High Prevalence of Ultrasound Detected Carotid Atherosclerosis in Subjects with Low Framingham Risk Score: Potential Implications for Screening for Subclinical Atherosclerosis


The cardiovascular (CV) risk assigned by the Framingham risk score (FRS) misses many subjects destined for CV events. Coronary artery calcification (CAC) as measured by computed tomography and carotid intima-media thickness (CIMT) and plaque assessment using B-mode ultrasound can identify subclinical atherosclerosis. The comparative relation of CAC and CIMT and carotid plaque after integration into the FRS is not established. The aim of this study was to develop a CV screening approach incorporating FRS, CAC, and CIMT.


The prevalence of subclinical atherosclerosis, defined as CAC score > 0, CIMT ≥ 75th percentile, or plaque ≥ 1.5 mm, was determined in the groups with low, intermediate, and high FRS among 136 asymptomatic subjects. The CIMT and CAC values were used to determine “vascular age” and “coronary calcium” age, respectively, with established nomograms.


In the 103 low-risk (FRS < 10%) subjects, 41%, 50%, 59%, and 66% had CAC scores > 0, CIMT ≥ 75th percentile, plaque ≥ 1.5 mm, and CIMT ≥ 75th percentile or plaque ≥ 1.5 mm, respectively. In the 33 subjects with intermediate (n = 14) or high (n = 19) FRS, 70%, 81%, 87%, and 87% had CAC scores > 0, CIMT ≥ 75th percentile, plaque ≥ 1.5 mm, and CIMT ≥ 75th percentile or plaque ≥ 1.5 mm, respectively. Fifty-two percent of subjects with coronary calcium scores of zero had carotid plaque. Adjusted for FRS, body mass index was an independent predictor of abnormal CIMT in the low-FRS group, but not of abnormal CAC. Mean vascular CIMT age was significantly higher than coronary calcium age (61.6 ± 11.4 vs 58.3 ± 11.1 years, P = .001), and both were significantly higher than chronologic age (56.9 ± 10.1 years) ( P < .0001 and P < .04, respectively). CIMT upgraded or downgraded FRS by >5% in more cases than CAC (42% vs 17%).


In asymptomatic patients without CV disease, CIMT and plaque assessment are more likely to revise FRS than CAC. Body mass index predicts increased CIMT in low-FRS subjects. These findings may have broad implications for screening in low-FRS subjects.

Atherosclerosis is a common and often lethal disease of the arteries of the heart, brain, and periphery. Many patients with atherosclerosis, even those with severe disease, are asymptomatic and thus unaware of its presence. In 40% to 60% of these individuals, the first indicator of atherosclerosis is an acute myocardial infarction or sudden death. Although potentially modifiable risk factors account for a majority of the population-attributable risk for an initial myocardial infarction, many patients with similar risk factors never develop myocardial infarctions. Similarly, cardiovascular (CV) events continue to occur in patients who harbor unrecognized atherosclerosis but are not receiving risk-reducing preventive therapies because they had been misclassified by conventional risk factors and assigned treatment goals not in line with their individual burden of atherosclerosis. Because there are many effective pharmacologic and nonpharmacologic therapies, the early detection of atherosclerosis itself before symptoms occur could provide a major opportunity to prevent many CV events.

Population-based screening tools such as the Framingham risk score (FRS) assign a risk probability, which is effective in assigning risk to populations but is not particularly helpful for individual risk assessment, especially in those at risk for near-term events. Several imaging modalities are currently available that can identify subclinical atherosclerosis long before CV events occur and have the potential to identify subjects at high risk in need of aggressive intervention and spare those at low risk from the unnecessary use of aggressive therapy. Unlike risk factors, anatomic measures of the intima-media thickness of the carotid artery wall (CIMT) and/or carotid plaque in adulthood and coronary artery calcification (CAC) measured by computed tomography, two of the most commonly used noninvasive techniques to detect subclinical atherosclerosis, represent the cumulative effects of an individual’s exposure over years to known and unknown (proatherogenic as well as atheroprotective) risk modifiers. The prognostic value of subclinical atherosclerosis detection has been shown to be greater than conventional risk factors such as low-density lipoprotein, high-density lipoprotein cholesterol, total cholesterol, systolic and diastolic blood pressure, diabetes, fasting glucose and insulin, reduced insulin sensitivity, active and passive smoking, and body mass index (BMI), and triglycerides. The purpose of this study was to determine comparative prevalence of carotid and coronary atherosclerosis and their respective utility in changing FRS in an asymptomatic adult population.


Patient Selection

This study included consecutive asymptomatic individuals aged >18 and <90 years who underwent both CAC and CIMT evaluation at our institution between October 2002 and October 2006. The mean time interval between CAC and CIMT evaluation was 1.1 ± 1.6 years. Those whose image quality for CIMT (n = 4) was suboptimal were excluded from the study. Those with histories of CV events in the coronary, peripheral, or cerebral vascular territory were excluded. The study protocol was approved by the institutional review board of Cedars Sinai Medical Center.

Computed Tomographic Coronary Calcium Score

Scanning was performed using either electron-beam tomography or multislice computed tomography, with acquisitions consisting of approximately 30 to 40 3-mm or 2.5-mm slices for the two tomographic systems, respectively. Foci of CAC were identified and scored by an experienced technician, blinded to patient characteristics, using semiautomatic commercial software on a Netra MD workstation (ScImage, Los Altos, CA) by the detection of ≥3 contiguous pixels (voxel size, 1.03 mm 3 ) of peak density >130 Hounsfield units within a coronary artery, with scoring verified by an experienced imaging cardiologist. CAC scores were calculated according to the method of Agatston et al, and age-adjusted and sex-adjusted CAC scores were determined according to the database of Raggi et al.


Carotid artery imaging was obtained using an ATL 5000 ultrasound system (Philips Medical Systems, Bothell, WA) with an 8-MHz to 15-MHz linear-array transducer. A depth of 4 cm was used and kept constant throughout study. Cine loops were acquired of the common carotid artery, bulb, and internal carotid artery on both the right and left sides at 4 different angles (180°, 150°, 120°, and 90° on the right and 180°, 210°, 240°, and 270° on the left). An additional 60° angle on the right side and 300° angle on the left side were used if CIMT was measurable in only 1 of the 4 angles on each side or if maximum disease was found at these very posterior angles. The near and far walls of the common carotid artery were acquired in the same image, whereas separate images were acquired each to show the near and far walls of the bulb and internal carotid artery at each angle on both sides. CIMT measurements were made offline by freezing frames at the R wave of the electrocardiogram and ensuring clear visualization of the boundaries between the intimal layer and carotid artery lumen and between the medial adventitial layer. Multiple measurements were made by the physician (T.Z.N.) and sonographer (M.G.) of the near and far walls of the distal 1 cm of the common carotid artery, the near and far walls of the carotid bulb, and the near and far walls of the proximal 1 cm of the internal carotid artery in each of the individual angles. A manual tracing was performed between the intimal layer and carotid artery lumen and between the medial adventitial layer, and mean, minimum, and maximum values were derived for each measurement. Any amount of CIMT thickness was included in the measurement. Offline analysis was performed on a Camtronics (Bothell, WA) workstation equipped with vascular analysis software. A mean value of CIMT was calculated on the basis of average of all the measurements (≥18 total measurements on each side) and was used for CIMT analysis. Using data from the Atherosclerosis Risk in Communities (ARIC) study, the composite CIMT from the far wall of all segments of the left and right carotid arteries was used to calculate an age-matched, sex-matched, and race-matched. In 15 randomly selected images used for CIMT measurement, intraobserver variability was 2.3 ± 4%, and interobserver variability was 3.2 ± 2%.

We defined plaque as a focal structure that encroached into the arterial lumen, demonstrating a thickness of ≥ 1.5 mm as measured from the media-adventitia interface to the intima-lumen interface and was 50% greater than the surrounding CIMT value. Bilateral composite mean CIMT only included plaques at the far wall of the common and internal carotid arteries and bulb (as done in the ARIC protocol), but separate evaluation of plaque presence ≥ 1.5 mm also included plaque in the near walls of the common and internal carotid arteries and bulb bilaterally. Thus, plaque assessment incorporated additional near wall segments to those assessed by CIMT. Investigators performing CIMT and plaque assessment were blinded to the CAC results.

Vascular Age

The CIMT and CAC values were used to determine “vascular age” and “coronary calcium” age, respectively, with established nomograms.


FRS and other variables were obtained through clinical and laboratory data review at the time of CAC and CIMT evaluation to evaluate the 10-year risk for coronary heart disease.

Statistical Methods

Mean composite far wall CIMT was converted on the basis of the ARIC study nomogram. Abnormal CIMT was defined as ≥75th percentile for age, race, and gender or by the presence of plaque ≥ 1.5 mm. Two cutoff values for abnormal CAC were defined, an absolute score > 0 as well as coronary calcium score > 75th percentile. Although CAC score > 0 indicates the presence of atherosclerosis, CIMT increases progressively with age, and hence no single cutoff value defines abnormal CIMT. We therefore used the most widely reported method denoting abnormal CIMT, that of ≥75th percentile for age, race, and gender. BMI was categorized as 20 to <25, ≥25 to <30, and ≥30 kg/m 2 . Obesity was defined as a BMI ≥ 30 kg/m 2 . Spearman’s ρ correlation of FRS was performed with CIMT and CAC score = 0 versus CAC score > 0 in the entire group as well as in subsets with low, intermediate, and high FRS defined as FRS 0 to <10, 10 to 20, and >20. One-way analysis of variance and Pearson’s χ 2 tests were performed to identify potential predictors of abnormal CAC or CIMT. Univariate predictors were then entered into a logistic regression model. Multiple models were tested adjusting for FRS, the presence or absence of a family history of coronary disease, the use of statins, BMI, and age. Odds ratios (ORs) and 95% confidence intervals (CIs) between predictor variables and outcome variables were derived. Vascular and coronary calcium ages were substituted for chronologic age to obtain CIMT-derived vascular FRS and CAC-derived FRS. Student’s t test and Spearman’s correlation were performed to compare and correlate differences in calculated ages and risk. Data are summarized as mean ± SD or frequencies and percentages.


Baseline Characteristics

Two hundred seventy-seven consecutive patients referred for CIMT evaluation at our institution were initially screened. Of these, 136 had no known CV disease and also underwent CAC assessment before or after the CIMT study. The mean age was 56.7 ± 10.9 years, and 57% were men. Ninety-two percent were Caucasian, 4% were black, 3% were Hispanic, and 1% were Asian. Baseline characteristics of the study population are listed in Table 1 . The mean FRS of the study population was 5.9 ± 6 (range, 0.1-29) and the mean CAC score was 152 ± 549 (range, 0-5205). The mean far wall composite CIMT was 0.821 ± 0.206 mm (range, 0.48-1.52 mm), and the mean CIMT percentile was 53 ± 21 (range, 10-95). There were 60 patients with far wall plaques and 30 with near wall plaques. There were 103 subjects with low FRS, 14 with intermediate FRS, and 19 with high FRS.

Table 1

Baseline characteristics of the study population (n = 136)

Variable Value
Age (y) 57.4 ± 11 (34-87)
LDL cholesterol (mg/dL) 109 ± 38 (43-219)
HDL cholesterol (mg/dL) 62 ± 20 (28-120)
Total cholesterol (mg/dL) 194 ± 46 (99-318)
Triglycerides (mg/dL) 109 ± 63 (35-398)
Men 78 (57%)
Hypertension 62 (46%)
Obese 27 (20%)
Smoker 12 (9%)
Prior tobacco use 44 (32%)
Family history 52 (38%)
On statins 59 (44%)

Data are expressed as mean ± SD or as number (percentage).

HDL , High-density lipoprotein; LDL , low-density lipoprotein.

Chronologic Age Versus Vascular Age

The mean vascular CIMT age (61.6 ± 11.4 years) and coronary calcium age (58.3 ± 11.1 years) were significantly different ( P = .001), and both were higher than mean chronologic age (57.4 ± 11 years) ( P < .0001 and P < .04, respectively). CIMT upgraded or downgraded FRS by >5% in more cases than CAC (42% of CIMT cases vs 17% of CAC cases). The correlation of FRS with mean far wall CIMT was 0.49 ( P < .0001), and the correlation of FRS with CAC was 0.47 ( P < .0001). Table 2 shows the effect of CIMT and CAC evaluation on the distribution of FRS category. As shown, CIMT upgraded more subjects to intermediate-risk and high-risk categories among the 103 low-FRS subjects. Both tests readjusted FRS in the intermediate-FRS and high-FRS groups. There was complete agreement between CIMT-adjusted FRS and CAC-adjusted FRS in 91 subjects (76 at low risk, 11 at intermediate risk, and 4 at high risk by CAC and CIMT).

Table 2

Effect of IMT and CAC on FRS

FRS<10% FRS 10-20 FRS>20% Total
FRS<10% 78 23 2 103
FRS 10-20% 3 7 4 14
FRS>20% 8 6 5 19
Total 89 36 11 136

FRS<10% FRS 10-20 FRS>20% Total
FRS<10% 88 14 1 103
FRS 10-20% 5 7 2 14
FRS>20% 9 6 4 19
Total 102 27 7 136

IMT FRS FRS<10% FRS 10-20 FRS>20% Total
FRS<10% 76 11 2 89
FRS 10-20% 24 11 1 36
FRS>20% 2 5 4 11
Total 102 27 7 136

FRS , Framingham Risk Score; IMT FRS , Vascular age-adjusted FRS; CAC FRS , CAC age-adjusted FRS.

FRS, Coronary Calcium Score, and Carotid Plaque in the Overall Population

Table 3 shows the prevalence of atherosclerosis in the study population. The mean FRS for those with carotid plaque versus those without plaque was 7.0 ± 7.0 versus 4.0±3.0 ( P < .01). The OR for FRS to predict carotid plaque was 1.13 (95% CI, 1.03-1.23; P = .01). Carotid plaque was present in 52%, 81%, and 89% in those with CAC scores of 0, 1 to 99, and >100, respectively ( P for trend = .002; Figure 1 ), but the prevalence of CIMT ≥ 75th percentile did not increase with progressively increasing CAC ( Figure 2 ). In multiple logistic regression analysis, adjusted for FRS, the odds of carotid plaque were greater for a CAC score of 1 to 99 (OR, 3.22; 95% CI, 1.14-9.11; P < .03) and CAC score > 100 (OR, 5.35; 95% CI, 1.37-20.8; P < .02) compared with a CAC score of 0. The areas under the receiver operating characteristic curve for predicting the presence of carotid plaque were 62% for FRS and 72% when CAC was added to FRS ( P < .02). BMIs were significantly higher in subjects who had atherosclerosis by both CAC score > 0 and CIMT ≥ 75th percentile (29.2 ± 5.1 kg/m 2 ) compared with those without atherosclerosis by both tests (25.6 ± 3.8 kg/m 2 ). Those with CAC scores > 0 only (26.1 ± 4.2 kg/m 2 ) or CIMT ≥75th percentile only (26.5 ± 5.6 kg/m 2 ) had intermediate BMIs ( P < .01 within groups). In the subgroup with BMIs < 25 kg/m 2 , there was a progressive decrease, and in subgroups with BMIs ≥ 30 kg/m 2 (obese subjects), there was a progressive increase in subclinical atherosclerosis detected by both CAC and CIMT ( Figure 3 ).

Table 3

Distribution of subclinical atherosclerosis in the study population

Variable n (%)
Plaque ≥ 1.5 mm 90 (66)
Plaque ≥ 1.5 mm (CAC score = 0) 26 (52)
CIMT ≥ 75th percentile 43 (32)
CIMT ≥ 75th percentile or plaque ≥ 1.5 mm 97 (71)
CAC and IMT< 75th percentile 78
CAC ≥ 75th percentile only 15
Both CAC&CIMT>75th percentile 11
CAC score = 0 71 (52)
CAC score = 1-99 33 (24)
CAC score > 100 32 (24)

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jun 16, 2018 | Posted by in CARDIOLOGY | Comments Off on High Prevalence of Ultrasound Detected Carotid Atherosclerosis in Subjects with Low Framingham Risk Score: Potential Implications for Screening for Subclinical Atherosclerosis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access