About the SA Blog Network



Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

Are Babies Dying in the Pacific Northwest Due to Fukushima? A Look at the Numbers

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

A recent article on the Al Jazeera English web site cites a disturbing statistic: infant mortality in certain U.S. Northwest cities spiked by 35 percent in the weeks following the disaster at the Fukushima Daiichi nuclear power plant. The author writes that "physician Janette Sherman MD and epidemiologist Joseph Mangano published an essay shedding light on a 35 per cent spike in infant mortality in northwest cities that occurred after the Fukushima meltdown, and [sic] may well be the result of fallout from the stricken nuclear plant.” The implication is clear: Radioactive fallout from the plant is spreading across the Pacific in sufficient quantities to imperil the lives of children (and presumably the rest of us as well).

The article doesn’t link to the Sherman/Mangano essay, but a quick search reveals this piece that begins "U.S. babies are dying at an increased rate.” The authors churn through recently published data from the Centers for Disease Control and Prevention (CDC) to justify their claim that the mortality rate for infants in the Pacific Northwest has jumped since the crisis at Fukushima began on March 11. That data is publicly available, and a check reveals that the authors’ statistical claims are critically flawed—if not deliberate mistruths.

The authors gather their data from the CDC’s Morbidity and Mortality Weekly Reports. These reports include copious tables of data on death and disease in the U.S., broken down by disease (U.S. doctors reported 62 cases of rabies the week ending June 11) and location (30 of those were in West Virginia). Sherman and Mangano tally up all the deaths of babies under one year old in eight West Coast cities: Seattle; Portland, Ore.; Boise, Idaho; and the California cities of San Francisco, Sacramento, San Jose, Santa Cruz and Berkeley. They then compare the average number of deaths per week for the four weeks preceding the disaster with the 10 weeks following. The jump—from 9.25 to 12.5 deaths per week—is "statistically significant," the authors report.

Let’s first consider the data that the authors left out of their analysis. It’s hard to understand why the authors stopped at these eight cities. Why include Boise but not Tacoma? Or Spokane? Both have about the same size population as Boise, they’re closer to Japan, and the CDC includes data from Tacoma and Spokane in the weekly reports.

More important, why did the authors choose to use only the four weeks preceding the Fukushima disaster? Here is where we begin to pick up a whiff of data fixing. Though the CDC doesn’t provide the data in its weekly reports in an easy-to-manipulate spreadsheet format (that would be too easy), it does provide a handy web interface that allows individuals to access HTML tables for specific cities. I copied and pasted the 2011 figures from the eight cities in question and culled all data aside from the mortality rates for children under one year old. You can see those numbers in a Google doc I’ve posted here. (Note: Because I use the most recent report, my mortality figures are slightly higher than the Sherman and Mangano’s, as some deaths aren’t reported to medical authorities until weeks afterward. The small difference doesn’t change the analysis.)

Better still, take a look at this plot that I’ve made of the data:

Chart of infant deaths by week for eight U.S. cities in 2011

The Y-axis is the total number of infant deaths each week in the eight cities in question. While it certainly is true that there were fewer deaths in the four weeks leading up to Fukushima (in green) than there have been in the 10 weeks following (in red), the entire year has seen no overall trend. When I plotted a best-fit line to the data (in blue), Excel calculated a very slight decrease in the infant mortality rate. Only by explicitly excluding data from January and February were Sherman and Mangano able to froth up their specious statistical scaremongering.

This is not to say that the radiation from Fukushima is not dangerous (it is), nor that we shouldn’t closely monitor its potential to spread (we should). But picking only the data that suits your analysis isn’t science—it’s politics. Beware those who would confuse the latter with the former.

Photo by o5com on Flickr


Rights & Permissions

Comments 37 Comments

Add Comment
  1. 1. TechnoTigger 6:26 pm 06/21/2011

    Such a shame, the Fukushima incident is a big enough disaster already, without the abuse or careless use of statistics. It just muddies the water, and adds emotion to an already, and inevitably, emotion laden debate.

    (Doesn’t help the cause of stats either. They already have a bad enough press: "There are three kinds of lies: lies, damned lies, and statistics.")

    Link to this
  2. 2. onlinetoomuch 6:43 pm 06/21/2011

    I agree with you that the data pool was way to small, however, I wanted to point out that they did review the entire US and saw a 2.5% increase in the infant mortality rate.

    Link to this
  3. 3. ocifferdave 7:31 pm 06/21/2011

    Credit to Joviation at for clarifying that the increase is actually higher than 35% when compaired to same weeks in 2010. Darn! :(

    The second item in this Counterpunch article [not a pro-nuke mouthpiece] discusses the data manipulation. They had their own statistician “Pierre Sprey” review the selected set, and expand it.

    “Sprey reviewed the data and confirmed that if you accept the sample cities that Sherman and Mangano picked, and also if you accept comparing 4 weeks before the March 11 Fukushima disaster and 10 weeks afterwards, then Sherman and Mangano’s calculations are valid.

    One of the CounterPunch critics pointed out that using four weeks before and ten weeks afterwards “looked like cherry-picking the data.” To overcome this potential bias, Sprey collated the death numbers for the ten week period before, then did the calculations comparing infant deaths for ten weeks before and ten weeks afterwards for the same eight cities. His result was a statistically insignificant difference in deaths per week before and after- an increase of infant deaths of only 2.4 per cent. To further guard against the possibility of some seasonal effect due to comparing a period earlier in the spring with one later in the spring, Sprey also compared the ten weeks after Fukishima with the identical weeks in 2010. He found exactly the same result: a 2.4 per cent increase in infant deaths over the prior year which, given 128 deaths in the ten week sample, is entirely insignificant statistically.

    But then Sprey went further and looked at the Sherman/Mangano selection of eight cities from the 122 reporting to CDC….Sprey elected to look at smaller, geographically consistent groupings of cities. The results were striking.

    Simply by moving the boundary line northward from Santa Cruz Sprey found that the four northernmost Pacific Northwest cities in the CDC sample – Portland, Tacoma, Seattle and Spokane – show remarkably significant results – a larger infant mortality increase than the original Sherman-Mangano results.

    During the ten weeks before March 11 those four cities suffered 55 deaths among infants less than one year old. In the ten weeks after Fukushima 78 infants died – a 42 per cent increase and one that is statistically significant. To confirm once again that these results were not due to seasonality Sprey compared these infant deaths in the ten weeks after Fukushima to the deaths in the equivalent ten weeks a year earlier. (continued in the article link).

    Link to this
  4. 4. ocifferdave 7:32 pm 06/21/2011


    The results were almost identical with the ten weeks before Fukushima in 2011. Within the equivalent ten weeks of 2010 53 infants died in these four cities.

    The post-Fukushima deaths are 47 per cent higher than they were in the same period a year before – once again statistically significant. If you add Boisie, Idaho to the four city sample the results remain almost unchanged.”

    Link to this
  5. 5. ocifferdave 7:34 pm 06/21/2011

    BTW, you have to scroll down a screen or so to get to the baby death article that is on the same page as the Weiner article.

    Link to this
  6. 6. Pugsley 8:43 pm 06/21/2011

    They could have chosen all cities on the West Coast for which data was available, and they could have looked at data for a solid year.

    They didn’t. ‘Nuff said.

    Link to this
  7. 7. Pugsley 8:48 pm 06/21/2011

    By the way, I believe that Fukushima is an ongoing disaster, one that has been covered up enough so that few people know it’s much worse than Chernobyl already, and over the next century or so will add up to many times the radiation release of Chernobyl (which is also still emitting radiation). I’m not downplaying Fukushima at all! But this sort of easily-debunked thing doesn’t help the cause, it just makes concerned people look like tinfoilers.

    Link to this
  8. 8. episto 9:19 pm 06/21/2011

    This criticism is also flawed, as it assumes by omission that the level of radionuclides from Fukushima were deposited equally on all cities where data was collected. Without this as a variable in a study, any analysis is flawed. Oddly enough, the highest spike in infant mortality was in Pennsylvania, where the highest bequerel count was noted by the EPA.

    Go back and cross-reference EPA data on post-Fukushima monitoring and the mortality rates, and the study doesn’t seem as preposterous as portrayed here.

    Link to this
  9. 9. kukuiawana 10:56 pm 06/21/2011

    And what are the causes of these deaths? Acute death by radiation poisoning takes a huge dose, much larger than would reasonably have been expected given the distances involved. Lower doses can result in chronic problems (i.e. higher cancer rates) which will impact long term mortality, but it will obviously take a lot more time to see any impact on statistics.

    Has anyone proposed a reasonable mechanism by which sub-acute radiation exposure can increase infant mortality?

    Link to this
  10. 10. Martin Wirth 1:29 am 06/22/2011

    Based on CDC data from the Pacific region, average infant mortality since Jan 2007 was 27.034 with a slight upward trend of 0.007 deaths per week increase. The standard deviation of 6.074 suggests the trend that I calculated is not significant. In all of the deaths recorded since the radiation release of week 10 in 2011, there were three weeks, 13, 14, and 22 that were outside the standard deviation and these were below it at 16, 17, and 18 deaths respectively.

    While I’m convinced that boiling or pressurized water nuclear reactors (BWR, PWR) are a form of borderline insanity in their basic design, I don’t believe that cherry-picking data to make a political point is a good approach to solving the problem.

    In order to satisfy the growing energy requirements of humanity, we need to keep an open mind about ways to meet those demands. Solar and wind are intermittent and need backup storage. Having seen a few molten salt nuclear reactor designs, there are probably some good scalable designs that would be resistant to fire, have a negative reactivity coefficient, and not require massive amounts of external power just to cool the system on shutdown. Some refer to these variously as thorium salt reactors. These are concepts worthy of research because, unlike BWR and PWR reactors, they nearly eliminate the presence of oxidative and possibly combustible combinations of materials. This should be a essential requirement of any nuclear power reactor design.

    Attempts to exaggerate the effects of radioactive material releases from what was certainly a catastrophe are not a productive entry into the public discussion. There is enough nonsense on both the pro and con sides already to easy drown out reasonable ideas in the noise.

    Link to this
  11. 11. rwerkh 1:43 am 06/22/2011

    Your own graph shows some higher than normal figures post Fukushima.

    The trend line heading down adds to the significance of the higher peaks.

    The evidence from Chernobyl shows that fallout varies significantly between adjacent houses – let alone cities.

    The evidence from Fukushima and Chernobyl is that the fallout is not necessarily higher closer to the source.

    I am not satisfied yet that there is enough data to prove a link, but your arguments against it are mostly very weak and incorrect – one might question your truthfulness in putting up such arguments as logical proof and providing only a few more samples to try and prove lack of data.

    Clearly a year to year comparison is the only valid thing here – adding a few earlier samples is cherry picking.

    Why do I keep seeing bad science on these topics from those claiming the scientific higher ground – and always throwing accusations around at nuclear dissenters?

    Link to this
  12. 12. rwerkh 1:48 am 06/22/2011

    You are responding to something that said they did what you accuse them of not doing.

    Statistics is not always about sampling all data even if not relevant.

    The fallout pattern is not even – so the data won’t be. A larger sample can just hide the truth.

    I suspect you want the truth hidden.

    Link to this
  13. 13. rwerkh 1:53 am 06/22/2011

    The dose required for damage is minimal.

    I do not think we are discussing acute radiation poisoning.

    I believe we are discussing the deaths of infants that are triggered by the effects of the fallout.

    Are you really claiming not to understand the way radiation affects human cells when coming from an internal source?

    Link to this
  14. 14. rwerkh 1:56 am 06/22/2011

    When I logged in to post I believed this article disproved the claim – and intended to post about the flaws in it’s reasoning whilst still agreeing with it.

    Upon re-reading and reading the comments I realise that it does not disprove anything.

    Doubt in this situation is good – flawed logic to deny something is not good.

    Link to this
  15. 15. timbo555 11:01 am 06/22/2011

    No, Pugsly, they didn’t use all the cities on the West coast because they weren’t interested i Scientific integrity, they were interested in making a political statement in the guise of a scientific determination.

    Sounds remarkably like another "Scientific" body with the initials IPCC.

    Link to this
  16. 16. timbo555 11:31 am 06/22/2011

    Why would they even bother to do a study on deaths other than those already known to be caused by radiation? Nowhere in the article does it list the cause of death of any of these infants. The reason for this, and I may just be spitballing here, is because NONE of the children, NOT ONE, died of radiation sickness or anything remotely correlated to radiation.

    If there were a FIVE HUNDRED percent increase in infant mortality in those four cities but it had already been established they were caused by bed-bug bites, what purpose would it serve to do a study on the possible effects of radiation on infants in those cities.

    This article, like the study, is more leftist hyperbolic crap served up as "science". I keep reading though. I’m waiting for the difinitive peer reviewed study and subsequent SCIAM Article on the "Science" of Justin Bieber’s hair.

    Link to this
  17. 17. Soccerdad 11:57 am 06/22/2011

    Fukushima is an economic disaster, not a public health disaster, or even really an environmental disaster. There is no health effects for residents living outside the exclusion zone, much less across the Pacific.

    Link to this
  18. 18. Dr_Zinj 12:54 pm 06/22/2011

    No one has died from the Fukushima Daiichi disaster. No one has been seriously injured from the Fukushima Daiichi disaster. This was worse than Three Mile Island. This isn’t even in the same category as the Chernobyl Disaster. I get more radiation spending the evening in my basement in New Hampshire than was received by any of these babies from the F-D fallout.

    Fukushima Daiichi was an example of a test to destruction. The earthquake and tsunami was about as bad as you can get. Yet the plant mostly survived, is containable, and is economically cleanable. And this was a 1st generation power plant design that should have been retired and disassembled years ago.

    We have 3rd and 4th generation plans now. Much more efficient, much safer. What isn’t safe is the storage of spent fuel at all the reactor sites around the country. A better situation to cause a catastrophic disaster you could not deliberately plan. This is the direct fault of the anti-nuclear crowd, and the constant lobbying against the Yucca Mountain repository.

    Link to this
  19. 19. Michael Moyer 1:41 pm 06/22/2011

    It’s great to read all these incisive comments. Certainly the health effects of the Fukushima disaster–whatever they may be–will be studied for months and years to come (both in Japan and worldwide). As they should be. I didn’t mean to suggest the case was closed one way or another, only that this single "35 percent increase" statistic–which was beginning to get unthinkingly picked up and repeated by reporters who should know better–was specious.

    Link to this
  20. 20. Soccerdad 4:10 pm 06/22/2011

    Including Obama who installed as the head of the NRC a man whose sole intent is to shut down Yucca Mountain. He was recently found by the Inspector General as having concealed information in order to lead the NRC to a no-go decision on Yucca.

    Link to this
  21. 21. Lantzelot 4:52 pm 06/22/2011

    If you go and play with the data yourself you can probably find larger variations than 2.5% if you set the boundary conditions the way Sherman and Mangano did. And you can do it for any arbitrary year, you just need to find a suitable time frame (with or without a nuclear disaster to fit). Here’s a link for you, go ahead and play:

    Link to this
  22. 22. Lantzelot 5:01 pm 06/22/2011

    There is something wrong with the re-analysis done for Counterpunch, I have no idea where Pierre Sprey got it wrong, but we are several people who independently have checked it. No increase of infant mortality this time either.
    My version here:

    Version by Antonio Rinaldi (in Italian):

    One would think that the anti-nuclear movement would have enough material to work on after Fukushima, why do they feel that they need to make up fake effects in this way? Shame on Sherman and Mangano, they are a disgrace to the medical profession.

    Link to this
  23. 23. scientific earthling 8:33 pm 06/22/2011

    Martin: There is one point you make I cannot agree with: In order to satisfy the growing energy requirements of humanity-

    No we do not have to satisfy growing energy demands. As long as you do, they will keep increasing. The homo sapien is an incredibly greedy stupid animal.

    Link to this
  24. 24. kukuiawana 10:43 pm 06/22/2011

    All I’m saying is show me a viable mechanism by which such low doses would cause these deaths. I have trouble seeing how exposure to levels just above background would cause even statistically significant chronic effects, much less acute.

    Link to this
  25. 25. Pugsley 1:27 am 06/23/2011

    "No one has died from the Fukushima Daiichi disaster. No one has been seriously injured from the Fukushima Daiichi disaster. This was worse than Three Mile Island. This isn’t even in the same category as the Chernobyl Disaster. I get more radiation spending the evening in my basement in New Hampshire than was received by any of these babies from the F-D fallout."

    I urge you to update your numbers. This indeed seemed to be the case in the early weeks of the disaster, but you should have noticed something peculiar about your neighbor Massachusetts getting 80 times the safe limit of radioactive iodine in its rainwater, all those thousands of miles away! In fact Tepco has admitted at least twice now that they lied about the early figures in order to prevent panic.

    At this time, if you add the atmospheric fallout and the contamination of the ocean (which is accelerating), you will find that the total number of becquerels has bypassed Chernobyl, and the incident is ongoing – not contained in the least, not decreasing. Do a bit of homework tonight!

    Link to this
  26. 26. ocifferdave 3:07 am 06/23/2011

    Need help doing your "homework"? Then head on over to! Also, do a daily "fukushima" word search at Also try "fukushima, today" at youtube daily. But start at enenews. Tons of articles around the world…daily articles linked, lots of commentors on each article. Fun.

    Link to this
  27. 27. spitalfields 5:54 am 06/23/2011

    Thanks for setting the records straight with your analysis. When I first came across the Al Jazeera article, the apparently convincing statistics to support the claims made by Sherman & Mangano were shocking. I immediately sought to obtain the base data to verify their claims when I came across your article.

    It is a relief to learn that Scientific American, after all these years, can still be relied upon to uphold the standards of scientific journalism, and to swiftly dispel pseudo-scientific claims that are based on inaccurate use of data.

    Well done Michael.

    Link to this
  28. 28. spitalfields 5:54 am 06/23/2011

    Thanks for setting the records straight with your analysis. When I first came across the Al Jazeera article, the apparently convincing statistics to support the claims made by Sherman & Mangano were shocking. I immediately sought to obtain the base data to verify their claims when I came across your article.

    It is a relief to learn that Scientific American, after all these years, can still be relied upon to uphold the standards of scientific journalism, and to swiftly dispel pseudo-scientific claims that are based on inaccurate use of data.

    Well done Michael.

    Link to this
  29. 29. spitalfields 5:55 am 06/23/2011

    Thanks for setting the records straight with your analysis. When I first came across the Al Jazeera article, the apparently convincing statistics to support the claims made by Sherman & Mangano were shocking. I immediately sought to obtain the base data to verify their claims when I came across your article.

    It is a relief to learn that Scientific American, after all these years, can still be relied upon to uphold the standards of scientific journalism, and to swiftly dispel pseudo-scientific claims that are based on inaccurate use of data.

    Well done Michael.

    Link to this
  30. 30. Michael Moyer 3:56 pm 06/23/2011

    Thanks spitalfields, that’s very kind of you.

    Link to this
  31. 31. linlithgow 10:27 pm 06/23/2011

    kukuiawana 10:56 PM 6/21/11

    Good point. Also, you have to be careful about causality. What proof is there that the Fukushima disaster is responsible even if there is increased mortality, and the increase is not due to other factors? Unless you can show that some of the deaths are caused by increased exposure to radionuclides that Fukushima released, it’s just a guess.

    Link to this
  32. 32. linlithgow 10:53 pm 06/23/2011

    Now who is being disingenuous? The shift manager at the Chernobyl plant did not die until nearly two weeks after the disaster and he was in the bloody control room at the time, and died from acute poisoning. Louis Slotin, who was exposed to a fatal amount of radiation when an experiment with the demon core at Los Alamos went awry died over a week later.

    Yet I’m supposed to believe that the accident that occurred at Fukushima in Japan in mid- March was the direct cause of an increase in infant mortality in the Pacific Northwest, despite the fact that the EPA stated there was no dangerous increase in levels of radiation 10 days after the accident? I find Fukushima’s culpability difficult to swallow.

    Link to this
  33. 33. KaiGeologist 12:08 am 06/24/2011

    Low level radiation dose is not dangerous, it can even be good for your health…
    Quick doses under 100 mSv and longer term doses under some thousand mSv/year are not dangerous.
    Chernobyl, etc experience tells that.
    In comparison: Small sub-micrometer sized particles from coal, oil, wood, biomassa burning,.. are much more dangerous- millions of people die every year because of these…
    See more

    Link to this
  34. 34. mvpel 12:41 am 06/24/2011

    It’s not "just a guess," LinLithgow, it’s just shameless, base, reprehensible propaganda.

    What’s the motive this propaganda, I wonder?

    Well, from 1973 to 1987, nuclear power displaced 11.7 billion barrels of oil worldwide, and prevented $323 billion in US oil purchases. 11.7 billion barrels is $1,404,000,000,000 at $120 per barrel – $1.4 trillion, with a T. That’s a lot of money at stake against the growth of nuclear power. Hmmm…

    Link to this
  35. 35. pmagn 4:23 am 07/16/2011

    Btw there was an almost 50% jump in infant mortality rate in BC Canada for the first six months of the year over last year! And that’s official from the corners office reported on in the general media….

    Taking this into account there was definitely and incident on the west coast especially the NW. I think this debunking here does not actually cover the incident properly. What is the overall incident rate across each of the states from Washington down to California? I think you will find There was a spiking incident.

    This is not surprising and has been observed for other nuclear incidents previously not only in humans but animals as well.

    Link to this
  36. 36. pmagn 4:44 am 07/16/2011

    Here is some more on infant mortality radiation and NW birds…

    Link to this
  37. 37. udevilish 6:41 pm 06/3/2013

    I live in the pacific Northwest, and really the mortality rate is the least of it. I can think of 22 people diagnosed with “food allergies” myself included. Could it have possibly gotten into our food supply? I don’t see how it couldn’t, it must have in some way. Could this be the cause of the new found allergies… I have never been allergic to anything in my life. I have been here 4 yrs, I am now a celiac, lactose intolerant, allergic to bees, cottonwood trees, mold & birch. How? for 30 years I was all good.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article