There is paucity of research looking at variations in carotid artery intima–media thickness (CIMT) during the cardiac cycle in children. The aim of this study was to ascertain variations, if any, in CIMT during the cardiac cycle in a population of high-risk children.
Forty-nine children aged 6 to 19 years with dyslipidemia and other atherosclerosis-promoting risk factors underwent a carotid ultrasound. CIMT was measured using commercially available, semiautomated edge-detection software. The region of interest was the far wall of the common carotid artery. CIMT was measured at various points during the cardiac cycle using the electrocardiogram (EKG) as a reference. CIMT measurements two frames before, during, and after the QRS complex (end diastole) were analyzed separately (designated as “QRS CIMT”) from the other CIMT measurements (designated as “non-QRS CIMT”). Demographics, heart rate, blood pressure, anthropometric measures, lumen diameter, family history, and presence of other atherosclerosis-promoting risk factors were documented.
“QRS CIMT” was significantly thicker than “non-QRS CIMT” ( P = .01), with the age group 10 to 14 years showing the most significant variation between “QRS CIMT” and “non-QRS CIMT” ( P = .005). CIMT values between right and left carotid arteries differed by 2.5%. Age, systolic blood pressure, and blood glucose were significant predictors of mean CIMT by simple linear regression; systolic blood pressure was the only significant predictor of mean CIMT by stepwise multiple linear regression analysis.
CIMT measurements vary during the cardiac cycle in children. It is thicker during the QRS complex on EKG. Carotid ultrasound should be performed with an EKG, and CIMT should be measured at the same point on the EKG to overcome this variation. Furthermore, we recommend that CIMT be measured at the R-wave on EKG because this is an easily discernible point in the cardiac cycle.
Several risk factors, such as dyslipidemia, hypertension, diabetes, and increased body mass index (BMI), can accelerate atherosclerosis and cause early cardiovascular disease in youth. Carotid artery intima–media thickness (CIMT) is a marker of atherosclerosis, and progression in carotid atherosclerosis correlates with progression of coronary atherosclerosis. In children, epidemiologic studies have noted the association of atherosclerosis-promoting risk factors with premature thickening of CIMT as they become young adults. Furthermore, several clinical and case-control studies have shown risk factors in children to be associated with increased CIMT, including familial hypercholesterolemia, hypertension, diabetes, and increased BMI. Because children rarely develop clinical cardiovascular disease, evaluating subclinical markers of atherosclerosis such as CIMT can be useful in further risk stratification and management, especially when multiple risk factors with poorly understood interplay coexist as is seen in obese children. Thus, CIMT may add additional information to risk assessment beyond traditional risk factor measures alone and aid in the management of these high-risk children.
Using carotid artery ultrasound to estimate CIMT has several advantages, including being noninvasive, easily repeatable, relatively inexpensive, and requiring no radiation. A scanning reading protocol that is clinically reproducible and valid will have utility in an outpatient clinic setting to further stratify high-risk children on the basis of the atherosclerotic burden they may have. Such abbreviated CIMT scanning protocols used to facilitate clinical screening have proven to be as valid in adults as more extensive research-based protocols that measured several segments of the carotid arteries. The common carotid artery is close to the skin, relatively straight, and well imaged, allowing CIMT to be measured easily. Thus, most studies in children have focused on common carotid artery far wall CIMT.
Many standard ultrasound machines are equipped with high-resolution probes that enable B-mode scanning of the carotid arteries. Edge-detection software enables efficient reading of CIMT in multiple frames of digitally acquired clips, thereby enhancing accuracy and decreasing bias. Recent studies have shown that measurement of CIMT using edge-detection software is accurate and more easily reproduced when compared with manual reading. Semiautomated edge detection software is also easy to use and can be used in an outpatient clinic setting. These not only are less time-consuming and more efficient in reporting CIMT because of their semiautomated nature but also can be used by less-experienced readers. A recent study compared a novice reader with a research laboratory using semiautomated edge-detection software and found comparable results with high reproducibility.
Several factors, including hemodynamic measures of heart rate, blood pressure, and elasticity of the vessel wall, may potentially have an impact on CIMT. A recent study in 50 men and 50 women aged 18 to 25 years using an edge-detection software looked at CIMT measurements of lumen diameter during R-wave electrocardiogram (EKG)-triggered frames. The investigators contrasted CIMT values obtained from the right and left carotid arteries in each patient. CIMT was higher in frames with a narrower lumen diameter, suggesting the possible effect of arterial size, compliance, and elasticity on CIMT. The investigators also noted a 2% to 3% higher CIMT in the right carotid artery and the R-wave EKG-triggered frames. Other published data in adult subjects have confirmed differences between CIMT measured from the common and internal carotid arteries, and between right and left carotid arteries.
Children have a more compliant vasculature, and changes in CIMT during the cardiac cycle may be more pronounced because of larger fluctuations in lumen diameter. Therefore, knowing when exactly in the cardiac cycle CIMT should be measured may be as important as uniform standards for acquisition and reading of carotid images. Thus, this study’s aim was to ascertain if the thickness of CIMT varies in children according to when the measurement was taken during the cardiac cycle. To our knowledge, none of the articles in the pediatric literature have examined differences in CIMT measurements during different phases of the cardiac cycle. Such differences may be clinically relevant, especially if follow-up readings are obtained to assess effectiveness of interventions.
Materials and Methods
This study was a retrospective chart review of 49 children aged 6 to 19 years who were referred to the Preventive Cardiology Clinic at our institution (University of Missouri-Kansas City School of Medicine) for evaluation of familial dyslipidemia, obesity and dyslipidemia, or family history of premature coronary artery disease. Family history, history of exposure to tobacco smoke, and demographic and anthropometric data (including age, gender, race, height, weight, and BMI) were collected. In addition, blood pressure, fasting lipid profile, glucose, and insulin were obtained. Approval from the institutional review board was received before data collection.
Carotid artery ultrasound for estimation of CIMT had been offered as part of atherosclerosis risk assessment to all these children. The carotid arteries were imaged using a standard ultrasound machine (Philips iE 33, Bothell, WA) equipped with a high-resolution L9-3 MHz linear array transducer. The carotid imaging was performed by 2 trained sonographers who followed a standardized institutional scanning protocol that had been established for pediatric imaging from modifications of a previously published protocol ( Appendix 1 ). The common carotid artery intima–media was the focus of interest, and the sonographers aimed to acquire images where both the far and near wall intimae were clearly delineated. The images were stored digitally as clips for offline reading.
The far wall of the common carotid artery was used for measurement of CIMT ( Figure 1 ). The digital clips acquired from both the left and right common carotid arteries were analyzed offline by the primary author using commercially available, semiautomated edge-detection software (QLAB, Philips iE 33). We used this software to measure CIMT within a 10-mm–wide box (region of interest) that was placed along the far wall of the common carotid artery within 2 cm proximal to the carotid bifurcation. Each measurement was assigned a “success rate,” which was defined by the percentage of the intima–media wall within the region of interest that was able to be accurately measured. We only accepted measured frames that had a success rate of 90% or greater. Each child’s study consisted of measuring CIMT on approximately 100 such frames of both the right and left carotid arteries.
CIMT measurements obtained starting 2 frames before the QRS complex through 2 frames after the QRS complex were separated from the other CIMT measurements and designated as “QRS CIMT” ( Figure 2 ). All other CIMT measurements were designated as “non-QRS CIMT” ( Figure 3 ). Measurements were tabulated in separate Excel (Microsoft Corp, Redmond, WA) spreadsheets as “QRS CIMT” and “non-QRS CIMT.” The grand mean “QRS CIMT” for each child was calculated by averaging all their “QRS CIMTs” of the right and left carotid arteries. The grand mean “non-QRS CIMT” for each child was calculated by averaging all their “non-QRS CIMTs” of the right and left carotid arteries. A paired t test was performed to compare the averaged grand mean “QRS CIMT” with the averaged grand mean “non-QRS CIMT” for all 49 children. Lumen diameters were measured at “QRS” and reported as the average of 2 measures from the right common carotid artery and 2 measures from the left common carotid artery. Heart rates were determined by averaging the heart rates documented at the time of these lumen diameter measurements.
CIMT was correlated to age, heart rate, lumen diameter, BMI, systolic blood pressure (SBP), tobacco smoke exposure, family history, total cholesterol, high-density lipoprotein cholesterol, triglycerides, insulin, and glucose using both simple and multiple linear regression using SAS version 9.1 (SAS Institute, Inc., Cary, NC). Differences in variables such as heart rate, lumen diameter, BMI, “QRS,” and “non-QRS CIMT” between age groups (6-9 years, 10-14 years, 15-19 years) were compared using analysis of variance and Tukey’s post hoc test.
The study population included 49 children with a history of obesity and atherosclerosis-promoting risk factors or history of familial dyslipidemia. Chronologic age was 13.2 ± 3.4 years. Thirty-seven participants (76%) were white. Twenty-seven participants (55%) were male. Mean body weight was 65.0 ± 24.4 kg, with 28 (57%) at > 95th percentile for age and gender. Mean BMI was 26.1 ± 6 kg/m 2 , with 30 (61%) at > 95th percentile for age and gender. Mean SBP was 119 ± 13 mm Hg, with 13 (27%) at > 95th percentile for gender, age, and height ( Table 1 ). Fasting lipid profile, insulin, and blood glucose data are shown in Table 1 . Of the 49 children, 40 (82%) had total cholesterol levels > 170 mg/dL, 28 (57%) had triglyceride levels > 120 mg/dL, 37 (76%) had low-density lipoprotein cholesterol levels > 110 mg/dL, and 28 (57%) had high-density lipoprotein levels < 45 mg/dL.
|n = 49||Mean ± SD|
|Age (y)||13.2 ± 3.4|
|Weight (kg)||65.0 ± 24.4|
|Height (cm)||155.1 ± 17.3|
|Body mass index (kg/m 2 )||26.1 ± 6.2|
|Systolic blood pressure (mm Hg)||119 ± 13|
|Diastolic blood pressure (mm Hg)||64 ± 8|
|Total cholesterol (mg/dL)||220 ± 53|
|Triglyceride (mg/dL)||149 ± 80|
|LDL (mg/dL)||148 ± 49|
|HDL (mg/dL)||45 ± 14|
|VLDL||30 ± 16|
|(n = 46)||88 ± 9|
|(n = 28)||12 ± 6|
No statistically significant deviations in CIMT readings were noted between the images acquired by the 2 sonographers independently on the same patient and at the same time. In addition, interobserver and intraobserver reliability of CIMT readings was obtained for a total of 106 frames (53 on the right common carotid artery and 53 on the left common carotid artery). A single reader read all these frames on 2 different occasions blinded to the other readings for the purpose of determining intraobserver reliability. Three readers read the same frames independently and were blinded to each other’s readings to establish interobserver reliability. The intraclass correlation was calculated for both cases using established guidelines where values greater than 0.75 indicate excellent reliability and values between 0.4 and 0.75 indicate good reliability. The intraobserver, intraclass correlation value was 0.81 (95% confidence interval, 0.74-0.87), indicating excellent reliability. The interobserver, intraclass correlation value was 0.67 (95% confidence interval, 0.57-0.76), indicating good reliability. However, the primary author read all the carotid scans for this study.
The overall mean lumen diameter was 5.2 ± 0.42 mm, and the overall mean heart rate was 74 ± 14 beats/min. To further account for age-related differences, the patient population was separated into 3 age groups (6-9 years, 10-14 years, and 15-19 years) ( Table 2 ). Significant differences were found between age group and mean lumen diameter (overall P = .002), and post hoc analyses found groups 6 to 9 years and 15 to 19 years to be significantly different ( P < .05) ( Figure 4 ). In addition, a significant age group difference was found for BMI between age groups 6 to 9 years and 15 to 19 years with a P value of .027, mean “QRS CIMT” between age groups 15 to 19 years and each of the other groups ( P = .004), and mean “non-QRS CIMT” between age groups 15 to 19 years and each of the other age groups ( P = .001). CIMT values between right and left carotid arteries differed by 2.5%. There were no age-related differences in mean heart rates.
|Age (y)||N||Mean ± SD|
|6-9||11||Mean lumen diameter (mm)||4.8 ± 0.38 a|
|11||Mean heart rate (beats/min)||78 ± 7|
|11||BMI||22.1 ± 5.3 b|
|10-14||22||Mean lumen diameter (mm)||5.1 ± 0.31|
|22||Mean heart rate (beats/min)||73 ± 14|
|22||BMI||26.4 ± 6.0|
|15-19||16||Mean lumen diameter (mm)||5.4 ± 0.43 a|
|15||Mean heart rate (beats/min)||73 ± 17|
|16||BMI||28.5 ± 5.9 b|