i have been seeing the nature thymic involution paper passed around as if a “healthy thymus on CT” is a new biomarker you can chase. it is not, at least not yet, and the way people are reading it is going to lead to some bad self-experiments. the myth: thymic fat fraction on CT, scored by their model, tells you your immune age and you can act on it. the reality: the model quantifies thymic tissue density across a large retrospective cohort and finds associations with downstream outcomes. association, in a population, adjusted for the covariates they had access to. that is a very different claim than “your individual score is actionable.” what is actually missing before anyone titrates anything off this: - no intervention arm. nothing in the paper shows that moving your score moves your outcome. - the cohort skews toward people who already had a CT for some clinical reason.
that is not a healthy adult baseline, it is a selected one. - thymic tissue on imaging is not the same as thymic output. naive T cell production is measured with TRECs, and the paper does not bridge the two cleanly enough to swap one for the other. - effect sizes look meaningful at the population level and small at the individual level. easy to confuse those. if you want to track something useful in your own data while this matures, a yearly CBC with differential and a lymphocyte subset panel will tell you more about your actual immune status than any imaging score will for the next few years. write the dates down.