Abstract

Seventeen thousand three hundred and fifty four mildly hypertensive people with diastolic blood pressures between 90 and 109 mm Hg at screening were randomised to active treatment, with bendrofluazide or propranolol, or to placebo tablets. They were followed for a maximum of five and a half years, giving a total of 85,572 patient-years of observation. There were 456 myocardial infarctions or sudden coronary deaths. Drug treatment did not affect the overall rate of coronary events. Rates per thousand person-years were 8.3 and 9.0 in men and 1.8 and 1.7 in women in the active treatment and placebo groups respectively. Event rates were much higher in smokers than in non-smokers on placebo treatment (12.6 and 7.5 in men and 3.5 and 1.0 in women in smokers and non-smokers respectively). An analysis of subgroup results showed a lower event rate in non-smoking men on propranolol than in non-smokers on placebo (5.0 and 7.5 per thousand person-years respectively). Bendrofluazide had no apparent effect on the event rate. The interaction between the type of treatment (propranolol, bendrofluazide, or placebo) and smoking in determining the coronary event rate was not statistically significant, however. The incidence of electrocardiographic changes of silent infarction--that is major Q/QS abnormalities--differed little with sex, smoking habit, or treatment with either active drug.