A Systematic Review of the Efficacy of Bioactive Compounds in Cardiovascular Disease: Carbohydrates, Active Lipids and Nitrogen Compounds

The prevalence of cardiovascular disease (CVD) is increasing by the day; it is the prime cause of death in all developed countries [1] and one of the most important for developing countries [2]. There are some risk factors that cannot be changed such as family history, ethnicity and age. However, detection and control of modifiable factors such as, blood pressure (BP), high cholesterol, obesity, type 2 diabetes (T2D) or unhealthy diets can help to prevent intermediate risk CVD processes like inflammation or oxidative stress. Thus, the primary prevention of CVD by identifying and treating at-risk individuals remains a major public-health priority. A healthy life style is the main preemptive approach [3,4].

Dietary habits are quite different around the world; nevertheless, certain consumption patterns are common worldwide like inclusion of fruits and vegetables or products like cocoa, coffee or condiments is a merging point. Bioactive compounds (BACs) are ‘extra nutritional' constituents that are present in small quantities in plant products and lipid-rich foods [5]. The growing body of scientific evidence indicates that certain BACs play a beneficial role in CVD prevention [5,6]. BACs oral supplements taken along with the usual diet can increase the intake of ingredients reputed to have clinical benefit. These supplements are, usually, an addition to a healthy diet, and not a substitution to conventional food or the sole item of a meal [7].

Putative beneficial biological effects such as antilipidemic, antihypertensive, anti-glycaemic, antithrombotic and anti-atherogenic effects are attributed to BACs. In this study, the main goal was to examine the scientific evidence of the role of BACs in the prevention and treatment of CVD by a systematized review of randomized clinical trials (RCTs). The vegetal BACs considered in this review were those derived from carbohydrates, active lipids and organosulfur compounds.

Epidemiologic studies have suggested that dietary fibres such as β-glucan, a non-starch viscous polysaccharide derived from oat, are known to exhibit hypercholesterolemic effects, furthermore protecting against CHD [8].

Concerning active lipids, we can find that the conjugated linoleic acids (CLAs) are a group of minor fatty acids from ruminant origin, which has long been recognized as a body fat-lowering agent. Dairy food products are rich in CLA, which is considered an important component due to its potential health benefits such as anti-carcinogenic, anti-atherogenic, antidiabetic and anti-adipogenic properties [9]. Additionally, the terpenes compounds are constituted by monoterpenes, which are naturally occurring isoprenoids found in the essential oils of citrus fruits, cherries, mint, and herbs [10]. Diterpenoids form a large class of plant-derived secondary metabolites that possess a wide spectrum of important biological activities; an example is the extract from seeds and bark of horse chestnut (Aesculus hippocastanum L.) that is used as a herbal medicine to treat chronic venous insufficiency [11]. Moreover, plant sterols (or phytosterols) are natural constituents of cell membrane of plants. Their role in plants is similar to that of cholesterol in humans [12]. It has been known since the 1950s, that plant sterols inhibit cholesterol absorption, and since the 1980s that plant stanols inhibit cholesterol absorption [13].

Other compounds included are those derived from the Allium sativum, commonly known as garlic, which is a plant of the family Alliaceae and is widely consumed as a spice in Korea. Garlic contains high levels of organosulfur compounds and flavonoids as well as a variety of compounds that function synergistically to provide various health benefits [14].

Along with tea, coffee is the most common hot drink prevalent in many cultures. Caffeine (1,3,7-trimethylxanthine), a purine alkaloid contained in the seeds of the coffee plant, represents the most abundant component in coffee, and its potential effects on health still represent a controversial issue.

We thoroughly examined the effects of BACs on BP, lipids profile (triacylglycerol, cholesterol, high and low density lipoproteins), carbohydrate (CHO) metabolism (glucose, insulin, and insulin resistance), oxidative stress, inflammation and endothelial function (EF). Furthermore, we gave a recommendation for its consumption based on the evidence grade according to Scottish Intercollegiate Guidelines Network (SIGN) [15].

We developed a literature search in Medline by PubMed (U.S. National Library of Medicine and the NIH), just as in LILACS and EMBASE, including publications in English, Spanish and Portuguese until December 2014. Studies eligible for this review included randomized controlled trials (RCTs) in healthy and unhealthy adults, with prospective, parallel or crossover designs, with full text, and whose primary outcomes were vascular homeostasis, BP, oxidative stress and/or inflammatory biomarkers; we excluded those studies with a cohort, ecological or case-control design, which analyzed whether a drug or BACs were combined with other compounds. However, there was no restriction on the publication type or sample size.

Due to the diversity of their chemical structures (fig. 1), BACs can present different effects in CVD; consequently, in this section, we included all those related to carbohydrates, active lipids and nitrogen compounds. We included different keywords for the search, which are considered bioactive compounds, including active lipids (linoleic acids, conjugated (CLAs), fatty acids, hydro fatty acids, fatty acids esters, alpha-pinene) and carbohydrates (beta-glucans); terpenes (monoterpenes: iridoids, menthol, norbornanes, pyrethrins, thymol; diterpenes: aonitine, aphidicolin, atractyloside, colforsin, abietane, clerodane, kaurane, ginkgolides, phorbols, phytanic acid, phytol, ryanodine, taxoids; triterpenes: cucurbitacins, lanosterol, limonins, pentacyclic triterpenes, quassins, sapogenins; polyterpenoids: phytosterols, ellagic acid, squalene); nitrogen compounds (caffeine, allium, glucosinolates, xanthine, pseudoephedrine, theobromine, theophylline, ephedra); alkaloids (betalains, indole alkaloids, physostigmine, secologanintryptamine alkaloids, staurosporine, peptide alkaloids). We combined the MeSH term ‘cardiovascular diseases' with each bioactive compound as a MeSH major topic, together with NOT ‘review' [Publication Type] in PubMed. However, equations in Spanish language were used when the search was carried out in LILACS, that is, (tw:(ácido linoleico)) AND (tw:(enfermedad cardiovascular)) AND NOT (tw:(revisión)). When consulting the database EMBASE, the equations were elaborated as ‘catechols'/mj AND ‘cardiovascular diseases'/mj NOT ‘review'. Articles published before 1990 were discarded because they were not relevant any more.

First, both titles and abstracts were identified independently by two reviewers, for discarding those that did not fit with the language, date, subject matter, design and outcomes established. Then, full-text publications were classified based on pathologies according to outcomes analyzed in each study.

Moreover, RCTs were finally selected if they obtained a score between 3 and 5 according to the JADAD scale [16]. Such method pretends to minimize bias for RCTs, ensuring certain quality in the evidence; it took into account if they were randomized, blind and provided detailed information about patients.

Furthermore, we verified the quality of selected articles by the Scottish Intercollegiate Guidelines Network (SIGN) [15]. Diverse quality levels of scientific evidence are established according to two features: the design and bias risk of a study. The levels are from 1++, when the information is considered high quality, to 4 when the information is considered very low quality. Signs are used to reporting with reference to compliance degree of key criteria associated with potential bias (1++, 1+, 1-, 2++, 2+, 2-, 3, and 4). Additionally, we included a grade of recommendation, based on the evidence strength of the antecedents, whose levels are A, B, C, D, ‘A' being highly recommended and ‘D' not recommended. These grades of recommendation by SIGN guidelines are equivalent to those designated by FAO/ WHO, as evidence criteria: convincing, probable, possible and insufficient [17].

In total, 760 RCTs were found using the equations proposed in the different databases (EMBASE, LILACS and PubMed). We exclude 715 to obvious irrelevance leaving in 45 full papers to read (fig. 2). Finally, after reading the full paper and evaluating it using the JADAD scale, we selected 26 articles for the final review and they are included in tables 1 and 2.

Three articles of carbohydrates were selected (table 1), all related to β-glucan polysaccharide. The quality scores were between 3 and 4 according to the JADAD scale. Subjects were between 26 and 75 years, with a sample size of 18 and 97 participants [8,18]. Cugnet-Anceau et al. [8] evaluated the intake of 3.5 g/day of β-glucans in soup, in T2D patients (BMI of 20-35 kg/m²) for 2 months. Furthermore, Maki et al. [18] assessed the effects of consuming oat β-glucans between 2.1 and 3.2 g/day for breakfast in hypertensive patients during 3 months. Keogh et al. [19] assessed supplements of β-glucan incorporated daily into snacks and meals in mildly hypercholesterolemic men. The dose of both treatment and control was scaled to body size for all subjects (8.1-11.9 g/day of β-glucan). The main outcomes of these interventions were BP, lipids profile, CHO metabolism, and oxidative stress biomarkers. Cugnet-Anceau et al. [8] observed a decrease in triacylglycerols (TAG); besides, Maki et al. [18] demonstrated a decrease of mean peak insulin and of incremental area under the insulin curve in the β-glucans group, and an improvement of BP in BMI >31.5 kg/m² T2D patients. However, any significant effect was reported in mildly hypercholesterolemic men by Keogh et al. [19].

Four articles referenced to CLAs were selected (table 1). The quality scores obtained varied from 3 to 5 within the JADAD scale. Only one of those studies was conducted among healthy subjects, with being male as the inclusion criterion (19-35 years) [20]. The subjects recruited in the other two studies were BMI >25 kg/m², one of them with LDL phenotype B as requirement for participating, both between 35 and 65 years [21,22]. One article assessed CLA's effect in active rheumatoid arthritis patients [23]. Sample sizes were from 40 to 92. The administered doses ranged from 2-5.5 g/day in butter or capsules as oil, during periods from 5 weeks to 3 months. The main outcomes in these studies were related to BP, lipids profile, CHO metabolism, endothelial function, inflammatory and oxidative stress biomarkers, as well as leptin, adiponectin. Naumann et al. [22] did not observe significant changes in any parameters analyzed (table 1). Moreover, there were no significant changes in plasma lipids profile. In addition, Taylor et al. [21] observed an impairment of endothelial function and increase in markers of oxidative stress, since plasma 8-iso-PGF2α increased after 4.5 g/day of CLA administration. In addition, Raff et al. [20] observed higher urinary excretion of 8-iso-PGF2α, increasing lipids peroxidation. Regarding other parameters, only BP decreased with 2 mg/day CLAs supplementation, along with erythrocyte sedimentation rate (ESR) when CLAs were administrated in conjunction with vitamin E (400 mg/day) [23].

Three articles were reviewed and are included in table 2. Quality score allocated according to the JADAD scale was around 3-5. Only one intervention was long-term intervention, whereas two of them focused on acute intervention. The main outcomes of the interventions regarding norbornanes were BP and HR (table 2). One study developed two interventions with 57 orthostatic hypotensive patients >50 years [24]: 75 drops/day Korodins (Herz-Kreislauf-Tropfen®) for 7 days, and a single dose of 25 Korodins drops evaluated at 1, 3, 5, 8 and 10 min. Significant changes were observed in the first intervention; thus, orthostatic hypotensive decreased. All other studies were a single dose of 25 Korodins drops in hypotensive women [25] and a single dose of 5, 20, 80 Korodin drops in orthostatic hypotensive patients (18-40 years) in both sex [26], measured at 1-5 min (1 drop = 1 mg D-camphor and 38.62 mg crataegus berry extract). In both interventions, BP improved after Korodins acute administration and BP increased almost immediately after its application in hypotensive women, while orthostatic hypotension decreased at all times with 80 drops according to Belz's [26] study.

The three articles selected investigated their actions in hypotensive subjects, two of them in orthostatic hypotensive [24,25,26]. The combination of natural D-camphor and an extract of fresh crataegus berries (Korodin Herz-Kreislauf-Tropfen®) with a single dose of 25 drops increased BP almost immediately after application [25]; as the same pattern repeated with a dose of 80 drops after 1, 3 and 5 min [26]. After one week, 75 drops/day also increase BP [24] in hypotensive subjects.

Four papers were included and the quality measured according to the JADAD scale ranged between 3 and 5 scores (table 2). Almost all studies were developed in HT patients [27,28,29], except one in which participants suffered type 1 and type 2 diabetes with normal BP or low BP [30]. Ages varied from 20 to 75 and sample size around 40-174 subjects. The dosage interval was 300-1,500 mg/day of steviol glycoside stevioside from three months to two years. The main outcome in all studies was increased BP, but was significantly decreased only in two studies. Other parameters when analyzed (lipids profile, CHO metabolism, as well as left ventricular mass index (LVMI)) did not show significant changes.

One trial was chosen (table 2) and evaluated with a score of 3 according to the JADAD scale. Ottillinger and Greeske [31] separated the trial into two interventions. Both assessed the effect of 2 capsules/day HCSE (horse-chestnut seeds extract) (standardized to 100 mg aescin (Venostasin®)), in leg volume among 240 early chronic venous insufficiency (CVI) and 355 advanced CVI patients, for 12 and 16 weeks, respectively. The treatment caused a significant decline in leg volume in both interventions.

Two RCTs related to polyterpenoids were finally selected (table 2), since its score as per JADAD level was from 3 to 4. The trials were based on 17-104 volunteers in an age range 20-70. Allen et al. [32] carried out an intervention of 22 g/day cocoa flavanols with plant sterols (2.2 g sterol esters) provided with dark chocolate bars in hypercholesterolemic patients for a period of 8 weeks. Earnest et al. [33] examined the efficacy of ingesting 2.6 g/day of encapsulated phytosterol esters in hypercholesterolemic subjects for 12 weeks. Lipids profile was considered a major outcome by the two investigations. Among them, TC and LDLc decreased in both cases. With regard to BP, only Allen et al. [32] considered this aspect, resulting in the reduction of SBP. Concerning CVD risk biomarkers, changes were not observed.

Caffeine

Three articles were finally selected according to our criteria (table 3), as well as based on the JADAD scale, whose score varied between 3 and 4. Lane et al. [34] investigated the effects on 47 healthy habitual coffee drinkers of 500 mg/day of caffeine on ambulatory BP and HR for three days. After three days of caffeine administration, BP increased, but average HR was found to be decreased. Furthermore, Shechter et al. [35] assessed the effect of acute caffeine ingestion (200 mg) on endothelial function, nitro-glycerine-mediated dilation (NTG) and high sensitivity hsCRP, in 80 subjects with/without coronary artery disease (CAD) [35] after one hour of administration. The endothelial function was increased; also, hsCRP decreased both in with/without CAD patients. Duffy et al. [36] also evaluated one dose of caffeine (200 mg) in CAD patients and SBP increased after 2 hours. Moreover, the same subjects took 270 mg caffeine contained in black tea daily for 4 weeks. In both interventions, an improvement in the endothelial function was observed, but without significant effects in relation to other parameters analyzed (table 3).

Eight articles were definitely selected due to its score in JADAD scale, hovering between the values 4 and 5 (table 3). The ages of participants were between 35 and 82 and sample size ranged from 15-84 volunteers. One intervention was developed in 34 healthy subjects, where the efficacy of Aged Garlic Extract (AGE) was assessed at doses of 2.4-7.2 g/day for 44 weeks [37]. The main outcomes were inflammatory biomarkers, aggregation and adhesion platelet. The highest AGE supplementation increased threshold concentrations for epinephrine and collagen, fibrinogen and increased slightly threshold level of adenosine diphosphate-induced aggregation, platelet adhesion to collagen, adhesion to von Willebrand factor. Moreover, adherence to fibrinogen was potently inhibited at all levels of supplementation. Three clinical trials were related to coronary artery disease (CAD) patients: the effect of 4 ml/day of AGE was evaluated in 23 subjects [38], while Sobenin et al. [39] investigated the effect of allicor (garlic powder) in 51 subjects, both during a period of one year. However, Williams et al. [40] assessed 2.4 g/day of AGE only during two weeks. The main outcomes of these studies consisted of BP, lipids profile, endothelial function, oxidative stress, inflammatory biomarkers, and coronary artery calcification (CAC), platelet aggregation and adhesion just as changes in 10 years prognostic risk of myocardial infarction and sudden death. Within lipids profile measured by the two interventions [38,39,40], LDLc and TC parameters were significantly decreased after allicor treatments [39]. Reduction of CAC progression was detected in the study that assessed it [38] (table 3). Sobenin et al. [39] observed a decrease in multifunctional prognostic risk of myocardial infarction and sudden death prevents the rise of risk with aging. Williams et al. [40] only reported significant changes in the endothelial function. Ried et al. [41] also assessed the effect of AGE in capsules (960 mg/day containing 2.4 mg S-allylcysteine), as well as of different doses (240/480/960 mg/day (0.6/1.2/2.4 mg/day of S-allylcysteine)) [42] as an adjuvant to antihypertensive medication in patients with uncontrolled hypertension, both during 12 weeks. Sobenin and colleagues [43] separated the trial into two interventions. The first one included treatment that consisted of 600 mg/day of allicor (300 mg garlic), being a randomized design, whereas the other intervention was an open-label branch, where patients received higher doses of 2,400 mg/day, both patients with mild or moderate HT for 8 weeks [43]. Both authors studied the efficacy in the improvement of BP, and SBP was significantly reduced by 1.2 mg/day of S-allylcysteine [42]. Further, the effects of Bulbus allii macrostemi were tested on UA/NSTEMI patients, in doses of 6 g/day plus baseline therapy for eight weeks [44]. Markers such as ox-LDL, PAI-1 and activity of PLG were significantly improved (table 3).

Maki et al. [18] assessed the effects of oat β-glucans in hypertensive subjects and observed significant changes in CHO metabolism and BP (table 1), the latter being only in subjects with BMI >31.5 kg/m². Cugnet-Anceau et al. [8] observed that a 3.5 g dose for 2 months of β-glucans reduced TAG in T2D subjects. Nevertheless, other plasma lipids, as well as TAC, CHO metabolism and CRP were not significantly altered. They concluded that to improve the metabolic profile of T2D subjects in the long term, the quantity, the food vectors and the tolerability of β-glucan products might be reevaluated. Moreover, Keogh et al. [19] reported no clinically significant improvements in lipids profile or CHO metabolism; thus, they concluded that there was no evidence of improvement in CVD or T2D risk in mildly hypercholesterolemic men. With regard to previous bibliography, oat β-glucans have been used in an open-label clinical trial to reduce glucose, showing that oat β-glucans lowered postprandial glycaemia [45]. These results seem to be promising, but most antecedents were found in animal models. It is possible that subjects took a low dose in both studies, but it seems that an increase in the concentration of β-glucan negatively affects the palatability. In fact, Maki et al. [18] justified greater dropout more likely related to palatability than toleration of the study products. Chen and Raymond [46] concluded that no pure β-glucan has been used until the moment and thus, it is difficult to identify which components are effective. In the present review, authors did not use a pure extract, and consequently, clinical trials should be designed taking into account this reflection. Whether there are not systematic review and/or meta-analysis related to the effect of β-glucan along with the studies selected presented here that do not show enough evidence, we cannot recommend β-glucan in CVD.

According to the results obtained to active lipids, concretely CLA, along with previous reviews [47,48], evidence is scarce in relation of lipids profile and on inflammatory processes, as well as CHO metabolism. We must take into account that the background is mainly based on animals and cells studies. Recently, Brouwer et al. [47] reviewed the effect of CLA and concluded that it increases LDL cholesterol, plasma LDL/HDL and TC/HDL ratios. However, RCTs selected following our criteria, did not show any effect with any doses, both in healthy and unhealthy subjects. Aryaeian et al. [23] concluded that a supplementation of CLAs plus vitamin E might be profitable for heart disease prevention in RA patients. Furthermore, BP decreased after 3 months of CLAs administration [23]. Even so, there are not enough preceding publications and grade of evidence for making a recommendation in these patients. Furthermore, Taylor et al. [21], reported that CLA administration in men with BMI >27 kg/m² degraded endothelial function, apart from increasing plasma 8-iso-PGF2α. Additionally, Raff et al. [20] found an increase in urinary excretion of 8-iso-PGF2α, resulting in lipid peroxidation increased a relevant risk factor for atherosclerosis. Given the controversy and scarce evidence, further investigation is required, but it is important to consider the deleterious effect that seemed to be present in lipid peroxidation [20]. Thus, we cannot recommend CLAs against CVD.

Despite these positive results, there is no meta-analysis to support its effect. Nevertheless, it could be interesting to continue testing this product in combination with other BACs in patients with hypotension, both in acute and chronic interventions. In the future, Korodin may be applied for controlling hypotension.

Many reports have shown that several classes of diterpenoids exert significant cardiovascular effects in animal models [49,50,51]. BP decreased in two studies, while plasma lipids were not different when compared to the placebo group in any case. In the two trials with any significant change, Kaurane was assessed for short-term [27,30]. Although Chan et al. [29], found changes since the third month evaluating the same doses than Barriocanal et al. [30], the sample size and/or subjects' characteristics may be relevant. Only one intervention considered LVMI, from which changes were not found after intervention [28]. Because of all this, without evidence in relation to Kaurane effects in humans, recommendations may not be valid, requiring more evidence and solid results.

Leg volume was the main outcome in the two interventions that were part of study, as visible sign of CVI. In both, leg volume decreased after oral HCSE in early and advanced CVI patients. We can compare such results with antecedents, since in 2012, Pittler et al. [11] published in the Cochrane Library a systematic review; its aim was to assess the evidence for HCSE as treatment of CVI with similar results in regards to reduction of mean leg volume. The adverse events reported in the reviewed trials were mild and infrequent. Thus, authors suggested that HCSE could be an effective short-term treatment option for CVI. Hence, this herbal extract may be recommended, but more rigorous RCTs are required, especially for the long term.

Both studies selected in this review, evaluated lipid profile changes after plant sterols administration (table 2), with positive results [32,33]. Moreover, BP improved in the study that analyzed it [32]. Furthermore, two meta-analyses evaluated the effect of plant sterols and stanols, the first versus policosanol in hyperlipidaemia [52], concluding that the last one is a valuable addition to the usual therapeutic lifestyle changes and may be an attractive alternative to conventional lipid-lowering drugs, due to offers advantages of greater efficacy. The second study tried to minimize the potential bias due to uncontrolled confounders for guaranteeing that the calculated mean differences do not provide unbiased estimates of the effect of sterols on CVD [53]. Taking into account the rigorous analysis and their conclusions, together with our scarce evidence, sterol plants do not appear to have efficacy as a source to lower CVD risk.

Caffeine

Lane et al. [34] tested 500 mg/day of caffeine vs. placebo control for three days. Healthy, non-smoking, habitual coffee drinkers' subjects recruited and authors observed caffeine significantly raised average ambulatory BP during the workday and evening and reduced average HR. These results suggested that caffeine consumption may exaggerate the deleterious effects of stress in daily life and aggravate the damage to health that stress can cause. These results also demonstrated that this modest caffeine dose produced an amplification of the effects of stress on BP and HR during the day. The possible relation between caffeine and cardiac events is well known. Moreover, SBP also increased after two hour of caffeine administration in CAD patients in a trial carried out by Duffy et al. [36]. Particularly, BP is a strong risk factor for heart failure. Experimental studies have consistently shown that coffee and caffeine are associated with acutely raised BP [54]. Sixteen randomized controlled trials were collected in a meta-analysis, comprising 1,010 subjects, yielded an overall BP increase of 2.0/0.7 mm Hg for regular coffee or caffeine intake. Nevertheless, it is very important to note that 9 out of 11 coffee trials were not blinded and this may have influenced the observed effects on BP. In contrast, two studies enrolled by our criteria, were double blind, as absolute requirement (according to JADAD scale), since caffeine was administrated in capsules as placebo did. Shechter et al. [35], demonstrated that 200 mg of caffeine administered as a single dose improved brachial endothelial function in controls and patients with CAD. Duffy et al. [36] also observed an improvement in the endothelial function with the same dose and kind of patients, both after two hours and four weeks. Additionally, increased serum caffeine was associated with a decrease of inflammatory markers such as hsCRP in both with/without CAD, and an increase in serum adiponectin patients with CAD. Other meta-analyses indicate that there is a J-shaped relationship between coffee consumption and heart failure incidence, with a modest inverse association with moderate consumption [55]. Compared with no consumption, the strongest inverse association was seen for 4 servings/day and a potentially higher risk at higher levels of consumption. It is clear that caffeine should be administered in purified capsules to prevent influence from other ingredients. In light of these findings, further evidence is required for recommending clearly caffeine administrated in capsules as moderate protection against heart failure incidence; this also depends on the kind of patients.

Studies in animal models suggest that garlic has potential antilipidemic, antihypertensive, anti-glycaemic, antithrombotic and anti-atherogenic properties [56]. Nevertheless, evidence is scarce in relation to clinical trials. The seven trials collected in our review offered different outcomes. Gallic acid decreased BP, especially SBP in hypertensive patients [41,42,43]. Nevertheless, BP was also measured in CAD patients without hypertension and changes were not observed [39]. Ried et al. [57] developed a systematic review; and 11 articles were suitable for meta-analysis. Such rigorous bibliography review concluded that garlic preparations are superior to placebo in reducing BP in individuals with hypertension. Additionally, another meta-analysis developed by Reinhart et al. [58] associated gallic acid with BP reductions in patients with an elevated SBP although not in those without elevated SBP. Using the SIGN guidelines, both works obtained 1+ by level of evidence. Consequently, we could establish recommendations for allium as far as effect on BP in hypertensive subjects. On the other hand, both authors emphasize that it is required to investigate the effect in long-term trials. Stabler et al. [59] concluded that garlic reduces mean supine BP versus placebo. However, they identified that the confidence intervals for these effect estimates are not precise by means of meta-analysis; thus, there is difficulty in recommending its use. We should take into account that Stabler et al. [59] studies, were valuated with 1+ and they did not suggest the recommendation of allium due to insufficient evidence. Simons et al. [60] evaluated 32 studies by meta-analyses and from their critical point of view, the methodological quality of the studies was poor. The meta-analysis was very rigorous, obtaining 1++ score according to SIGN guidelines and level of evidence is considered high. In their view, the use of garlic cannot be recommended as antihypertensive advice for hypertensive patients in daily practice [60].

With regard to the lipids profile, a decrease in TC and LDLc after [61] consuming garlic powder [39] for one year was observed in CAD patients. Although, 4 ml/day of AGE not caused any changes in plasma lipids, only in CAC progression rate [38]. Zeng et al. [62] elaborated a meta-analysis and concluded that garlic was superior to placebo in reducing TC and TAG concentrations and concluded that garlic therapy should benefit patients with risk of CAD according. Nevertheless, other serum lipid parameters such as LDLc, HDLc, Apo-B and TC/HDLc were not significantly affected. The level of evidence through SIGN guidelines resulted in 1++, so that allium could be recommended in subjects with cardiovascular risk with a grade of recommendation of A. Based on these facts, allium (gallic acid, especially) is beneficial for CAD prevention. Furthermore, there is a great quality meta-analysis focused on BP and characterized by its rigorous methodology [60]. Additionally, Sobenin et al. [39] also observed a decrease of multifunctional prognostic risk of myocardial infarction and sudden death prevents the rise of risk with aging; the most significant effects were observed in men. We must also take into account the decrease of the CAC progression rate [38], which is an excellent marker of overall coronary atherosclerotic burden correlating with both the total plaque burden and the presence of obstructive disease [63]. With reference to Williams et al. [40], we must consider that the evaluation of AGE was only during two weeks, and this period is short to consider the changes. Even so, endothelial function improved after such intervention.

We must mention the changes in UA/NSTEMI using Bulbus allii macrostemi, which decreased ox-LDL, PAI-1, apart from to increase the activity of plasminogen [44]. Results are promising in relation to Allium sativum in the CVD prevention, but further evidence in long-term is needed for verifying such facts, especially Bulbus allii macrostemi alone, without baseline therapy.

Certain limitations need to be considered. First, MeSH terms are not often used by researchers. Such specific terms must be taken into account when articles are drafted, assuring a good indexation and more visibility, facilitating the evidence valuation. Second, the application of resources such as CONSORT (Consolidated Standards of Reporting Trials) statement or JADAD scale is highly scarce. Moreover, the existence of checklists helps authors and editors to improve the reporting of RCTs and consequently provides scientific quality in data report. We consider that a clinical trial is reliable when it is at least randomized and blinded. Then, trials included in this review have a high level of heterogeneity, making it more difficult to draw concrete conclusions in relation to the kind of subjects, format of analyzed product or its combination with other compounds.

The role of BACs as adjuvants in CVD is increasing and validation of its effects is essential. Evidence shows that BAC such as horse chestnut seed extract, sterol plants and allium derivatives are helpful in decreasing risk factors of CVD. However, it is necessary to develop better quality RCTs (crossover design, double-blinded, long term, placebo/controlled) and furthermore, elaborate rigorous meta-analysis of existence evidence to support the effect of BAC on prevention and treatment of CVD.

The authors are grateful for the support of Mrs. Mercedes Rodriguez del Castillo Martin in developing some technical assistances with the search in databases.

B.P.-V., O.D.R.-H., C.M.A., A.G. contributed to the planning of the search of the literature, designed the analysis and results presentation and created the tool for assessing the quality of the articles. B.P.-V., O.D.R.-H. were involved in the analyses of the articles. B.P.-V., O.D.R.-H. wrote the draft. All authors discussed and revised all drafts and approved the final manuscript.

There are no conflicts of interest to report.

B.P.-V. and O.D.R.-H. contributed equally to the present work.