Immune parameters associated with mortality in the elderly are context-dependent: lessons from Sweden, Holland and Belgium.

The pioneering Swedish OCTO/NONA-Immune longitudinal studies led by Anders Wikby in Jönköping in the 1990s established a cluster of simple baseline immune parameters associated with excess mortality in 85 year-old non-institutionalized individuals over 2, 4 and 6-year follow-up. We dubbed this cluster the "Immune Risk Profile" (IRP) consisting of poor proliferative responses of peripheral blood mononuclear cells to T cell mitogens, accumulations of CD8+ CD28- T-cells resulting in an inverted CD4:8 ratio, decreased amounts of B-cells, and seropositivity for Cytomegalovirus (CMV). The concept of the IRP has since been applied by others to many different populations in different circumstances and at different ages, but in general without specifically establishing whether the same risk factors were relevant in the tested subjects. However, our own later studies showed that risk factors in aged populations from The Netherlands and Belgium were markedly different, indicating that the IRP cannot simply be transferred between populations. Moreover, there was a striking sex difference in the Belgian study, which was the only one large enough to include sufficient numbers of old men. The reasons for these marked differences between populations which one might have assumed a priori to be quite similar to one another are not clear, and many candidates can be speculated upon, but the important lesson is that there is a marked context-dependency of immune biomarkers of ageing, suggesting that IRPs cannot be assumed to be identical in different populations.