Risk of coronary events by baseline factors during 28 years follow-up and three periods in a random population sample of men.

AIMS

To investigate the predictive value of risk factors for coronary events measured in midlife during three separate periods over a follow-up period extending through 28 years.

METHODS

A total of 7437 men aged 47-55 years and free of myocardial infarction at baseline were examined. Risk of coronary events (nonfatal myocardial infarction and coronary deaths) was analysed for the entire period and for 0-15, 16-21 and 22-28 years' follow-up, using age-adjusted and multiple Cox regression analyses.

RESULTS

Age, diabetes, elevated blood pressure and serum cholesterol were all independently associated with increased risk of coronary events for the entire 28 years as well as for each of the periods. A family history of coronary events amongst fathers, mothers and siblings was independently significant for the entire follow-up period, and the risk did not decline with extended follow-up. Effort-related chest pain was a strong and independent risk factor for the first 21 years but not thereafter. The importance of smoking decreased over time and was not significantly associated with outcome during the last period. Stress was also significant for the entire 28 years, but in selected periods only up to 21 years. Body mass index, low physical activity and low social class were inconsistently or not at all related to outcome in multiple analyses.