Abstract
In a longitudinal study, we determined interindividual and intra-individual variation in 20 plasma proteins and lipids and in other blood constituents by analysis of variance. Blood from 20 healthy subjects was sampled monthly for six months, a rigorous blood-sampling technique being applied. The mean proportion of interindividual variation differed for each blood constituent, ranging from 22 to 91% of the total variation. The possible clinical application of individual reference intervals of this homeostatic model was demonstrated by the fact that they were exceeded in individual cases of upper respiratory tract infection. Concordance between individual reference intervals in healthy controls and in patients--as exemplified in two chronic diseases, multiple sclerosis and chronic inactive pyelonephritis--suggests that the use of individual intervals in (chronic) disease is valid, even when derived from healthy persons. Additionally, sex- and age-related differences were significant for some constituents.