Abstract
INTRODUCTION: Plasma D-dimer is a useful clinical test for acute venous thromboembolism (VTE), and concentrations remain higher in VTE patients after treatment than in controls. Yet, evidence is limited on whether higher basal D-dimer concentrations in the general population are associated with greater risk of first VTE. OBJECTIVE: To assess the prospective association between D-dimer and incident VTE over a long follow-up. METHODS: We measured plasma D-dimer in 12,097 participants, initially free of VTE, in the Atherosclerosis Risk in Communities Study. Over a median follow-up of 17years, we identified 521 VTEs. We calculated hazard ratios of VTE using proportional hazards regression. RESULTS: The age, race, and sex adjusted hazard ratios of VTE across quintiles of D-dimer were 1, 1.5, 1.8, 2.1, and 3.2 (p for trend <0.0001). For the first 10years of follow-up, the hazard ratio for the highest versus lowest quintile was 3.5, and was 2.9 after 10years. In both whites and African Americans, VTE risk remained strongly associated with D-dimer after further adjustment for diabetes, body mass index, kidney function, and several thrombophilia genetic markers. D-dimer was associated with both unprovoked and provoked VTE, but more strongly with unprovoked. CONCLUSIONS: A higher basal level of plasma D-dimer in the general population, presumably reflecting a predisposition to thrombosis, is a strong, long-term risk factor for a first VTE.