Readers keep flagging this claim that radiation is causing cancer among pilots and flight attendants do to regular exposure inflight. But the underlying scientific work doesn’t support this conclusion at all.
Does radiation at cruising altitude cause deadly cancers in pilots and flight attendants? A new @JAMAInternalMed study suggests the answer is yes pic.twitter.com/vOj6StgExm
— Alex Berenson (@AlexBerenson) August 17, 2026
I read the piece with my morning coffee. It doesn’t show that pilots and flight attendants are 36% – 50% more likely to die of cancer, or that cosmic radiation caused their cancers.
- It finds that, among people who had already died, selected cancers accounted for an unusually large share of deaths in people whose usual occupation was pilot or flight attendant.
- That’s not a rate of cancer incidence, or a mortality rate since the study doesn’t treat this relative to overall crew or offer conclusions of years at risk per person.
Flight crew are exposed to more ionizing radiation than most people. They appear to have elevated incidence of skin cancers and female cabin crew appear to have elevated breast cancer incidence.
It’s unclear whether cosmic radiation is the cause of this, rather than UV exposure, circadian disruption, lifestyle, screening, or something else entirely is the cause of this and the piece doesn’t attempt to show that at all. And it’s certainly not enough to overturn previous literature on the subject.
Here’s the commentary piece and the underlying paper. It found that among people who died, the odds that the recorded cause was a selected cancer rather than something else was 36% – 50% higher. That does not mean, as is frequently being claimed, that working in aviation raised someone’s probability of developing or dying from cancer by those amounts.
That might also be exactly what’s expected. Pilots have to pass medical exams and have much lower all-cause and cardiovascular mortality. So cancer coud cause a larger share of deaths because pilots die less frequently from cardiovascular disease, smoking-related illness, and other causes.
In a 2014 study with two million person-years of observation, cancer made up 30% – 38% more crew deaths even though the actual cancer mortality rate was 27% below normal for male pilots and normal for female flight attendants. And that ratio is pretty close to this new paper’s 36% – 50% framing.
This paper doesn’t attempt to estimate radiation exposure tied to cabin crew and deaths – there’s nothing about flight hours, years worked, routes, altitudes, or age at exposure. Someone with one year of short haul flying is treated like a 35-year pilot flying polar routes.
Broadly speaking, research finds that death from cancer among airline crew is either lower than the population at large or at least not elevated, but that’s likely because they’re healthier than the general population at hiring and during employment. Melanoma and other skin cancers though tend to be elevated and so is breast cancer but radiation doesn’t appear to explain why.
This 2020 German study of 26,846 crew found no increased cancer mortality. Nordic pilots had no excess cancer deaths but about twice the incidence of skin cancer.
As it happens, in June there was a new National Academy of Sciences consensus report “Assessing Radiation Exposure, Health Outcomes, and Mitigation Strategies for Flight Crewmembers 2026” that finds:
- Crew exposure is real, cumulative, and important enough to track and manage.
- An individual flight’s dose is low, but a career can accumulate meaningful exposure.
- Existing studies don’t establish that cosmic radiation causes particular observed health outcomes.
- Pilot melanoma is fairly well documented. It’s likely due to UV exposure rather than cosmic radiation.
- Female cabin crew breast cancer is also well documented, though its cause is unclear.
- Evidence for other cancers is inconclusive.
This new piece isn’t an outlier in detecting greater skin cancer, but it doesn’t show radiation causing cancer or airline crew being more likely to die as some seem to be suggesting.
Well said, Gary !
I think you may be wrong on this one. I have seen several UK studies which do show that pilots and flight attendants do have a higher incidence of contracting cancer. Studies show that this particularly true for people flying the North Atlantic and North Pacific which is where cosmic radiation is at the highest levels
The study depends whether it’s from the “Sugar-Industry” and/or US government agency. Studies from overseas reflect a different view somehow.
One or maybe a few European airlines test crews regarding radiation exposure around the second half of the year. If levels in body are elevated, then they are done working for the year supposedly.
I heard growing concern amongst crews getting leukemia from flying composite aircraft like the B878. Not enough shielding perhaps.
What if the pilots and flight attendants are on aircraft that are spraying chemtrails? Are they at greater risk of cancer?
Thanks for sharing this, Gary. I certainly commented about it beforehand. (Oof, if @TexasTJ is one-side of something, I’m immediately skeptical of his and Gary’s skepticism). When a travel commentator (sorry, a “thought leader”) tears down a health study like this, it is worth asking, is Gary defending scientific rigor, or are you protecting ‘vibes’…
Sure ignoring this would be awfully convenient (gee shucks…) for an industry that has little interest in paying for better shielding, rewriting scheduling limits to combat circadian destruction, or admitting that the friendly skies come with an occupational tax. (‘Get back to work, y’all!’) Yet, no, it is probably not a cartoonish, Big Tobacco-style conspiracy meeting in a smoke-filled boardroom. But, the airlines do likely prefer a neat, data-driven excuse to ‘look the other way.’
It’s fine to, let’s say, get a “second opinion” on all this, but that shouldn’t mean letting airlines off the hook just because an observational study was ‘imperfect.’ If the data is too messy to prove the risk, management should not be saying, “see, nothing to worry about, get back in the galley.” They should be asking why, many decades into the jet age, we still have to rely on noisy proxy studies to figure out what chronic radiation and altitude are doing to the people working the front of the plane. My hope is studies like this are a starting-point, not the end.
@Matt — Please consider, on some airlines, those are ‘premium’ chemtrails… (@L737, vacation?)
You are spot on that the statistics would be different because pilots are screened for health issues and would show a much lower rate of cardio issues than the general population. Our preemployment physical and career long physicals rule out any pilot with an increased chance of heart disease. Nothing in the physicals checks for cancer. You can’t get hired if you have a risk factor for heart disease and once hired pilots often keep fit (not all of them!) so they can pass their physicals. So of course there would be a distorted ratio of cancer deaths compared to the general population. Cancer doesn’t care if you keep your blood pressure down or have normal cholesterol but the airline that hires you does.
Thanks for the compliment, @1990 ! For you to state “if @TexasTJ is one-side of something, I’m immediately skeptical of his and Gary’s skepticism” is something, given that you are almost never correct in your position. I’ll take it !
An observational study typically raises one or more questions (“hypothesis generating”) but doesn’t answer them. We see an associations between occupation and a cancer diagnosis at death, but don’t know if its causal. There are ways to answer the question with reasonable scientific rigor in an observational format, but it would require matching for a large number of confounding variables and perhaps conducting the observation prospectively rather tha retrospectively. Gary rightly points out what may be the most important confounder: dose (years of flying as a surrogate). Seeing a positive dose-response curve would be much more suggestive than what we see in this study.
Remember that when you see this sort of study looking for multiple signals in a large population that there will always be false positive signals, essentially by intent. This is a rapid, low cost throw of a broad net knowing that you’ll need to cull out the bad hypotheses with more rigorous studies afterwards. Remember that the standard for statistical significance is 95%. If you ask a hundred questions (look for 100 associations) as part of one of these broad net studies you would expect five false positives (give or take).
It all depends on the type of radiation exposure , whether waves , or particles . I was exposed to Agent Orange , which is a liquid , and not radiation .
@1990, @Texas TJ
It’s the old “ I’d agree with you, but then we would both be wrong “!
Correction to my earlier comment: if you FIND 100 associations, five would be expected to be false positive, not LOOK FOR 100 associations.
@Jay Gee: Sweet !!! Love it !
Finally! Something I have a little expertise on.
Crew members are getting upwards of 3-6 mSv a year. If that dose definitively causes higher fatal cancer rates, we’re all screwed anyway. Especially all you high altitude folks in Denver etc. For reference the average American is getting 3 mSv of annual background radiation (NRCP 160).
Having said that, mandating the FAA track crew doses is a legitimate debate to have. If airlines were under NRC jurisdiction and I was the Radiation Safety Officer, I’d be absolutely chewed out by federal/agreement state authorities for not having any evidence to prove crew members don’t need to wear dosimeters.
I bring radiation dosimeters on every trip I go on. Unfortunately they all go through TSA x-ray screenings but I have never gotten a measurable neutron reading from cosmic sources.
Radiation is scary to a lot of people so I’m hoping that proper context is disseminated. Thanks for helping with that, @Gary!
@1990 — I’ve been around! Next trip this weekend, just been enjoying the shows with a lot of popcorn – been some lively back and forths recently!
@TexasTJ, @Jay Gee — Now, it’s a party!
@L737 — Niccce… “we’re all screwed anyway.” Ok, so, “3.6 roentgen—not great, not terrible.”
As the saying goes, “Don’t bother me with facts, I’m trying to make a point here…”
@Denver Refugee — Bah, you got a mention!
@1990 @Denver Refugee — No joke, I originally had wrote “@Denver Refugee and all you high altitude folks”, should’ve kept it, bah!
@1990 – no one is suggesting ‘ignore this’ I’m simply saying that the viral claims around this aren’t backed by what the study says, there’s plenty else that would explain it, there’s a pretty robust body of data on it as well and we should try to understand that rather than act as though this is the first time anyone has asked this question.
Excellent dissection Gary. I read more research (published and seeking-to-be published) in my career than I could possibly count. I often read research outside my area as part of IRB§ screening. The ability of scholars not to fully control for covariates always astounded me. Folks, it ain’t finding correlations, but explaining them.
@Mark F, The standard is usually (certainly in areas I know) wriiten as 5%, not 95%, though that’s just as unimportant as ground beef being 95% lean or 5% fat. But, as it depends on area, there exist areas where the standard alpha level is less than 5%. That high wouldn’t fly in my area. But, as you know, if you divide the US into 10,000 “communities” of ~34,000 each, you’ll find that 200 communities will have a “statistically higher” cancer rate with a one-sided alpha of 2% (two-sided 4%). Residents will be certain it isn’t random.
§an IRB reviews proposed human-subjects research.
@This comes to mind says: 95% or greater certainty that the finding is not due to chance/5% chance or less that it is due to chance. Same coin different side.
We get a much higher radiation from indoor Radon along here in the upper Midwest.
That being said mountain West is probably a double whammy with altitude and radon
Can we talk about mmWave security scanners as well.
I personally feel zapped of energy after passing through, and it makes the science feel concerning.
Cell phone radiation is linked to cancer, why not going through a scanner every day?
@John when did cell phone radiation get linked to cancer?
@Gary Leff — Alrighty. Even if the existing data is ‘robust,’ it’s a shame the industry has spent decades waiting for noisy ‘viral’ (the social media kind) studies to force a serious conversation about what chronic occupational exposure actually does to crews. Either way, thank you for addressing the topic.
So John says, “I personally feel zapped of energy after passing through [mmWave security scanners], and it makes the science feel concerning.” Was everybody else picturing “Saul’s” brother in “Better Call Saul”? Particularly, when Saul had batteries snuck on him….
@This comes to mind — Actually, now that you mention it, great reference! (Unfortunately, you’ll never see this, because you ignore what I write, allegedly. That must be tough since you frequent most of the same blogs. Do you shield your eyes from my comments… you know, like Chuck’s electromagnetic hypersensitivity…)
In my years of riding in the cockpit, I see more and more pilots wearing baseball caps. In the summer, it’s not unusual to see one wearing a long sleeve cut off from an old winter shirt over their outboard arm. Both are used to protect from the enhanced UV light streaming in from the overhead, side and forward windscreens. I personally know a long time international FO that was treated with MOHS surgery for cancer lesions on his RIGHT side of the scalp. He wears his hair in a “crew cut” (no pun intended) which further exposes the danger of UV
not sure who to believe:
1. Gary (who apparently went to both law school and medical school)
2. A Harvard Medical School article
The fact that American Airlines previously granted AAdvantage elite status solely based on the number of miles flown must have effectively invited a scientific investigation to find out whether the more you fly, the more likely you are to glow in the dark—or at least to accumulate more deadly cancers caused by cosmic radiation. Isn’t it obvious that ConciergeKey members must be gathering more rare cancer diagnoses just as frequently as their frequent flyer miles are collected? It seems entirely logical (though by logical I mean entirely unscientific) to believe that the less worthy AA Executive Platinum elites should be dying off at a higher rate than the slightly less radioactive AA Platinum Pro elites. Since I’m only a Lifetime AAdvantage Platinum elite, I can conclude that I’m going to live longer and enjoy better health—at least until American Airlines comes up with a new tier for those who survive AA ConciergeKey, AA Executive Platinum, and AA Platinum Pro. Live long and prosper to all AAdvantage elite members.