We sought to determine the predictive ability of total white blood cell (WBC) count and its subtypes for risk of death or myocardial infarction (MI).

An elevated WBC count has been associated with cardiovascular risk, but which leukocyte subtypes carry this risk is uncertain.

Consecutive patients without acute MI who were assessed angiographically for coronary artery disease (CAD) and were followed up long-term were studied. The predictive ability for death/MI of quartile (Q) 4 versus Q1 total WBC, neutrophil (N), lymphocyte (L), and monocyte (M) counts and N/L ratio were assessed using Cox regressions.

A total of 3,227 patients was studied. Mean age was 63 years; 63% of patients were male, and 65% had CAD. In multivariable modeling entering standard risk factors, presentation, and CAD severity, the total WBC (hazard ratio [HR] 1.4, p = 0.01) and M (HR 1.3, p < 0.02) were weaker and N (HR 1.8, p < 0.001), L (HR 0.51, p < 0.001), and N/L ratio (HR 2.2, p < 0.001) were independent predictors of death/MI. When WBC variables were entered together, N/L ratio and M were retained as independent predictors. Risk associations persisted in analyses restricted to CAD patients or including acute MI patients.

Total WBC count is confirmed to be an independent predictor of death/MI in patients with or at high risk for CAD, but greater predictive ability is provided by high N (Q4 >6.6 × 10³/μl) or low L counts. The greatest risk prediction is given by the N/L ratio, with Q4 versus Q1 (>4.71 versus <1.96) increasing the hazard 2.2-fold. These findings have important implications for CAD risk assessment.