No one knows exactly how many people are dying from extreme heat

By Zahra Hirji and Preeti Soni

When officials in Miami-Dade County wanted to figure out how many people there were dying from extreme heat, they turned to Christopher Ueijo.

A geographer at Florida State University, Ueijo used temperature data and death records to determine that, between 2015 and 2019, heat killed an average of 34 people a year in the county who wouldn’t have died otherwise.

His estimate was more than 10 times higher than the official figures: Over the same time period, Miami-Dade had classified just two deaths as heat-related.

About a year later, Ueijo did a second analysis of the same data, this time factoring in the buildup to heat waves’ peaks to capture more of their early and cumulative health impacts. The newer findings, published in May, were wildly different: 600 heat deaths per year.

“They are both peer-reviewed studies,” says Jane Gilbert, the chief heat officer for Miami-Dade County who commissioned the two analyses. “I asked [Ueijo] which one I should use, and he said it’s somewhere between 34 and 600.”

In the space between two, 34 and 600 lies the central challenge of counting fatalities from heat. Officially this year, there have been 63 heat deaths in Thailand, at least 143 in India, at least 172 in Mexico, and more than 1,300 in a single week during the Hajj pilgrimage in Saudi Arabia.

But unofficially, these numbers are a fraction of heat’s true toll. They are often the most obvious cases — those where doctors or coroners could confidently draw a line between heat and death — and are much less likely to account for fatalities in which temperatures played an indirect role.

Some of this ambiguity has to do with the narrow way heat deaths are tracked, and some with the sheer number of health conditions exacerbated by rising temperatures. But the end result is the same: More people are dying from heat, while academics, medical professionals and government officials struggle to keep track.

“Heat’s tricky,” says Este Geraghty, chief medical officer at Esri, which produces mapping software used by researchers and government officials to track heat impacts. “It’s been called the silent killer, and for good reason.”

The most common way to track heat deaths is by tallying up fatalities where heat has been officially identified as the main or contributing cause of death.

This is Gregory Hess’s job. As chief medical examiner for Arizona’s Pima County, he and his team record the cause of death for many of the people who expire in one of the hottest parts of America. Last year, Pima County’s Tucson experienced nearly 90 days where temperatures hit at least 100F (38C).

“So, let’s say somebody dies because they shot themselves in the head,” Hess says. “Why the person died is pretty easy to keep track of. That doesn’t require a lot of sophistication.”

Heat is different. The most direct version of a heat death comes from hyperthermia, two extreme forms of which are heat exhaustion and heat stroke. Symptoms of heat exhaustion include dizziness, thirst, nausea and weakness, while heat stroke symptoms may include confusion and lost consciousness.

Heat stroke usually sets in when core body temperature reaches roughly 104F (40C). The body simply can’t handle such high temperatures for long and will start failing without rapid cooling and other treatment.

But there are limits to identifying hyperthermia. Not everyone who gets heat-sick is able to seek help. If someone dies alone at home or overnight in a park, core temperature will likely have decreased by the time their body is found, says Sarah Henderson, director of British Columbia’s Centre for Disease Control. In other words, the most obvious sign of heat stroke is erased.

Many official heat death records rely almost entirely on identifiable heat stroke. In Japan, heat stroke is the only type of heat death tracked by the country’s Fire and Disaster Management Agency.

In the US, the Centers for Disease Control and Prevention tracks heat deaths by compiling information from death certificates, where reported heat deaths are often tied to heat stroke. The data is limited to US citizens, and “there tends to be a one-year lag before complete data are available,” says CDC press officer Sharleta Stamps.

Heat stroke also makes up the vast majority of heat deaths tracked by India’s Ministry of Health and Family Welfare, though local hospital systems can also report heat-related deaths. Combined, the local numbers are often higher than the official national figures.

India has “a fairly high threshold for what can be classified as a heat-related death,” says Bhargav Krishna, a fellow at the research organization Sustainable Futures Collaborative.

The designation usually requires a recorded high body temperature and extensive documentation on a patient’s family and medical history. Impacted families may be “eligible for certain compensation,” Krishna says, which can result in physicians being “more conservative in declaring a heat death.”

Heat stroke fatalities are the most straightforward kind of heat death. Another, bigger category is what Hess refers to as “heat-contributed deaths.”

Imagine somebody has a heart attack on a hot day, says Geraghty at Esri: “Did they have it because it was hot or it just happened to be a death on hot day?” The answer isn’t always easy to figure out, and many places don’t have the resources or inclination to try.

The human body has two primary tools to avoid overheating: sweating, which releases heat through evaporation, and pumping the heart faster to increase blood circulation and transfer heat to the skin.

Any group unable to avoid the heat — athletes, the homeless, low-income families, and outdoor workers — is increasingly vulnerable as temperatures climb. So too are people whose bodies can’t easily adjust, including infants and young children, pregnant people, older people, and those with certain health issues.

Chronic obstructive pulmonary disease, heart disease and diabetes, for example, can limit the body’s ability to keep cool, as can some of the medicines used to treat them. But when people suffering from those conditions die in hot weather, the disease is “more likely listed as cause of death than heat,” says Ueijo.

“Covid is a good analogy,” says Kristie Ebi, a professor of global health at the University of Washington. Due to limited testing and investigation, many likely Covid-linked deaths were attributed only to comorbidities such as hypertension or diabetes.

The virus’s long tail means it also factors into deaths from weakened organs and health complications well after a person stops testing positive.

Even for the physically healthy, low levels of heat illness can impact concentration, balance and energy levels. That increases the risks for anyone impaired by drug or alcohol use, and for those suffering from mental illness.

“Why do we get so many more unhoused people dying in parks with drugs in July than we do in December? It’s because of the heat,” Hess says.

To accurately determine heat-contributed deaths, officials at every step of the process must be versed in heat’s impacts and willing to figure out whether they played a role. At minimum, there needs to be an official record of the death — not a given in rural parts of India, for example — that lists heat as a contributing factor.

But if governments are only reporting heat deaths based on death certificates and physical examinations of dead bodies, “then it’s highly likely to be an underestimation,” says Eunice Lo, a research fellow in climate change and health at the University of Bristol’s Cabot Institute for the Environment.

While Pima County has always tracked some heat deaths, Hess says they weren’t able to investigate a wider swath of cases until he secured enough funding to hire an epidemiologist in 2022. Since last year, Pima County has been sharing its heat death findings online and updating them monthly. According to that data, the county has identified 24 heat-related deaths in 2024 so far.

“I’m sure it’s still an undercount,” Hess says. “I think we’re doing it as well as we can with how the system works, and as well as we can means that’s going to be better than a lot of other places. But there’s going to be those deaths that slip through the cracks.”

Hess’s inspiration is neighboring Maricopa County, home to Phoenix and long a national leader in tracking heat deaths. During hot months, the county publishes weekly reports on known and suspected heat fatalities. For the first half of the year, Maricopa County reported 13 confirmed heat-related deaths and 162 more under investigation. In 2023, there were 645 heat-related deaths.

To get around how infrequently heat is listed as an official cause of death, a growing number of academics and public health officials are applying “excess mortality” studies to heat deaths.

While these analyses don’t normally help with real-time reporting, they use the same method that public health experts do to pin down how many people are dying directly or indirectly from Covid-19.

To complete an excess mortality study, experts apply statistical modeling to a location’s meteorological data and official death records over a certain period of time, whether that’s a week-long heat wave, a single summer, or multiple years.

The aim is to “find the relationship between temperature and people dying,” Lo says. One such study, published in 2021, found that heat caused roughly 489,000 excess deaths a year globally between 2000 and 2019.

After a heat wave in 2003 caused more than 70,000 deaths across Europe — a number determined by an excess mortality study — the UK upped its focus on heat risk. Since 2016, the country’s Health Security Agency has calculated excess heat mortality for every summer.

In 2023, officials estimated there were nearly 2,300 excess heat deaths over five hot periods. “Even having a heat warning system came about because of the 2003 heat wave,” Lo says.

In Europe, heat contributed to some 61,000 deaths in the summer of 2022, according to an excess heat mortality study published last year. One 2020 study looking at 297 US counties — nearly two-thirds of the population — found that there were about 5,600 excess heat deaths annually between 1997 and 2006, double the highest heat death toll found using death certificates.

Excess mortality studies are only feasible in regions that keep comprehensive records on all causes of death over time — a prerequisite that’s easier to meet in the UK, US and Europe than it is in Africa, India or Southeast Asia.

“For many countries, we simply don’t have digitized death records or registrations,” Lo says. “For some places, like in Southeast Asia, we’ll have some data but it’s only in the last few years. Whereas in Europe, you can have 40 years of data.”

The studies also leave plenty of uncertainy, as Ueijo’s team found with their analyses of Miami-Dade County. The disparate findings speak to “the lack of full understanding of the excess morbidity and mortality related to heat exposure and the need for improved data collection and more research,” Gilbert says.

Extreme heat isn’t going anywhere. Rising greenhouse gases in the atmosphere have already warmed the planet by about 1.2C compared to the pre-industrial era, and heat waves are becoming more frequent and more intense. That trajectory makes it increasingly critical to appreciate heat as a lethal threat.

“All those deaths are potentially preventable,” Ebi says. “Yet people don’t really think of heat as the killer it can be.”

Greater awareness could also keep the heat from becoming even more deadly. People can get in the habit of increasing hydration, dressing for high temperatures and shifting their daily routines to avoid the hottest part of the day. Governments can increase access to cooling at home and in public, plant trees to add shade, and implement other solutions geared at heat adaptation.

But the first step is understanding the scope of the problem. “You need to know these numbers so that you can design your infrastructure and policy accordingly,” says Avikal Somvanshi, head of the Urban Lab at India’s Centre for Science and Environment. Otherwise, other countries might think, “Oh, India is a country of 1.4 billion people and only 300 people died of heat stroke last year.”