7 ms·
This paper came up as a pre-print. You can't make the extrapolation that the headline is making - they're using gas chromatography to estimate quantities from 1
by timr 2y ago
This paper came up as a pre-print. You can't make the extrapolation that the headline is making - they're using gas chromatography to estimate quantities from 1-2mg samples, and then extrapolating to get to these scary sounding whole-organ estimates. If you look at the paper [1], you'll see that the microplastics in in situ samples are not discernible by light microscopy, and that there was a ~25% variation in within sample measurement of the GC [2], indicating a great deal of uncertainty in the precision of the fundamental measurement (the authors brush this off; see quote below).
Basically, you've got an extremely sensitive measurement system being used to make tiny measurements, and then they extrapolate these measurements by a huge factor to get to ug/g estimates. Further extrapolating (to the weight of an organ, say) when you know that there's 25% inter-sample variation, is just guaranteed to be nonsense.
[1] https://www.nature.com/articles/s41591-024-03453-1 https://www.nature.com/articles/s41591-024-03453-1
[2] "Both analytical laboratories (UNM and OSU) observed a ~25% within-sample coefficient of variation, which does not alter the conclusions regarding temporal trends or accumulation in brains relative to other tissues, given the magnitude of those effects."
- sympil 2y ago[flagged]
- idiotsecant 2y agoObviously parent post is not making the point you're trying to argue, only making the point that the measurement is flawed. If you are going to address their argument do it in good faith and address the strongest point. This isn't reddit
- pinkmuffinere 2y agoThankyou. I think I will quote your last two lines frequently.
- squigz 2y agoOr just quote the guidelines https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html > Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith. And > Please don't post comments saying that HN is turning into Reddit. It's a semi-noob illusion, as old as the hills.
- idiotsecant 2y agoI'm dying on that hill.
- sympil 2y agoI did address their point. The tendency with climate change, micro plastics, etc. is to deny, deny, deny, until it is too late. Perhaps we should err on the side of caution. In this situation its much better to err on the side of being overly cautious. Experts wrote the research paper. Until otherwise demonstrated I will give them, the experts, the most generous interpretation per site guidelines. The experts addressed the concern by OP but OP calls it “brushing off”. People can nitpick anything. The question remains, just how much microplastics are OK? If the answer is small then the study has merits despites its alleged issues.
- throwaway519 2y ago[flagged]
- morpheos137 2y ago[flagged]
- rthomas6 2y agoWell, it still tells us something. What are the upper and lower bounds of whole brain microplastic content, given that 25% variation?
- timr 2y agoI couldn't begin to tell you [1]. It's not a 25% variation, once you've extrapolated from the samples by 10,000x (or whatever). The 25% inter-sample error was on a few replicas of teeny tiny measurements. The post-extrapolation error bars are so wide that they're meaningless. The Smithsonian magazine article is garbage. Ignore it. The paper is saying that they see longitudinal trends in plastic bioaccumulation in various cadaver tissues, and this is plausible. But no, you don't have a plastic spoon in your head. That's just panic porn. [1] Actually...I just asked Gemini and it reminded me that the expected variance of a draw from a distribution, scaled by factor C, should have a variance of the scaled sample distribution that grows with C^2. So, if we scale up the measurement by a factor of 10,000, we'd expect the variance of the scaled estimate to be proportional to 10,000^2.
- drawkward 2y agoWhy yes, i think i will believe the random internet commenter.
- LoganDark 2y agoPractically every internet commenter is random.
- drawkward 2y agoWhich is fine in plenty of contexts. Rebutting this particular article? Nah.
- robertlagrant 2y agoEvaluate things based on reasoning, not source.
- userbinator 2y agoIMHO the more important part is they used pyrolysis gas chromatography, which breaks down all polymer chains. Besides man-made plastics, guess what else has long hydrocarbon chains, occurs naturally in humans and other biological matter, and behaves similarly under pyrolysis... https://en.wikipedia.org/wiki/Fatty_acid https://en.wikipedia.org/wiki/Fatty_acid Here's an interesting related article: https://www.oaepublish.com/articles/jeea.2022.04 https://www.oaepublish.com/articles/jeea.2022.04 Analysis of saturated, monounsaturated and polyunsaturated fats was demonstrated to form the same pyrolysis products as PE
- timr 2y agoTo their credit, they do discuss this in the results. I suspect the reviewers had the same concerns. Their response is not especially convincing, IMO, but they do at least discuss it.