Prediction of the risk of cardiovascular mortality using a score that includes glucose as a risk factor. The DECODE Study

Abstract

Aims/hypothesis. Risk scores have been developed to predict cardiovascular or coronary risk, and while most have included diabetes as a risk factor, none have included lower glucose concentrations, either at fasting or following a 2-h oral glucose tolerance test. This article develops 5- and 10-year risk scores for cardiovascular mortality that include glucose concentrations as well as known diabetes status. Methods. Data is from the DECODE cohort: 16,506 men and 8,907 women from 14 European studies. The risk factors studied were as follows: age, fasting and 2-h glucose (including cases of known diabetes), fasting glucose alone (including cases of known diabetes), cholesterol, smoking status, systolic blood pressure and BMI. For an absolute risk score the 1995 country-and sex-specific cardiovascular death rates were used. Results. In men, for both 5- and 10-year cardiovascular mortality, after adjusting for age and study centre, all studied risk factors, except BMI, were significantly associated with cardiovascular mortality (p<0.05). These results were unchanged in multivariate models with all factors included. In women, after adjusting for age and centre, glucose categories, systolic blood pressure and BMI were predictive of 5-year cardiovascular mortality. With all factors in the model, only age and glucose categories were predictive. In terms of 10-year cardiovascular mortality, smoking status and blood pressures were also predictive in the women. For men and women, the same scores were used for the risk factors, except for age and glucose categories where the hazard ratios differed significantly. Conclusions/interpretation. Including glucose concentrations as well as diabetic status provides quantitative information on cardiovascular risk prediction. © Springer-Verlag 2004.