ARIC hemostasis study--IV. Intraindividual variability and reliability of hemostatic factors. The Atherosclerosis Risk in Communities (ARIC).

We have recently reported the short-term intraindividual variability of several coagulation factors and inhibitors included in the ARIC study (Chambless et al. Ann Epidemiol 1992; 2:723). In this paper, we reported the intraindividual variability results of additional hemostatic factors. Blood samples were collected for hemostatic assays three times at 1-2-week intervals from 39 subjects recruited from 4 ARIC field centers. The contributions of within-person, processing and assay (designated "method") and between-person variances to the total variance were estimated and from them the reliability coefficient, R, was computed as the proportion of total variance in the between-person component. The R value was high for beta-thromboglobulin and tissue-plasminogen activator: 0.83 and 0.81, respectively; and intermediate for D-dimer and plasminogen activator inhibitor-1: 0.73 and 0.72, respectively. Protein S (total and free) and platelet factor 4 had low repeatability (R < 0.50) derived mostly from "method" variability while low R value (0.03) for fibrinopeptide A was attributed to high "method" and "within-person" variability. Gender, age and the level of hemostatic factors did not influence the intraindividual variability.