Short-term and long-term variability of plasma homocysteine measurement.

Within-person and methodological variability of a given analyte are important elements in determining whether an individual has altered concentrations of that analyte. We report the short-term (1 month) within-person, between-person, and methodological variability of plasma homocysteine in 20 healthy participants from whom samples were drawn weekly for 4 weeks. The short-term between-person variance was high, whereas within-person and methodological variances were relatively very low, giving a high reliability coefficient (R) for homocysteine (R = 0.94). The long-term (30 months) reliability coefficient was 0.65, but was greatly influenced by an outlier (R = 0.82 with the outlier excluded). The data suggest that an individual's plasma homocysteine concentration is relatively constant over at least 1 month, and a single measurement characterizes the average concentration reasonably well.