Your Artificial Sweeteners, Your Bacteria, and Your Health

It seems like one cannot help hearing about this paper throughout the microbiome and related fields.  The paper "Artificial Sweeteners Induce Glucose Intolerance by Altering the Gut Microbiota" was recently published in Nature, and it has had a lot of press.  Interest in the paper is partially due to its focus on two hot topics: the influence of food on the gut microbiome and aspects of obesity, as well as artificial sweeteners, which have long been a topic of debate.  I presented this paper at our student microbiome journal club a couple of weeks ago, so I wanted to go over it here too.

The big three artificial sweeteners. <Source>

First let's get the media hype out of the way.  This paper has been oversold by the media as a reason to stop eating artificial sweeteners, and the results do not actually support this claim (we will get to this below).  Furthermore, to be honest, the title of the paper oversold itself a bit too.  The results don't quite support the bold title, which also contributed to some of the perceived hype surrounding the paper (we'll get to this below too).  But beyond the hype, this is a pretty great and interesting paper and worth discussing in more detail.


The paper talks a lot about non-caloric artificial sweeteners (NAS), which are popular sugar substitutes because they can pass through the gastrointestinal tract without being digested.  This is beneficial because it avoids the calorie intake associated with sugar digestion.  Although the report generally discussed NAS, it is important to note that they really only look at saccharin (Sweet'n Low), and don't really discuss aspartame (Equal).  One of the reasons they don't focus on aspartame is because they see a minimal effect of the chemical on glucose intolerance early in their studies.  This is important to note because the paper title generalizes its claims to NAS, but there were no effects of aspartame (also commonly used) and they really only looked at saccharin.

The group showed that chronic exposure of mice to saccharin led to statistically significant (not sure about clinically significant?) glucose intolerance.  This was cool, but what was really interesting was that exposure of those mice to antibiotics resulted in a loss of the glucose intolerance phenotype, and this occurred whether the mice were given antibiotics targeting gram negative or positive bacteria.  This suggested that there could be a microbial component to the glucose intolerance observed in the mice, so they followed up on this with a few other experiments.

The chemical structure of saccharin. <Source>

To make a long story short, one of the highlights of the paper was that they supported their microbial theory with a standard fecal transplant experiment.  In short, they were able to take the feces from the saccharin treated mice, transfer it to the GI tracts of germ-free mice (these are special mice that have never been exposed to any bacteria in their lives), and observe the glucose intolerance phenotype in those germ-free mice.  They followed up on this experiment by culturing anaerobic bacteria out of the feces for over a week, transplanting that culture to the germ-free mice, and showing that the glucose intolerance phenotype is again transferred.  This was a pretty interesting finding, and again suggested a microbial component in the observed glucose intolerance.

The group followed up with some microbiome sequencing analysis of 16S rRNA genes and whole genomes (shotgun sequencing libraries).  While it was interesting to see gut bacterial dysbiosis after treatment with saccharin, the functional-potential analysis of the shotgun metagenomic libraries seemed to miss the point of the paper (for me reading it at least).  The focus of this paper is not really the microbiome of the gut following saccharin exposure (although it was an interesting observation), but rather the fact that they were able to transfer the glucose intolerance phenotype with fecal samples and cultures.  So while the functional genomic characterization was interesting, I was still left wanting to read about a biochemical experiment in which they fractionated those whole and cultured fecal samples and teased apart what chemicals are related to the phenotype transfer we see.  So the functional stuff was a little off topic, but it successfully peaked my interest in what the group finds next.  Hopefully I get to see them talk about their findings at an upcoming conference!

Finally, the group took this study from mice to humans, using a small human cohort (n=7) treated with high, long term levels of saccharin.  They found that they could categorize their subjects into a responder or non-responder group, meaning that one group had glucose intolerance following saccharin treatment, and the other group did not.  What was really cool was that the glucose intolerance phenotype experienced by the responder group was able to be transferred to germ-free mice via fecal transplant.  This suggests that there is some kind of metabolite, either present in the feces from the human or bacteria, that is causing the phenotype.  I was again left hoping for some more biochemistry to tease this apart, but we will have to wait for the group's next paper.  Despite this, I think this is an interesting case study observation (I say case study because of the small size and to avoid hyping this as something medically relevant) and may lead to interesting medical findings in the future.

Human gut microbes might be affected by artificial
sweeteners. <Source>

In the end, what can we say about these results?  Like I said above, the conclusions have been a bit exaggerated in the press, so I want us to go through what we can really say with these results.  Saccharin results in statistically significant (but possibly not clinically significant?) glucose intolerance.  The glucose intolerance phenotype is absent upon antibiotic treatment, and is transferable by whole and cultured fecal content transplant, which suggests a microbial metabolite is playing a role in this phenotype, although more work is needed to test this hypothesis.  Finally, high level saccharin administration affected human glucose intolerance in a responder subgroup, and this phenoype was transferable to germ-free mice through fecal transplant, which might also suggest a role for a microbial metabolite, but more work is needed.  I should also note that, although this has been sold as a microbiome paper, I left out most of the microbiome-ralted findings because they felt a little off the focus of what really made this paper cool.

Of course the question everybody is asking is whether saccharin is bad for you?  Based on this paper, I can firmly say I don't know.  This paper is not strong enough to make any medical conclusions, and more work will need to be done if that is their goal.  But scientifically, this is a very interesting paper with some interesting observations, and I will be interested to see what the group publishes next.  It will especially be cool to see some follow up work with how saccharin is affecting the git microbial communities they observed.

For further reading, check out the paper below at nature, or check out the brief article that appeared in The Economist.  As always, please also feel free to ask questions or express concerns in the comments below.

Works Cited

Suez, J., Korem, T., Zeevi, D., Zilberman-Schapira, G., Thaiss, C., Maza, O., Israeli, D., Zmora, N., Gilad, S., Weinberger, A., Kuperman, Y., Harmelin, A., Kolodkin-Gal, I., Shapiro, H., Halpern, Z., Segal, E., & Elinav, E. (2014). Artificial sweeteners induce glucose intolerance by altering the gut microbiota Nature DOI: 10.1038/nature13793

Sugar substitutes may mess with gut bacteria—causing obesity in the process. (2014) The Economist. <Source>