Gut microbiota in individuals with type 2 diabetes mellitus and obesity

Most gut micro-organisms inhabit the large intestine that contains an estimated 10¹¹⁻¹² bacteria per g⁽ Reference Moreno-Indias, Cardona and Tinahones ^{14 )}. Gut microbiota can influence host adiposity and regulate fat storage⁽ Reference Backhed, Ding and Wang ^{4 ,} Reference Ley, Backhed and Turnbaugh ^{7 ,} Reference Jumpertz, Le and Turnbaugh ^{15 –} Reference Shavakhi, Minakari and Firouzian ^{17 )}.

The Bacteroidetes and the Firmicutes are groups of bacteria dominant in the human gut⁽ Reference Ley, Turnbaugh and Klein ^{8 )}. A correlation between changes in gut microbiota composition and obesity was reported in obese human subjects⁽ Reference Ley, Turnbaugh and Klein ^{8 ,} Reference Turnbaugh, Hamady and Yatsunenko ^{18 )} and ob/ob mice⁽ Reference Turnbaugh, Ley and Mahowald ^{19 )}, with lower microbial diversity, increased Firmicutes and decreased Bacteroidetes, and this obesity-associated gut microbiota had an increased capacity for energy harvest from the diet⁽ Reference Turnbaugh, Ley and Mahowald ^{19 )}. Germ-free wild-type C57BL/6J mice colonised with caecal microbiota from obese donors showed a significant increase (47 (sd 8·3) %) in body fat, compared with 27 (sd 3·6) % in mice colonised with a microbiota from lean donors over 2 weeks. During the 2 weeks, obese microbiota recipients consumed 55·4 (sd 2·5) g chow and gained 1·3 (sd 0·2) g fat, while the lean microbiota recipients consumed 54·0 (sd 1·2) g chow and gained 0·86 (sd 0·1) g fat, a difference of 2 % of total energy consumed⁽ Reference Turnbaugh, Ley and Mahowald ^{19 )}.

Furthermore, composition of the intestinal microbiota in adults with T2DM was different from that in non-diabetic adults. The proportion of Firmicutes, Clostridia and bifidobacteria was significantly lower in diabetic adults than in non-diabetic adults⁽ Reference Larsen, Vogensen and van den Berg ^{20 ,} Reference Wu, Ma and Han ^{21 )}.

Mucin-degrading bacteria such as Akkermansia muciniphila and Desulfovibrio were enriched in samples derived from T2DM patients⁽ Reference Qin, Li and Cai ^{22 )}. In contrast, several recent studies showed that in mice, direct administration of Akkermansia muciniphila ⁽ Reference Everard, Belzer and Geurts ^{23 –} Reference Schneeberger, Everard and Gomez-Valades ^{27 )} or specific proteins isolated from the outer of membrane of Akkermansia muciniphila could prevent obesity, insulin resistance as well as atherosclerosis and a human study also showed that the abundance of Akkermansia muciniphila was associated with glucose homeostasis and body fat composition⁽ Reference Dao, Everard and Aron-Wisnewsky ^{28 )}.

Vrieze et al. ⁽ Reference Vrieze, Van Nood and Holleman ^{29 )} investigated the effect of altering the gut microbiota on insulin sensitivity in subjects with the metabolic syndrome. Obese subjects given a small-intestinal infusion of faeces from lean donors (n 9) have shown improved peripheral insulin sensitivity after 6 weeks (median rate of glucose disappearance changed from 26·2 to 45·3 µmol/kgpermin; P<0·05), as assessed by the two-step hyperinsulinaemic–euglycaemic clamp method. These subjects also have shown a 2·5-fold increase in butyrate-producing gut microbiota, Roseburia intestinalis, compared with obese subjects reinfused with their own faeces (n 9). The faecal microbiota of obese subjects had lower microbial diversity, and contained higher amounts of Bacteroidetes and lower amounts of Clostridium cluster XIVa compared with faecal microbiota after lean donor infusion at 6 weeks⁽ Reference Vrieze, Van Nood and Holleman ^{29 )}.

Germ-free mice which are protected from diet-induced obesity had increased phosphorylated 5'-AMP-activated protein kinase (AMPK) in skeletal muscle and liver and increased fatty acid oxidation enzymes (acetyl CoA carboxylase; carnitine palmitoyltransferase), while germ-free knockout mice deficient in fasting-induced adipose factor (FIAF), a circulating lipoprotein lipase inhibitor, were not resistant to diet-induced obesity. Germ-free FIAF-deficient animals fed a Western diet showed decreased expression of the peroxisomal proliferator activated receptor-γ coactivator 1α (PGC1α) which is known to increase genes encoding regulators of mitochondrial fatty acid oxidation in the gastrocnemius muscle compared with germ-free FIAF^+/+ littermates, while there were no differences in phosphorylated AMPK levels between both groups. Consequently, germ-free mice were protected from diet-induced obesity through increased FIAF by inducing PGClα and also through elevated AMPK activity, implicating that obese gut microbiota can be responsible for decreased fatty acid oxidation and decreased FIAF/AMPK within the adipose tissue and liver⁽ Reference Bäckhed, Manchester and Semenkovich ^{30 )}.

The lipopolysaccharide (LPS) endotoxins, found in the outer membrane of some species of Gram-negative bacteria (for example, Neisseria spp. and Haemophilus spp.), induce signalling by binding to Toll-like receptor-4 present on endothelial cells, macrophages and monocytes. This promotes pro-inflammatory cytokines, chemokines, adhesion molecules and reactive oxygen species. An increase in LPS has been directly associated with insulin resistance⁽ Reference Cani, Amar and Iglesias ^{31 )}. Cani et al. ⁽ Reference Cani, Bibiloni and Knauf ^{32 )} found in a mouse model that high-fat feeding changes gut microbiota with a marked reduction in some bacteria (Lactobacillus spp. and Bacteroides–Prevotella spp.), leading to an increased intestinal permeability, and LPS absorption. This increased metabolic endotoxaemia initiates adipose tissue inflammation (plasminogen activator inhibitor-1 and IL-1 mRNA), macrophage infiltration markers (macrophage chemoattractant protein-1 (MCP-1) mRNA, F4/80 mRNA) and oxidative stress (NADPHox mRNA, visceral adipose tissue and six transmembrane protein of prostate 2 (STAMP2, known to regulate nutrient-derived and inflammatory signals coordinately for metabolic homeostasis⁽ Reference Wellen, Fucho and Gregor ^{33 ,} Reference Waki and Tontonoz ^{34 )})).

Even though intestinal microbiota may play a role in the aetiology of obesity and insulin resistance, the relationship between the bacteria and these metabolic disorders remains a matter of debate and most publications merely report associations between intestinal microbial composition and metabolic disorders such as obesity and T2DM⁽ Reference Bouter, van Raalte and Groen ^{35 )}.

Probiotics and effects of probiotics on glucose metabolism in human interventions

According to the FAO/WHO, probiotics are ‘live microorganisms which, when administered in adequate amounts, confer a health benefit on the host’⁽ Reference Araya, Morelli and Reid ^{36 )}. The requirement of adequate amounts differs between countries: products must contain at least 10⁷ colony-forming units (CFU)/g of probiotic bacteria in Japan, at least 10⁸ CFU/g probiotic bacteria in USA and 10⁹ CFU/g probiotic bacteria in Canada. In general, >10⁶–10⁸ CFU/g, or >10⁸–10¹⁰ CFU/d of viable cells are regarded efficacious⁽ Reference Homayoni Rad, Mehrabany and Alipoor ^{37 ,} Reference Champagne, Ross and Saarela ^{38 )}, but the cell count levels recognised do not guarantee a health effect⁽ Reference Champagne, Ross and Saarela ^{38 ,} Reference Reid ^{39 )}. Moreover, it is suggested that the recommendation for CFU determination should be established using accurate and frequent assessments because the number of viable cells is reduced during production, processing and formulation⁽ Reference Champagne, Ross and Saarela ^{38 )}.

Some of the species are: (1) lactic acid-producing bacteria (Lactobacillus, Bifidobacterium, Streptococcus); (2) non-lactic acid-producing bacterial species (Bacillus, Propionibacterium); (3) non-pathogenic yeasts (Saccharomyces; for example, Saccharomyces boulfecesardii, a non-colonising lactic acid-producing yeast); (4) non-spore-forming and non-flagellated rod or coccobacilli⁽ Reference Saraf, Shashikanth and Priy ^{40 )}. Over 100 Lactobacillus and over thirty Bifidobacterium species have been identified⁽ Reference Saraf, Shashikanth and Priy ^{40 )}. Lactic-producing Bifidobacterium and Lactobacillus are the predominant and subdominant probiotic groups⁽ Reference Bermudez-Brito, Plaza-Diaz and Munoz-Quezada ^{41 )}.

Human interventions with probiotics are shown in Table 1. Ten interventions⁽ Reference Ejtahed, Mohtadi-Nia and Homayouni-Rad ^{10 ,} Reference Hove, Brons and Faerch ^{12 ,} Reference Laitinen, Poussa and Isolauri ^{42 –} Reference Barreto, Colado Simao and Morimoto ^{49 )} have shown positive effects of probiotics on glucose control. Ten interventions⁽ Reference Lindsay, Kennelly and Culliton ^{6 ,} Reference Naruszewicz, Johansson and Zapolska-Downar ^{50 –} Reference Ogawa, Kadooka and Kato ^{58 )} have shown no effect and two interventions have shown negative effects⁽ Reference Ivey, Hodgson and Kerr ^{59 )}.

Table 1 Summary of probiotic human intervention studies

CFU, colony-forming unit; C, control; PC, placebo–control; P, parallel; RD, randomised; DB, double blind; SBP, systolic blood pressure; BP, blood pressure; SB, single blind; TC, total cholesterol; LDL-C, LDL-cholesterol; HDL-C, HDL-cholesterol; HC, hypercholesterolaemia; HOMA-IR, homeostasis model assessment of insulin resistance; QUICKI, quantitative insulin sensitivity check index; ATCC, American Type Culture Collection; NGT, normal glucose tolerance; IGT, impaired glucose tolerance; T2DM, type 2 diabetes mellitus; IL-Ira, IL-1 receptor antagonist; hs-CRP, high-sensitivity C-reactive protein; HbA1c, glycated Hb; FOS, fructo-oligosaccharide; MetS, metabolic syndrome; VLDL-C, VLDL-cholesterol.

Patients with T2DM supplemented with probiotic yogurt experienced attenuated fasting glucose and glycated Hb (HbA1c) concentrations and increased erythrocyte superoxide dismutase (SOD), glutathione peroxidase (GPx) activities and total antioxidants, compared with the control group⁽ Reference Ejtahed, Mohtadi-Nia and Homayouni-Rad ^{10 )}. Pregnant women given a probiotic supplement (Bifidobacterium lactis Bb12 and Lactobacillus rhamnosus GG) and dietary counselling together had improved glycaemic control during and after pregnancy compared with the control/placebo group⁽ Reference Laitinen, Poussa and Isolauri ^{42 )}.

In sixty women with gestational diabetes, the daily supplement of a probiotic capsule, containing three viable freeze-dried strains including Bifidobacterium bifidum (2×10⁹ CFU/g), Lactobacillus acidophilus (2×10⁹ CFU/g) and L. casei (2×10⁹ CFU/g), for 6 weeks showed improved insulin sensitivity as assessed by the homoeostasis model assessment for insulin resistance (HOMA-IR) and quantitative insulin sensitivity check index (QUICKI), compared with a placebo capsule (cellulose). This probiotic supplementation also lowered TAG and VLDL levels, but this study did not measure HbA1c⁽ Reference Karamali, Dadkhah and Sadrkhanlou ^{48 )}. Moreover, oral supplementation of L. acidophilus NCFM (progenitor of the strain being used for complete chromosome sequencing in order to identify the relationship between genetics and probiotic functionality⁽ Reference Sanders and Klaenhammer ^{60 )}) for 4 weeks improved insulin sensitivity as assessed by hyperinsulinaemic–euglycaemic clamp, compared with a placebo group, without changes in inflammatory markers such as TNF-α, IL-6, IL-Ira and high-sensitivity C-reactive protein⁽ Reference Andreasen, Larsen and Pedersen-Skovsgaard ^{43 )}.

On the other hand, one study of overweight adults⁽ Reference Ivey, Hodgson and Kerr ^{59 )} has shown conflicting results in that the intake of probiotic yogurt increased HOMA-IR (P=0·038) and probiotic capsules significantly increased fasting glucose (P=0·037) with no change in HOMA-IR⁽ Reference Ivey, Hodgson and Kerr ^{59 )} (n 77 for the probiotic group and n 79 for the probiotic capsule; the specific study design is shown in Table 1). The probiotics used were Lactobacillus acidophilus La5 and Bifidobacterium animalis subsp. lactis Bb12 (dose of 3·0×10⁹ CFU/d)⁽ Reference Ivey, Hodgson and Kerr ^{59 )}.

A single-blinded clinical trial of thirty-four subjects with T2DM showed no effects on glycaemic control, lipid profiles and inflammatory markers between a placebo group (n 18; 1000 mg magnesium stearate) and a treatment group (n 16) who received 1500 mg probiotic capsules containing Lactobacillus acidophilus, L. bulgaricus, L. bifidum and L. casei twice daily for 6 weeks⁽ Reference Yousefinejad, Mazloom and Dabbaghmanesh ^{54 )}. However, this study provided no counts of probiotics used.

In summary, given the mixed results from human interventions, it is still unclear if probiotics favourably influence glucose control. More human interventions are needed with more comprehensive and dynamic measures of insulin sensitivity, as most studies did not use these. It is also required to investigate the treatment effects of specific strain(s) at different dosages and durations on insulin resistance.

Other fermented food

Kimchi, made with napa cabbage and various ingredients (garlic, red pepper powder, onion, ginger, radish, fermented fish sauce and starch syrup), is a fermented traditional Korean food. Kimchi can give health benefits due to its high nutritional value and abundant bioactive compounds including dietary fibres, minerals, amino acids, vitamins, carotenoids, glucosinolates and polyphenols. Kimchi can be improved by additional ingredients and altered fermentation conditions⁽ Reference Patra, Das and Paramithiotis ^{61 )}. Fermented kimchi mostly contains lactic acid-producing bacteria including Lactobacillus plantarum, Lactobacillus brevis, Pediococcus cerevisiae, Streptococcus faecalis and Leuconostoc mesenteroides, which could exert a probiotic effect⁽ Reference Patra, Das and Paramithiotis ^{61 )}.

Studies have shown beneficial effects of fermented kimchi on glucose metabolism in obese⁽ Reference Kim, An and Lee ^{62 )} and prediabetic individuals⁽ Reference An, Lee and Jeon ^{63 )}.

Fermented kimchi intake for 4 weeks decreased fasting glucose, fasting insulin, total cholesterol, MCP-1, leptin and the waist:hip ratio compared with fresh kimchi in a cross-over design of twenty-two overweight and obese patients. Fresh kimchi was defined as 1 d-old kimchi and fermented kimchi was defined as 10 d-old kimchi. The number of Lactobacilli in fermented kimchi was higher than in fresh kimchi (4·3×10⁹ (sd 1·2×10⁹)/ml v. 1·4×10⁷ (sd 3×10⁶)/ml)⁽ Reference Kim, An and Lee ^{62 )}. The consumption of fermented kimchi for 8 weeks decreased HbA1c, fasting insulin, HOMA and increased quantitative insulin sensitivity check index (QUICKI) and β-cell function compared with before fermented kimchi intake, in twenty-two adults with prediabetes⁽ Reference An, Lee and Jeon ^{63 )}.

A randomised controlled clinical trial in twenty-four obese women showed that fermented kimchi intake (180 g/d) for 8 weeks altered gut microbiota composition, with a decrease in genus Blautia and an increase in Prevotella and Bacteroides, compared with fresh kimchi intake and up-regulated expression of genes related to the metabolic syndrome such as acyl-CoA synthetase long-chain family member 1 (ACSL1; involved in enhancing fatty acid degradation) and aminopeptidase N (ANPEP; involved in the regulation of pain, angiogenesis, inflammation and apoptosis)⁽ Reference Han, Bose and Wang ^{64 )}.

Very recently, Shin et al. ⁽ Reference Shin, Kang and Jang ^{65 )} demonstrated metabolic pathways of kimchi action based on in silico modelling of published data. A total of 4351 genes were associated with kimchi metabolites. Of these, 283 genes were associated with carbohydrate metabolism. In all, 309 genes were associated with lipid metabolism and twenty-seven genes (especially GNAS, CTNNB1, EDN1, RAC1 and adenyl cyclases (ADCY1, ADCY2, ADCY5) known to act as regulators of metabolic and cardiovascular function) are directly related with CVD. Twenty-three genes (especially PTPRC, LCK, JAK3, ZAP70 and VEGFA) were related to immune diseases and twenty-five genes were related to endocrine and metabolic diseases⁽ Reference Shin, Kang and Jang ^{65 )}. In summary, these inconsistent findings with probiotic interventions might result from heterogeneity in probiotic strains and populations. Intervention studies should be designed with a specific group and a specific strain⁽ Reference Panwar, Rashmi and Batish ^{66 )}.

Potential mechanisms of action of probiotics

One potential mechanism of anti-diabetic effects is that certain probiotics facilitate production of SCFA (acetate (C₂), propionate (C₃) and butyrate (C₄)), leading to the secretion of incretin hormones which may influence glucose levels⁽ Reference Yadav, Lee and Lloyd ^{9 ,} Reference Belenguer, Duncan and Calder ^{67 )}. Yadav et al. ⁽ Reference Yadav, Lee and Lloyd ^{9 )} have demonstrated a potential mechanism of probiotics through butyrate-induced secretion of glucagon-like peptide 1 (GLP-1) in mouse models. In this study⁽ Reference Yadav, Lee and Lloyd ^{9 )}, VSL#3 consisting of Lactobacillus casei, L. plantarum, L. acidophilus and L. delbrueckii subsp. bulgaricus, Bifidobacterium longum, B. breve and B. infantis and Streptococcus salivarius subsp. thermophilus was used. The levels of the SCFA butyrate in the mouse faecal samples significantly increased after VSL#3 (daily oral dose of 5 mg/kg body weight), as measured by liquid chromatography–electrospray ionisation–tandem MS. Significantly increased plasma butyrate levels were observed in VSL#3-treated mice compared with PBS-treated control mice. For the measurement of butyrate-producing bacteria, gene expression of butyrate kinase was assayed after 2 and 4 weeks of oral administration. Gene expression of butyrate kinase increased at 2 weeks in VSL#3-treated mice. An increase in GLP-1 was observed in the human intestinal L-cell line NCI-H716 treated with butyrate⁽ Reference Yadav, Lee and Lloyd ^{9 )}. GLP-1, an incretin hormone secreted by L-cells mainly in the ileum and large intestine, increases insulin secretion while glucagon is suppressed. GLP1 secretion results in delayed gastric emptying and reduced appetite, food intake and body-weight gain⁽ Reference Drucker and Nauck ^{68 )}. Lactic acid produced by lactic acid-producing bacteria can be converted to acetate or propionate by Clostridium propionicum, Propionibacterium ssp., Desulfovibrio ssp, Veillonella ssp. and Selenomonas ssp.⁽ Reference Seeliger, Janssen and Schink ^{69 )} via methylmalonyl-CoA or acrylyl-CoA, and then to butyrate via acetyl-CoA by Eubacterium hallii (butyrate-producing species)⁽ Reference Belenguer, Duncan and Calder ^{67 )}.

Other potential mechanisms of anti-diabetic effects of probiotics could be associated with enhanced immunity and increased anti-inflammatory cytokine production, reduced intestinal permeability and reduced oxidative stress⁽ Reference Ejtahed, Mohtadi-Nia and Homayouni-Rad ^{10 ,} Reference Yadav, Jain and Sinha ^{11 ,} Reference Ma, Forsythe and Bienenstock ^{70 ,} Reference Paszti-Gere, Szeker and Csibrik-Nemeth ^{71 )}. In a randomised, double-blind, controlled intervention, patients with T2DM consumed 300 g/d of yogurt (L. acidophilus La5, B. lactis Bb12 with a dose of 3·98×10⁹ CFU) for 6 weeks and experienced a reduction in fasting glucose and HbA1c and an increase in GPx and erythrocyte SOD activities and total antioxidant status, compared with the control (300 g/d of conventional yogurt). GPx and SOD are scavengers of reactive oxygen species⁽ Reference Ejtahed, Mohtadi-Nia and Homayouni-Rad ^{10 )}.

Pre-incubation of HeLa cells with live Lactobacillus reuteri cells for 1–2 h inhibited translocation of NF-κB to the nucleus, inhibited degradation of IKKB (inhibitor of NF-κB kinase subunit β) and prevented expression of pro-inflammatory cytokines under NF-κB regulation. Live L. reuteri up-regulated nerve growth factor and inhibited constitutive synthesis and secretion of IL-8 induced by TNF-α in T84 and HT29 cells (human colonic adenocarcinoma)⁽ Reference Ma, Forsythe and Bienenstock ^{70 )}. Nerve growth factor is known to play roles in the regulation of inflammation⁽ Reference Lambiase, Bracci-Laudiero and Bonini ^{72 ,} Reference Ma, Wolvers and Stanisz ^{73 )} and proliferation of pancreatic β-cells⁽ Reference Pierucci, Cicconi and Bonini ^{74 )}. Metabolites of Lactobacillus plantarum 2142 down-regulated peroxide-induced elevation in proinflammatory cytokines IL-8 and TNF-α in the IPEC-J2 cell line (jejunal epithelia isolated from neonatal piglet)⁽ Reference Paszti-Gere, Szeker and Csibrik-Nemeth ^{71 )}. In streptozotocin-induced diabetic rats, probiotic dahi containing Lactobacillus casei and Lactobacillus acidophilus suppressed streptozotocin-induced oxidative stress in pancreatic tissues by preventing the depletion of glutathione, GPx and SOD, as well as by decreasing thiobarbituric acid-reactive substances and nitrite⁽ Reference Yadav, Jain and Sinha ^{11 )}. This finding implicates that probiotic dahi could delay streptozotocin-induced alteration in glucose homeostasis by exerting an antioxidant effect on β-cells⁽ Reference Yadav, Jain and Sinha ^{11 )}.

Prebiotics and effects of prebiotics on glucose metabolism in human interventions

Prebiotics are non-digestible food ingredients that are not metabolised or absorbed while passing through the upper gastrointestinal tract and are fermented by bacteria in the colon and selectively enhance the growth and/or activity of one or more potential beneficial bacteria (for example, Bifidobacterium and Lactobacillus) in the digestive system⁽ Reference Yoo and Kim ^{75 –} Reference Flesch, Poziomyck and Damin ^{77 )}.

Food sources of prebiotics are seeds, whole grains, legumes, chicory roots, Jerusalem artichokes, onions, garlic and some vegetables. Some prebiotics can be produced during the process of enzymic action or alcohol or cooking⁽ Reference Flesch, Poziomyck and Damin ^{77 )}.

Prebiotics include fructo-oligosaccharides (FOS), galacto-oligosaccharides, lactulose and large polysaccharides (inulin, resistant starches, cellulose, hemicellulose, pectin and gum)⁽ Reference Yoo and Kim ^{75 )}. Of these, researchers have given more attention to FOS⁽ Reference Flesch, Poziomyck and Damin ^{77 )}. Synthetic oligosaccharides such as galacto-oligosaccharies have shown better effects and fewer side effects than natural forms⁽ Reference Flesch, Poziomyck and Damin ^{77 )}. Oligofructose-enriched inulin can act across the whole colon. Oligofructose is a short-chain fructan (a polymer of fructose molecules) containing three to ten monosaccharides linked together. It is quickly fermented and completely metabolised in the ascending part of the colon, whereas inulin is a long-chain fructan containing nine to sixty-four monosaccharides linked together. It is fermented and metabolised in the descending colon⁽ Reference Flesch, Poziomyck and Damin ^{77 ,} Reference Dehghan, Pourghassem Gargari and Asghari Jafar-abadi ^{78 )}.

Inulin-type fructans of 10–20 g/d can normalise glucose tolerance or lipid profiles⁽ Reference Sanz and Santacruz ^{79 –} Reference Letexier, Diraison and Beylot ^{84 )}. FOS or inulin of 4 g/d is the minimal requirement for the enhancement of bifidobacteria growth but 14 g/d or more of inulin can cause intestinal discomfort⁽ Reference Flesch, Poziomyck and Damin ^{77 )}.

In various animal studies, prebiotics have shown improved glucose metabolism⁽ Reference Kok, Morgan and Williams ^{85 ,} Reference Everard, Lazarevic and Derrien ^{86 )}. However, a few human studies have demonstrated inconsistent findings. Human interventions of prebiotics are shown in Table 2.

Table 2 Summary of prebiotic human intervention studies

CO, cross-over; TC, total cholesterol; RD, randomised; DB, double blind; PC, placebo–control; P, parallel; HbA1c, glycated Hb; HOMA-IR, homeostasis model assessment of insulin resistance; PMR, partial meal replacement; C, control; L, longitudinal; HDL-C, HDL-cholesterol; HC, hypercholesterolaemia; GLP-1, glucagon-like peptide-1; PYY, peptide YY; FOS, fructo-oligosaccharide; LDL-C, LDL-cholesterol; OGTT, oral glucose tolerance test; LPS, lipopolysaccharide; hs-CRP, high-sensitivity C-reactive protein; T2DM, type 2 diabetes mellitus; SB, single blind; sc-FOS, short-chain fructo-oligosaccharide; AST, aspartate aminotransferase; GIP, glucose-dependent insulinotropic peptide; ALT, alanine aminotransferase; GGT, γ-glutamyltranspeptidase; RS, resistant starch; MTT, meal tolerance test; HAM-RS2, high-amylose maize type 2 resistant starch; MetS, metabolic syndrome.

Potential mechanisms of action of prebiotic-derived SCFA in insulin sensitivity

Microbial fermentation of prebiotics facilitates the production of SCFA (essential endproducts of carbohydrate metabolism) and enhances gut barrier function⁽ Reference Cani, Possemiers and Van de Wiele ^{109 )}. In a mouse model, a prebiotic treatment decreased intestinal permeability and increased GLP-2 secretion, and reducing the hepatic expression of inflammatory and oxidative stress markers and decreasing LPS during obesity and diabetes⁽ Reference Cani, Possemiers and Van de Wiele ^{109 )}.

Synbiotics and effects of synbiotics on glucose metabolism in human interventions

A combination of probiotics and prebiotics is described as a synbiotic⁽ Reference Flesch, Poziomyck and Damin ^{77 )}. Lactobacillus acidophilus DSM20079 induced 14·5-fold more butyrate in the presence of inulin or pectin than in the presence of glucose⁽ Reference Nazzaro, Fratianni and Nicolaus ^{159 )}.

Human interventions of synbiotics are shown in Table 3. Eleven of twelve studies of synbiotics have shown favourable effects on glucose metabolism⁽ Reference Tajabadi-Ebrahimi, Sharifi and Farrokhian ^{13 ,} Reference Shavakhi, Minakari and Firouzian ^{17 ,} Reference Moroti, Souza Magri and de Rezende Costa ^{160 –} Reference Mofidi, Poustchi and Yari ^{169 )}. Three⁽ Reference Shavakhi, Minakari and Firouzian ^{17 ,} Reference Eslamparast, Poustchi and Zamani ^{168 ,} Reference Mofidi, Poustchi and Yari ^{169 )} of four⁽ Reference Shavakhi, Minakari and Firouzian ^{17 ,} Reference Wong, Won and Chim ^{167 –} Reference Mofidi, Poustchi and Yari ^{169 )} studies in subjects with non-alcoholic fatty liver disease have shown a positive effect on glucose control. The Asemi research group conducted a randomised, double-blind, placebo-controlled trial in subjects with T2DM; the consumption of synbiotic bread (containing the probiotic Lactobacillus sporogenes (1×10⁸ CFU) and 0·07 g inulin per 1 g as prebiotic) for 8 weeks improved insulin metabolism, lipid profiles and plasma NO and malondialdehyde levels, compared with the probiotic alone (Lactobacillus sporogenes; 1×10⁸ CFU) and a control bread⁽ Reference Tajadadi-Ebrahimi, Bahmani and Shakeri ^{164 ,} Reference Bahmani, Tajadadi-Ebrahimi and Kolahdooz ^{170 ,} Reference Shakeri, Hadaegh and Abedi ^{171 )}.

Table 3 Summary of synbiotic human intervention studies

CFU, colony-forming unit; FOS, fructo-oligosaccharide; T2DM, type 2 diabetes mellitus; RD, randomised; P, parallel; DB, double blind; PC, placebo–control; HDL-C, HDL-cholesterol; TC, total cholesterol; hs-CRP, high-sensitivity C-reactive protein; GSH, glutathione; NASH, non-alcoholic steatohepatitis; ALT, alanine aminotransferase; AST, aspartate aminotransferase; OL, open label; C, control; LDL-C, LDL-cholesterol; NAFLD, non-alcoholic fatty liver disease; HOMA-IR, homeostasis model assessment of insulin resistance; GGT, γ-glutamyltransferase; MetS, metabolic syndrome; QUICKI, quantitative insulin sensitivity check index; VLDL-C, VLDL-cholesterol; MDA, malondialdehyde; TAC, total antioxidant capacity; BP, blood pressure; CO, cross-over; SB, single blind.

In summary, a very limited number of interventions have shown beneficial effects on glucose metabolism. More pronounced effects of synbiotics on glycaemic control and inflammation have been observed than with the use of probiotics alone⁽ Reference Jumpertz, Le and Turnbaugh ^{15 )}.