4 ms·

The study : https://journals.physiology.org/doi/epdf/10.1152/japplphysiol.00276.2025 https://journals.physiology.org/doi/epdf/10.1152/japplphysio... From a qui

by maeln 1y ago

The study : https://journals.physiology.org/doi/epdf/10.1152/japplphysiol.00276.2025 https://journals.physiology.org/doi/epdf/10.1152/japplphysio... From a quick read : It was tested in a cell culture, so not in a human or animal. That does change a lot of things. For the dosage: > Thereafter, hCMECs were treated116 with regular media or media containing 6 mM erythritol (Sigma Aldrich, Cat #E7500; St. Louis MO), a117 dose equivalent to a typical amount of erythritol [30g] in a single can of commercially available118 artificially sweetened beverage, for 24 hours (N=5 experimental units)

sfjailbird 1y ago
> It was tested in a cell culture, so not in a human or animal. That does change a lot of things. The article points out that similar observations have already been made in human subjects: > Positive associations between circulating erythritol and incidence of heart attack and stroke have been observed in U.S. and European cohorts
- Eisenstein 1y ago
  One of the cited studies (Khafagy et al., 2024) directly contradicts such claims. The study explicitly said "we did not find supportive evidence from MR that erythritol increases cardiometabolic disease". The primary human study they reference (Witkowski et al., 2023) has a few issues: - All subjects had a "high prevalence of CVD and risk factor burden" and represented the sickest patients in the healthcare system - Erythritol was measured only once at baseline, despite data which shows that levels fluctuate dramatically with consumption - It did not differentiate between dietary intake and erythritol produced by the body - Seeing as they were already sick they the subjects may have been consuming more artificial sweeteners than the general population There are two more human studies referenced but I didn't read them.
ricardobeat 1y ago
It’s tiring to see these quick dismissals of scientific studies at the top of the comment section. They are more often than not based on technicalities or fallacies. Pitting a two minute reading vs months of work by a team of scientists is not a great move. In this case, the nature of the study is clearly acknowledged, it does not “change a lot of things”: > We recognize given the in vitro, isolated single cell nature of this study we cannot make definitive translational conclusions or assertions regarding erythritol and clinical risk. However, the markers and mediators of brain microvascular endothelial cell function studied herein have been shown to have strong causative links with the development cerebrovascular dysfunction, neuronal damage and injury, thrombosis and acute ischemic stroke These findings are a starting point for further understanding, not something to be immediately ranked as true/false.
- kreetx 1y ago
  I think what commenters are looking for is a reason where this study is relevant for them (us) as humans, and they assess whether it is definitive or not. As HN is more of a generic curiosity and engineering related site then these starting point for further understanding are unlikely to get more nuanced discussion than that. Thus, rather than submitting articles like the current, rather wait until anything more is available. We are tired of clickbait as well.
  - Eduard 1y ago
    > Thus, rather than submitting articles like the current, rather wait until anything more is available. How long more to wait? https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=Erythritol&sort=byDate&type=story https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
    - kreetx 1y ago
      Are any of them decisive?
      - lithocarpus 1y ago
        We shouldn't need to wait for a _decisive_ study showing that a novel compound is dangerous to consider avoiding the novel compound. There's rarely profit in demonstrating that novel compounds are dangerous so it's extremely unlikely for a given dangerous novel compound that there will be any decisive studies showing the danger. IMHO the novel compound should have to be decisively shown to be safe before being sold in food, but since they are not, I recommend everyone avoid novel compounds as much as it's practicable.
        kreetx 1y ago
        I initially reacted to the root comment with why these kinds of articles often receive dismissive comments. HN is mostly not a medical forum so a typical reader reader isn't going to want nor be able to discuss the technicalities - they just want to know whether to avoid a substance or not. As is often the case, the results are inconclusive, hence the dismissals. (And these dismissals as top comments are also useful for the typical reader as they pretty much want a yes/no verdict and move on.) But to your points, if there aren't any studies which can show that a compound is dangerous in any meaningful way, why would you want to avoid it? (Given there is a need or purpose, e.g, a low-calories sweetener.) Also, decisively showing something to be safe is impossible in a similar way that software tests can only show that you haven't found any bugs yet, it doesn't mean there are none. (Off-topic: That's a quote from Edsger Dijkstra, which the following can be added: he is right, but for unit tests - using types, property testing or by running through the entire argument space for a pure function you actually can show that there are no bugs.)

by maeln 1y ago

sfjailbird 1y ago

> It was tested in a cell culture, so not in a human or animal. That does change a lot of things. The article points out that similar observations have already been made in human subjects: > Positive associations between circulating erythritol and incidence of heart attack and stroke have been observed in U.S. and European cohorts

Eisenstein 1y ago
One of the cited studies (Khafagy et al., 2024) directly contradicts such claims. The study explicitly said "we did not find supportive evidence from MR that erythritol increases cardiometabolic disease". The primary human study they reference (Witkowski et al., 2023) has a few issues: - All subjects had a "high prevalence of CVD and risk factor burden" and represented the sickest patients in the healthcare system - Erythritol was measured only once at baseline, despite data which shows that levels fluctuate dramatically with consumption - It did not differentiate between dietary intake and erythritol produced by the body - Seeing as they were already sick they the subjects may have been consuming more artificial sweeteners than the general population There are two more human studies referenced but I didn't read them.

ricardobeat 1y ago

It’s tiring to see these quick dismissals of scientific studies at the top of the comment section. They are more often than not based on technicalities or fallacies. Pitting a two minute reading vs months of work by a team of scientists is not a great move. In this case, the nature of the study is clearly acknowledged, it does not “change a lot of things”: > We recognize given the in vitro, isolated single cell nature of this study we cannot make definitive translational conclusions or assertions regarding erythritol and clinical risk. However, the markers and mediators of brain microvascular endothelial cell function studied herein have been shown to have strong causative links with the development cerebrovascular dysfunction, neuronal damage and injury, thrombosis and acute ischemic stroke These findings are a starting point for further understanding, not something to be immediately ranked as true/false.

kreetx 1y ago
I think what commenters are looking for is a reason where this study is relevant for them (us) as humans, and they assess whether it is definitive or not. As HN is more of a generic curiosity and engineering related site then these starting point for further understanding are unlikely to get more nuanced discussion than that. Thus, rather than submitting articles like the current, rather wait until anything more is available. We are tired of clickbait as well.
- Eduard 1y ago
  > Thus, rather than submitting articles like the current, rather wait until anything more is available. How long more to wait? https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=Erythritol&sort=byDate&type=story https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
  - kreetx 1y ago
    Are any of them decisive?
    - lithocarpus 1y ago
      We shouldn't need to wait for a _decisive_ study showing that a novel compound is dangerous to consider avoiding the novel compound. There's rarely profit in demonstrating that novel compounds are dangerous so it's extremely unlikely for a given dangerous novel compound that there will be any decisive studies showing the danger. IMHO the novel compound should have to be decisively shown to be safe before being sold in food, but since they are not, I recommend everyone avoid novel compounds as much as it's practicable.
      - kreetx 1y ago
        I initially reacted to the root comment with why these kinds of articles often receive dismissive comments. HN is mostly not a medical forum so a typical reader reader isn't going to want nor be able to discuss the technicalities - they just want to know whether to avoid a substance or not. As is often the case, the results are inconclusive, hence the dismissals. (And these dismissals as top comments are also useful for the typical reader as they pretty much want a yes/no verdict and move on.) But to your points, if there aren't any studies which can show that a compound is dangerous in any meaningful way, why would you want to avoid it? (Given there is a need or purpose, e.g, a low-calories sweetener.) Also, decisively showing something to be safe is impossible in a similar way that software tests can only show that you haven't found any bugs yet, it doesn't mean there are none. (Off-topic: That's a quote from Edsger Dijkstra, which the following can be added: he is right, but for unit tests - using types, property testing or by running through the entire argument space for a pure function you actually can show that there are no bugs.)