BACKGROUND: The aim of this study was to build risk charts for the assessment of cardiovascular mortality of the CUORE project, an Italian longitudinal study, and to compare them with the systematic coronary risk evaluation (SCORE) project charts for low risk European countries. DESIGN: Random population samples enrolled in the 1980s and 1990s in Italy were included in the analysis: 7,520 men and 13,127 women aged 35-69 years without previous cardiovascular events and with a mean follow-up period of 10 years for cardiovascular disease. ICD-9 codes of death certificates similar to those of the SCORE project were considered when they appear as first cause of death. METHODS: Sex-stratified Cox proportional hazard model including age, systolic blood pressure, ratio between total and HDL cholesterol, and smoking habit as risk factors was used to assess cardiovascular mortality. RESULTS: Analysis showed that all risk factors included in the model were statistically significant. The corresponding area under the receiver operating characteristic curve was 0.825 (95% confidence interval: 0.803-0.846) for men and 0.850 (0.823-0.877) for women. The CUORE project charts yielded similar results to the corresponding charts of the SCORE project: Lin's coefficient was 0.929 for men and 0.935 for women. CONCLUSION: The comparison between CUORE and SCORE mortality risk charts shows that SCORE charts reflect quite well the Italian cardiovascular mortality and, correspondingly, Italian cohorts of the CUORE project are quite representative of European countries at low risk for cardiovascular mortality.
BACKGROUND: The aim of this study was to build risk charts for the assessment of cardiovascular mortality of the CUORE project, an Italian longitudinal study, and to compare them with the systematic coronary risk evaluation (SCORE) project charts for low risk European countries. DESIGN: Random population samples enrolled in the 1980s and 1990s in Italy were included in the analysis: 7,520 men and 13,127 women aged 35-69 years without previous cardiovascular events and with a mean follow-up period of 10 years for cardiovascular disease. ICD-9 codes of death certificates similar to those of the SCORE project were considered when they appear as first cause of death. METHODS: Sex-stratified Cox proportional hazard model including age, systolic blood pressure, ratio between total and HDL cholesterol, and smoking habit as risk factors was used to assess cardiovascular mortality. RESULTS: Analysis showed that all risk factors included in the model were statistically significant. The corresponding area under the receiver operating characteristic curve was 0.825 (95% confidence interval: 0.803-0.846) for men and 0.850 (0.823-0.877) for women. The CUORE project charts yielded similar results to the corresponding charts of the SCORE project: Lin's coefficient was 0.929 for men and 0.935 for women. CONCLUSION: The comparison between CUORE and SCORE mortality risk charts shows that SCORE charts reflect quite well the Italian cardiovascular mortality and, correspondingly, Italian cohorts of the CUORE project are quite representative of European countries at low risk for cardiovascular mortality.