Sweeteners and Blood Sugar: Stevia, Monk Fruit, Allulose, Sugar Alcohols, and Sucralose With Diabetes

Sweeteners are among the first things people reach for after a diabetes or prediabetes diagnosis. The aisle is crowded, the labels are written in marketing language, and the headlines change every year. This guide sorts out what the major sweeteners are, how each shows up on a label, what the evidence says about blood sugar, and where the real uncertainty is. I am not a physician, I am not telling you what to eat, and I will flag every place where I could not verify a number.

Quick answer: Most people with diabetes can use non-sugar sweeteners without a direct rise in glucose, and Mayo Clinic says the sweeteners themselves do not affect blood sugar. The catch is the product around the sweetener. Sugar-free foods are rarely carbohydrate-free, sugar alcohols carry calories and can raise glucose a little, and bulking agents like dextrose or maltodextrin sneak into packets and blends. The World Health Organization advised in 2023 against using non-sugar sweeteners to control body weight, but that guidance explicitly does not apply to people with pre-existing diabetes. Your own readings and your own clinician are the final test.

How to think about sweeteners before you compare them

Four different things get called “sweeteners,” and mixing them up causes most of the confusion.

First are sugars and syrups: table sugar, honey, maple syrup, agave. They are carbohydrate and they raise glucose. Honey and maple syrup sound gentler, but I found no evidence that they get a pass for people with diabetes. They count.

Second are high-intensity sweeteners, which are many times sweeter than sugar, so only tiny amounts are used. FDA lists saccharin, aspartame, sucralose, acesulfame potassium, neotame and advantame as approved food additives, and treats certain highly purified stevia and monk fruit extracts through GRAS notices, meaning FDA has not questioned the notifiers’ conclusion that they are generally recognized as safe for their intended uses. The same FDA page says whole-leaf and crude stevia extracts are not permitted as sweeteners, so “stevia leaf” supplements are a different product from the purified extract in a packet.

Third are sugar alcohols, also called polyols: erythritol, xylitol, sorbitol, maltitol, mannitol, isomalt and lactitol. They are not sugar and they are not alcohol in the drinking sense. FDA says they range from 25 percent to 100 percent as sweet as sugar. They contain some calories and they are the group that creates label math.

Fourth is allulose, sometimes called a rare sugar. It is a real sugar chemically, but FDA treats it differently on labels because the body uses very little of it for energy. I explain that below.

One more note on the WHO guideline, since it is cited everywhere. WHO’s definition of non-sugar sweeteners covers acesulfame K, aspartame, advantame, cyclamates, neotame, saccharin, sucralose, stevia and stevia derivatives. It does not count low-calorie sugars and sugar alcohols, because they contain calories. So when someone says “WHO says sweeteners are bad,” they are usually describing a narrower, more conditional finding than the headline suggests.

The sweetener comparison table

Read this table as a starting map. Where I could not verify a number from a source I could open, the cell says so rather than guessing. Calories and carbs are per typical serving, and packet blends are often different from the pure sweetener, so your label always wins.

Table notes: The sugar and honey serving figures are standard reference values that I did not re-fetch for this guide. The Truvia packet figure comes from a ReciPal ingredient entry (1 packet, 3.5 g, 1 calorie, 3 g carbohydrate, all sugar alcohol) and could differ from the box you hold. FDA’s allulose guidance is from October 2020. The sugar alcohol calories are FDA values reproduced by Kansas State University. GI values are left unverified on purpose; published GI numbers differ by study design and by who is tested.

Sugar alcohols: calories, glucose, and what they do to your gut

Sugar alcohols sit in an awkward middle. The ADA Diabetes Food Hub says sugar-free foods made with them are not carbohydrate-free, that the amount appears under Total Carbohydrate on the label, and that all types raise blood glucose less than regular sugar. It adds that some raise it a little, others not at all, and that erythritol may not raise it. Mayo Clinic is blunter: sugar alcohols such as mannitol, sorbitol and xylitol have about half the calories of sugar and can raise blood sugar, and for some people they cause diarrhea. Both statements can be true at once, because the sugar alcohols are different molecules.

The calories-per-gram column is why maltitol deserves a second look. At 3.0 calories per gram it sits closest to sugar’s roughly 4. Erythritol is the outlier at the low end, which is part of why it is popular in keto and low-carb baking blends, but I could not open a source with its exact figure, so treat “near zero” as a commonly cited range, not a verified number.

On digestion, the ADA Diabetes Food Hub notes that many sugar alcohols can cause gas, bloating and stomach aches, especially in large amounts, and that sensitivity varies. Kansas State’s summary lists reported laxative thresholds of roughly 50 to 80 grams per day for sorbitol and roughly 20 grams per day for mannitol. Those are population-level figures; you may react to far less. If a product has a warning like “excess consumption may have a laxative effect,” take it literally. It is also a practical reason to introduce a new sugar-free product one serving at a time, not three bars in a row.

Label math: how to count sugar alcohols

The Nutrition Facts panel lists sugar alcohols voluntarily under Total Carbohydrate. FDA’s interactive label guidance says manufacturers may list them, may name a single sugar alcohol if only one is used, and must list them if the package makes a claim about sugars or health effects. FDA’s page does not tell people with diabetes how to subtract anything, and I could not find the ADA Diabetes Food Hub page giving a subtraction rule either.

The rule most people have heard, “subtract half the sugar alcohol grams,” is taught by the University of California San Francisco Diabetes Teaching Center. Their example: a product with 29 g total carbohydrate and 18 g sugar alcohol is counted as 20 g, because 18 divided by 2 is 9 and 29 minus 9 is 20. I want to be careful about attribution. I verified the half-rule at UCSF, not at the ADA or FDA, so please do not repeat it as an ADA or FDA rule. Some educators use different approaches for erythritol or for fiber, and I did not verify those variations. Your dietitian or diabetes educator can tell you which method fits your plan and your medication.

Two more label points. Allulose is the exception to “it is on the sugar line”: FDA said in October 2020 that it would not enforce including allulose in total sugars and added sugars, pending future rulemaking, but it must still be counted in total carbohydrate, and FDA allows 0.4 calories per gram. That means an allulose-sweetened treat can show low sugar and still show real total carbohydrate. And “net carbs” is a marketing phrase, not an FDA-defined term on the panel. I did not find an FDA definition of it, so do not assume two brands calculate it the same way. For a full walkthrough of the panel itself, see my guide to how to read a nutrition label for carbs.

Worked example: a “sugar-free” bar

This is a made-up label, not a real product, calculated in code so the arithmetic is checked. Suppose a bar lists 24 g total carbohydrate, 9 g fiber, 2 g sugars and 10 g sugar alcohol, with maltitol as the main sugar alcohol.

Under the UCSF half-rule, the bar counts as 19 g, not 0 and not 24. If the sugar alcohol were maltitol, the 10 g contribute about 30 calories; if xylitol, about 24. I did not subtract fiber in this example because I did not verify a fiber rule, and some plans handle fiber differently. The takeaway is that “sugar-free” on the front still leaves a real number on the back, and the half-rule is a rough estimate that your own glucose reading can confirm or contradict.

A second computed example, for packets. Four Truvia-style packets in a day, at 3 g sugar alcohol and 1 calorie each per the ReciPal entry, add up to 12 g sugar alcohol and 4 calories. That is small, but it is not zero, and it matters if you also eat a sugar-free dessert the same day. The point is stacking: individually trivial items can add up.

Safety, side effects, and the recent headlines

Erythritol and the 2023 heart-event study

In 2023, Cleveland Clinic researchers published a study in Nature Medicine reporting that higher blood levels of erythritol were associated with a higher risk of major adverse cardiovascular events such as heart attack, stroke or death. According to a summary of the paper, the team screened blood samples from 1,157 people undergoing cardiac risk assessment, then checked the finding in two validation cohorts, 2,149 people in the United States and 833 in Europe, all being evaluated for heart disease. The first cohort was followed for about three years.

The same work reported laboratory results: at physiological levels, erythritol increased platelet reactivity in vitro and promoted clotting in vivo, and in eight healthy volunteers who drank erythritol, blood levels stayed elevated for more than two days. This is where caution is needed. The human association is observational. It does not show that eating erythritol caused the events. The participants were people already being evaluated for cardiac risk, not a general population of people with diabetes using a sweetener. The authors themselves said further studies on long-term safety are warranted, especially in people with kidney disease or existing cardiovascular disease. They noted that human safety trials had lasted up to four weeks. A sensible reading is “worth discussing with your clinician if you have heart or kidney risk,” not “erythritol causes heart attacks.”

The WHO 2023 non-sugar sweetener guideline

In May 2023 WHO advised against using non-sugar sweeteners to control body weight or reduce the risk of noncommunicable diseases. The recommendation is rated conditional, WHO’s own wording says the review found no long-term benefit for reducing body fat in adults or children, and it describes possible long-term harms in adults, such as type 2 diabetes, cardiovascular disease and mortality, with hedged language (“may be”). WHO also noted that the observed links might be confounded by participants’ baseline characteristics and complex patterns of use. Importantly for this audience, WHO states that the recommendation does not apply to people with pre-existing diabetes. I could not open the full guideline, so I cannot say how much of the evidence was observational versus randomized; the WHO news page itself does not label the study designs.

Aspartame

In July 2023 the International Agency for Research on Cancer classified aspartame as Group 2B, “possibly carcinogenic to humans,” citing limited evidence for liver cancer in humans. In the same announcement, the joint WHO/FAO expert committee (JECFA) reaffirmed an acceptable daily intake of 40 mg per kilogram of body weight. WHO’s worked example is that a 70 kg adult would need to drink more than 9 to 14 cans of a diet drink containing 200 to 300 mg of aspartame to exceed that limit, assuming no other sources.

I recomputed that: 40 mg times 70 kg is 2800 mg per day, and 2800 divided by 300 is about 9.3 cans, while 2800 divided by 200 is 14 cans. So the numbers hold. People with phenylketonuria must avoid aspartame, and FDA requires a phenylalanine statement on products that contain it.

Sucralose and insulin sensitivity

A 45-person study led by Yale researchers and published in Cell Metabolism had adults aged 20 to 45 drink beverages containing sucralose alone, sucralose with sugar, or sucralose with maltodextrin over about a week or two (the news summary I read gave both). The sucralose-with-maltodextrin group showed changes in insulin sensitivity and glucose handling, while the maltodextrin-only group did not. The authors proposed that sucralose may change how the body handles a carbohydrate consumed at the same time. This was a small, short study in adults who did not usually consume low-calorie sweeteners, and the news summary did not give the sucralose dose or study limitations. It is a hypothesis-generating result, not proof that a sucralose packet in your coffee harms you.

Stevia, monk fruit and allulose

For stevia, a randomized trial in people with type 2 diabetes exists, but I could not open it, so I will not quote its results. For monk fruit, I found no trial data in people with diabetes that I could verify. For allulose, I found a meta-analysis of clinical trials in type 2 diabetes and a separate meta-analysis in healthy people, but could not open either, so I will not claim an effect size. What I can say is that these are the less-studied options, and the honest statement is that long-term outcome trials are limited.

What labels to look for

Before looking at the sweetener claim on the front, flip the package and run this checklist.

1. Total carbohydrate, not “sugar-free.” A cookie can be sugar-free and still contain flour. Total carbohydrate is the number your meal plan was probably built around.

2. The sugar alcohol line. Grams of sugar alcohol under Total Carbohydrate tell you how much of that total may be only partly absorbed. Check the ingredient list for which one: erythritol, maltitol and sorbitol behave differently.

3. Allulose and “0 sugar” claims. Because FDA lets allulose stay off the sugar lines, check total carbohydrate and ingredients when you see a “low sugar” claim.

4. Bulking agents. Dextrose and maltodextrin appear in many packet and blend products. A calorie-free sweetener plus a bulking agent is not calorie-free.

5. Serving size. The label may describe a teaspoon when you pour a tablespoon, or one cookie when the package holds six.

6. Other carbohydrate in the product. Mayo is explicit that artificial sweeteners themselves do not affect blood sugar but other ingredients in products that contain them might. Flour, starches, fruit and syrups still count.

Using sweeteners at home: coffee, baking and cooking

The table below mixes label-reading advice with practical baking notes. The baking notes reflect general food-science and kitchen practice, not clinical trials, and the way a sweetener behaves in a recipe depends on the brand and blend.

A practical point about baking: sugar does more than sweeten. It adds bulk, moisture and browning. When you swap it out, a “1 for 1” blend can still change the texture, and the total carbohydrate per slice depends on the flour, fruit and other ingredients as much as on the sweetener. If you want the bigger picture of building a plate, my diabetes food guide and the type 2 diabetes diet guide cover the other pieces.

Scenarios

“I only want sweetener in my coffee”

This is the lowest-stakes case. A liquid or powder stevia, monk fruit or sucralose option generally adds very little carbohydrate. The things to check are whether your packet is a blend with dextrose or sugar alcohol, how many you use, and whether your coffee also includes flavored creamer or syrup, which can contain sugars or sugar alcohols.

“I drink diet soda instead of regular”

Mayo says the sweeteners themselves do not affect blood sugar. WHO says its caution is about weight control and long-term disease risk in the general population, and says it does not apply to people with diabetes. If you notice cravings or intake creeping up, that is worth discussing with your care team. I would not frame diet soda as either a health food or a danger on the evidence I could verify.

“I buy sugar-free desserts and bars”

This is where most people get surprised. The sugar alcohol amount, the starch, and the serving size drive the result. Use the label math above for a rough count, then check your own reading. For a method, see my post-meal glucose tracker and the guide to reading CGM data. Because my goal is education, not a personal plan, I will not tell you what your reading “should” be.

“I take insulin or a sulfonylurea”

Sugar-free products are not used to treat a low. If your regimen can cause hypoglycemia, how you treat lows and how you dose for a sugar-free dessert are conversations for your care team. Please do not change doses based on a blog.

“I have heart or kidney disease”

This is the group the erythritol authors specifically called out for further study. It is a reasonable question for your clinician: how much sugar alcohol, if any, fits your situation?

“I am cooking for family”

Whole-family cooking often goes better with small swaps than with a separate “diabetic” version. Reducing the sugar in a recipe, using fruit for some sweetness, and serving a modest portion can be simpler than a full substitution, and none of that needs a special product. See the broader complete guide to type 2 diabetes for how food fits with the rest of care.

Common mistakes

Treating “sugar-free” as “carb-free.” The label total decides.

Subtracting all sugar alcohols. The only rule I verified, from UCSF, is to subtract half. Some sugar alcohols have meaningful calories and can raise glucose.

Quoting the half-rule as an ADA or FDA rule. I could not verify that attribution.

Ignoring the bulking agent. Packets and blends sometimes contain dextrose or maltodextrin.

Eating several servings at once. Sugar alcohols, especially sorbitol and maltitol, are the classic cause of cramping and diarrhea when overdone.

Assuming “natural” means no effect. Honey, agave and coconut sugar are still sugars; stevia in a blend is only as good as the rest of the blend.

Using a single headline to decide. Observational studies like the erythritol cohorts show associations, not proof, and trials show specific effects in specific conditions.

Myth versus fact

Myth: Honey or maple syrup is safe for diabetes because it is natural. Fact: They are sugars and count as carbohydrate.

Myth: Artificial sweeteners spike blood sugar. Fact: Mayo says they do not affect blood sugar themselves. Product ingredients can.

Myth: WHO banned sweeteners. Fact: WHO issued a conditional recommendation not to rely on non-sugar sweeteners for weight control, and said it does not apply to people with pre-existing diabetes.

Myth: Erythritol is proven to cause heart attacks. Fact: The 2023 findings were associations in cardiac-risk cohorts with supportive lab data, and the authors called for more long-term studies.

Myth: Net carbs is an official label number. Fact: I did not find an FDA definition; calculations vary by brand.

Questions to bring to your dietitian or care team

Which sweeteners, if any, fit my medications and my kidney and heart history? How should I count sugar alcohols, and does the same method apply to erythritol? How should I treat a low if I eat sugar-free products? Can we test a product together with my glucose readings? Is a sugar-free product actually better for me than a smaller portion of the regular one?

How to test a sweetener product yourself

If you check glucose with a meter or a sensor, a simple approach is to eat the product in a normal serving by itself, note the time, and compare a before reading with a reading a couple of hours later, as the ADA Diabetes Food Hub suggests (it names 1.5 to 2 hours after eating). One reading is a clue, not a verdict, because meals, stress, sleep and activity all move glucose. Repeat it, keep the portion the same, and bring the log to your clinician.

Frequently asked questions

What is the best sweetener for diabetes?

There is no single best one. Purified stevia, monk fruit, sucralose and aspartame are not carbohydrates, and Mayo Clinic says artificial sweeteners themselves do not affect blood sugar. The product around the sweetener matters more, so check total carbohydrate and bulking agents, and ask your care team what fits your situation.

Can I use honey or maple syrup instead of sugar?

They are sugars and count as carbohydrate. I found no evidence that they are a free pass for people with diabetes, so they should be counted like other sugars.

Do sugar alcohols raise blood sugar?

Usually less than sugar, according to the ADA Diabetes Food Hub, but not always zero. Mayo Clinic says they can raise blood sugar. Erythritol may not, per the ADA Food Hub. Maltitol has the highest calories per gram of the common ones at 3.0.

How do I subtract sugar alcohols on a label?

One widely taught approach, from the UCSF Diabetes Teaching Center, is to subtract half the grams of sugar alcohol from total carbohydrate. I verified it at UCSF, not at the ADA or FDA. Ask your dietitian which method fits your plan.

Is allulose counted as a carbohydrate?

Yes on the label. FDA said in October 2020 that allulose must still be counted in total carbohydrate, though it will not enforce including it in total sugars and added sugars, and allows 0.4 calories per gram.

Is erythritol safe?

Erythritol is widely used. A 2023 Nature Medicine study found higher blood erythritol levels associated with more major cardiovascular events in people being evaluated for heart disease, plus supportive lab findings. It is observational and the authors called for more long-term research, especially for people with kidney or heart disease. Discuss it with your clinician if that applies to you.

Did WHO say people with diabetes should avoid sweeteners?

No. WHO advised against relying on non-sugar sweeteners for weight control in 2023, and states the recommendation does not apply to people with pre-existing diabetes. The recommendation is conditional.

Is aspartame dangerous?

In July 2023 IARC classified it as possibly carcinogenic (Group 2B, limited evidence), while the WHO/FAO committee reaffirmed an acceptable daily intake of 40 mg per kg of body weight. People with PKU must avoid it.

Can sugar-free candy cause stomach problems?

Yes. Sugar alcohols such as sorbitol and maltitol can cause gas, bloating and diarrhea, especially in larger amounts, and sensitivity varies from person to person.

Should I use sugar-free products to treat low blood sugar?

No. Sugar-free products are designed to avoid sugar. If your medicines can cause lows, ask your care team how to treat them.

Sources

• WHO: advises not to use non-sugar sweeteners for weight control (May 15, 2023)
• WHO: aspartame hazard and risk assessment results released (July 14, 2023)
• ADA Diabetes Food Hub: What are sugar alcohols?
• UCSF Diabetes Teaching Center: counting sugar alcohols
• FDA: allulose labeling guidance (October 2020)
• FDA: high-intensity sweeteners
• FDA: interactive Nutrition Facts label, sugar alcohols
• Kansas State University: reduced-calorie sweeteners (polyols)
• Mayo Clinic Care Network: artificial sweeteners and blood sugar
• Nutraceuticals World summary of the 2023 Nature Medicine erythritol study (Cleveland Clinic)
• Endocrine News: sucralose with carbohydrates and insulin sensitivity (Cell Metabolism)
• ReciPal: Truvia packet nutrition entry

About the author: Charles Kirkland is the founder of Metabolic Clarity Labs and author of The Journey Back. He is not a physician. This is general education, not a medically reviewed treatment plan.

Sources checked October 31, 2026. Discuss personal decisions with a qualified clinician. Do not start, stop, or change prescribed treatment based on this article.