To further the study of the relationship between risk factors and atherosclerosis and to avoid the biases involved in relying solely on clinical disease outcomes, objective and unbiased measures of cardiovascular disease (CVD) have been introduced in NHLBI-initiated epidemiologic studies of CVD etiology. Two well-developed examples include echocardiography and carotid ultrasound. The observation that risk of subsequent cardiovascular events and mortality increased 1.5 to 2-fold per 50 gm/m of echocardiographically-defined left ventricular mass in Framingham led to inclusion of echocardiographic LV mass assessment in several population samples. Development and validation of duplex ultrasound imaging as a measure of atherosclerosis provided for the first time a method for visualizing the atherosclerotic process at the level of the vessel wall and suitable for application in population-based studies.
Use of carotid ultrasound in the Atherosclerosis Risk in Communities (ARIC) Study has demonstrated strong relationships of carotid wall thicknesses to clinical CVD and to traditional CVD risk factors such as age, sex, lipids, hypertension, smoking, diabetes, obesity and dietary fat. Using the powerful nested case-control design, in which atherosclerosis cases and non-atherosclerotic controls are selected from the entire population distribution of wall thickness, ARIC has also demonstrated strong associations of carotid atherosclerosis with Lp(a) and Lp(a) phenotypes, antibodies to infectious agents, fibrinolytic factors and inflammatory factors, cellular adhesion molecules, homocysteine levels, and dietary magnesium and antioxidants.
Carotid ultrasound and echocardiography were also included in the Cardiovascular Health Study (CHS) and were quickly demonstrated to bear strong relationships with overt CVD in this older population sample, to be associated with many conventional CVD risk factors, and to be strong predictors of subsequent overt cardiovascular events. More importantly, short-term risk of clinical events was shown to be low to non-existent in the absence of subclinical disease.
Subclinical disease measures have thus enhanced studies of CVD risk by examining the early stages of CVD in an objective manner free of biases related to severity, diagnostic suspicion, or completeness of medical investigation. Because subclinical disease is asymptomatic and previously unknown to participants, it is unlikely to have any direct impact on health behavior, such as lifestyle modification or medication use, which may limit the detection of risk relationships with clinical disease. In addition, the continuous nature of most subclinical measures greatly increases power to detect risk associations compared to discrete measures, such as presence or absence of clinical events. Most importantly, subclinical measures now permit the epidemiologic investigation of CVD risk to focus on the biology of CVD rather than on vagaries in its diagnosis. Application of non-invasive measures in ongoing prospective, population-based studies has demonstrated a surprisingly high prevalence of subclinical disease. Atherosclerotic plaque is extremely common in the extracranial carotid arteries of United States men and women older than 45 years and can be detected in subjects at low risk as defined by traditional risk factors. Coronary calcium, a specific marker for coronary atherosclerosis, is present in 10 percent of women and 30 percent of men aged 30-39 and increases in prevalence with age. One-sixth of men and nearly one-fifth of women in Framingham have echocardiographically defined left ventricular hypertrophy, and over 30 percent of persons over age 65 in CHS have MRI-defined cerebral infarcts.
Standardized measures of retinal arteriovenous nicking, which are strongly associated with hypertension and cerebral infarction, are present in 14 percent of middle-aged ARIC participants. Quantitative retinal arteriolar narrowing increases across the entire range of blood pressure, even in persons receiving effective antihypertensive therapy, and may provide an integrated measure of duration and severity of hypertension.
Risk associated with subclinical disease measures has been shown to be graded and continuous (similar to risk associated with conventional CVD risk factors such as hypertension and cholesterol) rather than demonstrating a threshold level at which risk increases sharply. This suggests that interventions yielding even modest reductions in levels of subclinical disease should be explored for their potential impact on reducing CVD risk. To design such interventions, factors contributing to the development and progression of subclinical disease must be identified.
The recognition that plaque rupture is a key event in coronary thrombosis and that plaques that rupture tend to be subcritical stenoses associated with lipid-laden lesions has shifted the focus of etiologic research to factors leading to formation and rupture of unstable plaque, such as inflammation and impaired endothelial function. Inflammatory and infectious factors have long been known to be associated with CVD in epidemiologic studies and recognition of the importance of plaque rupture provides a plausible mechanism for this relationship. Continued research on inflammation and CVD risk in populations thus provides a promising avenue for elucidating mechanisms of plaque rupture. It also exemplifies the cyclical and complementary nature of risk factor identification in population and experimental studies, progressing from clinical events at the population level to the tissue and cellular level experimentally, and back to the tissue and population level in studies of subclinical disease.
The project originated in the Epidemiology and Biometry Program of the Division of Epidemiology and Clinical Applications. The concept derived from an NHLBI Task Force on Research in Epidemiology and Prevention convened in 1993, in which investigation of subclinical disease and its progression to clinical disease was recommended as a major focus of future NHLBI population studies. This was followed by a Special Emphasis Panel on Longitudinal Cohort Studies in June 1995, which strongly recommended studies based on subclinical disease measures, and the inclusion of underrepresented minorities, particularly Hispanics, in population based research. A subsequent Special Emphasis Panel on Use of Cardiac EBCT and MRI in Epidemiologic Studies of Cardiovascular Disease in June 1996, recommended inclusion of carotid and cardiac MRI and EBCT in elucidating the progression of subclinical to clinical disease and identifying subclinical disease characteristics most strongly associated with increased risk. In November 1997, the Requests for Proposals were released for the Coordinating Center, Field Centers, Special Laboratory Center, Ultrasonography Reading Center, Magnetic Resonance Imaging Reading Center, and Electron Beam Computed Tomography Reading Center.
Participants were recruited from the Field Centers and screened for evidence of subclinical coronary atherosclerosis, using electron-beam computed tomography (EBCT) or another noninvasive measure of subclinical disease, and for coronary risk factors. A sample of those at highest risk for coronary heart disease events, within defined age-, sex- and ethnicity-strata and an age-, sex-, ethnicity-, center-matched sample of participants representing the normal spectrum of subclinical coronary disease were reexamined in more depth approximately six months later for other evidence of subclinical disease and for characterization of possible precursors of disease events. The cohort was characterized with respect to coronary calcification (using computed tomographic scanning), ventricular mass and function (using cardiac magnetic resonance imaging), flow-mediated endothelial function, carotid intimal-medial wall thickness and distensibility in the carotid artery (using ultrasonography), lower extremity vascular insufficiency (using ankle-brachial blood pressure indices), electrocardiographic measures, standard coronary risk factors, sociodemographic factors, lifestyle factors, and psychosocial factors. Blood samples were assayed for putative biochemical risk factors and stored for case-control studies. DNA was extracted and lymphocytes immortalized for study of candidate genes and possibly genome-wide scanning. Four clinical examinations, 18 to 24 months apart, were planned. Exam 2, beginning in September 2002 and ending in January 2004, included repeat coronary calcification measures and carotid MRI in subsets of the cohort. Exam 3 is underway. A second cardiac computed tomography is performed on the cohort during Exams 2 and 3.
Because the study will provide examination results to participants and their physicians, there is expected to be a significant amount of clinical intervention in individuals, depending on the severity of the disease, the acceptance of the test results by the medical community, and the aggressiveness of the individual treating physicians. The study will collect information on the nature and magnitude of such interventions to control for them in analyses. Approximately 15 percent of the highest risk participants will be selected for this comprehensive examination and follow-up. Assuming that a large proportion of these high risk participants will be referred for an intervention that substantially modifies their risk, the effective sample size after accounting for the intervention effect is assumed to be 10 percent, rather than 15 percent. All participants will be followed for identification and characterization of cardiovascular disease events, including acute myocardial infarction and other forms of coronary heart disease, stroke, and congestive heart failure; mortality; and for any cardiovascular disease interventions received.
The study involves seven Field Centers, a Coordinating Center (which will subcontract for a Central Lipid Laboratory), a Special Laboratory Center; and separate Reading Centers for Ultrasonography, Magnetic Resonance Imaging, and Electron Beam Computed Tomography (EBCT).
The MESA was expanded in 2003 when the MESA Family Study was added. The purpose of the MESA Family study is to locate and identify genes contributing to sub-clinical cardiovascular disease in United States minority populations. These goals will be addressed in a study of 2700 individuals from 900 sibships (sibtrios or larger), evenly distributed among African-Americans and Hispanic Americans, utilizing the existing framework of the NHLBI Multi-Ethnic Study of Atherosclerosis (MESA). In Aim 1, the MESA Family Study will determine the extent of genetic contribution to variation in coronary artery calcification (CAC) (EBCT and helical-gated CT scan) and intimal medial thickness (IMT) (B-mode ultrasound) in these two populations. This aim will be accomplished by examination (phenotyping) of 1800 siblings from 900 MESA index cases (evenly divided between African-Americans and Hispanic-Americans). In Aim 2, biological candidate regions in the human genome linked to these quantitative sub-clinical cardiovascular disease traits (coronary calcium and IMT) will be identified by genome scan approaches, including fine mapping of the best regions. This aim will use the MESA Study resources (Data Coordination Center, Central Laboratory, CT and Ultrasound Reading Center, 6 Clinical Field Centers) and the combined resources and cardiovascular genetic epidemiology expertise at Cedars-Sinai Medical Center and Wake-Forest School of Medicine. In Aim 3, gene localization and identification will be accomplished by association studies of positional as well as biological candidate genes in the subjects from the 3 minority populations of MESA (African-Americans, Hispanic-Americans, and Chinese-Americans). Whereas the purpose of MESA (the parent study) is to assess sub-clinical CVD and identify epidemiological risk factors in multi-ethnic populations, the purpose of the MESA Family Study is to identify the genes (quantitative trait loci, or QTLs) that contribute to these sub-clinical CVD risk factors. In combination with traditional risk factor assessment, identified inherited markers should be useful in the management of patients with vascular disease. These results will permit targeted diagnostic testing and pharmacologic intervention, and in the identification of subjects who could benefit from specific prevention protocols that result in an increase in the overall well being of the US population.