Working Group Report: Future Research Opportunities in CARDIA

Meeting Summary

March 6, 2007

I. Purpose:

This meeting was convened by the NHLBI Division of Prevention and Population Sciences to obtain expert opinions about future research plans for a renewal of one of NHLBI's large contract-funded multi-center observational cohort studies, the Coronary Artery Risk Development in Young Adults (CARDIA) Study. The Working Group was charged with identifying new research opportunities in CARDIA and discussing how this research fits into the NHLBI Strategic Plan. The early part of the meeting included presentations from CARDIA investigators, who provided some background materials, research findings, and their ideas for future research directions. After addressing questions from the Working Group members, the CARDIA investigators then departed. The remainder of the meeting involved discussions and recommendations from the Working Group in the absence of the CARDIA investigators.

II. Background:

CARDIA began as a study to examine the evolution of cardiovascular disease (CVD) risk factors during young adulthood in 1985. It started with a cohort of 5115 healthy black and white men and women aged 18-30 years, selected to have approximately the same number of people in subgroups of race, gender, education (high school or less and more than high school) and age (18-24 and 25-30). These same participants were asked to participate in follow-up examinations during 1987-1988 (Year 2), 1990-1991 (Year 5), 1992-1993 (Year 7), 1995-1996 (Year 10), 2000-2001 (Year 15), and 2005-2006 (Year 20); the proportions of the surviving cohort who returned for these examinations were 90%, 86%, 81%, 79%, 74%, and 72%, respectively. In addition to the follow-up examinations, participants are reached regularly for their contact information and the ascertainment of hospitalization and clinical events. Within the past five years, 94% of the original surviving cohort has been contacted.

The current CARDIA contract ends on August 30, 2008. The Project Office envisions a potential renewal to include a follow-up examination in 2010-11, when participants will be 43-55 years old, with an estimated cohort size of about 3500 (~70% of the living); surveillance for events would continue through 2013, after which the possibility of another contract renewal might be considered. Three potential research topic areas for this Year 25 examination include: 1) identifying predictors of early development and rapid progression of CVD risk factors, subclinical atherosclerosis, and impaired lung function, to accurately inform prevention and intervention efforts from the very beginning of adulthood; 2) examining young adult antecedents and consequences of obesity and inflammation in the four CARDIA groups: black/white by male/female; and 3) assessing racial differences in severity and progression of early CVD risk factors and subclinical disease.

III. Discussions and Recommendations:

As a healthy cohort at baseline, and now, with over two decades of collected data and stored specimens, the Working Group thought that CARDIA has provided a rich resource for NHLBI. They strongly encouraged NHLBI and other NIH Institutes to continue to invest in this study to examine topics related to cardiovascular as well as non-cardiovascular outcomes. The Group affirmed the value of continuing in-person examinations in CARDIA and suggested specific research topics that CARDIA is well-positioned to study in the future.

A. Year 25 Exam and Rationale:

The Working Group highly recommended that an examination be conducted in Year 25 and in future years as funding allows. This cohort is entering a critical period in which the originally young adult cohort is transitioning into mid-life, when key risk factors such as obesity, hypertension, and glucose intolerance become more prevalent and clinical events begin to emerge. Event ascertainment alone would be inadequate to collect traditional or contemporary measures that are either more accurately or only feasibly obtained via an in-person examination. An in-person examination would allow the study to collect subclinical disease measures as surrogates for clinical endpoints, since few clinical events will have accrued by then. It would also allow the limited DNA stores to be replenished for further genetics research. If renewed, the Year 25 examination could be strengthened by the following efforts:

• Develop goals for translating CARDIA findings relevant to health promotion and disease prevention; conduct research related to translating key findings into policy issues
• Increase the breadth of scientific expertise involved in the study; further promote in-depth ancillary studies spearheaded by outside investigators
• Make concerted effort to perform in-depth studies and analyses to explain the observed racial/gender/socioeconomic status (SES) differences in CARDIA
• Provide a concrete plan for systematically conducting in-depth analyses; establish expert writing groups to thoroughly address the Y25 scientific objectives

B. CARDIA and the NHLBI Strategic Plan:

CARDIA's past and present research activities already fit with several elements of the NHLBI Strategic Plan. To further advance its contribution to the Strategic Plan, the Working Group recommended the following:

• To enhance transmission of knowledge between basic science and clinical research, NHLBI could require a stronger basic science component in CARDIA
• To enhance translation of research into clinical practice, CARDIA should:
• Explore novel findings on health disparities to explain its observed black/white differences
• Generate hypotheses for clinical trials, including behavioral interventions, from its findings
• Make concerted effort to publicize its findings that contain strong public health messages
• To support the development of multidisciplinary teams and expand the expertise involved in the CARDIA study, NHLBI could encourage ancillary studies by issuing parallel Requests for Applications (RFAs) concurrent with the core contract renewal

C. Future Research on Obesity, Inflammation, and Diabetes:

The Working Group recommended that obesity research should remain a primary objective in CARDIA.

• Study the consequences of obesity, including its effects on renal function, inflammation, muscle metabolism and the development of diabetes, atherosclerosis, and vascular abnormalities. Electron beam computerized tomography (EBCT) is not recommended to measure atherosclerosis at Year 25, as it inadequately captures non-calcified plaques and CARDIA has already collected coronary calcium scores at Year 15 and Year 20
• Perform imaging (such as computed tomography (CT) or magnetic resonance imaging (MRI)) of abdomen and/or quadriceps to measure adiposity (visceral fat and skeletal muscle fat) and muscle metabolism in relation to obesity trends and obesity outcome
• Perform cognitive function testing and brain MRI in all and functional MRI in a subgroup; this can serve as baseline for future follow-up on vascular dementia
• Identify psychosocial and behavioral determinants on weight gain and weight maintenance
• Administer questionnaires on home environment (e.g. spouse behavior, children food habits, weight of spouse and children)
• Administer questionnaires on stressors, job strain, and depression
• Perform challenges or provocative tests (consider this for at least subsets of persistently thin, weight gain, and/or extreme obese groups), such as
• Oral glucose tolerance test
• Dexamethasone suppression test
• Feeding behavior study
• Effects of exercise on biology (inflammation)
• Consider obtaining biological tissues (probably in a subgroup), such as
• Adipocytes
• Immune cells
• Consider studying neurohormonal elements related with obesity, inflammation, and diabetes.
• Salivary cortisol
• Endogenous hormones
• Identify determinants of the black/white and gender differences observed in CARDIA and examine new aspects of risk factors for obesity and their implications
• Explore gene-by-environment interactions of all the above topics using pre-existing CARDIA data, analyses of stored specimens, and new data collected in Year 25
• Study the impact of the obesity epidemic on the next generation: consider obtaining basic phenotypic data on the offspring

D. Future Research on Vascular Diseases:

The Working Group identified several research opportunities for evaluating vascular abnormalities in CARDIA as the prevalence of hypertension, diabetes, and other cardiovascular risk factors increases.

• Study whether and how elevations of blood pressure, changes in blood pressure determinants including sodium intake, and specific aspects of blood pressure (e.g. arterial waveforms and nocturnal blood pressure) in young adulthood are etiologically related to decline in cognitive function, incident brain lesions (i.e. white matter disease), and decline in kidney function. Further characterize the development of hypertension, particularly important as the prevalence of hypertension is increasing in this cohort. Consider obtaining the following measures:
• Cognitive function interviews (in Year 25 and Year 35)
• Brain MRI (in Year 25 and Year 35)
• Sophisticated blood pressure measurements (e.g. tonometry, ambulatory blood pressure monitoring)
• Retinal photographs
• Improve phenotypes of early kidney disease and determine subsequent risk of subclinical CVD, particularly interactions with other CVD risk factors. Consider obtaining:
• 24-hour urine for sodium, potassium, creatinine, albumin, and protein (provide participants with additional monetary reimbursement for this)
• Spot urine to enhance clinical relevance
• Consider measuring early, non-calcified asymptomatic cardiovascular disease. Multi-detector computer tomography (MDCT) with angiography could serve this purpose if feasibility issues are appropriately addressed. Potential risks include radiation exposure (levels of which would likely decrease as technologies advance) and allergic reactions (including the possibility of anaphylaxis) to intravenous contrast dye. Repeat scanning of only EBCT is not recommended as it inadequately captures non-calcified plaques.

E. Future Research on Pulmonary and Ventricular Function:

The Working Group identified several research opportunities for the evaluation of pulmonary function abnormalities and characterization of ventricular parameters in CARDIA.

• Compare peak ("plateau") lung function in young adulthood with incident COPD in mid-life and measure regional distribution of percent emphysema. COPD is defined/diagnosed by a post-bronchodilator FEV1/FVC ratio <0.70. Self-reports of COPD are unreliable due to both reporting and diagnostic bias; considerable misclassification occurs with the use of only pre-bronchodilator measures. CT scans with full-lung view provide valid estimates of regional distribution of percent emphysema and total lung capacity (TLC). TLC is essential to help define restrictive disease, particularly in relationship to obesity. Full-lung sections could be added to a cardiac scan protocol relatively easily on MDCT machines with little additional radiation exposure. They would probably be feasible on EBCT machines in a middle-aged, relatively healthy population (the main relevant limitation of EBCT scans is the longer breath-hold). Consider obtaining the following tests in Year 25:
• Pulmonary function test (PFT) pre- and post-bronchodilator challenge to measure early incident COPD
• CT with full lung views to measure lung capacity and density
• Measure left ventricular mass, structure, and atrial size. Diastolic function indices may be particularly useful for examining precursors to heart failure in this biracial cohort. Diastolic function, as well as valvular structure and function, can be characterized by echocardiography. Given that 2-D echocardiography was conducted in CARDIA in prior exams (Year 5 and Year 10), repeating it at Year 25 offers the opportunity to study changes from young adulthood to mid-life and the ability to identify factors associated with this progression. Further, 2-D echocardiography could potentially be more clinically relevant given its wide use in clinical practice. More advanced measurements such as 3-D echocardiography or cardiac MRI should also be considered, as they allow for more detailed characterizations of phenotypes. Compared to 2-D, 3-D echocardiography provides more precise assessment of left ventricular regional systolic and diastolic function. Cardiac MRI has the advantage of providing more reproducible assessments of left ventricular mass and aortic morphology and distensibility; there is also less operator dependence in image acquisition. In summary, consider including in the Year 25 exam:
• Cardiac echocardiography (2D or 3D) or
• Cardiac MRI

F. Other Comments:

• Include more questions with policy implications (e.g. environmental effects on health, access to health care questions) in order to develop a prevention agenda
• Further expand research on women's reproductive health to include menopause transition; examine the development of subclinical and clinical diseases during the menopause transition and identify factors related to menopause that may impact this development
• Consider measuring Apo-B in Year 25, since estimated LDL may not accurately reflect actual LDL levels
• Consider measuring complete blood counts in Year 25, especially hematocrit and hemoglobin level for anemia
• Highly encourage the investigators to publish a summary of CARDIA findings with strong public health messages in well-known science journals
• Consider linking participants who need follow-up with regional consortiums of Clinical and Translational Science Awards (CTSAs) - there is a need for Institute-wide policy on reporting results back to participants, including addressing whether care should be provided to high-risk participants who are without health insurance
• Consider more frequent exams than every five years for some subgroups and for some risk factors that emerge at a rapid rate
• Work harder to bring back those who did not return for the recent Year 20 exam; missing data, which were most common in black men, could be problematic
• Encourage interaction and collaboration with other studies, especially for policy and prevention messages and data-sharing; NHLBI could take the lead in establishing standardized definitions of phenotypes collected in its cohort studies and provide financial support for coordination among the studies
• Perform secular trend analyses comparing and contrasting with the findings from the Atherosclerosis Risk in Communities (ARIC) study, given that there is now overlap in age between some of the CARDIA and ARIC participants
• Consider establishing an Ethics Advisory Board; include participants among the Board members

Working Group Panel Members:

Allen Taylor (Chair), Lawrence Appel, R. Graham Barr, Emelia Benjamin, (Ingrid Borecki, absent), Frederick Brancati, Josephine Briggs, Mary Cushman, Caroline Fox, Lewis Kuller, Shiriki Kumanyika, Lenore Launer, Rena Wing

CARDIA Investigators:

Stephen Hulley, Cora E. Lewis, Kiang Liu, O. Dale Williams

NHLBI Staff:

Jane Harman, Cay Loria, Denise Simons-Morton, Hanyu Ni, Jean Olson, George Papanicolaou, Peter Savage, Paul Sorlie, Gina Wei

Last Updated May 2007