In the present study, we investigated the relationship between cardiovascular risk factors and arterial stiffness, with emphasis placed on novel risk factors, as ADMA and EPCs, among subjects with pre-diabetes.
From current knowledge, arterial stiffness represents the combined cumulative effect of traditional cardiovascular risk factors, with the full repertoire of factors implicated to be greatly obscure though. Arterial stiffness is an established predictor of cardiovascular disease, and its increase is associated with a 48% increase of cardiovascular disease risk.
According to the results of the present study levels of ADMA, LDL-C, as well as exercise, significantly and directly correlated with PWV in pre-diabetic subjects. The finding of LDL-C correlation with PWV is in accordance with the findings of others. In contrast with the above, other researchers failed to demonstrate any association between aortic stiffness and plasma lipoprotein levels. These conflicting results can be attributed to the different atherosclerotic burden among studied populations, the evaluation of different fractions of lipoproteins, the confounding effect of hypertension, and the use of different techniques/indices for the evaluation of large artery stiffness.
At the molecular level, high LDL-C levels induce endothelial cell dysfunction with subsequent decreased NO bioavailability due to impaired L-arginine transport and metabolism and eNOS uncoupling. Peroxynitrite, the major uncoupling byproduct of eNOs, has been reported to directly damage elastin[11, 29, 30]. The ensuing decreased NO bioavailability, also increases vascular smooth muscle cell tone, which further aggravates arterial stiffness.
Although arterial stiffness could potentially be improved with HMG-coA enzyme inhibition by statins in the present study, a lack of association between statin therapy and PWV was observed. However our findings are in agreement with previously published studies.
Several crucial studies have yielded contradictory results in relation to the association of ADMA with a PWV. In our study, ADMA increased arterial stiffness and this is also in agreement with the results of others. Given that NO participates in maintaining the structure of blood vessels, it is not surprising that increased expression of a natural nitric oxide synthase (eNOS) inhibitor, such as ADMA, has been linked to vascular stiffening.
In contrast, in the PREVENCION study, ADMA and NMMA (N-MonoMethyl Arginine) were not associated with large artery stiffness. The discrepancies between our analysis and PREVENCION study could be attributed to differences in study design and population characteristics. For example, the sample population in our analysis consisted of individuals with pre-diabetes, carrying a higher atherosclerotic burden. Moreover, in our study we investigated the correlation of cf-PWV with conventional risk factors after multivariate selection of variables including exercise, number of EPC, HOMA index, and use of antihypertensive or hypolipidemic treatment, which were not performed in the PREVENCION study.
In the present study, we did not observe any correlation between EPC count and arterial stiffness. In agreement with our findings, Palombo et al. reported no significant correlation between CD34+/KDR+ counts and c-f PWV in type 1 diabetics. In the same study LDL-C was proved to be independent predictor of c-f PWV in agreement with our findings. Although Yang et al. reported that impaired EPC activity was associated with reduced elasticity among hypertensive subjects, no significant correlation was again demonstrated between the number of circulating EPCs and arterial elasticity indices. It should be underlined though, that our results are not fully comparable with those of the previous study, since subjects with diabetes and smokers were excluded.
In contrast with the above, reduced numbers of circulating EPC have been associated with impaired arterial elasticity among diabetic patients[14, 35]. Given that EPC recruitment from bone marrow is NO mediated, it is expected that, as endothelial dysfunction worsens in parallel with glucose levels, EPC count would also decrease. Consequently our results are not totally comparable to others, since our study population consists of pre-diabetic individuals.
According to our findings, HbA1c levels did not correlate with PWV at univariate analysis. Fasting glucose levels positively and significantly correlated with PWV at univariate analysis but this correlation was lost at multivariate analysis. A trend for increased PWV across prediabetic categories was observed that did not reach statistical significance though and the above is most possibly due to the small sample size in the present study. It should be noted though that previous studies have shown that hyperglycemia and hyperinsulinemia explain only 30% of arterial changes associated with glucose intolerance and in non-diabetic individuals with lower levels of hyperglycemia no definitive correlation has been proved between arterial stiffness and IGT. From the theoretic point of view it has also been proposed that arterial stiffness in pre-diabetic individuals is, at least partially, NO mediated. Insulin resistance is associated with decreased NO bioavailability while hyperinsulinemia contributes to proliferation of arterial smooth muscle cells.
Our results demonstrate that among subjects with pre-diabetes moderate physical activity for at least 150 min per week can reverse arterial stiffness. In agreement with this finding Tanaka et al. have also reported that high levels of physical activity may prevent increases in arterial stiffness in elderly women. In the same study, it was also reported that in addition to aerobic fitness, plasma total cholesterol and plasma LDL-cholesterol levels were significant determinants of central arterial stiffness.
In our sample of prediabetic patients age significantly correlated with increased arterial stiffness at univariate analysis but at multivariate analysis this correlation lost significance. However as previously reported, arterial elasticity is not determined only by the structural components of the vessel wall, but NO can also modify arterial stiffness. Indeed Ngo et al. reported that NO synthesis may represent the basis for increased arterial stiffness in ageing individuals. Even though age is a major determinant of arterial stiffness, several studies reported that increase in large artery stiffness, follows an non linear evolution, implying that an increase in PWV may be observed over the age of 55 years; indeed, in our sample mean age was 50.2 years, well below the 55 year-threshold[41–43].
Although blood pressure is traditionally considered a major determinant of the arterial stiffness, such correlation did not persist in the present study at the multivariate analysis. High levels of systolic blood pressure are believed to increase arterial stiffness due to structural and functional alterations in the arterial media and higher systolic blood pressure levels have been connected with greater increases in arterial stiffness in several prospective studies. Although the relationship of blood pressure with vascular stiffness seems to be bidirectional, investigators of the Framingham Offspring cohort, concluded that vascular stiffness is a precursor rather than the result of hypertension. Other studies have also failed to document SBP as an independent predictor of increased arterial stiffness[46–49] and it has been proposed that increased PWV lays in the basis of increased arterial pressure and not the opposite on a long-term basis. From this point of view the cross sectional nature of our study might have obscured causality between arterial stiffness and MBP that needs time to develop.
A limitation of our study pertains to the relatively small population size. Due to the small sample size, certain secondary observations may not have reached statistical significance; nevertheless, the fact that statistically significant associations emerged and persisted at the multivariate analysis, points to the validity and robustness of the findings presented herein. In addition, the cross-sectional study design does not allow the substantiation of causality, the latter necessitating prospective cohort studies.
Unfortunately in this analysis, qualitative alterations of EPC, such as colony formation, migration or adhesion, were not examined, but in most cases, the number and the functionality of EPC are essentially two sides of the same coin; quantitative and qualitative EPC properties are indeed co-regulated by the same molecular pathways, and consequently decreased EPC count is commonly accompanied with dysfunction.
Finally, the use of a control group was not deemed necessary, since the main objective of the present study has been the investigation of correlations between novel and traditional cardiovascular risk factors and arterial stiffness in a specific subpopulation (prediabetics), in which the major, independent determinants of PWV were sought. The validation and replication of previously documented differences (such as comparisons of individuals with pre-diabetes versus controls) was considered redundant and was not performed. Indeed according to the findings of the Hoorn study, severity of arterial stiffness seems to correlate with degree of glucose metabolism impairment with IFG and IGT patients to exhibit increased arterial stiffness that is intermediate between type II frank diabetics and subjects with normal glucose tolerance[5, 37] and the above findings have been verified by other studies.
Regarding EPC counts, pre-diabetes and diabetes have been directly connected with reduced numbers of circulating EPCs compared with normal glucose tolerance status in several studies[13, 14]. Numerous studies have also established that ADMA levels are increased in states of impaired glucose metabolism, in relation to normal subjects, and this increase seems to positively correlate with glycemic category and metabolic control[51, 52].
Nevertheless the lack of a control population may be considered a major limitation of the present study. Indeed, an age matched control group may have significantly contributed to the comparison of novel markers (such as EPC and ADMA) between patients with pre-diabetes and healthy subjects. The control group could also have proven that all methods yielded results comparable with the literature. Future studies adopting control groups seem necessary in our setting so to further elaborate and validate the aforementioned findings. In our knowledge, this is the first study to investigate the correlation of traditional and novel cardiovascular risk factors, as AMDA and endothelial progenitor cells and arterial stiffness, as evaluated by PWV, in pre-diabetic individuals.