Modulation of the Gastrointestinal Microbiome with Nondigestible Fermentable Carbohydrates To Improve Human Health

ABSTRACT

There is a clear association between the gastrointestinal (GI) microbiome and the development of chronic noncommunicable diseases, providing a rationale for the development of strategies that target the GI microbiota to improve human health. In this article, we discuss the potential of supplementing the human diet with nondigestible fermentable carbohydrates (NDFCs) to modulate the composition, structure, diversity, and metabolic potential of the GI microbiome in an attempt to prevent or treat human disease. The current concepts by which NDFCs can be administered to humans, including prebiotics, fermentable dietary fibers, and microbiota-accessible carbohydrates, as well as the mechanisms by which these carbohydrates exert their health benefits, are discussed. Epidemiological research presents compelling evidence for the health effects of NDFCs, with clinical studies providing further support for some of these benefits. However, rigorously designed human intervention studies with well-established clinical markers and microbial endpoints are still essential to establish (i) the clinical efficiency of specific NDFCs, (ii) the causal role of the GI microbiota in these effects, (iii) the underlying mechanisms involved, and (iv) the degree by which inter-individual differences between GI microbiomes influence these effects. Such studies would provide the mechanistic understanding needed for a systematic application of NDFCs to improve human health via GI microbiota modulation while also allowing the personalization of these dietary strategies.

The GI MICROBIOME AND CHRONIC NONCOMMUNICABLE DISEASES

Vertebrates have evolved with dense microbial populations in their gastrointestinal (GI) tract (referred to as the GI microbiome) that contribute to performance and health of the host (1). Although symbiotic in nature, animal experiments have established that the GI microbiota plays a causative role in the development of chronic noncommunicable diseases (CNCDs) such as obesity, diabetes, cardiovascular disease, colon cancer, autism, autoimmune diseases, allergies, and other atopic diseases including asthma (Fig. 1) (2). CNCDs are often associated with microbial dysbiosis, which is typically characterized by a reduced diversity, a bloom of facultative taxa (such as enterobacteria), and a lower output of beneficial metabolites (3). These associations provide a clear rationale for the development of strategies that modulate GI microbiome structure and function for the prevention of CNCDs (4).

FIGURE 1.

CNCDs that are associated with the GI microbiome and diets low in NDC. An industrialized lifestyle is associated with an increased prevalence of multiple CNCDs (15). Most of these diseases have now clearly been associated with the GI microbiome (pathology in animal models is dramatically different under germfree conditions, and the GI microbiome displays a dysbiosis in humans suffering from the disease). The Venn diagram designates CNCDs that are associated with the GI microbiome (2, 221, 247, 248) and a diet low in NDCs (13, 107, 244). NAFLD, nonalcoholic fatty liver disease.

Various strategies have been developed for the introduction of live microbes to modulate the GI ecosystem, through either probiotics, live biotherapeutics, or fecal transplants. These approaches have generated tremendous interest (5). However, diet has also been shown to readily alter GI microbiome structure and function (6–8), making it a particularly promising modifiable lifestyle factor of interest for the treatment of CNCDs. In addition, dietary supplements that employ nondigestible carbohydrates (NDCs) have been developed for several decades as prebiotics, targeted to support growth of beneficial GI microbiota in an attempt to improve health (9).

Interestingly, the incidence of most microbiota-associated CNCDs has substantially increased in recent decades in industrialized countries (10, 11), suggesting that practices associated with industrialized lifestyles predispose to disease. Although the exact factors that drive disease development are unknown and likely complex, the lack of NDCs is one contributing factor (12, 13). Humans likely evolved consuming more than 100 grams of NDCs daily, and nonindustrialized communities still exist today that have intakes that parallel those of our ancestors (14). In the 1970s, Denis Burkitt and colleagues compared industrialized (United States and United Kingdom) and nonindustrialized (rural Africa) populations and found strong epidemiological links between urbanization and a reduction of dietary NDC intake (13, 15). These lifestyle shifts were associated with increased prevalence of CNCDs (13, 15). Recent research demonstrated that a diet low in NDCs that are accessible to the GI microbiota (i.e., nondigestible fermentable carbohydrates [NDFCs]) resulted in reduced production of fermentation end products that provide important physiological and immunological functions to the host (16, 17). Moreover, a diet low in NDFCs has been implicated in the depletion of microbiome diversity observed in industrialized societies (18–22) and the loss of bacterial species that rely on them for growth (23). Therefore, a possible explanation for the rise in CNCDs is that by shifting away from a diet in which our human-microbiota interrelationship evolved, we have essentially disrupted this symbiosis, ultimately reducing or removing the evolutionary routed benefits provided by the microbes (12).

The connections described above provide a rationale for the application of NDFCs to modulate the composition and/or function of the GI microbiome to benefit host health. In this article, we discuss (i) the concepts by which the GI microbiome can be modulated through the intake of NDCs that are fermentable (or accessible to the microbes), (ii) the effects of these strategies on GI microbiota, (iii) the mechanisms by which these strategies promote health, and (iv) the future research needed to optimize these strategies for their use in the field of human nutrition. We will focus on strategies that allow the systematic increase of NDFCs in the human adult diet by means of supplements such as prebiotics, fermentable dietary fiber, and microbiota-accessible carbohydrates (MACs), and not through whole-food sources such as fruits and vegetables. Supplemental NDFCs could provide a promising avenue for targeted modulation of the GI microbiota once they are understood at a mechanistic level (24). They could also provide an applicable strategy for increasing NDFC consumption within the context of an industrialized lifestyle, since current whole-food-focused strategies encouraged by nutritional organizations have shown little success in increasing NDFC intake (25, 26). Furthermore, even though NDFCs such as human milk oligosaccharides are important for infant growth and development and have tremendous potential to be included in infant formula (27), we will focus this article on the application of NDFCs in weaned children and adults.

MODULATION OF THE GI MICROBIOME THROUGH NDFCs: THREE CONCEPTS

Prebiotics

Although the potential to modulate the human GI microbiome through NDFCs had been recognized decades earlier (especially in Japan), the term “prebiotics” was coined by Gibson and Roberfroid in 1995. A prebiotic was initially defined as a “nondigestible food ingredient that beneficially affects the host by selectively stimulating the growth and/or activity of one or a limited number of bacteria already resident in the colon, and thus attempt to improve host health” (9). The definition has been adjusted various times (28), with the International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus panel proposing the most recent definition: “a substrate that is selectively utilized by host microorganisms conferring a health benefit” (29). Currently, the most commonly used definition is “a selectively fermented ingredient that results in specific changes in the composition and/or activity of the gastrointestinal microbiota, thus conferring benefit(s) upon host health” (30). Although most definitions do not restrict prebiotics to carbohydrates, currently only lactulose, inulin-type fructans, and trans-galactooligosaccharides are considered prebiotics (30). Polydextrose, glucooligosaccharides, lactosucrose, soybean oligosaccharides, and xylooligosaccharides have been further proposed as “candidate prebiotics” (30). However, the criteria that these carbohydrates would have to fulfill to move beyond the status of mere candidates have not been clearly established (28).

To date, most definitions, including the recent definition brought forward by the ISAPP consensus panel (29), require that prebiotics have to be “specific” for or “selective” toward health-promoting taxonomic groups. Conceptually, the idea is to shift the GI microbial community toward a more “healthy” state. This concept is derived from early findings from culture-based and later probe- and primer-based studies that found bifidobacteria and lactobacilli (putatively health-promoting organisms) to be selectively stimulated through prebiotics such as inulin and galactooligosaccharides (31). However, even though there is a strong rationale to specifically target the beneficial components and functional attributes of the GI microbiome, the prebiotic concept as it currently stands has been repeatedly criticized for being poorly defined (28, 32–34) and/or scientifically outdated (28, 35, 36). Some scientists have suggested revisions to the concept (36), while others consider it completely obsolete (5, 35). Criticism is primarily focused on the concept of selectivity and the question of how to identify beneficial microbiota that should be targeted (28, 36).

First, based on the most current scientific understanding, it is too simplistic to categorize GI microbes as either “good” or “bad.” Microbes can possess both beneficial and detrimental traits, strains of one species can differ widely in their attributes, and their role is highly context-dependent (e.g., host genetic predisposition, host physiology, GI microbial ecology, diet, etc.). Second, it has been argued that the specific targets should go beyond that of Bifidobacterium and Lactobacillus species, because many bacterial genera (Akkermansia, Eubacterium, and Fecalibacterium) have been linked to health benefits (37–40), including taxa previously considered to be detrimental, such as Clostridia and Bacteroides (41, 42). Third, given that the GI microbiota functions as a complex community, it may be imperative to support community characteristics such as diversity, stability, and ecosystem functionality (e.g., short-chain fatty acid [SCFA] output), which have all been positively correlated with health (43–45). To this end, the formation of SCFAs does not rely on selective fermentation. Moreover, although some strains of bifidobacteria and lactobacilli have been reported to produce butyrate and propionate (46), which are the SCFAs with the most evidence for health effects (17), from the metabolism of amino acids, the amounts produced (<150 μM) are less than 1% of what bacteria produce from the fermentation of carbohydrates (46, 47). Bifidobacteria and lactobacilli lack the biosynthetic pathways to produce butyrate and propionate from the fermentation of carbohydrates (48–51) and therefore cannot be the target of prebiotics that aim at boosting these SCFAs. Fourth, human studies using next-generation sequencing have shown that the response of the GI microbiome to NDFCs that are currently regarded as prebiotics, such as inulin, is not as selective as previously believed (52), while NDFCs that were considered to be broadly fermented result in restricted shifts of the GI microbiome (53–55). In this respect, it is important to consider that no carbohydrate is fermented solely by one or two species (partially because bacterial traits are shared between bacteria through horizontal gene transfer [56]), and no carbohydrate is broadly fermented, especially not under the competitive conditions within the GI tract (57). So the recent proposition by the ISAPP consensus panel that selectivity “could extend to several microbial groups, just not all” (29), would essentially mean that any carbohydrate would qualify as a prebiotic.

The notion that carbohydrates currently accepted as prebiotics are not utilized differently by the GI microbiota when compared to “regular” dietary fibers has recently been demonstrated using in vitro fecal fermentations of inulin (a well-described prebiotic) and pectin (not considered a prebiotic) (58). Both carbohydrates induced multiple substrate-specific compositional shifts. The fermentation of inulin resulted in the increased abundance of five taxa, while the fermentation of pectin resulted in the enrichment of seven different taxa. Nevertheless, both carbohydrates resulted in comparable amounts of SCFAs (58). This illustrates that both inulin and pectin can lead to a specific enrichment of different bacterial species among the GI microbiome, which would allow targeted modulation of GI microbiota. In spite of this, the rationale for why only one of them is considered a prebiotic is not obvious from these findings, especially considering that selectivity, according to the ISAPP consensus panel, “could extend to several microbial groups” (29). What constitutes a specific fermentation therefore remains ill-defined.

In this respect, it is important to point out that specificity of a prebiotic has, to date, been exclusively established only by the determination of compositional shifts. However, species that utilize a prebiotic might produce metabolites without becoming enriched. For example, Bacteroides numbers often decrease after the administration of prebiotics even if they are able to utilize them (59, 60), likely since their growth is negatively affected through a reduction in pH that results from the production of SCFAs (58, 61, 62). To truly establish whether a prebiotic is selectively fermented would require the use of techniques such as stable isotope probing (63) that allow for the identification of all microbes that are able to utilize the substrate within a complex microbial community, including those that do not become enriched. Finally, there are often substantially different inter-individual responses within the GI microbiome toward a prebiotic, yet this variation has not been considered in the concept at all (36).

Overall, we agree with recently published statements that the prebiotic concept is ill defined and based on outdated scientific views (5, 28, 32–35). To address these inconsistencies, Bindels and colleagues proposed updating the definition of a prebiotic to “a non-digestible compound that, through its metabolization by microorganisms in the GI tract, modulates composition and/or activity of the GI microbiota, thus conferring a beneficial physiological effect on the host” (28). By removing the requirement of specificity, this definition embraces the complexity of host-microbe metabolic interactions and focuses on ecological and functional features of the microbiota that are more likely to be relevant for host physiology, such as the production of SCFAs. Still, consensus on the definition of a prebiotic has not been reached (32), and some scientists prefer to abandon the term altogether (35). However, it is the opinion of the authors that the term remains helpful, as it has become well known in the scientific community and is recognized by regulators, industry, consumers, and health care professionals. Therefore, the prebiotic concept could remain valuable after the inconsistencies in the definition have been resolved.

Fermentable Dietary Fiber

Based on current definitions, most prebiotic carbohydrates are dietary fibers, but not all dietary fibers are considered to be prebiotics (64, 65). Nonetheless, a clear-cut distinction between prebiotics and nonprebiotic dietary fiber is not possible (33), and it is increasingly recognized that the fermentation of dietary fiber by the colonic microbiota contributes to human health.

The term “dietary fiber” was coined in 1953 by Eben Hipsley to describe the nondigestible components of the plant cell wall (66). Later, in the mid-1970s, Trowell and colleagues refined the definition to “remnants of plant cells resistant to hydrolysis by the alimentary enzymes of man, the group of substances that remain in the ileum but are partially hydrolyzed by bacteria in the colon” (67). Since then, the definition has undergone several revisions. The most widely used definition was put forth in 2009 by the Codex Alimentarius Commission, a joint principal branch of the Food and Agriculture Organization and the World Health Organization, which defines dietary fiber as “carbohydrate polymers with ten or more monomeric units, which are not hydrolyzed by the endogenous enzymes in the human small intestine and belong to the following categories:

1. Edible carbohydrate polymers naturally occurring in the food as consumed,

2. Carbohydrate polymers which have been obtained from raw materials by physical, enzymatic, or chemical means and which have been shown to have a physiological effect of benefit to health as demonstrated by generally accepted scientific evidence to competent authorities, and

3. Synthetic carbohydrate polymers which have been shown to have a physiological effect of benefit to health as demonstrated by generally accepted scientific evidence to competent authorities” (68).

The decision to consider carbohydrates with three to nine monomeric units (oligosaccharides) as dietary fiber was left to individual country’s authorities (26). Recently, the Food and Drug Administration in the United States updated its food labeling regulations to model the Codex definition by requiring that “isolated” and “synthetic” NDCs added to foods must first demonstrate a beneficial physiological effect in humans prior to being permitted as a dietary fiber on a food label (69).

Dietary fibers can be found in plants, bacteria, and fungi and can be chemically synthesized (Fig. 2) (70, 71). Plant-derived nonstarch polysaccharides display a substantial variety of chemical structures due to the diverse functional roles that they play in plants (71). Dietary fibers also include starches that are resistant to human digestion (resistant starches [RSs]), which are divided into five subtypes based on the mechanism responsible for their inaccessibility to host digestion (72). Given the heterogenic chemical structures found in dietary fibers, their utilization within the GI tract often requires a diverse array of enzymes distributed among various microbial members. The extent by which different fiber types are utilized or fermented by the GI microbiota is therefore structure-dependent and relies on the metabolic capabilities of an individual’s microbiome, which ultimately determines the bacterial metabolites produced from the fermentation (24).

FIGURE 2.

Categories of NDCs. NDCs are a heterogeneous group of compounds that display diverse chemical structures, which is the basis for their categorization alongside their origin (24, 70, 71). Nondigestible oligosaccharides are NDCs composed of three to nine monosaccharides and are from either plant or animal origin, as well as chemically synthesized.

The chemical structure of a dietary fiber also determines other important physicochemical properties such as solubility and viscosity (73), which influence its accessibility to microbes. Most linear fibers form crystalline structures (such as cellulose), which significantly reduces their solubility in water. Meanwhile, dietary fibers with charged moieties or structural irregularities in the sugar backbone and side chains (such as β-glucan) tend to increase in solubility, which is often correlated with fermentability (74). However, one common exception is RSs, which are typically insoluble in water yet highly fermentable (33). Furthermore, most soluble dietary fibers tend to also be viscous in water (33), which relates to their ability to render a solution gelatinous via the absorption of water (74). Solubility and viscosity are important characteristics that influence the functionality of a dietary fiber, but for the purpose of modulating the GI microbiota, fermentability is of particular relevance.

“Dietary fiber” is a widely accepted and useful term to describe NDCs that influence health benefits. Most consumers recognize the term, because it is a required part of the nutrition facts label of processed food. However, to describe NDFCs that are intended to modulate the GI microbiome, the term has obvious limitations. Various mechanisms (e.g., reduced absorption through viscosity, bile acid binding, stool bulking, etc.) have been identified by which dietary fiber can benefit human health completely independently of their effects on the GI microbiome (73). In addition, as described above, not all dietary fibers are fermented by the microbiota. The definition of a prebiotic proposed by Bindels and colleagues would encompass dietary fibers for which there is evidence that they improve health via the GI microbiota (28), but as described above, there is no agreement on the definition of a prebiotic (32). What is clear is that the portion of dietary fiber that is fermentable or “accessible” by an individual’s microbiota is what determines its ability to modulate the GI microbiome. This notion is central to the concept of microbiota-accessible carbohydrates (MACs).

MACs

To address the limitations of the concepts discussed above, especially as they relate to the inter-individual differences of the human GI microbiome, Sonnenburg and Sonnenburg introduced the concept of MACs (16). MACs can be divided into dietary MACs (NDFCs such as prebiotics and dietary fiber) and host-derived MACs (such as mucosal glycans) (16). Determining whether an NDFC is considered a MAC is not entirely dependent on the NDFC’s physicochemical characteristics, because an individual’s GI microbiota must also have the enzymatic capacity to metabolize it (75, 76). For example, cellulose does not qualify as a MAC for humans, because the capacity of the human GI microbiota to ferment cellulose is extremely low (77), while it would qualify as a MAC for other host species (e.g., hindgut and foregut fermenters). On the other hand, RS type 3 would be considered a MAC for most individuals, as it is readily metabolized by their GI microbiota. However, individuals who lack the keystone species Ruminococcus bromii do not have the enzymatic capacity to metabolize RS type 3, and thus for these individuals, it would not be considered a MAC (75). The concept of MACs is therefore particularly applicable to efforts aimed at personalizing human nutrition, which could customize dietary recommendations toward the goal of incorporating specific NDFCs that are known to be accessible to an individual’s GI microbiota.

The concept of dietary MACs is essentially equivalent to that of fermentable dietary fiber, with the additional criterion regarding the individuality of the GI microbiota in their capacity to utilize certain NDFCs. Both concepts include the same types of NDFCs (16), and according to Bindels and colleagues, these carbohydrates could be considered prebiotic if they exert a health benefit via the GI microbiota (28). Because there is currently no unifying concept that describes NDFCs that target the GI microbiota for health purposes, we will continue to refer to these carbohydrates as NDFCs for the purpose of this review.

HOW DO NDFCs MODULATE GI MICROBIOTA COMPOSITION AND FUNCTION?

Modulation of GI Microbiota Composition and Diversity

Consumption of NDFCs has the potential to improve human health by changing both the composition (structure and diversity) and the function (metabolism) of the microbial communities that reside in the GI tract. Dietary administration of NDFCs alters the nutritional niches in the GI tract by providing substrates for microbial growth. Thus, in general, species that are able to utilize these substrates can expand their populations (78). For example, administration of RS has been shown to enrich specific bacterial groups (Bifidobacterium adolescentis, R. bromii, and Eubacterium rectale) in a subset of individuals (53–55). The taxa shown to be enriched differ between RS type 2 (which appears to be similar to RS type 3 [75]) and RS type 4 (55), indicating that shifts are dependent on the carbohydrate’s chemical structure. Accordingly, the consumption of galactooligosaccharides mainly induces Bifidobacterium species that possess the enzymatic machinery to efficiently utilize this substrate (60). In addition to its enzymatic capacity, the ability of a microbe to “adhere” to a substrate and to tolerate the environmental conditions generated from fermentation (e.g., low pH) determines whether a microbe can become enriched. Although RS is utilized by many members of the human GI microbiota (57, 79, 80), the species that become enriched under competition (B. adolescentis, R. bromii, and E. rectale) have been shown to directly adhere to this substrate in the human GI tract (80).

Furthermore, the degradation of many complex NDFCs does require a different species with complementary enzymatic repertoires for their degradation, which establishes syntrophic interrelationships within the GI microbiota (81–83). Primary fermenters directly degrade NDFCs, leading to the release of partial breakdown products and the production of metabolites that can benefit themselves, as well as be beneficial or inhibitory (e.g., through acidity) to other taxa (84). Secondary fermenters benefit through cross-feeding on these partial carbohydrate breakdown products and the metabolic end products released by primary fermenters (82, 85). Through coculture experiments, R. bromii has been identified as a primary fermenter of RS types 2 and 3, whereupon RS degradation-reducing sugars are released that support the growth of secondary fermenters that are not able to degrade RS directly, including B. adolescentis, E. rectale, and Bacteroides thetaiotaomicron (75). Another cross-feeding relationship exists for hydrogen-consuming species such as Blautia hydrogenotrophica, which flourish in the mouse GI tract only if in bi-association with B. thetaiotaomicron by utilizing the hydrogen generated as an end product of fermentation by the latter (86). Removal of hydrogen maintains the redox balance (NAD⁺/NADH ratio) in the GI tract, providing conditions favorable for fermentation, and can result in the formation of methane by methanogens (such as Methanobrevibacter smithii) and acetate by acetogens (such as B. hydrogenotrophica), which is further converted to butyrate by other taxa, including Roseburia intestinalis (87–89). Lactate is another metabolic end product of NDFC fermentation that can also be converted to butyrate through cross-feeding between lactate-producing species and lactate-utilizing butyrogenic species (83, 90).

While the processes described above lead to the enrichment of GI microbes through the fermentation of NDFCs, with different species being either directly or indirectly stimulated, there are taxa that simultaneously become inhibited through the metabolites produced. For instance, Bacteroides species often decrease in number after the administration of NDFCs (37, 59, 60), even though this group of bacteria is well equipped to utilize these substrates. These inhibitory effects are due to Bacteroides having a low tolerance for acidic conditions generated from the SCFAs produced during the fermentation of NDFCs (58, 62).

The impact of NDFCs on the GI microbiome composition displays several consistent characteristics that are important for therapeutic applications. First, the magnitude of the induced changes can be substantial, with specific species becoming enriched to constitute more than 30% of the fecal microbiota (54, 55, 60), thus providing a potential strategy for the enrichment of minority members of the GI microbiome to become dominant members. However, these changes are only maintained as long as the substrate is consumed. Once the substrate is no longer available, resilience of the microbiota results in a return to the original state. Second, the microbial response to NDFCs is highly individualized, with some individuals showing substrate-specific shifts, while others do not respond at all (55, 60). The reason for this individuality is not yet understood. Individuals might lack keystone species (75) or contain strains with varying enzymatic capacity for the substrate (91). Third, although individualized, compositional changes observed after administration of NDFCs remain restricted to certain groups of microbes. This is true for classic prebiotics such as inulin and galactooligosaccharides that are supposed to be selectively fermented (60, 92), but it also applies to substrates that were assumed to be broadly utilized, such as RSs and pectin (54, 55, 58). The reason for these observations stems from the highly competitive conditions within the human GI tract, which allow for only certain microbes to benefit directly from the NDFCs (87). Although central to the original prebiotic concept (30), whether such specific shifts are related to health outcomes still remains to be established.

Most CNCDs are associated with a dysbiosis that displays decreases in bacterial diversity and/or genetic richness of the GI microbiome (3, 43, 44). Although it is difficult to prove whether these patterns are the cause or an effect of disease, community ecology theory postulates high diversity as a beneficial trait that is attributed to stability and functionality of the ecosystem (45, 93). Several independent research groups have consistently shown that individuals from nonindustrialized populations in various parts of the world have greater GI microbiome diversity when compared to individuals from industrialized regions (94). This increased diversity is reflected not only in the number of bacterial species (or operational taxonomic units) (18–22), but also in the abundance of genetic functions encoded in the GI microbiome (20). While there are many possible factors that could cause reduced GI microbial diversity through industrialization (e.g., sanitization and antibiotic use), dietary diversity is considered a key mediator (16, 95). Research using humanized mice (mice that have been colonized with a human GI microbiota) demonstrated that a maternal diet low in MACs induced significant depletions in the GI microbiota diversity of the offspring within only a few generations and that this depletion was irreversible even after the reintroduction of MACs (23). Enrichment of the maternal diet with MACs maintained the GI microbiota diversity in the offspring over multiple generations (23). Moreover, the transition of non-human primates from wild to semicaptive to captive conditions led to a reduction in bacterial operational taxonomic unit richness. An in-depth analysis of the chloroplast sequences found within the 16S sequences obtained from fecal samples suggested that this depletion was in part driven by a reduction in the diversity of dietary plant content, and particularly by a reduction in the consumption of NDFCs (96).

Overall, there is convincing evidence that a depletion of NDFCs in the diet results in a reduction of microbial diversity within the human GI microbiome, thus providing a rationale for targeting microbial diversity through dietary modulations. An enrichment of both the amount and structural diversity of NDFCs could in theory enhance microbial diversity and gene richness by generating niche opportunities (24). Cross-sectional assessments of long-term dietary intake in overweight humans have shown that long-term consumption of fruit and vegetables, and therefore NDFCs, is associated with higher GI microbial gene richness and diversity (44, 97). Conversely, short-term dietary intervention studies have produced conflicting results (37, 98, 99). Studies supplementing fiber-rich whole foods have been shown to increase GI microbial diversity (37, 93), while most short-term feeding trials with purified NDFCs (55, 60, 98, 100) or even whole-plant-based diets (6) had no effect.

In summary, although clear associations exist between the consumption of NDFCs, GI microbial diversity, and improved metabolic and inflammatory markers of CNCDs, rigorously controlled human intervention studies with NDFCs that assess well-defined clinical and microbial endpoints are needed to determine (i) if diversity can be enhanced by NDFCs and (ii) to what extent this constitutes a microbial-dependent mechanism by which NDFCs can improve human health.

Impact of NDFCs on GI Microbiome Function

The provision of NDFCs not only impacts microbiota composition as described above, but also changes the profile of microbial-derived metabolites within the GI tract (101). The fermentation of NDFCs results in the production of beneficial metabolites (e.g., SCFAs) and microbial gases (H₂, CO₂, CH₄), with SCFAs being the main focus of recent research (17, 101). Acetate, propionate, and butyrate are the dominant SCFAs (at >95%) (102). A significant portion of acetate (around 24%) can support the production of butyrate through microbial cross-feeding (103). The total amount and proportion of individual SCFAs produced is dependent on the type of NDFC (104), as well as the individual microbiota (105), further reinforcing that the response of human GI microbiota to NDFCs is individualized.

Fermentation of NDFCs within the GI tract leads to various systemic effects on the host, including an influence on energy homeostasis and metabolism (17). A majority of the SCFAs produced are rapidly absorbed in the GI tract, with only around 5 to 10% being excreted in the feces (101). Upon absorption, the majority of butyrate is metabolized by the colonocytes and serves as their major form of energy (106). Propionate reaches the liver via portal circulation, where it is primarily utilized for hepatic gluconeogenesis (GNG). Acetate, on the other hand, reaches peripheral circulation at extensively higher concentrations than the other SCFAs, where it is metabolized by peripheral tissues for energy, in addition to being utilized by the liver for lipogenesis (101–103). NDCs provide the host with 0 to 2.5 kcal/g (with digestible carbohydrates providing 4 kcal/g), dependent on their level of fermentability by the GI microbiota (107).

A low intake of NDFCs leads not only to a reduction in SCFAs, but also to shifts in GI microbiota metabolism toward the utilization of less favorable nutrients, particularly dietary and endogenously supplied proteins (108). For instance, moving the diet of human volunteers from a weight-maintenance diet to a high-protein, low-carbohydrate diet not only significantly reduced the production of total SCFAs and butyrate (109), but also led to an increase in potentially detrimental metabolites derived from the fermentation of amino acids, including branched chain fatty acids, ammonia, amines, N-nitroso compounds, phenolic compounds including p-Cresol, and sulfides (Fig. 3) (110, 111). These metabolites are thought to directly contribute to the development of CNCDs, particularly colon cancer (111). In addition, depletion of NDFCs within the diet subsequently causes the GI microbiota to shift their glycan-foraging behavior toward utilizing host-derived substrates such as mucins by upregulating the expression of bacterial genes necessary for the metabolization of mucosal glycoproteins (112). This shift toward the fermentation of mucosal glycans leads to a significant depletion of the epithelial mucus layer (113), which can cause GI inflammation and increases the host’s susceptibility to pathogen invasion (114).

FIGURE 3.

Mechanisms by which the metabolism of NDFCs by the GI microbiota modulates host health. NDFCs are fermented by the GI microbiota to SCFAs, which upon absorption into enterocytes can activate intestinal GNG, leading to improved satiety and glucose homeostasis. SCFAs can further stimulate enteroendocrine L-cells to secrete PYY, GLP-1, and GLP-2. Both PYY and GLP-1 act as satiety hormones, while GLP-1 also promotes glucose tolerance. Meanwhile, the secretion of GLP-2 enhances intestinal barrier function by upregulating the expression of tight junction proteins. SCFAs further enhance the intestinal barrier by stimulating mucin secretion from goblet cells, which aids in reducing the translocation of LPS through the intestinal epithelium, consequently reducing inflammation. Additionally, SCFAs exert immunomodulatory effects by regulating the production of antimicrobial peptides, the expansion of regulatory T-cells, and myeloid cell function to inhibit inflammation. Moreover, SCFAs signal to organs distant from the GI tract, such as white adipose tissue, where they may act on adipocytes promoting the secretion of leptin, another anorectic hormone. Furthermore, the presence of NDFCs inhibits the production of potentially detrimental metabolites from the fermentation of dietary proteins through lowering intestinal pH. AMP, antimicrobial peptides, BCFAs, branched-chain fatty acids; CVD, cardiovascular disease; GLP, glucagon-like peptide; GNG, gluconeogenesis; LPS, lipopolysaccharides; PYY, peptide tyrosine tyrosine; SCFAs, short-chain fatty acids; T2D, type 2 diabetes; Tregs, regulatory T-cells.

In summary, the metabolic effect of NDFCs is central to their importance in human nutrition and their effects when used as supplements (115) in the form of prebiotics or fermentable dietary fiber (65). Although the fermentation of NDFCs is considered beneficial, this subject is not without its controversies. Individuals with obesity tend to have increased fecal SCFAs when compared to their lean counterparts (116), and SCFAs might contribute to weight gain by providing energy. In addition, butyrate has a controversial role in the induction of colon cancer, because it has been shown to fuel the hyperproliferation of colon epithelial cells in a mouse model of the disease (117, 118). However, epidemiological studies consistently report a negative association between dietary fiber consumption, which would increase SCFAs, and both obesity (119) and colon cancer (120, 121). Furthermore, as discussed below, the majority of the physiological effects associated with SCFAs are considered beneficial.

PHYSIOLOGICAL EFFECTS OF NDFCs ON THE HOST

Two primary signaling mechanisms have been described by which SCFAs influence the biological responses of the host. First, SCFAs can impose epigenetic regulation through direct inhibition of histone deacetylase activity and expression (122). Histone deacetylase inhibition has been indicated as a central mechanism by which SCFAs modulate the immune system and inhibit the development of colon cancer (122). Second, SCFAs can bind to G-protein-coupled receptors (GPRs), with the primary receptors activated by SCFAs being GPR41, GPR43, and GPR109A (122). GPR41 and GPR43 are coexpressed locally on colonic enteroendocrine L-cells (46, 47), as well as systemically expressed in white adipose tissue, skeletal muscle, and the liver (46–48). GPR109A (which is also commonly referred to as niacin receptor 1 since niacin is its primary ligand [123]) has been shown to be expressed on ileal and colonic enterocytes, adipocytes, and immune cells (123–125). Agonists for GPR41 can be ranked in the following order based on potency: propionate greater than or equal to butyrate, butyrate greater than acetate; however, these SCFAs exhibit similar potencies for GPR43 (126, 127). Butyrate, on the other hand, is the only SCFA known to bind to GPR109A. Overall, SCFAs are a primary microbiota-dependent mechanism by which NDFCs modulate host health (17) through the regulation of satiety, glucose and lipid metabolism, as well as systemic inflammation (Fig. 3).

Regulation of Satiety

The regulation of the balance between hunger and satiety, which ultimately impacts energy intake, is highly complex and is influenced by multiple physiological, psychological, and environmental factors (128). SCFAs act as physiological regulators of satiety by primarily functioning as signaling molecules through the enhanced production of key anorectic hormones such as peptide tyrosine tyrosine (PYY) and glucagon-like peptide-1 (GLP-1) (Fig. 3) (47). By means of activating GPR41 and GPR43, SCFAs induce the release of both PYY and GLP-1 from colonic enteroendocrine L-cells into systemic circulation (129, 130). PYY has been shown to promote satiety by acting on the arcuate nucleus within the hypothalamus to suppress neuropeptide Y neurons and activate pro-opiomelanocortin neurons while also delaying gastric emptying (131, 132). GLP-1 similarly influences the hypothalamus by binding to the GLP-1 receptor (133, 134) while also inhibiting gastric emptying and the secretion of gastric acid (135, 136). Furthermore, the SCFAs acetate and propionate have been shown to act on white adipose tissue to stimulate the production of leptin, another anorectic hormone involved in the regulation of satiety (Fig. 3) (137, 138).

SCFAs can also regulate the interplay between hunger and satiety independent of anorectic hormones. Acetate has been shown to induce satiety through directly eliciting hypothalamic appetite suppression (139). Furthermore, propionate and butyrate promote an upregulation of intestinal GNG gene expression, which positively influences energy homeostasis and promotes satiety by portal vein glucose sensors (140, 141). Besides influencing energy intake, SCFAs may also affect energy expenditure. In rodents, SCFA supplementation promoted an increased rate of oxygen consumption while also enhancing mitochondrial function, adaptive thermogenesis, and fat oxidation (142, 143). However, the extent of this influence and its relevance to humans is not currently known.

Glucose and Lipid Metabolism

Obesity, cardiovascular disease, type 2 diabetes, and other CNCDs are associated with an altered glucose and lipid metabolism (144, 145). Human intervention studies have shown that increased consumption of NDFCs can improve glucose and lipid metabolism (146–148), thereby providing a mechanism by which they reduce the risk of developing cardiovascular disease, type 2 diabetes, and other CNCDs (149, 150). As discussed above, activation of enteroendocrine L-cells by SCFAs stimulates the release of GLP-1, which directly acts on pancreatic β-cells to promote insulin and inhibit glucagon secretion (133, 151, 152). GLP-1 is also known to improve β-cell responsiveness to glucose, even in glucose-resistant β-cells (153, 154). Additionally, SCFAs themselves, specifically propionate, have recently been shown to act directly on β-cells to stimulate insulin secretion, independent of GLP-1 (155). This would subsequently increase the uptake of glucose by skeletal muscle and adipose tissues while also decreasing hepatic-associated GNG (156). Further, SCFAs have been shown to inhibit hepatic-associated GNG via upregulation of intestinal GNG (140). This reduction in hepatic GNG is critical, because enhanced hepatic production of glucose is linked to insulin resistance and the development of some CNCDs (157, 158).

SCFAs may also directly act on the liver to further influence hepatic glucose and lipid metabolism, which may in part be mediated by GPR41 and GPR43 signaling (127). Dietary administration of SCFAs decreased lipid accumulation within the liver in mice by means of increased lipid utilization (159). This occurred primarily by downregulation of peroxisome-proliferator-activated receptor-γ expression and activity, which stimulated AMP-activated protein kinase-associated fatty acid oxidation within the liver. This SCFA-induced increase in fat oxidation was also observed in adipose tissue (159). SCFAs may play a further role in lipid metabolism by acting on adipose tissue through both intracellular and extracellular mechanisms. The acute administration of acetate and propionate in humans led to a significant reduction in serum free fatty acid levels (160). This may be in part due to a GPR43-dependent decrease in the intracellular lipolytic activity of adipocytes (161). In addition to this, propionate may also have extracellular lipolytic properties by enhancing the activity of adipose tissue lipoprotein lipase (162). Collectively, the mechanistic involvement of SCFAs in modulating glucose and lipid metabolism has been well established in animal and in vitro models, although further research is needed to clarify if these findings translate to humans.

Systemic Inflammation

Most CNCDs are characterized by a state of systemic, low-grade inflammation, which contributes to disease progression (144, 145, 163). Although the findings are to some degree variable, recent research indicates that NDFCs have anti-inflammatory effects (164, 165), primarily through SCFA-dependent mechanisms involving intestinal barrier function and regulation of immune cell responses (17).

Barrier function and endotoxemia

Diminished GI barrier function enhances the translocation of lipopolysaccharide (LPS) and other microbial-derived proinflammatory molecules across the GI epithelial layer. LPS interacts with LPS-binding protein and CD14 to stimulate Toll-like receptor-4 and subsequently promotes a proinflammatory response, including a systemic increase in acute-phase proteins and proinflammatory cytokines (166). Elevated plasma LPS in humans (endotoxemia) is positively correlated with percent body fat, excessive energy intake, metabolic inflammation, insulin resistance, and dyslipidemia (37, 167–169). In mice, endotoxemia has been demonstrated to cause obesity and insulin resistance without influencing food intake (170). Considering this, promoting GI mucosal barrier function could constitute a strategy to address CNCD-associated metabolic abnormalities.

SCFAs have been shown to enhance GI barrier function through upregulating the expression of tight junction proteins, which occurs through at least two mechanisms. First, NDFC-induced SCFA production increases GLP-2 secretion by enteroendocrine L-cells (171, 172), which leads to an upregulation in the expression of tight junction proteins (zonula occludens-1 and occludin) and, concomitantly, a reduction in LPS and systemic inflammation (Fig. 3) (173). Recently, Kelly and colleagues illustrated another mechanism by which SCFA metabolism (mainly butyrate) within intestinal epithelial cells creates a state of hypoxia, which enhances intestinal barrier integrity (174), likely through enhanced tight junction protein expression (175).

Other mechanisms by which SCFAs have been shown to improve GI barrier function include butyrate-driven upregulation of epithelial cell proliferation and differentiation (176), enhanced production of antimicrobial peptides including secretory immunoglobulin A (IgA) and intestinal alkaline phosphatase (177), stimulation of goblet cells to secrete mucus that fortifies the GI mucus layer (178, 179), and GPR109A-dependent regulation of enterocyte-derived interleukin-18, which is essential to maintain mucosal homeostasis and ultimately GI barrier function (180, 181).

Immunoregulation

Subpopulations of immunosuppressive regulatory T-cells (Tregs) play an essential role in both the maintenance of immunotolerance and, ultimately, the prevention of many CNCDs (182, 183). Microbiota-derived SCFAs, particularly butyrate and propionate, have been demonstrated to encourage immunotolerance through expansion and differentiation of Tregs (182). This process has been shown to occur through GPR43-dependent signaling (184), as well as through the epigenetic regulation of the Foxp3 promoter through inhibition of histone deacetylase (185, 186). This systemic immunoregulatory effect is thought to be determined largely by the immunologic context. For example, in a state of infection, SCFAs enhance the differentiation of proinflammatory T-helper cell subsets (i.e., Th1 and Th17 cells) instead of Tregs (187). This suggests that SCFA production from the fermentation of NDFCs may play an intricate role in host immune responses to infection beyond lowering intestinal pH, which inhibits colonization by pathogens (9).

The immunoregulatory effect of SCFAs is, however, not solely reliant upon an expansion of Tregs and can occur in organ systems throughout the body (e.g., lungs, skin) (188, 189). SCFAs have also been shown to regulate immune cells of the myeloid lineage (Fig. 3) through facilitating the polarization of macrophages toward an M2 phenotype involved in downregulating an inflammatory response (190), specifically that of GI macrophages to LPS (191). Moreover, SCFAs modulate neutrophil activity through suppressing their migratory behavior (192) while shifting their microbial killing phenotype toward increased phagocytic activity with diminished proinflammatory cytokine production (193, 194). The activity of dendritic cells has similarly been shown to be influenced by SCFAs, primarily through enhancing their ability to induce the differentiation of Tregs (180) and to promote the production of IgA by plasma cells (195). Furthermore, SCFAs act directly on β-cells to enhance the production of IgA, both by acting as an energy source and by upregulating the expression of genes necessary for plasma cell differentiation (196). SCFAs can also act on nonimmune cells to stimulate the release of antimicrobial peptides. Within the pancreas, SCFAs act on pancreatic β-cells in a GPR-dependent manner to enhance the secretion of cathelicidin-related antimicrobial peptide, which ultimately induces Treg cell expansion and protects against the development of autoimmune diabetes (197). Although additional research is needed, these animal and in vitro models elegantly illustrate the central mechanisms by which microbiota-derived SCFAs reduce inflammation.

EVIDENCE FOR MICROBIOTA-MEDIATED HEALTH EFFECTS OF NDFCs

There is substantial evidence from well-designed animal studies that NDFCs, or the metabolic products that result from their fermentation (e.g., SCFAs), have beneficial effects and that these effects are in part due to the GI microbiota. Work originating from Patrice Cani and Nathalie Delzenne’s research groups has repeatedly demonstrated, although primarily through associations and not causal assessments, that the GI microbiota is implicated in the ability of NDFCs to improve obesity-associated low-grade inflammation and insulin resistance (40, 171–173, 198). Research employing fecal microbiota transplantation methodology supports these findings and further suggests a causative link, in that upon transferring the GI microbiota from mice fed an NDFC (resistant maltodextrin) to antibiotic-treated db/db mice (a model of type 2 diabetes), the improved glucose homeostatic phenotype was likewise transferred (199).

Work with GPR41 and GPR43 knockout (KO) mice, or their specific receptor antagonists, provides additional evidence that the NDFC-associated metabolic health effects described above are mediated by the GI microbiota and specifically through the production of SCFAs. For instance, the ability of SCFAs to stimulate insulin secretion through increased GLP-1 release (200, 201), promote satiety through enhanced PYY secretion (202), reduce systemic free fatty acid levels (161), or regulate blood pressure (203) is lost in GPR41 and/or GPR43 KO mice. Moreover, the combination of GPR41 KO mice with a model of allergic asthma showed that the production of SCFAs was essential for NDFCs to protect against allergic airway inflammation, which occurred by impairing the ability of lung dendritic cells to promote allergen-reactive Th2 responses (188). A similarly designed study compared the effect of guar gum, a viscous NDFC, in wild-type, GPR43 KO, and GPR109A KO mice treated with dextran sulfate sodium, which induces colitis. The study showed that signaling through GPR43 and GPR109A was essential for the NDFC to protect against the development of colitis (204).

Although these animal studies do illustrate a clear causative link between NDFCs, host health, and GI microbiota composition and function (e.g., SCFAs), knowledge of the exact role of the GI microbiota in the health effects of NDFCs remains incomplete. One experimental approach by which this could be resolved is through the comparison of the physiological effects of NDFCs in conventionalized and germfree animals (28). However, germfree animals are not without limitations since there are important physiological features that differ from their conventional counterparts, including an immune system that is not fully developed (205, 206). Furthermore, the pathological progression of CNCD-like symptoms in several animal models is often vastly different under germfree conditions, preventing direct comparison between germfree and conventional animals (207, 208). To address these limitations, one could alternatively test if the phenotypic effects of NDFCs are able to be transferred through cohousing or fecal microbial transfers. This is based on the premise that if the beneficial effect of the NDFC is caused by a shift in the microbiome, then the same beneficial effect should be seen in the recipient animal (28).

In humans, most microbiota-associated CNCDs are also linked to a low intake of dietary NDFCs, with a substantial degree of overlap between the two (Fig. 1). This suggests that NDFCs may prevent CNCDs through modulation of the GI microbiome (209). Epidemiological studies have established convincing associations between dietary fiber intake and health (149, 150, 210), and although findings from human intervention studies are less consistent (210), health claims have been approved for dietary fibers (cancer and cardiovascular disease) (211) and prebiotics (constipation) (212). Moreover, multiple human intervention studies have been conducted that assess the effect of NDFCs on well-defined clinical outcomes while also characterizing the microbiome for compositional and functional signatures that correlate with these health outcomes (37, 52, 98, 213–216). For instance, treatment with the prebiotic inulin led to an increase in both Bifidobacterium and Faecalibacterium prausnitzii, which was inversely correlated with serum LPS levels, indicating that the prebiotic enhanced GI barrier function through modulating the GI microbiota (52). Furthermore, systematic meta-analyses have shown that prebiotic supplementation is capable of restoring bowel function (217), while also reversing multiple metabolic abnormalities associated with CNCDs, including reducing fasting insulin, triglycerides, and low-density lipoprotein cholesterol levels (146, 148, 218, 219). Supplementation with lupin kernel fiber, a viscous NDFC, also led to a significant reduction in low-density lipoprotein cholesterol levels, and this response was inversely correlated with the fecal excretion of SCFAs (220). Although these studies do detect microbial signatures that closely correlate with clinical outcomes of NDFC supplementation, they do not provide direct evidence for a causative role of the GI microbiota (28, 221).

Still, some studies have been conducted that do indicate a causative link between the GI microbiota and NDFCs. Work from Fredrik Bäckhed’s group paired a human study with a humanized germfree mouse model to demonstrate that improved glucose metabolism due to whole-grain barley intake was dependent on the presence of Prevotella copri within the participant’s GI tract (222). Furthermore, by providing SCFAs directly through colonic infusions, studies have demonstrated that SCFAs do promote a systematic increase in PYY and GLP-1, while also benefiting markers of inflammation (223, 224).

Clinical research that establishes clear connections between the health benefits of NDFCs and the GI microbiota, as well as clear evidence for the role of the GI microbiota in the health effects of NDFCs, including the underlying mechanisms involved, is altogether limited. As described above, NDFCs can exert health benefits through microbiota-independent mechanisms such as a reduction in nutrient absorption through viscosity or binding of bile acids and cholesterol (73). This clearly illustrates a need for future studies that pair rigorously designed human studies with animal models, because this would constitute a unique tool to assess causation in humans (209, 222).

FUTURE DIRECTIONS

Assessing the Clinical Efficacy of NDFCs and the Role of the GI Microbiome

Until now, most of the clinical evaluation of NDFCs has occurred through one-sided studies that solely assess either the GI microbiome or the host while completely overlooking the other (59, 225). Clinical research is needed that assesses the effect of NDFCs in rigorously designed randomized controlled trials with relevant clinical endpoints and a parallel characterization of the role of the GI microbiota in these effects. Doing so would require close collaborations between nutritionally and microbiologically focused research groups to facilitate truly interdisciplinary research.

Despite clear evidence of the benefits of NDFCs from epidemiological studies, results from human intervention studies remain inconsistent (210). These inconsistencies could stem from a variety of reasons. Most of the dietary fiber assessed in epidemiological studies is derived from whole foods (fruits, vegetables, and whole grains), in which the NDFCs are consumed intact within a food matrix that also includes components such as phytochemicals and bioactive lipids (226). These bioactive compounds are likely to act synergistically within the food matrix (227, 228), and once purified to be used as a supplement, the health effects of the NDFC might be lost or reduced (229). However, purified fibers and prebiotics, as well as their fermentation products (SCFAs), have been repeatedly shown to be beneficial in mouse models (209). The variability in human intervention studies could arise from both the highly inter-individualized nature of the human GI microbiome (53, 55, 60) and the variability in the host’s metabolic response to NDFCs (39, 222). These two factors are inherently higher in humans, because mouse colonies are often composed of inbred mice housed in a highly standardized environment and are fed homogeneous diets. Given that the GI microbiotas of human subjects differ in the degree by which they are able to utilize specific NDFCs (75, 76), inter-individual variation is likely to be more pronounced in studies using a single purified substrate instead of a mixture of substrates or a whole food containing multiple fiber chemistries. In this respect, studies should characterize the chemical structure of the NDFCs to establish structure-function relationships between NDFC-chemistry and GI microbiome gene content, because this could be used to personalize approaches (24).

What is needed are human intervention studies with clinical endpoints that compare single NDFCs and their mixtures to those effects observed from whole foods, while also including a multi-omics approach for the analysis of the GI microbiome. Metagenomic analyses can identify specific shifts in GI microbiome composition and structure (e.g., diversity) that correlate with health outcomes. Such shifts, if they exist, would suggest that health benefits are due to selective changes of the GI microbiota in accordance with the original prebiotic concept (30). However, such studies should also include predictive modeling to determine how inter-individual differences in microbiome composition and functional capacity impact clinical outcomes. These studies could provide an explanation for the high variation observed in intervention studies with NDFCs and would establish a basis for personalized NDFC applications. Pairing this approach with a metabolomic analysis can further identify the compounds that are associated with health outcomes, which could range from microbial metabolites that originate from the fermentation of NDFCs to phytochemicals and their metabolic derivatives. Research has shown that phytochemicals present in fruits, vegetables, whole grains, and many fiber extracts are also metabolically transformed by the GI microbiota and absorbed by the host, correlating with health benefits (230).

Considering that the human-microbiota symbiosis evolved with a supply of NDCs beyond 100 g/day, future human intervention studies should clearly be performed with more physiologically relevant doses (14). Supplementation of 10 to 15 grams in intervention studies ensures that participants meet current dietary fiber recommendations of around 30 g/day (26), but this intake is still far below that of our ancestors (12, 14), potentially hindering the opportunity to detect evolutionary routed interrelationships between NDFCs, the GI microbiota, and health. Although experiments in nonindustrialized populations are clearly confounded by the possibility that specific fiber-degrading bacteria have been lost (16, 23), human studies that used NDC doses greater than 50 g/day did detect health benefits through the assessment of CNCD markers. For instance, by switching African Americans over to a more traditional South African diet that consisted of 55 g/day of NDCs, markers of colon cancer were improved in only 2 weeks (231). Furthermore, following a 2-week dietary intervention that resembled an ancestral diet, with around 143 g/d of NDCs provided as green leafy vegetables, fruit, and nuts, a 25% reduction in total cholesterol was observed, which is a response comparable to cholesterol-lowering medications (232, 233). These studies clearly provide a rationale for the use of higher doses of fiber in clinical research.

The approach described in this paragraph would ultimately help to identify putative mechanisms by which NDFCs exert their health effects, which would allow the formation of hypotheses that would inform the design of clinical studies and assist in the development of dietary strategies and targeted applications of NDFCs to improve human health.

Elucidating the Exact Mechanisms

Although well-conducted human studies with clinical endpoints would be sufficient to establish efficiency of NDFCs and obtain health claims for dietary fibers or prebiotics (28, 234), it is important to point out that such research would establish only correlations and not causation. Correlations can be misleading, because directionality cannot be established, especially since host parameters altered through NDFCs (e.g., inflammation, metabolic outcomes) can in themselves have an effect on the GI microbiota (235, 236). In addition, intake of RS has clearly established that the health effects of an NDFC can be completely independent of the GI microbiota, even though clear correlations between diet-induced shifts in the microbiome and host markers exist (237).

In this respect, it is important to consider that human studies have unavoidable limitations when it comes to establishing mechanisms, because the health effects of NDFCs can be completely microbiota-independent. For example, NDCs increase fecal bulk and decrease colonic transit time (238), which in turn influences GI microbiota composition (239). Ingestion of viscous NDCs may also increase the viscosity of digesta, interrupting the rate of nutrient absorption and ultimately promoting an improvement in clinical markers, especially postprandial glycemic response (240, 241). Furthermore, NDCs sequester compounds such as sterols, bile acids, and carcinogens (242–244), inhibiting their absorbance and enhancing their excretion. Modulation of the bile acid pool alone may have systemic implications for glucose and lipid metabolism, as well as systemic inflammation through action on the farnesoid X receptor and the GPR TGR5 (245). These microbiota-independent mechanisms can be difficult to distinguish from microbiota-dependent mechanisms, especially because they still might lead to strong correlations between diet-induced shifts in microbiome composition and host markers (237).

Elucidation of the role of the GI microbiota in the health effects of NDFCs and the underlying mechanisms will require innovative experimental approaches such as utilizing animal studies in parallel with human intervention studies to directly test the role of the GI microbiota in health outcomes (209, 222). Although establishing the contribution of the GI microbiota in the health outcomes of an NDFC would in theory be required to establish its prebiotic action, such studies would be extremely difficult and costly, and probably an unrealistic demand for the purpose of defining if an NDFC qualifies as a prebiotic. Therefore, despite the limitations discussed above, the establishment of correlations between the microbiome features and health benefits of an NDFC should probably be considered sufficient from a practical standpoint to establish which NDFCs constitute prebiotics (28). Establishing the causal role of the GI microbiota and the underlying mechanisms would remain essential information for the development of improved nutritional strategies. Only an in-depth mechanistic understanding will allow for the selection of NDFCs, or mixtures thereof, to systematically target specific features of the GI microbiome (i.e., specific taxa, diversity, metabolites) with the goal of correcting both immunometabolic abnormalities and dysbiotic features that underlie CNCDs.

CONCLUSION

The human diet has clearly changed over the last millennia, resulting in diminished potential to support our GI microbial community with growth substrates due to the dramatic reduction in the intake of NDFCs to a mere fraction of what was present in the diet of our ancestors (14). Evidence points to this “fiber gap” as being one prominent driving force behind the increased prevalence of CNCDs (13, 26). Although increased consumption of NDFC-rich whole foods such as fruits, vegetables, and whole grains is preferable from a nutritional perspective, efforts to increase their consumption have been ineffective to date, because dietary fiber levels remain low despite substantial efforts from nutritional agencies (25). Humans resist long-term changes of their dietary habits (246), illustrating a need for NDFC sources that can readily enrich the standard western diet (12). A broad array of NDFCs, which have vast potential to enhance the food supply, already exist on the market; however, product development and consumer research, in combination with clinical research, is needed to determine practical and cost-effective means by which these NDFCs can be incorporated into the food supply (12).

We know that there is a high degree of inter-individual variation in GI microbial response to NDFCs (53, 55, 60), and we have the tool set available to determine to what degree this individuality affects health outcomes in clinical studies (39). This essential knowledge would ultimately support the development of a framework by which interventions with NDFCs could be personalized. This ensures tremendous potential for growth within this field, promising exciting new developments as the focus of human nutrition shifts toward targeted nourishment of our symbiotic microbial communities as a way of preventing and treating CNCDs through supplementation with NDFCs.