Introduction

Cardiovascular disease (CVD) is the number one cause of death and accounts for 32.3 % of (or one of every three) deaths in the USA [1]. Large artery stiffness called arteriosclerosis has been identified as an independent predictor of CVD mortality and morbidity [24]. Arterial stiffness is characterized by structural remodeling and functional changes in the arterial wall. Major changes, typically seen with “physiologic” arterial aging, include fragmentation and calcification of elastic fibers, and increased collagen deposition and collagen cross-linking in the media [5, 6]. Oxidative stress, decreased bioavailability of endothelial-derived vasodilators (e.g., nitric oxide), and low-grade inflammation process have also been recognized to promote arterial stiffness [7]. Diminished elasticity of arterial walls results in earlier arrival of wave reflection, and reflected waves that arrive during left ventricular ejection increase left ventricle afterload, the amplification of systolic blood pressure, and pulse pressure [8, 9]. The exposure of small arterioles to high pulsatile pressure caused by arterial stiffness also explains the microvascular damage in the brain and the kidneys [10, 11]. Arterial stiffness has been found to be an independent and consistent predictor of left ventricular hypertrophy [12], hypertension [13, 14], myocardial infarction [15] and stroke [16], and cardiovascular mortality in patients with end-stage renal disease [17, 18].

Despite the tremendous emphasis on the role of arterial stiffness in the development of CVD, relatively little is known about the biological mechanisms and pathways underlying the development of arterial stiffness. Studies on heritability of arterial stiffness suggest that genes explain a moderate (range from 23 to 50 %) proportion of the variability in arterial stiffness. This has been reported from populations in a number of studies including: the Framingham Heart Study [19], the Strong Heart Family Study [20], twin studies [21, 22], and the Erasmus Rucphen Family study [23]. The influence of a genetic component in arterial stiffness has been shown to be independent of well-known cardiovascular risk factors such as age and blood pressure [20, 23]. Identification of genetic variations related to arterial stiffness may provide novel insights into its pathophysiology. Moreover, potential preventive measures may be developed to benefit those with a high genetic risk of arterial stiffness.

Arterial stiffness can be noninvasively assessed by determining the speed of energy conduction through the arterial wall, analyzing pulse waveforms, or relating diameter change to pressure change [24]. These methods produce parameters including pulse wave velocity (PWV), augmentation index (AIx), arterial compliance, arterial distensibility, elastic modulus, Young’s elastic modulus, and stiffness index [24]. Pulse pressure measured using a standard sphygmomanometer is also recognized as a surrogate marker of arterial stiffness in persons aged more than 50 years [25].

In this paper, we reviewed genetic variants that have been reported for their association with arterial stiffness (Table 1). We next discussed several potential factors that contribute to the inconsistent results of the reviewed studies. Lastly, we proposed the model (Fig. 1) explaining arterial stiffness pathophysiology with the genes reviewed in this study.

Table 1 Single nucleotide polymorphisms (SNP) related to arterial stiffness
Fig. 1
figure 1

AGT angiotensinogen, ACE angiotensin-converting enzyme, AGTR1 angiotensin II type 1 receptor, CYP11B2 aldosterone synthase gene, ADRB1, B2, B3 adrenergic receptors B1, B2, B3, ECE endothelin-converting enzyme, ETAR endothelin A receptor, ETBR endothelin B receptor, VSMC vascular smooth muscle cell, eNOS endothelial nitric oxide synthase, TXNIP thioredoxin interaction protein, IL-6 interleukin-6, SELP P-selectin, ICAM1 intercellular adhesion molecules-1, VCAM1 vascular cell adhesion molecules-1, FAS tumor necrosis factor (TNF) receptor superfamily member 6, CRP C-reactive protein, TNF-α tumor necrosis factor alpha, RAGE receptor for advanced glycation endproducts, GLO1 glyoxalase1, ER α and β estrogen receptor alpha and beta, CYP19A1 cytochrome P450 family 19 subfamily a polypeptide 1, ATP2B1 calcium transporting ATPase 1, ENPP1 ectonucleotide pyrophosphatase/phosphodiesterase-1, ADIPOQ adiponectin, ELN elastin, COL4A1 collagen type 4 alpha 1, FBN1 fibrillin-1 glycoprotein, LMNA lamin A/C, MMP matrix metalloproteinase, GNB3 G protein B3 subunit

Method

We performed a PubMed literature search for relevant articles with no date restriction. The keywords used included single nucleotide polymorphism (SNP), gene, genetics, association, arterial stiffness, pulse wave velocity, augmentation index, distensibility, elastic modulus, and arterial stiffness β. In order to make the findings comparable, studies that examined the associations between combined variables and arterial stiffness (e.g., haploblock) were excluded. We focused on cross-sectional association between SNP and commonly used measures of arterial stiffness. A total of fifty articles that met our criteria were selected for the review.

Genetic Variants Related to Arterial Stiffness

Renin-Angiotensin-Aldosterone System-Related Genes

Renin-angiotensin-aldosterone system (RAAS) is involved in blood pressure regulation by sodium-fluid balance and is known to play an important role in arterial tone [81]. Studies that have investigated the angiotensinogen gene (AGT) have frequently examined the Met235Thr (T704C) polymorphism in exon 2. Although one study with hypertensive subjects reported that those homozygous for the threonine allele in the Met235Thr polymorphism had significantly higher arterial stiffness [29], this relationship was not found in the Rotterdam Study [28] or in two other studies with hypertensive subjects [26, 27].

For the angiotensin-converting enzyme gene (ACE) which converts angiotensin I to vasoactive angiotensin II, the insertion/deletion polymorphism of intron 16 has been extensively investigated. Previous studies have shown that the number of D allele was positively related to the level of ACE expression in T-lymphocytes [82] and serum levels of ACE [83]. The 287-bp deletion in intron 16 has been shown to influence arterial stiffness in both the general population and diabetic and hypertensive patients [28, 3033, 35]. However, the allele related to increased arterial stiffness varied between general population and patient groups. While those with the D allele had significantly greater arterial stiffness in general populations [28, 30, 31], type 2 diabetes [33] and untreated hypertensive [36] patients with the D allele had less arterial stiffness. Possibly, the activity of the D allele on arterial stiffness may be modified in hypertension or diabetes through its interaction with mechanisms involved in their pathophysiology.

As for the A1166C polymorphism of angiotensin II type 1 receptor gene (AGTR1) in the 3′ untranslated region, the C allele is related to increased arterial stiffness in hypertensive patients [27, 35, 37], in contrast to the A allele which is related to higher arterial stiffness in more severe hypertensive patients (50 % in grade 3 hypertension) [38]. In two studies with normotensive White subjects [35] and the random general population [40], arterial stiffness measured by carotid-femoral PWV was not different across the genotypes. However, arterial stiffness measured by femoral–dorsalis pedis or tibialis posterior PWV was significantly associated with the C allele of the A1166C polymorphism [40], suggesting that the A1166C polymorphism may have different effects on central elastic arteries and peripheral muscular arteries. There is also a potential combined effect of this genotype with age. For example, a recent longitudinal study showed that the C allele carrier in the A1166C polymorphism had a 35 % more pronounced increase in carotid-femoral PWV over a 16-year period than the AA alleles subjects, and this difference in PWV was only observed after the age of 55 years [84].

In another study examining polymorphisms in the aldosterone synthase gene (CYP11B2), the C allele of the T-344C polymorphism in the promoter region was associated with higher peripheral and central augmentation index than T allele homozygotes in a healthy random population [34]. Inconsistent results in both arterial stiffness and age-related changes of arterial stiffness were found with this T-344C polymorphism in studies with hypertensive patients [27, 41].

β-Adrenergic Receptor Genes

β-adrenergic receptors, which are a class of G-protein-coupled receptors, play an important role in cardiovascular function. Chronic stimulation of β-adrenergic receptors by the sympathetic nervous system may contribute to physiological and pathological remodeling of vascular function and structure. There are three β-adrenergic receptors: β1 (ADRB1), β2(ADRB2), and β3(ADRB3) [85]. Compared with the studies on RAAS-related genes, studies that investigated the effect of genetic variants of β-adrenergic receptor genes on arterial stiffness have been conducted in racially more diverse groups, including Asians, White, and Black adults. In one study with White and Black young adults aged 19–44 years, the glycine allele of the Arg389Gly polymorphisms in ADRB1 was associated with arterial stiffness in White young adults, but not in Blacks. In contrast, the arginine allele of the Arg16Gly polymorphism in ADRB2 and the arginine allele of the Trp64Arg polymorphism in ADRB3 were associated with arterial stiffness in Black young adults, but not in Whites. These three polymorphisms were not associated with blood pressure and heart rate [42], suggesting the independent role of β-adrenergic receptor genes on arterial elasticity in young adults. Interestingly, the alleles of the Arg16Gly polymorphism in ADRB2 may have an opposite influence on arterial stiffness in different racial groups. The glycine allele was associated with a higher pulse pressure in European American youth [44]. In Japanese hypertensive patients, arterial stiffness and these three polymorphisms (Arg389Gly, Arg16Gly, and Trp64Arg) were not associated; instead, homozygotes of the glycine allele of the Ser49Gly polymorphism in ADRB1 had higher arterial stiffness [43].

Endothelin-Related Genes

Endothelin-1 (ET-1) is a potent vasoconstrictor produced in endothelial cells. It binds to two G-protein-coupled receptors, endothelin A receptor (ETAR) and endothelin B receptor (ETBR), which are expressed mainly on vascular smooth-muscle cells and endothelial cells, respectively [86]. Whereas ETAR plays a role in vasoconstriction of smooth muscle cells, ETBR facilitates endothelial synthesis of nitric oxide (NO) and prostacyclin [87]. In a study with untreated hypertensive subjects, women with the G allele of the A-231G polymorphism in ETAR and with the GG alleles of the Leu277Leu polymorphism (without amino acid change) in ETBR had significantly higher carotid-femoral PWV than other groups, but these results were not found in men [39]. In another study, the A958G polymorphism in ETAR, and the Leu277Leu polymorphism in ETBR did not contribute to the levels of arterial stiffness measured by brachial-ankle PWV in a general Japanese population [45]. Endothelin-converting enzyme (ECE) converts a 39-amino-acid precursor to a 21-amino-acid endothelin-1 protein, thus it may be important in regulating vascular tone. However, the polymorphisms, A2013(+289)G in ECE-1 and T669(+17)C in ECE-2, were not significantly related to arterial stiffness in Japanese subjects [45].

Endothelial Nitric Oxide Synthase Gene

Endothelial NOS (eNOS), also known as nitric oxide synthase 3 (NOS3), plays an important role in modulating vascular smooth muscle tone. In the presence of oxygen, reduced nicotinamide adenine dinucleotide phosphate (NADPH), flavin adenine dinucleotide (FAD), flavin mononucleotide (FMN), and calmodulin, eNOS released from endothelium catalyzes l-arginine to nitric oxide (NO) which is a short-lived vasodilatory gas [88]. The results on the relationship between eNOS polymorphisms and arterial stiffness have shown inconsistencies in different races, sex, and various arterial sites for measuring arterial stiffness. The T allele of G894T (Asp298Glu) polymorphism was related to more arterial elasticity in African Americans, but not in White Americans [46]. This common missense mutation of G298T was associated with changes in central pulse pressure in females but not in males [47]. Furthermore, the C allele of the T-786C polymorphism in eNOS was significantly associated with increased PWV measured in peripheral arteries, but not in central elastic arteries [40]. In these studies, none of polymorphisms in eNOS were significantly associated with arterial stiffness measured by carotid-femoral PWV which is a direct measure of central elastic arteries. It was also reported that both eNOS polymorphisms, G10T and G298T (Glu298Asp), were not associated with aortic PWV in hypertensive or non-hypertensive Europeans [49]. Despite promising evidence of eNOS in endothelial function and vascular function, the effect of eNOS polymorphisms on arterial stiffness remains unclear.

Oxidative Stress-Related Genes

The crucial role of NADPH oxidase system in vascular oxidative stress and atherosclerosis has been well described in previous studies. NADPH oxidase has been identified to be a major source of reactive oxygen species (ROS), such as superoxide anion, which causes oxidation of low density lipoprotein (LDL) [89]. Studies have shown that the C242T (Tyr72His) polymorphism of gene encoding NADPH oxidase p22phox subunit is significantly related to arterial stiffness in Caucasian male runners [51] and in the general population of Brazil [90]. However, the allele related to arterial stiffness was different in each study. While runners with the CC alleles had higher carotid-femoral PWV (measured with the PulsePen device) than runners with the CT and TT alleles, individuals with the TT alleles had higher carotid-femoral PWV (measured by the Complior device) than those with the CC or CT alleles in the Brazilian population.

Thioredoxin-interacting protein (TXNIP) is a protein known to increase the production of ROS and oxidative stress [91]. The G allele carriers of rs7212 in TXNIP showed significantly higher carotid-femoral PWV than CC homozygotes in diabetic patients in Brazil, but not in the non-diabetic group, suggesting a possible interaction of TXNIP and pathophysiology in diabetes [50].

Inflammation-Related Genes

Although inflammation may contribute to arterial stiffness, biomarkers of inflammation are not always related to arterial stiffness, even in patients with coronary artery disease [92]. One population-based study showed that the C allele of the G-174C polymorphism in the interleukin-6 gene (IL-6) promoter was significantly associated with increased carotid-femoral PWV and pulse pressure [52]. A study in young Americans also demonstrated that polymorphisms, Ser29Asn and rs2244529 in the P-selectin gene (SELP), Gly241Arg in the intercellular adhesion molecules-1 gene (ICAM1), and Asp693Asp (C>T) in the vascular cell adhesion molecules-1 gene (VCAM1) were significantly related to aorto-foot PWV, suggesting that genetic variations of adhesion molecules may be involved in the process of arterial stiffness [56].

However, another population-based study showed that among the polymorphisms in 12 inflammatory biomarkers genes, only the polymorphism rs10509561 of the TNF receptor superfamily member 6 gene (FAS) was significantly related to central pulse pressure. This polymorphism was not associated with either carotid-femoral PWV or augmentation index [55]. The C1444T polymorphism of C-reactive protein gene (CRP) and G-308A of the tumor necrosis factor alpha gene (TNF-α) were found to be determinants of carotid stiffness in patients with a history of Kawasaki disease, but not in control subjects [54]. Two studies reported that circulating levels of inteleukin-6 were significantly associated with carotid-femoral PWV and with its genetic variants; nevertheless, the genetic variants were not associated with carotid-femoral PWV [53, 55]. Inflammation does not seem to be a direct causal factor of arterial stiffness, but it may influence arterial stiffness indirectly through different mechanisms such as endothelial dysfunction enhanced by inflammation [93].

A recent meta-analysis of genome-wide association data has reported that the SNP, rs7152623, located in the B-cell lymphoma/leukemia 11B gene (BCL11B) desert was significantly associated with carotid-femoral PWV in nine community-based European ancestry cohorts and two other European cohorts [80]. Since BCL11B is known to participate in interleukin-2 gene (IL2) expression [94] and maybe important in the inflammatory process, further studies are needed to define the association of this genetic variation with inflammation.

Advanced Glycation Endproducts-Related Genes

Advanced glycation endproducts (AGE) is known to contribute to the development of vascular complications in diabetes. Binding AGE to the receptor for AGE (RAGE) may initiate the production of reactive oxygen species (ROS), resulting in pro-inflammatory cellular responses [95]. Data from two Dutch population-based cohort studies demonstrated that the AA genotype of the T-374A and the TT genotype of the C5878T in RAGE were associated with less arterial stiffness in individuals with normal glucose metabolism. However, in individuals with impaired glucose metabolism or type 2 diabetes, these genotypes were associated with higher pulse pressure and higher arterial stiffness, respectively [58]. This result suggests that the effect of the polymorphism under the condition of normal glucose metabolism may be different in altered glucose metabolism such as diabetes. An analysis on the part of the data from a Dutch population explored the glyoxalase1 (GLO1) gene which is used to detoxify methylglyoxal, a main precursor of AGEs. In 1289 participants with 33.3 % of DM2, the Ala111Glu of GLO1 was only weakly associated with pulse pressure, but not with other arterial stiffness measures [57].

Estrogen-Related Genes

Estrogen is considered a cardioprotective hormone which enhances NO production and inhibits inflammation and vascular smooth muscle cell (VSMC) proliferation [96]. Estrogen receptors (ER) are expressed in vascular smooth muscle cells [97, 98]. The associations between polymorphisms and arterial stiffness have been examined in ER-α, ER-β, and aromatase (CYP19A1) genes. One study conducted in healthy older adults in Japan reported that in the T-401C and the T30C (Ser10Ser) polymorphisms of ER-α, women with the T allele had higher brachial-ankle PWV as compared with women with the C allele. This relationship was not found in men [59], showing a sex-specific effect of the ER-α polymorphism on arterial stiffness. Another population-based study demonstrated that these polymorphisms, T-401C and Ser10Ser, were related to augmentation index (AI), but with the CC alleles relating to higher AI. In this study, the polymorphisms, T-401, Ser10Ser, (TA)n, and rs9340799 in ER-α, and rs944460 and rs1256034 in ER-β, were significantly related to AI but not to carotid-femoral PWV in the sex-pooled analysis [60]. Further, the polymorphisms, rs4646, Var80Val, (TTTA)n, and rs726547 in CYP19A1 were neither associated with augmentation index nor carotid-femoral PWV.

Adiponectin Gene

Adiponectin (ADIPOQ), produced by adipose tissue, is identified as an important adipocytokine that improves insulin resistance through its anti-inflammatory and anti-atherogenic properties [99, 100]. Adiponectin is implicated in the metabolic syndrome by regulating glucose levels and reducing fatty acids [101]. In a Japanese cohort study, hypertensive patients with the T allele of the G276T had significantly lower HDL cholesterol, higher HbA1c, and higher brachial-ankle PWV [61]. Further studies are required to replicate these results and to elucidate the direct relationship between the G276T of ADIPOQ and arterial stiffness in healthy individuals.

Apolipoprotein E Gene

Apolipoprotein E (APOE) is known to facilitate plasma cholesterol homeostasis by the hepatic uptake of lipoproteins and stimulation of cholesterol effluxes from macrophages [102]. APOE has four alleles (ε1, ε2, ε3, and ε4) according to the combination of C in rs429358 and T in rs7412 (ε1 is very rare) [102]. Many studies have shown the association between the ε4 allele (CC) and coronary disease [103]. However, in a study in the general population in urban Brazil, carotid-femoral PWV was not different among subjects with the ε2, ε3, and ε4 alleles. These APOE polymorphisms did not seem to have a direct effect on arterial stiffness, although, the ε4 allele was significantly associated with an unfavorable lipid profile [62].

Elastin Gene

Medial degeneration is a key feature of arterial stiffness and is characterized by a decrease in elastin fibers and an increase of collagen [5]. Elastin is one of the major components of the extracellular matrix. In a study on the A/G polymorphism of exon 16 (Ser422Gly) in the elastin gene (ELN), the A allele carriers had significantly lower distensibility of carotid arteries than the GG homozygous subjects. But this association was not found in radial arteries from the same cohort, which suggests that this polymorphism is associated with distensibility of elastic arteries only, but not with muscular arteries [63]. Another polymorphism, deletion/A in the ELN 3′ untranslated region, was associated with higher brachial-ankle PWV in A allele carriers from a Japanese population [64], showing the variant’s effect on muscular arteries. Given the important role of central arteries in the development of CVD, further studies are recommended to examine the influence of these ELN polymorphisms on carotid-femoral PWV.

Fibrillin-1 Glycoprotein Gene

Fibrillin-1 glycoprotein (FBN1) is an essential connective protein for the formation of microfibrils which constitute arterial elastic fibers [104]. Previous studies have examined the relationship between arterial stiffness and the variable nucleotide tandem repeat (VNTR) polymorphism in intron 28 of FBN1, but their results have not been consistent. While one study reported that the FBN-1 VNTR polymorphism was not associated with either carotid-femoral PWV nor other measures of stiffness in a healthy general population [70], other studies have shown that participants with the 2-3 genotype of the VNTR polymorphism had greater pulse pressure [74] and arterial stiffness (measured by aortic stiffness β and elastic modulus) [67] in healthy men. Another study in participants with coronary artery disease also reported that those with the 2-3 genotype of the VNTR polymorphism had greater arterial stiffness (measured by aortic impedance) and higher central pulse pressure than those with the 2-2 and 2-4 genotypes [73].

Collagen Genes

Collagen is another important determinant of arterial wall mechanics along with elastin [5]. Collagen type 4 is the major structural component of basal lamina. A genome-wide association from the SardiNIA study showed that the minor C allele of A4002C (Gln1334His) in the collagen type 4 alpha 1 gene (COL4A1) is associated with higher carotid-femoral PWV. This association was replicated in 1828 Sardinians and 813 Amish individuals in the oldest category [65]. Another study in the UK reported that the G+2046T polymorphism in COL1A1, was associated with a higher PWV (measured in three segments; aortoiliac, aortoradial, and aorto-dorsalis-pedis) in young healthy individuals with the T allele [66]. However, this polymorphism was not associated with aortic distensibility in men from Sweden (mean age 55.3 years) [67]. The role of COL1A1 polymorphisms in arterial stiffness should be reexamined in further studies due to these inconsistent results.

Matrix Metalloproteinase Genes

Matrix metalloproteinases (MMPs) are a family of proteases that degrade extracellular matrix molecules such as elastin and collagen. It was reported that healthy subjects with the T allele of the C-1562T and the glutamine allele of the Arg279Gln in MMP-9 had higher serum MMP-9 and aortic PWV [105]. The association with the C-1562T polymorphism was also reported in untreated hypertensive patients [71] and patients with coronary artery disease [72]. The C-1562T polymorphism may have a sex-specific effect on arterial stiffness. In a population-based study conducted in Taiwan, this polymorphism was associated with arterial stiffness (measured by arterial stiffness β, elasticity modulus, and one-point PWV) in menopausal women, but not in men [69].

As for the human stromelysin-1 (MMP-3) gene, a common promoter polymorphism with one allele containing 5 adenosines (5A) and the other allele containing 6 adenosines (6A) was associated with stomelysin-1 gene expression [106]. A study showed that 5A homozygotes and 6A homozygotes have higher ascending aortic input impedance as compared with the heterozygotes in the older group (>60 years of age). However, this relationship was not found in the younger group (30–60 years of age), suggesting an age-related effect of this polymorphism. The gene and protein expression data from this study showed that those with the 5A/6A alleles had intermediate MMP-3 levels which are lower than those with the 5A/5A alleles and higher than those with 6A/6A. These results show that the optimal level of MMP-3 may be an important factor for the homeostasis between degradation and deposition of matrix proteins [68].

Lamins Gene

Lamins are intermediate filament-type proteins which are major components of the nuclear lamina [107]. Studies showed that lamins are involved in cardiac and muscular muscle dystrophy [108, 109]. In the lamin A/C gene encoding for lamin A and C, the C-1030T polymorphism in the promoter region may influence arterial stiffness. In a study of healthy Japanese men, the CC homozygotes of the C-1030T polymorphism had greater brachial-ankle PWV than other genotypes, and the CC alleles were reported to be a significant and independent predictor of brachial-ankle PWV after controlling for other CVD risk factors including age and blood pressure (BP) [75].

Calcium-Related Genes

Calcium phosphate metabolism may influence arterial stiffness both in structure and function. For example, vascular calcification can be accelerated by calcium accumulation in the extracellular matrix, and calcium is also involved in vascular cell function and smooth muscle contraction [110]. Calcium transporting ATPase 1, which is encoded by the ATP2B1 gene, moves calcium ions from the cells in effort to promote intracellular calcium homeostasis; thus, it may be involved in regulating vascular smooth muscle tone [111]. In untreated hypertensive patients, the A allele homozygotes of rs17249754 and the G allele homozygotes of rs1401982 had lower carotid-femoral PWV than heterozygotes or other homozygotes [76].

Fetuin-A is a systemic inhibitor of calcium-phosphate precipitation [112]. A study conducted in subjects with normal kidney function showed that the plasma fetuin-A level was independently and negatively correlated with aortic PWV. Subjects with the Ser/Ser alleles in the Thur256Ser polymorphism of the fetuin-A gene had significantly lower fetuin-A levels and higher aortic PWV. This relationship was found only in males, suggesting an interaction of this genetic variant with sex [77].

Ectonucleotide pyrophosphatase/phosphodiesterase-1 (ENPP1), which is involved in regulating pyrophosphate levels, is also known to inhibit calcification [113]. In a study in Austrian patients with end-stage renal failure, the patients with the Lys/Gln genotype of the Lys173Gln polymorphism in ENPP1 had a higher coronary calcium score and a higher carotid-femoral PWV than those with the Lys/Lys genotype after controlling sex, age, diabetes, and duration of dialysis treatment [78]. Further studies are needed to clarify the role of polymorphisms in calcium metabolism related genes in arterial structure and function.

G-Protein B3 Subunit Gene

The family of G-protein-coupled receptors interacts with heterotrimeric G protein complexes (α,β,γ subunits) to transduce extracellular signals to intracellular signals. Upon receptor-ligand binding, the α unit is separated from the receptor and β and γ subunits and then activates other molecules in the cell [114]. It was reported that in the C825T (Ser275Ser) polymorphism of the gene encoding the G-protein β3 subunit (GNB3), carriers of the T allele showed a significantly higher carotid-femoral PWV and augmentation index than those with the CC genotype in 72 young and healthy males [79] Further study is required to replicate this result in a large sample.

Factors Contributing to Inconsistent Results

The studies reviewed here illustrate conflicting results on the association between genetic polymorphisms and arterial stiffness. These inconsistent results can be attributable to several reasons. This review highlights the issues of different methods used to measure arterial stiffness and the selection of the study sample, which have impeded the progress of studies assessing the genetic component of arterial stiffness.

Phenotype of Arterial Stiffness

Various measures of arterial stiffness at different arterial sites can serve as a major source of inconsistent results. Commonly used methods in the studies reviewed include pulse wave velocity (PWV), augmentation index (Alx), arterial stiffness β, arterial compliance, arterial dispensability, and elastic modulus. Noninvasive measures of arterial stiffness can be grouped into three categories: measuring PWV, analyzing pulse waveforms, and measuring diameter change in relation to pressure change in arteries [24].

Determining PWV is generally accepted as the robust and reproducible method to measure arterial stiffness [25, 115]. PWV can be determined by the distance the pulse wave travels between two arterial sites divided by pulse transit time, PWV = D (meters)/Δt(seconds) [116]; thus, PWV provides regional stiffness between two arterial sites. Carotid-femoral PWV is considered the “gold-standard” measure of arterial stiffness because it is measured along the aortic and aorto-iliac pathway which provides the largest BP buffering function [116].

Pulse wave analysis (PWA) is a noninvasive method to generate an ascending aorta pressure wave from arterial pressure waveforms measured either at carotid or radial arterial site by mathematical transformation [117]. A pulse pressure waveform is a composite of forward pressure wave (incident wave) and reversed wave (reflected wave). The augmentation index (AI) is a commonly used measure of arterial stiffness and is obtained from PWA. AI is defined as augmented pressure (AP) by a reflected wave and is expressed as a percentage of pulse pressure (PP) [24]. Although AI is used to measure systemic arterial stiffness, because of differences in elasticity and length of arteries, pulse waveforms are not identical when measured in different arterial sites [118].

The methods to relate diameter change to pressure change (or reverse) include arterial compliance, distensibility, elastic modulus, Young’s modulus, and stiffness index β. These methods measure local arterial stiffness of measured sites. The issue regarding the difference between the site to measure diameter change and the site to measure pressure change has been pointed out in several studies [24, 119].

Taken together, methods to measure arterial stiffness represent either regional, systemic, or local arterial stiffness. Most studies have used different arterial sites, such as carotid, femoral, brachial, tibial, and radial arteries for arterial stiffness measurements. However, central arteries and peripheral arteries differ in their structure and function. While central arteries are more elastic containing smaller amount of smooth muscle, peripheral arteries are less elastic containing more smooth muscle cells and less elastin. Various arterial sites also respond differently to age, sympathetic nervous activity, hormones, and drugs [119, 120]. Thus, arterial stiffness measured even by the same method may not provide the same information when measured in different arterial sites. Due to heterogeneity of arterial system, polymorphisms may exert different effects on various arterial segments. Therefore, caution is required to interpret and compare the results from studies that used different methods to measure arterial stiffness in different arterial sites.

Furthermore, studies are recommended to standardize measures of arterial stiffness that best identify the functional and structural changes of arterial system over different age groups. According to the data from Framingham Heart Study offspring cohort, before age 50 years, carotid-femoral PWV was lower than carotid-brachial PWV, but after age 50 years, carotid-femoral PWV was higher, showing central and peripheral arteries have different progress over aging process [121]. The Anglo-Cardiff Collaborative Trial showed that while PWV was a good measure in older individuals (>50 years), AI is a more sensitive marker for arterial stiffness risk in younger individuals (<50 years). Measures of regional, systemic, or local arterial stiffness can be standardized for different age groups [122].

Selection of Study Sample

Arterial stiffness is a multifactorial condition which results from the interaction of several genes and environmental factors. Thus, identifying genetic determinants of arterial stiffness can be very challenging unless confounding factors (e.g., race, sex, pathological condition,) are well controlled. Studies reviewed in this paper provide suggestive evidence for racial differences in the associations between arterial stiffness and polymorphisms in ADRB1, ADRB2, ADRB3, and eNOS. Studies on the polymorphisms in ETAR, ETBR, eNOS, ER-α, MMP9, and fetuin A gene showed a sex-specific role for these polymorphisms in arterial stiffness. Furthermore, many studies have been trying to find novel determinants of arterial stiffness in patients with hypertension and atherosclerosis, in which their pathophysiological mechanisms are closely interrelated to ones of arterial stiffness.

Most studies on AGT, AGTR1, and CYP11B2 polymorphisms have been conducted in hypertensive patients but their associations have not been replicated by ones conducted in healthy individuals. Population-based longitudinal studies have shown that arterial stiffness may itself predict progression of hypertension in normotensive subjects [13, 14]. However, continuously elevated blood pressure will also contribute to arterial stiffness through structural and functional changes of the vascular wall, showing the “bi-directional” relationship between arterial stiffness and hypertension [123]. In addition, arterial stiffness progressively increases with age, and isolated systolic hypertension affects 50 % of people over age 60 [122]. In order to exclude the influence of hypertension on arterial stiffness, young and non-hypertensive subjects would need to be selected for the study populations. In addition, in all studies that measure arterial stiffness, mean arterial pressure and age which are major determinants of arterial stiffness should be adjusted in data analysis.

The terms of arterial stiffness and atherosclerosis have been used interchangeably, which present challenges in current arterial stiffness research. While atherosclerosis is characterized by a local inflammatory process and accumulation of fatty plaque in the intima, arterial stiffness (called arteriosclerosis) is more likely due to medial degeneration characterized by elastin degradation and collagen deposition [7]. Although several studies have shown the significant relationship between atherosclerosis and arterial stiffness, these studies have been conducted in patients with atherosclerosis [124], or elderly people who may already have progressive subclinical atherosclerosis [125]. Since currently available noninvasive measures of arterial stiffness assess both intimal and medial calcification, data from these subjects would not provide accurate information about the levels of arterial stiffness. Indeed, the significant relationship between atherosclerosis and arterial stiffness was not demonstrated in the studies with middle-aged US populations [126], healthy volunteers [127], middle-aged patients referred for transesophageal echocardiography [128], asymptomatic middle-aged men at cardiovascular risk [129], and women in the general population [130]. In their editorial commentary, Wilkinson and McEniery pointed out that “an important first step is the realization that athGenetic Variants Related to Arterierosclerosis and arteriosclerosis are different conditions.” [131] Studies on normotensive population without atherosclerosis may help to answer research questions of whether or not polymorphisms influence arterial stiffness independently of blood pressure and atherosclerosis.

A Proposed Conceptual Model of Arterial Stiffness Pathophysiology

Mechanisms contributing to arterial stiffness were proposed in Fig. 1. Although not many genetic variants have shown definitive causal associations with arterial stiffness, this model includes the genes that were investigated for their potential associations with arterial stiffness in the previous studies. Age, sex, and race, along with genetic makeup, can be important demographic factors that may influence arterial stiffness. There is evidence that psychological stress, diet, and physical activities are behavioral factors that intervene stiffness of arteries. Uncontrolled cholesterol, glucose, and blood pressure are known as critical factors that may predispose vasculature to be easily insulted by oxidative stress and inflammation [95, 123, 132].

Stiffness of medium and large arteries called arteriosclerosis can be attributable to alterations in function (e.g., endothelial function) and structure (e.g., vascular remodeling). Arterial walls consist of the endothelium, intima, media, and adventitia. The innermost layer of vascular wall, endothelium, releases endothelium-derived relaxing factor (EDRF). One EDRF is NO, which has been identified to play an important role in regulating vascular smooth muscle tone. Endothelial dysfunction from oxidative stress and inflammation process decrease the production of NO and may consequently increase arterial stiffness [133, 134]. Although NO influences smooth muscle tone, its effect on arterial stiffness seems to be relatively less than the stiffness caused by structural change in medial layer of the arterial walls. For example, it was demonstrated that inhibition of NO and cytochrome-related endothelial-hyperpolarizing factor (EDHF) significantly decrease smooth muscle tone and arterial wall stiffness assessed by local measure [134]. However, other studies could not find a significant association between endothelial function and carotid-femoral PWV in young healthy subjects [135], or in adults with type 1 diabetes [136].

Oxidative stress and inflammation are considered key mechanisms in the progression of atherosclerosis [132]. While normal endothelium does not bind leukocytes for a prolonged time, when endothelial cells undergo inflammatory activation, they increase the expression of various leukocyte adhesion molecules (e.g., vascular cell adhesion molecule (VCAM)-1) [137]. Monocytes bound to VCAM-1 penetrate into the intima in the presence of monocyte chemoattractant protein-1 (MCP-1), and become intimal macrophages. Macrophage scavenger receptors bind oxidized LDL, and these lipid-laden macrophages, called foam cells, secrets pro-inflammatory cytokines (e.g., interleukin-6, tumor necrosis factor-α, CD-40) and reactive oxygen species (ROS) [138], exacerbating oxidation and inflammation process in endothelium. As continued process of inflammation, vascular smooth muscle cells (VSMC), which are abundant in medial layer of the arterial wall, migrate to the intima, proliferate, and excrete extracellular matrix proteins that contribute to the development of fibrous plaques. This process explains the progress of a simple fatty streak becoming atheromatous plaques and calcificated lesions [139]. This intimal calcification is closely associated with atherosclerosis.

On the other hand, calcification in the medial layer of the arterial wall directly increases arterial stiffness [7]. Medial calcification is characterized by fragmentation of elastin lamellae and collagen accumulation [140]. These structural changes in the medial layer occur both in the presence or absence of atherosclerosis [139] and increase with age and metabolic diseases such as diabetes [141] In vivo studies have demonstrated that catecholamine released by sympathetic function plays an important role in regulating vascular smooth muscle contraction and arterial stiffness [142144]. Although the studies reviewed in this paper have not clearly demonstrated whether or not polymorphisms in RAAS genes play a dominant or initiating effect on arterial stiffness, several cultured cells and in vivo studies demonstrated that angiotensin II and aldosterone increase collagen synthesis in vascular smooth muscle cells [145], and that ACE inhibitors delay the accumulation of collagen in aorta [146, 147]. It was also reported that aldosterone increases arterial stiffness in absence of oxidative stress and endothelial dysfunction in an in vivo study [148]. Together, these analyses suggest that arterial stiffness may be induced independently of blood pressure and inflammation.

Although atherosclerosis and arterial stiffness (called arteriosclerosis) denote different conditions, current measures of arterial stiffness do not distinguish arteriosclerosis from atherosclerosis [7]. Furthermore, stiffened arteries increase stress within arteries and make the vascular environment more prone to atherosclerosis. In order to identify genetic variants of arteriosclerosis in clinical setting, careful selection of the study sample and proper adjustment of confounding factors should be considered.

Conclusion

In this paper, we reviewed the genetic association studies in arterial stiffness, and most findings showed inconsistent associations. These inconsistent results may have originated for several reasons. Among these existing studies, several methods have been used to measure local, regional, and systemic arterial stiffness at various arterial sites. The arterial system is not identical throughout body and the various arterial sites are different in their structure and characteristics. As such, the effect of genetic variants on arterial stiffness may not be same in different arterial sites. Thus, interpreting and comparing the results from different studies requires caution. Futures studies are also recommended to standardize local, regional, and systemic measures of arterial stiffness for different age groups. Furthermore, in the arterial stiffness studies, selection of participants with confounding underlying conditions such as hypertension and atherosclerosis can induce a biased conclusion. Arterial stiffness is very closely aligned with hypertension and atherosclerosis and they may share some pathophysiological pathways. Research findings would be more convincing when studies are conducted in a normotensive population with normal cholesterol levels. Lastly, the proposed model of arterial stiffness pathophysiology (Fig. 1) was developed to separate the concept of “measurement of arterial stiffness” and “arteriosclerosis.” This is useful to clarify that the current measures of arterial stiffness assess arteriosclerosis (e.g., medial calcification) which is combined with atherosclerosis (e.g., intimal calcification). Considerate adjustment of confounding factors along with careful selection of the study population may progress arterial stiffness studies.