Determinants of 40-year all-cause mortality in the European cohorts of the Seven Countries Study

If a few risk factors had predictive power for all-cause mortality in different geographical-cultural areas, then preventive efforts might be concentrated on these. Thirteen potential risk factors were measured in 6,554 men aged 40–59 around 1960 in Northern, Southern and Eastern European areas of the Seven Countries Study. In 40 years 85.3% of men died in the pooled areas (87.9, 81.8 and 87.9% in Northern, Southern and Eastern Europe, respectively). Six risk factors were significant predictors of events in all three areas: directly for age, smoking habits, mean blood pressure, heart rate and ECG abnormalities; inversely for forced expiratory volume. In a pooled model also father and mother life status, socio-economic status, and arm circumference (the last one in an inverse way) had significant coefficients that were not heterogeneous across areas (except for socio-economic status). Serum cholesterol was around significance. ROC curves had values of 0.833, 0.806 and 0.819 respectively in Northern, Southern and Eastern Europe, and 0.827 in the pooled areas. Correlation coefficients between observed and expected cases in deciles of estimated risk were between 0.98 and 0.99. Survivors after 40 years in the lower half of the estimated risk were 10.7, 23.6 and 13.3% in Northern, Southern and Eastern Europe, respectively. Under- or over-estimate of cross-applying risk functions did not exceed 15%. All-cause mortality and survival in middle aged men during 40 years were strongly associated with a few, mainly cardiovascular, risk factors, whose predictive power was similar in different cultures across Europe.