Replying to M. P. Rozing, T. B. L. Kirkwood & R. G. J. Westendorp Nature 546, 10.1038/nature22788 (2017)

In the accompanying Comment1, Rozing et al. provide the interesting analogy of sports performance with ageing to question our finding of a limit to human lifespan2. We thank them for giving us the opportunity to clarify several misunderstandings.

First, the long-jump records merely show that, while the occasional athlete may set a new record, there is still a fundamental limit to human performance. As another commentary comparing our findings with the world of sport observed3, although leading athletes might be able to reduce the world record for the 100-metre sprint by few milliseconds, they will never be able run the same distance in, for example, 5 seconds. Similarly, somebody might jump a few extra centimetres, but they will never span a kilometre in a single leap. Whether that reflects a mechanism that is similar to the process of ageing is unclear, but certainly not supported by any concrete evidence.

The criticism that we use the same dataset both to propose a hypothesis and to test it is simply not true. First, a possible end to the continuous life expectancy increases of older and older cohorts is suggested by our figure 1 (ref. 2), apparently overlooked by Rozing et al.1, who mistakenly believe that survival probabilities at old age show no deceleration. Second, our analysis of the International Database on Longevity (IDL) and the Gerontological Research Group (GRG) database (independent datasets, albeit based in part on the same individuals) is data-driven and generates a model that provides strong evidence for a limit to lifespan (not a decline, as Rozing et al.1 mistakenly infer from our paper). The fact that a significant increase in the maximum reported age at death (MRAD) value is obtained when all data from the 1960s to 2015 are considered is obvious and does not preclude a finding that the MRAD value reaches a plateau in the 1990s. Indeed, the fact that such an increase has been lacking for more than 20 years (not a decade, as Rozing et al.1 state) in spite of the fact that the number of centenarians over that same time period has increased exponentially speaks for itself.