The Shape of U.S. Mortality 2: 1900-2022
Extending and comparing against the Spanish flu pandemic
At the end of October, I did a podcast episode called The Shape of U.S. Mortality, where I looked at the long-term trends of U.S. mortality, going back to 1900 and even before to see how Americans died and how things have improved since then.
The data I looked at stopped at 2018 because my point was about earlier improvements and not what’s happening now.
So let’s add back 2019-2022, and see the big picture.
The shape of U.S. mortality 1900-2022
Let’s start high-level and look at the age-adjusted death rate, 1900-2022.
The Spanish flu pandemic is difficult to miss in this graph. The spike sticks out in 1918.
But there are other spikes and local maxima to think about — throughout the 1920s there were a lot of ups & downs — diphtheria was pretty endemic, and killed off many children (and horses).
In 1936, there was a flu epidemic in the U.S., and it was also the end of the Dust Bowl.
I marked off two mid-20th century flu epidemics, which had been major and I posted about back in 2020.
But I will note that pre-1950, mortality trends were volatile, with infectious disease having many one-year effects, and though penicillin was discovered in 1928, it wasn’t until post-WWII that it became widely used in the civilian population. It is rather amazing how smooth the mortality curve was from about 1970 to 2019.
We got complacent pre-COVID, thinking that those decades were “normal”, when that smoothness was unusual, historically. The spike of the 1918 Spanish flu pandemic was also extreme. Still, you can see the smaller spikes — those were pretty substantial, and frequent, before we got effective treatment for bacterial pneumonia, say.
Yes, there is a COVID mortality spike, but other things are going on other than just COVID. Let’s break it out.
Shape of mortality by cause of death, 1900-2022
Speaking of bacterial pneumonia, let’s look at the shape of the mortality pattern by major causes of death.
Let’s get the full shape stacked up:
Heart disease and cancer are the top two causes from the mid-century onward. They didn’t provide what the top causes were in 1900, but obviously, there was a lot more infectious disease as the cause of death in 1900. That has come way down with vaccines and antibiotics.
But look at the shape — the peak of heart disease deaths came around the 1950s, partly due to having no good treatments/surgeries, and widespread smoking (specifically among men) driving much of that problem. Cancer has only slowly improved.
You can see that flu/pneumonia dropped off as a cause of death (mostly pneumonia) once widespread antibiotics became available.
But wait - let’s do this as a line graph.
Focusing on the top 5 causes of death
I’m omitting “all other causes”, because it is on a very different scale, and what dominates in that “all other” changed greatly over the years.
With the line graph, we can see the rise and fall of heart disease death rates (all of these are age-adjusted, to avoid the changing demographics of the U.S.)
But then we see cancer had a peak around 1990 — that relates to my Movember fundraiser — please donate! — and let’s focus on 2010-2022.
Top causes of death 2010-2022
Now that we look at the corner of the graph, we can see that heart disease had stalled out in improvement from 2010-2019.
Accidents had been getting worse the entire period - just got worse during the pandemic, but already had been doing poorly (this is the drug overdose issue, primarily).
Stroke did get worse during the pandemic, but compared to the others, it doesn’t look as bad.
To make it simpler, let’s make a table:
Recall that accidents include both accidental drug overdoses and motor vehicle accident deaths, both of which contributed to this result.
However, see that heart disease and stroke also had substantial increases over the pandemic.
In the year of the Spanish flu pandemic, it was a one-year impact, and the other causes of death decreased, not increased. That’s not what we’re seeing here.
That said, in 1907, we did see all causes increase. Sometimes that does happen. We also saw increases over all major causes during the 1920s. That was some volatility in the mortality experience during that decade.
To be sure, some of that could be iffy data, but there had been massive disruptions post-World War I, and we see that heart disease was climbing as a cause of death.
Don’t have to sit there and take it
Perhaps we are in a reversal of mortality trends, at least for some causes of death.
This is one of the reasons I’m involved in the Insurance Collaboration to Save Lives.
As I mentioned above, we’ve gotten complacent, assuming that mortality rates generally go down. That’s not necessarily true.
Many of the vulnerabilities in health in the U.S. have been uncovered in the pandemic, some of which were somewhat seen before 2019. But they got worse during the pandemic.
My contribution to the effort can be seen above — my expertise as a life actuary is in analyzing mortality trends and also considering the impact in life insurance and related areas.
Check it out!
