The biological variation represents the physiological variability of the concentrations of a measurand in the bio logical fluid of interest, for example, serum, plasma, whole blood, urine, or others. Two types of biological variation are recognized: intraindividual (concentration variations within the same individual) and interindividual (dispersion of average concentration values among different individuals belonging to a population with defined characteristics).

Intraindividual Biological Variation (CVI)

CVI is the random fluctuation of a constituent of the organ ism, measured at different times in the same individual, around its homeostatic point. The extent of this variation depends on the conditions that underlie the control of the concentrations of a given analyte in the biological fluid of interest. If the analyte levels are critical for the proper functioning of the organism as a whole (e.g., sodium, whose variations in plasma lead to a shift of water from the intracellular to the extracellular compartment, with potential problems in the central nervous system), the homeostatic control will be very tight, and the concentrations of the analyte will fluctuate within a narrow range. Conversely, if the measure and level does not have specific functions in the evaluated biological fluid (e.g., the enzymes indicating cytolysis) or its variations in the fluid are related to tissue needs or dietary intake (e.g., the plasma concentrations of iron or triglycerides), the intraindividual variation will be higher. The CVI is therefore characteristic of a specific measurand (analyte in a specific biological fluid), although there may be significant differences between individuals.

Interindividual Biological Variation (CVG)

CVG is defined as the difference in the average results of the same constituent obtained in different individuals, all under the same physiological conditions, due to the diversity of their homeostatic points. It represents the variability due to the different characteristics of the individuals belonging to a specific population. The extent of this variability is represented by the width of the range of values that a given measurand can assume in a population of appropriately selected subjects with criteria similar to those employed for defining a “reference population”.

An example of the two types of biological variation is presented in Fig1, showing serum creatinine values obtained in 90 subjects without renal dysfunction, whose blood samples were drawn once a week for 10 consecutive weeks. It is noteworthy that, while the CVI in different individuals is narrow, medians of the various subject results are dispersed over a wide range, with different values between males and females. The case of creatinine, presented in Fig.  1, is typical for a measurand that has sex-specific differences and a very low ratio between CVI and CVG, that is, an individuality index [II] equal to ∼0.3. In this situation, when the CVI is lower than the CVG, the population-based reference intervals lose their sensitivity in identifying individuals who may already present variations indicating possible renal dysfunction. In Fig.  2, the individual with the lower mean creatinine (∼65 μmol/L) will almost have to double his/her creatinine value to overcome the upper limit of the reference interval (104 μmol/L), which for this individual may already imply a relevant reduction of renal function.

Fig1. Biological variability of creatinine in serum of 90 subjects with physiological renal function, subjected to sampling once a week for 10 consecutive weeks. The data are divided by gender and, within the same group, sorted by increasing age (from 20 to 69 years for women and from 20 to 59 years for men). The horizontal line indicates the median, and the vertical line indicates the range of dispersion of the values (maximum-minimum) of each subject

Fig2. Example of the impact of biological variability on the interpretation of the creatinine level. The intraindividual biological variability is much lower than the population reference range (in other words, the interindividual biological variability). Consequently, variations in the serum creatinine concentration in a subject, evaluated by the estimated critical difference (~12%), can signal a change in glomerular filtration even when the results are still within the reference range of the population

Knowing the biological variation of the various measurands is important because, in addition to being one of the models for the definition of APS (see below), it allows one to estimate the reference change value (RCV), that is, the reference to statistically evaluate the significance of the variation between two consecutive measurements on the same individual beyond the total variation of the measurement. The RCV estimate allows one to establish if two consecutive measurements of the same measurand in the same subject can be considered different from the statistical point of view (with a given level of probability) from the variation only due to analytical and biological sources.

The RCV is calculated as follows:

In the formula, the total variation (which derives from the square root of the sum of the analytical variance and the intra-individual biological variance) is multiplied by the factor 2.77, which is obtained from 1.96 (the statistical fac tor to estimate the 95% probability) multiplied  (because two measures are compared). From the formula, it appears that the smaller the analytical variation, the smaller the RCV is. It is also clear that if the intraindividual biological variation is high, the RCV value will be larger and, therefore, it will be more difficult to assess the actual clinical significance of a result variation. A limitation of the use of RCV is that the available biological variation data are usually obtained as a mean of a group of individuals, but the bio logical variation of the individual on which the measurements were performed and the RCV employed could be significantly different than the average of the data used to derive RCV.