Recently a friend posted a link to the article “JANETTE D. SHERMAN, MD and JOSEPH MANGANO” – Is the Dramatic Increase in Baby Deaths in the US a Result of Fukushima Fallout? Calling it an article is a bit of an exaggeration, it’s basically just a blog post. However, because the authors are an MD and a statistician, people in the press apparently think this puts the post in the same category as an article that would appear in a medical or science journal.
Just to be clear, it’s not the same.
The authors look at some public data from the CDC and suggest it’s evidence that radioactive fallout from the Fukushima nuclear power plant in Japan is causing a 35% increase in infant mortality in the Western US. The crux of their ‘report’ and the entirety of their evidence:
The recent CDC Morbidity and Mortality Weekly Report indicates that eight cities in the northwest U.S. (Boise ID, Seattle WA, Portland OR, plus the northern California cities of Santa Cruz, Sacramento, San Francisco, San Jose, and Berkeley) reported the following data on deaths among those younger than one year of age:
4 weeks ending March 19, 2011 – 37 deaths (avg. 9.25 per week)
10 weeks ending May 28, 2011 – 125 deaths (avg.12.50 per week)
This amounts to an increase of 35% (the total for the entire U.S. rose about 2.3%), and is statistically significant. Of further significance is that those dates include the four weeks before and the ten weeks after the Fukushima Nuclear Power Plant disaster.
Where do I start?
They offer no compelling evidence, as far as I’m concerned.
To support their conclusions, they should address whether there are similar changes in the rate at other timepoints to rule out seasonal variations, stochastic changes in rate, etc. They present no information on their statistical analysis- just say the change is statistically significant. Details on their statistical analysis might help to rule out some of those variations, but since it is not provided, a reader cannot form an informed opinion as to the validity of their conclusion. Also, what about data from Japan? If the minuscule amounts of radiation in the Western US are causing such dramatic changes, their hypothesis would predict that the infant mortality rate in Japan, where the radiation levels are much higher, would be even worse than here in the US. They make no mention of seeking or analyzing any such data.
While those numbers are scary, we MUST remember, correlation does not equal causation!! Just because it rains every time you forget your umbrella doesn’t mean that you forgetting your umbrella causes it to rain. A friend of mine who is a professor of community psychology gives this example to demonstrate that point: “In my class I tell [my students] that there is a positive correlation between number of churches and amount of prostitution… this is a fact… I’ve [had students suggest] the priests are pimps, prostitutes are religious, etc… Oh, the joys of teaching! Then, they realize there is a third variable that “causes” both… as cities grow, number of churches increases, as does prostitution!”
What about the general biological feasibility of the hypothesis? The 10 week time frame is just too tight for the cause of the increase in infant mortality to be radiation. The radiation was released over several days/weeks after the earthquake and tsunami, then, the fallout would take several days to even reach the US- thus it wouldn’t make sense to examine infant mortality rates the week after the earthquake/tsunami. Further, the radiation levels that have been detected in the mentioned cities are minute. The chance that such a minuscule increase could so rapidly result in infant deaths is ridiculous. A lot of the exposure comes from ingesting contaminated foods. What are the chances that the Fukushima radiation entered the food chain, was eaten in by expectant moms and transmitted to their fetuses or eaten by infants in quantities sufficient to be hazardous in only 10 weeks? Pretty slim. This post on Boing Boing does a great job (with reputable, cited sources) of explaining the risk of fallout to the West Coast.
Blog posts aren’t peer reviewed. That ‘publication’ isn’t peer reviewed at all, just a little news blurb on a website. A rigorous review process, like that required for publication in scientific and medical journals, would probably point out the same issues I did with the hypothesis, methods, and conclusion.
The authors of the post pose the question, “Is the Dramatic Increase in Baby Deaths in the US a Result of Fukushima Fallout?” so I guess they aren’t really saying for sure it does. However, they make no effort to qualify their ‘report’ to discourage readers from putting undue weight on their faulty hypothesis and their unsupported conclusions.
As one might expect of the lazy press, the ‘article’ is being cited as a “report” by other news sources- and since the author is an MD reporters are acting like this is a properly conducted, peer-reviewed scientific study.
Northwest sees 35% infant mortality spike post-Fukushima
Medical professionals publish report highlighting post-Fukushima mortality spike.
It’s not a ‘report,’ conducted by medical professionals. It’s a blog post written by an MD and a statistician. It’s a question/hypothesis, and a poorly founded one at that. The news article above is just a regurgitation of the blog post- no attempt to vet the source, verify the conclusion, seek a supporting/dissenting opinion. It’s not journalism, it’s just plagiarism.
It’s irresponsible of that doctor to even write the blog post in a way that could be interpreted as factual- and not purely hypothetical. It’s also terrible that shoddy/lazy news agencies/sources are citing that article as a scientific report!
Edited to add: My husband e-mailed me a link to this response from Scientific American: Are Babies Dying in the Pacific Northwest Due to Fukushima? A Look at the Numbers by Michael Moyer. It does a better job than I have in enumerating the flaws of Sherman and Mangano’s “report,” including plotting the data for all West Coast cities for which data were available (not just the select few that support their forgone conclusion) showing no change in infant mortality rates.