Cancer Alley

[ttweed@wildrockies.org (Tony Tweedale)]

>From: DrBillC@aol.com
>To: dioxin-l@essential.org
>Subject: Cancer Alley
>Message-ID: <961213090708_1920266561@emout07.mail.aol.com>
>
>Anyone see the recent article in the Journal of the Louisiana State Medical
>Society called "Is there a Cancer Corridor in Louisiana"? Since this site has
>been a locus for this type of dicussion I thought you'd be interested.
> Here's the abstract:
>
>"Cancer mortality rates in South Louisiana are higher than the national
>averages leading to the area's designation as a "cancer corridor".  This
>study was conducted to assess whether incidence data substantiate the
>reputation derived from mortality statistics.
>
>Age-adjusted cancer incidence rates for 1983-1987 were calculated for South
>Louisiana as a whole, for five regional divisions of it, and for the combined
>nine areas of the Surveillance, Epidemiology and End Results (SEER) Program.
> Significantly lower (p<0.0001) incidence rates were found in South Louisiana
>among white females, black males and black females for cancers of all sites
>combined; among women of both races for cancer of the breast; amopng men of
>both races for cancers of the colon and prostate; and among whites of both
>sexes for melanoma and rectal cancer.  South Louisiana incidence rates were
>significantly higher than the SEER rates only for lung and larynx cancers in
>white males.  The excess of lung cancer was significant only in the New
>Orleans area.
>
>The excessive mortality rates reported for South Louisiana are not the result
>of excessive incidence.  These results indicate poorer cancer prognosis in
>this region, a phenomenon that deserves more scrutiny by the health
>profession."
>
>Groves, Andrews, Chen, Fontham, Correa  J. La. State Med Soc 148, 155 (1996)


this abstract is a mess.  what are the results excluding the smoking
confounder; i.e. adjusting the data for the relative # of smokers in seer
data vs. s. louisiana?--this is how epi results are typically reported.
efforts at adjusting for differences amongst the 2 data sets in many other
known risk factors, e.g. diet for colon cancer should also be made.

wish i hadn't trashed another notice of this article, because i think it
said the database for the cancer alley cancer stats involved the *whole*
state.  anyway, i'd look real close at the method used to determine the
poupaltion studied for 'south louisiana'.  properly including just one or
two high cancer areas (or excluding one or two low cancer areas) could
easily reverse the results.

how many & which cancer types showed no signif. increase?

any analysis of possibility of endocrine toxins providing a protective
effect for endocrine associated cancers, as the epi studies indicate may be
going on at seveso?

p.s.: as far as i know, seer has 15 or so areas it collects comprehensive
data from, not 9, demographically representing over 10% of the us
population.