Fermented Foods: Definitions and Characteristics, Impact on the Gut Microbiota and Effects on Gastrointestinal Health and Disease

Fermented foods are defined as foods or beverages produced through controlled microbial growth, and the conversion of food components through enzymatic action. In recent years, fermented foods have undergone a surge in popularity, mainly due to their proposed health benefits. The aim of this review is to define and characterise common fermented foods (kefir, kombucha, sauerkraut, tempeh, natto, miso, kimchi, sourdough bread), their mechanisms of action (including impact on the microbiota), and the evidence for effects on gastrointestinal health and disease in humans. Putative mechanisms for the impact of fermented foods on health include the potential probiotic effect of their constituent microorganisms, the fermentation-derived production of bioactive peptides, biogenic amines, and conversion of phenolic compounds to biologically active compounds, as well as the reduction of anti-nutrients. Fermented foods that have been tested in at least one randomised controlled trial (RCT) for their gastrointestinal effects were kefir, sauerkraut, natto, and sourdough bread. Despite extensive in vitro studies, there are no RCTs investigating the impact of kombucha, miso, kimchi or tempeh in gastrointestinal health. The most widely investigated fermented food is kefir, with evidence from at least one RCT suggesting beneficial effects in both lactose malabsorption and Helicobacter pylori eradication. In summary, there is very limited clinical evidence for the effectiveness of most fermented foods in gastrointestinal health and disease. Given the convincing in vitro findings, clinical high-quality trials investigating the health benefits of fermented foods are warranted.


Introduction
Fermented foods are defined as "foods or beverages produced through controlled microbial growth, and the conversion of food components through enzymatic action" [1]. Many foods have historically undergone fermentation, including meat and fish, dairy, vegetables, soybeans, other legumes, cereals and fruits. There are several variables in the fermentation process including the microorganisms, the nutritional ingredients and the environmental conditions, giving rise to thousands of different variations of fermented foods. Historically, food fermentation was performed as a method of preservation, as the generation of antimicrobial metabolites (e.g., organic acids, ethanol and bacteriocins) reduces the risk of contamination with pathogenic microorganisms. Fermentation is also used to enhance the organoleptic properties (e.g., taste and texture), with some foods, such as olives, being inedible without fermentation that removes bitter phenolic compounds.
There are two main methods through which foods are fermented. Firstly, foods can be fermented naturally, often referred to as "wild ferments" or "spontaneous ferments", whereby the microorganisms

Kefir in Gastrointestinal Health and Disease
Several studies have been carried out in humans investigating the effect of kefir consumption on gastrointestinal function and dysfunction. Kefir has been suggested to be well tolerated by people with lactose malabsorption since it contains β-galactosidase expressing bacteria (e.g., Kluyveromyces marxianus), which hydrolyses lactose, thus reducing lactose concentrations in the drink. Kefir contains 60% more β-galactosidase than plain yoghurt, while a 30% reduction in lactose content has been shown in kefir compared with unfermented milk [57]. Despite reportedly greater β-galactosidase concentrations in kefir than yoghurt, a small cross-over randomised controlled trial (RCT) in 15 people with lactose malabsorption showed that although kefir produced a significantly lower breath hydrogen concentration compared to milk, it was similar following kefir and plain yoghurt, suggesting that kefir and plain yoghurt improved lactose digestion to a similar degree [57]. Kefir also led to a significantly lower flatulence severity compared to milk, but no differences were seen for flatulence frequency, abdominal pain and diarrhoea [57]. Overall, this study suggests kefir results in lower flatulence severity than milk, and is as well tolerated as yoghurt, in people with lactose malabsorption.
Several non-randomised studies have also explored the impact of kefir in constipation [58][59][60] (Table 2). A non-randomised cross-over study in 42 hospitalised patients with constipation and mental and physical disabilities showed that 6 g of lyophilized kefir had no impact on laxative use, stool consistency and stool volume compared to control powdered milk, however the number of patients not requiring any laxatives was higher 12 weeks following the kefir intervention compared to baseline (Table 2) [59]. Another small non-randomised, uncontrolled trial in 20 people with functional constipation showed that 500 mLs of kefir for 4 weeks significantly increased stool frequency, improved bowel satisfaction score and reduced gut transit time compared to baseline [60]. Considering the small sample and limitations in study design (difference in kefir form, no randomisation, uncontrolled, limited use of validated procedures), further high-quality trials are required to establish the impact of kefir on constipation.
In a RCT, kefir led to a significantly greater increase in stool Lactobacillus concentration and in blood haemoglobin concentration in patients with Crohn's disease (n = 10) compared to control (n = 20) ( Table 2); no change was shown however in ulcerative colitis (n = 15) [56].
Another double-blind RCT investigated the impact of 500 mL/day kefir, compared to 250 mL/day milk, on Helicobacter pylori eradication rates in patients with dyspepsia and diagnosed H. pylori infection who were taking a triple antibiotic therapy for 2 weeks [61]. The study found that the rate of H. pylori eradication was significantly higher in the kefir group (78%) compared to the control group (50%; p = 0.026) [61]. The occurrence of diarrhoea, abdominal pain and nausea were also significantly lower in the kefir group compared to control, suggesting kefir may be beneficial adjunct therapy during treatment for H. pylori infections (Table 2).
Another double-blind RCT in 125 children prescribed antibiotics for upper respiratory infections examined the effect of 150 mL/day kefir for 14 days, compared to 150 mL/day heat-treated kefir, in preventing antibiotic-associated diarrhoea [62]. This study showed that kefir with live microorganisms did not improved the rates of antibiotic-associated diarrhoea compared to control (relative risk 0.82, 95% confidence interval (CI) 0.54-1.43), and no differences were found for any of the symptoms assessed, including abdominal pain, loose stools, constipation and vomiting (Table 2) [62].
There are currently no RCTs investigating the effects of kefir in functional bowel disorders. To conclude, there is evidence from RCTs demonstrating kefir may be beneficial for lactose malabsorption, and H. pylori eradication. However, an important limitation of kefir studies is that each batch may consist of different microorganisms. This may explain some of the heterogeneous findings. Further high-quality RCTs are needed to establish the impact of kefir on the gut microbiota and its impact on other gastrointestinal disorders, such as constipation. Higher breath H 2 AUC in milk compared with plain kefir (p = 0.001), plain yogurt (p = 0.001), or flavoured yogurt (p = 0.005). Higher breath hydrogen AUC in flavoured kefir compared to plain yogurt (p = 0.043) or plain kefir (p = 0.008). No difference in breath hydrogen AUC between flavoured kefir and milk (p = 0.425) or flavoured yogurt (p = 0.331).
No difference in flatulence severity and frequency, diarrhoea and abdominal pain.

Kombucha
Kombucha is a fermented tea beverage reported to have originated in Northeast China in around 220 BC and consumed extensively during the Qin Dynasty. Similar fermented tea beverages subsequently became popular in Russia and Eastern Europe [63]. In modern societies, a range of kombucha beverages are available commercially, although the microbial and metabolite composition of these products along with methods of production are rarely reported [15].
Traditional kombucha is produced through aerobic fermentation of black tea (green tea may also be used) and white sugar by a combination of bacteria and yeast, known as the symbiotic culture of bacteria and yeast (SCOBY). The yeast convert sucrose to ethanol (in addition to organic acids and carbon dioxide) which acetic acid bacteria convert to acetaldehyde and acetic acid [64]. The microbial and metabolite composition of kombucha varies according to the exact composition of the SCOBY, the type and concentration of tea and sugar [65,66], oxygen concentrations, fermentation time [67,68], temperature [67,69] and storage duration [65]. The low pH of kombucha, owing mainly to the production of high concentration of acetic acid, has been shown to prevent the growth of pathogenic bacteria such as Helicobacter pylori, Escherichia coli, Salmonella typhimurium and Campylobacter jejuni [70]. Even at neutral pH and after thermal denaturation, kombucha was able to inhibit the growth of pathogens in vitro, suggesting that compounds other than acetic acid exert antimicrobial effects [70].
Kombucha has been shown to exert effects in animal studies on blood glycaemia [72], oxidative stress [73], diabetes-induced weight loss [74], chemically-induced nephrotoxicity [75], hypercholesterolaemia [72,76] and indomethacin-induced gastric ulceration [77]. Compounds hypothesised to play a role in these beneficial effects include d-saccharic acid-1,4-lactone (DSL). This is produced by Gluconobacter during fermentation [78,79], and in rats, inhibits oxidative stress and diabetes-induced renal damage [80] and acetaminophen-induced hepatic injury [81]. Nonetheless, there is no human data on DSL to confirm this proposed mechanism of action. Polyphenol and flavonoid content of tea increases with fermentation [15]. Furthermore, in vitro superoxide radical scavenging ability, reducing ability and total phenolic compound concentration increases during kombucha fermentation [82].
Although kombucha is a rich source of acetic acid and lactic acid bacteria and yeasts [15], there are no published studies exploring the effect of kombucha consumption on the gastrointestinal microbiota composition or function in either animals or humans. Interestingly, kombucha has been shown to have antimicrobial effects in vitro [83,84]. It is currently unknown whether the proposed physiological effects of kombucha are mediated by the gastrointestinal microbiota or other direct immunological pathways.
Despite evidence of physiological effects of kombucha consumption in animals, the effects in humans remain largely unknown. A recent systematic review [85] did not identify any RCTs of kombucha on gastrointestinal disorders, including any of the functional bowel disorders.
To conclude, there are no studies of the effects of kombucha on gastrointestinal health and microbiota in humans.

Sauerkraut
Sauerkraut is one of the most common forms of preserved cabbage originating in the 4th century BC. Sauerkraut is eaten frequently in Germany, but also in other European and Asian countries and the United States [86]. Sauerkraut is produced from a combination of shredded cabbage and 2.3%−3.0% salt, which is left to undergo spontaneous fermentation, generally involving Leuconostoc spp., Lactobacillus spp., and Pediococcus spp. The low pH of the final product results in a preserved cabbage [87].
Sauerkraut (homemade and shop-bought) has been shown, through culture-dependent techniques, to contain Bifidobacterium dentium, Enterococcus faecalis, Lactobacillus casei, Lactobacillus delbrueckii, Staphylococcus epidermidis, Lactobacillus sakei, Lactobacillus curvatus, Lactobacillus plantarum, Lactobacillus brevis, Weissella confusa, Lactococcus lactis and Enterobacteriaceae [17,18,88]. Adding a starter culture of Lactobacillus casei 11MZ-5-1 produced a sauerkraut containing predominantly Lactobacillus and Lactococcus, compared to spontaneous sauerkraut which, along with Lactococcus and Lactobacillus, also contained significant Enterobacter and Pseudomonas and was more variable in microbial composition [89]. Sauerkraut has also been shown to predominantly contain Leuconostoc and Lactobacillus spp. [18,19,90,91]. Certain Lactobacillus species isolated from sauerkraut demonstrate probiotic potential, with tolerance to low pH, adherence to Caco-2 cells and antimicrobial activity against pathogens in vitro [92]. Lactobacillus paracasei HD1.7, commonly found in sauerkraut, has been shown to produce a broad-spectrum bacteriocin that may play a role in sauerkraut preservation [93].
Oral administration of sauerkraut juices in Wistar rats led to increased activity of glutathione S-transferase (GST) and NAD(P)H:quinone oxidoreductase 1 (NQO1), key liver and kidney detoxifying enzymes [94]. Certain lactic acid bacteria contained in sauerkraut generate conjugated linoleic acid [95], for which there is evidence of anti-carcinogenic and anti-atherosclerotic activity in animals [96,97]. Furthermore, Lactobacillus plantarum P2 isolated from sauerkraut significantly induced TNF-α and IL-12 expression and prevented adhesion and invasion of Caco-2 cells by Salmonella enteritidis [98]. Sauerkraut contains glucosinolate breakdown products including kaempferol, (a flavonoid) isothiocyanates, indole-3-carbinol, goitrin, allyl cyanide and nitriles [99]. The relevance of such phytochemicals to human health is unclear, however kaempferol has been shown to have radical scavenging activity, to protect from oxidative damage and to attenuate cytokine-induced reactive oxygen species in vitro [100]. Isothiocyanates have been shown to have antimicrobial properties, preventing the growth of a range of species, including E. coli, C. difficile, C. jejuni and C. perfringens [101].
Sauerkraut is one of the few fermented foods for which there is a clinical trial in functional bowel disorders. A randomised double-blind trial compared the effects of sauerkraut containing viable lactic acid bacteria (LAB) on gastrointestinal symptoms and microbiota in 58 patients with irritable bowel syndrome (IBS) of any subtype diagnosed using Rome III criteria [18]. Patients were randomised to consume 75 g/day pasteurised (control) or unpasteurised (intervention) sauerkraut containing LAB for 6 weeks. There was a significant reduction in IBS Severity Scoring System (IBS-SSS) score between baseline and end of trial in both study groups, however there was no difference in symptoms between the diet groups; 16S rRNA sequencing revealed no difference in microbiota composition between study groups or between baseline and end of trial in either group (Table 3). This may suggest that the perceived health benefit of sauerkraut is independent of the live microbes. A limitation of this study is the per protocol analysis in that only patients who completed the study (n = 34) were included in the analysis of the primary outcome. Furthermore, because there was no raw cabbage arm, it is not possible to determine whether improvement in gastrointestinal symptoms was related to the fermentation-derived products or the cabbage itself.
Another study in Chinese participants suggested larger amount of sauerkraut may in fact be associated with poor health outcomes in gastrointestinal cancers. This case-control study found that the highest compared to the lowest quintile of sauerkraut intake was associated with a greater risk of laryngeal cancer (odds ratio (OR) 7.27) [102]. One possible mechanism may relate to the high salt content of sauerkraut, although another case-control study of dietary risk factors for laryngeal cancer in China showed no associations with salt-preserved vegetables [103]. Similarly, the high potassium content of sauerkraut is thought to counter the hypertensive effects of added salt. Taking the limited evidence for sauerkraut into account, one trial indicates that both pasteurised and unpasteurised sauerkraut reduced IBS severity, this effect does not appear to be mediated by gastrointestinal microbiota. Further studies are required to elucidate the mechanisms of this effect on gastrointestinal symptoms. There is little evidence for effects of sauerkraut on other health conditions.

Fermented Soy Products (Tempeh, Natto, Miso)
The first known fermented soy products originated in China and Japan, including fermented black soybean and red fermented tofu [107]. There are many fermented soybean products from different parts of Asia, including tempeh, natto, miso, sufu, douche, soy sauce and doenjang. This review will focus on tempeh, natto and miso.

Tempeh
Tempeh is a traditional Indonesian food produced by fermenting boiled and dehulled soybeans with a starter culture of Rhizopus oligoporus fungal species at room temperature for 35-37 hours [107,108]. This produces a soft white cake with a chewy texture and mushroom-like flavour. The microbial composition of tempeh varies according to variations in production [109]. Tempeh contains lactic acid bacteria [2,110], Enterococcus faecium [110], and Rhizopus filamentous fungi. Fermentation of soybeans has been shown to reduce concentrations of protease inhibitors, phytic acid and phenols [10], antinutritional factors that are high in raw soybeans, which may relate to phytases expressed by Rhizopus species in tempeh [111].
In Sprague-Dawley rats, stool Bacteroidetes, Firmicutes, Clostridium leptum and Bacteroides fragilis abundance increased following tempeh supplementation compared to rats fed non-fermented soybeans [112]. Application of soybean and bean tempeh to human microbiota in an in vitro gut simulator model increases Bifidobacterium, Lactobacillus, Escherichia coli and Enterococcus abundance [113]. In an open-label uncontrolled study of 10 healthy human volunteers, tempeh consumption led to greater stool Akkermansia muciniphila abundance and immunoglobulin A concentrations, suggesting that tempeh may influence gut microbiota in humans [114]. However, a larger, controlled study is needed to establish the effects of tempeh on gastrointestinal microbiota composition.
Fermented soy products have been proposed to have beneficial effects on health, including purported "anti-carcinogenic", "anti-diabetic", "antioxidant", "anti-inflammatory" and "anti-hyperlipidaemic" effects, although much of the existing evidence is limited to in vitro and animal studies [107]. Tempeh has been associated in vitro with greater free-radical and superoxide scavenging ability than unfermented soybeans [115], which may relate to changes in polyphenol content and digestibility in soybeans following fermentation [116,117].
To date, there are no RCTs of the effects of tempeh consumption in humans. Proposed health effects listed above require investigation in human trials.

Natto
Natto is a traditional Japanese fermented soybean, of which Itohiki-Natto is the most commonly consumed [108]. Natto is produced through fermentation of cooked yellow soybeans with Bacillus subtilis var. natto. This produces a viscous food with a distinct flavour and strong odour [118]. Natto characteristics vary according to soybean steaming time, relative humidity, fermentation time and temperature [108]. The fermentation of Natto produces a number of bioactive factors, including nattokinase, bacillopeptidase F, vitamin K 2 and dipicolinic acid [108]. Furthermore, the quantity of the isoflavone genistein, with purported associations with metabolic and inflammatory disorders and carcinogenesis [119], is greater in Natto compared to unfermented soy products [120]. A peptide with antibacterial activities against Streptococcus pneumoniae and Bacillus subtilis has been isolated from Natto [121], with potential clinical importance in treating S. pneumoniae infections, although this has not been investigated in humans.
Nattokinase is an enzyme of the subtilisin family produced by Bacillus subtilis var. natto [122], and can be isolated from Natto [123]. Nattokinase has direct in vitro [123] and in vivo [124] fibrinolytic activity, in addition to increasing tissue plasminogen activator [125] and reducing platelet aggregation [126]. Anti-thrombotic and anti-hypertensive activities of nattokinase have been demonstrated in small RCTs in humans [127,128].
There is limited evidence on the effect of Natto on the human GI microbiota. Consumption of Natto-containing miso soup for two weeks led to increased stool Bacilli and Bifidobacteria and decreased Clostridia and Enterobacteriaceae in 8 healthy volunteers [104]. Furthermore, stool short-chain fatty acids increased, and ammonia and sulphide declined. However, it is impossible to separate the effect of the miso soup (also a fermented soy product) from possible effects of Natto. In individuals with 3-5 bowel movements per week, consumption of 50 g/day Bacillus subtilis K-2 containing Natto for two weeks resulted in greater stool frequency and proportion of stool Bifidobacteria compared to consumption of 50 g/day of boiled beans [106], although only the abstract is available and no sample size is provided.
To date, there is limited evidence from RCTs suggesting Natto might positively influence stool frequency in patients with infrequent bowel motions and influences gastrointestinal microbiota. However, these require confirmation in high quality trials.

Miso
Miso is a traditional Japanese paste of fermented soybean used to make miso soup. Miso is produced by fermenting soybeans with 'Koji', produced from a mould Aspergillus oryzae, although Saccharomyces cerevisiae and lactic acid bacteria may additionally be used. As with other fermented soy foods, miso production varies greatly in terms of ingredients, temperature and fermentation time, salt concentration and the strain of A. oryzae used.
A microbial analysis of miso at different time points following the start of fermentation revealed Bacillus subtilis, Bacillus amyloliquefaciens, Staphylococcus gallinarum and Staphylococcus kloosii to be present during fermentation, with only the Bacillus species remaining in the final product [129]. A range of miso samples have also been shown to contain Lactococcus sp. GM005, which produces a bacteriocin with strong antibacterial activity that inhibits the growth of a range of bacteria, including Bacillus subtilis, Pediococcus acidilactici and Lactobacillus plantarum [130,131].
There is little evidence on the effect of miso intake on gastrointestinal disorders. One cross-sectional study reported an inverse relationship between miso soup intake and subjective gastro-oesophageal reflux disease, functional dyspepsia and reflux scores when adjusted for other dietary factors [132]. This association was hypothesised to relate to histidine, glutamate and aspartate found in miso soup, although no animal or human studies have investigated this mechanism to date.
The high soy intakes in China and Japan have historically been hypothesised to contribute to the relatively low rates colon and prostate cancers in these countries [133]. One proposed mechanism to support this hypothesis is the high concentrations of isoflavones genistein and daidzein found in soybeans [134]. Genistein is structurally similar to oestrogen and may influence breast cancer risk through oestrogen receptor binding, which has been demonstrated in vitro [135,136]. Further in vitro studies have demonstrated that genistein may exert effects on cancer risk through promoting cell-cycle arrest [137] inducing apoptosis [138] and reducing cancer cell migration [139]. Genistein and daidzein may be higher in fermented soybean products (miso and natto) compared to unfermented products such as soy milk and tofu [120,140,141], as previously discussed. Numerous Japanese cohort studies have investigated associations between miso intake and cancer risk. These studies are limited in the assessment of dietary intake (food frequency questionnaires, often with limited intake responses) and the presence of a multitude of possible confounding factors. With these methodological limitations in mind, cohort studies have observed inverse associations between frequent miso soup intake and stomach cancer risk in Japanese men [142]. In contrast, cohort and case-control studies have shown positive correlations between frequent miso soup intake and single and multiple stomach cancers in Japanese adults [143]. Furthermore, some cohort studies have shown no association between miso soup intake and risk of various types of cancer [144,145].
There are currently no RCTs investigating the effects of miso in functional bowel disorders. There is therefore limited evidence for the effects of miso on gastrointestinal conditions and microbiota. There are several observational studies demonstrating associations between miso intake and the risk of stomach cancer, however the strength and direction of these associations remains unclear.

Kimchi
Kimchi, which originates from Korea, is a term used for a group of salted and fermented vegetables. It consists of Chinese cabbage and/or radishes, and various flavouring ingredients (e.g., chili, pepper, garlic, onion, ginger), seasonings (e.g., salt, soybean sauce, sesame seed), and other additional foods (e.g., carrot, apple, pear, shrimps) [20]. To produce kimchi, the cabbage is brined and drained, then the rest of the seasonings, spices and food products are added and mixed with the cabbage, and finally, fermentation takes place (134). The fermentation occurs spontaneously by the microorganisms naturally found on the cabbage and foods included in the mixture, although starter cultures may be used for commercial production of kimchi [20].
Prior to fermentation, the kimchi mix contains a variety of different bacterial species within the Leuconostoc, Lactobacillus, Pseudomonas, Pantoea and Weissella genera [21] (Table 1). However, once fermentation has started, the bacterial diversity decreases and the bacterial community is rapidly dominated by the genus Leuconostoc within only three days of fermentation [21]. Within this genus, Leuconostoc citreum is the most abundant species prior to fermentation, but it is present in only a minor proportion after three days of fermentation, at which time point Leuconostoc gasicomitatum and Leuconostoc gelidum become dominant [21]. As kimchi can comprise a variety of ingredients, the microbial composition varies depending on the type and amount of the foods included. For example, a higher Lactobacillus concertation has been found when kimchi contains a higher garlic quantity [146], while the addition of red pepper powder leads to higher Weissella and lower Leuconostoc and Lactobacillus proportions [147]. Several archaea (e.g., Halococcus, Natronococcus) and yeast (e.g., Saccharomyces, Candida, Trichosporon) genera have also been identified in commercially available kimchi [22] (Table 3). An animal study showed that kimchi consumption, which contained Leuconostoc mesenteroides DRC 0211, may exhibit potential weight control properties in mice via reducing hepatic mRNA expression of adipogenesis-related genes and inflammation-related monocyte chemotactic protein-1 and interleukin-6 in epididymal fat tissue [148]. Reductions in serum total cholesterol, triglycerides, low-density lipoprotein cholesterol levels and atherogenic index have also being demonstrated in rats following consumption of kimchi fermented by Leuconostoc kimchi GJ2 [149]. A human study demonstrated that consumption of kimchi fermented for 8 weeks led to changes in expression of genes related to metabolic pathways and immunity [150]. In a mouse colitis model, Lactobacillus paracasei LS2, a strain isolated from kimchi, decreased cytokine production, myeloperoxidase activity, and the number of macrophages and neutrophils in the lamina propria lymphocytes, suggesting a potential anti-inflammatory effect [151]. Anti-carcinogenic properties have also been attributed to kimchi with an in vitro study demonstrating inhibition of gastric cancer cell growth [152]. Notably, as kimchi comprises a variety of ingredients, its impact on the gut microbiota and health is thought to result from a synergic effect of the microorganisms it contains, as well as the nutrient content (e.g., phytochemicals, fibre, vitamins) of the foods used in the preparation. For example, antimicrobial and antioxidant effects have also been attributed to food constituent of kimchi, such as red pepper seeds and garlic [153,154].
Several studies have investigated the impact of kimchi on the gut microbiota. A study on diet-induced obesity murine models showed that Lactobacillus plantarum HAC01, isolated from kimchi, resulted in a higher Adlercreutzia and lower Bacteroides, Mucispirillum and Ruminococcus proportions compared to control mice [155]. In humans, a non-randomised study of 6 people in South Korea showed that consumption of 300 g/day kimchi for 4 weeks increased stool concentrations of Lactobacillus and Leuconostoc, and decreased stool pH, compared to 60 g/day of kimchi [105]. Similar findings have also been identified in several other non-randomised human studies [156,157]. In a RCT of fermented compared to fresh (unfermented) kimchi in 24 women with obesity, women who were randomised to receive 180 g/day fermented kimchi for 8 weeks showed a decrease in Blautia abundance and increases in Prevotella and Bacteroides abundance compared to baseline, but both groups (fermented and unfermented kimchi) experienced increases in Proteobacteria and Actinobacteria abundance [150]. Another RCT compared two different kimchi preparations made of different ingredients and quantities; kimchi I contained baechu cabbage, radish, red pepper powder, green onion, garlic, ginger, anchovy juice, and sugar, while kimchi II contained organic baechu cabbage, radish, red pepper powder, green onion, garlic, ginger, sugar, mustard leaf, Chinese pepper, pear, Lentinus edodes juice and sea tangle juice, mistletoe extract powder, and 10 6 CFU/g of Lactobacillus plantarum PNU [158]. Both kimchi groups resulted in an increased abundance of short-chain fatty acid producing Faecalibacterium, Roseburia, and Phascolactobacterium, and reduced Clostridium and Escherichia coli compared to baseline. Although no direct comparisons between the two kimchi groups were made, kimchi I was shown to increase the relative abundance of Actinobacteria and decrease that of Proteobacteria compared to baseline, whereas kimchi II had the opposite effect suggesting that different types and quantities of ingredients in kimchi may impact the microbiota differently [158].
With regards to the effects of kimchi on gastrointestinal health and disease, a very small study attempted to investigate its impact on H. pylori eradication. Six people with H. pylori infection consumed 300 g (high dose) or 60 g (low dose) of kimchi for 4 weeks in a non-randomised trial [105]. H. pylori infection, assessed via 13 C urea breath test, was not eradicated in any of the six participants at the end of the intervention [105] (Table 2). Kimchi has also been investigated with regards to its association with gastric cancer in epidemiological studies. A number of epidemiological studies have shown an increased risk of gastric cancer in the Korean population with higher kimchi intake (OR 2.2, 95% CI 1.3-3.8), which has been suggested to be due to its nitrite, nitrate and salt content [159,160]. However, a case control study of 136 patients diagnosed with gastric cancer and 136 healthy controls showed that different types and preparations of kimchi were associated with different levels of gastric cancer risk; for example, moderate baiechu kimchi (prepared with salted Chinese cabbage) intakes were associated with a lower gastric cancer risk (OR 0.5, 95% CI 0.3-0.9), while moderate intakes of kkakduki (prepared with salted radish) were associated with a higher gastric cancer risk (OR 2.0, 95% CI 1.03-3.8) [161]. These differences could potentially be attributed to the different food and nutrient composition and preparation methods of the different types of kimchi.
There are currently no RCTs investigating the effects of kimchi in functional bowel disorders. To conclude, there is preliminary evidence that kimchi may have an impact on the gut microbiota composition. No evidence exists to date on the impact of kimchi on gastrointestinal health and disease, while the association between kimchi consumption and gastric cancer risk warrants further research.

Sourdough Bread
The sourdough starter culture is produced through the fermentation of flour by lactic acid bacteria and yeasts, that originate from the flour and surrounding environment. Making the sourdough starter takes on average seven days and involves replenishing the microbes with fresh flour and water daily. Once the starter is ready, a small portion is added to the sourdough base ingredients to initiate the sourdough fermentation process-this method is commonly referred to as "backslopping" [24,160]. Unlike standard bread, which is produced through a rapid yeast-only fermentation process, the symbiotic sourdough fermentation of both bacteria and yeast is thought to improve bread quality, including texture, flavour, nutritional content and shelf-life, and replace additives [24]. During fermentation, microbial and enzymatic-led conversions of cereal carbohydrates, proteins, lipids and phenolic compounds occur [24,162]. The microorganisms' and enzymes' activity are interlinked; for example, lactic acid bacteria result in a pH reduction, modulating the activity of cereal enzymes and solubility of substrates (e.g., gluten), and in turn the enzymes can provide substrates to allow the growth of microorganisms [24].
The microbial content of the sourdough starter depends on the traditional practices used and, therefore, not only the taste and texture, but the nutritional profile of the final product can vary considerably [25]. In general, several species within the Lactobacillus, Leuconostoc, Weissella, Pediococcus and Streptococcus genera have been identified in sourdough starters [162]. Lactobacillus species are the most prevalent, and Lactobacillus sanfransiscensis is a key bacterium isolated from most starters [25]. Saccharomyces cerevisiae is the most abundant yeast species, followed by Candida milleri, C. humilis, Saccharomyces exiguous and Issatchenkia orientalis. Only limited data exist on the microbial composition of sourdough bread, likely due to the impact of heat during baking, with only one study showing a gene copy number of 7 to 10 log gene copies/gram of sourdough bread [163].
The mechanisms through which sourdough bread may confer health benefits is primarily through the impact that the sourdough process has on the nutritional content of bread. For example, the sourdough process can lower the bread's content of non-digestible oligosaccharides fructans and raffinose (types of FODMAPs), resulting in the bread being better tolerated by patients with IBS [11,164]. This change in the carbohydrate content occurs due to the degradation of oligosaccharides by the sourdough microorganisms, especially the yeasts Saccharomyces cerevisiae and Kluyveromyces marxianus [165]. Sourdough and its constituent microorganisms have also been suggested to exhibit anti-microbial, anti-hypertensive, and cholesterol lowering properties, however these are based on in vitro studies examining the impact of sourdough-extracted bacteria, rather than of baked sourdough bread [166][167][168].
The effect of sourdough bread on the gut microbiota has been assessed in vitro and in vivo. The impact of sourdough wheat breads fermented for different lengths of time on the human gut microbiota was assessed using in vitro batch cultures with stool samples from 3 patients with IBS and 3 healthy donors [169]. A significant increase in bifidobacteria was shown in the healthy control samples after addition of sourdough bread fermented for 8 h compared to a non-fermented bread. Significant decreases in δ-Proteobacteria and Gemmatimonadetes was shown after inoculation of sourdough bread fermented for 8 hours in both patients with IBS and healthy donors, compared to baseline (prior inoculation) [169]. In addition, sourdough bread that was fermented for 8 h resulted in significantly lower gas production after 15 h of inoculation of IBS samples, compared to a non-fermented bread and bread fermented with yeast for 16 h. The authors suggested this may indicate that this sourdough bread was fermented more slowly by the gut microbiota [169]. In vivo, however, a randomised cross-over study in 20 healthy adults showed no significant differences in stool microbiota composition when consuming 145 g wholegrain wheat sourdough bread per day for 1 week, compared to 110 g white wheat bread, with the microbiota remaining resilient throughout both bread interventions [170].

Sourdough Bread in Gastrointestinal Health and Disease
Several studies have examined the impact of sourdough bread in gastrointestinal function and disorders. In a double-blind, cross-over RCT, 17 healthy adults were randomised to consume a single meal of 2 sourdough croissants or 2 brewer's yeast croissants, followed by magnetic resonance imaging analysis of gastric emptying [171]. The total gastric volume was significantly reduced by 11% and hydrogen production by 30% following sourdough croissants compared to brewer's yeast croissants [171]. Abdominal discomfort, bloating and nausea were significantly milder, suggesting sourdough croissants are better tolerated than brewer's yeast croissants [171]. In addition, a very small randomised cross-over trial of 7 participants reporting minor gastrointestinal symptoms showed a significantly different exhaled breath volatile organic compound profile following sourdough rye bread compared to wheat bread enriched with bioprocessed rye bran, however the impact on gastrointestinal symptoms was not measured. This suggests that the potential health effects of sourdough rye bread may indeed be mediated by the gut microbiota [172] (Table 4). The utility of adjusting the sourdough process to modify the nutritional content of bread was shown in a double-blinded, cross-over RCT in 87 patients with IBS that examined the impact of a sourdough rye bread prepared using a specific sourdough system that lowers the FODMAP content of the bread versus a traditionally-made sourdough bread in gastrointestinal symptoms [11]. This study showed a significantly lower breath hydrogen level, and significantly milder flatulence, abdominal cramps, rumbling and total gastrointestinal symptoms following a 4-week consumption of a low FODMAP sourdough rye bread, compared to a traditional sourdough rye bread [11]. In contrast, a pilot study of 26 people with IBS who were randomised to a 7-day consumption of sourdough wheat bread (low in FODMAPs) or yeast-fermented wheat bread showed that not only were there no differences for any gastrointestinal symptom or inflammation markers, but, unexpectedly, symptoms of tiredness, decreased alertness and joint symptoms were significantly worse in the sourdough wheat bread group, compared to the yeast-fermented wheat bread. It is important to note however that this was a pilot study with a small sample size that used non-validated questionnaires to assess symptoms [164].
Sourdough microorganisms also contain enzymes (e.g., proteases) that hydrolyse proteins, such as gluten. As a result, studies have attempted to study the impact of sourdough bread, fermented by specific combinations of bacteria and yeasts that hydrolyse gluten into amino acids [175], in patients with coeliac disease. Eight paediatric patients with coeliac disease with normal values for total serum IgA were asked to consume 200 g/day of sweet baked products made from fermented wheat flour (containing < 10 ppm residual gluten), while continuing their usual gluten-free diet, for 60 days; six out of eight patients completed the trial, all of whom had normal IgG and IgA-AGA and IgA-tTG antibodies values at the end of the intervention [173]. Similarly, a study in 20 patients with coeliac disease who were randomised to consume sourdough wheat bread (fermented with sourdough lactobacilli and yeast proteases to hydrolyse gluten) or traditional wheat bread for 3 days, found no significant changes in INF-γ secretion in patients consuming the fermented wheat bread, whereas the consumption of traditional wheat bread mobilised INF-γ-secreting cells in blood [174]. However, the duration of the intervention was only 3 days, which is not long enough to assess its impact on inflammatory markers related to coeliac disease. An important consideration should be whether sourdough bread is indeed gluten-free. An in vitro study showed that the level of gluten degradation depends on the strain of the sourdough microorganisms used, and that fermentation of wheat flour does not sufficiently decrease transglutaminase 2 binding sites on gliadin [176]. Therefore, not all gluten is hydrolysed during fermentation and, therefore, sourdough bread made from gluten-containing flour is not considered safe for consumption in coeliac disease.
To conclude, a small study showed no impact of sourdough bread on the stool microbiota composition, while preliminary evidence of the impact of sourdough bread in managing gastrointestinal symptoms stems from low-quality studies, with study sample sizes ranging from 7 to 26 participants. Further high-quality adequately powered studies are needed in order to establish the impact of sourdough bread on gastrointestinal health.

Conclusions
In summary, there is only very limited evidence on the effectiveness of most fermented foods in gastrointestinal health, with the majority of studies being of low quality. Kefir is the fermented food most commonly investigated in terms of its impact in gastrointestinal health, with evidence suggesting it may be beneficial for lactose malabsorption and H. pylori eradication. No human studies have been conducted on the impact of kombucha, tempeh and kimchi in gastrointestinal health. It is worth noting the difficulty in undertaking and replicating fermented food studies given the significant variability of cultures and ingredients present even within food categories, which may partly explain heterogeneous findings. To conclude, there is insufficient evidence to determine the impact of fermented foods in gastrointestinal health and disease.
Author Contributions: All authors planned the structure and content of the manuscript and reviewed the literature. E.D. and S.R.C. wrote the initial manuscript. M.R. and K.W. critically revised the manuscript. All authors reviewed and approved the final manuscript.
Funding: Publication of this manuscript was made possible through an unrestricted educational grant from Danone, Paris, France to the European Society of Neurogastroenterology and Motility (ESNM), Vienna, Austria.
Conflicts of Interest: E.D. has received an education grant from Alpro, received speaker fees from Yakult and research funding from Nestec Ltd, the Almond Board of California and the International Nut and Dried Fruit Council. M.R. has received speaker fees from Ryvita, Biokult, Symprove and Alpro and research funding from the Almond Board of California and the International Nut and Dried Fruit Council. K.W. has served as a consultant for Danone, has received speaker fees from Alpro and Yakult and research funding from Clasado Biosciences, Nestec Ltd, Almond Board of California and the International Nut and Dried Fruit Council, and is the coinventor of a mobile app to support patients following the low FODMAP diet. SC reports no conflicts of interest.