Research Article

Relation of low diastolic blood pressure to coronary heart disease death in presence of myocardial infarction: the Framingham Study.

BMJ 1991; 303 doi: (Published 17 August 1991) Cite this as: BMJ 1991;303:385
  1. R B D'Agostino,
  2. A J Belanger,
  3. W B Kannel,
  4. J M Cruickshank
  1. Department of Mathematics, Boston University, College of Liberal Arts, MA 02215.


    OBJECTIVE--To examine the hypothesis that a J curve relation between blood pressure and death from coronary heart disease is confined to high risk subjects with myocardial infarction. DESIGN--Cohort longitudinal epidemiological study with biennial examinations since 1950. SETTING--Framingham, Massachusetts, USA. SUBJECTS--5209 subjects in the Framingham study cohort followed up by a person examination approach. MAIN OUTCOME MEASURES--Coronary heart disease deaths and non-cardiovascular disease deaths in men and women with or without myocardial infarction relative to blood pressure. RESULTS--Among subjects without myocardial infarction non-cardiovascular disease deaths were twice to three times as common as coronary heart disease deaths. Furthermore, there was no significant relation between non-cardiovascular disease death and diastolic or systolic blood pressure. Also coronary heart disease deaths were linearly related to diastolic and systolic blood pressures. Among high risk patients (that is, people with myocardial infarction but free of congestive heart failure) death from coronary heart disease was more common than non-cardiovascular disease death. There was a significant U shaped relation between coronary heart disease death and diastolic blood pressure. Although there was an apparent U shaped relation between coronary heart disease death and systolic blood pressure, it did not attain statistical significance when controlling for age and change in systolic blood pressure from the pre-myocardial infarction level. None of the above conclusions changed when adjustments were made for risk factors such as serum cholesterol concentration, antihypertensive treatment, and left ventricular function. The U shaped relation between diastolic blood pressure and high risk subjects existed for both those given antihypertensive treatment and those not. CONCLUSIONS--These data suggest that an age and sex independent U curve relation exists for diastolic blood pressure and coronary heart disease deaths in patients with myocardial infarction but not for low risk subjects without myocardial infarction. The relation seems to be independent of left ventricular function and antihypertensive treatment.