Fascinating study.

The stats warrant some caution, though. The main finding is based on figure 4 [1] and I wouldn't be surprised if the number and location of these 'eras' varied a lot if the authors use 40,000 people instead of 4,000.

[1] https://www.nature.com/articles/s41467-025-65974-8/figures/4

Especially the last era - over 83 - is suspicious. With 4000 people and ages 1-90, how certain can we be about this? But I don't want to cast unjustified doubt, I'm sure they did the math.