There is an ongoing discussion of the safety of sweeteners – partly because some studies show an increased risk for chronic diseases among those who consume sweeteners, especially in soft drinks.

It is therefore important to investigate the physiological effect of sweeteners, and this has been done in this study.  Unfortunately, the authors decided to use an amount of sweetener that is – at least in the UK and Europe – unrealistic.  The sweetener concentration they used was 10-fold higher than the permitted amount in drinks and the single dose they use was more than most of us would eat during an entire day.

The authors also find other physiological effects of erythritol but mainly at plasma concentrations that are higher than those found in the study cohort and presumably also the general public.

These results suggest a potentially adverse effect of erythritol when consumed a amounts above what is generally consumed in Britain or the EU – and this is one of the reason why regulators set limits for the use of food additives in sweeteners: to protect the public and ensure intake is in a safe range.

In contrast to these data in this new study, observational studies can provide much more reliable data about links between sweetener intake and health as they reflect normal diet.  In this context, those studies measure actual intake which provides much better information on actual intake.  Unfortunately, even those studies are affected by confounding and an association between high sweetener intake and ill health might not establish a casual relationship.

Unfortunately, the authors do not provide any information about the erythritol concentration found in the study populations and this makes an interpretation difficult.  The study’s findings are surprising as only the group with the highest plasma erythritol concentration has a higher risk of disease – but not the others.  This group also has the oldest participants, and when the authors adjust for this, the observed association with risk becomes much smaller.

The interesting question here is the origin of plasma erythritol, as this could potentially explain some of the results: erythritol is not only used in foods but also other products such as tooth paste and drugs.  Especially the latter might affect the observed results as higher plasma concentrations could indicate that participants receive different medical treatments.

The press release is very cautious in making a bold claim, and I agree that the information is interesting and useful, but definitely not suitable to cause anyone to worry.

(The paper claims that erythritol has 'generally regarded as safe' [GRAS] status in Europe and this is incorrect.  EFSA does not have a GRAS system.)

This is a thought-provoking study based on good science.  There were a high number of study participants and the data has been well analysed.

The authors not only found a potentially important association between levels of the artificial sweetener erythritol in the blood and health impacts, but they also took the trouble to show a viable way by which the effect they found could occur.  Although this study can’t itself show whether or not this is the case, the authors suggest that higher erythritol levels may increase the risk of blood clots forming and this in turn may increase the risk of disease.

We should be mindful that correlation is not causation.  As the authors themselves note, they found an association between erythritol and clotting risk, not definitive proof such a link exists.

Because the people in the study already had a lot of cardiovascular risk factors it can’t be shown that it wasn’t one of these other factors that caused the increased clotting risk rather than the erythritol.  While some effects were seen in a test with healthy volunteers, this study only had 8 people in it and only lasted a few days.  This is not enough to draw firm conclusions from, and the authors don’t try to do so.

While I think the finding certainly warrants further investigation don’t throw out your sweeteners just yet.  This study only looks at erythritol and artificial sweeteners are generally considered safe.  Any possible (and, as yet unproven) risks of excess erythritol would also need to be balanced against the very real health risks of excess glucose consumption.

All scientific data, whether from studies, reviews or meta-analyses, must be reviewed by the European Food Safety Authority (EFSA), in this case as part of its re-evaluation programme for additives (Regulation 257/2010), which includes the 19 sweeteners authorised to date in the European Union. All studies, whether negative, neutral or positive, must be taken into account in order to adopt the appropriate decisions and positions, according to scientific validity. EFSA is the body that has to approve the scientific evidence and issue its scientific positions on food, nutrition and diet (more than 9,000 since it was created); in the case of sweeteners, it has not changed its positions and data to date, even in the totally exhaustive re-evaluations of sweeteners that it has carried out.  

On the other hand, in the European Union, polyols, such as erythritol, xylitol and others, cannot be used, for example, in any type of beverage due to conditions of approval for use in EU legislation (Regulation 1333/2008). In turn, when they are used in other foods, according to Regulation 1169/2011 on food information provided to the consumer, if they exceed 10 % of the ingredient content, the following legend must appear on the label: "excessive consumption may cause laxative effects".  

Erythritol also has two health claims approved and authorised by both EFSA and the EU authorities through Regulation 432/2012. 

Polyols occur naturally in many fruits and vegetables, as well as in other plants, such as birch and its well-known birch sugar (xylitol).  

This study, although the n=8 of the pilot study is very small, at least in the title of the paper it clearly states that its conclusions cannot be extrapolated to other sweeteners beyond erythritol. It should also be useful, even if it is based on an observational study, to be able to go deeper, especially with clinical trials of a sufficiently large size and with use below the ADI (Acceptable Daily Intake). This implies long-term studies.