Risk of Coronary Heart Disease in Subjects with Chest Discomfort: The Framingham Heart Study JOANNEM. MURABITO, M.D., FraminghamandBoston, Massachusetts, KEAVEN M. ANDERSON, Ph.D., framingham, Massachusetts, andBethesda, Maryland, WILLIAM 8. KANNEL, M.D., JANEC. EVANS, M.P.H., Boston, Massachusetts, DANIELLEVY, M.D., Framingham, Massachusetts, and Bethesda, Maryland PURPOSE: To examine the risk of coronary heart disease (CHD) events in subjects of the Framingham Study reporting new chest discomfort. SUBJECTSANDMETHOD&th@id cohortsub- jecta with chest discomfort were classified by their history into three groups: definite angina, possible angina, or nonanginal chest discomfort. Subjects were followed for 2 years for CHD events, including coronary insufficiency, myo- cardial infarction, or CHD death. RESULTS: Compared to that in subjects without chest discomfort, the relative odds of a CHD event was 3.7 (95% confidence interval [CI] 2.11, 6.60) in men with definite angina and 3.0 (95% CI 1.33,6.69) in men with possible angina. Com- parable increased CHD risk was also observed in women with definite or possible angina, with relative odds of 5.4 (95% CI 3.08,9.30) and 2.9 (95% CI 1.13,7.17), respectively. The increase in CHD risk associated with definite or possible angina persisted after adjustment for cardiac risk factor profile. There was no increase in risk associated with nonanghml chest discomfort. CONCLUSION: CHD risk is increased in subjects with new chest discomfort that on the basis of history is consistent with definite or possible an- gina, whereas CHD risk is not increased in sub- jects with nonanghml chest discomfort. The presence of chest discomfort and its characteris- tics facilitate the classification of subjects into meaningful categories that offer prognostic in- formation beyond that provided by traditional CHD risk factors. From the Framrngham Heart Study, Framingham, Massachusetts (JMM, KMA. DL), Division of Epidemiology and Clinical Applications of the National Heart, Lung, and Blood Institutes. Bethesda, Maryland (KMA, DL). and the Section of Preventive Medicine and Epidemiology of Boston University School of Medicine, Boston, Massachusetts (JMM, WBK, JCE). The Framingham Study is supported through NIH/NHLBI Contract NOl-HC-38038. Requests for reprints should be addressed to Joanne M. Murabito, M.D., Framingham Heart Study, 5 Thurber Street, Framingham. Massa- chusetts01701. Manuscript submitted March 30. 1990, and accepted in revised form June 14. 1990. C hest discomfort is a commonly reported com- plaint in the outpatient setting, where estab- lishing a definitive diagnosis is difficult. Prognosis in patients with clinically determined angina based on complaints alone has been well studied [l-4]. However, observations of subjects with other types of chest discomfort have been limited to reports correlating type of chest discomfort with angio- graphically documented coronary artery disease [5- 10]. In addition to the character of the chest dis- comfort reported, age and sex have proven useful in predicting the probability of anatomic coronary disease [&lo]. In these reports, subjects were se- lected for coronary arteriography and may not be representative of the general population with re- gard to disease presentation or severity. The inci- dence of coronary heart disease (CHD) events by chest discomfort presentation has rarely been re- ported [11,12] and has not been studied in a general population setting. The aim of the current study was to determine the incidence of CHD events in subjects of the Fra- mingham Heart Study reporting new chest discom- fort, to analyze patterns of conversion among types of chest discomfort, to examine the validity of the clinical determination of angina, and to determine whether chest discomfort presentation offers prog- nostic information beyond that provided by tradi- tional CHD risk factors. PATIENTS AND METHODS Study Population The Framingham Heart Study was initiated in 1948 when a sample of residents of Framingham, Massachusetts, between the ages of 30 and 62 years was selected to undergo follow-up examination every 2 years to identify determinants of cardiovas- cular disease [13,14]. Surviving original cohort sub- jects of the Framingham Study attending any fol- low-up biennial examination(s) from examination cycles 9 through 17, which occurred from 1964 to 1984, were studied. Subjects with a diagnosis of an- gina, coronary insufficiency, or myocardial infarc- tion prior to examination cycle 9 were excluded. A diagnosis of coronary insufficiency was assigned when an episode of prolonged chest discomfort was September 1990 The American Journal of Medicine Volume 89 297