Daily Archives: February 28, 2009

Public Health and Nuclear Power

Right below this piece, you’ll see an event announcement.  

Today’s Rutland Herald discusses this event with a little detail on the science behind it:

A study of cancer occurring in children living near Germany’s 16 nuclear power plants show that the reactors pose a serious risk to young children, a German pediatrician said Thursday.

Dr. Winfrid Eisenberg of Herford, Germany, has been speaking throughout Vermont for the past two weeks, citing a German study in 2007 that showed a definite link between nuclear power and childhood cancer, particularly leukemia.

What’s being referenced here is the KiKK study, and I strongly encourage people to read about it, discuss it, and think about it.  It’s highly technical and several hundred pages long, and in German, so I don’t expect everyone to read the whole thing (I haven’t), but the data itself is interesting.  There’s a good background site on the study here.

After the fold, I’ll be explaining a little bit about the scientific method and how it applies to this study.

Okay, I’m going to summarize a little here, but still try to get the gist of it:

On a basic level, we have three tools to use in assessing the nature of a scientific phenomenon.  One is case studies, which are detailed interviews of individuals or small groups.  They can’t be used to assess cause and they can’t be used to generalize, but they can be used to get a lot of detail which can give us ideas about other approaches.

Correlational research (such as surveys) will allow you to look at collections and links between a variety of factors.  A lot of the research on nuclear power and health effects is correlational.  A problem with this is that we can’t really determine cause from correlational work.  We can see that two factors tend to occur in coincidence with one another but we don’t know if one causes the other.

The example I use in my classes is this:

It’s widely known that there’s a high correlation between churches and bars in a given community.  I.e., the more bars you have, the more churches you have.  Now– this can be explained several ways.  It could be that churches are repressive and drive people to drink.  It could be that bars get more people to drink and do things they regret which drives them to churches.  

The more likely explanation however is that neither causes the other but simply that the larger a population you have the more of a perceived need there is for both factors.

So, in reality, correlations can’t really prove anything other that some sort of connection exists.

We’re still in that place.  The KiKK study tells us that the closer people live to power plants and that the younger those people are, the higher the incidence of leukemia is, but it doesn’t tell us that the nuclear power causes leukemia to nearby residents.  

The third approach is experimental design, which is actual manipulation of factors in order to see what changes.  I.e., (we would never do this, so it’s just an example) intentionally increasing the radiation levels in some homes to see how it impacts the health of its inhabitants vs. those whose radiation levels haven’t been increased.  In this particular research we can actively assess cause because we’ve changed something in order to see what factors are affected by that change.

Now let me get back to the KiKK study: what it can’t do is say for certain that nuclear power plants cause leukemia.   What it does do, however, is make a very compelling case that the power plants are the likely culprit.  The populations they investigated are small (which means that other factors could skew the data a bit) but leukemia is very rare and the number of children with leukemia in residence near nuclear power plants more than doubled what would have been expected based on statistical averages.

One thing I want to caution people about when discussing this: please don’t use this study to claim that nuclear power causes leukemia.  We don’t know this.  We suspect it, but we don’t know it.  What we do know is that nuclear power plants tend to have more leukemia surrounding them than places without power plants, but a power plant is an extremely complex system and its effect on the surrounding environment is not limited in itself to simply the nuclear materials.

All this provides context for surrounding issues: the decommissioning fund, whether or not to extend the license, etc., but in the meantime, we’ve got some fascinating research which bolsters the case against using nuclear power, but is not conclusive.

Science is a process, and one that teaches us over time if we allow it to.  This is good science and fascinating research, but for anyone who thinks that they can grasp the gist of this research from a six-paragraph news piece, please yourself a favor: if you use it to argue for a position, take some time to investigate it, read about it and understand the science before you just jump in and do yourself more harm than good by arguing for something you don’t understand.