Introduction

Cardiovascular disease (CVDs) is the leading cause of death globally, 17 million people die to CVDs each year and it is estimated to reach 24.8 million deaths in 2030 in the world [1]. Dyslipidemia refers to a number of lipid abnormalities including elevated total cholesterol (TC), triglyceride(TG), low-density lipoprotein-cholesterol (LDL-C), and decreased high-density lipoprotein-cholesterol (HDL-C) [2]. In modern society, due to bad dietary habits, the dyslipidemia has been increasing. It is estimated that prevalent cases of dyslipidemia increased at the rate of 1.76% per year to surpass 500 million in 2022. Furthermore, data from epidemiological studies suggest that dyslipidemia is an important predisposing factor for CVDs, including coronary heart disease(CHD), myocardial infarction, stroke, hypertension and other chronic diseases [3, 4]. LDL-C has always been known to be the main cause of atherosclerosis [5], elevated TC can also increase the risk of ischemic heart disease [6] and other lipid indicators are also proved to play important roles in CVDs development. So it is urgent to take effective measures to improve lipid profiles and prevent the CVDs including the use of effective food products.

In recent years, several alternative approaches including different medicinal plants or dietary approaches with the aim of reducing the risk of cardiometabolic diseases have been investigated, some of which them have also shown promising results [7,8,9,10]. Garlic, also known as Allium sativum L.(Liliaceae), is a member of the family Alliaceae. It has been used as an important medicinal ailment in many countries for long history [11, 12]. Ancient Indian and Chinese medicine recommend that garlic can be used to help respiration and digestion, and to treat leprosy and parasitic diseases. Besides, garlic plays important roles in the treatment of many diseases such as infectious diseases, gynecologic diseases, toothache, arthritis and hypertension [13,14,15,16]. Because of the widespread effects of garlic in maintaining good health, it has attracted particular attention of modern medicine.

The effect of garlic in anti-hyperlipidemia has been confirmed in some animal studies [17, 18]. However, in human studies, the association between garlic consumption and blood lipids have yielded inconsistent results. Earlier reviews strongly suggested that garlic was effective as a lipid-lowering agent [19, 20]. And some clinical experimental studies have shown that garlic has the ability to inhibit cholesterol biosynthesis, reduce lipids and fibrinogen concentrations [21, 22], decrease LDL oxidation and increase fibrinolitic activity [23]. However, a non-systematic review in 2003 gave contradictory results [24]. And some other studies have concluded that garlic has no significant effects on blood lipids [25,26,27]. Recent several meta-analysis published in 2014–2023, but they mainly focus on the impact of garlic on coronary heart disease(CAD) and metabolic syndrome(MetS). Hence, in their researches that participants were diagnosed as hypertension, hypercholesterolemia, obesity, CVDs or insulin resistance, and results also showed an inconsistency on TG, LDL-C and HDL-C [28,29,30].

With the increase of sample size, we carried this meta-analysis to further evaluate the effects of garlic on dyslipidemia. Unlike the previously published meta-analysis, all the enrolled participants in this study were diagnosed with dyslipidemia and not taking any hypolipidemic drugs. Our aim is to explore whether garlic is an effective and safe alternative approach for the treatment of dyslipidemia and prevention of CVDs.

Materials and methods

Literature search strategy

This meta-analysis was planned and conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [31], registration code: CRD42024506637. A systematic search of published articles was conducted in the PubMed, Embase, Scopus, Web of Science and Cochrane Library databases up to June 2024, and the publication language limited to English. Following search terms are used: (“garlic” OR “garlicin” OR “garlic oil” OR “garlic extract” OR “garlic cloves” OR “garlic powder” OR “allicin”) AND (“Serum lipids” OR “blood lipids” OR “lipid profile” OR “lipidemia” OR “hypercholesterolemia” OR “hyperlipidemia” OR “dyslipidemia” OR“ hyperlipemia” OR “total cholesterol” OR “TC” OR “triglyceride” OR “TG” OR “high density lipoprotein”OR “HDL-C” OR “low density lipoprotein”OR “LDL-C”) AND (“randomized controlled trial” OR “RCT” OR “controlled clinical trial” OR “controlled trial” OR “Cross-over” OR “Parallel”). To enhance the comprehensiveness of literature search, we conduced backward and forward snowballing searches to identify additional relevant articles [32].

Inclusion and exclusion criteria

Studies that met the following criteria were included in this meta-analysis: (1) Randomized controlled trial (RCT) in human with either parallel or crossover design; (2) concerning the effects of garlic consumption on blood lipid; (3) outcome is at least one of the lipid parameters (LDL-C, HDL-C, TC, TG); (4) participants with hyperlipidemia and ⩾18 years of age; (5) means and standard deviation (SD) or standard error of the mean (SEM) or 95% confidence intervals(CIs) had been provided.

The exclusion criteria were the following: (1) duplicate studies; (2) reviews, reports, conferences, animal studies; (3) insufficient data; (4) garlic combined with other lipid-lowering drugs or foods; (5) Intervention duration < 2 week; (6) Studies that were not randomized or no placebo control group; (7) Non hyperlipidemia patients. Two of the authors independently searched all references, using a kappa statistic to assess the agreement between the authors [33]. When kappa value > 0.7, we accept the decision, otherwise reject the decision and request a third-party ruling.

Data extraction

Two authors performed data extraction, again any disagreements were discussed and resolved by consensus. The following information was collected : first author’s name, publication year, country, study design, age and gender as well as health status of participants, sample size, daily dose of garlic consumption, intervention duration, blood lipids index( LDL-C, HDL-C, TG, TC ). In each trial, the means and standard deviation(SD) of blood lipids at baseline and endpoint were extracted. For studies that had multiple time points measurements, only the last end point was used for analysis. Blood lipid levels were collated in mmol/L. Therefore, blood lipid values reported in mg/dl were converted to mmol/L, where 1 mg/dL = 0.02586 mmol/L for TC, LDL-C, HDL-C, and 1 mg/dL = 0.01129 mmol/L for TG [34].

Quality assessment and quality of evidence

The study quality of RCTs was assessed with the Cochrane Tool, which includes : selection bias, performance and detection bias, attrition bias, reporting bias and other possible sources of bias. Each domain was judged to have a “low risk”, “high risk” or “unclear risk”. Any disagreements were resolved by consensus. Quality of evidence GRADE (Standards for the Development, Evaluation, and Evaluation Working Group) criteria were used to evaluate the overall certainty of the evidence in the meta-analyses. The quality of the evidence was categorized according to four assessment criteria: high, moderate, low, and very low [35].

Statistical analysis

Weighted mean differences (WMD) with 95% confidence interval (CI) were used to evaluate LDL-C, HDL-C, TG and TC changes levels. Net changes in measurements (change scores) for the trials were calculated using the following formula: (measure at the end of follow-up in the treatment group-measure at baseline in the treatment group) – (measure at the end of follow-up in the control group – measure at baseline in the control group). Change-from-baseline SD was estimated using the equation:

$${\rm{S}}{{\rm{D}}_{{\rm{change}}}}{\rm{ = }}\sqrt {{\rm{SD}}_{{\rm{baseline}}}^{\rm{2}}{\rm{ + SD}}_{{\rm{final}}}^{\rm{2}} - {\rm{2}} \times {\rm{R}} \times {\rm{SD}}_{{\rm{baseline}}}^{\rm{2}} \times {\rm{SD}}_{{\rm{final}}}^{\rm{2}}}$$

the R is the correlation coefficient, to be conservative, a minimum correlation coefficient of 0.5 was used. If SDs were not reported directly, we can calculate them from SEM or 95% CI using following formula: (1)\({\rm{SD = SEM}} \times \sqrt n\) (where standard error of the mean (SEM) was only reported); (2) \({\rm{SD}} = {\rm{ }}({\rm{upperlimit}} - {\rm{lowerlimit}}) \times \sqrt n \div 3.92\) (If the outcome measures were reported in median and range (or 95%CI), where n is the number of subjects [36]. Heterogeneity among the studies was estimated using the I2 and Q statistic. For the Q statistics, P < 0.10 indicated statistically significant heterogeneity. And the I2 values of 25%, 25–50%, 50–75%, and > 75% were classified as indicating no, small, moderate, and significant heterogeneity, respectively. Pooled WMD were obtained using either a fixed-effects model, if I2 < 50% or random-effects model if I2 > 50% according to Cochrane handbook. In addition, we performed subgroup analysis using sample size (≤ 60 and > 60), age (≤ 50 years old and > 50 years old ), types of garlic(garlic powder, garlic tablets, garlic oil) and intervention duration (≤ 12 weeks and > 12 weeks) to explore the potential heterogeneity.

To assess the potential publication bias, we used funnel plots and Egger’s linear regression test [37]. Sensitivity analyses were carried out by excluding each studies and reanalyzing the data. All statistical analyses were performed with STATA 12.0 (Stata Corp, Texas, USA). Significance was set at a P < 0.05 and all statistical tests were two-sided.

Results

Literature search

An electronic literature search identified 8623 studies concerning effects of garlic consumption on blood lipids level, 8602 of which were excluded based on the a series of reasons (firstly, exclusion of 1711 duplicated articles, then 6912 articles were screened by title and abstract, leading to exclusion of 3634 irrelevant studies, 2164 reports, reviews or meta-analysis, or letters, 972 non-human studies. Finally, leaving 142 articles for full-text review, where 46 articles do not provide available data, 21 articles are not RCT, 19 articles are healthy individuals, 31 articles garlic combined with other lipid-lowering food and 4 articles intervention duration < 2 week), resulting in inclusion of a total of 21 studies in the meta-analysis. The details of literature search are shown in Fig. 1.

Fig. 1
figure 1

Flowchart of study procedure

Full size image

Study characteristics

The characteristics of included studies are shown in Table 1. A total of 21 RCT (20 parallel design, 1 crossover design) [38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58] that met the inclusion criteria for our meta-analysis were published between 1981 and 2021. Of these studies, twelve were carried out in Asian countries, including Korea(n = 2), India(n = 2), Iran(n = 2), Nepal(n = 1), Iraq(n = 1), Pakistan(n = 2), Thailand(n = 1) and Turkey (n = 1); Five in American countries, including USA (n = 4), Canada (n = 1); Others including Australia (n = 2), and Russia (n = 2). In total, 1603 subjects included in this meta-analysis. Mean ages of participants ranged from 25 to 70 years. Except two studies [44, 46] were done on male, one study in female [38], others all were conducted on both sexes. Sample size ranged from 23 to 258, the intervention period ranged from 8 weeks to 36 weeks. Five different garlic preparations were tested in the trials, including garlic powder [40, 44, 45, 52, 53, 57], raw garlic [46, 54], garlic tablets [47, 48, 51], aged garlic [43, 56], garlic oil [42, 58], six studies did not provide detailed information about the garlic [38, 39, 41, 49, 50, 55]. All studies surveyed subjects with dyslipidemia.

Table 1 Characteristics of 21 RCT studies included in the meta-analysis
Full size table

Methodological quality of included trials

Documents quality evaluation were done based on standard methods recommended by Cochrane Collaboration. All of the included studies were randomized, placebo-controlled trials, including thirteen double-blind trials [38, 39, 42,43,44,45, 48, 50,51,52,53, 56, 57] and two single-blind trial [40, 47]. Thus they were considered as low risk of bias and others as unclear risk. Although randomization was declared in all studies, none of the trials reported how the randomization was conducted. Therefore, they were regarded as unclear risk of bias. None of the trials provided incomplete outcomes and selective reporting, and were assessed as low risk; In addition, none of the trials provided a clear description of allocations concealment, so they all were assessed as “unclear” risk(Table 2).

Table 2 Quality assessment of included RCT studies based on the Cochrane guidelines
Full size table

Effect of garlic consumption on blood lipids

TC

Effect of garlic consumption on TC levels was explored from 20 studies ( Intervention group = 661, control group = 668 ). Heterogeneity test results found significant heterogeneity in these.

studies (Q = 136.98, P < 0.001, I2 = 86.1%). And a random-effects model was applied, the summary WMD for 20 studies showed that garlic consumption significantly reduced TC levels(WMD = -0.64mmol/L, 95%CI = -0.75 − -0.54, P < 0.001 ). (Fig. 2) Subgroup analyses indicated that there was no significant difference in TC decrease between sample size and intervention duration. However, TC significantly decreased in patients age > 50 years(WMD = -0.66mmol/L, 95%CI = -0.73 − -0.59 ) compared to age ≤ 50 years(WMD = -0.45mmol/L, 95%CI = -0.69 − -0.20 ). For different types of garlic, a greater reduction in garlic oil group(WMD = -0.73mmol/L, 95%CI = -0.91 − -0.56 ) compared to garlic powder(WMD = -0.51mmol/L, 95%CI = -0.80 − -0.23 ). The quality of evidence related to TC was downgraded to moderate due to serious limitations in indirectness (Table 3).

Fig. 2
figure 2

Effects of garlic consumption on changes in TC(mmol/L)

Full size image
Table 3 Summary of findings and quality of evidence assessment using GRADE approach
Full size table

TG

Effect of garlic consumption on TG levels was explored from 19 studies ( Intervention group = 658, control group = 670 ). Combined results through the random-effects model revealed that garlic consumption significantly reduced TG levels(WMD = -0.17mmol/L, 95%CI = -0.26 − -0.09, P = 0.000 ) with moderate heterogeneity in these studies (Q = 88.64, P < 0.001, I2 = 79.7%). (Fig. 3) Subgroup analyses showed that a significant reduction in people age > 50 years(WMD = -0.19mmol/L, 95%CI = -0.28 − -0.04 ) compared to age ≤ 50 years(WMD = -0.07mmol/L, 95%CI = -0.14 − -0.01 ). However, there was no significant effect on TG in male patients, sample size > 60 and > 12 weeks intervention duration. According to the GRADE approach, TG was considered to have a moderate quality of evidence (Table 3).

Fig. 3
figure 3

Effects of garlic consumption on changes in TG(mmol/L)

Full size image

LDL-C

Effect of garlic consumption on LDL-C levels was explored from 18 studies (625 cases and 637 controls ). We observed significant heterogeneity in these studies (Q = 133.59, P < 0.001, I2 = 88.9%). And a random-effects model was applied, the summary WMD for 18 studies showed that garlic consumption significantly reduced LDL-C levels (WMD = -0.44mmol/L, 95%CI = -0.57 − -0.31, P < 0.001). (Fig. 4) Subgroup analyses indicated that garlic oil group resulted in a greater reduction in LDL-C levels(WMD = -0.66mmol/L, 95%CI = -0.86 − -0.46 ). However, there was no significant difference in LDL-C decrease between age, sex and sample size (WMD ranged between − 0.41 and − 0.48). The LDL-C quality of evidence was rated as moderate using the GRADE approach (Table 3).

Fig. 4
figure 4

Effects of garlic consumption on changes in LDL-C(mmol/L)

Full size image

HDL-C

Effect of garlic consumption on HDL-C levels was explored from 20 studies (690 cases and 699 controls). Combined results through the random-effects model revealed that garlic consumption had significant effects on HDL-C (WMD = 0.04mmol/L, 95%CI = -0.00 − 0.08, P < 0.001) with significant heterogeneity across these studies (Q = 146.27, P < 0.001, I2 = 87.7%). (Fig. 5) However, all subgroup analysis results are consistent, they indicated that garlic consumption has no significant effect on HDL-C. The quality of evidence related to HDL-C was downgraded to low due to serious limitations in indirectness and imprecision (Table 3).

Fig. 5
figure 5

Effects of garlic consumption on changes in HDL-C(mmol/L)

Full size image

Sensitivity analysis and publication bias

Sensitivity analysis results showed that removed each studies did not significantly alter the overall effect of garlic consumption on TC (WMD altered between − 0.7 and − 0.49), LDL-C (WMD altered between − 0.61 and − 0.27) and TG (WMD altered between − 0.21 and − 0.06). No publication bias was detected in current meta-analysis, although slight asymmetries were observed in the funnel plots, Egger’s linear regression test were not statistically significant (TC: P = 0.465; TG: P = 0.105; LDL-C: P = 0.079; HDL-C: P = 0.609).

Discussion

This meta-analysis evaluated the blood lipid-lowering effects of garlic with a sample size of over 1600 from 21 RCTs. The evidence from this meta-analysis revealed that garlic consumption significantly reduced TC, TG, LDL-C and slightly increased HDL-C concentration.

Studies have shown that a 1mmol/L reduction in TC and LDL-C levels results in a 26.6–29.5% decrease for any cardiovascular disease–related event [6]. And our study found a 0.64, 0.17, 0.44mmol/L reduction in TC, TG, LDL-C separately and a 0.04mmol/L increase in HDL-C after 3 weeks to 12 weeks of garlic consumption. This conclusion about effects of garlic consumption on TC is consistent with the most previous studies [38, 39, 43]. However, the conclusion in TG is contrary to two studies [46, 47]. And a significantly reduced effects on TC, TG and LDL-C in our study compared to garlic consumption had no any significantly effects in some studies [27,28,29,30]. In addition, most previous studies have not found that garlic has a significant effect on HDL-C. But we found garlic consumption slightly increased HDL-C concentration. This difference may be caused by several reasons: First, in previous studies, they repeatedly used the data of different time points measurements in the same article. And in our study, we only used the data of the highest garlic consumption and the longest observation time. Second, most previous studies did not limit the health status of participants. And in our meta-analysis, all studies involved hyperlipemia patients. Third, we include the latest data from 2014 to 2023. So, we confirmed the significant blood lipid-lowering effect of garlic.

Subgroup analyses indicated that there was not significant difference in TC, TG and LDL-C changes between sample size. However, TC, TG and LDL-C significantly decreased in people aged > 50 years. It indicated that older people were more susceptible to the protective effects of garlic; In addition, we found TC significantly decreased in male patients, while not for TG and LDL-C. The study by Zhang et al. suggested that women may benefit more than men from garlic treatment [59], while the study by Zeng et al. showed that the effects of garlic in men were greater than those in subjects of both gender [60]. Besides, TC, LDL-C significantly decreased in subgroup of ≤ 12 weeks intervention duration while for TG in subgroup of > 12 weeks intervention. Some studies showed significant reduction in TC vs. placebo over 4–12 weeks intervention with unclear effects after 20 weeks [61, 62]. So age, sex and intervention duration may be important sources of heterogeneity.

In addition, garlic oil and garlic tables greatly reduced TC and LDL-C compared with garlic power. So, the effect of garlic consumption on blood lipid levels may depend on the method of preparation and the type of garlic. Studies have shown that garlic in non-powder form which lowered TC and TG more significantly than powder form garlic preparation over 1–3 months [63].

Several mechanisms of garlic as hypolipidemic action has been proposed, including: (1) Depressed activity of hepatic lipogenesis and cholesterolegenic enzyme such as malic enzyme, fatty acid synthetase(FAS), glucose 6–phosphate dehydrogenase and 3-hydroxy, 3-methyl glutaryl CoA (HMG-CoA) reductase [64]; (2) Inhanced excreation of acidic and neutral steroid into bile after garlic feeding [65] and increased loss of bile salt in faeces and mobilization of tissue lipid into circulation as garlic has a profound effect in post-prandial hyperlipidemia [30]; (3) Suppressed LDL oxidation by garlic preparation, especially by aged garlic extract (AGE) and aqueous garlic extract, thus having anti-atherogenic effect [66, 67]. Allicin present in garlic has been identified as the active compound responsible for anti-atherosclerotic effects. Recent in-vitro studies revealed that water soluble organosulfur compounds especially S-allyl cysteine (SAC) present in aged garlic extract and diallyldisulphide (DADS) present in garlic oil are also potent inhibitors of cholesterol synthesis [64, 68]; (4) Garlic is a potential stimulant of lipase enzyme thereby, decreasing blood triglyceride level [69].

This meta-analysis has its own strengths. First, this is the latest study that explore the blood lipid-lowering effects of garlic in human. And, we synthesized and quantified evidence from 21 independent RCTs which from more strict inclusion criteria. Second, in data analysis, we only used the data of the highest dose garlic consumption and the longest observation time in each trial. The results found that garlic consumption significantly decreased TC, TG, LDL-C and slightly increased HDL-C concentration. Third, participants of the studies reviewed had a very low heterogeneity and were similar in levels of primary cholesterol and health status. All studies involved dyslipidemia patients. And the results were unlikely to be affected by publication bias.

Meanwhile, several potential limitations of our study deserve mention. Firstly, we only searched studies from English databases and missed non-English articles may affect the final results. Secondly, moderate to high heterogeneity was present for the TC, TG, LDL-C, and HDL-C analyses and was only partly explained by subgroup analyses. Third, in each trial, the type and dose of garlic are different, this may be also an important sources of heterogeneity. Thereby, the correlations between garlic consumption and serum lipid levels require further investigation through epidemiological studies with a larger sample size.

Conclusions

In conclusion, results from this meta-analysis indicate that garlic consumption is beneficial for the treatment of dyslipidemia and prevention of CVDs. This result from a significant reduction in TC, TG, and LDL-C. These findings also suggest that garlic can be used as an effective herbal medicine in clinical practice of hyperlipidemia.