Do Artificial Sweeteners Break a Fast?

Quick answer: On calories, no — these sweeteners are used in milligram quantities. The real argument is whether sweet taste alone triggers an insulin response, and the human evidence there is mixed and mostly weak. The best experiment separated sucralose taste from swallowing and found taste alone did not raise insulin. Skip them on a strict fast anyway; on a fat-loss fast they are a reasonable tool.

This is the most argued-about question in fasting, and most of the arguing is done with one study quoted badly. Here is each sweetener on its own.

First, the calorie question is settled

The FDA reports these sweeteners as hundreds to thousands of times sweeter than sugar, so the amounts used are tiny: a 60 kg (132 lb) adult would need 23 packets of sucralose, 45 of saccharin or 75 of aspartame in a day to reach the acceptable daily intake.

One rule matters more. Under 21 CFR 101.9, calories "less than 5 calories may be expressed as zero" per serving. "Zero calorie" is a rounding convention, not a measurement — which is why the bulking agent in a packet, not the sweetener, is usually the only thing carrying energy.

Does aspartame break a fast?

No, on any calorie definition, and the insulin evidence is the cleanest of the four.

Aspartame is an approved food additive — a dipeptide of phenylalanine and aspartic acid, about 200 times sweeter than sugar. The FDA notes it "contains calories, but consumers will likely use less than they would of table sugar" for precisely that reason: the quantity in a diet drink or a packet is measured in milligrams. On glucose metabolism, a 2025 systematic review and meta-analysis in Advances in Nutrition found "little to no effects of aspartame consumption on glucose metabolism over the short term or the long term."

The safety noise is a separate question from the fasting one. IARC classified aspartame as "possibly carcinogenic to humans" (Group 2B) on limited evidence in 2023, while the FAO/WHO expert committee, in the FDA's words, "did not raise safety concerns for aspartame under the current levels of use and did not change the Acceptable Daily Intake." Group 2B is a hazard category, not a risk estimate, and none of it bears on fasting.

Verdict: clean for a fat-loss fast. Out for a strict fast, on principle rather than physiology.

Does sucralose break a fast?

No on calories. This is the one sweetener with a real published insulin signal — and the finding is narrower than its reputation.

Sucralose is the sweetener behind the famous study, so read it carefully. In Diabetes Care in 2013, Pepino and colleagues gave 17 adults with obesity, who did not normally use sweeteners, 48 mg of sucralose or water 10 minutes before a 75 g glucose load. Against water, sucralose produced a 20 ± 8% greater insulin area under the curve (p<0.03), a 22 ± 7% greater peak insulin secretion rate (p<0.02), a higher peak glucose and a 23% fall in insulin sensitivity.

Two things are constantly misreported. First, the 20% figure is the insulin AUC, not a "20% insulin spike" — the paper reports no percentage change in peak insulin concentration at all. Second, this was sucralose followed by 75 g of glucose: a study of how sucralose changes your response to a sugar load, not of sucralose in black coffee at hour 14 of a fast.

The same lab then ran the experiment that answers our question. In 2019, Nichol and colleagues gave 21 adults three glucose tolerance tests preceded by sucralose swallowed, water, or sucralose tasted and spat out. Ingested sucralose raised the glucose AUC by 30 ± 10%. Sucralose tasted but not swallowed did not — and sham-fed sucralose actually decreased insulin in the first 60 minutes. The tasting-alone hypothesis, tested directly, failed.

Balance requires the other side: two null trials of sucralose — 13 weeks in 128 people with type 2 diabetes, and 12 weeks in 47 healthy men, both finding no effect on HbA1c, fasting glucose or insulin — were led by an author affiliated with the maker of Splenda.

Verdict: does not break a fast on calories. Taken with carbohydrate it may worsen that meal's response — a reason to watch it in your eating window, not your fasting one.

Does monk fruit break a fast?

No on calories — but the tub in your kitchen is mostly not monk fruit.

Monk fruit sweetness comes from mogrosides in Siraitia grosvenorii Swingle fruit extract, reported by the FDA as 100 to 250 times sweeter than sugar. Like stevia it is not an approved additive; the FDA reviewed GRAS notices and "has not questioned the notifiers' GRAS conclusions."

The human evidence is thin — a few small acute trials in which monk fruit is one arm among several. The best is a 2017 randomized crossover in 30 healthy men comparing monk fruit, aspartame, stevia and sucrose beverages: no difference in glucose AUC (p=0.960) or insulin AUC (p=0.216) over three hours. There is no long-term trial of monk fruit alone.

The practical catch is stevia's. Retail "monk fruit sweetener" is overwhelmingly erythritol with a trace of monk fruit extract — erythritol is listed first on the major brands' panels. So the honest answer here is usually an answer about erythritol.

Verdict: the extract is clean. Check what it is bulked with.

Does erythritol break a fast?

No on calories — US labeling law assigns erythritol 0 calories per gram — and the human glucose data is consistently null. The reason to be cautious is a different one.

Erythritol is a sugar alcohol rather than a high-intensity sweetener, so it is used by the spoonful. 21 CFR 101.9 sets its caloric value at 0 kcal/g for labeling, and the FDA notes sugar alcohols "do not promote tooth decay or cause a sudden increase in blood glucose." EFSA's 2023 re-evaluation of the controlled human feeding studies concluded the evidence "consistently showed no short-term effect of erythritol on postprandial glucose homeostasis in human."

The 2023 cardiovascular headlines deserve an honest summary rather than alarm or dismissal. A Nature Medicine study found people in the highest quartile of blood erythritol had two to three times the three-year risk of heart attack and stroke across three cohorts (hazard ratios 2.95, 1.80 and 2.21). But those cohorts measured erythritol in blood, never intake — and the body makes erythritol from glucose, so a high level may mark metabolic dysfunction rather than sweetener use. The authors state plainly that "these studies can only show association, and not causation."

Verdict: does not break a fast. Whether to use it daily is a separate, still-open question.

The cephalic phase insulin argument, honestly

Almost every "sweeteners break your fast" article rests on the cephalic phase insulin response — the idea that tasting sweetness makes the pancreas release insulin in anticipation. It is a real phenomenon, and much smaller than the argument requires.

A 2020 systematic review in Advances in Nutrition screened 582 papers and analysed 50. An insulin rise appeared in 41% of tested conditions and was statistically significant in only 22%, with a median rise of 2.5 μIU/mL, about 30% above baseline, at five minutes. Its conclusion: "CPIRs are small compared with spontaneous fluctuations… CPRs do not seem to be biologically meaningful in daily life."

A second 2020 systematic review reached the opposite headline — that cephalic phase insulin release does exist in humans, pooled effect size 0.47. Both are real and both can be true: the response exists and is small. Neither has a subgroup for non-nutritive sweeteners, so nobody has pooled the exact question.

What has been tested directly is the sucralose taste-versus-swallow experiment above, where taste alone did not raise insulin. And the broadest acute evidence — a network meta-analysis of 36 trials in 472 participants covering aspartame, sucralose, acesulfame K, saccharin, cyclamate, stevia and blends — concluded these drinks "have no acute metabolic and endocrine effects, similar to water."

The honest bottom line: no primary human study we could verify supports the claim that a zero-calorie sweetener produces an insulin response large enough to interrupt a fast.

Three fasts, three answers

Our does it break a fast lookup grades each item against three separate goals.

A strict clean fast — broken, by definition. A strict fast means nothing but water: a rule about inputs rather than calories, whose whole value is that you never have to litigate it.

A fat-loss fast — not broken. No meaningful calories and no reliable insulin signal. If a sweetened coffee carries you to noon, it is doing more good than harm.

An autophagy fast — unknown. No human evidence exists in either direction, and no test could tell you your status. Unknown is not the same as safe.

The things that actually matter more

• What it is bulked with. In packets, tubs and "natural" blends, the filler is the calories.
• What it does to your appetite. If sweet taste makes your last fasting hours harder and your first meal bigger, that costs more than any insulin argument on this page.

Track where your window really begins and ends with the fasting timer, and see what breaks a fast for the full ruleset.

Related reading

• Does stevia break a fast? — evidence pointing both ways
• Does diet soda break a fast? — the same chemistry in a can
• Do vitamins and supplements break a fast? — gummies are sweetened too

Frequently Asked Questions

Do artificial sweeteners break a fast?

Not on calories — they are used in milligram amounts. The insulin argument against them is weaker than commonly claimed: the direct test of sweet taste without swallowing found no insulin rise, and a network meta-analysis of 36 trials found these drinks behave acutely like water. Out for a strict fast on principle; fine for a fat-loss fast.

Does sucralose spike insulin?

In one specific setting. Given 10 minutes before a 75 g glucose load, 48 mg of sucralose raised the insulin area under the curve by 20% and the peak insulin secretion rate by 22% against water. That is sucralose plus sugar, not sucralose alone — when the same lab had people taste sucralose and spit it out, insulin did not rise.

Does erythritol break a fast?

No. US labeling law assigns erythritol 0 calories per gram, and EFSA's 2023 review found the evidence consistently showed no short-term effect on postprandial glucose. The 2023 cardiovascular headlines came from observational data on blood erythritol levels, not measured intake.

If sweeteners are fine, why does WHO recommend against them?

Because it answers a different question. The WHO's 2023 guideline is a conditional recommendation, on low-certainty evidence, against using non-sugar sweeteners for long-term weight control or to reduce disease risk. It says nothing about insulin during a fasting window, and it is not a safety finding.