From soft drinks to yogurt, artificial sweeteners have become commonplace in the food and beverage industry and are recognized as safe by the FDA. Yet a 2014 study found that artificial sweeteners are able to alter your gut microbes, and your health as a result. Read on to learn exactly what the researchers found and how artificial sweeteners might be contributing to the modern epidemic of metabolic disease.
A little over two years ago, I wrote a post titled “The Unbiased Truth about Artificial Sweeteners.” At the time of writing, the scientific literature did not really suggest any significant negative effects of artificial sweeteners. A few studies had found negative effects, but many others showed no correlation at all. While I urged caution in the consumption of these sugar substitutes, there was really no solid evidence at the time to recommend their strict avoidance.
Fast forward to today, and enter the gut microbiota. The trillions of bacteria that inhabit your gastrointestinal tract have received a tremendous amount of attention in recent years. A comprehensive research study has now shown (almost unequivocally) that artificial sweeteners can in fact impact health via altering gut microbes. (1).
Non-caloric artificial sweeteners
A sugar substitute is any food additive that provides a sugary taste but has significantly less associated calories, or food energy. Some sugar substitutes are natural, like stevia, while others are synthetic, termed “artificial sweeteners.” In the United States, six artificial sweeteners have been approved for use: aspartame, sucralose, neotame, acesulfame potassium (Ace-K), saccharin, and advantame. All have been deemed Generally Recognized as Safe (GRAS) by the FDA (2).
As high-fructose corn syrup continues to receive opposition from consumers and health organizations alike, the food and beverage industry is increasingly turning to artificial sweeteners instead. According to BCC Research, the global market for high-intensity sweeteners is expected to reach almost $1.9 billion in 2017, with the non-nutritive category growing particularly rapidly (3). The most commonly used non-caloric artificial sweeteners (NASs) are saccharin (Sweet’n Low), aspartame (Equal, NutraSweet and Canderel), and sucralose (Splenda).
Did you know artificial sweeteners can cause glucose intolerance?
Most NASs pass through the human GI tract without being digested by the human host. They therefore come in direct contact with microbes in the colon. As we’ll see from the results of this study, this has dramatic implications for the health of the host.
Prolonged consumption of non-caloric artificial sweeteners makes mice glucose intolerant
In the first part of the study, described in Nature, researchers from the Weizmann Institute of Science in Israel added saccharin, sucralose, or aspartame to the drinking water of three different groups of lean 10-week-old mice. They also had several control groups, including mice drinking only water or mice drinking water supplemented with glucose or sucrose, to see how the artificial sweeteners compared to normal sugars.
The researchers then performed a glucose tolerance test. As this is a crucial aspect of the study, I’m going to briefly describe how this is done. All of the mice are given only water for six hours prior to the test so that they are in a fasted state (and the NAS, glucose, and sucrose mice do not continue to eat sweeteners). They are then given 40 mg of glucose orally. Blood from the tail vein is used to measure glucose immediately before and at 15, 30, 60, 90, and 120 minutes after the glucose is given. From these data, the researchers can create a glucose tolerance curve to determine (a) fasting glucose levels, (b) how high the blood glucose spikes, and (c) how quickly glucose is cleared from the bloodstream.
So what did they find? At week 11 of feeding, the three control groups (water, water+glucose, water+sucrose) had comparable glucose tolerance curves, whereas all three NAS groups (water+saccharin, water+sucralose, water+aspartame) had developed significant glucose intolerance. Saccharin had the largest effect. They duplicated the experiment, this time in mice with diet-induced obesity, and observed the same result: NAS made the obese mice more glucose intolerant.
Treating mice with antibiotics abolished glucose intolerance
The researchers next wanted to see if the microbiota was responsible for the glucose intolerance. One of the simplest methods for determining if the microbiota plays a role in a particular trait or condition is to administer antibiotics and observe any changes. They did exactly this: the researchers treated mice with a combination of ciprofloxacin and metronidazole or vancomycin alone for four weeks and found that antibiotic treatment abolished glucose intolerance in both the lean and obese models.
This suggests that NAS induction of glucose intolerance is very likely mediated by the gut microbiota. However, there is always the slight possibility that off-target effects of the antibiotics on the host abolished the glucose intolerance, rather than the changes in microbiota composition. To rule this out and confirm the causality of the relationship, the researchers turned to germ-free mice.
Transferring NAS microbiota to germ-free mice transfers the glucose-intolerant phenotype
I’ve written about germ-free (GF) mice before in several of my blog articles on the microbiota (4, 5, 6, 7). Raised in sterile conditions, GF mice lack any microbes at all but can be selectively recolonized for experiments. Researchers can therefore perform an intervention in normal mice, take their fecal material, and transplant it into GF mice to determine if a particular effect of the intervention is mediated by the microbiota. If it is, the simple act of transferring the fecal material will elicit the effect in the recipient mice.
In this case, the researchers had two groups of mice that would serve as their fecal donors for the experiment, both of which were kept in normal (non-sterile) housing. One group was fed normal mouse chow + saccharin, and the other was fed normal mouse chow + glucose as a control. The amount of saccharin given was the mouse equivalent of the acceptable daily intake (ADI) of saccharin in humans (5 mg/kg, suggested by the FDA).
Fecal pellets from these two groups of mice were then transplanted into two groups of naïve GF mice. The GF recipient mice were therefore colonized by microbes from saccharin-fed or glucose-fed donors but were maintained on normal chow themselves. Six days after the transfer, the researchers performed a glucose tolerance test on the recipients. They found that those mice that received the saccharin-fed microbiota had developed significant glucose intolerance compared to those that received the glucose-fed (control) microbiota. This confirmed the findings in the antibiotic model, suggesting that alterations in the microbiota were in fact responsible for the differential glucose tolerance.
NAS alters microbiota composition and function
The most pressing research question then was, how had the microbiota been altered? Using 16S sequencing technology, they characterized the microbiota of each group. They found that mice consuming saccharin had a distinct microbiota composition from all three control groups. The authors reported over 40 operational taxonomic units (groups of bacteria) that were significantly altered in abundance, an indication of considerable dysbiosis.
Which bacteria changed? In the saccharin group, the Bacteroides genus increased, while Lactobacillus reuteri and Akkermansia muciniphila (two microbes generally considered to be beneficial) decreased. Several members of the Clostridiales order increased, while other Clostridiales decreased.
They next wanted to look at microbial function, and this is where the technology gets really cool. The researchers performed shotgun metagenomic sequencing, which allows for mass sequencing the genome of virtually every microbe in the gut. They did this sequencing on fecal samples collected before and after the 11 weeks for all of the groups of mice. The results showed that saccharin-fed mice had a strong increase in genes associated with glycan degradation pathways, which have been linked to enhanced energy harvest and metabolic disease (8, 9).
As if their prior findings in mice weren’t robust already, the researchers also cultured naïve mouse fecal material with NAS. They then transferred the NAS-exposed fecal material, or control fecal material, into GF mice. Recipients of the NAS-exposed fecal material showed increased glucose intolerance compared to GF mice given the control culture and similar alterations in microbial composition to the previous experiment.
What about humans? NAS consumption in humans is associated with impaired glucose intolerance
I discussed many of the human studies looking at NAS consumption in my previous article. When that was published, most human studies had shown mixed results with NAS. After finding such robust effects on the microbiota and glucose intolerance in mice, Suez and colleagues decided to see if the findings would translate to humans in their hands. They first performed a cross-sectional study, using a food frequency questionnaire to determine approximate NAS consumption, and also looked at several measures of metabolic disease.
Contrary to previous studies, they found significant positive correlation between NAS consumption and a number of clinical parameters related to metabolic syndrome: weight, waist-to-hip ratio, fasting blood glucose, hemoglobin A1c, glucose tolerance test, and ALT. You might be thinking: well, sure, it’s correlated because people with excess weight and metabolic syndrome are the most likely to use NAS in place of sugar as a means of losing weight. Luckily, the researchers took this into account and the effect still remained after correcting for BMI.
More human evidence
Correlations are great and all, but that’s all they are: correlations. We can’t infer causation from simple associations between consumption of a substance and outcomes. Fortunately, Suez et al. took it to the next level and performed a longitudinal study. Seven healthy volunteers who do not normally consume NAS or NAS-containing foods were followed for one week. On days two through seven, participants consumed the FDA’s maximal acceptable daily intake of saccharin (5 mg/kg body weight) as three divided daily doses. They were monitored by continuous glucose measurements and daily glucose tolerance tests.
Astoundingly, in just this short week-long period, four out of seven individuals had already developed significantly poorer glycemic responses (NAS responders) and pronounced changes in microbiota composition. The remaining three individuals had no change (NAS non-responders). Transfer of NAS-responders’ day seven stool into GF mice resulted in glucose intolerance (compared to day 1 stool transfer), while transfer of NAS-non-responder day 7 stool did not.
Takeaways from this study
Phew. Hopefully you stayed with me through all that! This was an extremely robust study with lots of different components, but it provides an incredible wealth of evidence against the use of NAS. If you glazed over some of the details, here are the main takeaways:
- NAS consumption in both mice and humans increases the risk of developing glucose intolerance and metabolic disease.
- The adverse metabolic effects are mediated by alterations in the composition and function of the microbiota.
- This study calls for a serious need to reassess the ubiquitous and ever-increasing use of NAS in the food and beverage industry.
Now I’d like to hear from you. Had you seen this study? Will this change your approach to sweeteners? Share your thoughts in the comments!