Direct answer
Purified steviol glycosides have product-specific regulatory safety assessments, while therapeutic and lasting microbiome benefits remain uncertain. Ingredients, manufacturing, comparators and commercial involvement differ across trials; whole-leaf and crude stevia are not interchangeable with purified glycosides.
- "Stevia" covers different ingredients. FDA has not questioned GRAS conclusions for specified high-purity steviol glycosides under their notified uses; whole-leaf and crude extracts are not permitted as sweeteners in the US (FDA sweetener guidance).
- A 2015 review reported nine studies and 756 participants, with tentative stevioside glucose/BP signals and no significant Reb A benefit. Variable quality, extreme heterogeneity and incomplete trial-finance reporting limit confidence (Onakpoya & Heneghan 2015).
- Human microbiome findings vary: an exploratory 12-week drops analysis found no overall diversity change but some taxon differences; a separate two-week trial found functional changes without significant stevia-related glycemic impairment. These short, product-specific studies do not establish lasting clinical harm or a universal absence of effects.
- EFSA and JECFA separately use an ADI of 4 mg/kg body weight/day as steviol equivalents. Conservative exposure models can exceed it in some high-consuming infants and toddlers; JECFA considers those modeled exceedances unlikely to pose a health concern (EFSA 2024; JECFA current entry).
- The plant is native to parts of Brazil and Paraguay (Kew). Commercial ingredients can be leaf-extracted, enzyme-converted or fermentation-derived. Cargill’s GRAS 882 dossier describes one engineered-yeast process; it is not a census of today’s global supply or production methods.
- Regulatory safety conclusions concern specified purified ingredients and uses. Evidence for therapeutic benefits is Limited; equivalent safety of whole-leaf/crude stevia is unestablished.
Table of contents
- Evidence summary
- What stevia is
- Common forms and product differences
- How it works
- Benefits by claim
- What works and what does not
- Risks and reported side effects
- Medication considerations
- Precautions and intake
- Dosage and how to take
- Preclinical evidence and its limits
- Funding and conflicts
- Frequently asked questions
- Sources and funding notes
| Claim | Evidence | Source | Funding/conflict | Strength |
|---|---|---|---|---|
| Stevioside may lower fasting blood glucose in some trial settings (Primary source) | Nine-study review reported 756 participants overall; small glucose signal, with inconsistent longer-term findings (Primary source) | Onakpoya & Heneghan 2015 | No specific review grant; three trials reported government/public-institution support and six did not specify funding. Reb A trial authors had manufacturer affiliations or support. | Limited by heterogeneous trials (Primary source) |
| Stevioside may lower diastolic blood pressure in some trial settings (Primary source) | Same review: aggregate systolic result was nonsignificant; stevioside subgroup favored lower BP, with extreme heterogeneity (Primary source) | Onakpoya & Heneghan 2015 | No specific review grant; three trials reported government/public-institution support and six did not specify funding. Reb A trial authors had manufacturer affiliations or support. | Limited by heterogeneous trials (Primary source) |
| Reb A affects blood pressure or cardiometabolic risk | No significant effect found | Onakpoya & Heneghan 2015 | No specific review grant; three trials reported government/public-institution support and six did not specify funding. Reb A trial authors had manufacturer affiliations or support. | Weak (null finding) |
| Stevia vs. sucralose in type 2 diabetes (eight weeks; 39 randomized, 34 completers) (Primary source) | No significant difference in fasting glucose, HbA1c, insulin, or lipids | PMC7103435 | NNFTRI (Tehran) support; commercial extract supplied by Parmida Mehr Pasargad (Primary source) | Limited: small trial, null results (Primary source) |
| Exploratory 12-week stevia drops microbiome analysis | 27 participants; no overall alpha/beta diversity change. Some exploratory taxon differences, several already present at baseline. | PubMed 38257188 | N8 AgriFood and BBSRC CASE support; no conflicts declared. CASE partner and supply terms not specified in this report. | Limited: exploratory analysis (Primary source) |
| Whole-leaf/crude stevia is safe for regular human consumption | US sweetener use is not permitted; equivalent safety to purified glycosides is not established (Primary source) | FDA Import Alert 45-06 | Regulatory determination | Insufficient |
| Regulatory ADI for specified steviol glycosides | 4 mg/kg/day as steviol equivalents retained; conservative high-consumer exposure estimates can exceed it. JECFA considers the modeled exceedance unlikely to pose a health concern. | EFSA 2024; JECFA | Official regulatory assessments; EFSA evaluated an International Stevia Council dossier. | Regulatory conclusion for specified ingredients and exposure assumptions |
What stevia is
Stevia rebaudiana is a shrub native to parts of Brazil and Paraguay (Kew botanical record), with traditional Guaraní use described in the 2015 Public Eye report. Its sweet compounds are steviol glycosides; commercial ingredients may be extracted from leaves or produced through other specified processes. FDA describes these ingredients as roughly 200–400 times sweeter than sugar, while the sweetness and concentration of finished drops or blends vary (FDA). In the US, whole-leaf and crude stevia are not permitted as conventional-food sweeteners. FDA has not questioned GRAS conclusions for certain specified high-purity glycoside ingredients under their notified uses. Purity, manufacturing process and intended use all matter.
Common forms and product differences
Stevia reaches consumers in several distinct product forms with very different regulatory and evidentiary status. The table below summarizes the main forms found on the market.
| Form | Composition | Purity/grade | Regulatory status | Sweetness vs. sucrose |
|---|---|---|---|---|
| Whole-leaf / green stevia powder | Dried, ground leaves; full phytochemical matrix (steviol glycosides plus flavonoids, chlorophyll, other plant compounds) | Not purified; variable glycoside content | Not permitted as a sweetener in the US. Rules in other jurisdictions require a separate product-specific check (FDA sweetener guidance). | Variable, generally lower/less consistent than refined extract |
| Refined white steviol glycoside extract | Purified stevioside, Reb A, Reb D, Reb M, or enzyme-modified glycosides | At least 95% steviol glycosides in the specific high-purity preparations described by FDA (Primary source) | Specified high-purity glycosides have no-questions GRAS responses for their notified uses; check the actual formulation (FDA sweetener guidance). | FDA describes steviol glycoside ingredients as roughly 200–400 times sweeter; potency varies with composition (FDA). |
| Liquid stevia drops/extract | Steviol glycoside ingredient diluted in a liquid; glycoside composition and concentration vary by product. | An ingredient may meet a ≥95% glycoside specification; this is not the glycoside percentage of the finished liquid. | Specified high-purity glycosides have no-questions GRAS responses for their notified uses; check the actual formulation (FDA sweetener guidance). | Product-specific; use the label’s sweetness equivalence. |
| Erythritol/maltodextrin stevia blends | Purified glycosides combined with bulking ingredients such as erythritol, maltodextrin or dextrose; exact formulations vary. | Check the glycoside ingredient specification separately from the composition of the finished blend. | Specified high-purity glycosides have no-questions GRAS responses for their notified uses; check the actual formulation (FDA sweetener guidance). | Product-specific; a 1:1 volume replacement cannot be assumed. |
| Tabletop packets | Purified glycosides, often with carrier or bulking ingredients; check the actual packet. | The glycoside ingredient specification is separate from the entire packet’s purity. | Specified high-purity glycosides have no-questions GRAS responses for their notified uses; check the actual formulation (FDA sweetener guidance). | Follow the product’s labeled sugar equivalence; packet sizes and formulas vary. |
Glycosides: stevioside vs. Reb A vs. Reb M vs. Reb D
Stevia leaves contain a family of steviol glycosides that differ in sweetness intensity, bitterness, and natural abundance:
- Stevioside and rebaudioside A (Reb A) are major conventional leaf glycosides. The cited human reviews mainly assessed stevioside or Reb A; that evidence does not establish benefits for every glycoside or blend (EFSA 2024; 2015 review; 2019 review).
- Reb M and Reb D occur in lower quantities in conventional leaves. Cargill’s GRAS 882 dossier describes predominantly Reb M, with smaller amounts of other glycosides, produced using engineered Yarrowia lipolytica yeast. It reports identity data matching leaf-derived Reb M and Reb D. Leaf extraction and enzyme conversion are other production routes; this dossier does not establish which route dominates the market (GRAS 838 dossier; EFSA 2024).
- GRAS notices concern specified ingredients, manufacturing processes, specifications and intended uses. Examples include the manufacturer dossiers for Notice 702 and Notice 838, and FDA’s response to Notice 790 for specified jerky uses. Meeting a purity threshold alone does not establish every product’s status. Whole-leaf and crude stevia are not permitted as US conventional-food sweeteners (FDA guidance).
How it works
Steviol glycosides are not digested in the upper gastrointestinal tract; colonic microbes hydrolyze them to steviol before absorption and further metabolism (EFSA 2024). A laboratory study found Reb A stimulated GLP-1 release in murine STC-1 cells and human-derived HuTu-80 cells; earlier work used mouse organoids and pig tissues. The human-cell experiment is a simplified model, not a human supplementation trial or proof of clinical glucose, satiety or weight benefits (Food & Function laboratory paper).
Benefits by claim
Blood glucose and insulin
The 2015 Onakpoya and Heneghan review reported nine studies and 756 participants. Some stevioside subgroup results favored lower glucose or blood pressure, but effects were small and highly heterogeneous; longer-term glucose analyses were inconclusive. This does not establish a reliable benefit from retail stevia products (Onakpoya & Heneghan 2015).
Ajami’s double-blind tea trial randomized 39 people with type 2 diabetes and analyzed 34 completers over eight weeks. It detected no between-group difference in fasting glucose, HbA1c, insulin or lipids versus sucralose; a small null result does not prove equivalence or absence of harm (original paper). Anker’s 2019 review of stevioside trials also found no significant pooled fasting-glucose or HbA1c benefit, with variable study quality and heterogeneity (original review). No reliable therapeutic dose is established.
Blood pressure
Across all glycosides, systolic BP was not significantly lower (MD −2.98 mmHg, 95% CI −6.23 to 0.27); the stevioside subgroup differed and favored lower systolic BP. The review reported no specific grant, Onakpoya no interests, and Heneghan teaching/consulting income and book royalties. Most underlying studies omitted funding; Reb A trial authors had manufacturer affiliations or support (Onakpoya & Heneghan 2015).
Gut microbiome
Singh’s exploratory analysis used stored faecal samples from 27 participants in the same 12-week drops trial described below. Overall alpha/beta diversity did not differ; several taxon differences already existed at baseline (Singh 2024). Suez’s separate two-week, unblinded study reported stevia-related functional microbiome changes without significant stevia-related glucose-tolerance impairment; its stevia arm had 20 analyzed participants and used a glucose-carrier product (Cell 2022). Kwok’s double-blind four-week study reported 59 healthy adults and no significant between-group microbiome or faecal SCFA difference versus sucrose. Cargill sponsored it and several authors had Cargill affiliations (original abstract; registry). These finite protocols do not establish lasting clinical harm or a universal absence of effects.
Weight management
Stamataki’s 12-week open-label drops trial randomized 31 healthy normal-weight adults, with 28 completers. Its primary glucose/insulin-response outcomes were null; the weight-maintenance and self-reported energy-intake signals were secondary and the trial was not powered to establish a weight-loss benefit. It reports N8 AgriFood/BBSRC CASE support and no conflicts; the CASE partner and product-supply terms are not specified (full drops paper). SWEET compared three biscuit formulations for two weeks each in 53 completers with overweight/obesity. Appetite did not differ; hormone analyses had 40 participants and glucose/insulin analyses 47. The Reb M formulation lowered postprandial glucose and insulin versus sucrose, but contained more polyols, so it did not isolate Reb M. The trial had Horizon 2020 funding, donated industry ingredients, a Cargill employee author and other disclosed food-industry relationships (SWEET original paper).
Taste and adverse effects in trials
Bitterness can affect palatability, but the cited meta-analysis does not establish aftertaste as the most common clinical complaint or compare Reb M against other glycosides. It reports gastrointestinal symptoms and dizziness; one trial stopped stevioside treatment because of severe nausea, fullness or vomiting (Onakpoya & Heneghan 2015; Reb A laboratory paper).
What works and what does not
| Claim | Verdict | Evidence basis |
|---|---|---|
| Stevioside may lower fasting glucose in some trial settings (Primary source) | Tentative signal; not consistently sustained in longer trials (Primary source) | Nine-study review reporting 756 participants overall (Onakpoya & Heneghan 2015) |
| Stevioside may modestly lower diastolic blood pressure (Primary source) | Tentative BP signal; aggregate and stevioside subgroup results differ (Primary source) | Onakpoya & Heneghan 2015 |
| Reb A lowers blood pressure or improves cardiometabolic risk | Not supported | No significant effect found (Onakpoya & Heneghan 2015) |
| Stevia is glycemically superior to sucralose in type 2 diabetes | Not supported — no significant difference found | PMC7103435 |
| Stevia disrupts gut microbiome balance | Lasting clinical harm unestablished; one small exploratory drops analysis found no overall diversity effect. | PubMed 38257188 |
| Whole-leaf/green stevia powder is as safe as purified extract | Not established — purified-glycoside safety conclusions cannot simply be transferred to whole leaf (Primary source) | FDA Import Alert 45-06 |
| Fermentation-derived Reb M is chemically different from leaf-extracted Reb M | Cargill's dossier reports matching chemical-identity data; manufacturing and specifications still matter (Primary source) | FDA-hosted Cargill GRAS 882 dossier |
Risks and reported side effects
| Side effect | Frequency/context | Source |
|---|---|---|
| Bitter/licorice-like aftertaste | Palatability issue; this laboratory paper does not establish comparative clinical complaint rates (Primary source) | Reb A laboratory paper |
| Mild GI discomfort (abdominal fullness, epigastric pain, bloating) | Reported in trials; some GI symptoms caused treatment discontinuation (Primary source) | Onakpoya & Heneghan 2015 |
| Flatulence and bloating in reformulated biscuits | Reported more often with the Reb M/polyol formulation than sucrose. Different polyol contents prevent attributing this solely to stevia. | SWEET original trial |
| Dizziness | Reported in the review; a reliable product-specific incidence is not established. | Onakpoya & Heneghan 2015 |
| Excess intake beyond ADI, particularly in children | Conservative modeled high-consumer exceedances, not measured adverse-event rates; current JECFA considers those estimates unlikely to pose a health concern. | EFSA 2024; JECFA |
| Unknown long-term safety of whole-leaf/crude extract | Not permitted as a US sweetener; do not extrapolate refined-glycoside safety to this form (Primary source) | FDA Import Alert 45-06 |
Trials generally found similar adverse-event frequencies to placebo, but reported gastrointestinal symptoms and some treatment discontinuation. Small studies cannot exclude rare serious events or establish every product's long-term safety. Safety assessments apply to specified purified glycosides and exposure levels (Onakpoya & Heneghan 2015; FDA sweetener guidance).
Medication considerations
| Drug/substance class | Mechanism of concern | Severity/guidance | Evidence status |
|---|---|---|---|
| Antihypertensive medications | BP effects from some concentrated stevioside trials could theoretically add to a medicine’s effect; ordinary food use was not tested as an interaction. | Discuss concentrated supplements with the prescriber; monitor as advised. | Theoretical extrapolation; these reviews do not establish a dedicated drug-interaction effect (2015; 2019). |
| Antidiabetic medications (insulin, metformin, sulfonylureas) | Some concentrated stevioside trials suggest glucose effects, but pooled findings are inconsistent. They do not establish an interaction at ordinary food intakes. | Discuss concentrated supplements with the prescriber; monitor as advised. | Theoretical extrapolation, not demonstrated in a dedicated interaction trial in these reviews (2015; 2019). |
| Other medicines | The cited trials do not provide a comprehensive medication-interaction profile. Absence of a reported interaction is not proof of compatibility. | Have a pharmacist or clinician review the actual product and medicine list. | Unresolved evidence gap (2015 review; 2019 review). |
Precautions and intake
- Avoid treating whole-leaf or crude ("green") stevia as interchangeable with purified glycosides; these forms are not permitted as sweeteners in the US (FDA sweetener guidance).
- People taking glucose- or blood-pressure-lowering medicines should discuss concentrated stevioside supplements with their prescriber. Tentative biomarker effects do not establish a contraindication to ordinary food-level stevia use or a confirmed interaction (2015 review; 2019 review).
- For children or other frequent consumers, consider total intake across products and distinguish steviol equivalents from the weight of a blend. EFSA’s conservative model exceeded the ADI in some high-consuming infants/toddlers; this was not observed clinical harm, and current JECFA considers those modeled exceedances unlikely to pose a health concern (EFSA 2024; JECFA).
- Pregnant or breastfeeding people should distinguish conventional-food sweetener use from concentrated supplements and discuss the latter with their clinician. The cited 2014 pregnancy review reported no human stevia pregnancy-outcome data at that time; it is not a current exhaustive pregnancy evidence review. General ADI assessments do not establish supplement safety in pregnancy, and whole-leaf/crude stevia is not permitted as a US conventional-food sweetener (2014 pregnancy review; FDA).
Dosage and how to take
| Parameter | Value | Source |
|---|---|---|
| Acceptable Daily Intake (ADI) | 4 mg/kg body weight/day, expressed as steviol equivalents | EFSA 2024; JECFA current database |
| Regulatory basis | Separate EFSA/JECFA assessments agree on the ADI. Product specifications include glycoside composition, process and contaminants, not purity alone. | EFSA 2024; JECFA |
| Population at risk of exceeding ADI | Conservative high-consumer models can exceed the ADI in some infant/toddler groups; JECFA judges those modeled exceedances unlikely to pose a health concern. | EFSA 2024; JECFA |
| Doses studied for glucose/blood-pressure effect (stevioside) | Varied across the included studies; substantial heterogeneity noted by authors (Primary source) | Onakpoya & Heneghan 2015 |
| Whole-leaf/crude form dosing guidance | No conventional-food sweetener intake allowance established here. US whole-leaf/crude sweetener use is not permitted; dietary-ingredient and other technical uses have separate rules. | FDA Import Alert 45-06 |
Preclinical evidence and its limits
Human efficacy conclusions rely on human studies with their funding and design limitations. Regulatory safety assessments also use preclinical evidence, including chronic animal toxicology. Animal or cell findings can inform hazards and mechanisms, but cannot by themselves establish a clinical benefit or harm from a particular human intake (EFSA 2024).
- Mouse toxicology. Farid’s study reported liver/kidney and blood changes after exposure to a commercial stevia product. Its small mouse groups, product characterization and dose translation limit extrapolation; these results do not establish organ damage from ordinary human stevia intake (original mouse study).
- Excluded from human efficacy conclusions: the ACS study tested bacterial isolates and a human gut microbial community in vitro, plus a Cebus apella animal model. “Human microbiome” in its title does not make it a trial in people (original abstract).
The mechanistic section includes human-derived HuTu-80 cell experiments, explicitly separated from murine STC-1 and earlier mouse/pig work. Cell models do not establish clinical benefit. This study reports public and public-private support and no declared conflicts (Food & Function laboratory paper).
Funding and conflicts
Research funding at a glance
9 disclosure entries. The counts below summarize independence tiers explicitly assigned in this article. They count disclosures, not studies, funding amounts or evidence quality.
Consult this article’s source and funding notes for named funders, countries, relationships and exceptions where available. Institutional backing, researcher interests and trial sponsorship are separate questions. Public funding alone does not establish independence; commercial ties alone do not prove a claim false. This overview is not a new financial audit.
| Source | Funding/affiliation | Independence rating |
|---|---|---|
| Onakpoya & Heneghan 2015 | No specific review grant. Heneghan disclosed teaching/consulting income and royalties; Onakpoya no interests. Three included trials reported government/public-institution support, six omitted funding; Reb A trial authors had manufacturer links. | Review funding differs from underlying trial independence (Primary source) |
| Singh 2024: exploratory microbiome analysis of the 2020 drops trial | N8 AgriFood Pump Priming and BBSRC CASE stipend BB/M011208/1; no conflicts declared. CASE partner and product-supply terms are not specified in this report. | Declared support; underlying partner/procurement gaps remain. |
| ACS steviol-glycoside/erythritol study | USDA/academic and Alpha Genesis author affiliations; publisher states no competing financial interest. Full funding details remain unverified. Microbial cultures and primate experiments, not a human trial. (Primary source) | Preclinical; excluded from human efficacy conclusions |
| SWEET consortium crossover trial | Horizon 2020 grant 774293; Cargill employee author, industry ingredient provision and disclosed beverage/sweetener/food-industry relationships. (Primary source) | Public-funded trial with commercial involvement and declared conflicts (Primary source) |
| Suez 2022 stevia arm | Public, charitable and private-donor support; a Strauss Institute fellowship and other support are acknowledged. Authors disclose DayTwo/BiomX founder roles and other commercial relationships. | Full relevant original declarations read; private contracts and product-supply terms remain unverified. |
| Kwok 2024 beverage trial | Registry names Cargill as sponsor and INQUIS as collaborator; several original author affiliations are Cargill. | Original abstract, indexed product-method passages and registry checked; complete main-paper disclosures and private terms were not recovered. |
| EFSA 2024 steviol glycosides opinion | EU food-safety authority; original assessment lists an International Stevia Council application/dossier. | Official regulatory assessment; not certification that all submitted studies are financially independent. |
| JECFA current chemical entry | WHO/FAO joint expert committee | Official WHO/FAO expert assessment; underlying study finances require their own checks. |
| FDA Import Alert 45-06 | US federal regulator | Official enforcement guidance; not clinical-trial funding clearance. |
Frequently asked questions
Is stevia the same thing regardless of which product I buy?
No. Ingredient purity, composition and manufacturing process matter. FDA has not questioned GRAS conclusions for certain high-purity glycosides under specified uses; whole-leaf and crude stevia are not permitted as sweeteners in the US (FDA sweetener guidance).
Does stevia lower blood sugar?
Some stevioside trials suggest lower glucose, but heterogeneity and longer-term null analyses limit confidence. Reb A studies did not establish a significant benefit. A small comparison with sucralose also found no glycemic advantage; none of this establishes stevia as a diabetes treatment (Onakpoya & Heneghan 2015; PMC7103435).
Does stevia harm gut bacteria?
One small 12-week study found no overall diversity effect but some exploratory taxon/classification differences. The separate four-week Reb M/D beverage trial was Cargill-sponsored. These protocols do not establish lasting clinical harm or prove every stevia product leaves gut bacteria unchanged (Singh 2024; Kwok 2024; trial registry).
What is the difference between stevioside, Reb A, Reb M, and Reb D?
Stevioside, Reb A, Reb M and Reb D are steviol glycosides with differing abundance and sensory properties. The cited human BP/glucose reviews mainly concern stevioside and Reb A. Reb M/D ingredients can come from leaf extraction, enzyme conversion or a specified fermentation process; Cargill’s dossier describes one example, not a market-wide trend (human review; Cargill dossier; enzyme-conversion dossier). FDA’s no-questions responses concern specified ingredients and uses; purity alone does not establish every product’s status (FDA).
How much stevia is safe to consume daily?
EFSA and JECFA separately use 4 mg/kg body weight/day as steviol equivalents for specified glycosides. This is not 4 mg of every retail blend. Conservative high-consumer models can exceed the ADI in some infants/toddlers; JECFA considers those estimates unlikely to pose a health concern. The ADI is a long-term intake benchmark, not a target dose or a threshold at which harm inevitably begins (EFSA 2024; current JECFA entry).
Where does commercial stevia actually come from?
The plant is native to parts of Brazil and Paraguay (Kew). A 2019 Dialogue Earth report described China as a major producer and documented a plan to process Paraguayan leaves there; it does not establish today’s global shares. The 2015 Public Eye advocacy report documented traditional Guaraní use and argued that commercialization failed to share benefits fairly. Those dated reports should not be treated as current supply-chain audits.
Sources and funding notes
- Onakpoya & Heneghan 2015, nine-study review of steviol glycosides: no specific review grant; trial funding and manufacturer ties require separate assessment.
- Ajami 2020 tea comparison in type 2 diabetes. NNFTRI support; Parmida Mehr Pasargad manufactured the stevia extract; no conflicts declared. Supply/payment terms are not reported.
- Anker 2019 stevioside biomarker meta-analysis. No external review funding and no conflicts declared; those statements do not clear financing or procurement for every included trial.
- Singh 2024 exploratory microbiome analysis of stored samples from the same 2020 drops trial. N8/BBSRC CASE support; partner and supply terms unreported here.
- News-Medical January 2024 coverage of Singh’s analysis; secondary reporting, not another trial or the basis for clinical conclusions.
- Kwok 2024 four-week Reb M/D versus sucrose trial; NCT05264636 names Cargill sponsor and INQUIS collaborator. Original abstract and indexed product/method sections checked; complete main-paper delivery remained limited.
- Impact of Steviol Glycosides and Erythritol on the Human and Cebus apella Gut Microbiome — laboratory microbial-community and primate experiments, not a human supplementation trial.
- Stamataki 2020 full drops trial; earlier conference abstract reports the same study, not a replication. N8 AgriFood/BBSRC CASE support and no conflicts declared; commercial partner and supply terms not specified here.
- SWEET biscuit crossover trial (eBioMedicine 2024): Horizon 2020 funding, industry ingredient provision and declared commercial relationships.
- FDA Import Alert 45-06, stevia leaf and crude extract.
- GRAS 702 manufacturer dossier — Xinghua GL Stevia leaf-derived purified glycosides. FDA hosting is not independent authorship or a universal product approval.
- FDA response to GRAS 790 — GLG Life Tech’s specified purified-glycoside jerky uses, up to 2,500 mg/kg. The no-questions letter is not an affirmative GRAS determination for every use.
- GRAS 838 manufacturer dossier — Jiang Su Svetia’s SweetUni enzyme-converted high-purity glycosides; commercial filing, not an independent trial.
- Cargill GRAS 882 dossier — specified fermentation-derived glycosides, predominantly Reb M. Commercial notifier evidence; private expert-panel terms not independently verified.
- EFSA 2024 original opinion retained the ADI after reviewing the International Stevia Council application. The 2010 press release is historical; current claims use the newer opinion.
- JECFA current steviol glycosides entry — WHO/FAO assessment and current intake interpretation. Official source role is distinct from underlying-trial financial independence.
- 2014 pregnancy review — relevant stevia section checked; no conflicts declared. Its historical absence of human pregnancy data is not a current exhaustive evidence census.
- Food & Function (2023), Reb A GLP-1 laboratory study using murine STC-1 and human-derived HuTu-80 cells; public/public-private support, no declared conflicts. Not a clinical trial.
- Farid 2020 mouse toxicology — relevant original methods, organ/blood results and disclosures checked. No specific grant and no competing interests declared; product-supply terms not reported. Not proof of ordinary human-intake harm.
- Kew Plants of the World Online — botanical identity and native distribution; replaces Wikipedia as the origin citation.
- Public Eye and co-publishers’ 2015 report — original advocacy report on traditional knowledge and benefit sharing; publisher identities checked, report-specific private funding not fully established.
- Dialogue Earth 2019 stevia report — historical reporting on Paraguay, China and trade; does not establish today’s global supply shares.
Last reviewed: July 4, 2026.
Have a question — or want us to cover something?
Ask about anything on this page, or request the next deep dive: an ingredient, a supplement, or a health concern. We use published research, evidence syntheses, and regulatory guidance, with clear source links.
One daily research roundup
Get the topics, key findings and links from our new articles in one email. At most one digest a day, only when there is something new.