An epidemiological perspective of sudden death. 26-year follow-up in the Framingham Study

Abstract

Over 26 years of follow-up of 5209 Framingham subjects there were 147 sudden deaths in men and 50 in women. Half the sudden deaths in men and 70% in women occurred without prior coronary heart disease. The incidence doubled with each decade of age, with women lagging men by 20 years. At any age in either sex the risk varied widely in relation to their risk factor make-up. By incorporating systolic pressure, serum cholesterol, vital capacity, cigarette smoking, relative weight, heart rate and ECG abnormalities into a multivariate formulation, a composite estimate of risk was obtained over a wide range. The multivariate sudden death prediction was as efficient in women as in men, identifying 53% and 42% of the sudden deaths, respectively, in the upper decile of multivariate risk. At any level of multivariate risk, women had only a third of the sudden death incidence of men, suggesting a unique biological resistance to sudden death. Ventricular premature beats, associated with an excess sudden death incidence, often occurred concurrently with ECG evidence of left ventricular hypertrophy, intraventricular block, and nonspecific ECG abnormalities. These associated ECG abnormalities were more highly predictive of sudden death than ventricular premature beats. The key to prevention of sudden death would appear to be reduction of the risk of coronary attacks. Once overt coronary heart disease or cardiac failure appears, the risk of sudden death escalates 9-fold, and at this stage the standard coronary heart disease risk factors have little predictive value. Only ECG abnormalities, arrhythmia and poor cardiac function appear to be predictive.