Buckwheat: the lipids move, the blood sugar argues

Pooling 13 randomised human trials, buckwheat lowered blood glucose by 0.85 mmol/L (95% CI -1.31 to -0.39) and total cholesterol by 0.50 mmol/L (95% CI -0.80 to -0.20). Triglycerides fell by 0.25 mmol/L (95% CI -0.49 to -0.02), the reviewers report high unexplained heterogeneity in every one of those analyses, and nine of the human studies were done in China, in a review that also covered 21 animal studies. The largest single trial disagrees about the glucose. In 165 patients with type 2 diabetes who replaced part of their staple food with tartary buckwheat for four weeks, total cholesterol went from 5.08 to 4.79 mmol/L and LDL from 3.00 to 2.80 mmol/L against the control diet. Blood glucose and glycated haemoglobin did not differ at all. That trial was open label, ran four weeks, and found insulin resistance falling only in the subgroup eating more than 110 g a day. So the lipid effect repeats and the glucose effect does not. The pooled estimate rests on short trials whose heterogeneity the reviewers could not account for, the biggest trial in patients found nothing on glucose, and we found no trial long enough to settle which reading is right.

The numbers, per 100 g

How much is actually in it

• Across eight sampled lots of whole grain buckwheat USDA puts magnesium between 186 and 225 mg per 100 g, but molybdenum between 31.6 and 101 mcg, a spread of 126 percent in the same grain. — USDA Foundation Foods record Buckwheat, whole grain, FDC ID 2512378, eight sampled lots, release of 2026-04-30 (n = 8 samples) Evidence E source
  Magnesium 203 mg (186 to 225, spread 19%), copper 0.4764 mg (0.404 to 0.63, 47%), calcium 13.64 mg (11.3 to 16.7, 40%), total fibre 4.046 g (3.24 to 5.08, 45%), iron 2.444 mg (2.08 to 2.8, 29%), molybdenum 55.25 mcg (31.6 to 101, 126%)
• USDA gives whole grain buckwheat 203 mg of magnesium per 100 g and the UK table gives buckwheat groats 48 mg, a fourfold gap that is far too wide for sampling and most likely means the two rows are not the same food. — USDA Foundation record Buckwheat, whole grain (FDC ID 2512378) against the CoFID 2021 record Buckwheat, groats (code 11-006); the two rows were paired automatically by name and nobody has confirmed the pairing (n = 2 records) Evidence E source
  Magnesium USDA 203 against CoFID 48 mg (76% apart), copper 0.476 against 0.7 mg (32%), thiamin 0.539 against 0.28 mg (48%)

What a real portion looks like

• Only 0.04 percent of US adults reported any buckwheat on either of two recall days, so rare that the study could not examine its health associations at all. — US adults in NHANES 2013 to 2018 who reported barley, buckwheat or quinoa in at least one of two 24-hour dietary recalls (n = NHANES 2013-2018 adult sample) Evidence C source
  Prevalence of consumption 0.36% for barley, 0.04% for buckwheat and 1.07% for quinoa; buckwheat had to be dropped from the association analysis for lack of consumers
• A cup of cooked buckwheat groats weighs 170 g in the US survey, while a cup of dry groats yields 500 g cooked, so the same cup means three different amounts depending on when it is measured. — USDA FNDDS survey food Buckwheat groats, WWEIA category Pasta, noodles, cooked grains, FNDDS code 56200510, four household measures (n = 4 household measures) Evidence E source
  1 cup cooked 170 g, 1 cup dry yields 500 g, 1 oz dry yields 85 g, quantity not specified 170 g

What your body absorbs

• Buckwheat crackers shifted three gut satiety hormones against rice crackers in healthy adults and in people with type 2 diabetes, yet post-prandial glucose, insulin and C-peptide did not change, and a week of daily eating moved no fasting marker. — healthy participants and people with diet-controlled type 2 diabetes, single-blind crossover, buckwheat or rice crackers with blood at 15, 30, 45, 60, 120 and 180 minutes, then one serving of buckwheat crackers daily for a week (n = not reported in the abstract, and the article has no PMC full text, so it cannot be checked) Evidence B source
  Post-prandial GLP-1, glucose-dependent insulinotropic peptide and pancreatic polypeptide altered after buckwheat versus rice; no change in post-prandial glucose, insulin or C-peptide; no effect of one week on fasting glucose, lipids or apolipoproteins

What it does to your health

• Pooling 13 randomised human trials, buckwheat lowered blood glucose by 0.85 mmol/L (95% CI -1.31 to -0.39) and total cholesterol by 0.50 mmol/L (95% CI -0.80 to -0.20), with high unexplained heterogeneity in every analysis. — 13 randomised controlled human studies and 2 cross-sectional human studies in free-living subjects, reviewed alongside 21 animal studies; nine of the human studies were done in China (n = 15 human studies of which 13 randomised, plus 21 animal studies) Evidence A source
  Blood glucose -0.85 mmol/L (95% CI -1.31 to -0.39), total cholesterol -0.50 mmol/L (95% CI -0.80 to -0.20), triglycerides -0.25 mmol/L (95% CI -0.49 to -0.02); HDL inconsistent; weight, blood pressure, insulin and LDL left unclear
• Replacing all the meat with buckwheat for one week in 20 healthy adults lowered plasma homocysteine (p less than 0.01) and left them exactly as full as their usual diet (p = 0.234). — 20 healthy volunteers, 6 men and 14 women, mean age 42.1 years and BMI 24.7, randomised crossover with two provided seven-day diets and at least two weeks of washout between them (n = 20) Evidence B source
  Plasma homocysteine down (p < 0.01) on the buckwheat diet with salicylic acid and 2,3-dihydroxybenzoic acid up in plasma, urine and faeces; satiety no different from the habitual diet (p = 0.234); plasma urea and faecal deoxycholic acid down on both diets

Safety, limits and interactions

• A 57-year-old man lost consciousness twice in the same day after eating buckwheat noodles, with buckwheat specific IgE of 81.5 kU/L, which is what biphasic anaphylaxis looks like. — single case reported alongside a systematic review, 57-year-old male patient, two episodes of loss of consciousness within one day of eating buckwheat noodles (n = 1) Evidence D source
  Buckwheat specific IgE 81.5 kU/L and Dermatophagoides pteronyssinus IgE 42.4 kU/L, both unusually high; the second episode came within a day of the exposure
• A health food shop worker was sensitised by inhaling buckwheat flour, first noticing itchy eyes and sneezing while handling it, and then reacting with anaphylaxis after eating a slice of buckwheat bread. — single case, a health food store employee in Europe, positive skin prick test with buckwheat flour and raised buckwheat specific IgE (n = 1) Evidence D source
  Airborne exposure produced rhinoconjunctival itching and sneezing, ingestion produced anaphylaxis with urticaria, angioedema and dyspnoea

What people believe that the data does not support

• Among 640 naturally gluten-free flours bought across Canada, buckwheat averaged 153 mg/kg of gluten, and 9.5 percent of all samples sat above the 20 mg/kg limit for gluten-free products. — 640 samples of naturally gluten-free flours and starches bought in eight Canadian cities and over the internet between 2010 and 2012 and analysed for gluten (n = 640 samples) Evidence C source
  61 of 640 samples (9.5%) above 20 mg/kg with a range of 5 to 7995 mg/kg; composite means soy 902, millet 272, buckwheat 153 mg/kg; among products actually labelled gluten-free only 3 of 268 exceeded the limit, range 5 to 141 mg/kg
• The EU register holds exactly one health claim naming buckwheat, for a flavonoid and zinc extract sold as Coldizin, and its status is non-authorised. — EU Register on nutrition and health claims, snapshot of 2026-09-21, 2330 claims of which 262 authorised and 2067 rejected (n = 1 claim) Evidence E source
  Claim POL-HC-6831 for buckwheat extract containing a flavonoid-mineral troxerutin and zinc complex, status Non-authorised

Who this works differently for

• In 165 patients with type 2 diabetes, swapping part of the staple food for tartary buckwheat for 4 weeks moved total cholesterol from 5.08 to 4.79 mmol/L and LDL from 3.00 to 2.80 mmol/L against the control diet. Blood glucose and HbA1c did not move at all. — 165 patients with type 2 diabetes in a parallel randomised open-label controlled trial, daily replacement of a portion of the staple food with tartary buckwheat food for 4 weeks against systematic diet plans with intensive nutritional education (n = 165) Evidence B source
  Fasting insulin 2.46 to 2.39 Ln mU/L, total cholesterol 5.08 to 4.79 mmol/L, LDL 3.00 to 2.80 mmol/L versus control (P < .05); no difference in blood glucose or glycated haemoglobin; insulin resistance fell only in the subgroup eating more than 110 g a day

What is still unknown

• Buckwheat carries phototoxic fagopyrins and no reliable quantitative toxicity figure for them exists. Seeds, flour and teas are safe in normal amounts, while diets built on sprouts, herbs, flowers or fagopyrin-rich extracts may cause fagopyrism. — systematic review of analytical and phototoxicity data on fagopyrins and protofagopyrins in buckwheat seeds, flour, tea, sprouts, herb and flowers; no human trial Evidence D source
  No reliable quantitative data on fagopyrin toxicity and no reference standard; the risk attaches to plant part and to extracts rather than to the grain

The bottom line

Two smaller trials fill in what buckwheat does not do. Replacing all the meat in the diet with buckwheat for one week in 20 healthy adults lowered plasma homocysteine (p less than 0.01). It left people exactly as full as their habitual diet (p = 0.234), with every meal provided across seven days and biomarkers as the outcome. Fasting glucose and insulin fell in the fava bean arm of the same crossover, while the buckwheat arm left them alone. Buckwheat crackers shifted GLP-1, glucose-dependent insulinotropic peptide and pancreatic polypeptide against rice crackers, in healthy adults and in people with diet-controlled type 2 diabetes. Post-prandial glucose, insulin and C-peptide did not change, and a week of eating one serving a day moved no fasting marker. That trial never states how many people took part and has no full text to check, so its size is unverifiable. The numbers on a buckwheat table are softer than they look. USDA sampled eight lots of whole grain buckwheat and published the spread. Magnesium came out at 203 mg per 100 g across a range of 186 to 225, and molybdenum at 55.25 mcg across a range of 31.6 to 101, a spread of 126 percent in the same grain. Those eight lots mix farming variation with analytical variation. The UK table gives buckwheat groats 48 mg of magnesium against USDA's 203 mg, and a fourfold gap is far too wide for sampling, so the likeliest reading is that the two rows describe different foods. Those rows were paired on the word buckwheat and nobody has confirmed the pairing, which leaves this grain without an independent second figure. Portions go wrong at home. The US survey puts a cooked cup of groats at 170 g, while a dry cup yields 500 g cooked, so a cup measured in the packet and a cup measured on the plate are different amounts. Both are recall conventions rather than weighed servings. Hardly anyone in America eats it. Some 0.04 percent of adults reported any buckwheat on either of two recall days in NHANES 2013 to 2018, in a survey design that undercounts an occasional food. The researchers dropped buckwheat from their health analysis because there were too few consumers to analyse. Two beliefs need correcting. Buckwheat is not wheat, and that alone does not make a bag of it gluten free. Among 640 naturally gluten-free flours bought across Canada between 2010 and 2012, buckwheat averaged 153 mg/kg of gluten as a composite mean, and 9.5 percent of all 640 samples sat above the 20 mg/kg limit. Flours that actually carried a gluten-free label exceeded it in 3 of 268. The EU register holds exactly one health claim naming buckwheat, for a branded flavonoid and zinc extract, and the regulator refused it. A refusal says that this wording was not permitted rather than that the extract was tested and failed. Three different harms are on record. Buckwheat carries phototoxic fagopyrins, and a systematic review found no reliable quantitative data on their toxicity and no reference standard to measure them against, so no threshold dose exists. That review puts seeds, flour and teas in normal amounts on the safe side, and attaches fagopyrism to diets built largely on sprouts, herbs, flowers or fagopyrin-rich extracts, and fagopyrism is the phototoxicity that review is about. Allergy is the sharper risk. A 57-year-old man lost consciousness twice in one day after eating buckwheat noodles, with buckwheat specific IgE of 81.5 kU/L. In that one case the useful part is the timing, because an apparent recovery relapsed within the day. A health food shop worker was sensitised by inhaling buckwheat flour, noticing itchy eyes and sneezing while handling it. That worker then had anaphylaxis with urticaria, angioedema and dyspnoea after eating a slice of buckwheat bread. Both are single patients, so they say nothing about how common this is. Both allergy reports here are adults. Nothing in this material speaks to children, to pregnancy or to medicines.

Sources

1. bkwt-r3-11 · food composition dataset
   Buckwheat, whole grain, Molybdenum, Mo 55.25 UG, range 31.6-101.0, median 50.9, 8 samples, spread 126 percent
   USDA FoodData Central, Foundation Foods, release 2026-04-30, FDC ID 2512378 · checked 2026-09-24
2. bkwt-r3-12 · food composition dataset comparison
   USDA Buckwheat, whole grain Magnesium 203 mg per 100 g against CoFID 2021 Buckwheat, groats code 11-006 Magnesium 48 mg per 100 g
   USDA FoodData Central FDC 2512378 and CoFID 2021, Public Health England, OGL v3.0, code 11-006 · checked 2026-09-24
3. bkwt-r3-09 · cross-sectional national survey analysis
   The prevalence of barley, buckwheat and quinoa consumption among US adults were 0.36%, 0.04% and 1.07%, respectively...Due to the very low prevalence of buckwheat consumption, the associations of consumption with diet and health markers were examined only for barley and quinoa
   Nutrients, 2025 · checked 2026-09-24
4. bkwt-r3-13 · food survey portion dataset
   Buckwheat groats, 1 cup cooked 170 g, 1 cup dry yields 500 g, 1 oz dry yields 85 g, FNDDS code 56200510, fdc_id 2708362
   USDA FNDDS, Buckwheat groats, fdc_id 2708362 · checked 2026-09-24
5. bkwt-r3-04 · randomised crossover trial
   Although the buckwheat cracker did not modify acute glycemia or insulinemia, it was sufficient to modulate gastrointestinal satiety hormones.
   Metabolism, 2013 · checked 2026-09-24
6. bkwt-r3-01 · meta-analysis of randomised controlled human trials
   Using random-effects models, the weighted mean difference of post-intervention concentrations of blood glucose, total cholesterol and triglycerides were significantly decreased following buckwheat intervention compared with controls [differences in blood glucose: -0.85 mmol/L (95% CI: -1.31, -0.39), total cholesterol: 0.50 mmol/L (95% CI: -0.80, -0.20) and triglycerides: 0.25 mmol/L (95% CI: -0.49, -0.02)]...All exhibited high unexplained heterogeneity
   Nutrients, 2018 · checked 2026-09-24
7. bkwt-r3-03 · randomised crossover feeding trial
   Both intervention diets were as satiating as the volunteers' habitual diets (p = 0.234)...Buckwheat diet decreased plasma homocysteine (p < 0.01) and increased plasma, urine and faecal concentrations of salicylic acid and 2,3-dihydroxybenzoic acid
   Eur J Nutr, 2025 · checked 2026-09-24
8. bkwt-r3-06 · case report
   Here, we report a case of biphasic buckwheat anaphylaxis in a 57 year-old male patient who lost consciousness twice in the same day after having buckwheat noodles. The serum test shows that Dermatophagoides pteronyssinus (Dp)immunoglobulin E (IgE) (42.4 kU/L) and buckwheat-specific IgE (81.5 kU/L) are unusually high.
   Asian Pac J Allergy Immunol, 2018 · checked 2026-09-24
9. bkwt-r3-07 · case report
   A health food store employee experienced anaphylaxis with urticaria, angioedema, and dyspnea after eating a slice of buckwheat bread. Prior to this reaction, the patient had repeatedly noticed rhinoconjunctival itching and sneezing when handling buckwheat.
   Allergol Select, 2025 · checked 2026-09-24
10. bkwt-r3-08 · cross-sectional survey of retail products
    The results showed that 61 of the 640 (9.5%) samples were contaminated above the Codex-recommended maximum level for gluten-free products (20 mg kg⁻¹) with a range of 5-7995 mg kg⁻¹.
    Food Addit Contam Part A Chem Anal Control Expo Risk Assess, 2013 · checked 2026-09-24
11. bkwt-r3-10 · register of health claims, European Commission
    POL-HC-6831 Buckwheat extract containing flavonoid-mineral (troxerutin - zinc) complex (Coldizin) Coldizin helps support the immune system Non-authorised
    EU Register on nutrition and health claims, claim POL-HC-6831, snapshot 2026-09-21 · checked 2026-09-24
12. bkwt-r3-02 · randomised open-label controlled trial
    The TB group decreased fasting insulin (2.46-2.39 Ln mU/L), total cholesterol (5.08-4.79 mmol/L), and low-density lipoprotein cholesterol (3.00-2.80 mmol/L) compared with the DC group at 4 weeks (P<.05). No significant differences in blood glucose or glycated hemoglobin levels were noted between the TB and DC groups.
    Nutr Res, 2016 · checked 2026-09-24
13. bkwt-r3-05 · systematic review of analytical and in vitro phototoxicity data
    A systematic review of fagopyrins and the phototoxicity of buckwheat found that reliable quantitative data on fagopyrin toxicity are not yet available. Generally, buckwheat seeds, flour, and teas are safe in normal amounts. Diets extensively composed of buckwheat sprouts, herbs, and particularly flowers or of fagopyrin-rich buckwheat extracts may cause fagopyrism.
    J Agric Food Chem, 2015 · checked 2026-09-24