Muscle stem cells in old organisms activate slowly, repair tissue sluggishly, and produce weaker regeneration than young ones. The standard interpretation: aging damages the cells, degrading their function. The implication: restore youthful function and you restore repair capacity.
Reader note (2026-11-16): The UCLA study (Science, 2026; Kang, Benjamin and colleagues per the release) is in mice, comparing muscle stem cells from young and old mice; this essay states it for old organisms generally and says 'at 80', which is the essay's own framing. The release matches the NDRG1 account (3.5 times higher in old cells, suppresses mTOR, blocking it made old cells behave like young ones and sped repair after injury) and adds that fewer stem cells survived over time, impairing regeneration after repeated injuries. 'Cellular survivorship bias' is the authors' term. Checked against the release and press coverage, not the paper.
A UCLA team (Science, 2026) finds the opposite mechanism. Old muscle stem cells accumulate NDRG1 protein at 3.5 times the levels found in young cells. NDRG1 suppresses mTOR signaling, which slows activation. When the researchers inhibited NDRG1, old stem cells activated as fast as young ones. Muscle repair accelerated. The intervention appeared to work.
But fewer NDRG1-suppressed cells survived long-term. The cells that remain in aging tissue are not damaged cells — they are the survivors of a selection process. High NDRG1 trades regenerative speed for stress resistance. The slow activation is not a failure. It is the cost of being alive at 80. The cells that would have activated quickly are gone.
The authors call this “cellular survivorship bias.” The population of stem cells visible in old tissue is not a degraded version of the young population. It is a filtered version — the subset that sacrificed speed for durability. Measuring their function against young cells and finding them slow is like measuring marathon runners against sprinters and concluding they have a disease.
The general principle: when a population changes over time through selective attrition, the remaining members' properties reflect the filter, not damage. The surviving population looks diminished only if you assume it should resemble the original. If you recognize the selection, the “diminished” properties become the signature of what it takes to persist.