Keys A, Menotti A, Aravanis C, Blackburn H, Djordevic B S, Buzina R, Dontas A S, Fidanza F, Karvonen M J, Kimura N
Prev Med. 1984 Mar;13(2):141-54. doi: 10.1016/0091-7435(84)90047-1.
Among 11,579 men ages 40-59 without evidence of cardiovascular disease, 2,289 died in 15 years, 618 from coronary heart disease. The 15 cohorts in seven countries (four regions) differed in all-causes death rate, mainly reflecting great differences in coronary mortality. Among characteristics of entry, only mean blood pressure helped to explain cohort differences in all-causes death rate. Three-quarters of the variance in coronary death rate was accounted for by differences in mean serum cholesterol and blood pressure of the cohorts. The mortality risk for individuals was examined in each of the regions. For coronary death, age, serum cholesterol, blood pressure, and smoking were highly significant in all regions except Japan, where coronary deaths were too few for evaluation. Relative weight was not significant anywhere. Physical activity was significant only in southern Europe, where differences are associated with socioeconomic status. For all-causes death, age and blood pressure were highly significant risk factors in all regions as was smoking habit, except in Japan. Relative body weight tended to be a negative risk factor everywhere, significantly so in southern Europe. Expectations for coronary death from the experience in the United States and northern Europe greatly exceeded observed deaths in southern Europe for men of their age, serum cholesterol, blood pressure, smoking habits, physical activity, and relative weight. The reverse, prediction of coronary deaths in America and in northern Europe from the southern European experience, greatly underestimated the deaths observed. Similar cross-predictions between the United States and northern Europe were good for all-causes deaths, excellent for coronary deaths. Analysis of time trends in relationships of mortality to entry characteristics showed continued importance of age, blood pressure, and smoking and a tendency for the importance of cholesterol to fall in the last 5 years of follow-up.
在11579名年龄在40至59岁且无心血管疾病证据的男性中，有2289人在15年内死亡，其中618人死于冠心病。七个国家（四个地区）的15个队列在全因死亡率上存在差异，主要反映在冠心病死亡率上的巨大差异。在入组特征中，只有平均血压有助于解释队列在全因死亡率上的差异。队列中平均血清胆固醇和血压的差异解释了四分之三的冠心病死亡率方差。对每个地区的个体死亡风险进行了检查。对于冠心病死亡，年龄、血清胆固醇、血压和吸烟在所有地区都具有高度显著性，除了日本，在日本冠心病死亡人数太少无法进行评估。相对体重在任何地方都不显著。身体活动仅在南欧显著，那里的差异与社会经济地位有关。对于全因死亡，年龄和血压在所有地区都是高度显著的风险因素，吸烟习惯也是如此，除了日本。相对体重在各地往往是一个负风险因素，在南欧尤为显著。根据美国和北欧的经验对冠心病死亡的预期大大超过了南欧同年龄、血清胆固醇、血压、吸烟习惯、身体活动和相对体重的男性的实际观察死亡人数。反之，根据南欧的经验对美国和北欧冠心病死亡的预测则大大低估了实际观察到的死亡人数。美国和北欧之间类似的交叉预测对于全因死亡情况良好，对于冠心病死亡情况极佳。对死亡率与入组特征关系的时间趋势分析表明，年龄、血压和吸烟仍然具有持续的重要性，并且在随访的最后5年中胆固醇的重要性有下降趋势。
