Beneficial effects of legumes on parameters of the metabolic syndrome: a systematic review of trials in animal models

Abstract

Legume consumption plays a pivotal role in the prevention and treatment of the metabolic syndrome (MetS). This systematic review aimed to highlight the beneficial effects of legume interventions for the prevention and/or improvement of parameters related to the MetS and the implicated metabolic pathways so far reported. The methodology involved a search in four electronic databases (Medline, Web of Science, Scopus, Cochrane Library) from January 2007 to December 2014, considering as descriptors ‘Metabolic Syndrome’ and ‘Fabaceae’ and adequately adjusting the equation in each one of them. In total, forty-one studies were finally included. The majority of the studies described a regulating effect on glucose and lipid metabolism due to legume administration, whereas effects on blood pressure and renal parameters are not fully described. Regarding the metabolic pathways involved, they include the up-regulation of genes related to β-oxidation and acetyl-CoA degradation and the down-regulation of glycolytic and lipogenesis genes, as well as those associated with the acetyl-CoA synthesis. The ameliorating effects of legume consumption on the alterations associated with the MetS are clearly reported and coincide with changes in the expression of protein and genes involved in lipid and glucose metabolism. More research needs to be conducted including more legume species that are highly consumed as part of a healthy dietary pattern.

The metabolic syndrome (MetS) represents a clustering of several metabolic disorders among which central obesity and insulin resistance are considered as causative factors^{( 1 , 2 )}, affecting one-quarter of the world’s adult population^{( 3 )}. The initial concept of ‘Syndrome X’ was described by Reaven^{( 4 )}, whereas the most recent diagnostic criteria, as established by the International Diabetes Federation in 2005^{( 5 )}, include obesity (waist circumference≥102 cm in men or ≥88 cm in women), dyslipidaemia (TAG≥150 mg/dl, HDL<40 mg/dl in men or <50 mg/dl in women), hypertension (≥130 mmHg systolic or ≥85 mmHg diastolic) and alterations of glucose metabolism (>100 mg/dl; includes diabetes)^{( 6 )}. Although the diagnostic criteria seem to be clear enough, the mechanisms underlying its pathology are not fully understood.

Preventing the development of the MetS requires a multidisciplinary approach, whereas the first step on the treatment of this pathology is focused on the amelioration of the related metabolic alterations and includes mostly lifestyle modifications^{( 7 )}. Nevertheless, in case such modifications prove to be inadequate, the next movement includes the prescription of appropriate pharmacological agents^{( 8 )}. Among lifestyle strategies, low-fat/low-glycaemic-index diets and regular physical exercise are encountered^{( 7 )}. For this reason, legumes have gained increasing interest given that their frequent consumption can help in the control of lipid homoeostasis and, consequently, reduce the risk of CVD. In addition, their consumption is associated with a better glycaemic control in diabetic patients and has exhibited hypolipidaemic effects by reducing the absorption of cholesterol. Their contribution to weight management because of their beneficial effect on appetite-regulating hormones and satiety has also been demonstrated^{( 9 , 10 )}.

The bioactive compounds that legumes contain such as resistant starch, α galactoside oligosaccharides, phytate, polyphenols and saponins may act as potential physiological modulators of metabolism, given that they inhibit the activity of angiotensin-converting enzyme and exhibit prebiotic effects, as well as antioxidant and bile acid-binding properties^{( 11 , 12 )}, thus showing promising potential as functional ingredients.

Taking into account that the actual lifestyle is at the same time leading to the increase of the prevalence of risk factors that induce the MetS and the undervalued consumption of legume foodstuff, as reflected by epidemiological nutritional surveys, there is a clear need to reinforce lifestyle strategies in order to better prevent the development of the MetS. The present review aimed at gathering the outcomes of recent intervention studies by putting together the beneficial effects that the consumption of different legumes exert on different alterations associated with the MetS.

Methods

Study eligibility

Considering that the aim of the present review was to collect the most recent and representative data for the effects of the legumes on the MetS, we performed a bibliometric analysis in the field of nutrition, which established the period of 7 years as the obsolescence period of the results of these studies^{( 13 )}. This period assured that more than half of the actual scientific production would be included (Burton–Kebler index: obsolescence according to median age/median production)^{( 14 )}. Therefore, the cut-off point for the publication date was established from January 2007 to December 2014. Although the present review focused on collecting data of animal trials, no filters were used at this point in order to prevent losing any entry not properly registered. Therefore, further exclusion of the entries was performed manually.

Thus far, the eligibility of the publications was confirmed by fulfilling the following inclusion criteria:

1. – The research articles should be recent intervention studies published after the year 2007, in which consumption of legume or administration of the legume-derived product was tested against different alterations related to the MetS.

2. – The research articles should be published in peer review journals, and the ones with complete text access were selected.

Data sources

A comprehensive and systematic review of literature was conducted using four electronic databases: MedLars Online International Literature, via PubMed^©, Web of Science, SCOPUS and the Cochrane Library Plus. The first step included the definition of the search terms through the use of Medical Subject Headings (MeSH) and considering as descriptors ‘Metabolic Syndrome’ and ‘Fabaceae’, in all the possible forms used by the indexed publications in PubMed. The final equation was (‘Metabolic Syndrome X’[Mesh] OR ‘Metabolic Syndrome X’[Title/Abstract] OR ‘Metabolic Syndrome’[Title/Abstract] OR ‘Insulin Resistance Syndrome X’[Title/Abstract] OR ‘Syndrome X, Metabolic’[Title/Abstract] OR ‘Syndrome X, Insulin Resistance’[Title/Abstract] OR ‘Metabolic X Syndrome’[Title/Abstract] OR ‘Syndrome, Metabolic X’[Title/Abstract] OR ‘X Syndrome, Metabolic’[Title/Abstract] OR ‘Dysmetabolic Syndrome X’[Title/Abstract] OR ‘Syndrome X, Dysmetabolic’[Title/Abstract] OR ‘Reaven Syndrome X’[Title/Abstract] OR ‘Syndrome X, Reaven’[Title/Abstract] OR ‘Metabolic Cardiovascular Syndrome’[Title/Abstract] OR ‘Cardiovascular Syndrome, Metabolic’[Title/Abstract] OR ‘Cardiovascular Syndromes, Metabolic’[Title/Abstract] OR ‘Syndrome, Metabolic Cardiovascular’[Title/Abstract]) AND (‘Fabaceae’[Mesh] OR ‘Leguminosae’[Title/Abstract] OR ‘Legume’[Title/Abstract] OR ‘Legumes’[Title/Abstract] OR ‘Beans’[Title/Abstract] OR ‘Amorpha’[Title/Abstract] OR ‘Andira’[Title/Abstract] OR ‘Baptisia’[Title/Abstract] OR ‘Callerya’[Title/Abstract] OR ‘Ceratonia’[Title/Abstract] OR ‘Clathrotropis’[Title/Abstract] OR ‘Colophospermum’[Title/Abstract] OR ‘Copaifera’[Title/Abstract] OR ‘Delonix’[Title/Abstract] OR ‘Euchresta’[Title/Abstract] OR ‘Guibourtia’[Title/Abstract] OR ‘Machaerium’[Title/Abstract] OR ‘Pithecellobium’[Title/Abstract] OR ‘Pithecolobium’[Title/Abstract] OR ‘Stryphnodendron’[Title/Abstract] OR ‘Tachigalia’[Title/Abstract] OR ‘Afzelia’[Title/Abstract]). The same search strategy was applied for the other three databases, and the equation was suitably adapted. The repeated studies found in the different databases were considered only once in the total list of the studies. The list of eligible studies was completed by the search in the reference list of the publications selected and respecting the a priori inclusion criteria established.

Study selection

Two of the authors (R. M. and G. K.) carried out the first screening of the eligible studies separately, which included the review of the abstracts of the studies and the selection of the suitable ones for full-text examination. At this point, bibliographic reviews, epidemiological studies, editorials, case reports and book chapters were excluded. There were no language restrictions. At the second stage of the selection process, the same authors examined the full-text articles and then selected the adequate studies to include. As the aim of the study was to review the existing data on animal intervention studies, the two authors manually excluded the clinical trials in humans. The decisions for the inclusion/exclusion were taken following mutual discussion and consensus. If consensus was not possible, two (M. L.-J. and J. M. P.) more authors examined the articles and the consensus was established after the discussion between the four authors.

Data extraction

After the conclusion of the study selection process, R. M. and G. K. independently reviewed and extracted the data of the selected studies. The overall inter-rater agreement rate before correcting discrepant items was determined using Cohen’s κ statistic^{( 15 )} and established to be superior to 0·80^{( 16 )}. Any discrepancies were resolved after consensus between the two or four authors (R. M. and G. K.) or between the four of them (including M. L.-J. and J. M. P.) if necessary. The quality of the studies selected was determined by the use of a specific questionnaire for the clinical trials (Scientific studies-clinical trials quality-evaluation questionnaire, CACEC-EC), which is divided into two parts: the first part includes filter questions that determine whether the study fulfils the methodology premises of a clinical trial (score>6) and the second part finally determines the quality (0–6, low; 7–14, good; 15–20, excellent) of the study in its different parts (intervention, sample manipulation, results and conclusions).

The extracted data were grouped in a table and classified according to the legume studied. In the different columns, the reference of the publication, the animal model (number, age and type of animals, experimental groups) used, the intervention (legume type and quantity consumed, technological process and experimental period) followed and the principal beneficial results achieved are noted, in order to facilitate the comprehension of the selected studies.

Results

The initial systematic search in the different electronic databases resulted in 417 references. After the exclusion of duplicated references (n 150, among which forty-three clinical trials, forty-nine epidemiological studies, fifty-eight reviews), there were 267 potentially eligible studies remaining. The first screening resulted in exclusion of bibliographic reviews (n 92), epidemiological studies (n 76) and other types of studies such as book chapters, case reports or editorials (n 6 in total). The possibly eligible studies were then reduced to ninety-three. The second screening, which was manually performed, resulted in the exclusion of: trials that studied parameters not relevant with the MetS alterations (n 27); clinical trials performed in humans (n 30); in vitro studies (n 8); or finally, animal studies that used legume diet intervention but obtained only negative results due to the specific intervention. After the second screening, twenty-six eligible studies remained, to which fifteen new were added after reviewing the reference lists of the studies already selected. After the whole process was completed, we retracted forty-one eligible studies, which included only in vivo experiments in different experimental animal models making use of a legume as part of the diet intervention. The entire process followed is represented in Fig. 1. In total, sixteen different legumes were reported in these studies. The beneficial effects on several parameters of the MetS were collected and are presented in Tables 1 and 2. The interobserver raw agreement was calculated at 95·12 % (k=0·725).

Fig. 1 Flow diagram of the eligible studies included in the systematic review.

Table 1 Beneficial effects of legumes on several parameters of the metabolic syndrome

BW, body weight; M, model; A/W, age/weight; LA, legume administration; SBO, soyabean oil; LLO, low α-linolenic soyabean oil; EP, experimental period; –, no effect; √, positive effect; FSD, soya protein concentrate; SPI, soya protein isolate; SBP, high content isoflavone soya protein; SP, soya protein; SPIs, isoflavone-depleted soya protein isolates; SPC, soya phytochemicals extract; FSP, fenugreek seed powder; EAES, ethyl acetate extract from seeds; FPEt, polyphenols from seeds; GAL, galactomannan; MCC, mycrocrystalline cellulose; CEL, cellulose; OFS, oligofructose; PF, yellow pea fibre; PFL, yellow pea flour; PS, yellow pea starch; JQ-R, refined JinQi-JiangTang tablet; GA, glycyrrhizic acid; LFO, licorice flavonoid oil; HFD, high-fat diet; LP, lentil protein; CPr, chickpea protein; AEPS, aqueous extract of Pterocarpus santaniloides; MEPS, methanolic extract of Pterocarpus santaniloide; SP-SHR, stroke prone – spontaneously hypertensive rat.

Table 2 Beneficial effects of legumes on different parameters of the metabolic syndrome expressed as numerical data (Mean values and standard deviations; mean values with their standard errors)

BR, bibliographic reference; CT, control diet; SBO, soyabean oil; LLO, low α-linolenic soyabean oil; F-diet, fructose diet; F-PL, 60 % fructose diet + phospholipids from soyabeans; G6PDX, glucose-6-phosphate dehydrogenase; TC, total cholesterol; FASN, fatty acid synthase; ACACA, acetyl-Coenzyme A carboxylase alpha; SCD1, stearoyl-CoA desaturase-1; CCD, starch and casein; FCD, fructose and casein; FSD, fructose and soya protein; CSD, starch and soya protein; HOMA-IR, homoeostatic model assessment for insulin resistance; ACE, angiotensin-converting enzyme; TBARs, thiobartituric acid-reactive substances; CAS, casein; SPI+, soya protein isolate; SPI−, soya protein isolate (negligible levels of phytochemicals); ACO, acyl-CoA oxidase; CPT-1, carnitine palmitoyltransferase I; HADHA, hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase/enoyl-CoA hydratase; PPAR, peroxisome proliferator-activated receptors; CYP/A-1, cholesterol 7 alpha – hydroxylase; ABCG5, 8, ATP-binding cassette sub-family G members 5, 8; LXRα, liver X receptor alpha; EW, protein of egg white; SP, soya protein; fa/fa, obese phenotype; BW, body weight; 6-Keto PGF1α, 6-keto prostaglandin F1α; LIS, low isoflavone soya protein; HIS, high isoflavone soya protein; CR, casein + rosiglitazone; Co, cocoa; S, soya; O, oats; Ω, fish oil, ASM, after metabolic syndrome; BSM, before metabolic syndrome; STD, standard diet; HFS, high-fat high-sucrose diet; Fen, fenugreek group; CON, starch diet; FRU, high-fructose diet; FRU + FPEt, high-fructose diet with fenugreek seed polyphenolic extract (200 mg/kg); FRU + Quer, high-fructose diet with quercetin (50 mg/kg); FRU + Met, high-fructose diet with metformin (50 mg/kg); QUICKY, quantitative insulin sensitivity check index; ISI_0,120, insulin sensitivity index at 0 and 120 min; GP, glycogen phosphorylase; Glu, glucose; ICDH, isocitrate dehydrogenase; SDH, succinate dehydrogenase; PTP, protein tyrosine phosphatases; PTK, protein tyrosine kinases; FSP, fenugreek seed powder; Allx, alloxan; CHOL, cholesterol; ALAT, alanine transaminase; ASAT, aspartate transaminase; ALP, alkaline phosphatase; GAL, galactomannan; NC, normal CT; Or, orlistat; MCC-PS, microcrystalline cellulose-potato starch; H, hyperlipidaemic; HMCC-PS, hyperlipidaemic diet and composite of MCC-PS; C, control; A, control + 1 % adzuki bean; CF, high-fat diet; AF, high-fat diet +1 % adzuki bean; EtEx, ethanol extract of adzuki beans; PF, yellow pea fibre; PFL, yellow pea flour; PS, yellow pea starch; OFS, oligofructose; SREBP-1c, sterol regulatory element binding protein 1; ACC, acetyl-CoA carboxylase; WPF, whole pea flour; FPF, fractionated pea flour; LFO, licoride flavonoid oil; ND, normal diet; HFD, high-fat diet; LDP, low-dose pigeon pea; MDP, medium dose; HDP, high dose; PC, post-control; LDLr, LDL receptor; HMG-CoA, HMG-CoA reductase (3-hydroxy-3-methyl-glutaryl-CoA reductase); MDA, malondialdehyde; RCD, regular chow diet; HFD, high-fat diet; CSEE, Casia seed ethanol extract; PG, pioglitazone; AMPK, adenosine monophosphate protein kinase; WAT, white adipose tissue; CP, chickpea; L, lentils; AEPS, aqueous extract of P. santaniloides; MEPS, metanolic extract of P. santaniloides; Ovex, ovariectomised; F, fructose-fed rats; F-T, fructose-fed Tamarindus indica seed aqueous extract; TpALet, aqueous extract of Tephrosia purpurea leaves.

* References number^{( 20 , 25 , 30 , 35 , 31 , 39 , 37 , 45 , 46 , 48 , 56 , 60 )}: no exact numeric data available, data represented by charts or figures.

† μmol of glucose phosphorylated/h per mg protein.

‡ μmol of pyruvate formed/min per mg protein.

§ μg of Pi liberated/min per protein.

|| μmol of Pi liberates/h per protein.

¶ Total cholesterol:HDL-cholesterol ratio.

** LDL-cholesterol:HDL-cholesterol ratio.

Mainly, as observed from the present review, the majority of the experiments were carried out using rats as an experimental model (n 30), followed by those that used mice (n 6). Focusing on the studies that used rats as the experimental model, different strategies for the induction and study of the MetS can be observed. Among them, the most common one is the induction of this pathology by diet in Wistar rats (n 15) followed by its induction on Sprague–Dawley (n 7) rats, another animal model that has been proven to be adequate for the study of this pathology. The most commonly used legume was Glycine max or soyabean (n 11), followed by Trigonella foenum gracecum or fenugreek (n 8) and Phaseolus vulgaris or beans (n 4), whereas in the rest of the studies a variety of legumes was used. The most common form of legume administration was in the form of an extract (n 11) or protein/fibre flour (n 7). It is worth mentioning that besides the study of the principal factors involved in the development of the MetS, the research is focused on the effects of the legume administration on the expression of several genes related to lipid, glucose and energy metabolism, as well as peptides and hormones associated with food intake, inflammatory markers and antioxidant status.

Glycine max/soyabean

Among the studies that used Glycine max as part of the diet intervention, one of them^{( 17 )} studied the effects of soyabean protein administration on pups of pregnant rats. The results of this study point out lower body weight and lipoprotein expression of the hepatic lipoprotein cytochrome P450, subfamily 2, polypeptide 11 in the pups that consumed soya protein isolate. In addition, the specific intervention positively influenced genes involved in peroxisomal and mitochondrial fatty acid β-oxidation such as acyl-CoA oxidase (COA), the mitochondrial trifunctional protein α subunit (hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase/enoyl-CoA hydratase) and fatty acid transport into the mitochondria by carnitine palmitoyltransferase 1A (CPT-1A) by increasing their expression in the liver. Further improvements on hepatic and serum lipid metabolism parameters due to soyabean administration were described in other studies^{( 18 – 23 )}. Specifically, among the mentioned studies, Barrios-Ramos et al. ^{( 18 )} and Potu et al. ^{( 21 )} indicated that the administration of powder and oil of soyabean induced improvements on hepatic steatosis and the hepatic inflammation marker c-reactive protein, respectively. In addition, proteins involved in lipid synthesis pathways (fatty acid synthase (FAS), acetyl-coenzyme A carboxylase α, Stearoyl-CoA desaturase-1, fatty acid elongase 6, sterol regulatory element binding protein 1 (SREBP1) and carbohydrate-responsive element-binding protein) were down-regulated as a consequence of soyabean administration, thus suggesting an improvement in lipid metabolism pathways^{( 19 , 24 )}.

Regarding glucose metabolism, the majority of the studies suggest a clear improvement induced by the specific legume. A decrease in plasma glucose, leptin and insulin concentration, as well as an improvement in insulin sensitivity index^{( 17 , 18 , 20 , 24 , 25 )}, has been reported. Such a beneficial action of soyabean is further supported by increased expression of key enzymes and genes linked to glucose metabolism such as insulin I (INS1), insulin II (INS2), GLUT2^{( 20 )} and PPARα and PPARγ ^{( 17 , 23 , 24 )} in pancreas, liver, muscle and adipose tissue.

Two of the retrieved studies pointed out positive effects of Glycine max on blood pressure^{( 18 , 26 )}, whereas Hwang et al. ^{( 27 )} observed a decrease of renal glomerular size and the improvement in parameters associated with glomerular filtration in the groups of rats fed soya protein. In this regard, Davis et al. ^{( 24 )} and Palanisamy et al. ^{( 26 )} reported a lower kidney weight, urinary volume and creatinine concentration, as well as proteinuria, because of the administration of this legume in Zucker diabetic and Wistar rats with MetS, respectively. Regarding oxidative stress in this tissue, the levels of thiobarbituric acid-reactive substances (TBARS) and GSH were restored and brought back to normal levels after the administration of Glycine max ^{( 26 )}.

The study of Zhou et al. ^{( 25 )} focused on the effects of this legume on white adipose tissue, demonstrating a decrease of the weight of this tissue in male and female mice.

Trigonella foenum gracecum/fenugreek

The use of fenugreek in all its different forms – that is, seed powder^{( 28 , 29 )}, extract^{( 30 – 32 )} isolated polyphenols^{( 33 )} or polysaccharide galactomannan^{( 34 , 35 )} – points out to the beneficial changes in glucose metabolism, as demonstrated by lower levels of blood insulin, glucose, AUC, as well as higher homoeostatic model assessment for insulin resistance (HOMA-IR) index. Moreover, the re-establishment of the enzymes that play an integral role within the insulin signalling cascade back to normal levels highlights this potential action^{( 33 )}. Specifically, Srichamroen et al. ^{( 35 )} demonstrated that galactomannan of fenugreek reveals its function at the intestinal level by reducing the in vitro uptake of glucose in both jejunum and ileal segments. Moreover, the hypolipidaemic properties of fenugreek are clearly demonstrated by lower levels of lipid fractions in blood^{( 28 – 31 , 34 )} and TAG in epidydimal adipose tissue^{( 34 )}, the weight of the latter being significantly lower after combining high-fat diets with powder of fenugreek seeds^{( 28 )}. Liver function markers such as alanine transaminase (ALT), aspartate transaminase (AST) and alkaline phosphatase activities^{( 29 , 31 )}, concentration of TBARS, as well as the activities of antioxidant enzymes such as catalase and superoxide dismutase^{( 25 )}, decreased because of the administration of this legume. In addition, serum parameters of renal functionality such as urea, uric acid and creatinine were reduced by fenugreek extract administration^{( 31 )}. Regarding the action of the specific legume on the immune system, Ramadan et al. ^{( 29 )} investigated the effects of fenugreek seed powder using an immunosuppressive rat model and demonstrated its potential by decreasing abnormalities of the immune system such as leucopenia, neutropenia and lymphopenia while increasing spleen-weight:body weight ratio and cellularity of lymphoid organs.

Phaseolus vulgaris/beans

The administration of P. vulgaris revealed a decrease in daily food intake and body weight, as well as improvements in plasma lipid parameters such as total cholesterol (TC), TAG, phospholipids and phosphorus phospholipids^{( 36 – 38 )}. Moreover, bean consumption caused a decrease in acetyl-CoA carboxylase (ACC) and increments in cholesterol 7 α-hydroxylase levels^{( 38 )}. Specifically, the study of Zaru et al. ^{( 39 )} demonstrated a decrease in the seeking behaviour of chocolate-flavoured beverage of animals fed P. vulgaris extracts compared with the animals in the control group. Regarding plasma glucose metabolism parameters, only blood glucose, plasma leptin and AUC were determined, which were all lower after the administration of this legume^{( 36 , 37 )}.

Vigna angularis/adzuki beans

In the three studies retrieved^{( 40 – 42 )}, the dietary intervention with Vigna angularis/Adzuki beans included the administration of this legume as an extract. The studies focused on glucose and lipid metabolism, indicating a reduction in glucose, insulin, glycated Hb and microalbumin:creatinine ratio in the plasma of the animals. In addition, concentrations of TC, TAG, as well as lipid content of the liver, were reduced as a consequence of the administration of this legume. Similar reductions were produced in liver weight. In contrast, faecal weight and lipid excretion were found to be increased.

Pissum sativum/yellow pea

The two studies retrieved^{( 43 , 44 )} demonstrated a reduction in blood glucose and insulin concentrations due to yellow pea administration, as well as decreased hepatic TAG, decreased ACC and increased SREBP mRNA levels.

Astragalus membranaceus/huáng-QÍ (translated as yellow leader)

The two studies retrieved^{( 45 , 46 )} used male diabetic animal models and aimed to study the effects of this legume on parameters related to glucose and lipid metabolism. Body weight was reduced resulting from legume administration, as well as parameters such as serum glucose and insulin concentrations, AUC and HOMA-IR index. In contrast, glucose infusion rate, after the performance of a hyperglycaemic clamp test, and hepatic glycogen content increased. Similar improvements were also found in parameters of lipid and energy metabolism represented by reduction of plasma TC and fatty acid concentration, as well as ACC and adenosine monophosphate activated protein (pAMPK) expression in the liver. The study of Gao et al. ^{( 45 )} performed histology and immunohistochemistry analyses of pancreas, demonstrating reduced pathological changes, stain intensity and area in the groups administered with the legume. Inflammation markers studied by Hoo et al. ^{( 46 )} were reduced in the adipose tissue of the treated groups.

Glycyrrhiza glabra/liquorice

The administration of Glycyrrhiza glabra lowered blood glucose, HOMA-IR index, serum insulin and leptin levels^{( 47 , 48 )}. Moreover, the 18-week administration of liquorice flavonoid oil (LFO) (1 %) led to lower body weight and periuterine and white adipose tissue of female C7BL/6J mice, whereas LFO (2 %) decreased adipocyte diameter and number of lipid droplets. In addition, it caused the up-regulation of genes related to β-oxidation and acyl-CoA degradation and down-regulation of glycolytic lipogenesis genes and those associated with acetyl-CoA synthesis^{( 47 )}. Increases in PPARγ and lipoprotein lipase (LPL) relative expressions after the administration of G. glabra were reported by the study of Yoke et al. ^{( 48 )}.

Other legumes

Other legumes in addition to the previously described ones have shown different effects on parameters associated with the MetS. The administration of amorfrutins of Glycyrrhiza foetida and Amorpha fruticosa (false indigo^{( 49 )}), Cajanus cajan (pigeon pea) powder, Pterocarpus santaniloides (Mututi^{( 49 )}) leaf extract, Pueraria lobata (Kudzu^{( 50 )}) root extract and Tamarindus indica (tamarind tree^{( 51 )}) aqueous extract^{( 52 – 56 )} decreased blood glucose, insulin content, as well as glucose and insulin AUC. The above-mentioned legumes in addition to Tephrosia purpurea, Amorpha administrated as a leaf extract^{( 57 )} have also shown their beneficial effect on parameters of lipid metabolism by lowering the serum levels of different lipid fractions. Tzeng et al. ^{( 58 )} demonstrated that an ethanol extract of Cassia tora (Foetid cassia^{( 59 )}) reduced the size of white adipose tissue, as well as the expression of enzymes such as FAS and SREBP in this tissue. In addition, it up-regulated the expression pAMPK, pACC and CPT1, all enzymes related to energy metabolism, and improved parameters of cardiovascular function such as atherogenic index and coronary risk index. Focusing on hepatic lipid metabolism, legumes such as Aspalathus linearis (Roibos), Lens culinaris (Lentils), C. cajan, G. foetida and T. purpurea ^{( 52 , 56 , 60 )} improved liver functionality by reducing liver weight, hepatic cholesterol and TAG content in addition to the reduction of lipid droplet accumulation and expression of TNFα, a widely used inflammation marker. According to the results of the present systematic review, only one study by Peng et al. ^{( 54 )} pointed out the beneficial effects on blood pressure after the inclusion of the root extract of P. lobata in the diet of the pups of an animal model of spontaneously hypertensive rats.

Discussion

The present systematic review was undertaken to give a comprehensive overview of the benefits of legume consumption on parameters related to the MetS and collect the existent mechanisms of action so far reported in animal experimental trials. In addition, it aimed to identify scarcities or abundancies with respect to legume consumption and its potential beneficial influence on the MetS alterations.

After the screening of the papers, data of forty-one studies were extracted. To our knowledge, this is the first systematic review gathering together the beneficial effects that a wide variety of legumes, most of them of common use, exert on the MetS, and include data on the way that legumes affect specific metabolic pathways involved in this pathology. The mechanistic emphasis of this review implies that preferentially animal studies were chosen.

Although some studies in humans indicate possible undesired effects due to the consumption of legumes, no such effects were reported in the studies collected for this review. Moreover, no toxicity effects by the administration of legumes in any form were reported. However, an increase of hepatic phospholipids was induced by the administration of adzuki beans^{( 41 )}, chickpeas and lentils^{( 61 )}, in addition to a decrease of LPL activity in epididymal fat reported by the latter study. In addition, in the study of Shahraki et al. ^{( 55 )}, an elevation of AST and ALT was observed in the group that consumed the aqueous extract of T. indica. As for the insulin resistance, Wagner et al. ^{( 23 )} concluded that after soya isoflavone administration, insulin responses significantly increased and were accompanied by decreased plasma adiponectin concentrations. In a similar manner, administration of soyabean oil in Ossabaw pigs^{( 21 )} resulted in elevated concentrations of glucose and insulin concentrations in plasma, as well as elevated blood lipids. Nevertheless, despite the negative effects of legume consumption in the above-mentioned studies, the majority of the studies gathered by the present systematic review highlight the beneficial effects of legume administration on the development and progression of the MetS and its related pathologies.

According to the results of the CACEC-EC questionnaire, the quality of the retrieved studies was good (Fig. 2), although there was great heterogeneity among them. In addition to the variety of legumes used, they were administered in different forms such as seed powder, extract or different fractions of the legume (protein, fibre). There was also great heterogeneity regarding the experimental period of the studies finally selected, which varied from 2 h^{( 32 )} to 40 weeks^{( 23 )}. However, all of them were randomised intervention studies according to the inclusion criteria established.

Fig. 2 Quality of the included studies of the systematic review. , Excellent; , good; , bad.

The frequent use of Glycine max/soyabean in the studies retrieved can be explained because of the declaration of its protein as a good substitute for animal products, offering a ‘complete’ protein profile and its protective action against CVD^{( 62 , 63 )} by the US Food and Drug Administration^{( 64 )}. Most of the studies included the investigation of various metabolic parameters simultaneously trying to offer evidence on more than one metabolic pathway. The most widely mentioned parameters related to glucose, lipid and renal metabolism are included, whereas inflammation, oxidative status, blood pressure, body weight and body composition were studied in fewer studies. Only one study focused on the anorectic effects of legumes by reducing appetite and craving for food^{( 39 )}.

As impairments of glucose metabolism are directly related to the MetS, these alterations are widely studied. Therefore, lowering glucose concentration, HOMA-IR index or increasing insulin response are among the most reported findings. Such positive effects seem to be independent from the intervention duration, as even the shortest intervention^{( 32 )} induced an improvement in blood glucose. However, it is worth mentioning that in this study T. foenum graecum extract was directly injected in alloxan-induced diabetic animals. In general, twenty-nine of the retrieved studies showed improvements in glucose metabolism and included several legumes such as Glycine max ^{( 17 – 20 , 24 , 26 , 38 )}, T. foenum graecum ^{( 28 , 29 , 31 – 33 )}, P. vulgaris ^{( 36 – 38 )}, V. angularis ^{( 41 )}, Pisum sativum ^{( 43 , 44 )}, Astragalus membranaceus ^{( 45 , 46 )}, G. glabra ^{( 47 , 48 )}, C. cajan ^{( 52 )}, G. foetida and A. fruticosa ^{( 56 )}, P. santaniloides ^{( 53 )}, P. lobata ^{( 54 )}, T. indica ^{( 55 )} and T. purpurea ^{( 57 )}. No such effects were reported for A. linearis ^{( 60 )}, C. tora ^{( 58 )} and L. culinaris/Cicer arietinum ^{( 61 )}. It seems that legumes influence the mechanistic pathways involving the expression of genes related to glucose metabolism such as GLUT2, GLUT4, INS1 or INS2^{( 20 , 24 )}, although the expression of more genes need to be studied. One of the retrieved studies also measured the activities of glucose- and glycogen-metabolising enzymes, therefore demonstrating the beneficial effect that polyphenols of T. foenum graecum exert on glucose metabolic pathways^{( 33 )}. Moreover, the study of Srichamroen et al. ^{( 35 )} revealed that another possible mechanism explaining glucose regulation is possible through the action of a galactomannan of the same legume in the reduction of the uptake of glucose by jejunum and ileal segments of the intestine.

In a manner similar to glucose metabolism, lipid parameters seem to be positively influenced by the administration of all sixteen different legumes that have been included in this review. Among the most widely mentioned beneficial improvements, the reduction of different lipid fractions in plasma, such as total-, LDL-, HDL-cholesterol and TAG^{( 20 , 22 , 28 – 31 , 34 , 36 , 38 , 46 , 48 , 54 – 56 , 58 , 61 )}, hepatic TAG and phospholipid content^{( 17 , 19 , 60 )}, or both of them^{( 18 , 24 , 40 – 43 , 45 , 52 , 61 )}, is reported. Other improvements associated with lipid metabolism and body composition are the decrease of body fat mass and white adipose tissue by Glycine max ^{( 17 , 25 )}, as well as the reduction of hepatic steatosis induced by this same legume^{( 18 )}. In this regard, the administration of G. glabra and A. linearis also reduced the number of lipid droplets in the liver^{( 47 , 60 )}. Moreover, the studies of Aoki et al. ^{( 47 )} and Tzeng et al. ^{( 58 )} used the determination of mesenteric, perirenal, periuterine, inguinal and epidydimal fat as a marker of increased lipid adiposity in animals and further improvement of this parameter by the administration of G. glabra and C. tora, respectively. It is quite clear that the MetS is related to impaired fat excretion, whereas the administration of V. angularis extract^{( 42 )} and Pigeon pea ^{( 52 )} improves such alteration. The results of the collected studies demonstrate that a great number of genes related to β-oxidation and acyl-CoA degradation are up-regulated by the administration of several legumes, whereas glycolytic lipogenesis genes are down-regulated. In particular, Glycine max ^{( 17 , 19 , 23 , 24 )}, P. vulgaris ^{( 38 )}, P. sativum ^{( 43 )}, A. membranaceus ^{( 45 )}, G. glabra ^{( 47 , 48 )}, C. cajan ^{( 52 )}, C. tora ^{( 58 )}, G. foetida/A. fructicosa ^{( 56 )} are among the encountered legumes with such action. Still, collected data indicate that more research needs to be developed on these and other potential mechanism related to the beneficial influence of legumes on lipid metabolism, whereas a greater range of legume species needs to be tested.

It is well known that renal alterations can occur with the development of the MetS. However, as demonstrated by the results of this systematic review, only six of the collected studies mention beneficial results on renal metabolism in which only four different legumes are included: Glycine max ^{( 24 , 26 , 27 )}, T. foenum graecum ^{( 31 )}, V. angularis ^{( 41 )} and C. cajan ^{( 52 )}. In this regard, legume administration managed to restore the augmented kidney weight, urea level, uric acid and creatinine derived from the administration of a high-fructose diet. The presence of glucose and protein in urine are also linked to alterations of renal metabolism and were improved by the administration of V. angularis ^{( 41 )} and Glycine max ^{( 24 )}. Worth mentioning is the study by Palanisamy et al. ^{( 26 )} that described a simultaneous reduction of blood pressure together with concomitant improvements in renal metabolism, as soya protein reduced glucose levels and produced the inhibition of the angiotensin-converting enzyme. Still, there is a lack of information in this field for the majority of the legumes gathered by this review.

The process of inflammation is highly involved in the development of the MetS and can be determined by the concentration of oxidative markers or the activity of antioxidant enzymes in different organs. As observed by this systematic review, only five of the legumes collected have been so far used to investigate these parameters. Among them, Glycine max ^{( 21 , 26 )}, T. foenum graceum ^{( 29 , 30 )}, A. membranaceus ^{( 46 )}, C. cajan ^{( 52 )} and Glythirrhiza foetida/A. fructicosa ^{( 56 )} are encountered. Two clear tendencies are observed for the evaluation of these parameters: on the one hand, the simultaneous determination of oxidative damage, as well as antioxidant enzymes^{( 26 , 30 , 52 )}, and on the other hand^{( 29 , 46 , 56 )}, the study of the level of cytokines involved in the process of inflammation.

Overall, legume administration positively affects glucose and lipid metabolism, which include the most widely studied parameters. Fewer studies have been focused in renal metabolism and the properties of legumes as antioxidant and anti-inflammatory agents. A possible limitation of the present review is that the bibliographic search was carried out based on the definition of search terms through the use of MeSH, not followed by all studies. It is important that the same rules be followed for the establishment of key words so that the inclusion of all available studies would be ensured.