Median and maximum lifespan respond to different things, and the difference is informative

Two ways a survival curve can change

A survival curve plots the fraction of a population still alive against time, and interventions can change its shape in structurally different ways. One change moves the curve rightward in its middle portion while leaving the point where it finally reaches zero more or less fixed. Another pushes the far tail outward, meaning the longest-lived individuals live longer than the longest-lived individuals ever did before.

These correspond to shifting median survival and to shifting maximum lifespan, and they arise from different underlying causes. Treating the two as one measurement is the most common way a modest result gets reported as a dramatic one.

Why the median moves more easily

Removing or delaying a common cause of early death raises the median substantially without altering the biology of the oldest survivors. In laboratory colonies this frequently means delaying the strain-typical tumour that kills a large share of the animals before old age. In human populations the historical rise in life expectancy was driven overwhelmingly by reductions in death during infancy and youth.

That is an enormous achievement in human terms and it is not evidence that the underlying rate of ageing changed at all. The two claims are routinely merged in popular writing, which produces the impression that ageing itself has already been slowed.

Maximum lifespan is statistically fragile

The maximum observed lifespan in any group is determined by a very small number of individuals, sometimes only one. Estimates based on extremes are unstable, since a single unusually long-lived animal can shift the reported maximum considerably. Statistical approaches exist that look at the survival of the longest-lived tenth rather than the single oldest individual.

Those approaches are more robust and they require larger groups, which is one reason many experiments are underpowered for tail effects. A claim about maximum lifespan from a small colony should be read with that fragility firmly in mind.

Rectangularisation and what it implies

When early deaths are eliminated without the tail moving, the curve becomes more rectangular, with most of the population dying within a narrow window. This shape is what a population approaching a fixed biological limit would look like, and it is what human survival curves have been doing. Whether an actual limit exists is contested, with analyses of very old individuals supporting different conclusions depending on the model chosen.

The disagreement is partly about statistics and partly about the reliability of age records at extreme ages, which are hard to verify. It is an open question rather than a settled fact in either direction, and confident statements on either side outrun the evidence.

Reading an experimental claim

A useful first question about any lifespan result is whether the reported effect appears in the median, the tail, or both. The second is what the animals actually died of, since a change in cause distribution explains many median-only effects. An intervention that moves both the median and the tail is making a stronger claim about ageing than one that moves only the middle.

Neither result is worthless, and they license different conclusions about what the intervention was doing to the underlying biology. Summaries that quote only a headline improvement without specifying which of the two measures actually moved are withholding the part of the result that carries the meaning.