Garlic: a real effect, a small one, and mostly from supplements

Garlic carries the reputation of a drug you can buy in the produce aisle. The strongest evidence says it does move the numbers, by a little. Pooling 108 randomised trials in 7137 adults, mostly small and short, on supplements of varied form and dose, garlic lowered total cholesterol by 10.21 mg/dL (95% CI 6.74 to 13.69 lower) and systolic pressure by 3.71 mmHg (95% CI 2.36 to 5.07 lower). The gains were largest in adults who started with unfavourable numbers, and the authors of that pooling ask for larger and longer trials. What those trials tested was mostly supplements. When 192 adults with LDL cholesterol between 130 and 190 mg/dL ate the equivalent of one average clove of raw garlic six days a week for six months, their LDL finished 0.4 mg/dL higher than it started. A powdered supplement finished 3.2 higher, an aged extract 0.2 higher, and placebo 3.9 lower. That trial ran at one dose in one population and measured lipids only, so it says nothing about blood pressure.

The numbers, per 100 g

What your body absorbs

• Allicin availability from enteric coated garlic tablets ranged from 36 to 104% of raw garlic, and fell to 22 to 57% when the tablet was taken with a high-protein meal. — 13 subjects, bioassay of 13 garlic supplements and 9 garlic foods against homogenised raw garlic as the 100% reference (n = 13) Evidence B source
  Enteric tablets 36 to 104% and 22 to 57% with a high-protein meal, non-enteric tablets 80 to 111%, garlic powder capsules 26 to 109%, Kwai powder tablets 80%

What cooking and storage change

• Cooking kills the enzyme that makes allicin, yet boiled garlic still delivered 16% of the allicin equivalent of raw, roasted 30%, pickled 19% and black garlic only 5%. — 9 garlic foods measured in the same 13-subject bioassay against homogenised raw garlic (n = 13) Evidence B source
  Boiled 16%, roasted 30%, pickled 19%, acid-minced 66%, black garlic 5% of raw garlic allicin bioequivalence

What it does to your health

• Pooling 108 randomised trials in 7137 adults, garlic lowered total cholesterol by about 10 mg/dL and systolic blood pressure by about 3.7 mmHg, with the largest gains in people who started with poor numbers. — 7137 adults across 108 randomised controlled trials, searched to June 2024 (n = 7137) Evidence A source
  Total cholesterol -10.21 mg/dL (95% CI -13.69 to -6.74), LDL -5.90 mg/dL (95% CI -10.51 to -1.29), systolic blood pressure -3.71 mmHg (95% CI -5.07 to -2.36), fasting glucose -2.77 mg/dL, with no change in body mass index, weight, apolipoprotein A-1 or B
• Across 27 randomised trials in 1619 adults, garlic supplements cut fasting glucose by 2.76 mg/dL and HbA1c by 0.53 percentage points, and did nothing measurable to insulin or HOMA-IR. Our record of those trials does not say whether the adults in them had diabetes, so the numbers cannot be read as what garlic does to a healthy person. — 1619 adults in 27 randomised controlled trials, search to January 2026; neither the abstract nor our record states whether the participants had diabetes or another metabolic condition (n = 1619) Evidence A source
  Fasting blood glucose WMD -2.76 mg/dL (95% CI -5.11 to -0.41), HbA1c WMD -0.53% (95% CI -0.86 to -0.20), insulin and HOMA-IR not significant, larger reductions at doses of 0.5 g per day or less
• Across 19 trials, aged garlic lowered systolic blood pressure by 2.49 mmHg and LDL by 4.41 mg/dL, while diastolic pressure and HDL did not move, and the significant subgroup effects appeared only in people with cardiovascular disease or high cholesterol. — 19 clinical trials of aged garlic in adults, search to May 2024 (n = 19 trials) Evidence A source
  Systolic blood pressure WMD -2.49 mmHg (95% CI -4.02 to -0.95), LDL WMD -4.41 mg/dL (95% CI -8.28 to -0.54), diastolic pressure and HDL not significant, total cholesterol borderline

Safety, limits and interactions

• In 12 healthy men given a single 25 mg dose of warfarin after two weeks of garlic, garlic changed neither how the body handled warfarin nor how strongly it thinned the blood, while cranberry in the same trial raised the INR curve by 30%. — 12 healthy male volunteers of known CYP2C9 and VKORC1 genotype, open-label three-treatment crossover, two weeks of pretreatment before a single 25 mg warfarin dose (n = 12) Evidence B source
  No significant change in warfarin pharmacokinetics or pharmacodynamics with garlic, against a 30% increase in the area under the INR-time curve with cranberry
• Wild garlic has a lookalike that can put a forager in intensive care, and one 58-year-old man ate Colchicum autumnale by mistake and left hospital after 7 days with liver and kidney damage. — one 58-year-old man, accidental poisoning (n = 1) Evidence D source
  Moderate renal and liver impairment, treated with intravenous lipid emulsion, vitamin K and N-acetylcysteine, discharged after 7 days

What people believe that the data does not support

• In 192 adults with high LDL cholesterol, a clove of raw garlic a day for 6 months moved LDL by 0.4 mg/dL, and two popular supplements did no better than placebo. — 192 adults with LDL cholesterol 130 to 190 mg/dL, four arms, dose equal to an average clove, 6 days a week for 6 months (n = 192) Evidence B source
  6-month change in LDL +0.4 mg/dL raw garlic, +3.2 powdered supplement, +0.2 aged extract, -3.9 placebo, with no significant effect on HDL, triglycerides or the cholesterol ratio

Who this works differently for

• In 12 healthy volunteers aged 60 to 76, taking garlic oil for 28 days held back the drug-processing enzyme CYP2E1 by about 22%, and left CYP1A2 unchanged. — 12 healthy volunteers aged 60 to 76 years, mean age 67, each botanical supplement taken for 28 days with a 30-day washout between them (n = 12) Evidence B source
  CYP2E1 activity inhibited by about 22% with garlic oil, no effect on CYP1A2, and by comparison St John's wort induced CYP3A4 by about 140%
• The US National Center for Complementary and Integrative Health draws the line at the plate: garlic in food amounts is not flagged in pregnancy, while supplement doses larger than that may not be safe in pregnancy or breastfeeding. — pregnant and breastfeeding women, agency guidance rather than a study population Evidence E source
  No number given by the agency, the statement separates food amounts from larger oral doses
• Among 102 people living with parosmia, garlic was named a trigger by 62.9%, second only to onion at 68%. — 102 people with current parosmia, online survey (n = 102) Evidence C source
  Onion 68%, garlic 62.9%, coffee 49.5%, beef 48.5%, with sewage and burnt as the most reported perceptions

What is still unknown

• What we hold on garlic and chronic liver disease is 4 trials, 212 patients and an abstract with no effect sizes in it, and that is the whole basis of the meta-analysis that calls garlic effective for the condition. — 212 patients with chronic liver disease across 4 randomised controlled trials, search to September 2021 (n = 212) Evidence A source
  Reported reductions in ALT, AST, total cholesterol, LDL and weight, with no change in adverse events; the abstract gives no effect sizes or confidence intervals

The bottom line

Cook it however you like. Measured in the same 13-person bioassay against raw garlic, boiled garlic delivered 16 percent of the allicin equivalent, roasted 30 percent, pickled 19 percent, acid-minced 66 percent and black garlic 5 percent, and the authors call those values higher than they expected. Enteric coated tablets in that study ranged from 36 to 104 percent of raw garlic, and dropped to 22 to 57 percent when swallowed with a high-protein meal, in 13 subjects measured through a breath metabolite standing in for allicin. Aged garlic has its own pooling of 19 trials: systolic pressure 2.49 mmHg lower (95% CI 0.95 to 4.02) and LDL 4.41 mg/dL lower (95% CI 0.54 to 8.28). Diastolic pressure and HDL did not move, and the significant subgroup results showed up only in people who already had cardiovascular disease or high cholesterol. Across 27 trials in 1619 adults, garlic supplements cut fasting glucose by 2.76 mg/dL (95% CI 0.41 to 5.11 lower) and HbA1c by 0.53 percentage points (95% CI 0.20 to 0.86 lower), and moved insulin and HOMA-IR not at all. Our record of those trials does not say whether the participants had diabetes, so that figure cannot be read as what garlic does to a healthy person. For chronic liver disease we hold four trials, 212 patients and an abstract with no effect sizes in it, and that is the whole basis on which a meta-analysis calls garlic effective there. The cautions here are narrow and specific. The US National Center for Complementary and Integrative Health draws its line at the plate: garlic in food amounts is not flagged in pregnancy, while oral doses larger than food amounts may not be safe in pregnancy or breastfeeding, and the agency attaches no number to that. In 12 healthy volunteers aged 60 to 76, garlic oil taken for 28 days held back the drug-processing enzyme CYP2E1 by about 22 percent and left CYP1A2 alone. That is a shift in drug handling read off metabolic ratios, and not a harm anyone recorded. In 12 healthy men, two weeks of garlic before a single 25 mg dose of warfarin changed neither how the body handled the drug nor how strongly it thinned the blood. Twelve healthy men on one dose of warfarin cannot rule out trouble in someone anticoagulated for years. Foraging carries a risk the supermarket does not. One 58-year-old man mistook Colchicum autumnale for wild garlic and left hospital after seven days with liver and kidney damage. If you live with parosmia, garlic matters for another reason: among 102 people with the condition in an online survey, 62.9 percent named it a trigger, behind onion at 68 percent, which records what people notice rather than any measured dose.

Sources

1. gar-r3-04 · randomised controlled trial
   For enteric tablets, ABB varied from 36⁻104%, but it was reduced to 22⁻57% when consumed with a high-protein meal, due to slower gastric emptying.
   Nutrients, 2018 · checked 2026-09-23
2. gar-r3-05 · randomised controlled trial
   Cooked or acidified garlic foods, which have no alliinase activity, gave higher ABB than expected: boiled (16%), roasted (30%), pickled (19%), and acid-minced (66%). Black garlic gave 5%.
   Nutrients, 2018 · checked 2026-09-23
3. gar-r3-01 · systematic review and meta-analysis
   Ultimately, 108 trials involving 7137 participants were incorporated into this meta-analysis...This meta-analysis demonstrates that garlic intake significantly improves key cardiovascular risk markers, including lipid profiles, glycemic indices, blood pressure, oxidative stress, and inflammatory biomarkers, particularly in adults with unfavorable baseline risk factors.
   Nutr Rev, 2026 · checked 2026-09-23
4. gar-r3-02 · systematic review and meta-analysis
   Twenty-seven RCTs (n = 1,619 participants) were included. Garlic supplementation significantly reduced FBG (WMD = -2.76 mg/dL; 95% CI: -5.11, -0.41) and HbA1c (WMD = -0.53%; 95% CI: -0.86, -0.20), with no significant effects on insulin (WMD = -1.18 μIU/mL; 95% CI: -2.91, 0.55) or HOMA-IR (WMD = -0.30; 95% CI: -0.72, 0.12).
   Diabetes Res Clin Pract, 2026 · checked 2026-09-23
5. gar-r3-06 · systematic review and meta-analysis
   Nineteen trials were included in this analysis. AG consumption significantly reduced SBP (WMD: -2.49 mmHg; 95% CI: -4.02 to -0.95; I2 = 29.76%) and LDL (WMD: -4.41 mg/dL; 95% CI: -8.28 to -0.54; I2 = 55.78%).
   Phytother Res, 2025 · checked 2026-09-23
6. gar-r2-10 · randomised controlled trial
   An open-label, three-treatment, randomized crossover clinical trial was undertaken and involved 12 healthy male subjects of known CYP2C9 and VKORC1 genotype...Co-administration of garlic did not significantly alter warfarin pharmacokinetics or pharmacodynamics.
   Br J Pharmacol, 2008 · checked 2026-09-24
7. gar-r3-08 · case report
   We report the case of a 58-year-old man who experienced moderate renal and liver impairment after accidental poisoning with Colchicum autumnale, which he confused with wild garlic (Allium ursinum).
   Toxics, 2026 · checked 2026-09-23
8. gar-r3-03 · randomised controlled trial
   There were no statistically significant effects of the 3 forms of garlic on LDL-C concentrations...None of the forms of garlic used in this study, including raw garlic, when given at an approximate dose of a 4-g clove per day, 6 d/wk for 6 months, had statistically or clinically significant effects on LDL-C or other plasma lipid concentrations in adults with moderate hypercholesterolemia.
   Arch Intern Med, 2007 · checked 2026-09-23
9. gar-r2-11 · randomised controlled trial
   Twelve healthy volunteers between the ages of 60 and 76 years (mean age 67 years) were randomly assigned to receive each botanical supplement for 28 days followed by a 30-day washout period...Garlic oil inhibited CYP2E1 activity by approximately 22%.
   Drugs Aging, 2005 · checked 2026-09-24
10. gar-r2-12 · government agency fact sheet
    Garlic may not be safe for use during pregnancy or while breastfeeding when taken orally in amounts greater than those found in foods.
    NCCIH, National Institutes of Health, Garlic · checked 2026-09-24
11. gar-r3-09 · cross-sectional survey
    The most frequently reported triggers were onion (68%), garlic (62.9%), coffee (49.5%), and beef (48.5%).
    Eur Arch Otorhinolaryngol, 2026 · checked 2026-09-23
12. gar-r3-07 · systematic review and meta-analysis
    Four RCTs and 212 patients are included in the meta-analysis...Garlic supplementation is effective to treat chronic liver disease.
    Afr Health Sci, 2023 · checked 2026-09-23