Age and Sex Distribution of Subclinical Aortic Atherosclerosis
A Magnetic Resonance Imaging Examination of the Framingham Heart Study
Autopsy data demonstrate a correlation between subclinical aortic atherosclerosis and cardiovascular disease. Therefore, noninvasive cardiovascular magnetic resonance (CMR) of subclinical atherosclerosis may provide a novel measure of cardiovascular risk, but it has not been applied to an asymptomatic population-based cohort to establish age- and sex-specific normative data. Participants in the Framingham Heart Study offspring cohort who were free of clinically apparent coronary disease were randomly sampled from strata of sex, quartiles of age, and quintiles of Framingham Coronary Risk Score. Subjects (n=318, aged 60±9 years, range 36 to 78 years, 51% women) underwent ECG-gated T2-weighted black-blood thoracoabdominal aortic CMR scanning. CMR evidence of aortic atherosclerosis was noted in 38% of the women and 41% of the men. Plaque prevalence and all measures of plaque burden increased with age group and were greater in the abdomen than in the thorax for both sexes and across all age groups. In addition, the Framingham Coronary Risk Score was significantly correlated with all plaque prevalence and burden measures for women but only for men after age adjustment. These noninvasive CMR data extend the prior autopsy-based prevalence estimates of subclinical atherosclerosis and may help to lay the foundation for future studies of risk stratification and treatment of affected individuals.
Atherosclerotic cardiovascular disease (CVD) remains the leading cause of morbidity and mortality in the Western world. Autopsy studies have demonstrated that aortic and coronary atherosclerosis have a long subclinical phase.1 Consequently, early diagnosis and treatment of atherosclerosis in the preclinical stage may reduce CVD sequelae such as myocardial infarction or stroke.
Clinical risk factors and summary measures such as the Framingham Coronary Risk Score (FCRS) can predict the probability of future CVD events.2 Although these models perform well on large populations, not all individuals with a heavy risk factor burden develop overt CVD, whereas other subjects with few risk factors develop symptomatic CVD. Common noninvasive methods (eg, exercise testing and nuclear imaging) for diagnosing atherosclerosis are dependent on the presence of flow-limiting lesions, but non-flow-limiting subclinical disease may predate overt disease by decades.3 In addition, these other methods are unable to directly image or quantify atherosclerotic burden.
Cardiovascular magnetic resonance (CMR) offers unique advantages over other imaging modalities, including the ability to quantify atherosclerotic plaque burden4– 6⇓⇓ and lack of ionizing radiation, and provides highly reproducible measures of aortic anatomy and atherosclerosis.7 However, CMR has not been applied to a longitudinally followed population-based cohort free of clinical heart disease to establish age- and sex-specific prevalence of subclinical atherosclerosis in the general population.
The study design of the Framingham Heart Study (FHS) has been detailed elsewhere.8 Subjects considered for this investigation were participants in the offspring cohort, which was initiated in 1971. Of the 3799 offspring participants in examination cycle 5 (1991 to 1995), those who were free of overt CVD were stratified by sex, quartiles of age, and quintiles of FCRS.2 The highest risk quintile was double-sampled. Because subjects were stratified by sex and age quartile, these indices were not used to calculate the FCRS for deriving the study sample. From the resulting strata, 331 participants were randomly selected and invited to participate in the present study, of whom 318 elected to participate (aged 60±9 years, 51% women). The present study was approved by the Institutional Review Boards at Boston University Medical Center and the Beth Israel Deaconess Medical Center. Written informed consent was obtained from all participants. All subjects were in sinus rhythm and without contraindication to CMR.
Cardiovascular Magnetic Resonance
Subjects underwent thoracoabdominal aortic CMR with the use of a commercial 1.5-T whole-body CMR system (Gyroscan ACS-NT, Philips Medical Systems) with a PowerTrak 6000 gradient system (peak gradient 23 mT/m, rise time 219 ms).
Twenty-four transverse slices spanning the aortic arch to the aortoiliac bifurcation were obtained by using a free-breathing, ECG-gated, T2-weighted turbo spin-echo sequence. Black-blood aortic images were obtained by using time delays of 75 ms (thoracic) and 125 ms (abdominal) after the R wave. Spectral presaturation of the lipid signal was used to minimize chemical shift artifact from periadventitial fat. Other imaging parameters included the following: repetition time 3 heart beats, echo time 45 ms, turbo spin-echo factor 14, field of view 264 mm×330 mm, matrix size 256×512 (in-plane spatial resolution of 1.03 mm×0.64 mm), and slice thickness 5 mm, with a 10-mm slice gap. Thoracic aortic images were obtained with a commercial 5-element cardiac phased-array receiver coil with 4 signal averages. Abdominal aortic images were obtained with the body coil as a receiver with 6 averages. Total aortic imaging time was <20 minutes.
Aortic and Atherosclerotic Plaque Analysis
CMR data were transferred to a commercial EasyVision Work Station (version 3.0, Philips Medical Systems) and analyzed by reviewers blinded to all clinical data. Images were analyzed if they (1) were contained within the descending thoracic aorta and above the origin of the common iliac arteries and (2) had sufficient signal to noise and contrast to noise to allow clear visual definition of >50% of the inner circumference of the aortic wall. All data from a subject were excluded if ≥50% of the slices did not meet these criteria.
Aortic and plaque measurements were classified as thoracic or abdominal according to their location above or below the diaphragm, respectively. With the use of an interactive drawing tool, the inner aortic perimeter (a bright medial layer on T2-weighted images,4 Figure 1) and cross-sectional area were recorded.7 For aortic contour tracing, the semiautomatic interactive tool used an elliptical framework that was elongated to intersect the inner circumferential border of the aorta. For the plaque contour tracing, an interactive free-hand manual drawing tool was used. Intraobserver variability of aortic and plaque planimetry in our laboratory has been shown to be excellent.7
Atherosclerotic plaque was defined as characteristic luminal protrusions4 of ≥1 mm in radial thickness7 that could be visually distinguished from the minimal residual blood signal. For each plaque, intimal perimeter, cross-sectional area, and maximal radial thickness were measured (Figure 1). Plaque perimeter was defined as the outer circumference of the plaque in contact with the inner border of the aorta (Figure 1), an index analogous to plaque surface area determined in autopsy studies.9–12⇓⇓⇓ Atherosclerotic lesions were further classified as type A (maximal radial thickness ≤2.5 mm) or type B (thickness >2.5 mm). This threshold was chosen to allow ≈4 pixels in the readout direction across a given plaque. Plaque prevalence (percentage of subjects with plaque), type and number of plaques, intimal plaque perimeter, and total cross-sectional area were assessed.
To quantify the magnitude of atherosclerosis in each subject, 3 measures of atherosclerotic plaque burden (slice plaque burden [SPB], perimeter plaque burden [PPB], and area plaque burden [APB]) were determined as follows: SPB (%)=100 · (number of aortic slices with plaque/number of total aortic slices); PPB (%)=100 · (Σ intimal plaque perimeter/Σ intimal aortic circumference); and APB (%)=100 · (Σ plaque cross-sectional area/Σ aortic cross-sectional area).
To account for double sampling of the highest risk group, all analyses used a weighting variable of 0.5 for subjects sampled from the highest risk category and 1.0 for subjects from other categories. Plaque prevalence is reported as a percentage, whereas plaque burden (SBP, PPB, and APB) is reported as mean and SD. Age-related trends were tested for plaque prevalence and plaque burden measures. Age-adjusted analyses were used when sex differences were assessed. Comparisons within subjects between abdominal and thoracic aortic plaque were made by using methods appropriate for paired data. Spearman correlations were calculated between FCRS and plaque prevalence and burden. All statistical analyses were conducted with the use of SAS/STAT (Release 6.12, SAS Institute Inc). A value of P<0.05 was considered significant.
Of the 318 FHS offspring subjects who presented for CMR, studies were completed in 312 (98%) subjects. Premature termination in 6 subjects was due to claustrophobia or inadequate ECG gating, and 14 subjects were excluded because of insufficient image quality. Therefore, 298 (94%) subjects were eligible for plaque analyses. From the average 19.1 aortic images per subject, 1.1 (5.7%) image per subject was excluded from plaque analysis because of an image artifact, such as thoracic respiratory artifact, residual blood signal, or insufficient signal-to-noise ratio.
We determined the interobserver variability of aortic and plaque planimetry by using the interactive drawing tool. Two trained observers (F.A.J., W.J.M.) independently analyzed 10 subjects for aortic and plaque cross-sectional areas. Interobserver variability was very high for aortic cross-sectional area (r=0.98) and plaque cross-sectional area (r= 0.94).
Aortic plaque of ≥1-mm radial thickness was identified in 38% of the women and 41% of the men (Figure 2). After adjustments were made for age, the prevalence of aortic plaque did not differ between men and women (P=0.38). In both sexes, there was an increasing trend in aortic atherosclerosis across age groups (P<0.01 for both sexes). For the entire group, aortic plaque was more prevalent in the abdomen than in the thorax (38.8% versus 6.3%, respectively; P<0.01). This difference was present for both sexes (both P<0.01, Figure 2a and Table 1). There was an increase in abdominal and thoracic aortic plaque across age groups in men (P=0.02 and P=0.01, respectively) and in women (both P<0.01). After adjustments were made for age, the prevalence of aortic atherosclerosis in the abdomen and in the thorax was similar (P=0.49 and P=0.22, respectively) among men and women.
All quantitative measures (SBP, PPB, and APB) of atherosclerosis burden (Table 2) increased with age group adjusted for sex (P<0.01) and were greater in the abdomen (versus thorax) in both sexes (Figure 2b and Table 3). The average FCRS for the population was 2.25±3.82. Age-adjusted partial correlations of the FCRS with plaque prevalence (by magnetic resonance imaging) and burden measures were significant for men and for combined men and women (Table 4).
To our knowledge, the present report represents the first population-based CMR study of aortic atherosclerosis burden. We found that 40% of asymptomatic FHS offspring subjects had evidence of subclinical aortic atherosclerosis. For men and women, atherosclerotic plaque was more prevalent in the abdominal aorta compared with the thoracic aorta. In both sexes, plaque prevalence and all quantitative CMR plaque burden measures (SBP, PPB, and APB) increased with age group. Last, CMR atherosclerotic plaque prevalence and burden measures were positively associated with the FCRS.
Previous population-based atherosclerotic imaging studies have been largely limited to carotid ultrasonography for intimal-medial thickness13–15⇓⇓ and calcium-based imaging techniques, such as chest x-ray of the thoracic aorta16 and electron-beam computed tomography of the epicardial coronary arteries.17–19⇓⇓ As with our CMR data, carotid intimal-medial thickness and electron-beam computed tomographic atherosclerotic measures increase with age.20,21⇓ At present, the optimal imaging method to assess for subclinical atherosclerosis has not been determined.22
In previous large autopsy studies of young (aged <35 years) trauma victims, aortic fibrous or complicated plaques (raised lesions) ranged from 0% to 12% by surface area and increased with age.9–12⇓⇓⇓ Consistent with our findings, the 2 studies that distinguished between thoracic and abdominal aortic plaque found substantially greater plaque in the abdominal aorta.10,12⇓ One study examined sex differences in aortic atherosclerosis and found that women in the highest age group (30 to 34 years) had greater surface involvement of raised lesions than did the corresponding men.12 Although not directly comparable because of age differences, our results are in accord with these autopsy data. Interestingly, all plaque burden measures in the oldest women were greater than those in the corresponding men, and the rate of increase of plaque burden was greater in women. These data may explain the postulated protective benefit of estrogen that lessens in older (postmenopausal) women.23
Autopsy data from a more analogous older population were reported as part of the International Atherosclerosis Project.24 Among white individuals from New Orleans and Oslo, advanced atherosclerotic lesions were less common in young women than in young men but were similar in older men and women. Although these data are consistent with our findings, raised atherosclerotic lesions were identified in a larger percentage of subjects than in the present report, suggesting a potential reduced sensitivity of this CMR method for minor lesions.
The present study was not designed to characterize atherosclerotic plaque components. Many CMR-detected aortic plaques demonstrated a dark central core underneath a bright rim of tissue, consistent (but not diagnostic) of a lipid pool or calcium underlying a fibrous cap (Figure 1). Although plaque characterization and detailed vessel wall measurements may be desirable, 6 this would require additional CMR imaging using T1-weighted and proton density-weighted sequences.5 This was not possible because of external time constraints. Also, given the time constraints, we chose to obtain 24 slices with a 10-mm gap rather than to perform a more dense sampling of the aorta. In a prior study using similar indices,7 we confirmed that this approach had excellent intraobserver reproducibility for aortic anatomy and total subject plaque burden. Improvements in image quality may be obtained by newer pulse sequences, such as dual-inversion black-blood methods,25 or by respiratory motion correction. From an epidemiological perspective, the FHS offspring cohort is predominantly white. These results may not be applicable to other racial subgroups. Finally, the present study did not enroll adults aged <35 or > 90 years.
The present study reports the first CMR population-based study of subclinical aortic atherosclerosis in a free-living population. Among FHS offspring subjects free of clinical coronary disease, subclinical aortic atherosclerosis was seen in 40%. In both sexes, aortic plaque prevalence and all plaque burden measures were greater in the abdomen than in the thorax. Plaque prevalence and plaque burden increased with age and were correlated with the FCRS for men and women.
These CMR data add to prior autopsy-based prevalence estimates and previous CMR studies26,27⇓ and help lay the groundwork for future studies regarding diagnosis, risk stratification, and the treatment of individuals with subclinical atherosclerosis.
The Framingham Heart Study and this project were supported by a grant and subcontract from the National Institutes of Health (N01-HC-38038). Dr Jaffer was supported in part by the William A. Schreyer Fellowship, Cardiology Division, Massachusetts General Hospital, Boston, Mass.
Received December 26, 2001; revision accepted January 28, 2002.
- ↵Holman RL, McGill HCJ, Strong JP, Geer JC. The natural history of atherosclerosis: the early aortic lesions as seen in New Orleans in the middle of the 20th century. Am J Pathol. 1958; 40: 37–49.
- ↵Wilson PW, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation. 1998; 97: 1837–1847.
- ↵Toussaint JF, Southern JF, Fuster V, Kantor HL. T2-weighted contrast for NMR characterization of human atherosclerosis. Arterioscler Thromb Vasc Biol. 1995; 15: 1533–1542.
- ↵Yuan C, Beach KW, Smith LH Jr, Hatsukami TS. Measurement of atherosclerotic carotid plaque size in vivo using high resolution magnetic resonance imaging. Circulation. 1998; 98: 2666–2671.
- ↵Fayad ZA, Nahar T, Fallon JT, Goldman M, Aguinaldo JG, Badimon JJ, Shinnar M, Cheseboro JH, Fuster V. In vivo magnetic resonance evaluation of atherosclerotic plaques in the human thoracic aorta: a comparison with transesophageal echocardiography. Circulation. 2000; 101: 2503–2509.
- ↵Chan SK, Jaffer FA, Botnar RM, Kissinger KV, Goepfert L, O’Donnell C, Levy D, Manning WJ. Scan reproducibility of magnetic resonance imaging assessment of aortic atherosclerosis burden. J Cardiovasc Magn Reson. 2001; 4: 331–338.
- ↵Berenson GS, Wattigney WA, Tracy RE, Newman WP, Srinivasan SR, Webber LS, Dalferes ER, Strong JP. Atherosclerosis of the aorta and coronary arteries and cardiovascular risk factors in persons aged 6 to 30 years and studied at necropsy (The Bogalusa Heart Study). Am J Cardiol. 1992; 70: 851–858.
- ↵Natural history of aortic and coronary atherosclerotic lesions in youth: findings from the PDAY Study: Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Arterioscler Thromb. 1993; 13: 1291–1298.
- ↵Simons PC, Algra A, Bots ML, Grobbee DE, van der Graaf Y. Common carotid intima-media thickness and arterial stiffness: indicators of cardiovascular risk in high-risk patients: the SMART Study (Second Manifestations of ARTerial disease). Circulation. 1999; 100: 951–957.
- ↵Psaty BM, Furberg CD, Kuller LH, Bild DE, Rautaharju PM, Polak JF, Bovill E, Gottdiener JS. Traditional risk factors and subclinical disease measures as predictors of first myocardial infarction in older adults: the Cardiovascular Health Study. Arch Intern Med. 1999; 159: 1339–1347.
- ↵Arad Y, Spadaro LA, Goodman K, Lledo-Perez A, Sherman S, Lerner G, Guerci AD. Predictive value of electron beam computed tomography of the coronary arteries: 19-month follow-up of 1173 asymptomatic subjects. Circulation. 1996; 93: 1951–1953.
- ↵Detrano RC, Wong ND, Doherty TM, Shavelle RM, Tang W, Ginzton LE, Budoff MJ, Narahara KA. Coronary calcium does not accurately predict near-term future coronary events in high-risk adults. Circulation. 1999; 99: 2633–2638.
- ↵Raggi P, Callister TQ, Cooil B, He ZX, Lippolis NJ, Russo DJ, Zelinger A, Mahmarian JJ. Identification of patients at increased risk of first unheralded acute myocardial infarction by electron-beam computed tomography. Circulation. 2000; 101: 850–855.
- ↵Polak JF, Kronmal RA, Tell GS, O’Leary DH, Savage PJ, Gardin JM, Rutan GH, Bohrani NO. Compensatory increase in common carotid artery diameter: relation to blood pressure and artery intima-media thickness in older adults: Cardiovascular Health Study. Stroke. 1996; 27: 2012–2015.
- ↵Haberl R, Becker A, Leber A, Knez A, Becker C, Lang C, Bruning R, Reiser M, Steinbeck G. Correlation of coronary calcification and angiographically documented stenoses in patients with suspected coronary artery disease: results of 1,764 patients. J Am Coll Cardiol. 2001; 37: 451–457.
- ↵O’Rourke RA, Brundage BH, Froelicher VF, Greenland P, Grundy SM, Hachamovitch R, Pohost GM, Shaw LJ, Weintraub WS, Winters WL Jr, 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.
- ↵Corti R, Fayad ZA, Fuster V, Worthley SG, Helft G, Chesebro J, Mercuri M, Badimon JJ. Effects of lipid-lowering by simvastatin on human atherosclerotic lesions: a longitudinal study by high-resolution, noninvasive magnetic resonance imaging. Circulation. 2001; 104: 249–252.