6 ms·

I don't really trust this kind of science reporting. The link to the nature article doesn't even link to it for me: https://10.0.4.14/s41586-021-04075-0 https:/

I don't really trust this kind of science reporting. The link to the nature article doesn't even link to it for me: https://10.0.4.14/s41586-021-04075-0 https://10.0.4.14/s41586-021-04075-0, the real link is: https://www.nature.com/articles/s41586-021-04075-0#citeas https://www.nature.com/articles/s41586-021-04075-0#citeas I also cannot access the study on SciHub, but one should keep eyebrows raised with alarmist reporting like this. Palmitic acid is in everything. Ask questions like: 1. What's the strength of effect? Is it 50x more likely to see metastasis? 15%? 2. Is the diet a realistic diet? Are the mice or rats eating a significant amount more of palmitic acid than what would be reasonable in a human participant? 3. How does the mouse model differ from humans? 4. What are the systemic effects? People don't eat fatty acids alone, they eat them as (ideally) part of a diet composed of whole foods. Might the effect hold in a realistic diet with more palmitic acid? ... and so on and so forth. Without access to the actual article and some folks' willingness to critically look at methodology, this kind of writing just encourages people to draw simplistic conclusions about incredibly complex systems.

Ginden 5y ago
> What's the strength of effect? You are acting like anyone cares about effect sizes. As civilization, we transcended past such nuances, you can write alarmistic abstract about 3% effect size, as long magic p < 0.05.
- jjoonathan 5y ago
  My favorite in the breathless overreporting genre is when debate wins are described as "evisceration."
CoastalCoder 5y ago
I never know how to interpret the phrase "linked to" in various news headlines. Is it just shorthand for some relationship that's truly significant and meaningful in the context of the article? Or is it a weasel-phrase to click-bait people into reading a less significant finding and/or to impute guilt by association?
- gregwebs 5y ago
  In 99% of the news headlines you hear about the underlying study does not contain causal evidence in humans. It is almost always correlational in humans or a mouse model. So yes. Generally weak evidence that can conform to some existing view an organization holds is cherry-picked into the news. Although sometimes it is just in the news because it is odd/interesting.
chiefalchemist 5y ago
Good context rules. However, what always concerns me about 1 and 2 is that in our environment there are toxins of all types. Yeah, a penny here, a dime there, a nickle here again. Those all add up. And since there's no way to sort that out, each of those individual toxins gets a free pass for being an insignificant amount. Unfortunately, we can to be mindful of the sum of those toxins being greater than the individual parts. So now what? Is it reasonable and safe to ignore that context? And if we have kids?
- Tagbert 5y ago
  But “toxins” don’t all have the same effect on the same biology. You can’t just assume that they all contribute to the same disease. Some of them may cancel others out. Even the word “toxins” is so much abused in the popular medical press as to be not only useless but misleading. Any report like this needs to be evaluated for various factors, but one of the most important is the degree of effect. Some studies have identified effects that are are so small you would need to consume gallons of the product a day to see any difference from not consuming. The press is full of these scary sounding reports based on a study that found something that the report fails to explain has not been verified. Any one study should not be the basis of your life decisions. This would need to be verified or debunked by other studies until a consensus gives us enough information to determine how to act if needed.
  - chiefalchemist 5y ago
    > You can't just assume... That is the point. They are all able to hide behind a cloud of doubt and lack of clarity. That's the spin. What's not addressed is the possibility aggregate - or worse, exponential - effect of each individual "Not us our data says becoming something far more (read: worse). So A, B and C do their data. But reality is not a silo. Those things combine. Who is addressing that? Just because that's extremely difficult doesn't mean it doesn't exist, doesn't matter. Keep in mind, as I understand it. A company can put a new chemical in a product without approval. It's only after the fact if it's proven unsafe will it be taken off the market (and perhaps only in select countries). Conclusion: Yes, it easy to say "It's only X amount of B." But it's not only B we're being exposed to. It's a whole bleepin' 1,000 letter alphabet of unknowns. No one should be comfortable with that. Yet we are.
frgtpsswrdlame 5y ago
Your rules are good but this: >Palmitic acid is in everything. Should have no effect on our judgement regarding its safety. Probably the story of our times seen from the future will be how we failed to account for extremely widespread use of very minorly dangerous products. See stuff like microplastics which took a couple decades to go from research to ban. Take the famous example from Fight Club: >The car crashes and burns with everyone trapped inside. Now, should we initiate a recall? Take the number of vehicles in the field, A, multiply by the probable rate of failure, B, multiply by the average out-of-court settlement, C. ... If X is less than the cost of a recall, we don't do one. Except we've moved even one degree from that now, think of a variable Y which is the cost to even figure out what B is. If you're a company selling a dangerous product you don't just have to consider X, you can also factor in Y as a proxy for long/hard entities will have to fight just to figure out the harm you're doing. So we should think about both sides of the coin, that the research is good enough to show there is an effect but also that we're not increasing Y to completely unreasonable amounts.
- smsm42 5y ago
  It's not everything like microplastics. It's everything like everything. Palmitic acid is one of the most common fatty acids. Cocoa butter is 25% PA. Cow milk is 30% PA. Chicken is 20% PA. In human milk, 25% of its fat content is PA. Olive oil has tons of it too, so do soy oil and sunflower oil (and probably any vegetable oil you've eaten). But if you don't eat anything at all, human body is still about 2% PA - as I said, it's a very common chemical compound. To quote Wikipedia: Palmitic acid, or hexadecanoic acid in IUPAC nomenclature, is the most common saturated fatty acid found in animals. If it were dangerous for animals, there would be no animals left living. It's not like microplastics that are exogenous. It's part of what living organisms are made of. Instead, what happened is probably some kind of p-hacking - they gave this compound, lots of it to be sure, to lots of poor mice to which they gave cancer previously, and in some of them the cancer got worse, because it's cancer. So they wrote down which cancers those mice had, and declared that they linked the compound to "spread of cancer in mice". Easy peasy. Sources: https://www.sciencedirect.com/science/article/abs/pii/S030881469700160X https://www.sciencedirect.com/science/article/abs/pii/S03088... https://www.sciencedirect.com/science/article/pii/S0952327819301279 https://www.sciencedirect.com/science/article/pii/S095232781... https://orbi.uliege.be/bitstream/2268/240227/1/fatty-acids-composition-of-meat-of-five-native-chicken-gallus-gallus-ecotypes-of-benin-reared-under-organic-or-conventional-system.pdf https://orbi.uliege.be/bitstream/2268/240227/1/fatty-acids-c... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5682332/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5682332/
koprulusector 5y ago
It’s embarrassing The Guardian posted a link to a 10. Address. :facepalm
- dec0dedab0de 5y ago
  I wonder how that is even possible. I just checked, they have different parent companies. Maybe that is an internal proxy/cache address? Possibly for getting around the paywall? Or maybe they paid/arranged for an internally proxied SAML connection?
  - r00fus 5y ago
    It's clearly a mistake - either the article was downloaded onto the author's network (and internal link provided) or simply the wrong URL.
    - smnrchrds 5y ago
      It helps that if you are not paying attention, it looks like a DOI.
  - aaron695 5y ago
    Somehow 10.1038 resolves to 10.0.4.14 -> https://www.google.com/search?q=1038+%2210.0.4.14%22 https://www.google.com/search?q=1038+%2210.0.4.14%22 Puzzle for the day. (If you don't know DOI's - https://en.wikipedia.org/wiki/Digital_object_identifier https://en.wikipedia.org/wiki/Digital_object_identifier important for looking up science articles, SciHub is one example. It might not be mathematical, it might be a API - https://doi.org/10.1038 https://doi.org/10.1038 )
- labster 5y ago
  The Sun posted a 10 too, but she was on page 3.
em500 5y ago
It mostly reminds me of this old Vox article[1], which was nicely summarized in the caption of its second graph: "Everything we eat both causes and prevents cancer". [1] https://www.vox.com/2015/3/23/8264355/research-study-hype https://www.vox.com/2015/3/23/8264355/research-study-hype
- wyclif 5y ago
  If this is junk science, then where is reliable medical information on the use of vegetable oils in cooking?
ravenstine 5y ago
> 2. Is the diet a realistic diet? Are the mice or rats eating a significant amount more of palmitic acid than what would be reasonable in a human participant? I'll have to go digging for it, but I want to comment on something this reminds me of. Someone a few months ago linked me to a study that showed that MCT oil caused weight gain in mice (might have been rats?), despite how MCT oil is known to be much less fattening than other fats based on how it's metabolized. If eaten in great enough abundance it may even cause weight loss. Having learned my lesson from just relying on the abstract, I found the full text for the study using Sci-Hub; the diet fed to both groups of mice was a high sucrose diet, the experimental group having more MCT oil than the other. The experimental group had less weight gain than the control group. I'm pretty sure that detail was entirely left out of the abstract. My point is that this person was trying to persuade me that MCT can "still cause weight gain" even though the study they linked to contradicted their conclusion, including the conclusion of the study itself. EDIT: I think it was this one, although I'm not sure this is of interest to anyone here, but whatever: https://journals.physiology.org/doi/full/10.1152/ajpendo.00200.2018?rfr_dat=cr_pub++0pubmed&url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org# https://journals.physiology.org/doi/full/10.1152/ajpendo.002... Honestly, when people read studies and jump to conclusions, I have to assume that most of the time they're just trying to appear clever and are counting on the fact that few people are willing to actually read the full study. Worse yet, journalists have a terrible track record in my experience. I've lost count of how many referenced studies I looked up that either contradicted the journalist's conclusion or didn't have any mention of what they were even talking about. > 4. What are the systemic effects? People don't eat fatty acids alone, they eat them as (ideally) part of a diet composed of whole foods. Might the effect hold in a realistic diet with more palmitic acid? Also this. How the diet interacts with the Randle cycle can change how these nutrients are ultimately handled. Palm oil in the diet may have no appreciable negative effect until you start eating it along with sugar in the form of things like Oreos and Twinkies (or sucrose-laden mice chow).

Ginden 5y ago

> What's the strength of effect? You are acting like anyone cares about effect sizes. As civilization, we transcended past such nuances, you can write alarmistic abstract about 3% effect size, as long magic p < 0.05.

jjoonathan 5y ago
My favorite in the breathless overreporting genre is when debate wins are described as "evisceration."

CoastalCoder 5y ago

I never know how to interpret the phrase "linked to" in various news headlines. Is it just shorthand for some relationship that's truly significant and meaningful in the context of the article? Or is it a weasel-phrase to click-bait people into reading a less significant finding and/or to impute guilt by association?

gregwebs 5y ago
In 99% of the news headlines you hear about the underlying study does not contain causal evidence in humans. It is almost always correlational in humans or a mouse model. So yes. Generally weak evidence that can conform to some existing view an organization holds is cherry-picked into the news. Although sometimes it is just in the news because it is odd/interesting.

chiefalchemist 5y ago

Good context rules. However, what always concerns me about 1 and 2 is that in our environment there are toxins of all types. Yeah, a penny here, a dime there, a nickle here again. Those all add up. And since there's no way to sort that out, each of those individual toxins gets a free pass for being an insignificant amount. Unfortunately, we can to be mindful of the sum of those toxins being greater than the individual parts. So now what? Is it reasonable and safe to ignore that context? And if we have kids?

Tagbert 5y ago
But “toxins” don’t all have the same effect on the same biology. You can’t just assume that they all contribute to the same disease. Some of them may cancel others out. Even the word “toxins” is so much abused in the popular medical press as to be not only useless but misleading. Any report like this needs to be evaluated for various factors, but one of the most important is the degree of effect. Some studies have identified effects that are are so small you would need to consume gallons of the product a day to see any difference from not consuming. The press is full of these scary sounding reports based on a study that found something that the report fails to explain has not been verified. Any one study should not be the basis of your life decisions. This would need to be verified or debunked by other studies until a consensus gives us enough information to determine how to act if needed.
- chiefalchemist 5y ago
  > You can't just assume... That is the point. They are all able to hide behind a cloud of doubt and lack of clarity. That's the spin. What's not addressed is the possibility aggregate - or worse, exponential - effect of each individual "Not us our data says becoming something far more (read: worse). So A, B and C do their data. But reality is not a silo. Those things combine. Who is addressing that? Just because that's extremely difficult doesn't mean it doesn't exist, doesn't matter. Keep in mind, as I understand it. A company can put a new chemical in a product without approval. It's only after the fact if it's proven unsafe will it be taken off the market (and perhaps only in select countries). Conclusion: Yes, it easy to say "It's only X amount of B." But it's not only B we're being exposed to. It's a whole bleepin' 1,000 letter alphabet of unknowns. No one should be comfortable with that. Yet we are.

frgtpsswrdlame 5y ago

Your rules are good but this: >Palmitic acid is in everything. Should have no effect on our judgement regarding its safety. Probably the story of our times seen from the future will be how we failed to account for extremely widespread use of very minorly dangerous products. See stuff like microplastics which took a couple decades to go from research to ban. Take the famous example from Fight Club: >The car crashes and burns with everyone trapped inside. Now, should we initiate a recall? Take the number of vehicles in the field, A, multiply by the probable rate of failure, B, multiply by the average out-of-court settlement, C. ... If X is less than the cost of a recall, we don't do one. Except we've moved even one degree from that now, think of a variable Y which is the cost to even figure out what B is. If you're a company selling a dangerous product you don't just have to consider X, you can also factor in Y as a proxy for long/hard entities will have to fight just to figure out the harm you're doing. So we should think about both sides of the coin, that the research is good enough to show there is an effect but also that we're not increasing Y to completely unreasonable amounts.

smsm42 5y ago
It's not everything like microplastics. It's everything like everything. Palmitic acid is one of the most common fatty acids. Cocoa butter is 25% PA. Cow milk is 30% PA. Chicken is 20% PA. In human milk, 25% of its fat content is PA. Olive oil has tons of it too, so do soy oil and sunflower oil (and probably any vegetable oil you've eaten). But if you don't eat anything at all, human body is still about 2% PA - as I said, it's a very common chemical compound. To quote Wikipedia: Palmitic acid, or hexadecanoic acid in IUPAC nomenclature, is the most common saturated fatty acid found in animals. If it were dangerous for animals, there would be no animals left living. It's not like microplastics that are exogenous. It's part of what living organisms are made of. Instead, what happened is probably some kind of p-hacking - they gave this compound, lots of it to be sure, to lots of poor mice to which they gave cancer previously, and in some of them the cancer got worse, because it's cancer. So they wrote down which cancers those mice had, and declared that they linked the compound to "spread of cancer in mice". Easy peasy. Sources: https://www.sciencedirect.com/science/article/abs/pii/S030881469700160X https://www.sciencedirect.com/science/article/abs/pii/S03088... https://www.sciencedirect.com/science/article/pii/S0952327819301279 https://www.sciencedirect.com/science/article/pii/S095232781... https://orbi.uliege.be/bitstream/2268/240227/1/fatty-acids-composition-of-meat-of-five-native-chicken-gallus-gallus-ecotypes-of-benin-reared-under-organic-or-conventional-system.pdf https://orbi.uliege.be/bitstream/2268/240227/1/fatty-acids-c... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5682332/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5682332/

koprulusector 5y ago

It’s embarrassing The Guardian posted a link to a 10. Address. :facepalm

dec0dedab0de 5y ago
I wonder how that is even possible. I just checked, they have different parent companies. Maybe that is an internal proxy/cache address? Possibly for getting around the paywall? Or maybe they paid/arranged for an internally proxied SAML connection?
- r00fus 5y ago
  It's clearly a mistake - either the article was downloaded onto the author's network (and internal link provided) or simply the wrong URL.
  - smnrchrds 5y ago
    It helps that if you are not paying attention, it looks like a DOI.
- aaron695 5y ago
  Somehow 10.1038 resolves to 10.0.4.14 -> https://www.google.com/search?q=1038+%2210.0.4.14%22 https://www.google.com/search?q=1038+%2210.0.4.14%22 Puzzle for the day. (If you don't know DOI's - https://en.wikipedia.org/wiki/Digital_object_identifier https://en.wikipedia.org/wiki/Digital_object_identifier important for looking up science articles, SciHub is one example. It might not be mathematical, it might be a API - https://doi.org/10.1038 https://doi.org/10.1038 )
labster 5y ago
The Sun posted a 10 too, but she was on page 3.

em500 5y ago

It mostly reminds me of this old Vox article[1], which was nicely summarized in the caption of its second graph: "Everything we eat both causes and prevents cancer". [1] https://www.vox.com/2015/3/23/8264355/research-study-hype https://www.vox.com/2015/3/23/8264355/research-study-hype

wyclif 5y ago
If this is junk science, then where is reliable medical information on the use of vegetable oils in cooking?

ravenstine 5y ago

> 2. Is the diet a realistic diet? Are the mice or rats eating a significant amount more of palmitic acid than what would be reasonable in a human participant? I'll have to go digging for it, but I want to comment on something this reminds me of. Someone a few months ago linked me to a study that showed that MCT oil caused weight gain in mice (might have been rats?), despite how MCT oil is known to be much less fattening than other fats based on how it's metabolized. If eaten in great enough abundance it may even cause weight loss. Having learned my lesson from just relying on the abstract, I found the full text for the study using Sci-Hub; the diet fed to both groups of mice was a high sucrose diet, the experimental group having more MCT oil than the other. The experimental group had less weight gain than the control group. I'm pretty sure that detail was entirely left out of the abstract. My point is that this person was trying to persuade me that MCT can "still cause weight gain" even though the study they linked to contradicted their conclusion, including the conclusion of the study itself. EDIT: I think it was this one, although I'm not sure this is of interest to anyone here, but whatever: https://journals.physiology.org/doi/full/10.1152/ajpendo.00200.2018?rfr_dat=cr_pub++0pubmed&url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org# https://journals.physiology.org/doi/full/10.1152/ajpendo.002... Honestly, when people read studies and jump to conclusions, I have to assume that most of the time they're just trying to appear clever and are counting on the fact that few people are willing to actually read the full study. Worse yet, journalists have a terrible track record in my experience. I've lost count of how many referenced studies I looked up that either contradicted the journalist's conclusion or didn't have any mention of what they were even talking about. > 4. What are the systemic effects? People don't eat fatty acids alone, they eat them as (ideally) part of a diet composed of whole foods. Might the effect hold in a realistic diet with more palmitic acid? Also this. How the diet interacts with the Randle cycle can change how these nutrients are ultimately handled. Palm oil in the diet may have no appreciable negative effect until you start eating it along with sugar in the form of things like Oreos and Twinkies (or sucrose-laden mice chow).