Anorxia (was: CSICOP)

[Hugh Haskell <hhaskell@MINDSPRING.COM>]

At 22:43 -0500 8/9/01, brian whatcott wrote:
> On the other hand, astrological - er - I mean seasonal effects,
> are being noticed in schizophrenia and some eating disorders,
> according to this New Scientist piece:

This is an interesting article, and could point to something
important. Unfortunately, many of these studies have a tendency to be
non-replicable, and after a lot of initial publicity, sometimes
enough to make it part of the lay pubic's general fund of
"knowledge," nothing is heard of it again. I would guess that the
hypothesized link to illness of the mother during pregnancy would be
devilishly hard to get a handle on, due to the long lag time between
the pregnancy and the onset of the syndrome. How many mothers
remember if they were sick during a particular pregnancy 15 or so
years ago?

I would guess that a rather extensive longitudinal study of currently
pregnant women, which would gather their medical history during the
pregnancy, and then later correlate that with the incidence of
anorexia among their daughters. It would also be important to gather
data on girls born in all months. How do you blind such a study? I
sure wouldn't want to encourage people to diagnose anorexia when it
wasn't there (or the converse either, for that matter).

I wonder who would be willing to fund such a study, which would take
15-20 years to complete. Can you imagine hanging round as an
unemployed grad student for 15 years, waiting for your test subjects
to get old enough to start showing the symptoms you are looking for?

Hopefully, someone will come up with a better way.

I also wonder if the anomalous brain scans were there before the
girls developed anorexia, and thus could be related to its onset, or
if the anorexia was there and caused the anomalous indications in the
scans. Which is cause and which is effect, if either?

I also recall reading an article in SI some years back about how
weakly corrrelated data could be "over-analyzed" to show
periodicities when none are actually there. Don't remember the
details and I was never much good at statistics, so I don't know if
that sort of thing might be relevant here.

Hugh
--

Hugh Haskell
<mailto://haskell@ncssm.edu>
<mailto://hhaskell@mindspring.com>

(919) 467-7610

Let's face it. People use a Mac because they want to, Windows because they
have to..
******************************************************