6.4 Metabolic syndrome and early atherosclerosis
6.4.2 Metabolic syndrome as a risk factor beyond the Framingham
Few studies have elaborated the association of MetS and CIMT after taking the Framingham Risk Score into account. Hassinen et al. found in a prospective
study with elderly women that MetS is associated with accelerated progression in CIMT independently of this risk score (Hassinen et al. 2006). Teramura et al. also suggested that MetS, as defi ned by Japanese criteria, would have additive predictive value on CIMT beyond the Framingham score values (Teramura et al. 2007).
On the other hand, Ahluwalia et al. suggested that MetS would not have such an additive effect (Ahluwalia et al. 2006). To the best of the author’s knowledge, sex-related differences in the associations between MetS, the Framingham Risk Score and CIMT have not been studied previously. In the present study, these associations were modifi ed by sex. The study population was divided into four risk profi le categories according to the Framingham score value and MetS status.
In men, CIMT was signifi cantly higher if the score value was high, regardless of the MetS status. In women, however, CIMT was signifi cantly higher if MetS was present, regardless of the Framinghan Risk Score category. These fi ndings suggest that MetS has a different, and in some circumstances stronger, association with early atherosclerosis in women than in men.
In conclusion, the present results suggest that MetS is, in both sexes, associated with early atherosclerosis independently of the risk factors that are not included in its defi nition. These associations are, however, modifi ed by sex, and MetS may increase the risk of early atherosclerotic changes more in women than in men. In men, traditional risk factors, as defi ned by the Framingham Risk Score, are strong determinants of CIMT, and MetS appears to offer no additional information. In women, however, MetS may add to the risk of subclinical atherosclerosis beyond the score values. This additive relationship was observed particularly in women with a low risk score, indicating that MetS is an important risk factor for women with a low cardiovascular risk according to the traditional risk factors. The NCEP and IDF defi nitions were both similarly associated with PWV, which suggests that both defi nitions are able to identify subjects with increased arterial stiffness.
6.5 Coronary heart disease-associated locus on chromosome 9p21.3 and early atherosclerosis
The chromosome 9p21.3 locus has a robust association with CHD, which has been demonstrated with a wide range of populations (Burton et al. 2007, Helgadottir et al. 2007, McPherson et al. 2007, Samani et al. 2007, Broadbent et al. 2008, Schunkert et al. 2008). The SNP (rs1333049) included in the present study has had the strongest association with CHD in these genome-wide association studies. The region associated with CHD is located adjacent to genes that play a central role in the regulation of the cell cycle and may be related to atherosclerosis through
94
their role in transforming growth-factor-β-induced growth inhibition (Hannon and Beach 1994, Kalinina et al. 2004). In this context, it is relevant to examine the association of this locus with different stages of atherosclerosis. Interestingly, one study has also shown an association of the locus with abdominal aortic aneurysms as well as with intracranial aneurysms, suggesting that the mechanism of its effect on the vascular wall might be more complex than simple promotion of the development of atherosclerosis (Helgadottir et al. 2007). Furthermore, the SNP of interest has not been identifi ed in genome-wide association studies investigating stroke (Matarin et al. 2007, Ikram et al. 2009) or peripheral arterial disease (Koriyama et al. 2010). On the other hand, in some candidate gene studies this particular SNP has been signifi cantly associated with stroke (Smith et al. 2009) and peripheral arterial disease (Cluett et al. 2009).
The association of the 9p21.3 locus with early atherosclerosis has not been examined extensively. Studying the relationships between the risk factors and arterial wall thickness in a young predisease cohort provides a useful means of investigating the early impact of such determinants on atherosclerosis. Because genetic determinants could be active from a young age, the association of this polymorphism with CIMT was studied in a cohort of young adult subjects. It was also studied in a second cohort with a similar age range as in those in which the association of the locus with CHD was originally demonstrated. Somewhat surprisingly, no evidence was observed in support of an association between this polymorphism and any of the CIMT parameters in either age group. Similarly, there was no signifi cant association between the polymorphism and brachial FMD in the Young Finns Study. After the publication of this study, others have reported similar results (Cunnington et al. 2009).
Some possible explanations for the lack of association between the studied SNP and early atherosclerosis must be discussed: 1) Both cohorts were population-based, ethnically homogeneous, and of Caucasoid origin where the association of CHD with this polymorphism has been well demonstrated (Burton et al. 2007, Helgadottir et al. 2007, McPherson et al. 2007, Samani et al. 2007, Schunkert et al. 2008). Therefore, the population structure does not offer a plausible explanation for the fi ndings. 2) The measurements of CIMT and FMD were performed using standardised protocols by trained personnel. Therefore, it is not likely that the observations would be explained by imprecise measurements. 3) The power calculations in the present data have shown that there was 80% power at an alpha of 0.05 to detect 0.02 mm and 0.06 mm differences in CIMT between CC and GG subjects in the Young Finns Study and the Health 2000 cohort, respectively.
This means that there was 99% power at an alpha of 0.01 to detect a 0.2 mm difference in both cohorts. Previous literature has demonstrated that an increase in CIMT by 0.1 mm would elevate the risk for future CHD by 20% to 30% (Bots et al.
1997, Hodis et al. 1998, O’Leary et al. 1999). According to a recent meta-analysis each copy of the risk allele (allele C) of this polymorphism is associated with a 24% increase in the risk of CHD (95% CI: 20% to 29%) (Schunkert et al. 2008).
Therefore, the expected effect of this SNP on CIMT would be approximately 0.1 mm per allele, assuming that the association with CHD was mediated through a similar mechanism. Moreover, several previously reported associations between other cardiovascular risk factors and CIMT were easily detected in both cohorts.
Based on this knowledge, it does not seem likely that inadequate power to detect an effect would explain the lack of association between the 9p21.3 locus and CIMT.
One plausible explanation for the lack of association between this CHD-associated locus and CIMT would be that it is related to the risk of CHD through mechanisms that are not refl ected by changes in arterial wall thickness. Similarly, the fi ndings suggest that this polymorphism does not enhance the risk of CHD primarily causing endothelial dysfunction at a young age. If the mechanism of the effect of the 9p21.3 locus on CHD is related to cell growth and turnover, it is possible that it affects coronary plaque stability or vulnerability rather than its development per se.