Nuclear Blues

November 3, 2003 3:16 PM

Nuclear plant operation correlates with increased infant mortality rate. Correlation may not prove causation, but these numbers are pretty dramatic.
posted by alms (19 comments total)

This is so flawed I don't know where to begin. First, they are clearly cherry picking counties for what constitutes "downwind." Second, they are using absolute numbers instead of rates. Without knowing population statistics, there is no way to know whether those counties just didn't have less babies in the years that the plant was shut down. They only consider one plant, instead of multiple plants, to get more statistically valid data, and they provide no plausible means of causation. Nuclear plants produce a very tiny ammount of radiation under normal operation -- far less than a coal plant even.

If people want to be anti-nuke, try using some valid arguments, such as cost, difficulty of decommissioning, waste storage, or safety in extraordinary events. Personally, I'm in favor of nuclear energy, because I believe the cost and difficulty of decommissioning is worth it, and the waste problem can be solved without too much trouble using either breeder reactors or accelerator based inerting. The huge advantages you gain are no air pollution, no CO2, and a nearly unlimited supply of fuel which doesn't come from the middle east.

Environmentalists have to recognize that generating energy is not something that can be done wholly without compromise. We should strive to use energy sources that provide the best overall solution, including environmental, safety, cost, and availablity of fuel. If you look at the alternatives, nuclear comes out looking pretty good.
posted by cameldrv at 3:41 PM on November 3, 2003

Suppose being downwind lowers property prices. Suppose this leads to more low-income families in downwind areas. Given a correlation between socio-economic status and infant mortality, you would see exactly this, no radiation required.

Should we take action against railways because of the higher infant mortality on the wrong side of the tracks?
posted by i_am_joe's_spleen at 3:58 PM on November 3, 2003

Suppose the type of radiation from power plants is so toxic that any exposure no matter how small is enough to cause mortality in infants. There is no way to test for this except observation. Even if the study is flawed it deserves more attention from the Nuclear industry than to just brush it off as being impossible, that is the disconcerting part of the article.
posted by stbalbach at 4:08 PM on November 3, 2003

Suppose the type of radiation from house plants is so toxic that any exposure no matter how small is enough to cause mortality in infants. There is no way to test for this except observation. Even if the study is flawed it deserves more attention from the Nuclear industry than to just brush it off as being impossible.
posted by sonofsamiam at 4:18 PM on November 3, 2003

sonofasamiam: "Pfft. Facts. You can use them to prove anything."
posted by keswick at 4:57 PM on November 3, 2003

posted by quonsar at 4:59 PM on November 3, 2003

cameldrv: how do you conclude that the counties are cherry-picked? I agree that could be done and it would be a problem, but I don't see evidence that it was done. Same question regarding number of deaths versus death rates. That's certainly not the case in this related study.

stbalbach, you should pay more attention before you move your fingers. The rates went up when the plant began operating, they went down when the plant stopped operating, then they went back up again when the plant started operating again. That's got nothing to do with property prices or income levels.
posted by alms at 5:00 PM on November 3, 2003

cameldrv: how do you conclude that the counties are cherry-picked?

I am not cameldrv, but I play him on TV:

Which ones are downwind? Unless the relevant part of Illinois is somehow blessed with unidirectional wind, all of the counties nearby will be downwind some of the time. Are the "downwind" ones downwind 90% of the time? 55% of the time? 19% of the time? What happened in the counties that are nearby but not "downwind?"

The rates went up

As cameldrv notes, the studies don't seem to address rates but only the absolute number of infant deaths. If there are more kids born in a given year, you'd expect more infant deaths, and so on.
posted by ROU_Xenophobe at 5:34 PM on November 3, 2003

Suppose the type of radiation from power plants is so toxic

Are you seriously positing that power plants somehow emit something other than alpha and beta particles and gamma rays, and maybe neutrons? That power plants somehow manage to create forms of radiation otherwise unknown or somehow unlike the radiation given off by every other form of radioactive decay known to man?
posted by ROU_Xenophobe at 5:40 PM on November 3, 2003

Googling for "coal power plant radiation", I found several sources which back my recollection that coal-fired plants release much more radiation than a normally operating nuclear plant. Given that, I'd like to see a study looking at infant mortality 'downwind' from all types of power plants. If the cause is radiation, then it's quite possible that other types of plants cause more disease than nuclear. On the other hand, perhaps the deaths have nothing to do with radiation. Perhaps they don't even have any relationship to power plants. I don't think that being against nuclear power "just in case" is a very valid response to this one study. Not unless there is a viable replacement for nuclear power that definitely solves the intended problem.
posted by jupiter at 5:41 PM on November 3, 2003

We should just use cold fusion power plants.
posted by The God Complex at 5:56 PM on November 3, 2003

If nuclear power plant operation --> infant mortality, then please explain Maggie Simpson.
posted by stonerose at 7:20 PM on November 3, 2003

A lot of the criticisms are good (as in "good reason for more study", not so much "good reason to dismiss the thing out of hand"), but I saw a lot of percentages, not a lot of absolute numbers in the news story. It's possible infant deaths went down (say, more people getting laid off when the plant is closed and deciding not to have kids), but would't the mortality percentage likely stay the same, if that's all there is to it? Perhaps one of the critics can link us to the actual study? I'm all for tearing apart studies, but not based on their "journalistic" representations.
posted by uosuaq at 8:04 PM on November 3, 2003

but I saw a lot of percentages, not a lot of absolute numbers in the news story

From what I can gather from the article, you're being mislead. The percentages seem to be increases in the absolute number of mortalities.

Case in point:
The same counties saw a 38 percent decrease in infant mortality when the Clinton plant was off-line for three years, falling from 145 cases in 1993-1995 to 88 in 1996-1998, when it was not operating.

But they don't indicate how many babies were born in those counties in the relevant years, which they need to.
posted by ROU_Xenophobe at 9:22 PM on November 3, 2003

Ah. Looks mostly like a stupid reporter. A press release up at http://www.radiation.org/closed.html (sorry, link button busted and I am lazy) shows that the actual study has rates, not absolutes.

But still, two things: first, ya gotta have some relevant causal mechanism, and they can have two. First, they can think that a really very trivial increase in radiation exposure is doing it -- but then we'd expect infant mortality to correlate strongly with geology (ie, people living on granite should die a lot). Or they can think that everyone is lying about the amount of radiation released from nuke plants.

Second, it's not the absolute safety that's important, it's the relative safety. Even if nuke plants are not perfectly safe, the relevant (realistic) question is whether or not they're safer than coal plants. If they remain safer than coal plants, then replacing a coal plant with a nuke plant is a good thing. Replacing a coal plant with a wind turbine would be good too of course; I just dunno that you can get mutliple terawatts of electricity out of it.
posted by ROU_Xenophobe at 10:00 PM on November 3, 2003

Nuclear plants produce a very tiny ammount of radiation under normal operation -- far less than a coal plant even.

I'd love to be able to trot this fact out. Do you have a source for this?
posted by bshort at 4:39 AM on November 4, 2003

So, babies must be dropping like flies in France?
posted by jfuller at 6:10 AM on November 4, 2003

Okay, time for the facts master to work his magic!

The measuring unit used below for comparing radiation dose is the millisievert (mSv) per year.

20 mSv/a: the Canadian limit for nuclear workers averaged over a period of 5 years.
3-5 mSv/a: is the typical dose rate (above background) received by uranium miners.
3 mSv/a: is the normal background radiation from natural sources in North America.
2 mSv/a: is the mean normal background radiation from natural background in Canada.
0.5 mSv/a: is a typical range of dose rates from artificial sources of radiation, mostly medical

Public dose limits due to licensed activities (NRC) 1 mSv/year

Some other reading, for your pleasure.

Risk / days of life expectancy lost by taking the risk
One diet drink/day / 2.5
Consuming nuclear electricity (if it were all nuclear) / 0.05

posted by shepd at 7:41 AM on November 4, 2003

You would also need to see the data for the past few decades to see whether the changes the researchers are detailing are possibly cyclical in nature, or may be related to another external explanation the researchers did not document or account for. That's why this type of research is fairly useless, because it doesn't adequately look at alternatives that could explain the findings, or ensure the findings are robust across time and geography. (For instance, did public health funding for child birth procedures increase or decreases during those same years; are the deaths correlated with SES; are the deaths correlated with anything else in the region; etc. etc.)
posted by docjohn at 7:50 AM on November 4, 2003