Nope—There’s No Thyroid Cancer Epidemic in Fukushima

A new study comes out with claims of a giant epidemic of thyroid cancer among kids exposed to radioactive iodine from the Fukushima nuclear accident. It’s disproven by another recent study showing that thyroid cancer rates are no higher in Fukushima than in distant regions uncontaminated by the accident. Which study gets lots of attention? And which one gets none?

If you think the alarmist study will get the spotlight while the reassuring one gets ignored, then you understand the malign nexus of anti-nuclear hysteria, ambitious fringe scientists and journal editors, and news media hungry for sensational stories—and how it distorts public understanding of science, health and the risks of radiation.

Dire warnings of the thyroid cancer epidemic—“The Nuclear Sacrifice of Our Children,” in the words of acitivist Helen Caldicott¹—have resounded since the accident in 2011. (Radiation experts counter that radio-iodine doses from Fukushima were too small to cause a noticeable uptick in thyroid cancer rates.) Last week the journal Epidemiology published a paper by ecologist Toshihide Tsuda and colleagues at Japan’s Okayama University that seemed to find proof positive of a cancer epidemic.² The researchers looked at statistics from Fukushima Prefecture’s thyroid cancer screening and treatment program. They found that over four years, 110 cancer cases developed among 298,577 kids younger than 19 at the time of the accident; that’s an incidence rate of 92 cancers per one million kids per year. They compared that with a baseline incidence rate in 2007 of just 3 cancers per million per year, and concluded that, after the accident, Fukushima kids were 30 times more likely than normal to get thyroid cancer. That finding made headlines and impressed eminent scientists like Columbia University radiation biophysicist David J. Brenner, who told the Associated Press that the thyroid cancer epidemic is “real."³

So is the cancer surge real? Nope. The Tsuda study’s conclusions are the product of bad methodology, flawed reasoning and egregious obfuscation of evidence.

The problem is that the diagnostic methods used to establish the study’s baseline incidence are radically different from the screening methods used to find thyroid cancer in Fukushima, which rely on super-sensitive ultrasound thyroid scans of hundreds of thousands of children. Normally, asymptomatic kids don’t get thyroid scans; cancer diagnosis proceeds only if a child has abnormal thyroid symptoms or a tumor big enough to feel by throat palpation. But in Fukushima, healthy kids got ultrasound scans that can detect tumors far too small to cause symptoms or be felt. The result is a “screening effect” that artificially inflates childhood cancer rates by finding small, slow-growing tumors that would normally not get diagnosed until adulthood, if ever. A proper baseline to compare to the Fukushima cancer rates should therefore be derived from the same screening protocol used in Fukushima, namely⁴:

Step 1. Give thousands of asymptomatic kids ultrasound thyroid scans to find potentially cancerous nodules (fleshy lumps) and cysts (fluid-filled sacs).

Step 2. Give nodules bigger than 5.1 millimeters in diameter, or cysts bigger than 20.1 mm, a second ultrasound exam. (Almost all of these were nodules.)

Step 3. Biopsy the nastiest-looking nodules to see which ones are malignant.

Tsuda’s baseline incidence rate of 3 cases per million comes from ordinary symptoms-and-palpation diagnoses, which detect far fewer malignancies. That apples-to-oranges comparison makes Fukushima cancer rates seem alarmingly high.

But a proper apples-to-apples baseline has been established now in a recent study with a simple design: examine kids living far away from Fukushima—kids not exposed to excess radiation—using the same ultrasound scans as in Fukushima, and see what their thyroid cancer rate is. It turns out to be about the same as in Fukushima—which means the supposedly sky-high Fukushima rates are quite normal. That study, which definitively disproved the cancer epidemic theory, garnered no headlines and sank without a trace.

The study, sponsored by the Japanese government and conducted by Naomi Hayashida of Nagasaki University and colleagues, gave thyroid scans to 4,365 kids age 3 to 18 in distant Aomori, Nagasaki and Yamanashi prefectures. The first phase determined that the rate of potentially cancerous thyroid abnormalities—nodules and cysts—was higher in these control areas than in Fukushima.⁵ Using identical screening standards, Fukushima thyroids looked as healthy as the controls.

In a crucial follow-up paper published last March—and ignored by the press—the Hayashida group looked for thyroid cancer proper.⁶ Over two years, one thyroid cancer was diagnosed among the 4,365 kids screened. (Some may have been missed; because of the thyroid hysteria in the prefecture, 23 percent of Fukushima kids with nodules over 5.1 millimeters got biopsies,⁷ compared to just 4.5 percent of kids in the Hayashida baseline study.) One is not many, but it’s an incidence rate of 115 cancers per million kids per year—38 times higher than the Tsuda study’s “normal” baseline rate of 3 per million, and higher than the 30-fold elevation Tsuda found in Fukushima kids.⁸ So by the Tsuda standards, Hayashida’s uncontaminated control group had a somewhat higher thyroid cancer incidence rate than did Fukushima kids.

The Tsuda researchers mention the Hayashida study in passing and even confirm a key numerical result (and then bury that number in a table in a supplemental information document separate from the main paper and say no more about it).⁹ What they don’t do is point out to readers what the Hayashida study demonstrates: that their claims of elevated Fukushima cancer rates are wildly inflated. Ultrasound screening campaigns always find rates of thyroid abnormalities and cancer many times higher than “normal” baseline diagnosis rates, even among healthy kids who have not been exposed to excess radiation. Claims of a Fukushima cancer epidemic depend on this deceptive comparison with an invalid baseline.

None of this will surprise anyone who knows the literature on “occult” thyroid tumors. One typical Japanese autopsy study found that 11.3 percent of people have undiagnosed thyroid cancer when they die.¹⁰ (Roughly 0.5 percent of people in Japan are diagnosed during their lifetimes).¹¹ Most of these tumors were small, but one percent of autopsies reveal tumors larger than 5 millimeters, the Fukushima diagnostic threshold. This suggests that the true prevalence of large thyroid tumors, diagnosed and occult, is perhaps 1.5 percent in the general population; in the Fukushima cohort it is currently just 0.042 percent.¹² That prevalence will rise as the cohort ages, but there’s a long way to go before it becomes abnormally high.

The Hayashida study is impeccable, and confirms the expert consensus on radiation and thyroid cancer. It agrees with many previous studies showing that hidden, innocuous thyroid tumors are much more common than diagnosis rates suggest. It should have driven a stake through the heart of the thyroid cancer myth. By contrast, the Tsuda study finding elevated Fukushima cancer rates was poorly designed, glaringly at odds with previous research and based on a comparison with diagnosis rates that is clearly invalid. It deliberately swept under the rug evidence that contradicted its conclusions. The study is so flawed it should never have passed peer review. Yet it was published and widely covered, while the Hayashida studies languish in obscurity.

So why the perverse difference in reception and impact? All through the food-chain of scientific publicity there is a bias towards finding scary health effects from radiation and other potential toxins. Scientists know that journals are more likely to publish such studies, journal editors know reporters are more likely to write stories about them, and journalists know fretful readers are more likely to click on those stories. The bias towards strong effects seriously distorts the scientific literature, with many—by some counts most—such results turning out to be wrong when researchers try to replicate them.¹³

Dogmatic radiophobia is also a factor. Some high-minded but wrong-headed scientists and journalists feel that the truth of terrifying radiation effects must be told, no matter what the facts say. Scientists are supposed to be hardened rationalists, but their ability to trick themselves into confirming their favorite hypothesis is well-documented,¹⁴ and seems rampant in Tsuda’s research. Yuri Kageyama, an Associated Press reporter who wrote a story about the Tsuda study that was picked up by many news outlets,¹⁵ feels so strongly about Fukushima thyroid cancer that she wrote this poem about it for her spoken-word stage act¹⁶ :

Please listen and tell the world.
How our children in Fukushima are getting thyroid cancer, one by one.
My daughter is one of them.
Pediatric thyroid cancer is rare.
The chance for getting it is under one in a million.
One in a million.
But in Fukushima, it’s 112 out of 380,000 children tested, and the tally is growing.
This is Fukushima after Three-Eleven.
Beautiful Fukushima, where rice paddies stretch between lazy mountains...

Perhaps no amount of journalistic objectivity could overcome Kageyama’s emotional attunement to the Tsuda study.

There have been skeptical notes in the coverage of Tsuda’s paper, and some scientists, like Geraldine Thomas, who studies thyroid cancer and radiation at London’s Imperial College, have offered a needed corrective to media hoopla.¹⁷ But the attention lavished on the study, while the Hayashida studies and others that demolish it go unremarked, points to a breakdown in scientific and journalistic vetting.

Scientists need to do some soul-searching about the preconceptions they bring to their research, and take steps to correct them.¹⁸ Journal editors and peer reviewers—especially at Epidemiology!—need to tighten their standards and give closer scrutiny to studies that defy the scientific consensus on highly politicized topics like radiation risks. Reporters need to be honest about their own ideologies and preferences for scare-stories, and work harder to develop an expertise in their beats so they can put panic-mongering studies in context. Genuine experts who know better—and most do—need to speak out against agenda-driven junk science. And while there’s no need to write poems about studies that find normal cancer rates, there should be a commitment, from lab-bench to news-site, to giving them a fair hearing.

Otherwise the public conversation over health risks will continue to be warped by ignorance and myth. That has a high cost. There is no cancer surge from radiation, but there is a tsunami of anxiety and unnecessary thyroid surgery. Fear is the worst epidemic now plaguing Fukushima.