Fermented soybean foods and diabetes.

INTRODUCTION

Type 2 diabetes is a major determinant of mortality, with an increasing prevalence worldwide 1 . In 2021, 537 million individuals were living with diabetes 2 . Therefore, the prevention and management of type 2 diabetes is an important public health concern.

Several dietary factors are reportedly involved in the pathogenesis of type 2 diabetes, including insulin secretion capacity, insulin resistance, fatty liver disease, inflammation, gut microbiota, and weight gain 1 . Although studies with high‐quality evidence are still lacking, Westernized diets, such as insufficient whole grains, excess refined rice and wheat, processed meats, and unprocessed red meats, are known to be associated with the development of type 2 diabetes 3 , 4 .

Fermented foods are foods transformed by the action of microorganisms to produce beneficial effects in humans 5 . Historically, many foods, including meat, fish, dairy products, vegetables, soybeans, legumes, grains, and fruits, have undergone fermentation. The production and consumption of fermented soybean foods are widespread in Asian countries. The main fermented soybean foods include natto, miso, douchi, cheonggukjang, doenjang, tempeh, and fermented soy milk 6 , 7 . Recent research has shown that fermentation can be applied to improve the health benefits of the bioactive components in soybeans. This review focuses on fermented soybean foods and summarizes their effects on diabetes.

NUTRITION CHANGE OF SOYBEAN DURING FERMENTATION

Raw soybeans contain dietary fibers, phospholipids, and isoflavones (genistein and daidzein), phenolic acids, saponins, trypsin inhibitors, and phytic acids.

The differences in fermented soybean foods are mainly due to the microorganisms used, with some using only bacteria, mainly Bacillus, for fermentation and others using only filamentous fungi, mainly Aspergillus, and often both of these microbial groups. For example, natto, kinema, and cheonggukjang use only bacteria; douchi, tempeh, miso, and tofu use only filamentous fungi; and doenjang uses both 6 . Fermentation improves soybean quality by increasing the digestibility, nutrition, and isoflavone content 8 . Fermentation increases the nutritional value of soybeans by increasing the content of vitamins, essential amino acids, and fatty acids and by enhancing detoxification. In addition to proteins and isoflavones, soybean contains numerous functional and nutritional substances. Microorganisms with the ability to produce specific hydrolytic enzymes play an important role in enhancing their functional properties 8 , 9 , 10 . The modification of isoflavones is carried out by β‐glucosidase; therefore, the amount of aglycone isoflavones depends on the amount of β‐glucosidase 11 , 12 . Aglycone isoflavones, such as genistein and daidzein, are deglycosylated by intestinal hydrolytic enzymes and microbial glycosidases, leading to an increase in their bioactive potential 13 . In addition, fermentation by various microorganisms can improve the antioxidant capacity and the levels of vitamin B2, vitamin B12, vitamin K2, and gamma aminobutyric acid 7 , 14 , 15 . Angiotensin‐converting enzyme inhibitory peptides are produced by the breakdown of soy protein 16 . Furthermore, several nutritional changes have been reported to result from soy fermentation, including increased total soluble iron, folate levels, tocopherol composition, and β‐, γ‐, and δ‐tocopherol levels 17 .

Miso, which is a traditional Japanese spice used to add flavor to soups and dishes, is a fermented soybean paste malted with ‘koji’, produced from Aspergillus oryzae. There are various types of miso depending on the koji used, including rice miso, which is made by adding rice koji to soybeans, barley miso, which is made by adding barley koji to soybeans, and soybean miso, which uses only soybeans. As with other fermented soybean foods, the characteristics of miso vary with ingredients, temperature, fermentation time, salt concentration, and the strain of A. oryzae used. It has been reported that amino acids such as glutamic acid, aspartic acid, and proline are included during the aging process of miso 18 , 19 .

Natto is one of the traditional Japanese fermented soybean foods. Among several varieties of natto, natto fermented with Bacillus subtilis is the most well known. It has been reported that during the fermentation process, proteins are cleaved by extracellular proteases produced by Bacillus natto, increasing the free amino acid content by 10–30%. Proteins derived from natto consist of at least 17 amino acids, including glutamic acid, glutamine, aspartic acid, leucine, proline, serine, lysine, methionine, threonine, glycine, isoleucine, tyrosine, phenylalanine, histidine, arginine, alanine, and valine 20 , 21 . The characteristics of natto vary with the steaming time, relative humidity, fermentation time, and temperature 22 .

In Korea, the various types of fermented soybeans are collectively called ‘jang’. Cheonggukjang is a traditional Korean fermented soybean food made from fermented boiled soybean rice straw. Cheonggukjang does not contain salt, but ingredients such as crushed green onions, garlic, and chili powder are added for flavor 23 . Various enzymes secreted by bacteria during fermentation have been reported to break down soybean hulls, cell membrane fibers, and intracellular sugars and proteins, improving digestibility and increasing free amino acid content, vitamin B2 and vitamin K2 23 . In addition, Cheonggukjang is reported to have higher protein and fat content than doenjang 24 .

Doenjang is a traditional Korean miso, made by fermenting boiled soybeans (meju) in salted water using yeast (A. oryzae) and Bacillus (Bacillus) to break down soybean proteins to produce organic acids, amino acids, and minerals 25 .

Kochujang is a mixture of meju powder, koji powder, and chili powder, which is fermented and aged for at least 6 months 26 . Kochujang contains a higher carbohydrate source than Doenjan and may contain a variety of Bacillus spp 26 .

Douchi is a Chinese fermented soybean food that has been used as a protein source and seasoning. Douchi is produced from two stages: pre‐fermentation, which consists of an aerobic process using several microorganisms, such as A. oryzae, Zygosaccharomyces rouxii, Lactobacillus plantarum, and B. subtilis, and post‐fermentation where salt and spices are added and mixed together for anaerobic fermentation 27 , 28 .

Fermented soy milk is a traditional Chinese vegetable protein drink, rich in bioactive substances such as saponins, polyphenols, isoflavones, and phytosterols in addition to soy nutrients 29 .

Tempeh is a food indigenous to Indonesia and has been consumed as a staple protein source 30 . Tempeh is usually made from soybeans fermented with Rhizopus oligoporus fungal species 31 . The microbial composition of tempeh varies depending on the variation in its production. Fermentation of soybeans is a source of large amounts of protein, vitamin B12, and bioactive compounds, along with reduced concentrations of protease inhibitors, phytic acid, and phenols, which are anti‐nutritional factors found in raw soybeans 32 , 33 .

Kinema is a traditional non‐salted fermented soybean food prepared in the eastern hills of Nepal, the Darjeeling hills, and Sikkim in India 34 . To make traditional kinema, soybeans are washed, soaked in water overnight, boiled for about 90 min, ground, wrapped in fern leaves and sack cloth, and fermented in a warm place for several days to 3 days. Bacillus subtilis is solely responsible for kinema production.

Kinema is rich in all essential amino acids with processing with B. subtilis leading to a 60‐fold increase in the free amino acid content of soybeans 35 .

EFFECTS OF FERMENTED SOYBEAN FOODS ON GUT MICROBIOTA

The role of gut microbiota in type 2 diabetes has been reported in recent years. Among the commonly reported genera, Bacteroides, Bifidobacterium, Roseburia, Faecalibacterium, Akkermansia, Lactobacillus, Ruminococcus, Fusobacterium, and Blautia have been associated with type 2 diabetes 36 . Metabolites produced by gut microbiota have been reported to have many functions, including energy conversion, signal transduction, epigenetic effects, and coenzyme activity, and are thought to be associated with diabetes 37 . However, it has been reported that the gut microbiota varies depending on the residence area and race 38 , 39 , and it is possible that fermented foods, including fermented soybean foods, also affect the gut microbiota and microbe‐associated metabolites 40 , including short‐chain fatty acids (SCFAs) and bile acids. It has been reported that SCFAs had a protective effect on diabetes by increasing energy expenditure, insulin sensitivity, and insulin secretion via GPR41and GPR43 41 and that secondary bile acids act protectively against diabetes by acting as endogenous ligands for the nuclear receptors FXR and TGR5, a member of G‐protein‐coupled receptor 42 , 43 . Although the effects of fermented soybean foods on bile acids have not yet been clarified, soy protein intake has been reported to regulate bile acid metabolism 44 .

The effects of each fermented soy food on gut microbiota are summarized in Table 1.

Summary of the effect of fermented soy foods on gut microbiota and diabetes

↑: phylum Bacteroidetes, genera Bacteroides and Lactobacillus, family Prevotellaceae NK3B31 and genus Desulfovibrio in rat. Families Prevotellaceae, Christensenellaceae, Dehalobacterium, Desulfitibacter; family Deferribacteraceae, order Deferribacterales, class Deferribacteres; and family Gemmatimonadaceae, order Gemmatimonadetes, and class Gemmatimonadales in mouse

↓: phylum Firmicutes, family Peptostreptococcaceae, genera Bifidobacterium and Turicibacter in rat. Family Microbacteriaceae, order Micrococcales, class Actinobacteria, and family Lactobacillaceae in mouse 45 , 46 , 47 , 48

Having the effect of decreasing of insulin resistance 48

↑: genera Bifidobacterium, Blautia, and Bacilli in human

↓: phylum Actinobacteria and genera Coriobacteriaceae_UCG‐002, Bacteroides, and Lactococcus in mice; class Clostridia and family Enterobacteriaceae in human 49 , 50 , 51

↑: genera Coprococcus, Bifidobacterium, and Ruminococcus in human; orders Bacillales, Lactobacillales, and Verrucomicrobiales (Akkermansia muciniphila) in rat

↓: genera Sutterella, Escherichia/Shigella, and Collinsella in human; order Enterobacterales in rat 52 , 53

↑: phylum Bacteroidetes in mice; genera Odoribacter, Akkermansia, and Lactobacillus in rat

↓: phylum Firmicutes in mice; families Ruminococcaceae and Lachnospiraceae in rat 54 , 55

↑: genera Alistipes, Lactobacillus, Faecalibaculum, Akkermansia, and Bifidobacterium in mice

↓: the phylum Firmicutes/Bacteroidetes ratio, phylum Firmicutes, genera Enterococcus and Oscillibacter, and family Deferribacteraceae in mice 56

↑: genus Bifidobacterium, Collinsella, and Prevotella in human

↓: unclassified genera of the families Ruminococcaceae and Lachonospiraceae in human 57 , 58

↑: phylum Actinobacteria in Zebrafish; Bacteroidetes, Firmicutes, Clostridium leptum in rat; Bacteroides fragilis, Bifidobacterium, Lactobacillus, Escherichia coli, Enterococcus, and Akkermansia in human

↓: phylum Proteobacteria in Zebrafish 59 , 60 , 61 , 62

Pyroglutamyl leucine, isolated from miso, has been shown to decrease the abundance of phylum Firmicutes and to increase the abundance of phylum Bacteroidetes 45 . Intake of Zygosaccharomyces sapae (strain I‐6), a probiotic yeast isolated from miso, changed the gut microbiota, such as increasing the abundance of genera Bacteroides and Lactobacillus and decreasing the abundance of family Peptostreptococcaceae and genera Bifidobacterium and Turicibacter, which activated the adenosine 5′‐monophosphate‐activated protein kinase pathway, thereby increasing the peroxisome proliferator‐activated receptor (PPAR)‐γ co‐activator‐1α and carnitine palmitoyltransferase 1α protein expression 46 . Furthermore, miso intake increased the abundance of Prevotellaceae NK3B31 and Desulfovibrio and decreased the expression of interleukin (IL)‐1β 47 .

Miso intake increases the abundance of the families Prevotellaceae, Christensenellaceae, Dehalobacterium, Desulfitibacter; family Deferribacteraceae, order Deferribacterales, class Deferribacteres; and family Gemmatimonadaceae, order Gemmatimonadetes, and class Gemmatimonadales and decreased the abundance of the family Microbacteriaceae, order Micrococcales, class Actinobacteria, and family Lactobacillaceae in a mouse model, which leads to increased SCFAs 48 .

Natto decreased the abundance of phylum Actinobacteria and genera Coriobacteriaceae_UCG‐002, Bacteroides, and Lactococcus 49 . Furthermore, natto supplementation increased the abundance of genera Bifidobacterium and Blautia 50 . The consumption of natto led to an increase in stool Bacilli and Bifidobacteria and a decrease in stool Clostridia and Enterobacteriaceae 51 .

Cheonggukjang increased the abundance of genera Coprococcus, Bifidobacterium, and Ruminococcus and decreased the abundance of genera Sutterella, Escherichia/Shigella, and Collinsella 52 . Furthermore, cheonggukjang increased the abundance of orders Bacillales, Lactobacillales, and Verrucomicrobiales (Akkermansia muciniphila) and decreased the abundance of order Enterobacterales, which led to an increase in the production of SCFAs and a decrease in the production of proinflammatory cytokines 53 .

Doenjang decreased the abundance of phylum Firmicutes and increased the abundance of phylum Bacteroidetes. In addition, it decreased the abundance of families Ruminococcaceae and Lachnospiraceae and increased the abundance of genera Odoribacter, Akkermansia, and Lactobacillus, which led to an increase of IL‐10, PPAR‐γ, and carnitine palmitoyltransferase‐1 expression and a decrease of lipopolysaccharide concentrations 54 , 55 .

One study investigated the effect of douchi, a Chinese fermented black bean, on gut microbiota. Supplementation of peptides VY and SFLLR, which are identified from douchi, decreased the Firmicutes/Bacteroidetes ratio (which was increased in those consuming the high‐fat diet); increased the abundance of family Deferribacteraceae, genera Alistipes, Lactobacillus, Faecalibaculum, Akkermansia, and Bifidobacterium; and decreased the abundance of phylum Firmicutes, genera Enterococcus, and Oscillibacter which led to an increase in the levels of SCFAs and IL‐10 56 .

Fermented soy milk increased the abundance of the genera Bifidobacterium, Collinsella, and Prevotella and decreased the abundance of unclassified genera of the families Ruminococcaceae and Lachonospiraceae 57 , 58 , which led to an increase in SCFA production.

Tempeh supplementation decreased the abundance of phylum Proteobacteria and increased the abundance of phylum Actinobacteria in Zebrafish 59 . Tempeh increased the abundance of Bacteroidetes, Firmicutes, Clostridium leptum, and Bacteroides fragilis in a rat model 60 . Tempeh supplementation also increased the abundance of Bifidobacterium, Lactobacillus, Escherichia coli, Enterococcus, and Akkermansia in humans 61 , 62 .

POTENTIAL BENEFICIAL EFFECT OF FERMENTED SOYBEAN FOODS ON DIABETES

A summary of the effects of each fermented soybean food on diabetes is shown in Table 1.

Miso exerts an antioxidant effect by improving the 1,1‐diphenyl‐2‐picryl‐hydrazyl (DPPH) radical scavenging ability 63 , 64 .

A study using an animal model reported that miso consumption suppresses fatty liver and visceral fatty obesity induced by a high‐fat diet 65 . Moreover, miso consumption suppresses insulin resistance 48 .

Cross‐sectional studies have reported that miso consumption reduces insulin resistance in individuals without diabetes 66 , 67 as well as the incidence of gestational diabetes 68 . Furthermore, cross‐sectional studies in patients with type 2 diabetes revealed better glycemic control and significantly lower relative muscle mass loss in miso soup consumers 69 , 70 .

It has been shown that natto inhibits glucose uptake by human intestinal cells 71 and that natto inhibits dipeptidyl peptidase IV 21 .

Animal study revealed that natto exerts antioxidant effects 72 and anti‐inflammatory effects by increasing IL‐10 levels 73 . A study using an animal model reported that natto consumption suppresses fatty liver 74 .

Moreover, cross‐sectional and randomized crossover studies have shown that natto intake is associated with the suppression of postprandial blood glucose and insulin resistance 75 , 76 , 77 .

Cheonggukjang has been reported to have protective effects against apoptosis, inflammation through nuclear factor‐kappa B activation, and oxidative stress through the formation of superoxide anion by removal of the tumor‐promoting factor hydrogen peroxide 23 .

Cheonggukjang extracts showed positive effects on glucose‐induced insulin secretion in β‐cells 78 , elevating the release of glucagon‐like peptide‐1 from L‐cells to increase serum glucagon‐like peptide‐1 concentrations 79 , 80 , and inhibiting lipid accumulation in 3 T3‐L1 cells 81 , 82 .

Several studies using animal models have shown that cheonggukjang intake improves blood glucose levels and suppresses body weight gain 79 , 80 , 83 , 84 , 85 . Furthermore, cheonggukjang intake improved blood lipid levels in animal models 86 , 87 .

Cheonggukjang intake improved fasting blood glucose levels in individuals with impaired fasting glucose levels 88 . Furthermore, several studies on individuals with overweight or obesity showed that compared with controls, plasma apoB levels and body fat decreased in participants who consumed cheonggukjang 89 , 90 .

Doenjang inhibited α‐glucosidase in an in vitro study 91 . Doenjang has been reported to have protective mechanisms against oxidative stress, including the inhibition of intracellular reactive oxygen species production and glutamate‐induced cytotoxicity 92 , 93 .

The increase in PPAR‐γ through changes in the gut microbiota may be associated with reduced insulin resistance 54 , 93 . Several animal model studies have shown that doenjang intake improves insulin resistance and visceral fat 94 , 95 .

It has been clarified that kochujang has antioxidant, fibrinolytic, and angiotensin‐I‐converting enzyme inhibitory activities 26 . Kochujang showed adipogenesis inhibitory activity in vitro and reduced 91 lipid accumulation through suppression of lipogenesis by downregulating sterol regulatory element‐binding protein‐1c and stimulation of lipolysis by increasing hormone‐sensitive lipase 96 .

Animal studies have shown that kochujang improves insulin resistance by suppressing hepatic glucose production and triacylglycerol accumulation and increasing glycogen accumulation 97 .

Several randomized controlled clinical trials involving individuals with obesity have shown that kochujang intake reduces blood lipid levels, body weight, and visceral fat 98 , 99 , 100 , 101 .

‘Jang’, such as cheonggukjang, doenjang, and kochujang, intake was associated with a low risk of metabolic syndrome in a cohort study 102 .

A previous study revealed that douchi has an antioxidant effect via DPPH radical scavenging activity and an increase in superoxide dismutase, glutathione peroxidase, and catalase, which are known antioxidant enzymes 103 , 104 . Douchi also has antioxidant capacity, as measured by protease and β‐glucosidase activities 105 .

Several previous in vivo animal and human studies have reported that douchi has α‐glucosidase inhibitory activity 106 , 107 , 108 , 109 , 110 , 111 . Moreover, a recent study showed that the peptides VY and SFLLR improve glucose homeostasis by activating glycogen synthase and inhibiting phosphoenolpyruvate carboxykinase and glucose 6‐phosphatase 56 .

Fermented soy milk exerts antioxidant effects by increasing nitric oxide secretion and other antioxidant enzymes and improving DPPH radical scavenging ability 29 , 112 , 113 . Fermented soy milk has also been reported to exert anti‐inflammatory effects 29 . Fermented soy milk can reduce blood glucose levels by modulating GLUT4 expression in muscles 114 .

Several studies with animal models have shown that serum lipid levels were decreased by fermented soy milk through inhibition of sterol regulatory element‐binding protein‐dependent cholesterol and triglycerides synthesis in the liver and enhancement of adiponectin signaling and PPAR‐α, fatty acid oxidation, and reverse cholesterol transport in adipose tissues 115 , 116 , 117 , 118 .

Tempeh has an antioxidant effect by improving the DPPH radical scavenging ability 119 , 120 .

Several studies using animal models have reported that tempeh improves blood glucose and lipid levels by altering the gut microbiota, leading to the inhibition of cholesterol synthesis, promotion of lipolysis, increased antioxidative capacity, and lowered reactive oxygen species levels 121 , 122 .

A prospective open‐label trial showed that tempeh consumption improves glycemic control and triglyceride levels 123 .

It has been reported that kinema exerts antioxidant effects by improving DPPH radical scavenging ability 124 , 125 , 126 , 127 . Thus, although no studies have shown an association between kinema and diabetes, kinema intake may affect diabetes.

CONCLUSION

In this study, we summarize the effects of fermented soybean foods on diabetes. Diet is associated with type 2 diabetes through changes in the gut microbiota.

Fermented soybean foods contain a variety of health‐beneficial components, and their continuous intake improves dysbiosis, which in turn promotes the production of gut microbiota‐related metabolites, including SCFAs and bile acids (Figure 1). Dietary habits may contribute to differences in gut microbiota by region of residence and race; thus, the consumption of fermented soybean foods specific to each region and race may more efficiently modify dysbiosis and alter gut microbiota‐related metabolites.

Mechanisms by which fermented soy foods affect diabetes.

There are still many unknowns regarding fermentation and fermented foods, including fermented soybean foods. In recent years, research on various fermented foods has increased worldwide, and new findings are expected to emerge in the future. We hope that this article will help readers to understand fermented foods and will be of some help in daily medical practice.

AUTHOR CONTRIBUTIONS

Yoshitaka Hashimoto: Conceptualization; Data curation; Investigation; Methodology; Project administration; Resources; Visualization; and Roles/Writing – original draft. Masahide Hamaguchi: Data curation; Supervision; Validation; Visualization; and Writing – review & editing. Michiaki Fukui: Funding acquisition; Project administration; Supervision; and Writing – review & editing. All authors have checked the final version and agree to be responsible for the work to ensure that any questions related to the accuracy or completeness of any work are appropriately investigated and resolved.

DISCLOSURE

Dr Hashimoto reports personal fees from Novo Nordisk Pharma Ltd, Sanofi K.K., Mitsubishi Tanabe Pharma Corp., Ono Pharma Co. Ltd, Nippon Boehringer Ingelheim Co. Ltd, Kowa Pharma Co. Ltd, and Sumitomo Dainippon Pharma Co. Ltd. Dr Hamaguchi received grants from Daiichi Sankyo Co. Ltd, Nippon Boehringer Ingelheim Co. Ltd, Astellas Pharma Inc., Mitsubishi Tanabe Pharma Corp., Novo Nordisk Pharma Ltd, Sanofi K.K., Takeda Pharma Co. Ltd, Sumitomo Dainippon Pharma Co. Ltd, Asahi Kasei Pharma, Kyowa Kirin Co. Ltd, and Eli Lilly Japan K.K. outside the submitted work. Prof. Fukui received grants from Taisho Pharma Co., Ltd, Mitsubishi Tanabe Pharma Corp, Novo Nordisk Pharma Ltd, Ono Pharma Co. Ltd, Kowa Pharma Co. Ltd, Sanofi K.K., Nippon Boehringer Ingelheim Co. Ltd, Daiichi Sankyo Co. Ltd, Kissei Pharma Co. Ltd, MSD K.K., Kyowa Kirin Co., Ltd, Sumitomo Dainippon Pharma Co. Ltd, Eli Lilly Japan K.K., Tejin Pharma Ltd, Takeda Pharma Co. Ltd, Nippon Chemiphar Co. Ltd, Astellas Pharma Inc., Abbott Japan Co. Ltd, Sanwa Kagagu Kenkyusho Co. Ltd, Johnson & Johnson k.k. Medical Co., and Terumo Corp., and received honoraria from AstraZeneca K.K., Taisho Pharma Co. Ltd, Ono Pharma Co. Ltd, Novo Nordisk Pharma Ltd, Sanofi K.K., Teijin Pharma Ltd, Takeda Pharma Co. Ltd, Astellas Pharma Inc., MSD K.K., Mitsubishi Tanabe Pharma Corp., Eli Lilly Japan K.K., Kissei Pharma Co. Ltd, Sumitomo Dainippon Pharma Co. Ltd, Daiichi Sankyo Co. Ltd, Mochida Pharma Co. Ltd, Kowa Pharma Co. Ltd, Arkray Inc., Abbott Japan Co. Ltd, Sanwa Kagaku Kenkyusho Co. Ltd, Kyowa Kirin Co. Ltd, Nippon Boehringer Ingelheim Co. Ltd, Medtronic Japan Co. Ltd, Bayer Yakuhin, Ltd, and Nipro Corp. outside the submitted work. The other authors declare no conflict of interest.

Registry and the registration no. of the study/trial: N/A.