Low-calorie sweeteners did not affect gut bacteria in one uniform way in a new laboratory study. Their effects varied by sweetener, bacterial species and the other substances present.
Researchers at the University of Cambridge tested 39 commercially used sweeteners against 25 bacterial species grown in the lab. They also examined what happened when the sweeteners were combined with compounds including caffeine, vanillin and several medications.
The findings, published in Molecular Systems Biology, suggest that studying sweeteners alone may miss potentially important interactions. They do not show that sweeteners damage the human gut, cause disease or create harmful medication interactions in people.
Researchers first grew each bacterial species separately and exposed it to the sweeteners. About three-quarters of the sweeteners changed the growth of at least one species, though the direction and strength of the effects differed considerably.
Some combinations slowed bacterial growth, while others weakened or reversed an effect seen when a sweetener was tested alone.
The researchers identified more than 100 instances in which the presence of another compound changed how a sweetener affected a bacterial species. In 34 cases, the combined effect was stronger. In 68 cases, it was weaker.
“Most of what we know about the potential impact of sweeteners on our health comes from animal research or from population studies,” said Professor Kiran Patil of the Medical Research Council Toxicology Unit at the University of Cambridge. “While these studies have indicated involvement of the microbiome in mediating the effect of sweeteners, it’s difficult to know how sweeteners act in the body — is it through direct interactions with our gut bacteria?”
One combination stood out in the experiments: isosteviol and duloxetine, an antidepressant also prescribed for conditions including anxiety and certain types of chronic pain.
Together, the compounds strongly suppressed the growth of Roseburia intestinalis and Parabacteroides merdae in isolated cultures. Both species have been studied for their possible roles in metabolism and intestinal health, but reducing their growth in a laboratory dish does not establish that the same change occurs in a person or leads to a health problem.
Isosteviol is related to steviol compounds associated with stevia-based sweeteners. The study did not test an ordinary serving of a branded stevia product as a person would consume it, so the findings should not be interpreted as evidence that stevia and duloxetine are a harmful combination.
The researchers also combined the 25 bacterial species into a simplified synthetic community. This allowed them to observe how the organisms competed and responded together rather than in isolation.
In that model, the isosteviol-duloxetine combination reduced microbial diversity and shifted which bacterial species became more or less abundant. Additional experiments found changes in how substances produced by the bacterial community affected cultured host cells involved in toxicity, inflammation and immune responses.
Those results provide possible biological pathways for future research. They do not demonstrate that the combination causes intestinal inflammation, immune dysfunction or other clinical effects in people.
The human gut contains hundreds of bacterial species, along with viruses, fungi and other microorganisms. Its composition also varies substantially from person to person and is influenced by diet, medication use, age, health conditions and many other factors.
The laboratory model could not reproduce digestion, absorption, liver metabolism or the concentrations that would reach the colon after a typical meal, drink or medication dose. It also could not show whether any microbiome changes would be temporary, harmful or offset by other organisms.
“Sweeteners are often marketed as metabolically neutral, but our study challenges this idea,” lead author Dr. Sonja Blasche said. “We found that they can directly affect gut bacteria, particularly when mixed with other compounds such as medication and food additives.”
The term “metabolically neutral” requires context, however. A substance can affect bacterial growth in a laboratory without causing a meaningful metabolic or health effect in a person. Sweeteners also differ chemically, making it misleading to treat them as one interchangeable category.
The study did not compare the health effects of sweeteners with those of sugar, nor did it test whether using a sweetener improves or worsens blood sugar, body weight, cancer risk or another clinical outcome.
It also provides no basis for people taking duloxetine to stop or change their medication. Prescription decisions should be made with a health care professional, and human studies would be needed before researchers could determine whether the laboratory interaction has any practical significance.
The main contribution of the study is more limited but still useful: the biological effects of food ingredients may depend partly on what they are consumed with.
“Artificial sweeteners don’t just pass through the body passively — they can interact with gut microbes, and these effects can be amplified or altered by other substances like medications,” Patil said.
Future studies will need to establish whether the concentrations used in the laboratory resemble real-world exposure, whether similar interactions occur in more complex microbiomes and whether any changes translate into symptoms or measurable health outcomes.
Until then, the findings should be viewed as a starting point for human research rather than guidance to avoid a particular sweetener, food or medication combination.
The research was funded by the European Union’s Horizon 2020 program and the United Kingdom Medical Research Council.