Subject-specific reference ranges
The reference range for many analytes, as mentioned previously, is commonly taken to be the middle 95% of values found in the population or in a stratified group based on age, gender, race etc. where these factors are known to influence the range of values found. The reference range depends on the intra-individual variation, the inter-individual variation, and the analytical imprecision. Unless the intra-individual variation is greater than the inter-individual variation, the reference range may be quite insensitive for detecting changes in a given subject. The formal comparison of these is known as the index of individuality, defined as the ratio of the CV(intra-individual)/CV(inter-individual). Values >1.4 suggest the reference range is sensitive for detecting changes in a subject, values <0.6 suggest it is not.
Maybe we should be paying more attention to intra-individual variation. A model for this is the study of hemoglobin concentrations in elite athletes. By regularly measuring their hemoglobin concentrations, it is possible to establish a subject-specific reference range that makes it easier to identify changes in hemoglobin concentration that may be associated with doping. A paper on this topic appeared some years ago.