Also called non-calorific artificial sweeteners, or NAS, the additives are found in diet sodas, cereals and desserts - a huge market for people worried about weight gain and sugar intake.
Some experts recommend NAS for people with Type 2 diabetes, a disease that has attained pandemic proportions, and for a pre-diabetic condition called glucose intolerance, with elevated blood-sugar levels.
But scientists reported in the journal Nature that experiments on lab mice and a small group of humans found NAS disrupted the makeup and function of gut bacteria, and actually hastened glucose intolerance.
"Artificial sweeteners were extensively introduced into our diets with the intention of reducing caloric intake and normalising blood glucose levels without compromising the human 'sweet tooth'," the paper said.
"Our findings suggest that NAS may have directly contributed to enhancing the exact epidemic that they themselves were intended to fight," it said bluntly.
Scientists led by Eran Elinav and Eran Segal of the Weizmann Institute of Science in Israel added three commonly-used types of NAS - aspartame, sucralose or saccharin - to the drinking water of mice in body-size appropriate doses equivalent to recommended maximum human intake.
Those rodents given NAS developed glucose intolerance, whereas mice that drank only water, or water with sugar, did not.
Next, the researchers transplanted faeces from NAS-fed and glucose-fed mice into rodents bred to have no gut bacteria of their own.
The blood-glucose levels of the NAS transplant recipients rose sharply, the team found - and their gut bacteria worked harder than the other group's at extracting glucose from nutrients.
The next step was to apply these insights to humans. Pouring over questionnaires and health data from 381 non-diabetic people, the team found a "significant" link between glucose intolerance and higher NAS consumption.
Finally, the researchers placed seven volunteers who did not normally use NAS on a seven-day regimen that included the maximum sweetener intake recommended by the US Food and Drug Administration (FDA).
Within five to seven days, four developed elevated blood-glucose levels and an altered gut bacteria mix, apparently mirroring the effect in mice.
DEMAND FOR A REVIEW
Past investigations into NAS have delivered mixed results. Some showed benefits in weight loss and glucose tolerance, while others suggested the opposite.
The picture is muddied by the fact that many NAS consumers are people who already have diabetes or are prone to it.
The new experiments are a red flag, the team said. "This calls for a reassessment of today's massive, unsupervised consumption of these substances," said Elinav.
Independent commentators praised the work for its innovation, but warned against overreaction. The human trial involved just seven people over a week, and wider and longer trials are needed to draw any firm conclusion, they said.
"Human diets are complex, consisting of many foods, the consumption of which can vary in amounts, and over time," warned John Menzies of the Centre for Integrative Physiology at the University of Edinburgh in Scotland.
"This research raises caution that NAS may not represent the 'innocent magic bullet' they were intended to be to help with the obesity and diabetes epidemics," Nita Forouhi, a University of Cambridge epidemiologist told Britain's Science Media Centre. "But it does not yet provide sufficient evidence to alter public health and clinical practice."