The deceleration of improvements in national life expectancy

Life expectancy has been increasing over time. A broad rule of thumb is that, on average across the world, every four years life expectancy has increased by about one year. That would seem to suggest that we are about to enter an age where young people (especially young women) can expect to live to 100 (although Riccardo Trezzi remains the only living immortal). However, as I noted in this 2017 post, it pays to take care with claims about future life expectancy.

Often, those making the claims are looking at some form of projection based on past changes in life expectancy. If we take a closer look at past improvements in life expectancy, what do they tell us? A 2018 article (ungated) by Carolina Cardona and David Bishai (both Johns Hopkins Bloomberg School of Public Health), published in the journal BMS Public Health, provides an answer. Cardona and Bishai used national-level data on 173 countries over the period from 1950-2009, and looked at decadal changes in national life expectancy. They found that:

When we compare countries at a similar starting point of LEB [Life Expectancy at Birth], we find that the pace of growth of LEB has slowed over the last 60 years for countries at all levels of starting LEB. Countries with the lowest LEB (LEB < 51) had the greatest slowdown...

In the lowest group, LEB decadal improvements in the 1960–69 decade were 4.171 (SE 0.473) years lower than in the 1950–59 decade (P < 0.001), and in the 2000–09 decade were 8.491 (SE 0.982) years lower than in the 1950–59 decade (P < 0.001). The deceleration of LEB growth among the group with the highest starting LEB was smaller, at 2.961 (SE 0.181) years gained per decade slower for the 1960s (P < 0.001) and 1.893 (SE 0.264) years gained per decade slower for the 2000s (P < 0.001), both compared to 1950s.

In other words, gains in life expectancy over time have been decelerating, and have been decelerating in all countries (on average), regardless of whether they have high, or low, life expectancy. These results control for a bunch of factors known to be associated with health and life expectancy. The growth of HIV/AIDS doesn't explain it, because the slowdown has happened in countries with and without generalised HIV epidemics. The slowdown has happened in every region of the world.

To illustrate further, here's Figure 1 from the article, which plots the change in life expectancy gains over time for countries with life expectancy below 51 years (on the left), and for those with life expectancy above 70 years (on the right). This slowdown is most apparent for countries with low life expectancy.

To me, the most disappointing thing about Cardona and Bishai's article is that they didn't extrapolate to see what their results imply about future life expectancy and where the limit may lie. That would be an interesting exercise for a future honours or Masters student, since the underlying datasets are all available online.

[HT: Marginal Revolution, last February]

Read more:

• Future life expectancy is good for NZ men, but be cautious
• The limit to human lifespan, or not?