On dying

Having recently published a post on ageing, it seems fit to tackle the question of dying. As I have already explained I am interested in the question of the maximal life span having provided estimates thereof based simply on the evolution of athletic performances. By a simple analysis of the existing data I reached the conclusion that there is a maximal lifespan of 125 years. This number is in agreement with the one provided by G. West (author of the monumental work "Scale") who obtained his estimate from physics-based considerations. 

Two recent works devoted to estimates of the maximal human lifespan attracted my attention and I would like to report on them here. 

The first paper, a work of the Singapore-based biotech company Gero, looked at human resilience, i.e. how well the organism bounces back from something that puts stress on it.  They found that an 80-year-old requirεs three times longer to recover from stresses when compared to a 40-year-old. The resilience keeps decreasing with age and disappears completely at some age between 120 and 150. They obtained this estimate by studying various indicators as a function of age and plotted the inverse of variance obtained. They remarked that by extrapolating there is an age where this quantity goes to 0. This means that the variability increases indefinitely, leading to an arbitrarily long recovery time, signaling thus the demise of the individual.

The authors concluded that their work explains why "even the most effective prevention and treatment of age-related diseases could only improve the average but not the maximal lifespan". Their work shows that the recovery rate is an important signature of ageing and this could guide the development of strategies aiming at slowing the process and lead to a higher quality of life across the lifespan.  

The second paper is a work led by a team of the Harvard Medical School. Their study, which was not limited to humans, investigated the modulation of lifespan of model organisms by genetic, dietary and pharmacological interventions. In the case of humans the mortality rate grows exponentially with age, as shown in the figures below. The authors posit that a natural soft bound on the maximal achievable lifespan does exist and is situated around 138 years. They do point out that this is higher than other estimates reported in the literature, which, the authors claim, were biased due to their choice of data based on the oldest cohort of the population. Still, I find a little bit ridiculous the precision announced: 138 rather than, say, 135-140.

Reading the critiques formulated in that paper led me to look up the existing literature. A first study of mortality, from a physics perspective, gives some interesting insights. The figure below presents the death rate in the US at the beginning and the middle of the past century. While the overall mortality did decrease substantially over the years from birth to roughly an age of 30, there was no major gain of life expectancy for ages beyond this.

It is even more interesting to examine the data from France, spanning almost two centuries.  Extrapolating from the available data using Gompertz' law one finds that the death rate lines converge at an age of 120. 

Perusing the literature I came upon a paper by Dong et al. that has been severely criticised. The main critique was that the authors have splitted their dataset arbitrarily into two groups and claimed that the maximum age of death had reached a plateau which they estimated at 115. Since there is one documented case of an individual living up to an age of 122 the maximum human lifespan cannot be smaller than this. A typical value advanced is 125. 

Since this is an athletics blog it is interesting to present in detail one of the rebuttals of Dong's work based on an athletics argument, showing that one can reach wrong conclusions by arbitrarily splitting the datasets. In the figure below are presented the results for the three medalists in men's long jump for the Olympics and the World Championships.  

By partitioning the data into two groups, before and after 1991, one can reach the conclusion that there was an improvement in performance up to 1991 and deterioration thereafter. However when one analyses the data as a whole, one finds a statistically significant increase in the winning long jump distances over time.

So, after all, what is the limit of the human lifespan? As far as I am concerned, I stand by my estimate of 125. It is reached by a very simple, sports-based argument, and to me this is enough. 

And if you wonder whether simple arguments may suffice in this domain, let me remind you that the Jewish birthday wish, since time immemorial, is 
or ‘may you live till 120’.