Elevated D-dimer levels are reportedly associated with coronary artery disease. It was the study objective to investigate the association of baseline D-dimer levels with strokes that occurred in the European Prospective Investigation into Cancer and Nutrition-Italy cohort. Using a nested case-cohort design, a centre--stratified sample of 832 subjects (66 % women, age 35-71) was selected as subcohort and compared with 289 strokes in a mean follow-up of nine years. D-dimers were measured by an automated latex-enhanced immunoassay (HemosIL-IL). The multivariable hazard ratios were estimated by a Cox regression model using Prentice method. Individuals with elevated D-dimer levels had significantly higher risk of incident stroke. It was evident from the second quartile (D-dimers > 100 ng/ml) and persisted almost unchanged for higher D-dimers (hazard ratio [HR] 2.10, 95 % confidence interval [CI]: 1.28-3.47; 2.42, 95 %CI: 1.44-4.09 and 2.10, 95 %CI: 1.27-3.48 for the second, third or fourth quartile compared with the lowest quartile, respectively). The association was independent of several confounders, including triglycerides and C-reactive protein. No differences were observed in men and women (P for interaction=0.46), in hypertensive or non-hypertensive subjects (P for interaction=0.88) or in subjects with low (< 1 mg/l) or elevated (≥ 1 mg/l) C-reactive protein (P for interaction=0.35). After stratification for stroke type, the hazard ratio for every standard deviation increase was statistically significant both for ischaemic (1.21; 95 %CI: 1.01 to 1.45) and haemorrhagic (1.24; 95 %CI: 1.00 to 1.65) strokes. In conclusion, our data provide clear evidence that elevated levels of D-dimers are potential risk factors not only for ischaemic but also for haemorrhagic strokes.
Keywords: Stroke / prevention; cerebrovascular disease; coagulation factors; epidemiological studies; risk factors.