This study demonstrates, for the first time, that SNP rs1333049 on chromosome 9p21.3 is an important determinant for the progression of increasing arterial stiffness, especially that of the aorta, in a Thai population. Increased afPWV and aaPWV are associated with increasing number of MetS risk factors, indicating that a clustering of MetS risk may interact synergistically to affect central arterial stiffness. Previous studies have suggested that arterial stiffening in MetS, particularly in diabetic individuals, is associated with elastic lamellae fragmentation and calcification, increased collagen cross-linking, low-grade inflammation, increased smooth muscle basal tone and endothelial dysfunction [22–25]. These processes may be influenced by both genetic and environmental factors [6–8, 23]. As the integrity of the aortic wall is dependent on effective load bearing by the elastin collagen network, one could speculate that the SNPs on chromosome 9p21.3 might affect the wall integrity either through increased elastin fracture, degradation due to enhanced matrix metalloprotease activity, calcification or enhanced collagen production.
A previous study has reported that central arterial stiffness, as measured by heart-femoral PWV (hfPWV) was more closely associated with CAD, CVD and peripheral arterial disease than peripheral arterial stiffness . hfPWV, carotid-femoral PWV (cfPWV) and afPWV, as used in this study may all be regarded as measures of ‘aortic’ stiffness since in all cases the major part of the path traversed by the pulse wave is the aorta and a recent consensus document confirms that cfPWV is an acceptable estimate of aortic stiffness . However, it should be noted that in older people the ascending aorta lengthens with age  as does the common iliac artery, which comprises approximately the distal third of the path , and this will add an error to the path length measurements, leading to an overestimation of PWV. Previously, a cut-off point of hfPWV at 11.18 m/s and cfPWV at 12 m/s showed good discrimination in the validation set on the Sphygmocor system and the Colins system, respectively . Deng et al. set a cut-off value of cfPWV at ≥ 9 m/s to diagnose arteriosclerosis in patients with essential hypertension  and more recently 10 m/s has been proposed as the threshold above which there is an increased risk of cardiovascular events . In the present study, we assessed aortic stiffness in all subjects using the Arterial Compliance Monitor which measures afPWV rather than cfPWV and chose to use our mean value of afPWV (11 m/s). This is close to the values recommended in the consensus document and revealed a clear difference in the association between afPWV and rs1333049 genotypes. We found that the frequency distribution of the rs 1333049 genotypes is significantly associated with afPWV groups (P<0.05). We conclude that, in spite of uncertainties in the precise value of PWV used to dichotomise the data, there is a clear association between polymorphism of rs1333049 and raised arterial stiffness.
In addition to the aforementioned association, we found increased aortic stiffness in subjects homozygous for the C-allele of rs 1333049 on chromosome 9p21.3, after adjusting for age and the number of MetS risk factors. We suggest the SNP rs1333049 might be involved with the progression of raised aortic stiffness in Thais, although the underlying mechanisms by which SNP rs1333049 on chromosome 9p21.3 contributes to this progression remain unclear. Previous studies have suggested that the chromosome 9p21 locus predicts CAD and involves the initiation or facilitation of atherosclerosis [32, 33]. Wang et al. demonstrated that rs 1333049 polymorphism is an independent determinant of coronary plaque progression in a Chinese Han population . Moreover, it was found that chromosome 9p21 locus is associated with recurrent myocardial infarction or cardiac death following an acute coronary event . Taken together, the genetic variants in this 9p21 locus increase the risk of CAD by promoting atherosclerosis development and plaque instability.
A large noncoding RNA gene named ANRIL(CDKN2B-AS) is expressed in several cells and tissues, where a number of SNPs located at the 5’ end including rs1333049 have been suggested to be associated with CAD . The ANRIL gene possibly coordinates transcriptional regulation of 2 cyclin-dependent kinase inhibitors (p16/CDKN2A and p15/CDKN2B) under both physiological and pathological conditions. A SNP rs1333049 was located in a 190 kb region of the high linkage disequilibrium, upstream both of p16/CDKN2A, p15/CDKN2B and p14/ARF that code for p15 and p16 and p14ARF . These proteins are involved in cell cycle regulation, and also affect the atherosclerotic process by inhibiting the transformation of tumour growth factor-β, a well known tumour suppressor protein . SNP rs1333049 has not only been reported to increase the risk of CAD [38, 39] but also of abdominal aortic aneurysms [16, 40], these diseases being preceded by altered arterial stiffness. Therefore, our finding of the homozygous CC genotype of SNP rs1333049 in subjects with high aortic stiffness is in good agreement with those observed in CVD patients.