Long-term mortality after a cardiovascular risk reduction program
ACP J Club. 1992 Jan-Feb;116:1. doi:10.7326/ACPJC-1992-116-1-001
Strandberg TE, Salomaa VV, Naukkarinen VA, et al. Long-term mortality after 5-year multifactorial primary prevention of cardiovascular diseases in middle-aged men. JAMA. 1991 Sep 4;266:1225-9.
To investigate the effects of a 5-year program of reduction of cardiovascular risk factors on 15-year mortality of middle-aged men.
10-year follow-up of participants after a randomized controlled trial.
Community study in Finland.
1222 healthy businessmen, born between 1919 and 1934. Each had ≥ 1 cardiovascular risk factor despite previous health education.
612 men were randomized to interventions for primary prevention of cardiovascular disease. Education and medications were provided at visits every 4 months to attempt to modify lipid levels, blood pressure, weight, fasting blood glucose, and tobacco and alcohol consumption. Drugs used included propranolol, pindolol, hydrochlorothiazide, probucol, and clofibrate. 610 men received no study intervention. After the 5-year assessment the men were advised to continue health checkups. Vital status was established for all participants in 1989.
Main outcome measures
Cumulative 15-year, all-cause mortality. Causes of death were determined by blinded review of death certificates without confirmation from other sources.
At the end of the 5-year intervention period, risk factors were significantly lowered in the intervention group compared with the control group. 5 years later, risk factors and medication were similar for both groups.
The 15-year cumulative number of deaths was 67 in the intervention group (10.9%) and 46 in the control group (7.5%): relative risk (RR), 1.45 (95% CI, 1.01 to 2.08; P = 0.048). In the intervention and control groups, 5.6% (34 deaths) compared with 2.3% (14 deaths), respectively, were attributed to cardiac causes (P = 0.004). There were more deaths attributed to suicide and accidents in the intervention group (13 deaths) than the control group (1 death; P = 0.002). There were no significant differences in deaths from cancer or stroke.
Primary prevention interventions used in the late 1970s to lower risk factors for cardiovascular disease in middle-aged businessmen were associated with unexplained excess mortality over a 15-year period.
Sources of funding: Sydäntutkimussäätiö; Suomen Kulttuurirahaston Kymenlaakson Rahasto; Kymenlaakson Terveyden Turva r.y.
Address for article reprint: Dr. T.E. Strandberg, Second Department of Medicine, University of Helsinki, SF-00290 Helsinki, Finland.
The initial report of this trial in 1985 showed a slight increase in the 5-year coronary incidence in the treated group, and the authors pointed to potential adverse effects of clofibrate and pindolol, a β-blocker with intrinsic sympathomimetic activity (ISA). In the WHO trial, clofibrate was associated with an excess mortality, and several randomized trials of secondary prevention using β-blockers with ISA were stopped early because of increased mortality in the treated group.
The findings from an additional 10 years of follow-up are difficult to explain. Even though post-trial levels of risk factors for intervention and control groups largely merged, the groups' mortality curves continued to diverge 10 years after the trial.
Recommendations for practice must be based on the bulk of the evidence. Although the pharmacologic treatment of high blood pressure with diuretics and non-ISA β-blockers has had less effect on the incidence of coronary disease than expected, the efficacy of treatment against stroke is clear (1). For cholesterol lowering (2), the primary prevention trials reduced coronary mortality (RR, 0.85; CI, 0.69 to 1.05) although not total mortality (RR, 1.07; CI, 0.94 to 1.21).
The effect of pharmacologic treatment on risk factors for coronary disease may be modest; selected agents may be associated with adverse events; and excessive lowering of diastolic blood pressure may increase the risk for coronary events (3). When pharmacologic therapy is appropriate, the randomized trials using major disease end points as their outcome suggest a benefit from the use of diuretics and non-ISA β-blockers for hypertension, and niacin, cholestyramine, and gemfibrozil for hypercholesterolemia.
Bruce M. Psaty, MD, PhD
David S. Siscovick, MD, MPH University of Washington Seattle, Washington