C-reactive protein (CRP) has been identified in prospective epidemiological research as an independent risk factor for cardiovascular disease. Using this acute phase protein in risk prediction in individuals is still under debate. In this paper short-term biological variation of CRP is examined and a strategy to test the confidence in a single CRP sample is proposed. Data were obtained from three different groups of healthy volunteers: male white-collar workers (n=92; age range 35-55 years), female nurses who did not use oral contraceptives (OC; n=75; age range 23-55 years) and female nurses who used oral contraceptives (n=56; age range 22-49 years). Blood samples were obtained 3 times in men and 2 times in women during a normal workweek (identical rotating shifts in the nurses). Subjects had to be in good general health, show no signs of overt disease or infection, and free of medication at the time of blood withdrawal. All sampling was done between 7:00 and 9:30 a.m.; subjects had to fast after 11:00 p.m. and refrain from intense physical activity the preceding day. CRP was determined using polyclonal antibody against CRP, both for catching antibodies and for conjugated antibodies (DAKO, Denmark). The intra-assay and inter-assay coefficient of variation were 1.0 and 10.0% for CRP.

On average CRP values were highest in the OC using women (1.95 mg/l; range 0.14-5.58), followed by the men (1.18 mg/l; range 0.02-7.98) and lowest in the non-OC using women (0.82 mg/l; 0.05-3.78). The 3-day test-retest correlation was 0.73 for men, 0.80 for OC using women and 0.82 for non-OC using women. In men, the additional 5-day test-retest correlation was 0.60. Between subject variance accounted for 65.6 % of total variance in men, 74.5 % in OC users, and 75.8 % in non-OC users. Within subject variance accounted for 34.4 % in men, 25.5 % in OC users, and 24.2 % in non-OC users. CRP correlated highly with body mass index (BMI) and fibrinogen in all three groups and with t-PA activity and PAI-1 antigen in men and women, but less with lipids and insulin. Correlation of CRP to age was absent or weak in all groups.

Using the repeated measures, individual samples were identified that reflected a "suspected" high value, i.e. a value that was more than 2 standard deviations higher than the lowest value obtained from the same subject. Three strategies were then tested to identify these suspected high CRP values, given a study design with only one single blood sample. The first most simple criterion was to label all values above an arbitrary absolute threshold level as suspect. Secondly, we predicted an individual's CRP value from age, gender, OC use, BMI, smoking behaviour and alcohol consumption, and labelled all CRP values suspect that deviated more than 2 standard deviations from this prediction. Thirdly, we used a similar regression procedure to obtain a predicted CRP value, this time including fibrinogen and the full spectrum of metabolic syndrome X risk indicators (insulin, HDL-C, LDL-C, triglycerides, t-PA antigen, t-PA activity, and PAI-1 antigen), as obtained in the same blood sample as well. We conclude that the most pragmatic strategy, using an absolute level criterion to decide whether a second sample should be obtained from a patient, also yields the best results in terms of maximising detection of suspected values and minimising unnecessary resampling. However, the absolute criterion should take gender and OC use into account. For healthy subjects in the age range from 25-55 years we propose 1.75 mg/l for men, 2.00 mg/l for OC using women and, 1.00 mg/l for non-OC-using women.