Isn´t this also an issue with World War's 1 and 2? They both dented life expectancy fairly significantly, and that cohort is now aging out of the figures. If the rate of increase is being measured, then that would presumably play a small part?
By doing so, historical events outside of the study period don't have a major effect. In this study, they deliberately avoided the covid pandemic as it would have caused a significant drop in life expectancy that is not representative of a general trend. Of course, the future is not taken into account either. Despite what the name may suggest, life expectancy is like a snapshot of the studied period (1990-2019 here), not a prediction.