Be Skeptical of the New Artificial Sweetener Scare

Does the artificial sweetener erythritol cause heart attacks and strokes? That’s the tenor of a spate of reporting on a new study published this week in the journal Nature Medicine. But there are numerous reasons to be cautious about drawing conclusions from the study’s purported findings.

The researchers report that “in patients undergoing cardiac risk assessment…circulating levels of multiple polyol sweeteners, especially erythritol, were associated with incident (3 year) risk for major adverse cardiovascular events,” including heart attack, stroke, and death. Erythritol—”one of the most widely used artificial sweeteners with rapidly increasing prevalence in processed and ‘keto’-related foods”—was “among the very top” molecules associated with these tragic outcomes.

“Subsequent analyses of stable patients undergoing elective cardiac evaluation confirmed this association,” with patients with the highest levels of circulating erythritol in their blood more likely to have cardiovascular disease and more likely to experience major adverse cardiovascular events. “Compared to participants in the lowest quartile of erythritol levels, those in the highest quartile had a substantially increased incident event risk in both” U.S. and European cohorts, the study states.

“The degree of risk was not modest,” lead study author Stanley Hazen, director of the Center for Cardiovascular Diagnostics and Prevention at the Cleveland Clinic Lerner Research Institute, told CNN—one of many major media outlets to report (uncritically) on the study’s findings. CNN summed it up like this: “People with existing risk factors for heart disease, such as diabetes, were twice as likely to experience a heart attack or stroke if they had the highest levels of erythritol in their blood.”

That sounds bad, right? But let’s dig a little deeper.

The first issue to jump out here is that all of the people studied already had metabolic issues of some sort. So it’s at least unclear that these results would generalize to populations without these issues.

Another red flag (raised by a neurobiology Ph.D. student on Twitter) is that the people in the highest quartile of erythritol levels look quite different than the people in the lowest quartile. In the first part of the study, for instance, those in the lowest quartile had an average age of 59.8, while those in the highest quartile had an average age of 70.2. The highest quartile had about a point heavier BMI, was less likely to be male, and was significantly more likely to have diabetes (36.7 percent versus 14.8 percent in the lowest quartile). They were more likely to have a history of heart attack and heart failure, and had significantly higher blood triglyceride levels, among other things. And similar differences can be seen in the U.S. and European study cohorts.

The researchers say they adjusted for these factors. But it’s difficult to adjust for such a wide range of differences and all the complex metabolic processes that they entail.

Perhaps the biggest issue in the study—noted by University of Oxford researcher Nicola Guess—is that human bodies produce some erythritol naturally, and yet the researchers didn’t control for endogenous versus exogenous erythritol.

Like cholesterol, the amount of erythritol in the body doesn’t directly correspond to erythritol consumed. Exogenous erythritol—which comes from our diet—is what all the reports on this study are concerned with. (For instance: “Study Suggests Possible Link Between Sugar Substitute and Heart Issues,” reported The New York Times in a typical headline about the study.) Endogenous erythritol is a molecule produced internally.

But the researchers in this study did not tease out how much circulating blood erythritol came from each source, leaving open the possibility that a) it’s not only or even mostly consumption of erythritol as a sugar substitute that led to high erythritol levels in the cohort with the worst outcomes and b) it’s not only or even mostly consumption of erythritol as a sugar substitute that is linked to higher instances of heart attack, stroke, and death.

It still could be—but we