Comments on: More on DeSoto and Hitlan (2007): Corrections and Clarifications http://epiwonk.com/?p=130 Epidemiology, Health, and Medical News Media Watchdog: A Blog for the General Public Fri, 10 Sep 2010 10:54:54 +0000 http://wordpress.org/?v=2.5 By: Alondra http://epiwonk.com/?p=130#comment-778 Alondra Mon, 10 Aug 2009 06:56:40 +0000 http://epiwonk.com/?p=130#comment-778 <strong>Alondra...</strong> this is why i always pick my nose.... Alondra…

this is why i always pick my nose….

]]> By: Alysha http://epiwonk.com/?p=130#comment-775 Alysha Mon, 10 Aug 2009 06:00:11 +0000 http://epiwonk.com/?p=130#comment-775 <strong>Alysha...</strong> an interesting take on a fun subject.... Alysha…

an interesting take on a fun subject….

]]> By: Ken http://epiwonk.com/?p=130#comment-438 Ken Sun, 27 Jul 2008 05:14:24 +0000 http://epiwonk.com/?p=130#comment-438 While the stats aren't perfect, I don't have much disagreement with their marginally significant p-value. However as you show it means there is little difference in mercury levels between the two groups showing that it could never be considered a major cause of autism. What is more important and should have been addressed in the papers are the non-statistical issues. Like any of these analyses there are a series of assumptions which I think are fairly dubious. Case-controls involve fairly stringent assumptions and it doesn't take much to introduce bias. All association studies have the risk of some unidentified factor inducing the association. I suspect that parents who take their children to an autism clinic are slightly more likely to feed their children fish. The big one though is the study measures mercury in subjects already diagnosed, possibly many years before. Is there any relationship between mercury tehn and now ? While the stats aren’t perfect, I don’t have much disagreement with their marginally significant p-value. However as you show it means there is little difference in mercury levels between the two groups showing that it could never be considered a major cause of autism.

What is more important and should have been addressed in the papers are the non-statistical issues. Like any of these analyses there are a series of assumptions which I think are fairly dubious. Case-controls involve fairly stringent assumptions and it doesn’t take much to introduce bias. All association studies have the risk of some unidentified factor inducing the association. I suspect that parents who take their children to an autism clinic are slightly more likely to feed their children fish.

The big one though is the study measures mercury in subjects already diagnosed, possibly many years before. Is there any relationship between mercury tehn and now ?

]]> By: False Controversy: Autism and Vaccines http://epiwonk.com/?p=130#comment-429 False Controversy: Autism and Vaccines Fri, 25 Jul 2008 21:37:25 +0000 http://epiwonk.com/?p=130#comment-429 [...] is a recent post by Epi Wonk on the 2007 DeSoto and Hitlan study, Blood Levels of Mercury Are Related to Diagnosis of Autism: A [...] […] is a recent post by Epi Wonk on the 2007 DeSoto and Hitlan study, Blood Levels of Mercury Are Related to Diagnosis of Autism: A […]

]]>