review and editing, F.P., A.F.G.C., C.B.; supervision, C.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** No new data were created or analyzed in this study. Data sharing is not applicable to this article.

**Acknowledgments:** We thank Professor Patrizia Brigidi, from the Department of Clinical and Surgical Sciences, University of Bologna, for kind support in manuscript conceptualization and revision.

**Conflicts of Interest:** A.F.G.C. is consultant to Roelmi and Viatris. C.B. is on the scientific board of Servier International and Menarini International. D.A. and F.P. declare they have no financial interest.

#### **References**

