Consecutive patients with SSc, according to the American College of Rheumatology criteria 14, attending our outpatient clinic were included. Patients were subclassified in subsets as defined by LeRoy and colleagues 1516. Forty-nine patients were included. Pregnancy was an exclusion criterion. Healthy subjects were included in this study as controls. Ethical approval for the study was obtained from the Medical Ethical Committee of the University Medical Center Groningen (University of Groningen, Groningen, The Netherlands). Informed consent was obtained from each participant.

Data were obtained from all subjects with respect to traditional risk factors for cardiovascular disease (CVD), including body mass index (BMI), smoking status, diabetes, blood pressure, lipid levels, and family history of CVD (considered positive if first-degree relatives suffered from CVD before 60 years of age).

In patients with SSc, we assessed disease-related factors as possible determinants of macrovascular disease. To assess disease activity, the preliminary European Scleroderma Study Group (EScSG) activity index (a score ranging from 0 to 10) was used. A score higher than 3 denotes active disease 1718. Also, the revised preliminary SSc severity scale (Medsger severity scale), a measure of activity, damage, and severity, was used. This scale is a nine-organ disease severity scale in which for each organ system a score of 0 to 4 is applied, with 0 being normal and 4 denoting end-stage organ involvement 19. We also recorded statin use, cumulative prednisolone dose, and current or former use of immunosuppressive agents.

B-mode ultrasonography was used to measure the common carotid IMT. Measurements were limited to the CCA and not extended to other segments. The prevalence of increased IMT and/or plaques is substantially higher in the carotid bulb or internal carotid artery (ICA), but the intended quantitative comparison between SSc patients and controls may be hindered by including these segments. An Acuson 128XP ultrasound device with a 7 MHz linear array transducer (Acuson Corporation, now part of Siemens Medical Solutions USA, Inc., Malvern, PA, USA) was used for measuring IMT in all patients and controls, as described before 720. With the subjects in the supine position, the left CCA wall segment was scanned from a lateral transducer position and recorded on s-VHS tape. The CCA wall segment was defined as 1 cm proximal to the carotid bifurcation. The far wall of the left common artery was assessed at three different positions. Off-line video analysis, using a previously described image analysis program 21, was performed by one reader unaware of patient or control group data or characteristics. The highest IMT value found in this segment was considered to be the maximum IMT, and the mean of three measurements in this segment was considered to be the mean IMT.

Lipid levels were measured by routine techniques. Additional serum and plasma samples for determination of markers of endothelial activation were stored at -20°C until analysis. Serum levels of vascular cell adhesion molecule-1 (s-VCAM-1) (R&D Systems Europe Ltd, Abingdon, UK) and thrombomodulin (TM) (Diaclone, Besançon, France) were measured according to the manufacturers' instructions. Serum was used to determine C-reactive protein (CRP) and plasma was used to determine von Willebrand factor (vWF) using in-house enzyme-linked immunosorbent assays, as described before 7.

Values are expressed as mean ± standard deviation when variables were normally distributed and as median with interquartile range (IQR) (25th to 75th percentile) in case of a non-normal distribution. Because the number of patients with diffuse cutaneous SSc (dSSc) was low in our patient group, subset analysis could not be performed. Differences between patients and controls were assessed by Student t test, Mann-Whitney U test, and χ² test (Pearson chi-square or Fisher exact test) as appropriate. Linear regression was used to assess the relationship between IMT and the different groups (SSc patients and healthy controls). An unadjusted analysis in which no corrections were made for possible confounders and an adjusted analysis in which corrections were made for possible confounders are presented. The variables were entered manually. The level of significance for the association between group and outcome variable was set at a P value of less than 0.05. The variables age, gender, BMI, and smoking were also studied as potential effect modifiers in the relationship of interest. The correlation between maximum IMT and disease-related factors and endothelial markers was assessed by Spearman rank correlation coefficient since maximum IMT was non-normally distributed. All analyses were carried out with the Statistical Package of Social Science, version 12.1. for Windows (SPSS Inc., Chicago, IL, USA).

The demographic characteristics and the traditional risk factors of SSc patients and healthy controls are shown in Table 1. Patients tended to be older (55.4 ± 11.6 versus 50.9 ± 10.1 years; P = 0.078) and used significantly more statins (14% versus 0%; P = 0.038) than healthy controls. Patients had lower levels of high-density lipoprotein (HDL) cholesterol (1.40 mmol/L, IQR 1.23 to 1.80 mmol/L, versus 1.68 mmol/L, IQR 1.48 to 1.89 mmol/L; P = 0.027) and higher levels of triglycerides (1.36 mmol/L, IQR 1.16 to 2.14 mmol/L, versus 1.17 mmol/L, IQR 0.77 to 1.67 mmol/L; P = 0.030) than controls. No significant differences were found in other cardiovascular risk factors. Four patients had a history of macrovascular events compared with none in the control group.

Patients with SSc had a median disease duration of 6 years and had experienced Raynaud phenomenon for almost 11 years. Limited cutaneous SSc (lcSSc) was present in 92% of patients and diffuse cutaneous SSc was present in 8% of patients. Patients had a median modified Rodnan skin score of 7.0 (IQR 4.5 to 14.0), a preliminary EScSG disease activity index of 0.5 (IQR 0.5 to 1.75), and a revised preliminary SSc severity scale score (Medsger severity scale score) of 6.0 (IQR 4.5 to 7.0). The preliminary EScSG disease activity index may have been an underestimation of reality since not all variables (especially erythrocyte sedimentation rate [ESR] and complement) were available. As shown in Table 1, CRP levels were not substantially elevated, suggesting normal ESR levels. However, patients had significantly higher CRP levels than controls (3.5, IQR 1.6 to 7.0, versus 0.8, IQR 0.3 to 2.0; P < 0.001). Forty-three percent of patients were current or former users of prednisolone, with a median cumulative dose of 3.6 g (1.9 to 16.1 g) (Table 2).

The median values for the mean IMT measurements in the CCA were 0.69 mm (IQR 0.62 to 0.79 mm) in patients and 0.68 mm (IQR 0.56 to 0.75 mm) in controls (Figure 1a). Also, IMT values were comparable between SSc patients and controls for each age decade. Among the four outliers in the group of SSc patients with an IMT of greater than 1.10 mm, one patient had a history of a cerebrovascular accident and one patient was known to have left ventricular dysfunction probably caused by coronary artery disease. Both had other cardiovascular risk factors like current or former smoking and hypertension. The other two patients were not known to have clinically manifest CVDs. One of these patients had a family history of CVD. Other risk factors were not present. Although plaques were not a primary endpoint, they were observed in three patients and not in healthy controls. Linear regression analysis of the mean IMT in the CCA between controls and patients, when not corrected for possible confounders, demonstrated no significant difference in mean IMT (B = 0.101; P = 0.067) (Table 3). Also, no significant differences were seen between the groups after correction for the strongest confounders. Both in the model with the confounders age, HDL cholesterol, and low-density lipoprotein (LDL) cholesterol (B = 0.042; P = 0.328) and in the model with the confounders age, HDL cholesterol, and history of macrovascular disease (B = 0.030; P = 0.474), no differences were found between patients and healthy controls. The possible relationship between mean IMT and SSc patients versus controls was lost when traditional risk factors were entered into the regression analysis. The addition of the confounder history of hypertension did not change the outcome. No correction for systolic and diastolic blood pressure, use of statins, smoking, diabetes mellitus, gender, BMI, total cholesterol, or triglycerides was necessary since these outcome parameters were not confounders in the linear regression model. No effect modification by age, gender, BMI, or smoking was found.

No significant differences were seen for the maximum IMT of the CCA between SSc patients (0.83 mm, IQR 0.70 to 0.97 mm) and healthy controls (0.77 mm, IQR 0.70 to 0.88 mm) by means of linear regression analysis before or after correction for the confounders age, HDL cholesterol, gender, and history of macrovascular disease (Figure 1b and Table 4). The addition of the other confounders (that is, history of hypertension, diastolic blood pressure, and LDL cholesterol) did not change the outcome. In patients, no correlations were found between maximum IMT, CRP, and disease-related variables.

Markers of endothelial activation were not increased in patients with SSc. Compared with controls, levels of VCAM-1 were even decreased (229 ng/mL, IQR 188 to 311 ng/mL, versus 287 ng/mL, IQR 236 to 350 ng/mL; P = 0.014). No differences between SSc patients and controls were found in levels of vWF (72%, IQR 34% to 125%, versus 71%, IQR 48% to 110%; P = 0.691) and TM (3.8 ng/mL, IQR 2.3 to 5.0 ng/mL, versus 2.9 ng/mL, IQR 2.1 to 3.7 ng/mL; P = 0.151) (Figure 2). Levels of endothelial markers were not correlated with maximum IMT.