Fetuin-A Does Not Explain Ethnic Disparity in Cardiometabolic Risk Factors and Subclinical Atherosclerosis Between Hispanics and Non-Hispanic Whites

Introduction

W e have previously demonstrated that among nondiabetic individuals without cardiovascular disease (CVD), mean coronary artery calcium (CAC) content, a measure of subclinical atherosclerosis and future CVD risk, ¹ was lower and the prevalence of those with a very high CAC score was reduced in Hispanics compared to non-Hispanic whites (NHW). ² This occurred despite the fact that several cardiometabolic risk factors such as insulin resistance, glucose intolerance and fasting triglyceride (TG) concentrations were higher in Hispanics compared to NHW. ² Subsequently, we reported similar findings in Hispanics with type 2 diabetes. ³ In agreement with our observations, lower CAC was documented in Hispanic participants of the large population-based Multi-Ethnic Study of Atherosclerosis (MESA) study. ⁴ This “Hispanic paradox” raises the possibility that one or more factors may be modulating both measures of insulin resistance and vascular calcification.

Fetuin-A (also named α2-Heremans-Schmid glycoprotein, AHSG) has been described as a serum-based inhibitor of calcification that forms a complex with calcium and phosphorus, increasing their solubility. ⁵ Genetic deficiency of fetuin-A in mice is associated with extensive extraosseal calcifications. ⁶ Low fetuin-A levels are associated with increased vascular and valvular calcifications in persons with end-stage renal disease. ⁷ In patients with coronary artery disease, low fetuin-A was associated with increased mitral calcification and aortic stenosis. ⁸ Fetuin-A is also a natural inhibitor of the insulin-stimulated insulin receptor tyrosine kinase. ⁹ Knockout of the fetuin-A gene in mice prevents high-fat diet-induced insulin resistance. ¹⁰ In humans, high plasma fetuin-A concentrations are associated with insulin resistance and predict worsening of insulin action ¹¹ and development of type 2 diabetes. ^12,13

We hypothesized that elevations in fetuin-A levels in Hispanics may explain both their reduced level of vascular calcification and their increased propensity toward insulin resistance and diabetes.

Materials and Methods

Serum fetuin-A concentrations were measured in samples from 76 of 80 age-matched, nonsmoking, healthy, 40- to 75-year-old male Hispanic and NHW subjects who participated in a previous study on ethnic differences in risk factors for atherosclerosis. ² Race and ethnicity were determined by self-identification. Hispanic ethnicity was confirmed if reported for at least two of their biologic grandparents. Fasting serum concentrations of fetuin-A were determined by a commercial sandwich enzyme-linked immunosorbent assay (ELISA; BioVendor, Modrice, Czech Republic). The intra- and interassay coefficient of variations (CV) were 3.5% and 5.4%, respectively. Other measurements included blood pressure, serum concentrations of fasting total cholesterol, low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C), TG, C-reactive protein (CRP), insulin, and glucose, as well as 2-h glucose concentrations during a 75-g oral glucose tolerance test (OGTT). ¹⁴ Serum concentrations of glucose and lipids were measured in the clinical laboratory of the Phoenix VA Health Care System using enzymatic assays in an automated analyzer (Abbott Laboratories, Abbott Park, Illinois). The Friedewald equation was used to estimate serum LDL-C concentrations. If serum TG levels were higher than 400 mg/dL, LDL-C was measured directly. Insulin and CRP assays were detailed previously. ² Insulin resistance was estimated by homeostasis model assessment (HOMA). Coronary scanning was performed using an Imatron C-150XL ultrafast computer-assisted tomographic scanner, as described previously. ² The lesions were scored using the algorithm developed by Agatston. ¹⁵ Intra- and interreader CV of CAC scoring was 1%. Statistical analyses were performed using the SAS program (v9.2; Cary, NC).

Results

Hispanics had higher fasting serum insulin and CRP, and 2-h glucose concentrations, as well as HOMA index values of insulin resistance (all P < 0.05), and tended to have higher TG and lower HDL-C levels (see Supplementary Table 1; Supplementary Data are available online at www.liebertonline.com/met). Hispanics demonstrated a trend for lower median CAC levels (17 vs. 64, P = 0.18) and had a lower prevalence of individuals with very high (>400) CAC (P = 0.008). In contrast to our hypothesis, serum fetuin-A concentrations were not significantly different between the two groups (Fig. 1A), and in fact trended slightly lower (not higher) in Hispanics. Fetuin-A levels were also similar between Hispanics and NHWs at both low and high levels of CAC (Fig. 1B). There was also no association between serum fetuin-A concentrations and CAC (Spearman correlation coefficient r = −0.02, P = 0.9), body mass index (BMI) (r = 0.10, P = 0.4), fasting insulin concentrations (r = 0.08, P = 0.5), fasting and 2-h glucose concentrations (r = 0.18, P = 0.1; r = 0.15, P = 0.2, respectively), HOMA (r = 0.10, P = 0.4), serum TG (r = 0.19, P = 0.10), or total and HDL-C levels (r = 0.04, P = 0.7; r = −0.13, p = 0.3, respectively), and serum CRP levels (r = −0.06, P = 0.6).

OPEN IN VIEWER

FIG. 1.

(A) Serum fetuin-A concentrations in Hispanics and non-Hispanic whites (NHW). (B) Serum fetuin-A concentrations by lower (n = 47) and higher (n = 29) coronary calcium categories in Hispanics (black bars) and NHW (white bars). Data are means ± standard error (SE). Group comparisons were performed by an unpaired Student t-test (A) or one-way analysis of variance (ANOVA) (B).

Discussion

Fetuin-A reduces both soft tissue and vascular calcification in several animal models. ^6,16 Moreover, low fetuin-A was associated with valvular calcification and atherosclerosis in patients with chronic kidney disease. ¹⁷ In contrast, Fiore et al. ¹⁸ demonstrated that fetuin-A levels were positively associated with the severity of peripheral atherosclerosis measured as intima media thickness of carotid and femoral arteries. Consistent with these later data, a recent analysis in a large European cohort with normal kidney function showed that higher fetuin-A levels were associated with an increased, not decreased, incidence of myocardial infarction and ischemic stroke. ¹⁹ Fetuin-A levels in our study were not associated with reduced CAC, perhaps indicating that fetuin-A may have little effect on the extent of atherosclerotic calcification in the setting of normal kidney function.

Previously, it has been demonstrated that increased fetuin-A concentrations are associated with insulin resistance and increased risk of type 2 diabetes. ^{11 –13} In contrast, we did not observe any association between measures of insulin resistance and fetuin-A in our study. One possible explanation is that the HOMA index is an inferior measure of peripheral insulin action compared to gold standard measures of insulin action such as the glucose clamp. However, the typical associations of CVD risk factors, such as TG, HDL-C, and CRP levels, with insulin levels and HOMA values were present, suggesting that the measures of insulin resistance were in fact reliable. It is also possible that because our participants were selected to be relatively healthy, the extent and variation of insulin resistance were reduced, thereby reducing our ability to detect a relationship.

In conclusion, differences in fetuin-A concentrations do not appear to explain the disparity in extent of coronary calcification and insulin resistance observed between nondiabetic Hispanics and non-Hispanic whites in this relatively small cross-sectional cohort. Further investigation is needed to identify the biological mechanisms underlying these important ethnic differences.