Study participants were residents living in Nonglin community in the center of Guangzhou, China. Residents were recruited in a community-based cohort aiming to prevent stroke by the organization of Department of Neurology, the First Affiliated Hospital, Sun Yat-sen University from July 2008 to December 2008. Residents who were volunteers, aged from 46 to 75, and willing to be reexamined in future were eligible. Residents were excluded if they lived in other districts, refused to complete necessary questionnaires and examinations in the study or had malignant tumors, acute myocardial infarction, and other critical illnesses which would possibly affect the follow-up in the future. Altogether 962 participants were recruited. The study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University and all participants gave written informed consent of the study. In the present study, participants were further excluded if they had diseases which corroboratively or potentially affect the function of the immune system including cardiovascular disease, cerebrovascular disease, chronic renal failure, chronic liver insufficiency, blood diseases, asthma, acute infectious diseases, or autoimmune diseases. After the present study criterion screening, 534 participants were eligible for our present study. Among these participants, there were 183 (34.3%) subjects with CAS, other 351 (65.7%) subjects without CAS were taken as controls. Eventually, random sampling was used to select the sample for the study. Finally 70 participants were respectively selected from the CAS and control groups. Totally 140 subjects were studied in our present study.

Demographic variables, histories of previous diseases, cigarette smoking, and alcohol consumption were obtained by well trained investigators by face to face interview in clinics using a standardized questionnaire. Cigarette smoking was defined as the subject had smoked more than 1 cigarette per day for more than half a year. Alcohol consumption was defined as the subject had drunk any kind of alcohol drinks more than once a week regularly for more than half a year. Previous histories of hypertension, diabetes, hyperlipidemia, heart disease or stroke were defined if the subject had been diagnosed with the disease in hospital before or had used medications regularly. Waist and hip circumference, weight and height were measured and then waist hip ratio (WHR) was calculated. Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m²). Blood pressure was measured with a mercury sphygmomanometer in the morning.

Fasting venous blood samples for fasting blood glucose (FBG), total cholesterol (TC), triglyceride (TG), low density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), high-sensitivity C-reaction protein (hs-CRP), and uric acid (UA) were collected in the morning and then analyzed by the enzyme method in autoanalyzing machine. The coefficients of variance of repeated measurements of the chemistry tests were 1.0% to 1.6%.

Carotid ultrasound measurements were performed with one B mode ultrasound system, APLIO XU equipped with a 7.5 MHz linear array transducer (TOSHIBA, USA) by 2 experienced vascular sonographers. Subjects were examined in supine position with their necks extended. The three bilateral defined locations including common carotid (20 mm proximal to the bifurcation), carotid bifurcation and internal carotid artery (ICA, 10 mm distal to bifurcation) were measured longitudinally and transversely. The scan and measurements were performed according to internationally recognized guidelines [15] . Intima-media thickness (IMT) was measured as the distance between the two parallel echogenic lines on the far wall of artery in longitudinal plane image frozen in the screen by electronic calipers [16] . Plaque was defined as localized thickening of IMT ≥ 1.3 mm which did not uniformly involve the whole wall of carotid artery [17] . Classification of ultrasound plaque appearance was performed independently from the videotape by 1 single operator blinded to all other data. 3 categories were classified as 1) no plaques; 2) echogenic plaques; 3) echolucent plaques. Echolucent plaques are mainly composed by the deposition of cholesterol, necrosis tissue and plaque bleeding, a type of plaque with instability [18] . The carotid stenosis measurements were performed according to internationally recognized guidelines [8] . The peak systolic velocity (PSV) and end diastolic velocity (EDV) were measured in the common and internal carotid arteries, and the degree of stenosis was then calculated according to St. Mary's Ratio which is the ratio of the PSV in the internal carotid arteries, a value that increases with degree of stenosis, divided by the EDV in the distal common carotid artery, a value that decreases with increasing resistance in the internal carotid arteries due to stenosis (see Table 1) [19] . We had assessed interobserver reproducibility of CAS detection by repeated scans on several randomly selected participants. The kappa value for the agreement of CAS was 0.738 (P < 0.01), which was acceptable.

Circulating serum neopterin concentrations were measured using a commercially available solid phase enzyme-linked immunosorbent assay (ELISA) kit (IBL, Hamburg, Germany) based on the basic principle of a competitive ELISA. An unknown amount of antigen in the sample and a fixed amount of enzyme labelled antigen compete for the antibody-binding sites (rabbit-anti-neopterin). Both antigen-antibody complexes bind to the wells of the microtiter strips coated with a goat-anti-rabbit antibody. Unbound antigen is removed by washing. The intensity of the color developed after the substrate incubation is inversely proportional to the amount of antigen in the sample. Results of samples can be determined directly using the standard curve. In our study, inter-assay and intra-assay coefficients of variation were <5.9% and <8.9% respectively.

Analysis of normality of the continuous variables was performed with the Kolmogorov-Smirnov test. Results for continuous variables normally distributed were expressed as mean value ± standard deviation (SD) and non-normally distributed (neopterin levels) was presented as median value (lower quartile-upper quartile). Dichotomized or categorized variables were described as numbers and proportions. In univariate analysis, continuous variables normally distributed were compared by student’s t test. Continuous variable non-normally distributed was compared by nonparametric Mann-Whitney U test. Dichotomized or categorized variables were compared by Chi-square tests. The correlation between serum neopterin and hs-CRP levels was assessed using Pearson’s coefficient correlation analysis. In order to adjust the confounding factors which is significant in univariate analysis, binary logistic regression analysis was performed to investigate the risk factors for CAS. Odds ratio (OR) was calculated by logistic model. The receiver operating characteristic (ROC) curve of the serum neopterin concentrations for predicting CAS and its area under curve (AUC) analysis were performed. All statistic analysis was calculated with SPSS 13.0 software system (SPSS Inc., Chicago, IL, USA). A two-sided P-value of less than 0.05 was considered as statistically significant.

In this research, totally 534 community subjects who had no diseases which corroboratively or potentially affected the function of the immune system including cardiovascular and cerebrovascular disease were eligible after screening of exclusion criterion, with an age of 57.1 ± 5.6 years. The overall baseline clinical characteristics and biochemical results are shown in Table 2. Carotid ultrasound revealed 183 (34.3%) subjects with CAS. In these subjects, most were mild stenosis (<50%), only 7 subjects were moderate stenosis (50% - 70%).

Variables of risk factors of the participants in different groups are shown in Table 3. There were significant differences in age, WHR, BMI, history of hypertension, SBP, HDL-C and neopterin level (P < 0.05). Neopterin level in the CAS and control group was 5.4 (3.3 - 7.3) nmol/L and 4.6 (3.6 - 5.4) nmol/L respectively (Figure 1). However, hs-CRP level in the CAS and control group was 0.91 (0.52 - 5.81) mg/L and 1.02 (0.73 - 7.04) mg/L respectively, and the difference was not significant (Table 3).

Neopterin level in the no plaque group, echogenic plaque group and echolucent plaque group were 4.6 (2.8 - 7.8) nmol/L, 4.7 (2.3 - 7.6) nmol/L and 6.0 (3.8 - 8.1)

nmol/L respectively (Figure 2). There were significant differences between no plaque group and echogenic plaque group, no plaque group and echolucent plaque group (P < 0.05).

Figure 3 shows that there was a no significant correlation between serum neopterin and hs-CRP levels (r = 0.121, P = 0.152)

In order to adjust the confounding factors which was significant in univariate analysis, binary logistic regression analysis was performed to investigate the risk factors for CAS (shown in Table 4). After multivariate adjusted, neopterin levels remained an independent factor. With per 1.0 nmol/L (approximately 1 IQR) increased, the OR value was 1.84 (95% CI: 1.25 - 2.71, P < 0.05).

Figure 4 shows the ROC curve of serum neopterin concentrations for the diagnosis of CAS. The AUC was 0.665 (P < 0.01). The serum neopterin cutoff value

*Approximately 1 SD; ^▲Approximately 1 IQR.

of 5.1 nmol/L had a sensitivity of 68.6% and specificity of 61.5% in detecting CAS.