The Autopsy / Case File · Environmental & Climate
The Heat Death the Certificate Doesn't Name
U.S. heat-related deaths have more than doubled since 1999. The official national count is still likely a severe undercount.
Heat rarely kills the way a death certificate is built to record. It triggers a heart attack, a stroke, kidney failure — the event that gets listed as the cause, while the heat that pushed a strained body past its limit goes unmentioned. Here's a sourced look at what heat actually does inside a body, why the national numbers are almost certainly too low, and what one county's own surveillance system reveals about the gap.
Join the discussion →
Heat doesn't need a heat wave headline to kill someone. It needs a body that can no longer keep pace with an environment demanding more cooling than that body can produce — and a coroner or medical examiner willing to write "heat" on a form built, historically, to record a single, specific, immediate cause of death.
The Trend Line
The most rigorous recent accounting of U.S. heat mortality comes from a peer-reviewed analysis of CDC National Center for Health Statistics data spanning nearly a quarter century, published in 2024. It tracked every death where heat was recorded as either the underlying cause or a listed contributing cause — the broadest definition the death-certificate system supports.
The trend wasn't a steady climb across the full 25 years. Researchers found a flat-to-declining pattern from 1999 through roughly 2016, followed by a sharp acceleration afterward — meaning the true story isn't simply "heat deaths have always been rising," it's that something changed around the mid-2010s to break a two-decade plateau and send the trend line sharply upward in less than a decade.
What Heat Actually Does
The lay understanding of a heat death — someone collapses outdoors on the hottest day of the year — describes only the most visible fraction of the mechanism. The physiological cascade that kills most heat victims is slower and, clinically, more indirect.
A rise in core body temperature of only a few degrees above normal is enough to seriously strain the cardiovascular system: the heart has to pump dramatically more blood to the skin to shed heat, on top of its ordinary workload, and in a person with existing heart disease, that added strain is frequently what triggers the actual event that kills them — a heart attack, a stroke, an arrhythmia — rather than heatstroke itself. Sustained hyperthermia also damages cells directly and can trigger a systemic inflammatory response that pushes multiple organs toward failure at once.
Younger adults can roughly double their cardiac output in response to heat stress. Older adults, whose cardiovascular systems can't mount the same compensatory response, are left disproportionately exposed to exactly the mechanism that kills.
That age-based vulnerability gap is a core reason heat mortality concentrates so heavily among older adults: it isn't just that older people are more likely to lack air conditioning or live alone, although both are real contributing factors — their cardiovascular systems are measurably less able to compensate for the same thermal load a younger body would tolerate without incident.
Why the Certificate Often Names Something Else
Death certificates are built around a single underlying cause, with a limited number of lines for listed contributing conditions — a structure well suited to a single traumatic event and poorly suited to a cascade where heat is the trigger for a cardiac event that gets recorded as the cause on its own. A medical examiner who wasn't told a decedent had spent hours in extreme heat before collapsing has no reason to write "heat" anywhere on the form; the immediate clinical cause, a heart attack, is accurate as far as it goes.
CDC itself has acknowledged the structural limitation for years. In 2017 the agency urged medical examiners to document a natural hazard like extreme heat on a death certificate even when its role was indirect rather than the immediate cause — guidance that depends entirely on an individual examiner's judgment and local investigative practice, which varies widely across roughly 2,000 separate coroner and medical examiner jurisdictions nationwide, most without any dedicated heat-surveillance protocol at all.
One County That Actually Counts Carefully
Maricopa County, Arizona — metro Phoenix — runs one of the only heat-mortality surveillance systems in the country built specifically to catch the cases a standard death-certificate process misses, investigating each candidate death individually rather than relying solely on cause-of-death coding. The county's own numbers, compared against the national picture, are the starkest illustration of the undercounting problem available.
Maricopa County isn't uniquely hot compared with the rest of the desert Southwest, and it isn't the only large county with a substantial elderly and unhoused population exposed to extreme heat. What it has that most of the country doesn't is a dedicated public health surveillance system that actively investigates ambiguous deaths for heat involvement instead of waiting for a death certificate to volunteer the information. The gap between what Maricopa finds when it looks specifically for heat's role and what most of the country's death-certificate-only system captures is, itself, the strongest evidence for how much the national tally is likely missing.
Some Neighborhoods Run Hotter Than Others
Exposure to dangerous heat isn't evenly distributed across a city, and the pattern isn't random. Neighborhoods with fewer trees, more pavement, and less green space absorb and retain far more heat than leafier, wealthier areas often just a few miles away — a well-documented effect known as the urban heat island, and researchers have traced its geography directly back to decades-old housing policy.
A federal mortality tally built from death certificates has no mechanism to capture any of this — two people can die of the same heat-triggered cardiac event on the same afternoon, one in a shaded, high-canopy neighborhood and one in a paved, treeless one that ran ten degrees hotter at the same hour, and the death-certificate record treats them identically. The undercounting problem this piece has traced through methodology and coding practice compounds with a geographic exposure gap that the same records were never built to see either.
The Smaller, Better-Documented Slice: Dying at Work
One corner of heat mortality is tracked with more precision than the general population figures above, because it flows through a separate federal system built for workplace deaths specifically. BLS's occupational fatality census recorded roughly 40 to 50 environmental-heat deaths a year in recent years — a fraction of the national total, but one where investigators know the exposure occurred on the job, in a specific industry, during a specific shift.
Even this narrower, better-tracked figure is widely regarded by occupational health researchers as an undercount of the same kind described above, since a worker who collapses from a heat-triggered cardiac event days after a shift, rather than visibly at the job site, may never be coded as an occupational heat fatality at all — the same gap between mechanism and paperwork, playing out again at a smaller scale.
And Then You're Dead
Heat doesn't announce itself on the paperwork the way a car crash or a gunshot does. It works through the heart, the kidneys, the blood vessels a person already had trouble with before the temperature climbed — and it leaves the record naming whichever of those failed first, not the heat that pushed it there. The count that says heat killed a few thousand Americans last year is probably the floor. Nobody currently running the count nationwide is confident it isn't a great deal higher.
That's it. That's the whole thing.
Sources
- Heat-related deaths rising from 1,069 in 1999 to a record 2,325 in 2023 (a 117% increase), 21,518 total deaths across the period, an age-adjusted rate of 0.26 per 100,000 person-years, and the post-2016 acceleration in the trend — "Trends of Heat-Related Deaths in the US, 1999-2023," peer-reviewed analysis of CDC NCHS mortality data, 2024
- Cardiovascular mechanism of heat mortality, including the blunted compensatory cardiac-output response in older adults — "The Physiology behind the Epidemiology of Heat-Related Health Impacts," American Physiological Society, Physiology, 2025 and "Human Cardiovascular Responses to Passive Heat Stress," PMC
- CDC's 2017 guidance urging death-certificate documentation of heat as a contributing factor even when indirect, and the agency's own acknowledgment that national heat-mortality tallies are likely undercounts — Scientific American, "U.S. Deaths from Heat Are Dangerously Undercounted," September 2024
- 645 confirmed heat-associated deaths in Maricopa County, Arizona in 2023, and 602 in 2024 (the first year-over-year decline in a decade, despite Phoenix's hottest recorded summer) — Maricopa County Department of Public Health, "2024 Heat-Related Deaths Report" and "2023 Heat Related Deaths Report"
- 4.5°F average temperature gap (up to 12.6°F in some cities) between formerly redlined and non-redlined neighborhoods across a 108-city study, and the 15.2% tree-canopy gap and roughly 2.7°F temperature gap between low- and high-income U.S. city blocks — National Community Reinvestment Coalition, "The Temperature of Disinvestment: Examining Urban Heat Islands and Historically Redlined Communities" and "The Tree Cover and Temperature Disparity in US Urbanized Areas," PLOS One
- 48 occupational environmental-heat-exposure deaths in 2024 and a roughly 40-per-year decade average — National Safety Council, "Work Safety: Exposure to Environmental Heat," citing BLS Census of Fatal Occupational Injuries data
Photo: "Extreme heat warning sign in Death Valley, California, USA" by Graeme Maclean, via Wikimedia Commons (CC BY 2.0).
Comments
Loading…