Scolaris Content Display Scolaris Content Display

大蒜治疗普通感冒

Collapse all Expand all

研究背景

据称大蒜因其抗菌和抗病毒的特性,在众多有益作用中,可以缓解普通感冒。故大蒜补充剂被广泛使用。普通感冒与严重的发病率和经济后果有关。平均而言,儿童每年有6到8次感冒,成年人有2到4次感冒。

研究目的

与安慰剂、不治疗或其他治疗方法相比,确定大蒜 (Allium sativum) 对预防或治疗普通感冒是否有效。

检索策略

我们检索了Cochrane对照试验中心注册库(Cochrane Central Register of Controlled Trials, CENTRAL)(2014年第7期),OLDMEDLINE(1950年至1965年)、MEDLINE(1966年1月至2014年7月第五周)、EMBASE(1974年至2014年8月)和AMED(1985年至2014年8月)。

纳入排除标准

比较大蒜与安慰剂、无治疗或标准治疗对普通感冒预防和治疗的随机对照试验。

资料收集与分析

两名综述作者从检索结果中独立检查和筛选试验,评估和评定研究质量并提取相关数据。

主要结果

在这次更新的综述中,我们从检索中确定了八项可能相关的试验。同样,只有一项试验符合纳入标准。该试验将146名受试者随机分配到含大蒜素的大蒜胶囊组(剂量未指定)或安慰剂组(每天一次),为期12 周。该试验报告大蒜干预组发生了24次普通感冒,而安慰剂组为65次(P<0.001),这导致大蒜组与安慰剂组相比患病天数更少(111天对366天)。两组从普通感冒中恢复的天数相似(4.63对5.63)。只有一项试验符合纳入标准,因此得出的结论有限。该试验依赖于自我报告的普通感冒发作,但在随机化和分配方案隐藏方面质量合理。不良反应包括皮疹和气味。

作者结论

关于大蒜预防或治疗普通感冒效果的临床试验证据不足。一项试验表明,大蒜可能预防普通感冒的发生,但还需要更多的研究来证实这一发现。对有效性的主张似乎主要依赖于质量差的证据。

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

大蒜治疗普通感冒

研究背景

人们普遍认为大蒜对普通感冒有用。这一观点是基于大蒜的传统用途和一些实验室证据,即大蒜具有抗菌和抗病毒的特性。平均而言,成年人每年有两到四次普通感冒。

研究特征

证据检索至2014年8月7日。在找到的八项研究中,只有一项符合综述标准。这项研究对146名受试者进行了为期三个月的评估。在此期间,一半的受试者服用安慰剂胶囊,另一半服用大蒜胶囊。当受试者出现感冒症状时,他们会在日记中记录。

主要结果

纳入的研究发现,连续三个月每天服用大蒜胶囊(而不是安慰剂胶囊)的人感冒较少。也就是说,在三个月的时间里,大蒜组发生了 24 次普通感冒,而安慰剂组发生了 65 次。当受试者患感冒时,两组的患病时间相似(分别为4.63天和5.63天)。

证据质量

与安慰剂组(1名)相比,大蒜组(4名)的受试者在打嗝时被注意到有气味的人更多,因此,对受试者的施盲可能是不够的。然而,其他潜在的偏倚得到了很好的控制。这唯一纳入的研究与综述问题直接相关。尽管试验规模很小,但有足够的受试者提供精确、可靠的结果。没有证据表明结局是选择性报告的。然而,这是可能的,因为结果似乎没有事先决定。考虑到补充剂公司进行积极试验的经济激励,那些显示大蒜没有效果的试验也有可能都没有发表。总体上证据质量中等。

副作用

在这项小型试验中可能出现的副作用包括气味和皮疹。需要更多关于大蒜可能的副作用的信息。

Authors' conclusions

Implications for practice

There is no conclusive evidence to recommend garlic supplements as a preventative or treatment option for the common cold. A single, small trial was found suggesting garlic might reduce the incidence of the common cold if taken continuously as a daily prophylactic (preventive treatment) but the results require validation. There is currently no evidence to help decide whether treating common colds with garlic will reduce symptom severity or days of illness. Anecdotally, adverse events reported include odour and minor skin or respiratory irritation. The frequency of adverse effects could not be determined from the evidence available.

Implications for research

Further research is needed to provide conclusive evidence of the efficacy of garlic for the common cold. Large, double‐blind randomised controlled trials should be conducted. Outcomes should be measured objectively, according to pre‐defined criteria, in a format that allows comparison.

Background

Description of the condition

The common cold is a heterogenous group of diseases caused by numerous viruses that belong to several different families (Heikkinen 2003). The viruses include picornaviruses (notably, rhinoviruses and enteroviruses), coronaviruses, adenoviruses, parainfluenza viruses, influenza viruses, metapneumoviruses and respiratory syncytial viruses (Fendrick 2003). They all cause the common symptoms of nasal stuffiness and discharge, sneezing, sore throat and cough. Other symptoms may also include hoarseness, headache, malaise and lethargy (Heikkinen 2003). The transmission of these viruses occurs via contact with secretions or small‐ or large‐particle aerosols (Heikkinen 2003). On average, children have six to eight and adults two to four colds per year (Heikkinen 2003).

The total annual economic impact of the common cold is estimated to be USD 40 billion in the USA, including the financial impact of medical costs, days off work and the possibility of severe complications in at‐risk groups (Fendrick 2003).

Due to the many different virus types, all with varying pathogenetic mechanisms, it is understandable that an effective universal treatment for the common cold has not been developed (Heikkinen 2003). Current treatments aim to relieve the symptoms of the common cold but Cochrane Reviews suggest that most commonly used treatments are of limited or uncertain effectiveness (Hemilä 2013; Karsch‐Völk 2014; Singh 2013).

Description of the intervention

Garlic (Allium sativum) has been traditionally used for both culinary and medicinal purposes (Rivlin 2001). Garlic remedies include raw garlic and commercial preparations such as powders, oil and aged extracts (Ruddock 2005; Staba 2001).

The exact usage of garlic for the common cold probably varies worldwide. A cross‐sectional population study conducted in Australia in 2007 found that 10.7% of participants used garlic; 29.8% of these for cold, flu or fever (Zhang 2008). In data from the USA in 2002, 3.76% of the population used garlic supplements (Barnes 2004). However, like other usage surveys, this study did not report the indication for use (Barnes 2004; Harris 2000; MacLennan 2006). Since many manufacturers of garlic supplements claim their products boost the immune system and assist in the prevention and treatment of the common cold, it is reasonable to assume that garlic supplements are commonly used by consumers for this purpose. The prevalence of herbal medicine use seems to be relatively consistent between Western countries (Harris 2000; MacLennan 2006).

How the intervention might work

Garlic is alleged to have antimicrobial, antifungal and antiviral properties (Ankri 1999; Ruddock 2005; Weber 1992). It is purported to lower cholesterol and triglyceride levels, reduce blood pressure, slow the development of atherosclerosis and act as an anticoagulant (Kyo 2001; NCCAM 2006; Tapsell 2006). Other studies have reported anti‐carcinogenic and immunomodulatory effects (Kyo 2001).

The mechanism of action of garlic as an antimicrobial and antiviral agent is unknown. However, its sulphur‐containing derivatives may exert an effect (Naganawa 1996; Weber 1992). Alternatively, the effects of garlic may be due to ajoene, a derivative of allicin which displays antiplatelet and antimicrobial activities in vitro (Naganawa 1996; Ruddock 2005; Weber 1992). When raw garlic is crushed, allicin is produced (Naganawa 1996; Weber 1992). Allicin has demonstrated antibacterial properties in vitro (Cavallito 1944), but some studies suggest it is an unstable compound that is not detected in the circulation after ingestion (Naganawa 1996). Fresh garlic is estimated to contain approximately 4.38 to 4.65 mg of allicin per gram of garlic; thus for one fresh clove of garlic, weighing approximately 4 g, there is approximately 17.52 to 18.60 mg of allicin (Ruddock 2005; Staba 2001; WHO 1999). It is important to recognise that commercial garlic preparations may contain different garlic‐derived compounds according to the process used to formulate the product (Miller 2000; Ruddock 2005; Staba 2001; Weber 1992), and that there may be substantial differences in the release of allicin from different preparations (Lawson 2001). There may, therefore, be differences in the effects between preparations and this should be taken into account when evaluating studies of effectiveness. In vitro studies do not indicate clinical efficacy.

Why it is important to do this review

Systematic reviews of garlic for lowering cholesterol and minimising hypertension have been conducted (AHRQ 2000; Silagy 1994; Tapsell 2006). However, prior to this Cochrane Review, no systematic review of garlic for the common cold had been conducted.

Objectives

To determine whether garlic (Allium sativum) is effective for the prevention or treatment of the common cold, when compared to placebo, no treatment or other treatments.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs) comparing garlic with placebo, no treatment or standard treatment for the common cold.

Types of participants

Trials eligible for inclusion were those involving adults or children (0 to 17 years) who had no other acute illnesses or severe chronic condition. In terms of common cold prevention, 'cases' were those who developed a common cold during the course of the study. For treatment trials, participants were required to have a common cold or non‐specific viral upper respiratory tract infection (URTI). Symptoms that were used to identify the common cold could include coryza, sore throat, rhinitis, headache and general malaise. We excluded studies of influenza or those in which the illness definition included myalgia and fever greater than 38 °C, as these are common distinguishing features of influenza.

Types of interventions

Trials of garlic in any medicinal formulation were included. However, we only assessed trials where garlic was the single active ingredient. Garlic extracts were acceptable but not trials where raw, unprocessed garlic was the intervention.

Types of outcome measures

Primary outcomes

For prevention trials, the outcome of interest was the number of occurrences of the common cold.

For treatment trials, the primary outcome of interest was the duration of symptoms of the common cold.

Secondary outcomes

Secondary outcomes included the duration of symptoms of the common cold (number of days), the number of days 'challenged' (where participants reported an occasional sneeze or felt that a cold was coming on) and the number of days to recovery.

We considered reported adverse effects.

Search methods for identification of studies

Electronic searches

For this 2014 update we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (2014, Issue 7) (accessed 7 August 2014), which includes the Cochrane Acute Respiratory Infections Group Specialised Register, MEDLINE (October 2011 to July week 5, 2014), MEDLINE in‐process and other non‐indexed citations (searched 6 July 2014), EMBASE (December 2011 to August 2014) and AMED (2011 to 2014). Details of the previous search are in Appendix 1.

We used the search terms described in Appendix 1 to search MEDLINE and CENTRAL. We did not use a filter to identify randomised trials in MEDLINE as there were too few results. We adapted the search terms to search EMBASE (Appendix 2) and AMED (Appendix 3). There were no language or publication restrictions.

Searching other resources

We searched the following trials registries: World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) and ClinicalTrials.gov (searched 8 August 2014). We handsearched the references of all identified studies. Sarah Thorning (ARI Group Trials Search Co‐ordinator) and two review authors (AB, EL) carried out the search. We also contacted the manufacturers of garlic supplements, experts in the field and the Cochrane Complementary Medicines Field.

Data collection and analysis

Selection of studies

Two review authors (AB, EL) independently reviewed and selected trials from searches, assessed and rated study quality and extracted relevant data. We resolved disagreements through discussion and consensus. We contacted trial authors to request missing data or to clarify methods whenever possible.

Data extraction and management

We extracted data using a standardised form. Information included:

  • age and gender of participant;

  • number of participants;

  • whether analysis was by intention‐to‐treat (ITT);

  • randomisation method;

  • method of blinding;

  • blinding of outcome assessment;

  • smoking or non‐smoking status;

  • pre‐existing chronic conditions;

  • exclusion criteria;

  • diagnostic criteria;

  • treatment setting;

  • duration of treatment;

  • outcomes;

  • duration of illness;

  • functioning (for example, time to return to normal activity);

  • severity of illness;

  • occurrence of illness (prevention trials);

  • adverse effects; and

  • other medicines being used, including those with potential drug interactions.

Assessment of risk of bias in included studies

As with any systematic review, trials of poor quality may overestimate the treatment effect. We assessed the following aspects of trial quality:

  • quality of randomisation;

  • quality of blinding;

  • quality of allocation concealment;

  • presence of selective reporting;

  • presence of incomplete outcome data; and

  • analysis by intention‐to‐treat (ITT).

Specification of the dose and standardisation of the garlic extract are important in order to generalise but should not affect quality. We did not conduct a sensitivity analysis as only one study met our inclusion criteria.

The authors independently assessed 'Risk of bias' using the tool available in Review Manager 5.3 (RevMan 2014), recommended by Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). There were no disagreements in assessments.

Measures of treatment effect

We reported the results as continuous data in days of common cold experienced. We did not conduct a meta‐analysis as there was only one trial.

Unit of analysis issues

There were no included studies that utilised non‐standard designs.

Dealing with missing data

There were no missing data to deal with.

Assessment of heterogeneity

Heterogeneity was not an issue as only one study was included.

Assessment of reporting biases

Reporting bias is possible. It was not possible to use a funnel plot to investigate this as there was only one included study.

Data synthesis

We did not conduct a meta‐analysis as only one study met our inclusion criteria. There were insufficient data to conduct a meta‐analysis of adverse effects and these were collected using different methods. We considered adverse effects reported in both included and excluded trials.

Subgroup analysis and investigation of heterogeneity

We did not conduct subgroup analyses as there was only one included study.

Sensitivity analysis

The exclusion of studies from the review was clearly objective and non‐contentious. In addition, only one trial was included. Therefore, a sensitivity analysis was not required.

Summary of findings and assessment of the certainty of the evidence

Results

Description of studies

Results of the search

In our first publication of this review, Lissiman 2009, we searched the following electronic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (2009, Issue 1), which includes the Cochrane Acute Respiratory Infections Group Specialised Register: 0 search results; OLDMEDLINE (1950 to 1965) and MEDLINE (January 1966 to March Week 3, 2009): 27 search results; EMBASE (1974 to March 2009): 10 search results; and AMED (1985 to March 2009): two search results.

In the 2012 update, Lissiman 2012, we searched the following electronic databases: CENTRAL (2011, Issue 4): one result; MEDLINE (Ovid) (1 February 2009 to November week 3 2011): six results; EMBASE (1 March 2009 to December 2011): 31 results; AMED from 2009 to December 2011: 0 results. We excluded one new trial in this update (Yakoot 2011).

In this 2014 update, we searched the following electronic databases: CENTRAL (2014, Issue 7): three results; MEDLINE (Ovid) (October 2011 to July week 5, 2014): three results; MEDLINE in‐process and other non‐indexed citations (searched 6 July 2014): one result; EMBASE (December 2011 to August 2014): two results; AMED (2011 to 2014): 0 results; clinical trials registries WHO ICTRP and Clinicaltrials.gov (searched 8 August 2014): 0 results. We excluded two new trials in this updates (Nantz 2012; Polanco‐Rojas 2013).

Included studies

Of the eight trials identified as potentially relevant from our searches, we eventually excluded seven. However, we included information about adverse effects described in these studies as additional anecdotal reports.

Josling 2001 randomly assigned 146 participants to either garlic (one allicin‐containing garlic capsule (dose unspecified) per day with the main meal) or a placebo, for 12 weeks. Participants kept a diary and the primary outcome measure was the number of occurrences of the common cold measured by participants' self rating. Other outcomes included the cold duration (number of days), the number of days 'challenged' (where participants reported an occasional sneeze or felt that a cold was coming on) and the number of days to recovery.

Excluded studies

Andrianova 2003 (Russian) was a randomised controlled trial (RCT) comparing Allicor (slow‐release garlic tablets) to benzimidazole or placebo for treating acute respiratory disease (ARD) in children. The definition of ARD included influenza, thus excluding it from our review. The trial was conducted in two stages; the first stage was a five‐month, open non‐RCT, which investigated the tolerability of Allicor and its effects on ARD morbidity; the second stage was a five‐month, double‐blind RCT, which assessed effects on morbidity. In the first stage, 172 children aged 7 to 16 years were given Allicor and were compared to 468 controls; there was no difference in the prevalence of gastrointestinal side effects between the groups. In the second stage, 42 children aged 10 to 12 years were treated with Allicor, compared to 41 placebo‐treated children. Allicor reduced ARD morbidity 1.7‐fold compared to placebo.

Rafinski 1974 (Polish) assessed the clinical course of recurrent upper respiratory tract infections (URTI) in 49 children aged 2 to 15 years, following treatment with Alliofil, a coated garlic tablet. We excluded this study because there was no comparison group and it was a non‐randomised controlled trial. Before treatment, swabs were taken from the patients and sensitivity tests were conducted for Alliofil and several major antibiotics (penicillin, streptomycin, terramycin, erythromycin, aureomycin, tetracycline, neomycin and sulphonamides). From the 49 cases of recurrent URTI, bacteria were sensitive to the tested antibiotics in only nine children. The authors report that the bacterial species isolated from the remaining 40 children were sensitive to Alliofil.

Ushirotake 2004 (Japanese) was not a RCT and was thus excluded. The study assessed the number of occurrences of the common cold and the severity of symptoms in 272 volunteer participants at drugstores. One hundred and thirteen had been taking Kyoleopin (containing aged garlic extract) for more than one year, 41 had been taking Leopin‐5 (containing aged garlic extract) and 118 had not been taking either. As the study was not randomised or blinded, there is a high risk of bias. Of interest, this study has been used to support claims by a nutritional supplement company that garlic is effective in preventing the common cold and decreasing the severity of symptoms. This emphasises the need for careful consideration before accepting claims of scientific evidence.

Hiltunen 2007 compared a cellulose nasal spray with a combination cellulose and garlic extract nasal spray to prevent airborne respiratory infections, including cold‐like symptoms. We excluded the study because the study outcome did not meet the definition of the common cold defined in our protocol. Our protocol required that studies include either a placebo control group or a standard treatment group for comparison. This study did not meet this criterion as the cellulose could not be considered a standard treatment or a placebo.

Yakoot 2011 assessed the efficacy of a multiherbal formula (including garlic) in the treatment of the common cold. We excluded this RCT for two reasons. Firstly, garlic was combined with several other active ingredients. Secondly, the trial included participants with myalgia and fever, which did not meet our inclusion criteria.

Nantz 2012 is a RCT that compared aged garlic extract (AGE) powder (four capsules: 2.56 g per day) to placebo capsules for cold and influenza symptoms. As documented symptoms included myalgia and fever we excluded this trial from our review. We contacted the authors to assess whether we could sub‐analyse the data by excluding the participants with influenza. However, unfortunately this was not possible. The primary outcome of the trial focused on measurement of immune cell proliferation. On analysis of the secondary outcomes of cold and flu symptoms, the incidence of people who experienced at least one episode of a cold or influenza during the time period was not statistically different (28 of 56 people in the placebo group, and 26 of 56 in the AGE group) (P = 0.848). However, the group consuming the AGE powder appeared to have reduced severity of symptoms.

The total number of symptoms reported during the study was lower in the AGE group (584 versus 737, or 21% fewer) (P < 0.001). Per episode of illness, this finding was not significant (11.9 in the AGE group versus 14.0 in the placebo group) (P = 0.536).

Participants were asked to report whether or not they had a decrease in desire or ability to carry out their normal routine (decrease in activity or DIA). The total number of days of DIA during the study was significantly different between groups: 53 in the AGE group versus 126 in the placebo group, or 58% fewer (P < 0.001). A significant difference was also noted per episode of illness (61% fewer days of DIA in the AGE group, P < 0.001).

Polanco‐Rojas 2013 evaluated the effect of garlic drops on children with an acute respiratory illness. We excluded the study as it was not a RCT. That is, participants were selected for the treatment or placebo group based on their clinical symptoms. Symptom scores were compared within the same group from day three to five, not between the control and treatment groups.

Risk of bias in included studies

We conducted the assessment of study quality according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The included trial was of reasonable quality (Josling 2001). The trial author reported (in correspondence) that the study was analysed by intention‐to‐treat (ITT); that is, participant results were analysed according to the treatment group to which they were randomised, regardless of whether they completed the study or changed treatment. Not analysing by ITT can affect the validity of results.

The overall risk of bias is presented graphically in Figure 1 and summarised in Figure 2.


'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.


'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Participants were matched for age, sex and previous use of garlic, then randomised to the active or placebo group with the use of a random number generator. Adequate methods of allocation concealment were used; the trial author reported that the research co‐ordinator was given plain white bottles marked A or B and these were provided to the patient according to the randomisation codes.

Blinding

The research co‐ordinator was blinded for the duration of the trial. As the outcomes were self reported and participants were recruited through advertising, poor blinding of participants may have biased outcome reporting. Reasonable measures were taken to blind participants to the intervention; the investigators reported using foil wrapping to prevent the active treatment from being identified by its smell. However, four of the participants in the active group and one in placebo group noticed a 'smell' when burping. The trial author was responsible for breaking the randomisation codes at the end of the trial after all diaries had been returned.

Incomplete outcome data

Four participants withdrew from the study; three from the active group and one from the placebo group. Criteria for participant inclusion and exclusion were not reported.

Selective reporting

There was no evidence of selective reporting of outcomes. However, the statistical analysis and primary outcomes do not appear to have been decided in advance. The analysis used may therefore have been chosen post‐hoc to maximise the chances of finding a statistically significant result.

Other potential sources of bias

The trial author reported he had no conflict of interest at the time of the study.

Effects of interventions

The included trial reported 24 occurrences of the common cold in the garlic intervention group compared with 65 in the placebo group (P < 0.001). There were fewer days of illness in the garlic group compared with the placebo group (111 versus 366). A cold could be defined as 'feeling low and beginning to exhibit symptoms' or 'full cold symptoms' (headache, sneezing, runny nose, tiredness). Statistical significance was not reported for the number of days to recovery from an occurrence of the common cold (4.63 versus 5.63) but these appear similar. The trial authors reported that 16 participants taking placebo had more than one full‐blown cold compared to two participants taking garlic, but no statistical analysis was reported. Adverse effects included rash and odour.

Discussion

Evidence

Garlic may prevent occurrences of the common cold. The published evidence for this is positive but limited, as it comes from one relatively small trial with subjective outcome measures.

Only one trial that met the selection criteria could be identified (Josling 2001), limiting the conclusions that can be drawn. The trial reported significant differences in effect between the placebo and intervention groups. Adverse effects reported were relatively minor (smell and skin rash). It is not certain whether the single case of gout reported could be reasonably attributed to the garlic.

While the results suggest that garlic may have an effect on preventing the common cold, the subjective nature of the outcome measure means that this result is somewhat uncertain. The outcome of having a cold was not confirmed by any objective observation and may be unreliable. Further, a five‐point categorical scale was collapsed for analysis; hence a cold was defined as a score of either 2 ‐ 'feeling low and beginning to exhibit symptoms' or 1 ‐ 'full cold symptoms' (headache, sneezing, runny nose, tiredness). The trial authors do not state whether this analysis was defined in advance and it is possible this was done to increase the likelihood of achieving a statistically significant result, since 'full cold symptoms' would seem to be the clearest definition of a cold. Inclusion and exclusion criteria were not reported, nor were differences in co‐morbidity or concurrent illnesses. These factors reduce the ability to generalise from this trial to other situations and may have introduced bias into the results.

No trial was identified that looked at whether treating symptoms of the cold with garlic reduces their severity or duration. However, in the included study the number of days to recover from a cold was similar in both groups.

Adverse effects

Josling 2001 reported that one participant allocated to receive the garlic capsule withdrew due to development of gout and another due to pruritic rash below the knees, which faded after the garlic capsule was discontinued. Four participants in the intervention group and one in the placebo group noticed a 'smell' when burping. Adverse effects reported in excluded studies were also considered, acknowledging that any adverse events reported could not be attributed to garlic, because of weaknesses in randomisation or the lack of a control group. In the Andrianova 2003 trial, there were no gastrointestinal side effects observed but it is unknown whether there were any other adverse effects. Rafinski 1974 reported that no side effects were observed and Yakoot 2011 reported that the frequency of side effects did not differ significantly between intervention and placebo groups (no data reported). In Nantz 2012, there was no difference in the incidence of reported adverse effects (one person reported gastrointestinal side effects in each group). It is not known whether Ushirotake 2004 or Polanco‐Rojas 2013 reported adverse effects.

The safety of consuming small quantities of raw garlic is evident in its worldwide use as a culinary spice (WHO 1999). Adverse events associated with garlic have been reported in non‐randomised studies, randomised trials in other conditions and in case reports. A review of other adverse effects reported in the literature included bad breath and body odour and allergic reactions, manifesting in minor respiratory or skin symptoms (AHRQ 2000; WHO 1999). There is a potential for high‐dose garlic to interact with antithrombotic drugs (for example, warfarin), increasing the risk of bleeding, but the few reported case studies are inconclusive (AHRQ 2000; Fugh‐Berman 2000; WHO 1999).

Summary of main results

The trial reported fewer occurrences of the common cold in people who took the garlic for 12 weeks (24) compared with the placebo group (65) (P < 0.001). Statistical significance was not reported for the number of days to recovery from an occurrence of the common cold (4.63 versus 5.63).

While this single small trial had a positive finding, there was insufficient evidence to confirm an effect of garlic on the common cold. No significant harms were reported.

Overall completeness and applicability of evidence

The one included trial addressed the objectives of the review. However, the small size of the trial limits the ability of this review to address the review question adequately. We identified no treatment trials.

Quality of the evidence

We included only one small trial. This trial also had several methodological limitations, including in blinding (high risk) and selective reporting (unclear risk). Therefore, the available evidence does not allow any robust conclusions to be drawn.

Potential biases in the review process

We conducted thorough searches of the literature, including of several large databases and the references of relevant studies. As few results were obtained, the searches were not limited to randomised controlled trials (RCTs). However, it is possible that not all relevant data were obtained, if published only in abstract format or in another language. The decisions to include and exclude studies were clearly objective and non‐contentious.

Agreements and disagreements with other studies or reviews

There are no other systematic reviews assessing the efficacy of garlic for the common cold.

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figures and Tables -
Figure 1

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Figures and Tables -
Figure 2

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.