Age-dependent normalisation functions for T-lymphocytes in healthy individuals

Lymphocyte numbers naturally change through age. Normalisation functions to account for this are sparse, and mostly disregard measurements from children in which these changes are most prominent. In this study, we analyse cross-sectional numbers of mainly T-lymphocytes (CD3+, CD3+CD4+ and CD3+CD8+) and their subpopulations (naive and memory) from 673 healthy Dutch individuals ranging from infancy to adulthood (0-62 years). We fitted the data by a delayed exponential function and received parameter estimates for each lymphocyte subset. Our modelling approach follows general laboratory measurement procedures in which absolute cell counts of T-lymphocyte subsets are calculated from observed percentages within a reference population that is truly counted (typically the total lymphocyte count). Consequently, we receive one set of parameter estimates per T-cell subset representing both the trajectories of their counts and percentages. We allow for an initial time delay of half a year before the total lymphocyte counts per {micro}l of blood start to change exponentially, and we find that T-lymphocyte trajectories tend to increase during the first half a year of life. Thus, our study provides functions describing the general trajectories of T-lymphocyte counts and percentages of the Dutch population. These functions provide important references to study T-lymphocyte dynamics in disease, and allow one to quantify losses and gains in longitudinal data, such as the CD4+ T-cell decline in HIV-infected children, and/or the rate of T-cell recovery after the onset of treatment.