HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chung, H. Articles by Budoff, M. J. Search for Related Content PubMed PubMed Citation Articles by Chung, H. Articles by Budoff, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

Repeatability Limits for Measurement of Coronary Artery Calcified Plaque with Cardiac CT in the Multi-Ethnic Study of Atherosclerosis

¹ Department of Biostatistics, Collaborative Health Studies Coordinating Center, University of Washington, 6200 NE 74th St, Building 29, Suite 310, Seattle, WA 98115.
² Department of Radiology, Wake Forest University School of Medicine, Winston-Salem, NC.
³ Los Angeles Biomedical Research Institute at Harbor–UCLA, Torrance, CA.

Received June 14, 2007; accepted after revision September 21, 2007.

 
Supported by contracts N01-HC-95159 through N01-HC-95165 and N01-HC-95169 from the National Heart, Lung, and Blood Institute.

Address correspondence to R. L. McClelland (rmcclell{at}u.washington.edu).

WEB

This is a Web exclusive article.

Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
OBJECTIVE. The purposes of this study were to examine the repeatability of the findings of coronary artery calcification (CAC) measured with CT on repeated scans, to estimate 95% repeatability limits for CAC, and to use these limits to quantify detectable change in CAC over time.

SUBJECTS AND METHODS. The Multi-Ethnic Study of Atherosclerosis is a prospective cohort study with 6,814 participants 45–84 years old and free of clinical cardiovascular disease at enrollment. Agreement for presence of CAC was assessed for 6,742 participants who had baseline replicate scans on which a CAC score of 0 indicated no coronary calcification. Among 3,380 participants with baseline CAC, the 95% repeatability limits were established with a quantile regression model. Detectable change in CAC during follow-up was defined by an increase or decrease beyond the baseline repeatability limit.

RESULTS. At baseline, 274 (4.1%) of the rescan pairs were discordant (presence or absence of CAC). Greater body mass index was associated with a discordant pair (trend, p < 0.05). The upper 95% repeatability limits were (0.17 x Agatston score) + (4.89 x Agatston score) + (0.44 x body mass index)–10.84 for Agatston score and (0.16 x volumetric calcium score) + (4.30 x volumetric calcium score) + (0.23 x body mass index)–5.00 for volumetric calcium score. Rescan repeatability was comparable for electron beam and 4-MDCT scanners. At 2.5 years of average follow-up (range, 0.9–5.0 years), a detectable increase in Agatston and volumetric calcium scores was observed in 1,027 (36.3%) and 1,020 (36.0%), respectively, of 2,832 participants with baseline CAC.

CONCLUSION. The repeatability limits derived can be used to evaluate whether an increase in CAC score exceeds that expected from measurement error alone.

Keywords: cardiac CT • coronary artery calcification • detectable change • repeated scan • repeatability • repeatability limit

Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
CT has been found to be a noninvasive radiologic technique for acquiring images of the coronary arteries and for assessing coronary artery calcification (CAC) [1–6]. Two types of cardiac CT are used for measuring CAC: electron beam CT (EBCT) and MDCT. EBCT was introduced in the late 1980s and has been extensively studied and used. MDCT was introduced in the late 1990s and has rapidly evolved to be the dominant CT system for measurement of CAC and for CT angiography. Despite the increasing role of cardiac CT in medical practice and ongoing improvement in CT technology, concerns remain about the clinical utility of cardiac CT beyond traditional cardiovascular risk factors and about the variability of CT measurements of calcified plaque on repeated scans [7]. For variability on repeated scans, the main challenges are continuous heart motion, varying plaque size, large patient size, and intraobserver and interobserver variability in calcium scoring. The degree of agreement between measurements obtained in quick succession, called repeatability [8], is important not only in comparisons of scanner types and CAC scoring systems but also in the evaluation of change in CAC over time. Our goal was to quantify rescan repeatability with a repeatability limit. For the repeatability limit, we used the International Organization for Standardization definition, which is described by Sevrukov et al. [9] as "the value less than or equal to which the absolute difference between two test results obtained under repeatability conditions may be expected to be with a probability of 95%."

The Multi-Ethnic Study of Atherosclerosis [10] is a prospective cohort study with 6,814 U.S. participants free of clinical cardiovascular disease. In the study, two sequential CT scans were acquired at enrollment with EBCT or 4-MDCT for nearly all participants (between years 2000 and 2002) [11]. The first follow-up CT examination was performed 1–5 years after the first. The Multi-Ethnic Study of Atherosclerosis affords an excellent opportunity to compare rescan variability across CT type and participant characteristics (body size, sex, race) and to explore temporal change in CAC exceeding that expected on the basis of baseline rescan variability. The objectives of our study were to construct 95% repeatability limits of CAC scores and to use the repeatability limits to examine detectable change in CAC over time in the Multi-Ethnic Study of Atherosclerosis.

Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Study Participants
In the Multi-Ethnic Study of Atherosclerosis, 3,213 men and 3,601 women 45–84 years old (mean age, 62.2 years) were recruited from six U.S. communities from 2000 to 2002. The communities were Winston-Salem, NC; New York, NY; Baltimore, MD; Minneapolis, MN; Chicago, IL; and Los Angeles, CA. Completion of a CT examination at baseline was one of the inclusion criteria. Exclusion criteria included severe obesity (weight > 300 pounds [136 kg]) and pregnancy. The study population was 38.5% white (n = 2,624), 11.8% Chinese American (n = 803), 27.8% black (n = 1,894), and 21.9% Hispanic (n = 1,493). A detailed description of the study has been published [10]. The Multi-Ethnic Study of Atherosclerosis was approved by the institutional review boards of all participating study sites. Each participant gave informed consent.

Multi-Ethnic Study of Atherosclerosis CT Examinations
A standardized Multi-Ethnic Study of Atherosclerosis CT and reading protocol has been documented [11, 12]. Briefly, participants were instructed about breath-holding and about the importance of immobility during scanning. All scans were acquired in a single breath-hold. Participants were asked to be relaxed on the table between scans. The protocol was designed to require less than 15 minutes in the CT room, and the average time between scans at baseline was 2 minutes. At the Multi-Ethnic Study of Atherosclerosis baseline, EBCT (Imatron C150 scanner, GE Healthcare) was used at three study sites, and 4-MDCT (LightSpeed scanner, GE Healthcare, or Volume Zoom scanner, Siemens Medical Solutions) at three sites. The field of view of CT images was 350 mm. The detailed CT acquisition has been described by Carr et al. [11]. A total of 6,742 participants had a pair of CT scans with valid Agatston and volumetric calcium scores at baseline, including 3,380 participants with CAC (i.e., a positive CAC score).

Follow-up CAC measurements were performed on one half of the cohort (randomly selected) at a second Multi-Ethnic Study of Atherosclerosis examination in years 2002–2004 and the other half of the cohort at a third examination in years 2004–2005, an average of 1.6 and 3.2 years, respectively, after the baseline examination. A total of 5,757 participants in the Multi-Ethnic Study of Atherosclerosis had a follow-up CAC score measured at either examination 2 or examination 3, and 5,733 of them underwent scanning twice consecutively. During the follow-up period, scanners were replaced with newer models at some study sites. At examinations 2 and 3, the following scanners were used for some of the study participants: LightSpeed Pro 16 (GE Healthcare) (n = 515), Sensation 16 (Siemens Medical Solutions) (n = 269), Sensation 64 (Siemens Medical Solutions) (n = 186), Aquilion 32 (Toshiba) (n = 39), and Aquilion 64 (Toshiba) (n = 130).

In our study, we focused on the Agatston score [13] and the volumetric calcium score [14]. Calibration phantoms with known calcium density were used in the Multi-Ethnic Study of Atherosclerosis to adjust for attenuation variability across study sites and participants [12]. In our study, however, we used raw CAC measurements without phantom calibration because the within-subject variability was of primary interest and because the Multi-Ethnic Study of Atherosclerosis phantom calibration algorithm is not used clinically. Our findings with raw CAC measurements would be easier to generalize to scans obtained in clinical practice.

Detectable Change in CAC over Time
For participants with CAC at baseline, we defined detectable change in CAC as a change exceeding the baseline repeatability limits. That is, a difference between the two CAC measurements exceeding that expected from measurement error alone. We examined detectable CAC change from baseline to follow-up using data on participants who had CAC at baseline and who had CAC follow-up data. At each time point (baseline or follow-up), a point estimate of the amount of calcium in each participant was presented as an average CAC score, mostly from dual-scan runs.

Statistical Analysis
Given two repeated scans (scan 1, scan 2), we used M to denote the average magnitude, or (scan 1 + scan 2)/2; D, the difference, (scan 1–scan 2); and |D|, the absolute value of the difference. Relative absolute difference is defined as the absolute difference divided by the average (A) with the result expressed as a percentage, or (|D|/A) x 100.

The distribution of CAC in asymptomatic participants was highly skewed, with a considerable proportion of participants having no measurable calcified plaque according to the scoring criteria (CAC score, 0). Among the Multi-Ethnic Study of Atherosclerosis participants, approximately 50% of participants had no CAC at baseline. Therefore, as in previous reports of CT in the Multi-Ethnic Study of Atherosclerosis [11, 15], we reported the proportions of discordant rescan pairs (presence or absence of CAC) using logistic regression models to examine whether this disagreement was associated with CT scanner type or participant characteristics. Body mass index (BMI) (weight in kilograms divided by height squared in meters) was used as a measure of body size. When appropriate, we further examined other indicators of body size, including weight, waist circumference, waist-to-hip ratio, and body surface area.

In participants with CAC at baseline, we first generated a Bland-Altman plot [8] for each calcium measurement and found a nonuniform variability of CAC measurements. We estimated the 95th percentile for the absolute difference (|D|) as a function of the average magnitude (M) and other covariates (e.g., BMI) by fitting a quantile regression model [16]. Results of previous work [9, 17] suggested that the rescan repeatability of CAC measurements is a function of square-root-transformed magnitude (M), so we included this value as a term in our quantile regression model. A common way [9, 18] of estimating CAC repeatability limits is to use the half-normal method, in which it is assumed that properly scaled differences follow a normal distribution. The quantile regression method we used has the advantage of not requiring a distributional assumption.

Comparisons of proportions and tests for interaction were conducted with logistic regression (Wald test). Tests of the linear trend across BMI categories were performed by treating the variable as continuous. A value of p 0.05 was considered significant. All analyses were performed with Stata version 9.2 software (Stata).

Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Agreement for Presence of CAC at Baseline
Among the 6,742 pairs of scans, 274 (4.1%) of the pairs were discordant. Table 1 shows that the proportions of discordant rescan pairs did not vary significantly by scanner type (4-MDCT vs EBCT), sex, or race. In contrast, greater BMI was associated with more discordant pairs. The body size effect seemed not to differ by sex or scanner (interaction with BMI, p = 0.26 and 0.40, respectively). With regard to other body size measures, waist-to-hip ratio was found significantly related to rescan discordance (p = 0.02). The possibility that body size effect might have been driven mostly by very obese participants was explored. When participants with a BMI of 40 or greater were excluded, p = 0.06 for linear trend across normal mass (BMI < 25), overweight (BMI, 25–29.9), and moderate obesity (BMI, 30–39.9).

View larger version (14K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1 —Bland-Altman scatterplot shows rescan difference versus rescan average of Agatston score in Multi-Ethnic Study of Atherosclerosis at examination 1.

View larger version (15K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2 —Scatterplot shows rescan relative difference (%) versus rescan average of Agatston score in Multi-Ethnic Study of Atherosclerosis at examination 1.

To estimate the repeatability limits for Agatston score, we initially modeled the repeatability limit of rescan absolute difference (|D|) using a square-root-transformed rescan average (M) based on the nonlinear relation between |D| and M and on previous results [9, 17]. There remained, however, a significant residual dependence between the resulting repeatability limit and CAC magnitude (p = 0.01). Thus we included in the quantile regression model a second order of M, which is identical to M; that is, (M)² = M. The resulting repeatability limit (Agatston score) was estimated as:

A total of 168 (5.0%) of the observations were outside this repeatability limit, and this proportion was not associated with CAC magnitude (p = 0.88). We examined the associations of scanner type and participant characteristics with this repeatability limit (Table 2) and found a significant association with BMI. We therefore added an adjustment term for BMI in the quantile regression model. The resulting repeatability limit (Agatston score) was expressed as follows:

With this model, the proportion of observations falling outside the repeatability limits was not associated with CAC magnitude (p = 0.89), BMI, or the other factors listed in Table 2.

Figure 3 compares the CAC-adjusted and CAC- and BMI-adjusted repeatability limits for Agatston score. The plot shows that these two repeatability limits agreed well with each other. Given the ranges of Agatston scores (from zero to several thousand) and BMI (generally between 15 and 60), the effect of BMI on repeatability is minimal. However, for persons with very small amounts of CAC (e.g., < 30 Agatston units), the effect of BMI correction is more noticeable.

View larger version (12K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3 —Graph shows coronary artery calcification (CAC)-adjusted and CAC- and body mass index (BMI)–adjusted repeatability limits for Agatston score less than 400 at Multi-Ethnic Study of Atherosclerosis examination 1. Solid line represents CAC-adjusted repeatability limits; dashed line, CAC- and BMI-adjusted repeatability limits for BMI of 25; dotted line, CAC- and BMI-adjusted repeatability limit for BMI of 40.

The proportions outside these repeatability limits were not associated with CAC magnitude (p = 0.77 and p = 0.73, respectively). Table 2 shows the effects of scanner type and participant characteristics on these repeatability limits. The rescan differences in volumetric calcium score appeared to be less affected by BMI than they were in the Agatston score.

Table 3 shows the CAC-adjusted repeatability limits, or detectable change, for Agatston and volumetric calcium scores. For example, for a participant with a baseline score of 50 Agatston units, a change of 45 units or more is unlikely to be due to measurement error. In contrast, for a baseline Agatston score of 300, changes of more than 139 units are unlikely to be due to measurement error alone.

View this table: [in this window] [in a new window]   TABLE 3: Detectable Change () for Agatston and Volumetric Calcium Scores

View larger version (10K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4 —Plot shows percentage of follow-up Agatston score beyond baseline coronary artery calcification (CAC)–adjusted repeatability limit for 2,832 Multi-Ethnic Study of Atherosclerosis participants with positive CAC scores at baseline. Samples are grouped into 3-month intervals. Point shading and size reflect group size. Black indicates n 100; range, 129–483. Gray indicates n < 100; range, 3–43. Larger points indicate greater number of participants. Dashed line represents overall proportion (36.4%) of detectable change.

Applying the CAC-adjusted repeatability limits to the observed changes in 2,832 participants with positive baseline CAC scores, we found that 1,031 (36.4%) and 1,027 (36.3%) of the participants had a detectable CAC increase for Agatston and volumetric calcium scores, respectively. As expected, the proportion with detectable change increased with time since baseline (Fig. 4). When the CAC- and BMI-adjusted repeatability limits were applied, 1,007 and 1,020 participants had a detectable CAC increase for Agatston and volumetric calcium scores, respectively. A total of 913 participants had a detectable CAC increase with respect to all four repeatability limits considered.

Only 11 of the 313 participants with a negative change in CAC score had a detectable decrease in CAC with respect to at least one repeatability limit. Four of the 11 participants had a moderate amount of CAC at baseline (Agatston score > 40; volumetric calcium score > 30) but had no CAC at follow-up. In particular, one participant had coronary revascularization between baseline and follow-up CT scans (baseline Agatston score, 294; follow-up Agatston score, 138).

Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
We examined rescan repeatability with regard to agreement on presence of CAC and the 95% repeatability limits for the Agatston and volumetric calcium scores. We established the Multi-Ethnic Study of Atherosclerosis baseline repeatability limits as a function of amount of CAC and, optionally, BMI and illustrated how the repeatability limits can be used to evaluate detectable CAC change over time.

We found that greater BMI is associated with a measurable decline in rescan repeatability (disagreement for presence or absence, wide repeatability limits). However, this body size effect would be practically meaningful only for persons with low CAC scores. In other words, early detection of true calcification with CT and current CAC scoring methods is more difficult for extremely overweight people (e.g., BMI > 40). Similar findings were obtained in CT phantom data analysis [12, 19, 20]. Possible explanations include soft-tissue attenuation [19], background image noise [21], and body tissue composition.

The baseline repeatability limits were used to quantify detectable change in CAC, and this approach has several advantages. The proportion of scans with detectable change increases with length of time between scans, as expected. In addition, the number of participants with detectable regression is very small (11 of 313 participants with a negative change), and in many of those cases the regression appears not to be due to measurement error.

Although we present results for raw CAC scores, the rescan variability of the phantom-adjusted CAC scores was almost identical to that of raw CAC scores. Calibration with the phantom did not entirely eliminate rescan variability due to obesity; however, this BMI effect was minimal compared with the effect of CAC magnitude.

We found no difference in rescan variability for the scanner types (EBCT and 4-MDCT) used in the Multi-Ethnic Study of Atherosclerosis baseline CT examinations between 2000 and 2002. We also found that 3–5% of the rescan difference with new MDCT scanners used at follow-up (examinations 2 and 3) was beyond our 95% repeatability limits, but the relatively small number of participants imaged with the new scanners must be considered. The CT data were obtained according to a standardized protocol for the Multi-Ethnic Study of Atherosclerosis, and we put our efforts into controlling for participant-specific covariates (most important, CAC magnitude). Nevertheless, ability to generalize the Multi-Ethnic Study of Atherosclerosis baseline 95% repeatability limits to different scanners from different vendors is limited, and further study is warranted.

A CAC score captures only part of the available pathobiologic information on calcified plaque available from cardiac CT images. Global CAC scores of the entire coronary circulation can mask or minimize substantial specific changes in vessels and plaque. In addition, different positions on the scanner table at different times can influence CT images and subsequent CAC scores. Thus an increase in CAC score over time does not necessarily imply clinical atherosclerotic progression and subsequent increased risk of cardiovascular events. The prognostic utility of tracking change in coronary calcified plaque remains unclear with limited evidence [6]. In evaluation of the prognostic utility of tracking calcification, understanding of and possible adjustment for within-person variability will likely be useful. The use of baseline rescan repeatability information may be one way to increase the sensitivity and specificity of change in calcification.

We emphasize the importance of controlling for CAC magnitude in repeatability analyses. The distribution of CAC differs substantially with age, sex, and race and ethnicity [22]. Unless the ranges of CAC magnitude are comparable, simple comparison of either absolute difference or relative difference will not be useful in comparisons of rescan variability.

Acknowledgments
 
We thank the investigators, staff, and participants of the Multi-Ethnic Study of Atherosclerosis. A full list of participating Multi-Ethnic Study of Atherosclerosis investigators and institutions is available at www.mesa-nhlbi.org.

References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. O'Rourke RA, Brundage BH, Froelicher VF, et al. American College of Cardiology/American Heart Association Expert Consensus document on electron-beam computed tomography for the diagnosis and prognosis of coronary artery disease. Circulation 2000;102 : 126–140[Free Full Text]
2. Schoenhagen P, Halliburton SS, Stillman AE, et al. Noninvasive imaging of coronary arteries: current and future role of multi-detector row CT. Radiology 2004;232 : 7–17[Abstract/Free Full Text]
3. Schoepf UJ, Becker CR, Ohnesorge BM, Yucel EK. CT of coronary artery disease. Radiology 2004;232 : 18–37[Abstract/Free Full Text]
4. Clouse ME. How useful is computed tomography for screening for coronary artery disease? Noninvasive screening for coronary artery disease with computed tomography is useful. Circulation2006; 113:125 –146[Free Full Text]
5. Raggi P. Noninvasive imaging of atherosclerosis among asymptomatic individuals. Arch Intern Med 2006;166 :1068 –1071[Free Full Text]
6. Budoff MJ, Achenbach S, Blumenthal RS, et al. Assessment of coronary artery disease by cardiac computed tomography: a scientific statement from the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology. Circulation 2006;114 :1761 –1791[Free Full Text]
7. Chen J, Krumholz HM. How useful is computed tomography for screening for coronary artery disease? Screening for coronary artery disease with electron-beam computed tomography is not useful. Circulation 2006;113 : 125–146[Free Full Text]
8. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet1986; 1:307 –310[CrossRef][Medline]
9. Sevrukov AB, Bland JM, Kondos GT. Serial electron beam CT measurements of coronary artery calcium: has your patient's calcium score actually changed? AJR 2005;185 :1546 –1553[Abstract/Free Full Text]
10. Bild DE, Bluemke DA, Burke GL, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol2002; 156:871 –881[Abstract/Free Full Text]
11. Carr JJ, Nelson JC, Wong ND, et al. Calcified coronary artery plaque measurement with cardiac CT in population-based studies: standardized protocol of Multi-Ethnic Study of Atherosclerosis (MESA) and Coronary Artery Risk Development in Young Adults (CARDIA) study. Radiology 2005;234 : 35–43[Abstract/Free Full Text]
12. Nelson JC, Kronmal RA, Carr JJ, et al. Measuring coronary calcium on CT images adjusted for attenuation differences. Radiology 2005;235 : 403–414[Abstract/Free Full Text]
13. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15 : 827–832[Abstract]
14. Callister TQ, Cooil B, Raya SP, Lippolis NJ, Russo DJ, Raggi P. Coronary artery disease: improved reproducibility of calcium scoring with an electron-beam CT volumetric method. Radiology1998; 208:807 –814[Abstract/Free Full Text]
15. Detrano RC, Anderson M, Nelson J, et al. Coronary calcium measurements: effect of CT scanner type and calcium measure on rescan reproducibility: MESA study. Radiology2005; 236:477 –484[Abstract/Free Full Text]
16. Koenker R, Bassett G Jr. Regression quantiles. Econometrica 1978;46 : 33–50[CrossRef]
17. Hokanson JE, MacKenzie T, Kinney G, et al. Evaluating changes in coronary artery calcium: an analytic method that accounts for interscan variability. AJR 2004;182 :1327 –1332[Abstract/Free Full Text]
18. Bielak LF, Sheedy PF 2nd, Peyser PA. Coronary artery calcification measured at electron-beam CT: agreement in dual scan runs and change over time. Radiology 2001;218 : 224–229[Abstract/Free Full Text]
19. Raggi P, Callister TQ, Cooil B. Calcium scoring of the coronary artery by electron beam CT: how to apply an individual attenuation threshold. AJR 2002; 178:497 –502[Abstract/Free Full Text]
20. Stanford W, Burns TL, Thompson BH, Witt JD, Lauer RM, Mahoney LT. Influence of body size and section level on calcium phantom measurements at coronary artery calcium CT scanning. Radiology2004; 230:198 –205[Abstract/Free Full Text]
21. Sevrukov A, Pratap A, Doss C, Jelnin V, Hoff JA, Kondos GT. Electron beam tomography imaging of coronary calcium: the effect of body mass index on radiologic noise. J Comput Assist Tomogr2002; 26:592 –597[CrossRef][Medline]
22. McClelland RL, Chung H, Detrano R, Post W, Kronmal RA. Distribution of coronary artery calcium by race, gender, and age: results from the Multi-Ethnic Study of Atherosclerosis (MESA). Circulation 2006;113 : 30–37[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. J. Gibbons, P. A. Araoz, and E. E. Williamson The year in cardiac imaging. J. Am. Coll. Cardiol., January 6, 2009; 53(1): 54 - 70. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chung, H. Articles by Budoff, M. J. Search for Related Content PubMed PubMed Citation Articles by Chung, H. Articles by Budoff, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?