I’m getting my A/C system replaced (yay!) so it seems like the perfect time to look at heat deaths in the U.S.!
Look, it’s not that many people who die explicitly from excessive natural heat in the U.S. (this is not including fire deaths).
But I have started including the “contributing cause of death” breakout as well as underlying cause of death, so that people can see the trends.
Long-term trend of heat deaths, 1968-2025
These are the straight counts, not rates.
I wrote about the historic heat wave of 1980 in my 2024 post. The very short version: there was an extremely nasty summer, lasting from the end of June through the beginning of August, where a large area of the Midwest/Southern Plains area saw temps in the 90s-100s … and a large number of homes had no air conditioning.
It was 1980, it wasn’t necessarily common.
I lived in Savannah, Georgia at the time, where it got to 100+ every summer, but we also didn’t necessarily have A/C everywhere. That’s why we were “lazy”. We sat in the shade of the oak trees and waited for it to cool down.
People have moved around since 1980. For example, I no longer live in Georgia. I live in NY.
Heat deaths, 1999-2025, UCD and CCD
Let’s look a little more recently.
In the prior section, I showed heat deaths only where they were the UCD (underlying cause of death).
Here, I’m including where natural heath where it’s a CCD (contributing cause of death) as well.
For UCD, that’s the one and only one cause of death on the death certificate, and that’s X30 (exposure to excessive natural heat).
But for CCD, I’ve got all these other codes I can grab:
P81.0 (Environmental hyperthermia of newborn)
T67.0 (Heatstroke and sunstroke)
T67.1 (Heat syncope)
T67.2 (Heat cramp)
T67.3 (Heat exhaustion, anhydrotic)
T67.4 (Heat exhaustion due to salt depletion)
T67.5 (Heat exhaustion, unspecified)
T67.6 (Heat fatigue, transient)
T67.7 (Heat oedema)
T67.8 (Other effects of heat and light)
T67.9 (Effect of heat and light, unspecified)
X30 (Exposure to excessive natural heat (hyperthermia))
And you can see in recent years how these extra heat deaths have contributed.
It went down a little in 2025.
Monthly pattern of heat deaths, 2018-2025
I didn’t do this graph before, and the result is unsurprising, but I thought you might like to see it.
Unsurprisingly, it peaks in July.
You can see more starkly the changes from year to year.
Heat deaths by State
I always “enjoy” this breakout, because some people get a little surprised….
Guys. There are deserts in Arizona and Nevada….
….but it also relates to a bunch of retirees moving there.
(In heat death posts in prior years I contrasted to cold-related deaths, but I am not in the mood to contrast these. No, I don’t know what’s going on in Arkansas. Yet.)
Heat-Related Death Rates by Age
This is unpleasant, but it must be done.
Perhaps I should have led with this graph, but the only other thing I will note: this is UCD only.
I will post two graphs — the longer-term and the shorter-term. There aren’t huge differences.
Death rates 1999-2025
Death rates 2021-2025
There are some differences — the death rates for the oldest ages increased for the 2021-2025 period compared to the longer-term period.
That is hardly surprising given the higher susceptibility of seniors to heat deaths (yes, infant and toddler heat deaths are mostly exactly what you think they are — left in vehicles). So most of the increase in heat deaths are coming from older adults, not children.
Why?
To be sure, some can be from climate change, but some can be from movement of people.
Note from the tile grid map, the highest heat death rates were in Nevada and Arizona. Hmmm, let’s take a look at the population graphs for those states since 1999.
Nevada population, 1999-2025
Arizona population, 1999-2025
Both Arizona and Nevada have shown robust growth over 1999-2024 (the data available).
To definitely cherry-pick (and yes, I’m doing this on purpose), let’s take a look on Illinois and New York over the same horizons:
New York, 1999-2025
Illinois, 1999-2025
Given what happened through the pandemic, one could question how much to trust these estimates.
But let’s just pretend.
In the case of New York, there’s an estimate of rising from 18.2 million to 20.0 million in population over 1999 (January) to 2025 (January) — a fairly anemic 0.4% per year growth rate.
Then there’s Illinois. Less than 600,000 in population increase over those 26 years. Barely 0.2% per year growth.
What’s my point here? The HOT PLACES were growing in population well beyond what was going in NY and IL.
From 1999-2025, separate from whatever the climate was doing, people in the U.S. were shifting from the midwest and north to the south and west in the U.S. Few people die of heat in NY or Illinois compared to the desert areas of the West.
The old folks, personally, like moving to Phoenix, Arizona. Even if there may be some extra risks due to the heat.
But the point is, that if we’re looking at heat death trends, we may also need to consider that people are moving to areas where they’re more likely to die from the heat.
