Abstract
The objective of this study was to evaluate the scientific evidence on garlic (Allium sativum), including expert opinion, folkloric precedent, history, pharmacology, kinetics/dynamics, interactions, adverse effects, toxicology, and dosing. This review serves as a clinical support tool. Electronic searches were conducted in 10 databases, 20 additional journals (not indexed in common databases), and bibliographies from 50 selected secondary references. No restrictions were placed on the language or quality of the publications. All literature collected pertained to efficacy in humans, dosing, precautions, adverse effects, use in pregnancy and lactation, interactions, alteration of laboratory assays, and mechanisms of action. Standardized inclusion and exclusion criteria were used for selection. Grades were assigned using an evidence-based grading rationale. The present study reports an abridged version, based on the complete review (published at www.NaturalStandard.com), discussing the application of garlic for the mitigation of hyperlipidemia. Evaluation of the literature has revealed a number of controlled trials examining the effects of oral garlic on serum lipids. Although most studies have been small (<100 subjects), with poorly described design and results, and have reported nonsignificant modest benefits, several overlapping meta-analyses pooling available data suggest that non-enteric-coated tablets containing dehydrated garlic powder (standardized to 1.3% alliin) elicit modest reductions in total cholesterol vs. placebo (<20 mg/dL) in the short term (4–12 weeks), with unclear effects after 20 weeks. Small reductions in low-density lipoprotein (LDL) (by <10 mg/dL) and triglycerides (by <20 mg/dL) have also been indicated in the short-term, although results were variable. High-density lipoprotein (HDL) levels were not found to be significantly affected. Long-term effects on lipids or cardiovascular morbidity and mortality remain unknown. Other preparations (such as enteric-coated or raw garlic) have not been well studied.
Brief Background
Garlic—a member of the lily family, which also includes hyacinth, tulip, onion, leek, and chives—is a culinary herb that has been employed medicinally since ancient times, and remains widely used for the treatment and prevention of cardiovascular disease and cancer today.1
The Codex Ebers, an Egyptian medical papyrus from 1550 BC, mentions garlic as a remedy for a number of ailments, including hypertension, worms, and tumors.
King Tutankhamen's tomb was provisioned with cloves of garlic.
King Tutankhamen’s tomb was provisioned with cloves of garlic.2 Garlic was used in ancient Greece (by Hippocrates), Rome, India, China, and Japan for multiple indications, including performance enhancement, pulmonary and digestive complaints, abnormal growths, cardiovascular health, emotional health and potency, and as an anti-infective agent.3 Garlic was also used in Native American medicine and by early European settlers in North America.
The widespread contemporary use and scientific interest in garlic likely stems from antibiotic applications developed by Louis Pasteur and Albert Schweitzer. It was used during both world wars to prevent gangrene.4 Research on garlic began with study of its antibacterial activity in the 1930s, with subsequent investigations into cancer inhibition beginning in the late 1940s.5 Currently, garlic is one of the most widely used herbal compounds in the United States, with ongoing research in several areas related to cardiovascular health, oncology, and infectious disease, with good evidence supporting its action in alleviating hypertension and hyperlipidemia. Its mechanism of action appears to be related to multiple compounds and not only to allicin, as was previously believed.
Scientific Evidence for Common/Studied Uses
| Hyperlipidemia | A |
| Hypertension | A |
| Cardiovascular disease risk | B |
| Alopecia areata | C |
| Angina | C |
| Antibacterial | C |
| Antifungal | C |
| Atherosclerosis | C |
| Athletic injuries | C |
| Athletic performance | C |
| Benign breast diseases | C |
| Cancer | C |
| Cardiac secondary prevention | C |
| Circulation | C |
| Familial hyperlipidemia | C |
| Gastritis | C |
| Helicobacter pylori infection | C |
| Mosquito repellant | C |
| Otitis media | C |
| Parasitic infections | C |
| Peripheral vascular disease | C |
| Pre-eclampsia | C |
| Sickle-cell anemia | C |
| Systemic sclerosis | C |
| Tick repellant | C |
| Tick repellant | C |
| Type 2 diabetes | C |
| Upper respiratory tract infection | C |
| Warts | C |
Brief Safety Summary
Likely safe: When consumed in amounts usually found in foods.
Possibly safe: When used as a dietary supplement in recommended doses in healthy adults. When used intravenously for less than 1 month.6
Possibly unsafe: When used in individuals with thyroid disorders or who take thyroid medications, as garlic may affect the thyroid.7 When used in patients with bleeding disorders or those taking anticoagulants, NSAIDs or antiplatelet agents, or herbs or supplements that increase the risk of bleeding, such as Ginkgo biloba, due to an increased risk of bleeding.8,9,10 When taking antilipemic agents.11,12,13,14,15 When used in diabetics or patients using hypoglycemic medications, as garlic may reduce blood sugar.16,17 When used in patients with peptic ulcer disease or in those prone to gastric irritation.18,19 When used in patients using antiretroviral agents, particularly protease inhibitors.20,21,22 When used intravenously for more than 1 month, as liver, kidney, and bone marrow damage were reported after that time.23 When used in patients using antihypertensives.24,25,26 When used in patients using cytochrome P450-metabolized agents,27,28 although other research has suggested garlic does not affect the metabolism of CYP450 substrates (alprazolam, P450 3A4 or dextromethorphan, P450 2D6).29 When used in patients using vasodilators,30,31,32 although conflicting evidence exists.33 When used in patients using fish oil or EPA (eicosapentaenoic acid).34
Likely unsafe: When used in patients with a known allergy to garlic or to other members of the Liliaceae (lily) family, including hyacinth, tulip, onion, leek, and chives.35 When used orally in patients prior to some surgical or dental procedures, due to an increased risk of bleeding.36,37,38,39,40
Note: Dishes containing generally acceptable doses of raw garlic are unlikely to increase the risk of perioperative bleeding.41
Dosing
Adult (age ≥18 years)
Oral
Administration of 600–1,200 mg daily of non-enteric-coated, dehydrated garlic powder in 3 divided doses, standardized to 1.3% allicin content, has been studied in multiple clinical trials of hyperlipidemia, peripheral vascular disease, and hypertension (many studies have used the product Kwai®).42,43,44,45,46,47,48,49,50,51 2.1 g of garlic powder daily for 3 months has been used.52 Enteric-coated garlic powder tablet equivalent to 300–400 mg of garlic, twice daily, has been used.53,54 Garlic supplements, equivalent to an average-sized garlic clove, 6 days of every week for 6 months, has been used.55
Pediatric (age <18)
Oral
Safety or efficacy of garlic supplementation has not been established in children. One study in 31 children (ages 8–18 years) with possible familial hyperlipidemia (total cholesterol levels >185 mg/dL) assigned subjects to receive 900 mg of dehydrated garlic powder tablets (Kwai®) in 3 divided daily doses or placebo.56 No significant effects of garlic were observed on lipid levels, although the trial may have been too small to detect effects.
Adverse Effects/Post-Market Surveillance
• A review of 45 randomized trials and 73 studies of garlic use found limited information relating to adverse effects.57 Common side effects of oral and intravenous garlic include malodorous breath, body odor, nausea, vomiting, flatulence, weight loss, facial flushing, tachycardia, dizziness, insomnia, and allergic reactions.58,59 Potential reactions of concern associated with oral garlic use include bleeding (multiple case reports) and hypoglycemia (likely not clinically significant).60,61,62,63,64 Topical exposure may elicit dermatitis or burns (multiple reports) in children and infants, as well as in adults exposed to topical garlic.65,66,67,68,69,70,71,72,73 Based on human study, administration of intravenous garlic was not tolerated for more than 1 month, as liver, kidney, and bone marrow damage were reported after that time.74
Pharmacology
Based on human study, the active principle of garlic for hyperlipidemia is reportedly the essential oil, which contains a combination of sulphur-containing compounds, mainly allyl propyl disulphide and diallyl disulphide.75 Garlic’s lipid-lowering effects may occur via inhibition of HMG-CoA reductase or other enzymes, possibly by diallyl di- and trisulphide components of garlic.76,77,78,79,80,81,82,83, 84,85 Other suggested mechanisms include increased loss of bile salts in feces and mobilization of tissue lipids into circulation, as garlic has a profound effect on postprandial hyperlipidemia.86,87 Wild garlic (Allium ursinum) has shown similar efficacy to garlic (Allium sativum) in decreasing hepatocyte cholesterol synthesis in vitro.88,89 Aged garlic extract and its constituents have been shown to inhibit Cu2+-induced oxidative modification of low-density lipoprotein.90 Aged garlic extract and its constituent S-allyl cysteine have been found to protect vascular endothelial cells from injury caused by oxidized LDL.91
Animal and human cell lines have demonstrated reductions in vascular tissue lipids,92,93 fatty-streak formation and atherosclerotic plaque size.94,95,96,97,98,99,100 The mechanism of action may include reduction of lipoprotein oxidation, as demonstrated in vitro101,102,103 and in vivo,104 possibly due to organosulfur compounds in garlic.105 However, this hypothesis has been in dispute based on a 6-month trial in moderately hypercholesterolemic volunteers, which failed to demonstrate any effects of garlic supplementation on lipoprotein oxidation.106
Evidence Discussion
Summary: Oral tablets containing dehydrated garlic powder appear to elicit modest reductions in total cholesterol vs. placebo (<20 mg/dL) in the short term (4–12 weeks), with unclear effects after 20 weeks.107 Small reductions in low-density lipoprotein (LDL) (by <10 mg/dL) and triglycerides (by <20 mg/dL) may also occur in the short term, although results have been variable. High-density lipoprotein (HDL) values have not been found to significantly change. Numerous controlled trials have examined the effects of oral garlic on serum lipids. Most studies have been small (<100 subjects), with poorly described design and results, and most have reported nonsignificant modest benefits of garlic therapy. Several overlapping meta-analyses have pooled studies and reported significant mean decreases in total cholesterol between 4–12 weeks, and variable significance of reductions in LDL and triglycerides.108,109,110,111,112 The statistical significance of these effects disappears after 20 weeks,113 which may be due to low statistical power. Singh et al. have reviewed clinical trials involving garlic and serum lipids and suggest a high incidence of effectiveness even in well-conducted studies.114 The optimal dose and preparation for maximal benefit is not clear, and most studies have used non-enteric-coated dehydrated garlic powder tablets standardized to 1.3% alliin (Kwai®), 900mg per day in 3 divided doses. One study of familial hyperlipidemia in children did not demonstrate significant effects on lipids, although this trial was not designed to detect small effect sizes. Criticisms of existing trials include the frequent use of the non-enteric-coated preparations, which may allow for degradation of alliinase by gastric acid (although there are scant reliable data on the efficacy of enteric-coated preparations).115,116 There is also preliminary evidence that aged garlic extract may elicit superior resistance to LDL oxidation than fresh garlic.117 The long-term maintenance of effects remains unclear, and the ultimate effect on cardiovascular morbidity and mortality is not known.
Meta-analyses and systematic reviews: Khoo et al. conducted a meta-analysis and systematic review to evaluate the effects of garlic on cholesterol levels in both healthy and hypercholesterolemic subjects.118 A comprehensive search of the Cochrane Library, MEDLINE, Embase, electronic publishing sites, and reference lists of relevant papers and manual searches of relevant journals from inception to March 2008 were performed. Randomized controlled trials of garlic ranging from 11 to 24 weeks in duration were included. Thirteen trials (N=1056 subjects) were eligible for review and meta-analysis. Overall, garlic administration did not show any significant difference in effects on all outcome measures (total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides, apolipoprotein B level) examined when compared with placebo.
Reinhart et al. conducted a meta-analysis and systematic review to determine the impact of garlic on total cholesterol, TAG levels, LDL-cholesterol, and HDL-cholesterol.119 A comprehensive search of MEDLINE, CINAHL, and the Cochrane Database from the earliest possible date through to November 2007 was performed. Medical subject headings (MeSH) and keywords included ‘garlic,’ ‘Allium sativum’ or ‘allicin’. Results were limited to randomized clinical trials in human subjects. Twenty-nine studies were included in the analysis. After meta-analysis, garlic was found to significantly reduce total cholesterol (-0.19; 95% CI -0.33, -0.06 mM/L) and TAG (-0.11; 95% CI -0.19, -0.06 mM/L) but exhibited no significant effect on LDL or HDL.
Knox et al. conducted a systematic review to evaluate the effects of dietary supplements on the prevention and treatment of coronary artery disease.120 ProQuest, MEDLINE, and the Cochrane Library between May 2004 and May 2006 were searched, followed by a manual search of the reference list of relevant retrieved papers. Supplements were identified by cross matching the MeSH term dietary supplements with the MeSH terms hypertension, hypercholesterolemia, and myocardial infarction. Only randomized, clinical, double-blind, placebo controlled trials of more than 1 week’s duration, published in English in a peer-reviewed journal after 1980, were included. The following supplements were identified: beta carotene, coenzyme Q10, calcium, black currant seed oil, hawthorn, magnesium, vitamin C, vitamin E, artichoke, chromium, guggul, magnesium-pyridoxal-5’-phosphate-glutamate, policosanol, and garlic. For garlic, 2 trials were identified for hypertension, and 18 were identified for hyperlipidemia. From the 2 trials evaluating hypertension, garlic was found to reduce systolic and diastolic blood pressure. Seven of the 18 studies found positive effects in reducing total and LDL cholesterol with little effect on HDL-cholesterol.
Singh et al. conducted a systematic review to evaluate Ayurvedic and commonly used herbals for the treatment of hyperlipidemia.121 The terms Ayurvedic, herbal, and hyperlipidemia, were exploded as Medical Subject Headings (MeSH) terms and searched as text words. The terms garlic, guggul, Terminalia arjuna, Abana, (a combination herbal remedy that contains Terminalia arjuna) guar gum, and fenugreek were searched only as text words. Fifty-one studies were identified: 10 for guggul, 7 for Terminalia arjuna, 3 for guar gum, and 26 for garlic. There were 5 herbals in the miscellaneous category that had only 1 article each: Maharishi Amrit Kalash, a proprietary formula by Maharishi Ayurveda (including Indian gooseberry, ashwagandha [winter cherry], and heart-leaved moonseed); Arogyavardhini, a traditional ayurvedic formula indicated for liver disorders (ingredients including Terminalia chebula and amalaki [Phyllanthus emblica]); silymarin, the active constituent in milk thistle; fenugreek; and Coccinia grandis, or ivy gourd. Randomized controlled trials improved by decade of publication. Effectiveness was reported in more than 50% of garlic, more than 80% of guggul, and 100% of Arjuna. Safety scores did not improve.
In 2003, Alder et al. published a systematic review of 10 randomized controlled trials evaluating garlic for hyperlipidemia.122 Six
studies reported positive effects, with a mean 9.9% reduction in total cholesterol, 11.4% reduction in LDL, and 9.9% reduction of triglycerides. The authors noted small sample sizes, lack of power calculations, and lack of control of diet during studies as limitations of the existing literature.
In 2001, the results of the Mulrow et al. analysis of lipid effects were slightly reformulated and published in Archives of Internal Medicine.123 This meta-analysis included 34 randomized trials representing 1,798 patient records. The authors noted similar effects of garlic vs. placebo as in the prior analysis: small reductions in total cholesterol at 1 month (range of average pooled reductions: 1.2–17.3 mg/dL), small reductions at 3 months (range of average pooled reductions: 12.4–25.4 mg/dL), and no significant differences between garlic and controls at 6 months. Changes in LDL and triglyceride levels paralleled total cholesterol level results, but no statistically significant changes in HDL were observed. The authors noted that their conclusions were limited by the small samples sizes and low overall quality of the included trials.
In 2000, Mulrow et al. prepared a high-quality systematic review and meta-analysis for the Agency for Health Care Research and Quality to assess the safety and efficacy of garlic for cardiovascular disease and cancer prevention.124 Literature searches were conducted in 11 electronic databases using accepted synonyms and were not restricted by language; selection criteria included randomized controlled trials of at least 4 weeks duration. The authors included 44 studies of lipid outcomes in 42 published articles of garlic for hyperlipidemia. There was variability between trials in terms of method and timing of lipid measurements as a primary or secondary outcome. Quality of design and reporting was also variable, with unclear reporting of randomization techniques in 82% of studies, and unclear blinding in 25%. The majority of trials included fewer than 100 participants. Most trials were conducted in Germany or the United States, and were sponsored by industry. Of these trials, 5 were not placebo-controlled and 5 evaluated combination products. Meta-analyses of placebo controlled trials that reported total cholesterol outcomes at 4–6 weeks, 8–12 weeks, and 20–24 weeks were conducted. Pooled data demonstrated a significant average reduction in total cholesterol levels of 7.2 mg/dL (95% CI, 1.2–13.2) after 4–6 weeks of any type of garlic therapy vs. placebo, which increased when studies only using standardized dehydrated preparations were analyzed, to 10.2 mg/dL (95% CI, 3.1–17.3). Combined data from studies with reported outcomes at 8–12 weeks showed a significant average reduction in total cholesterol levels vs. placebo of 17.1 mg/dL (95% CI, 13.0–25.4). In contrast, combining the results of 8 trials with reported outcomes at 20–24 weeks did not demonstrate significant cholesterol reductions vs. placebo (1.2 mg/dL; 95% CI, -8.2 to 10.7). This lack of significance may be due to true lack of effect, or to low statistical power. These results suggest a need for further study of outcomes beyond 12 weeks.
This publication also reported the results of an analysis of 13 trials reporting low-density lipoprotein (LDL) changes at 8–12 weeks. This analysis revealed a significant reduction of 6.2 mg/dL vs. placebo (95% CI, 0.8–11.7), with greater benefits when the analysis was limited to 10 trials using dehydrated garlic preparations (6.7 mg/dL; 95% CI, 0–13.5). Pooled data from 17 studies measuring triglyceride changes after 8–12 weeks showed a significant reduction of 19.1 mg/dL (95% CI, 7.6–30.4), with greater benefits when analysis was limited to the 13 trials of dehydrated garlic preparations (21.1 mg/dL; 95% CI, 8.3–34.0). Analysis of 14 trials that measured high-density lipoprotein (HDL) levels at 8–12 weeks found a nonsignificant reduction of 0.9 mg/dL (95% CI, -1.0 to 2.8).
Stevinson et al. conducted a 2000 meta-analysis of garlic therapy for hypercholesterolemia,125 using more stringent inclusion criteria, analyzing 13 trials including 796 subjects. Trials were excluded that were not randomized, double-blind, or placebo controlled; all studies administered oral garlic monotherapy, examined patients with mean baseline cholesterol levels ≥200 mg/dL, and used total cholesterol as an endpoint. Ten of the 13 studies used the same garlic product (Kwai®) in daily doses between 600–900 mg per day, administered over 8–24 weeks. Pooling of data revealed a significant reduction in total cholesterol levels vs. placebo of 15.7 mg/dL (95% CI, 5.7–25.6). Notably, this analysis included 1 study of children with familial hypercholesterolemia,126 a genetic defect which may be more refractory to treatment than other etiologies of hyperlipidemia, and may have reduced the calculated magnitude of effect. A sub-analysis was conducted which pooled data from the 6 included trials of highest methodological quality, and found a nonsignificant difference between garlic and placebo, although this analysis was of low power. Lawson127 criticized this meta-analysis, citing the use in most included studies of the product Kwai®, which does not have a protective coating to prevent degradation of alliinase by gastric acid. Lack of inclusion of other products and uncertain bioavailability were also noted as potential weaknesses.
Earlier meta-analyses reported more impressive benefits of garlic over placebo. In 1993, Warshafsky et al. analyzed 5 studies with 410 subjects,128 and in 1994, Silagy and Neil analyzed 16 studies with 952 subjects.129 Warshafsky et al. reported a significant reduction in total cholesterol of 23 mg/dL vs. placebo (95% CI, 17–29), while Silagy and Neil noted a reduction of 29.7 mg/dL vs. placebo, with no significant difference between daily garlic doses of 600 mg vs. 900 mg. Dried garlic was also associated with a moderate decrease in triglyceride levels vs. placebo. However, the trials included in these meta-analyses were of heterogeneous design and quality, including inappropriate randomization, lack of dietary control, short duration, and inadequate explanation of statistical analysis in most trials. A 1996 follow-up to the Silagy and Neil meta-analysis reported the results of a 6-month, randomized controlled trial in 115 patients with moderate hyperlipidemia (total cholesterol 6.0–8.5 mM/L and LDL ≥3.5 mM/L). Subjects received either placebo or 900 mg of daily dried garlic powder (standardized to 1.3% allicin). The authors reported no significant benefit of garlic vs. placebo in levels of total cholesterol, LDL, or HDL.130 A meta-analysis including this trial showed a slightly less significant reduction in total cholesterol levels vs. placebo than the prior Silagy and Neil meta-analysis.
