Hendriksen I J, Bezemer P D, Boerma G J, Zuiderveld U
Vrije Universiteit, vakgroep Theorie der Geneeskunde, Epidemiologie en Biostatistiek, Amsterdam.
Ned Tijdschr Geneeskd. 1992 Aug 1;136(31):1507-11.
The blood cholesterol level is not constant over time. Changing values may be found when rechecking high values, especially in patients on a diet or receiving medication. Repeated measurements are necessary. It will then be possible to estimate the 'true mean value'. This is the theoretical average of a large number of measurements taken from one person. The study concentrated on the intra-individual variation of the serum cholesterol and the consequences for screening and follow-up. For this purpose, during a period of four weeks, cholesterol levels were measured 12 times in 33 men aged 25-40. The mean coefficient of variation was 5.7%, with wide differences between participants, ranging from 2.9% to 9.8%. The position of the 'true mean value' was estimated (with 90% confidence), after I resp. 3 determinations. These findings have consequences for the classification of subjects in the different risk categories as defined in the Dutch Cholesterol Consensus. It is also possible to determine if, after a period of intervention, there is a significant decline in the cholesterol level. Roughly, a decline of 10-12% indicates a significant difference.
血液胆固醇水平并非随时间恒定不变。复查高值时可能会发现数值变化，尤其是在节食或接受药物治疗的患者中。重复测量是必要的。这样就有可能估算出“真实均值”。这是对同一个人进行大量测量的理论平均值。该研究聚焦于血清胆固醇的个体内变异以及对筛查和随访的影响。为此，在四周时间内，对33名年龄在25至40岁的男性进行了12次胆固醇水平测量。平均变异系数为5.7%，参与者之间差异较大，范围在2.9%至9.8%之间。在进行1次及3次测定后，估算了“真实均值”的位置（置信度为90%）。这些发现对按照荷兰胆固醇共识所定义的不同风险类别对受试者进行分类有影响。也能够确定在一段时间的干预后胆固醇水平是否有显著下降。大致而言，下降10 - 12%表明存在显著差异。
