Background: Despite a 5-year multifactorial cardiovascular disease (CVD) risk reduction in an intervention trial, mortality was unexpectedly higher in the intervention than the control group during the 15-year follow-up.
Purpose: In order to find explanations for the adverse outcome, we have extended mortality follow-up to 40 years. We have also re-examined baseline characteristics that contributed to total mortality.
Methods: The prevention period (1974–1980) included 1,222 originally healthy men (born 1919–1934) at high CVD risk, who were randomized into intervention (n=612) and control groups (n=610). The 5-year multifactorial intervention consisted of personal health education (diet, exercise, weight reduction, antismoking) and contemporary drug treatments for dyslipidemia and hypertension. In the present analysis we used previously unpublished data on baseline risk factors and lifestyle characteristics. Participants were followed-up for 40-year mortality using nation-wide death registers.
Results: The study groups were similar at baseline in 1974, and the 5-year intervention significantly improved CVD risk factors (body mass index, blood pressure, serum lipids and glucose), and total CVD risk by 46% in the intervention group. Despite this, total mortality was consistently higher in the intervention group than the control group during 25 years post-trial, but converged thereafter. Increased mortality was caused by CVD and accidental deaths. Specific factors, like drug treatments, did not explain increased mortality. Instead, of the newly-analyzed baseline factors, there was a significant interaction for mortality between intervention group and yearly vacation time (P=0.027): shorter vacation was associated with excess 30-year mortality in the intervention (hazard ratio 1.37, 95% CI 1.03–1.83, P=0.03), but not in the control group (P=0.5). This finding was robust to multivariable adjustments.
Conclusions: After a multifactorial intervention for healthy men with at least one CVD risk factor, there has been an unexpectedly increased mortality in the intervention group. This increase was especially observed in a subgroup characterized by shorter vacation time at baseline. Although this adverse response to personal preventive measures in vulnerable individuals may be characteristic to men of high social status with subclinical CVD, it clearly deserves further investigation.