[elmarks]

Both can still be true.

The majority of people do not die as children, but the effect of those who do is disproportionately large on "life expectancy" because it's an average.

Let's say that 23% of children die in their first year (true in Sweden in 1751). Let's say everyone else dies at 80. Life expectancy is 61.6 years.

Now let's say time passes and now 0.2% of children die in their first year (true in Sweden today) and everyone else dies at 80. Life expectancy is now 79.7 years. It jumped over 18 years (30%) with no one living any longer than before. Additionally, the percentage of children dying was never even 1/4 of the total, so the vast majority of people still died from other causes.

The point is that the linked chart represents numeric percentages of people dying, while life expectancy weights younger years much more heavily.

Most of the progress that has been made in life expectancy is from young children, but most of the progress in "what actually kills people" has been for older people.

[kenjackson]

On phone... Short reply. Life expectancy since 40 increased at a rate comparable to expectancy since birth. That is, we've made huge gains in keeping ppl alive even starting from age 40.