Haemostatic/inflammatory markers predict 10-year risk of IHD at least as well as lipids: the Caerphilly collaborative studies

Abstract

Aims: We compare the predictive values of plasma lipids (total and HDL-cholesterol, triglycerides) and three haemostatic/inflammatory risk markers for subsequent ischaemic heart disease (IHD).

Methods and results: Two UK populations totalling 4860 men were screened for evidence of IHD between 1979 and 1983. Men were followed over 10 years and validated coronary events were recorded. Risk estimates were made using relative odds, receiver operating characteristic (ROC) curves and deciles of risk. Regression dilution effects were also examined. By 10 years, 525 men had a coronary event (fatal, non-fatal or silent myocardial infarction, MI). Two alternative multivariate models were compared - a lipid model (total, HDL-cholesterol, triglyceride) and a haemostatic/inflammatory model (fibrinogen, viscosity and white cell count). 'Correction' for regression dilution increased relative odds for most risk factors. In the distribution of predicted risk, using established risk factors in conjunction with either lipid or haemostatic/inflammatory factors, the deciles of risk analysis showed that the observed 10-year risk of IHD was 34-35% in men in the top tenth, compared to 2-3% in the lowest tenth of the distribution.

Conclusion: At the 10 years' follow-up, major, haemostatic/inflammatory risk factors showed a graded relationship to incident IHD that was at least as strong as that given by plasma lipids. Haemostatic/inflammatory factors provide possible additional targets for intervention.