Heat disproportionately kills young people: A SIPA News Q&A with SIPA Climate Researchers

A study published by SIPA scholars in Science Advances challenges our traditional understanding of heat-related mortality. “Heat Disproportionately Kills Young People: Evidence From Wet-Bulb Temperature in Mexico” — led by a group of researchers including Andrew Wilson PhD ’24, a postdoctoral scholar at Stanford University's Center on Food Security and the Environment, R. Daniel Bressler, a PhD student at SIPA, and SIPA associate professor Jeffrey Shrader — reveals that people under 35 are surprisingly vulnerable to extreme heat, accounting for 75 percent of heat-related deaths.

Through their analysis of Mexico's mortality data and wet-bulb temperature measurements, the researchers found that occupational exposure may play a crucial role, particularly affecting outdoor workers. The findings challenge traditional early warning systems and protective policies, which have been designed primarily around protecting elderly populations. 

Their work also highlights what Shrader calls “silent killers.” While extreme weather events like hurricanes make headlines, heat and cold are enormous sources of mortality that often go unrecognized, even by medical professionals. These researchers’ insights are particularly urgent for developing countries that combine young populations, outdoor work, and extreme heat exposure.

In this SIPA News Q&A, Wilson and Shrader discuss their findings’ implications for public health, climate adaptation, and development policy, particularly for countries with younger populations and higher exposure to extreme heat.

Your research suggests we need different interventions for different age groups. Could you elaborate on how heat protection strategies might need to differ between children and young working adults?

Jeffrey Shrader: I think we’ve done a pretty good job over the last 20 years introducing effective interventions to help protect older adults from heat waves. In Canada, the US, and especially Europe, we have public health workers who check on older people during heat waves, make sure they can get to cooling shelters, and do early-warning systems so people can plan ahead.

For younger adults, our studies show there may be different exposure paths. In our results, many work in manual jobs exposed to temperature. So natural policy proposals would be things like on-the-job break time, access to water (especially for agricultural workers in the field), and training people to recognize signs of heat illness. California has policies protecting agricultural workers from heat, other states like Washington followed, and the Biden administration proposed a nationwide rule for this. For children, it’s different because they aren’t making their own decisions. The decisions about their heat exposure are made by caregivers — could be teachers on the playground, coaches during sports, or parents at home. So that’s a special challenge where education really needs to target that group of adults making decisions for kids.

The occupation angle is interesting. We’re raising this important point about needing specific policies to address temperature-related health risks for different age groups. I think it even raises some tensions around messaging, and that’s really important to keep in mind. It further reinforces that information with outreach, with education together is probably going to be more successful than any one of those three things separately.

Your study shows under-35s made up about two-thirds of Mexico’s population during your study period. How should this demographic profile influence our interpretation when looking at similar developing countries experiencing extreme heat?

Andrew Wilson: Most high-quality mortality data we have globally comes from high-income countries whose populations tend to be much older. So, existing research has looked a lot at what happens in older populations when they’re exposed to heat. Studies try to account for this by looking separately at age groups or age-standardizing the population. But just looking at results from these higher-income, older contexts and assuming they translate to lower-income, younger countries — without considering the different age structure — would probably underestimate the potential risks of climate change and health harm.

Shrader: One additional thing, Lower-income countries with younger populations also have larger shares of their population exposed to temperature. Larger fractions working outdoors, larger shares working in agriculture. So, there are these compounding risks that, as Andrew said, probably mean we’re understating the mortality risks related to climate change with our current estimates, which largely come from colder, older countries.

Wilson: And wealthier countries, too. The ways we protect ourselves from ambient temperature are typically by being indoors in a climate-controlled place, which is strongly associated with income. Lower-income places are just less able to invest in adaptation. As Jeff said, these places tend to have large numbers of people working in heat-exposed outdoor occupations, and compounding that, these workers tend to be younger on average.

Mexico has a lower-than-average population in agricultural work. For countries with a much higher percentage, would you expect to see a more prominent trend, and what protections could they take?

Shrader: I'm really curious about that. Combining my and Andrew’s points, I’m interested in looking at other locations with different mixes of age, occupation, humidity and heat, to understand how well this generalizes globally. I suspect it's going to be worse in places with higher agricultural work shares and more outdoor work in general. Even factory work inside can be extraordinarily hot and dangerous, with ambient heat contributing to that.

One of the biggest challenges to understanding this issue is that relevant health and mortality data simply don’t exist in many places where we most want to study this question. This is a major public health tragedy. In places across sub-Saharan Africa and Southeast Asia that are very hot, humid, and have many people exposed through their occupation or otherwise, we just don't have complete vital records to understand this issue.

Wilson: Yes - having lived in America most of my life, it feels almost obvious that a birth or death should result in a data point somewhere. But in many places – and this is a priority for the UN, though progress is quite slow — people are born, live their entire lives, and die without generating any information that could be used for research or understanding quality of life. In the US, we can look up relatives in census records from long ago. That sort of thing just doesn't exist for a large fraction of the world even today.

Based on your research findings, what specific development interventions should be prioritized?

Wilson: One thing I’ll raise – and this relates to another line of my research – is about early warning systems. We know from high-income countries such as those in the EU that early warning systems, alerts about heat waves, and weather forecasts can be really effective in helping people avoid mortality from temperature. Jeff works on this in the US, showing that more accurate weather forecasts really help people avoid mortality from temperature.

These high-quality forecasts and early-warning systems are not nearly as prevalent in lower-income developing countries – actually, the vast majority of low-income countries. It's something that's easy to take for granted in places like the US or Europe. So that's the sort of policy intervention that I think could be really effective, in addition to education about the dangers of heat, though many folks in these settings are already savvy about that sort of thing.

How do you design early-warning systems when you have different temperature thresholds – when 10 percent of days are 'too hot' for those over 50, while about half are ‘too hot’ for under-35s, according to your paper?

Shrader: One of the things we've learned from the European and US experience is that information by itself only gets you so far. Information is more effective when paired with effective outreach and communication strategies. 

In the US, the technical side of early warning systems — the physics and meteorology – is well understood and deployed. There's not much work needed there. But much of the ongoing policy and public health work is making sure the right groups understand the risks they're facing. Our work highlights an interesting tension in that communication challenge – though it's one that's faced already to some degree, just maybe not with quite this amount of specificity across different age groups.

Your paper has been getting a lot of buzz – both from researchers and in the news. What’s next on your research radar based on all the feedback you've received?

Shrader: I’m very curious to dig more into this occupational exposure channel. We know climate change poses grave risks for agricultural production, so thinking about how it affects agricultural workers is something I'm especially interested in. I do research on the construction sector — I've shown how climate change affects construction work, but I wasn't focused on health impacts. 

Now I'm interested in expanding that to study health impacts as well. This is a really big deal, but I don't think it’s widely appreciated. You see NOAA (The National Oceanic and Atmospheric Administration) saying heat is the most deadly natural disaster in the US, but it’s even more true in places like Mexico. Heat and cold are enormous sources of mortality, but they're kind of silent killers — they don't grab headlines like other natural disasters.

Wilson: Looking at other places where exposures aren’t quite so extreme but where we have high-quality mortality data would be top of my list. One surprising number from our analysis is that about 10 percent of deaths in Mexico are plausibly caused by ambient temperature exposure — in other contexts it’s between one and 10 percent, which is astronomical. I’m interested in understanding who's vulnerable and what these deaths look like — are they from cardiovascular stress, respiratory conditions? And interestingly, clinicians rarely code deaths as temperature-related on death certificates, even though we see this clear link in population-level data. If the people closest to these mortality events aren't recognizing temperature as a primary factor, there's clearly a lot left to explore.