Plasma fibrinogen has emerged as an important predictor of cardiovascular disease, but few data are available on its association with stroke. We sought to determine if plasma fibrinogen is a marker of increased risk or a direct causative risk factor for stroke.

Patients from the Bezafibrate Infarction Prevention Study, a placebo-controlled, randomized clinical trial of secondary prevention of coronary heart disease by lipid modification with bezafibrate retard (400 mg daily), were studied. Plasma fibrinogen levels were measured at baseline and yearly thereafter. Stroke, a prospectively monitored endpoint, was systematically assessed regarding stroke type, subtype, and functional outcome.

Mean baseline fibrinogen levels were significantly higher in patients subsequently having a cerebrovascular event (140 strokes, 36 transient ischemic attacks; mean follow-up, 6.2 years) than in patients who did not (375 vs. 349 mg/dL, P <0.0001). Fibrinogen levels did not differ significantly by the type, subtype, or severity of the cerebrovascular event. Risk of ischemic stroke increased from 3.3% in the lowest tertile (baseline fibrinogen <314 mg/dL) to 7.% in the middle tertile (fibrinogen 314 to 373 mg/dL) to 10% in the upper tertile (fibrinogen >373 mg/dL, P <0.001). Adjusting for age, blood pressure, and other covariates, fibrinogen levels in the upper tertile were associated with more than a twofold increase in risk of ischemic stroke compared with in the lowest tertile (hazard ratio = 2.6; 95% confidence interval: 1.5 to 4.3). We did not find fibrinogen change from baseline to be related to subsequent ischemic stroke events.

Plasma fibrinogen is a strong predictor of, rather than a direct causative factor for, subsequent stroke among patients at increased risk owing to manifest coronary heart disease.