This is a short narrative bibliography of several key references in cardiovascular epidemiology. This is not a comprehensive list, but a starting point for those interested in gaining a better understanding of the goals and methodologies of cardiovascular disease epidemiology.
The subject area is as vast as the public health issue of cardiovascular disease. A graphical overview of the magnitude and temporal trends of cardiovascular disease in the United States can be found in the NHLBI morbidity and mortality chart book. Every year the American Heart Association publishes updated statistics on heart disease and stroke.
An historical overview of cardiovascular epidemiology can be found by exploring this Web site (Preventing Heart Attack and Stroke: A History of Cardiovascular Disease Epidemiology). The American Heart Association Council on Epidemiology and Prevention has provided a home for the fruitful collaborations that advance this exciting field. We encourage the reader to follow their curiosity and read about the colorful history of the council (History of the Council on Epidemiology and Prevention, American Heart Association).
Historical Example — Low Density Lipoprotein Cholesterol
To make sense of where epidemiology fits into overall cardiovascular disease prevention, we will review the historical example of a prominent and major cardiovascular disease risk factor: low density lipoprotein (LDL) cholesterol.
In the late 19th century, careful study of pathologic specimens by Virchow and others led to better characterization of atherosclerosis and the identification of lipid accumulation as a central feature of this disease. Subsequent animal experiments in the early to mid 1900's revealed that atherosclerosis could be experimentally induced in a variety of species by cholesterol and fat enhanced diets (reviewed by Stamler J in Coronary Heart Disease Epidemiology, Marmot M ed, 1992:35-66). Additionally, from an epidemiologic perspective, there was the realization that fractions of circulating cholesterol were associated to the cardiovascular complications of atherosclerosis (Goffman et al., Circulation 14:689-741, 1956). In longer term follow-up of the Framingham Heart Study, the observation that levels of LDL cholesterol were risk factors predicting coronary heart disease among a relatively unselected community based cohort of men and women was confirmed (Kannel WB et al., Ann Intern Med 61:888-899, 1964). By the 1970’s the laboratory of Drs. Brown and Goldstein had done considerable basic biochemical work relating to the metabolism and transport of cholesterol and lipoprotein receptors. Their contributions to the basic understanding of these processes (reviewed by Brown and Goldstein, Science 191:150-154, 1976) led to their interests in genetic defects causing hypercholesterolemia (Goldstein and Brown, Proc Nat Acad Sci USA 70:2804-2808, 1973). Ultimately, these studies facilitated the development of HMG CoA reductase inhibitors (statins), the most potent cholesterol lowering agents. Over the subsequent decades, many large-scale drug trials were aimed at reducing levels of LDL cholesterol, and taken together these trials convincingly demonstrated that these interventions were indeed successful in lowering LDL cholesterol and preventing cardiovascular disease events (Law MR et al., BMJ 326:1423-1427, 2003). In this outline, we have dissected a linear progression from pathologic and animal studies, through small clinical studies and then larger observational studies, to basic science, finally culminating in definitive interventional studies. However, progress in understanding how to use all of these strategies continues in parallel with a fascinating give-and-take of ideas between the different streams of investigation with epidemiology playing a central and crucial role.
Whereas risk factors such as LDL cholesterol (along with other established risk factors for cardiovascular disease) are well established, several “novel” biomarkers for cardiovascular disease are being currently investigated. The vast subject of biomarkers for cardiovascular disease was recently reviewed (Vasan RS, Circulation 113:2335-2362, 2006).
The immediate generalizability of rigorous epidemiological research informs guidelines for clinical decision-making. Systematic summaries of multiple studies using meta-analyses often inform the collective wisdom, for example, as in the identification of prehypertension as a clinical category (Lewiston S et al., Lancet 360:1903-1913, 2002). The thought process involved in the translation of research to practice can be found in the JNC 7 guidelines for the treatment of hypertension and NCEP ATPIII guidelines for hypercholesterolemia.
Along with the traditional epidemiologic hypothesis development and confirmation approach that we have described above, a new revolution in cardiovascular prevention research is expected given the advent of emerging technologies that have made rapid scanning of the whole genome feasible. Thus, we may be at the threshold for finding entirely novel ways to predict risk and otherwise unforeseen opportunities for intervention (Hirschhorn JN, Nature Reviews Genetics 6:95-108, 2005).
Cardiovascular Epidemiology — Statistical Methods
The progress of cardiovascular disease prevention science incorporates many sterling intellectual faculties: intuitive understanding, clinical acumen, ability to generalize observations, and the devising of pragmatic interventions. Scientists always realize that though the flash of anecdotal insight is necessary, the ability to summarize and compare observations is essential to create a generalizable and usable corpus of experience. Thus, the statistical methods that summarize observations informatively and allow for fair comparisons of alternatives are an essential part of the cardiovascular epidemiologist’s toolkit. Journals such as Circulation and the BMJ have published tutorial articles on statistical methods from time to time, which are informative and targeted to clinician researchers.