Coronary Artery Risk Development in Young Adults (CARDIA) Study



This meeting was convened by the NHLBI Division of Cardiovascular Sciences, Prevention and Population Sciences Program to obtain advice about research areas that a possible continuation of the Coronary Artery Risk Development in Young Adults (CARDIA) Study could uniquely address. External experts in cardiovascular and pulmonary diseases, social sciences and behavior, lifestyle, diabetes, metabolism, genetics, cognitive functioning, neurology, and population studies were invited to discuss future research opportunities in CARDIA.

The Working Group was welcomed by the DCVS Associate Director, who presented an overview of cohort studies supported and conducted by the Epidemiology Branch, and the overall charge. The Project Officer provided an overview of the design of CARDIA, Project Office preliminary plans for the future of CARDIA, and questions from the Project Office for the Working Group to consider. CARDIA investigators presented select findings from published reports, data on cohort retention, scientific productivity, details on morbidity and mortality ascertainment and experience to date, preliminary results from the 25th year follow-up examination currently in progress, and future research possibilities. After addressing questions from the Working Group, the CARDIA investigators departed. The remainder of the meeting involved discussions and recommendations from the Working Group in the absence of the CARDIA investigators.


Discussion and recommendations:

Dr. Bruce Psaty chaired the Working Group. As a healthy cohort at baseline, and now, with nearly three decades of collected data and stored specimens, the Working Group thought CARDIA provides a rich resource for NHLBI as well as other NIH Institutes to continue investigating important research questions on cardiovascular as well as non-cardiovascular topics.

The Group affirmed the research value of a Year 30 examination. The cohort is continuing through a critical period in midlife when risk factors continue to emerge (e.g., obesity, hypertension, diabetes), subclinical disease becomes more prevalent, and clinical events become more frequent. A Year 30 examination would allow the study to measure the extent and progression of subclinical disease; identify additional surrogates and endpoints [e.g., albuminuria, diabetes, kidney disease, chronic obstructive pulmonary disease (COPD)]; update information on important exposures that change over the life-course; and collect new information not measured at previous exams.

The following is a summary of the research areas discussed and the Working Group prioritizations for the proposed renewal period classified into one of four categories based upon the discussion: high priority, medium priority (topics that did not make the high priority list, but received some level of discussion and enthusiasm), consider (topics mentioned in a positive light, but without much discussion), and low priority.

(1) Understanding racial disparities in CVD risk factors and outcomes:

CARDIA has repeatedly assessed established risk factors for CVD at each clinic examination, including adiposity, blood pressure, diabetes status, and kidney disease. In recent years, strong racial disparities in each of these risk factors have been observed with black men and women experiencing a greater burden than white men and women. Furthermore, in a study of the first 27 cases of heart failure in the CARDIA cohort appearing in a 2009 issue of the New England Journal of Medicine, all but 1 of the 27 heart failure cases were among black adults. To help explain these racial disparities in CVD risk factors and outcomes, CARDIA has measured both biologic (e.g., subclinical disease, heart structure/function, markers of stress, genetics) and non-biologic factors (e.g., socioeconomic status, discrimination, neighborhood characteristics, health care access and utilization). The following is a summary of the recommendations from the Working Group:

Medium Priority:

  • To help explain racial disparities, collect information on health care received, changes in social class, and drug abuse. In terms of biomarkers, measure vitamin D, natriuretic peptide metabolism, nitric oxide biology, and inflammatory factors. Also recommend studying lifestyle changes, plaque characteristics, blood pressure treatment/control, previously identified single nucleotide polymorphisms (SNPs), community-level influences, differences in heart structure/function, and epigenetics.
  • Examine racial disparities in cognitive functioning
  • Continue to examine racial disparities in women?s health

(2) Cardiovascular function and imaging:

CARDIA has collected a wealth of data on cardiovascular functioning and imaging of subclinical CVD over the last 25 years. A resting electrocardiogram was obtained at baseline, Year 7, and Year 20 (on most participants). An echocardiogram was performed at Years 5, 10 (on a subset), and 25. Electron beam computed tomography (CT) of the heart was performed at Years 10 (on a subset), 15, 20, and 25. Carotid intima-media thickness (IMT) was measured at Year 20 and a CT of the abdomen to assess abdominal aortic calcified plaque is being performed at Year 25. The following is a summary of the recommendations from the Working Group:

High Priority:

  • An assessment of cardiac structure/function. Echocardiography was preferred over magnetic resonance imaging (MRI) at the next exam. This was due to a number of reasons: (1) the inclusion of echocardiography in past exams and therefore the ability to document change in heart structure and function over time; (2) the performance of an echocardiogram in the research clinic affords the advantage of not having to schedule appointments at an outside imaging facility; and (3) a greater number of participants will be eligible for an echocardiogram due to MRI exclusion criteria.
  • A resting electrocardiogram
  • A measure of vascular compliance such as pulse wave analysis
  • Ultrasound to measure carotid IMT and plaque characteristics. However, the Working Group noted that documenting change in carotid IMT from Year 20 may be difficult.

Low priority:

  • Chest CT due to its frequent inclusion in past exams and exposure to radiation
  • Stress imaging
  • Ambulatory electrocardiography

(3) Diet, physical activity, body composition, nonalcoholic liver disease, and diabetes:

A wealth of information has been collected in the areas of diet, activity, and body composition in CARDIA. Body weight, waist circumference, and self-reported physical activity have been assessed at each clinic exam (hip circumference at most exams). At Year 25, an abdominal CT is measuring the extent of visceral, subcutaneous, liver, and intermuscular fat. In addition, the Year 25 exam includes questionnaires to assess sedentary behavior, sugar-sweetened beverage consumption, and regular diet practices (Years 7-20 as well). Diet has been measured with the CARDIA diet history and cardiorespiratory fitness with a graded treadmill exercise test at baseline, Year 7, and Year 20. Diabetes has been assessed at each clinic examination. The following is a summary of the recommendations from the Working Group:

High Priority:

  • Remeasuring diet due to changes in intake over time. Use of the CARDIA diet history was highly recommended. The Working Group suggested modifying the instrument to include contemporary food items. Administering the instrument over the phone was discussed, but there were concerns about comparability with past administrations and difficulty in using three-dimensional food items over the phone. Pre-testing was suggested if there is a change in the method of administration.
  • Re-administering questionnaires on physical activity, sedentary behavior, weight control, self-image, sugar sweetened beverage consumption, and diet practices
  • Repeating anthropometry with hip circumference

Medium priority:

  • Use of DXA to measure body composition and bone mineral density (since there are important changes in bone density in mid-life, particularly in women)
  • Remeasuring cardiorespiratory fitness since fitness was last measured at Year 20
  • An abdominal CT to measure changes in visceral, subcutaneous, liver, and intermuscular fat


  • Using an accelerometer to measure physical activity; however, opinions were mixed. Some members believed it was important to measure activity with an accelerometer because of changes in physical activity and sedentary behavior over time and the technological improvements that have been made to many of the devices. Other members believed the devices were not user friendly and their output was difficult to understand.
  • Measuring advanced glycation end products
  • Measuring liver enzymes and other markers of liver disease
  • Measuring biomarkers such as leptin and adiponectin

Low priority:

  • Repeating an oral glucose tolerance test

(4) Pulmonary function and disease:

Spirometry was performed in CARDIA at baseline and Years 2, 5, 7, 10, and 20. A questionnaire to assess conditions involving the chest, including coughing, phlegm production, wheezing, and breathlessness has been included at baseline and Years 2, 10, 15, and 20. The following is a summary of the recommendations from the Working Group:

High priority:

  • An assessment of sleep disordered breathing with a non-invasive home-based monitor and the administration of sleep questionnaires
  • Remeasuring lung function at Year 30. At a Year 30 exam, participants will be at an age when there is a significant increase in the number of adults diagnosed with COPD. Lung function should be measured pre- and post-bronchodilator therapy to differentiate between asthma and COPD. However, to reduce participant burden and exam time, bronchodilator therapy can be administered only to participants with reduced-to-moderately reduced lung function.

Low priority

  • A computed tomography scan of the chest due to the availability of images from past exams
  • A test of diffusing capacity due to the age of the cohort
  • A bronchial challenge test since it involves a high level of participant burden and is time-consuming. Self-reported information on asthma history was considered to be sufficient.

(5) Psychosocial factors, stress, social determinants of health, health measures/outcomes, and racial disparities:

CARDIA has offered a rich opportunity to investigate the role of psychosocial factors in the etiology of CVD and in the explanation of racial disparities in cardiovascular health and disease. Psychosocial constructs measured in previous CARDIA examinations include depression, social support, job strain, goal striving stress, quality of life, discrimination, chronic burden, care-giving stress, subjective social standing, social networks, type A personality, neighborhood cohesion, personal mastery, life orientation, anger-in, anger-out, hostility, optimism, socioeconomic status, and life events. Since many of these constructs are likely to change over the life-course, the Year 25 exam includes repeat assessments of depression, discrimination, chronic burden, quality of life, social networks, and socioeconomic status. The following is a summary of the recommendations from the Working Group:

High priority:

  • Repeating questionnaires on depression, discrimination, and neighborhood cohesion due to the potential for these constructs to change over time
  • Including new questionnaires to assess neighborhood stress, social engagement, and positive/negative affect


  • Initial or repeat assessments of chronic stress; neuroticism; multiple life roles and the changes, stresses, and rewards of those roles; relationship quality; loneliness; substance abuse; childhood trauma and early life exposures; and a mental health history

(6) Vascular contribution to cognitive impairment and dementia:

Instruments to measure cognitive functioning were included for the first time in CARDIA in the Year 25 examination. These instruments include the Rey-Auditory Verbal Learning Test, the Digit Symbol Substitution Test, and the Stroop Test. An MRI of the brain is being performed on a subset of participants. This scan will measure structural vascular and neurodegenerative lesions, in addition to novel markers of early pathology including cerebral blood flow, vascular reactivity, and tissue integrity. The following is a summary of the recommendations from the Working Group:

High priority:

  • Repeating the Rey-Auditory Verbal Learning Test and the Digit Symbol Substitution Test
  • Including a reading test, such as the National Adult Reading Test (NART). The Wechsler Adult Intelligence Scale-Revised (WAIS-R) can be considered as an alternative. The rationale for including tests such as these is to better understand the high school era achievement level of participants.
  • Performing a brain MRI. Increasing the number of participants with an initial MRI was considered higher priority than performing a repeat scan on the subsample at Year 25 because there is likely to be little change in brain structure or function over five years at this age.


  • Repeating the Stroop Test
  • Including a test of attention and reaction time, such as the Flanker Test
  • Performing a positron emission tomography scan; however, the scientific value of performing these scans is unknown because almost all available data have been obtained in older cohorts. The cost of performing these scans also may be prohibitive.

(7) Chronic kidney disease:

Renal function has been measured with the assessment of serum creatinine at baseline and Years 10-25. Cystatin C has also been measured in most participants at Years 10, 15, and 20. Albuminuria has been measured at Years 10-25 with a paired assessment of urinary creatinine. The following is a summary of the recommendations from the Working Group:

High priority:

  • Remeasuring creatinine, cystatin C, urinary albumin, and urinary creatinine

(8) Genetics:

The CARDIA Study has collected and stored buffy coats starting at Year 10. The contract has supported the measurement of about 100 SNPs and CARDIA has contributed to genome-wide association efforts through its participation in multiple collaborative efforts and consortia, including the Candidate gene Association Resource (CARe) study, the Blood Pressure Genome Wide Association Study (GWAS) in the Gene-Environment Interaction Initiative, Gene Environment Association Studies (GENEVA), Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE), Genome-wide Investigation of Anthropometric Measures (GIANT), and the NHLBI Grand Opportunity Exome Sequencing Project (GO ESP). In addition, CARDIA contributes to the investigation of context-dependent effects of SNPs influencing CVD risk, which have been identified through GWAS, as part of the Genetic Epidemiology of Causal Variants Across the Life Course ancillary study, which is part of a larger consortium: the Population Architecture using Genomics (PAGE).

High priority:

  • Continuing participation in collaborative genetic efforts and consortia


  • Studies in the area of epigenetics

(9) Other Comments:

High priority:

  • Collecting discharge summaries for all hospitalizations. These data can serve as a resource for generating ancillary study applications.
  • Collecting and storing biological specimens for use by ancillary studies, including saliva, serum, plasma, urine, buffy coats, red blood cells, and toenails
  • Using PAXgene tubes to collect blood for the measurement of RNA
  • Continuing support for early stage investigators


  • Collecting and storing fat samples
  • Establishing a CARDIA offspring cohort

The meeting was adjourned at 4:00 p.m.