Supplements · Sweeteners · Metabolic & Cardiometabolic
stevia does not increase blood glucose
In plain terms: Is stevia safe for your blood sugar?
Part of: • Stevia
Yes - this is stevia's most solid benefit. Your body doesn't turn steviol glycosides into glucose, so stevia doesn't raise blood sugar, and swapping it in for sugar modestly lowers the post-meal glucose and insulin bump. A few of the studies are industry-linked, but the finding is uncontested and makes mechanistic sense.
🧭 Read this score with how it was built in mindA study was counted as evidence for something it never varied
One of the studies removed here was the only one on this page pointing in more than one direction. It tested a drink sweetened with sucrose AND stevia together, so its blood-sugar rise cannot be attributed to the stevia. Two others tested stevia inside a multi-ingredient supplement, or measured only which drink rats preferred and no blood sugar at all. What remains is a set of trials in which stevia itself was the thing that changed, and they agree; the price of that cleanliness is that the page now shows no disagreement, because the disagreement lived in the studies that could not isolate stevia.
📅 Last reviewed: 2026-07-15 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: All trials, pooled (Meta-analysis)
How the studies fall
What the evidence shows
Stevia is blood-sugar-friendly: steviol glycosides aren't absorbed intact or converted to glucose (gut bacteria cleave them to steviol, which is excreted), so stevia doesn't raise blood sugar — and when it replaces sugar it modestly lowers the post-meal glucose and insulin rise. This is well established and mechanistically clear, the clearest of stevia's health claims.
The evidence (11)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Gregersen S, Jeppesen PB, Holst JJ, Hermansen K 2004 · Metabolism | RCT | supports | moderate | Acute cross-over trial in 12 people with type 2 diabetes: one standard test meal supplemented with 1 g of stevioside, compared against the same meal with 1 g of maize starch. Stevioside cut the incremental area under the post-meal glucose curve by 18% (P=.013), so blood sugar ran lower after the stevioside meal, not higher. Insulin, GLP-1 and GIP curves were not significantly changed. This is a single meal in a small group; the trial does not test regular stevia use or stevia replacing sugar. |
| Zare M, Zeinalabedini M, Ebrahimpour-Koujan S, Bellissimo N, Azadbakht L 2024 · Clin Nutr ESPEN | meta-analysis | supports | moderate | Meta-analysis of RCTs on stevia and glycemic indices: favorable/neutral glycemic profile, no glucose-raising effect. |
| Smeets PA, Veit R, Oosterink E, Meijboom S, Risso D, Preissl H, Kullmann S 2026 · Am J Clin Nutr | RCT | supports | moderate | Randomized crossover in 30 healthy fasted adults, six 500-mL drinks: water, or equisweet flavored water containing 25 g sucrose, sucralose, stevia extract, allulose plus stevia, or monk fruit extract. Only the sucrose drink raised serum glucose and insulin; the stevia and allulose-plus-stevia drinks did not, though allulose plus stevia slowed gastric emptying as much as sucrose did. Glucose and insulin were secondary measures (the primary outcome was brain blood flow) and the abstract gives no glucose or insulin values. A co-author is employed by sweetener maker Tate & Lyle. |
| Anker CCB, Rafiq S, Jeppesen PB 2019 · Nutrients | meta-analysis | supports | moderate | Pooled 7 trials (9 RCTs, 462 people) of stevioside 200-1500 mg/day, lasting 4 hours to 2 years. Fasting glucose was essentially unchanged versus placebo (-2.63 mg/dL, 95% CI -7.77 to 2.51, p=0.32) and HbA1c was flat (0.00%, 95% CI -0.24 to 0.25, p=0.98, with no disagreement between trials). No subgroup - diabetic or non-diabetic - showed a rise in blood sugar; the authors report no indication of reductions either. Most included trials were designed to test tolerability of stevia as a sweetener rather than to test glucose control. |
| Almiron-Roig E, Navas-Carretero S, Castelnuovo G, Kjølbæk L, Romo-Hualde A, Normand M, Maloney N, Hardman CA, Hodgkins CE, Moshoyiannis H, Finlayson G, Scott C, Raats MM, Harrold JA, Raben A, Halford JCG, Martínez JA 2023 · Appetite | RCT | supports | moderate | Double-blind randomised crossover trial (project SWEET) in 60 adults with overweight or obesity: a 330 mL zero-calorie beverage with one of three sweetener blends, or one with 8% sucrose (26 g), just before a carbohydrate-rich breakfast. All three blends cut the 2-hour insulin iAUC (p<0.001), but only the stevia RebA-thaumatin and sucralose-acesulfame-K blends also cut the glucose iAUC (p<0.05). Stevia was never given alone - RebA came with thaumatin, RebM with mogroside V. LDL-cholesterol ran 3% higher after stevia RebA-thaumatin than after sucrose (p<0.001). EU Horizon 2020 funding. |
| Núñez Martínez P, Argüelles Luis J, Perillán Méndez C 2016 · Psicothema | mechanism | supports | low | Two-bottle preference tests in Wistar rats. The rats 'displayed preferences for stevia extract and pure rebaudioside A solutions over water at a range of concentrations (0.001% to 0.3%), and their intake peak occurred at 0.1% concentration', and they also preferred a commercial rebaudioside A plus erythritol mixture (Truvia-type) to water, peaking at 2%. The study measured only voluntary drinking; it does not report blood glucose, insulin or any other glycemic measurement. The authors offer it as dosage guidance for later behavioural and physiological work. |
| Galicia-Ayala C, Klünder-Klünder M, Vilchis-Ordoñez A, López-Martínez B, Miranda-Lora AL 2026 · Nutrition | RCT | mixed | moderate | Randomized crossover trial in 20 healthy adults comparing 355 mL colas sweetened with sucrose, with aspartame/acesulfame-K, and with sucrose plus stevia against carbonated water. The only stevia exposure was a sucrose/stevia blend, with no stevia-alone drink, so nothing here isolates stevia: that blend showed a glucose-lowering effect at 60 minutes (-12.5 mg/dL, P<0.05) and a rise in pancreatic polypeptide peaking at 120 minutes (+73.4 pg/mL, P<0.05), and the abstract does not say which drink the fall was measured against. |
| Lifshitz Y, Saban R, Paz S, Gorzalczany Paz K, Samish I, Jenkins AL, Zeltser D, Eldor R 2026 · Food Chemistry | RCT | supports | moderate | Double-blind randomized crossover trial in 19 healthy adults (NCT06520293) comparing equally sweet beverages: the sweet protein sweelin (0.051 g), stevia Reb-M (0.225 g), or dextrose (75 g), with blood sampled over 120 minutes. Glucose and insulin after sweelin were significantly lower than after dextrose at every timepoint and by iAUC (p<0.0001), and no different from stevia. The paper reports no separate test of stevia against dextrose or baseline, so stevia's flat response rests on its equivalence to sweelin. Funded and mostly authored by sweelin's maker. |
| Angarita Dávila L, Durán Agüero S, Apaicio D, Parra K, Uzcátegui M, Céspedes V, Reina Villasmil N, López Miranda J 2017 · Nutr Hosp | RCT | supports | low | Randomized crossover test in 19 healthy adults of a nutritional supplement formulated with stevia and L-carnitine, compared against a glucose solution and white bread, with blood sampled to 120 minutes. The supplement's glucose incremental area under the curve was lower than both references (11,778.73 vs 13,724.06 and 13,153.56, p 0.005), its glycemic index 62 and glycemic load 16 against white bread's 69 and 18, with no difference in postprandial insulin. The authors state that the other nutrients in the formula mean the result cannot be attributed to the sweetener alone. |
| Magnuson BA, Carakostas MC, Moore NH, Poulos SP, Renwick AG 2016 · Nutr Rev | mechanism | supports | low | Narrative review of sweetener biological fate. 'The objectives of this review' are to compare 'status, and biological fate (including absorption, distribution, metabolism, and' excretion of five low-calorie sweeteners. No glycemic outcome is reported anywhere in the readable text, and the previously recorded detail (steviol glycosides cleaved by gut bacteria, excreted, not converted to glucose) does not appear in it. Off-scope on the outcome; would at best be a review-role vote even if the full text carries that statement. Industry-adjacent authorship: 'S.P. Poulos is with the Calorie Control Council, Atlanta, Georgia, USA.' |
| Stamataki NS, Crooks B, Ahmed A, McLaughlin JT 2020 · Nutrients | RCT | supports | moderate | 12-week randomised open-label trial in 28 healthy adults: daily stevia extract did not raise blood glucose. Glucose iAUC across an oral glucose tolerance test was 132 mmol/L x min at baseline and 133 at week 12 on stevia, versus 131 then 159 in the untreated control group, and peak glucose went 7.08 to 6.91 mmol/L on stevia. Insulin responses were likewise unchanged. Fasting glucose stayed inside the normal range (3.9-5.5 mmol/L) in every participant. |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.