Erythritol and CVD Event Risk

3/1/23-UPDATE!

This is a solid analysis from Dr. Soto-Mota. He goes through largely the same process I did and arrives at largely the same conclusion: there might be something concerning here. I’m not going to stop using erythritol, in fact I doubt this really influences anything I do as “keto treats” tend to be a highly enjoyable but fairly infrequent thing for me.

Adrian Soto-Mota @AdrianSotoMota

Not mutually exclusive, but most people said they'd prefer seeing rather than joining a discussion to analyze THAT erythritol paper, and I have some airport hours ahead... Here go🧵. In case you've been living under a rock, this is the paper: nature.com/articles/s4159…

Hey folks!

You have likely seen this piece published in Nature which casts a long shadow over the popular sweetener Erythritol. It suggests a fairly tight dose response relationship between erythritol consumption, increased clotting and platelet aggregation, which could dramatically increase CV events such as stroke and heart attack. If this had been some simple, retrospective cohort study asking folks to recall what they ate 3 years ago, I would have ignored it outright, but it has a lot more going on than just that. There are several features to the case these folks are making, I’ll briefly detail those elements below but encourage you to read the whole paper, it’s quite accessable even for the layperson:

Untargeted metabolomics and MACE, discovery cohort

A group of participants were screened for presence of various small molecules (including erythritol) and the incidence of major adverse cardiac events (MACE). According to that data, the authors suggested a link between erythritol levels and MACE.

Targeted metabolomics analyses of erythritol, validation cohorts

Like many organic compounds, erythritol can be difficult to identify by various lab techniques are structurally similar compounds can make it confusing just how much erythritol (in this case) is contributing to the signal. the authors developed an assay to be able to specifically quantify the amount of erythritol patients in various cohorts presented. This is good rigorous stuff as far as it goes, but this is an interesting point where the waters get a bit muddy. From the original paper:

In both validation cohorts, plasma levels of erythritol were higher among individuals with prevalent CVD (P < 0.0001 each; Extended Data Fig. 3); higher levels of erythritol were also observed among those who experienced an incident MACE over the ensuing 3 years of follow-up (P < 0.0001 each; Extended Data Fig. 3). Further, in both cohorts, higher incident event risk was observed with higher levels of erythritol in Kaplan–Meier analyses…

This is a big deal as it’s well known that in mammals erythritol levels appear to elevate in the early stages of metabolic syndrome development. (Mammalian metabolism of erythritol: a predictive biomarker of metabolic dysfunction). From the abstract of that paper:

Recent findings: Elevated serum erythritol predicts future central adiposity gain and type 2 diabetes mellitus in healthy adults. Erythritol is a newly recognized human metabolic product of glucose, synthesized through the pentose phosphate pathway. The final conversion of this metabolic pathway is catalyzed by the enzymes sorbitol dehydrogenase and alcohol dehydrogenase 1. Erythritol is also a well characterized nonnutritive sweetener. Recent studies show that dietary erythritol can be metabolized to erythrose or erythronate in humans before excretion.

Summary: Elevated serum erythritol predicts risk for cardiometabolic disease, but more research is required to maximize its utility as a biomarker, including characterizing the determinants of endogenous erythritol synthesis from glucose. New insights into dietary erythritol metabolism also highlight the need to evaluate the effects of long-term erythritol consumption.

So, this is a bit of a concerning piece to this story if we are to paint erythritol as a problem. We know for certain metabolic syndrome/insulin resistance is a major risk factor for CVD. Lots of mechanisms there…increased glycemic load alone is a problem (which I’ll get to in a bit) elevated systemic inflammation and all the confounders that go along with metabolic syndrome: generally inactive, poor sleep etc, etc, which no mater how much folks claim they can “adjust” for that, its just bollocks. Exercise is so foundational to metabolic health, so protective against CVD, that it’s a big lift to suggest a sedentary person really has healthy CVD profiles. Anyway, we have some interesting cases being made here, but also some concerning confounders. On to the next bit of investigation:

Erythritol exposure and platelet responsiveness

In this piece the authors looked at platelet aggregation (clotting potential) with “normal” physiological levels of erythritol. They discovered a trend which suggests increasing platelet aggregation with increasing levels of erythritol. This is some of the stuff that made me take this more seriously than a simple epidemiological correlation. There was an interesting wrinkle to this however, from the paper:

Across the physiologically relevant concentration range observed in fasting plasma samples, erythritol dose-dependently enhanced platelet aggregation in PRP (Fig. 3a). In contrast, no effect on platelet aggregation responses was observed with either glucose, the most common polyol, or 1,5-anhydroglucitol (AHG),

This is interesting in that it’s well known that normal glucose levels are benign to platelet aggregation, while hyperglycemia is quite thrombic. from that paper (which IS in a mouse model, but this has been established in humans):

In vitro analysis of platelets revealed that normal glucose reduces platelet activation whereas hyperglycemic conditions increase platelet activation. We therefore hypothesized that hyperglycemia increases platelet glucose utilization, which increases platelet activation to promote thrombosis. 

So, I guess the case could be made here that “normal” physiological levels of either erythritol or glucose should be regarded as relatively benign, but it seemed a bit of an odd comparison given what we know about all of this.

The authors did one additional assay looking at human whole blood and not just platelet activation but also adhesion, which precipitates clotting, and again, they found a strong correlation between erythritol levels and what we’d recognize as initiation of the clotting cascade.

Postprandial levels of erythritol in healthy individuals

In the final section of this paper the authors fed a rather large dose (30g) of erythritol in the form of a sugar free beverage…they noted this would be on par with consuming about a pint of “keto” ice cream, and I think this would be in the ballpark of what folks use for baked goods when using erythritol at home. Some have said “this is way more than people would normally consume” but I’m just not seeing that, this seems like reasonable intake. They found a remarkably increased level of erythritol above baseline- nearly a 1,000 fold increase which lasted for several hours, and most folks had reasonably elevated levels 2-days later. The case the authors make is this puts folks well within the prothrombic range that is observed via the other methods.

So, Robb…nothing sauce, or should I freak out?!

I’m honestly not sure. Some folks have dismissed that last piece regarding the feeding as it had a small number of participants. I don’t see it that way…large numbers are necessary for certain studies when the effect is subtle and we need to make sue we account for other random features in the system or study. The erythritol feeding is NOT subtle. Not at all. Folks consistently had a remarkably elevated. check this out:

That’s a big-ass bump off of baseline! Erythritol is not cleared from the system (for the most part) via metabolic breakdown, but rather via excretion in the urine. That’s not necessarily good or bad, but it does present a bottleneck as to clearance. Many small molecules act as signaling agents. It appears that erythritol may act to facilitate the release of calcium which initiates all this prothrombic activity. Glucose acts similarly at higher concentrations. There IS a concerning piece to this in that I suspect something very different is happening with metabolically healthy folks vs insulin resistant, but I’m speculating on that but I’d be surprised if it were not the case. We also have the confounder that other research has suggested erythritol is not only an antioxidant, but improves endothelial function. In that study (performed in type 2 diabeticcs no less) they fed folks 28g (in this case divided into three 12g doses via beverage) of erythritol per day for a month and documented only favorable responses, although they were NOT looking specifically at clotting potential. This too raises a host of questions: did the folks in this study eat less sugar after the drink? Fewer overall calories? I can;t tell from what’s available, but like all this stuff, you turn over one stone and create exponentially greater questions.

One thing that stands out is we’ve NEVER seen a signal like this. I can find no literature on a coagulation panel being done with erythritol, not in animals, not in humans. This is a really easy test and for me would be more compelling than the in-vitro studies.

This will annoy some and anger others, but I’d personally put the possibility that erythritol might promote unfavorable CV processes as a maybe. We need to poke around some more, folks smarter than myself might recommend something other than a coagpanel, but to me that seems like a simple, slam dunk kind of assay.

Ok, that’s all the time I have for now, let me know what you think and please do post any questions to the comments.