番茄社区

COVID-19 lowered life expectancy in the U.S. in 2020 and 2021. Last month, the CDC released new estimates based on 2022 data. The headline: life expectancy is finally creeping back up. There has been some cautious celebration around this. After all, it's nice to see that sudden jackknife giving way to an upward trajectory.

But there's a sad reality that we have to recognize. Among the living -- that is, we who have survived since COVID-19 erupted in 2020 -- life expectancy already was up. We just could not see it in the graphs. Why? Because it was outweighed by the downward pull of the 1.16 million COVID-19 deaths.

On Average, 80-Year-Olds Live Longer Than Most People Think

Let's use 80-year-olds to understand this phenomenon. If you live to age 80 in the U.S., your average life expectancy remains around 9 years. But it's a bell curve. Some will die this year, and some will die in 20 years.

To illustrate this, I instructed ChatGPT to make the following graphs.* As you can see in the first graph below, about as many 80-year-olds are expected to die in the next 2 years as there are people who will live for 15 more. So, as the vertical red line shows, the average 80-year-old will live around 9 more years (108 months), but half will live longer and half will die sooner.

What Would We Expect a Pandemic to Do?

Pandemics cause shifts in these distributions. It's easy to see why. We know that COVID-19 is more likely to be fatal in people with dangerous medical conditions.

In the graph above, those people are going to be more commonly found to the left of the red mean life expectancy line. (The medical conditions these individuals have are why they would be expected to die sooner.)

So while COVID-19 lowered the survival curve across the spectrum of this graph (i.e., the whole population of 80-year-olds), it would be expected to affect it most dramatically to the left of the red line (and in fact, the further left you go, the bigger the change in survival would be).

What would that effect look like in a post-pandemic world? Take a look at what we might expect to see for a post-pandemic "survivor" cohort:

If you look closely, you'll notice that the post-pandemic curve indicates a higher fraction of 80-year-olds surviving beyond 10 years (and, as the blue vertical line shows, the mean life expectation has increased by around 6 months).

The reason is something called the "healthy survivor effect." Any 80-year-old who got COVID-19 and survived probably already had a somewhat longer life expectancy than an 80-year-old who got the virus and died. (That is, the differences in their risk was already present in 2019; the pandemic just unveiled the situation.) COVID-19 didn't off people randomly. It offed people who had higher risks more frequently. (Sadly, it kills some people with no appreciated risks at all -- albeit this was relatively rare.)

The result is a mean life expectancy (blue vertical line) that has moved to the right in comparison to the red line in the previous graph, meaning that life expectancy is now higher.

Finally, here's a graph that combines the two scenarios above. (Again, "original" indicates what might have happened if there had been no pandemic while "pandemic adjusted" shows the results of plucking people off who had actually been on the left part of the "original" curve all along, whether they knew it or not.)

So, Why Is Life Expectancy Still Below 2019 Rates?

If COVID-19 killed over a million high-risk people and created a healthy survivor cohort, why hasn't life expectancy returned to normal, or even leapfrogged to all-time highs?

The reason is straightforward: COVID is still killing too many people. It's a dangerous enough disease that it is killing people in the broad middle of these curves (and maybe even some on the right side). In a life expectation tug-of-war between COVID-19 and the healthy survivor effect, the virus is still winning.

All of this means something profound: if we were to somehow get rid of COVID (or acquire enough immunity to render its effect on mortality negligible in all age groups), we'd promptly expect the post-pandemic survivor cohort to have a pretty dramatic increase in life expectancy. That would not be because we're healthier. It would be because we allowed a novel virus to kill a million higher-risk people (i.e., people on the left of the above curves). This is called the "pull forward effect."

Many viruses have a habit of doing just what I've described COVID as having done (just less dramatically). So, imagine what life expectancy curves for older people could look like if we were able to knock out COVID, influenza, RSV, and a few other bad pathogens that deliver fatal hits to aging people. Now that would be something to celebrate.

*The power of ChatGPT is pretty crazy. I made all of these graphs by giving the app written instructions. I gave it absolutely no numbers at first! (I eventually asked it to make it so that the mean of the second graph would be 6 months higher than the original.) Thought you'd enjoy that detail.

Comment