SoylentNews is people
SoylentNews
Celiac disease linked to common chemical pollutants:
According to NYU Grossman School of Medicine researchers who led the study, people with the immune disorder have severe gut reactions, including diarrhea and bloating, to foods containing gluten, a protein found in wheat, rye and barley. The only treatment is a gluten-free diet, with no bread, pasta, or cake, says lead investigator and doctoral student Abigail Gaylord, MPH.
Reporting in the journal Environmental Research online May 12, the NYU Langone team found that children and young adults with high blood levels of pesticides -- and with high levels of pesticide-related chemicals called dichlorodiphenyldichlorethylenes (DDEs) -- were twice as likely to be newly diagnosed with celiac disease as those without high levels.
The study also found that gender differences existed for celiac disease related to toxic exposures. For females, who make up the majority of celiac cases, higher-than-normal pesticide exposure meant they were at least eight times more likely to become gluten intolerant. Young females with elevated levels of nonstick chemicals, known as perflouoroalkyls, or PFAs, including products like Teflon, were five to nine times more likely to have celiac disease.
Young males, on the other hand, were twice as likely to be diagnosed with the disease if they had elevated blood levels of fire-retardant chemicals, polybrominated diphenyl ethers, or PBDEs.
Abigail Gaylord, Leonardo Trasande, Kurunthachalam Kannan, Kristen M. Thomas, Sunmi Lee, Mengling Liu, Jeremiah Levine. Persistent organic pollutant exposure and celiac disease: A pilot study. Environmental Research, 2020; 109439 DOI: 10.1016/j.envres.2020.109439
(Score: 2, Insightful) by khallow on Tuesday May 19 2020, @07:05PM (2 children)
Is there anything you can't prove with a small sample size? Assuming at least two races, we have at least eight categories to further marginalize any statistical significance of this study. And of course, correlation doesn't imply causation. Let's read about the sample from the actual paper [sciencedirect.com]:
So first thing, this wasn't an unbiased sample. 88 (not 90) children/young adults came to this hospital for some reason which is ignored by the study. Second, "(OR = 13.0, 95% CI = 1.54, 110)" which is given without explanation means [ucla.edu]
Notice how huge those intervals are and how close the bottom is to 1. This is meat and drink for the anti-NHST (null hypothesis significance testing) guy. Where is he?
(Score: 1, Informative) by Anonymous Coward on Wednesday May 20 2020, @09:57AM (1 child)
They went there because they thought they had digestive problems. That is not in itself a problem.
Sadly, this is what happens when you get studies like these and especially without access to the actual paper. But yes, 90 is not very many subjects. This does not mean there is no effect but you need to get more data. The unfortunate thing is that then people jump to conclusions.
This reminds me of a research paper from Taiwan by some anti-nuclear folks. You could actually look in the author's history and just find more of these useless papers. He had things like counting incidence of cancer in buildings that had Cobalt-60 source in the steel by accident. The total number of cancers were actually down by 2 standard deviations (something like 45 vs. 64 expected) and the only cancer that was slightly increased was leukemia. I think it was 2 or 3 standard deviations above background incidence. But what was his conclusions? That radiation causes (I make up numbers here, but you'll see uselessness of comparisons here) eye cancers (3 vs. 0.2 expected) or ovarian cancer (4 vs. 1.2 expected)... and of course you can't do that. Might as well say that radiation prevents penile cancer because there was none detected but 0.1 was expected :)
Anyway, most researchers are rather bad at this. They have pressure to have results which means they should have paper published, even if they don't have enough data to draw conclusions some conclusion must be drawn. So basically, this is probably not a bad paper but you probably can't draw conclusions from it like in the article. You need MORE data and research like this should be more prioritized.
PS. khallow is not a troll here but he's getting trolled by moderators that maybe have expectations bias about results of science.
(Score: 0) by Anonymous Coward on Wednesday May 20 2020, @10:38PM
Not a troll, but doesn't know enough to launch the criticisms he thinks he is. The most obvious example is "how close the bottom is to 1." On the scale that he is referencing, an OR of 1.54 is farther away from 1 than an OR of 110 is from 167. Another is exactly what you point out about this being a pilot study. They are designed differently than the kind of studies he is thinking about because they have a different purpose.