Gee, heard anything about erythritol recently?
What You Need to Know
Should I Stop Eating Erythritol? The sweetener erythritol and cardiovascular event risk
For those of you who are shaking your heads in puzzlement, let me first explain: on February 27th, so three days ago as I write this, a medical journal article (Nature Medicine) regarding erythritol was published online. It stated that erythritol, a sweetener very popular with low carbers, me included, increased the risk for “major adverse cardiovascular events.” This was reported in major media outlets, including the Washington Post, so the story has spread rapidly. This news has understandably caused a great deal of consternation among the low-carb community.
So what do I think?
I confess, my initial response was skepticism, for several reasons. There has been a lot of hysteria about non-sugar sweeteners, artificial and natural, for. . . well, my whole life, and I ain’t young. Cyclamate was taken off the market in the United States in 1970, though it remains in use in 130 countries and the FDA has since stated that all the available evidence does not implicate cyclamate as a carcinogen. There were rumors at the time that the money for the study that torpedoed cyclamate in the US came from the sugar industry, though I do not know if that is so.
Saccharine (sold as Sweet’N Low for the last 50 years) was branded a carcinogen; the pink packets bore a warning for decades – until 2000, when it became clear that humans metabolize saccharine differently from rats and mice. (Am I too cynical in suspecting that saccharine was not taken off the market like cyclamate because it doesn’t taste as good and therefore was not a big threat to the sugar industry?)
Aspartame (sold as Equal) was branded a neurotoxin despite repeated rigorous testing showing it is not; I just find it tastes terrible. Also, aspartame doesn’t work for cooking since heat breaks it down, making it less than useful for a cookbook author.
Sucralose is made from sugar with chlorine compounds replacing three of the hydroxy groups; this makes it 300-1000 times sweeter than sucrose, allowing tiny quantities to be used. That chlorine is the thing people use to tar sucralose, though we surely get far more from other sources, starting with tap water. A fair criticism of granular sucralose products is that they are bulked with maltodextrin, a carbohydrate. The “0 carb” listing on the label is an artifact of labeling law and serving size. (Since anything with <0.5g of carbohydrate per serving can be labeled “0 carbohydrates” and the serving size is 1 teaspoon, I count 24 grams per cup of granular sucralose products, or, for that matter, any sugar replacement bulked with maltodextrin: 8 fluid ounces per cup, times 2 tablespoons per ounce, times 3 teaspoons per tablespoon: 8x2x3=48, 48×0.5=24. This may be high, but without truly accurate statistics it’s the closest I can get.)
Too, I have seen posters on social media claiming that all artificial sweeteners are “neurotoxic,” apparently because of that claim about aspartame. This particularly tweaked me because, people, these compounds are all chemically distinct. They don’t all do anything except taste sweet.
In short, I have come to consider most claims about the purported dangers of sugar-free sweeteners to be overblown, to say the least. So you can understand why I raised an eyebrow at this report.1:
There were a couple of things about this report in particular that pinged my BS meter.
One is that it is in a journal called “Nature Medicine,” a title which made me suspect that the researchers were automatically anti-any sugar substitute. The other was that, while stating that erythritol occurs naturally in plants and is even endogenous to the human body (made inside of us), they insisted on calling it an “artificial sweetener.” (However, “It’s natural!” is a crap argument for safety. According to the study, “when incorporated into processed foods, it is typically added at levels 1,000-fold higher than endogenous levels…” There is nothing natural about that intake.)
Still, I was not going to dismiss it out of hand. Happily, I got ahold first of an abstract of the article and then of the whole article itself. I read them, of course, though I confess that, not having a medical education, I did not understand all of the jargon. Still, I got enough to answer several of my most pressing questions.
The Credibility of the Study
First of all, the editorial board seems to have serious credentials. That right there gained my respect.
Secondly, this article is based on a series of studies. One is vanishingly small – eight healthy volunteers who were given erythritol. The size of this study has drawn derision from some and were that the only study, it would be deserved. But that small study was inspired by previous, much larger studies of erythritol intake in people undergoing cardiac risk assessment – a US study of 1,157 subjects, another US study of 2,149 subjects, and a European study of 833 subjects.
In those larger studies, “incident risk” – the risk of actual negative events – over three years was higher in subjects with higher levels of multiple polyols (sugar alcohols) “especially erythritol” in their blood. This was in people who were “stable,” but chose to have their risk assessed because of risk factors that had already been noted.
In the eight healthy volunteers, the researchers found, first of all, that consuming erythritol raised blood erythritol levels for more than two days. For a substance that is largely peed out unchanged, that seems a long time to me. More to the point, they found that erythritol increases platelet reactivity and aggregation, aka clotting. Given the number of people who are on blood thinners for heart disease and stroke risk, and the fact that those risks are already higher in those of us with carbohydrate intolerance issues, this is no small thing.
Comparing Erythritol to Glucose
Reading the whole article answered another question for me: the researchers did, indeed, compare erythritol to glucose. Increased glucose did not increase clotting risk, while erythritol did.2
It is worth noting that the tests on the eight healthy volunteers involved them drinking a beverage with 30 grams of erythritol before testing. I looked at the ingredient labels of Zevia, a popular erythritol/stevia-sweetened soda. A can of Zevia Black Cherry contains 7 grams of erythritol, while a can of Zevia Ginger Ale contains 4 grams of erythritol. This makes the 30-gram dose between 4.3 to 7.5 times that of a can of Zevia, a not-insignificant difference in dose. Still, given that in my youth I drank 12-18 cans of Tab per day (Tab, children, was Diet Coke before Diet Coke was Diet Coke, and yes, I had a serious Tab monkey on my back), it is likely some people are drinking 30 grams of erythritol’s worth of such sodas per day.
I also looked at the label on a popular “keto” dessert. If you ate a whole pint of the product in one sitting, and I have no doubt some people do, you’d get just upward of 30 grams of erythritol. The label on a bag of granular erythritol listed a serving as 2 teaspoons or 8 grams. That means just four cups of coffee or tea with 2 teaspoons of erythritol would also give you more than 30 grams.
So the experimental dose, while high-ish, is not unrealistic.
More Questions Need More Research
I still have questions. I want to know the actual number of increased incidents. If it’s in the article, I could not find it; this may be because of my lack of familiarity with the jargon. But if the risk went from, say, five incidents per thousand subjects to six incidents per thousand subjects, while that is a 20% increase in risk, it doesn’t terrify me. On the other hand, if the risk went from five incidents per thousand to twenty-five incidents per thousand, that’s a whole lot more worrisome.
I’d like to know if the usual over-the-counter blood thinners – aspirin and fish oil – are sufficient to counteract the effect of erythritol on clotting. Since I take a fish oil capsule daily, I’d like to know if and how this mitigates the risk.
How does the risk scale with dosage? The first rule of toxicology is “the dose makes the poison.” What is the risk of, say, a half-cup serving of a keto dessert instead of a whole pint?
Too, the larger studies were observational studies, not controlled in any way. Such studies are useful for suggesting avenues for further, controlled studies, but do not, in and of themselves, prove anything.
That said, I will be adjusting my approach to erythritol. I am glad that I have for years used products that combine erythritol with high-intensity sweeteners like stevia, monk fruit, or sucralose so that only half the quantity is needed to achieve the desired degree of sweetness. I am heartbroken about my beloved brown-sugar-style Swerve, of which I just bought another bag. I will be experimenting with allulose, starting first with blood sugar and ketone tests. I may go back to combining a small quantity of molasses with a sugar-free sweetener to get a brown sugar flavor. English toffee stevia is also useful in this regard.
I will likely not shun erythritol completely, at least until more research is done. I don’t eat a lot of sweeteners anyway – a few drops of stevia in a mixed drink, or yogurt, or a dressing or marinade now and then. But I don’t drink soda at all; it has long since come to revolt me. I drink tea with no sweetener, sparkling water, and dry red wine. And while I eat low-carb, sugar-free ice cream now and then, I usually consume about that half-cup serving on the label, and I eat it now and then, not daily. As for homemade sweets, I often go months without making any except when working on a book.
And this would be my best advice to you, and in fact has been all along: back away from sweet stuff in general. Not “DO NOT EAT ANYTHING SWEET EVER!!!” Just… back away.
On October 18, 2022, CarbSmart’s Dana Carpender interviewed William Davis, MD for her “3 Questions with” series. Dr. Davis spent an informative hour talking with CarbSmart’s managing editor. Read Dana’s review of Wheat Belly and Super Gut. Includes link to video on YouTube.
And Dr. William Davis, whom you may know as the author of Wheat Belly, Undoctored, and Super Gut (and if you don’t, get on it!), informed me that artificial sweeteners in general are hard on the gut microbiome. Since I spend good money on those little bugs, I’d rather not wantonly kill them. That doesn’t mean I never use sucralose, but I don’t use it often. I take Bill Davis’s advice seriously.
Speaking of Dr. Davis, here is his take on the issue.
A recent study from the Cleveland Clinic raises questions about the safety of erythritol, a natural non-caloric sweetener that I have included in some of my recipes. This comes from the laboratory of Dr. Stan Hazen, who has–in my view—released suspect evidence in past regarding the atherogenicity (atherosclerotic plaque-causing) and cardiovascular event-promoting effects of trimethylamino...
1 I have heard two criticisms of artificial sweeteners from authorities I trust. One is Dr. William Davis’s information regarding artificial sweeteners and the gut microbiome. The other is this: Dr. Mary Vernon, a diabetologist and former President of the American Society of Bariatric Physicians explained years ago on the Low-Carb Cruise that it had been discovered that we have tastebuds in our small intestines, the same tissue as we have on our tongues. She went on to say that when those intestinal tastebuds sense anything sweet, sugar or artificial sweetener (or stevia or monk fruit), they signal the body to release a little insulin in preparation for an anticipated rise in blood sugar. If the blood sugar does not rise, more insulin will not be released, but there is a mild rise in insulin even from non-caloric sweeteners. (return)
2 Though glucose did not, in this small study, increase platelet aggregation, that does not let sugar off the hook where heart disease is concerned. I earned my I’m A Big Geek badge one day more than a decade ago on the Low-Carb Cruise. There I was on a big, shiny party boat in the middle of the Caribbean on a beautiful day, sitting in a conference room well below decks, listening with rapt attention to a cardiologist explaining the effect of sugar on the endothelium.
“The what?!” I hear you cry. The endothelium is the single-cell-thick lining of the blood vessels, and it controls what our blood vessels do — dilate or contract, how they direct the blood, and all sorts of things. Because it is in direct contact with the blood, the endothelium cannot defend itself from high blood sugar by becoming insulin resistant as so many other cells do. As a result, high blood sugar levels will damage the endothelium, setting up inflammation. This in turn causes damage to the blood vessel walls, and the body creates a plaque — the plaques that cause coronary artery disease — to patch the damage. Sugar also causes the naturally slick endothelium to become roughened, making it more likely that a clot or a piece of plaque will catch and block the vessel, causing a heart attack, stroke, or deep vein thrombosis.
My point is that just because glucose doesn’t increase platelet aggregation doesn’t exonerate sugar where cardiovascular disease is concerned, not by a long shot. (return)
More Low Carb Recipes & Articles by Dana Carpender
What surprised me was that they didn’t talk about the fact that erythritol is synthesized endogenously in humans from glucose (and I would guess fructose since that converts to glucose) via the pentose-phosphate pathway, so how much of the elevated levels are caused by that? I do think that you’re spot on about just avoiding the sweet stuff or making it a very rare treat is a good idea.
They did mention the endogenous production of erythritol, but not that it’s made from glucose. Good catch!
Where does BochaSweet stand in all of this?
Honestly, I have no idea. I haven’t tried it. Perhaps I should!
hi. I drink Zevia sodas and was concerned when you said they are sweetened with erythritol. I looked on their website and in the faq’s it says “Zevia products are solely sweetened by highly purified stevia extract. Unlike non-beverage stevia products, Zevia does not contain any “bulking agents” such as erythritol, and is naturally sweetened.”
Thank you for your insights about the journal article.
Thank you! Zevia did, indeed, include erythritol for a long time; the cans I have in my cellar, awaiting a visit from someone who likes soda, do list erythritol. I wonder when and why they got rid of it? I find referring to it as a “bulking agent” odd; it’s a sweetener.
Blue Sky’s website still showed erythritol in their Blue Sky Free sugar free soda, at about the same levels as were formerly in Zevia, so it’s still quite possible that an avid soda-drinker could get 30 grams or more of erythritol in a day.
Lakanto Liquid Monk Fruit Extract Drops lists ingredients as: Water, Monk Fruit Extract, Natural Citrus Flavors (Preservative)….
Have you tried this? Is it comparable (in taste) to the granulated form? Do you think this would be an acceptable replacement, since it has no Erythritol?
I dislike all such products but see no reason not to use this within reason, if that’s the whole ingredient list. If you’d like a soda, combine it with sparkling water.
Nature Medicine is NOT an “alternative” or disreputable journal. Nature is one of the most respected scientific journals in the world Nature Medicine is the medical journal in their family. It’s comparable to the Lancet in reputation. You still need to critically read every article, of course, as none are perfect, but that first step of initial credibility has been met just by publication there.
The lack of long term randomized controlled trials of dietary factors is not a flaw. Observational studies are typical for these studies, though they do have flaws, because you can’t control every aspect of people’s diet for 10 years. You can do it for a few days or weeks, but it’s not practical or affordable beyond that.
Article critique is a real skill. It was one of my favorite things to do in school, and I still enjoy reading them critically.