Why Do Climbers Really Die on Everest?

The fact that two experienced climbers died near the summit of Everest last week is sad but unsurprising. As Alan Arnette pointed out, expeditions on the Nepal side of the mountain alone have been averaging almost four deaths a year since the turn of the century. But the situation this year is a little more fraught, with a severe wave of coronavirus ripping through Nepal and a worsening outbreak at Everest Base Camp.

Authorities in Nepal were quick to dismiss any link between the deaths and the virus. “Reaching to that height is impossible if someone is infected with the COVID,” the director general of Nepal’s tourism department, Rudra Singh Tamang, told the New York Times. The head of Seven Summit Treks, which was guiding both of the deceased climbers, said the same thing, attributing the deaths instead to altitude illness. On the surface, that seems like a reasonable claim (and I have no specific information to either refute or support it), but it prompts a question: what is it, exactly, that does kill climbers on Everest?

There’s plenty of data on this question, thanks to the comprehensive Himalayan Database started by the late Elizabeth Hawley. And there have been several attempts by researchers to analyze the patterns in this data. Sometimes the causes of death are clear. There’s no ambiguity about the 15 people who died at Everest Base Camp in the 2015 avalanche. But when someone collapses in the so-called Death Zone above about 26,000 feet (8,000 meters), it’s much harder to distinguish between the various forms of altitude illness, cold-related injuries, and straightforward exhaustion, all of which leave them stranded to die of exposure. Even if they fall off a cliff, you don’t know whether it was a consequence of impaired balance and cognitive function due to altitude illness, or perhaps a loss of coordination from frostbite.

With those caveats in mind, here are some stats. In 2008, a team led by anesthesiologist Paul Firth published an analysis in the British Medical Journal of 192 deaths among more than 14,000 Everest climbers and Sherpas between 1921 and 2006. Of that total, 59 percent of the deaths were attributable to trauma either from falls or hazards such as avalanches. In 14 percent of the cases, the bodies were never found so details are unknown. The remaining 27 percent are the most interesting ones, attributed to non-trauma causes like altitude illness and hypothermia.

When you restrict the data to the 94 people who died above 8,000 meters, some interesting details emerge. Even among those who fell to their deaths, many were described as showing signs of neurological dysfunction, such as confusion or loss of balance. This is significant, because altitude illness comes in several forms. The mild version is acute mountain sickness (AMS), which mostly just manifests as feeling like crap. The two more serious versions, either of which can be fatal, are high-altitude cerebral edema (HACE, meaning swelling in the brain) and high-altitude pulmonary edema (HAPE, or swelling in the lungs).

One dog-that-didn’t-bark detail, according to the study, is that “respiratory distress, nausea, vomiting, and headache” were rarely noted in those who died above 8,000 meters. That may be, in part, because those symptoms—characteristic of AMS or HAPE—might be unambiguous enough to prompt you to turn back before it’s too late. In contrast, if your thinking is a little cloudy thanks to incipient HACE, that may not seem like such a big problem—and your ability to recognize the problem is compromised by the cloudiness of your thinking.

I’ll admit that I’m skeptical of the assertion that no one with COVID can get to 8,000 meters. Depending on the timing and severity of your infection, you might be healthy enough to get to the highest camp, and just start showing very mild respiratory symptoms on the day of your summit push—not enough to realize that you’re in trouble, but just enough to put you in danger as the day wears on. But the data above suggests that, for the most part, it’s not lung problems that kill people near the summit. That doesn’t rule out the possibility that COVID was involved in this year’s deaths, but it certainly lowers my index of suspicion.

There’s a more recent analysis that’s also worth digging into, published last year in PLOS ONE by a team co-led by biologist Raymond Huey of the University of Washington and statistician Cody Carroll of the University of California, Davis. Huey and his colleagues had published an earlier analysis of all 2,211 climbers making their first attempt to ascend Everest between 1990 and 2005, looking for patterns in who succeeded and who didn’t. The new paper updates that analysis with another 3,620 first-time climbers between 2006 and spring 2019, and there are some notable insights about the differences.

Of course, there have been plenty of changes on Everest since 2006. As the viral photographs and permit numbers reveal, it’s way more crowded. The standard critique is that guiding companies are hauling rich, inexperienced dilettantes up the mountain who create traffic jams and make bad decisions, putting everyone at greater risk. Interestingly, the death rate has decreased a bit, from 1.6 percent in the earlier period to 1.0 percent in the more recent period. That said, since the number of climbers has quadrupled, the actual number of deaths has increased. The more recent climbers were also twice as likely to reach the summit: “This supports (I think) the idea that better logistics, weather forecasting, fixed ropes, experience (of expedition leaders and high-altitude porters) have improved success rates and slightly lowered death rates,” Huey told me in an email. “But we have no direct data to evaluate these suspicions.”

The role of crowding is a little trickier. Nepal issued a record 408 climbing permits to foreigners this year, and more than 100 climbers summited on May 11 and 12 alone. Huey and his colleagues compared the summiting and death rates on crowded and uncrowded days, and didn’t see any differences. But that doesn’t mean crowding doesn’t matter. “Perhaps the ‘uncrowded days’ had relatively bad weather or poor snow conditions, and climbers waited for better conditions,” Huey says. “If that is the case, then the crowded days would be crowded because conditions were favorable, and favorable conditions compensated for any detrimental effects of crowding.”

Indeed, it’s hard to imagine that crowding doesn’t make a difference. It inevitably causes delays, and your risk of getting caught by an avalanche or rock fall is directly proportional to how long you’re out there—one of Reinhold Messner’s rationales for rapid alpine-style climbing, Huey notes. Perhaps even more importantly, the longer you’re at extreme altitude the more the effects of altitude illness may accumulate.

The 2008 BMJ analysis notes that there are two main explanations for why climbers would develop balance and cognitive impairments. One is that you’re not getting enough oxygen to the brain, either because you run out of supplemental oxygen or because you’re exercising really hard. But there were no apparent differences in patterns of death for those with or without supplemental oxygen, and there were very few deaths while ascending just below the summit, when the physical demands of the ascent are greatest. So the more likely explanation is that these climbers are suffering from the brain-swelling effects of HACE.

Back in 2006, a British doctor named Andrew Sutherland wrote an opinion piece for BMJ titled “Why are so many people dying on Everest?” He’d recently summited Everest, and had paused to help a climber with HAPE at 23,000 feet—and then, farther up the mountain, passed the bodies of four less fortunate climbers.

“I think it is likely that we all develop a certain degree of pulmonary and cerebral oedema [i.e. swelling] when going to the summit,” he wrote, “and that it is only a matter of time before we succumb to it.” The mild disorientation from HACE leads to bad decisions and a slower rate of climbing, which in turn (along with factors like crowding) lengthens the amount of time you’re exposed to extreme altitude, causing the symptoms to worsen. This root cause, he argued, likely contributes to many deaths whose final blow is dealt by a fall or hypothermia or exhaustion.

After his own climb, Sutherland had to visit to the French consulate in Kathmandu to identify the body of a Frenchman who’d reached the summit but been too exhausted to descend, managing only about 150 feet in six hours before being abandoned by his expedition partners. The consul shook his head. “He didn’t reach the summit until 12:30; that is a 14-hour climb—it is too long. All the files we get of those that die on the mountain, c’est toujour la même chose—they take too long to reach the summit.”

For more Sweat Science, join me on Twitter and Facebook, sign up for the email newsletter, and check out my book Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance.