Intra-individual variability of high-sensitivity C-reactive protein - Age-related variation over time in Japanese subjects

Background Prospective studies have demonstrated a direct association between the baseline level of high-sensitivity C-reactive protein (hsCRP) and the risk of developing cardiovascular disease. The potential for hsCRP variability during the follow-up interval is of great interest, so the repeatability of measurements of hsCRP in men and women, and in different age groups, was investigated in the present study. Methods and Results The measurement of hsCRP was performed in 899 male and 780 female Japanese civil servants aged 18-60 years in the spring of 2004 and 2005. Within-subject repeatability and intra-class correlation coefficient of reliability were used to characterize the variation in measurements. Error of the measurement process was negligible, as the analytical variance was 0.0003 [ln(mg/L)] [In(mg/L)] with a reliability coefficient of 0.997. In contrast, the within-subject variation of hsCRP was considerable with a reliability coefficient of 0.61 in both sexes and repeatability coefficients of 5.25 and 5.38 in men and women, respectively. The variation differed with age, especially in women, as less variation was observed in subjects over 40 years of age during the follow-up interval. Conclusions The results suggest that a single measurement of hsCRP is subject to considerable but equal within-subject variation over time in men and women, and this variation differs among the age groups, which may affect the risk assessment of hsCRP in association with future cardiovascular events.