Advances in understanding dietary fiber: Classification, structural characterization, modification, and gut microbiome interactions

Abstract

Gut microbiota and their metabolites profoundly impact host physiology. Targeted modulation of gut microbiota has been a long‐term interest in the scientific community. Numerous studies have investigated the feasibility of utilizing dietary fibers (DFs) to modulate gut microbiota and promote the production of health‐beneficial bacterial metabolites. However, the complexity of fiber structures, microbiota composition, and their dynamic interactions have hindered the precise prediction of the impact of DF on the gut microbiome. We address this issue with a new perspective, focusing on the inherent chemical and structural complexity of DFs and their interaction with gut microbiota. The chemical and structural complexity of fibers was thoroughly elaborated, encompassing the fibers’ molecular composition, polymorphism, mesoscopic structures, porosity, and particle size. Advanced characterization techniques to investigate fiber structural properties were discussed. Additionally, we examined the interactions between DFs and gut microbiota. Finally, we summarized processing techniques to modify fiber structures for improving the fermentability of DF by gut microbiota. The structure of fibers, such as their crystallinity, porosity, degree of branching, and pore wettability, significantly impacts their interactions with gut microbiota. These structural differences also substantially affect fiber's fermentability and capability to modulate the composition of gut microbiota. Conventional approaches are not capable of investigating complex fiber properties and their influences on the gut microbiome; therefore, it is of the essence to involve advanced material characterization techniques and artificial intelligence to unveil more comprehensive information on this topic.

1. INTRODUCTION

The terminology of “dietary fiber” (DF) was first proposed in the 1950s (Williams et al., 2019), and since then, several workable definitions were proposed. To date, the most widely accepted definition for DF was defined by Codex Alimentarius, referring to “carbohydrate polymers with $\ge$10 (or 3, depending on the jurisdiction) monomeric units that are resistant to enzymatic hydrolysis during human small intestinal digestion on a condition that they are either: 1) naturally occurring edible carbohydrate polymers ingested during food consumption; 2) naturally occurring carbohydrate polymers extracted from foods that demonstrate scientifically proven physiological health benefits; and/or 3) synthetic carbohydrate polymers that demonstrate scientifically proven physiological health benefits” (Wang et al., 2019). From a chemistry perspective, DF is a mixture of different components, including but not limited to cellulose, hemicellulose, lignin, oligosaccharides, inulin, resistant starch, pectin, and gums (Tang et al., 2020; Williams et al., 2019). DF plays a critical role in modulating the human digestive system and is associated with numerous health benefits. Early studies found that a high intake of DF is related to a lower risk of developing coronary heart disease, stroke, hypertension, diabetes, obesity, and certain gastrointestinal diseases (Anderson et al., 2009). The most updated 2020–2025 Dietary Guidelines for Americans emphasize the importance of fiber intake, recommending a daily DF intake varying from 14 to 34 g for different age groups and genders (Phillips, 2021). However, the average American's fiber intake is less than half of the recommended amount (Anderson et al., 2009). More recent data from USDA show that US consumers on average consumed 8.1 g of fiber for each 1000 calories in 2017–2018, which is 58% of the recommended 14 g per 1000 calories (USDA, 2023). Therefore, there is a need to better incorporate DFs into American foods and diets.

Recently, scientific interest has focused on the prebiotic effects of DFs for modulating gut microbiota. Prebiotics, defined as substrates selectively utilized by host microorganisms conferring health benefits, represent a subset of DFs that specifically promote the growth of beneficial bacteria in the gut, though not all DFs qualify as prebiotics due to their varying fermentability and selective utilization patterns. Researchers have been increasingly investigating the interaction between fiber and the gut microbiome, as new findings increasingly suggest that gut microbiome health is closely associated with immune response, epithelial integrity, electrolyte re‐absorption, and gut motility, as well as the function of peripheral organs including the liver, pancreas, brain, and muscles (Poeker et al., 2018). Generally, DF impacts human health and prevents chronic diseases through the following mechanisms: lowering cholesterol, improving glycemic control, normalizing stool form, and producing short‐chain fatty acids (SCFAs). However, depending on the types of fiber being taken, the health benefits vary. For example, high‐viscosity DFs (psyllium, and raw guar gum) provide laxative effects that lower cholesterol and improve glycemic control, whereas nonviscous soluble dietary fibers (SDFs) (inulin, fructooligosaccharides, and wheat dextrin) are highly fermentable in the lower intestine and promote SCFA production (McRorie & McKeown, 2017).

It is estimated that roughly 1 trillion bacteria live in the gut, responding to their environment via diffusible small molecules or metabolites, which in turn affect their neighboring microbes and human hosts. Studies have attempted to establish an understanding of the linkage between bacterial population evolution and host health conditions (Henke & Clardy, 2019). For example, inflammatory bowel disease, colorectal inflammation, obesity, and metabolic syndrome are all closely associated with gut microbiota health (Fan & Pedersen, 2021; Gomaa, 2020; Sekirov et al., 2010; Yin et al., 2019). Consequently, shifting the gut microbiota population toward the enrichment of health‐promoting bacteria is of great interest. One way to shift the bacterial population is through diet, especially through a diet rich in DFs. By modulating DF intake, researchers have been able to shift the microorganism ecosystem in the large intestine (Hamaker & Tuncil, 2014). Therefore, it is crucial to understand the impact of DF on gut microbiota to optimize a shift toward a more beneficial bacterial population; however, the chemical and structural complexity of DF, together with intriguing gut microbial composition, presents a grant challenge to precisely modulate the gut microbial composition. Recently, it was reported that even subtle variations in fiber structure (e.g., backbone lengths and branching units) led to different compositions of gut microbiota (Tuncil et al., 2020). Although contradictory results have been reported in fiber–gut microbiome research, some findings reached a wide agreement. For instance, the consumption of pectin can increase the abundance of beneficial intestinal microbiota and reduce the abundance of pathogens or mucin‐degrading bacteria (Beukema et al., 2020). Inspired by promising findings like this, more ambitious efforts have been undertaken, aiming to precisely modulate the gut microbial composition using designated DFs (Hamaker & Tuncil, 2014).

Most review articles have focused on the physiological and health benefits of different types of fibers or fibers from various plant sources (Anderson et al., 2009; Gill et al., 2021; He et al., 2022); however, there has not yet been a comprehensive review to summarize and elucidate the complex structural and chemical properties of DFs and their impact on fiber–gut interactions. As the food science and nutrition community becomes increasingly interested in function‐specific DFs to precisely modulate the gut microbiome and improve human health, an insightful connection between fiber properties and gut microbiome shifting will unequivocally contribute to the advancement of DF research. This review aims to elucidate the complexity of DFs with an emphasis on structural properties and their impact on fiber digestibility and the gut microbiota. The DF structures are elucidated from the molecular to the micro‐scale to the macro‐scale levels. Advanced characterization techniques (e.g., X‐ray powder diffraction, nitrogen adsorption/desorption, small‐angle X‐ray scattering [SAXS], fluorescence probe) are discussed in depth with related examples to understand how these techniques are used to unveil the complex structures of DFs. Potential processing strategies for modifying structures and physicochemical properties of fibers to improve their accessibility and fermentability are also examined. We conclude this review by identifying the knowledge gap and other challenges in fiber research and suggesting the need for future research.

2. THE CLASSIFICATION, CHEMICAL COMPOSITION, AND STRUCTURAL INSIGHTS OF DFs

The classification of DF is not universally congruent. Conventional classification differentiates fibers by water solubility, fermentability, source, monomer composition, chain length, and degree of polymerization (DP). A summary of common DF information is listed in Table 1. The water solubility‐based classification is the most commonly used, dividing DF into SDF and insoluble DF (IDF). In the past, the water solubility and fermentability of DFs were presumed to be linked. For example, Dhingra et al. (2012) categorized cellulose, hemicellulose, and lignin as “water insoluble and less fermented” fiber, and they categorized pectin, gums, and mucilage as “water soluble and well fermented” fiber. However, as the understanding of DF advanced, such a simplified classification became inadequate, as water solubility is not the only factor determining fiber's fermentability by gut microbiome. A recent study by Nsor‐Atindana et al. (2020) reported the significant dependence of fermentation potential on the molecular weight (MW) and particle size of DF. Using cellulose as a model DF, researchers found that gradually reducing the cellulose size to a nanometric scale resulted in an increased fermentation and elevated Bifidobacterium population. The fermentation of DF also highly depends on DF's crystallinity, porosity, and bacterial affinity. In this section, we thoroughly elucidate varying fiber structures at different levels and discuss their features and potential impacts on the gut microbiome. An overview of the breakdown structure is presented in Figure 1 using cellulose as an example. The structural features of plant fibers will be covered from the molecular composition (angstroms), polymorphism (nanometers), mesoscopic structures (tens of nanometers), porosity (nanometers to microns), architecture (tens of microns), to particle size (hundreds of microns to millimeters).

TABLE 1.

A summary of common dietary fibers and their monomer constituents, molecular weight, and water solubility.

Fiber	Monomer	Molecular weight (kDa)	Solubility in water a	Reference
Cellulose	d‐Glucose	3–875	Insoluble	Kraemer, 1938
Hemicellulose	d‐Xylopyranose, d‐glucopyranose, d‐galactopyranose, l‐arabinofuranose, d‐mannopyranose, d‐glucopyranosyluronic acid, d‐galactopyranosyluronic acid	10–2000	Insoluble	Zhou et al., 2017
Lignin	Guaiacyl, syringyl, p‐hydroxyphenyl	1 ‐ 400	Insoluble	Liu et al., 2022
Pectin	Heteropolysaccharide	50‐150	Soluble	Mudgil et al., 2017
Guar gum	Mannose, galactose	1000–2000	Soluble	2016 Gelardi et al., 2016
Inulin	d‐Fructosyl, α‐d‐glucose	0.5 ‐ 13 c	Soluble	Mensink et al.,2015
Resistant starch	d‐Glucose	14–1400	Both b	Ma & Boye, 2018
β‐Glucan	d‐Glucose	21–3100	Soluble	Du et al., 2019

^a The solubility was determined by unmodified fibers.

^b Depending on different types of resistant starch. ^c some bacterially produced inulin has higher molecular weight.

FIGURE 1.

Hierarchical structure of plant fiber from plant cell macroscale to molecular nanoscales. (Reprinted from Miyashiro et al., 2020). No special permission is required to reuse all or part of article published by MDPI, including figures and tables.

2.1. Molecular compositions of DFs

Although fibers are classified and named according to their monomer constituents, differences are still present at the molecular level such as DP and MW. For example, depending on the source, the DP of cellulose could vary drastically, resulting in its highly variable MW ranging from 2 to over 875 kDa (Kraemer, 1938). Hemicellulose usually has an MW ranging from 10 to 200 kDa (Zhou et al., 2017). Hemicellulose is composed of different sugar monomers (e.g., d‐xylopyranose, d‐glucopyranose, d‐galactopyranose, l‐arabinofuranose, and d‐mannopyranose), and the fraction of each monomer varies in hemicellulose depending on its plant source. For example, xylose constitutes 81% of monomers in hemicellulose from corn stover but less than 1% of monomers in hemicellulose from spruce wood (Zhou et al., 2017). A recent study found that wood‐derived hemicellulose (rich in galactoglucomannan and arabinoglucuronoxylan) showed excellent prebiotic properties for promoting the production of SCFAs, which play a pivotal role in gut health (La Rosa et al., 2019). Lignin provides exceptional mechanical strength and coherence to biomass due to its three‐dimensional network of aromatic substructures. It is an amorphous polymer of phenylpropane units containing three types of monolignols: p‐hydroxy‐phenyl (H subunit), guaiacyl (G subunit), and syringyl (S subunit). As the ratio among the H, G, and S subunits in lignin varies in different fiber sources, their physicochemical properties are also different. A previous study found that the presence of guaiacyl‐rich lignin is an indication of high mechanical strength, poor hydration capability, and reduced enzymatic accessibility (Del Río et al., 2011; Rosado et al., 2021).

In addition to heterogeneous monomer composition, the difference in DFs is also based on their molecular size, which is typically measured by DP—the number of monosaccharide units in the molecule (Table 2). The accurate definition of DF is still debated, particularly whether oligosaccharides with the DP between 3 and 9 should be considered as DF or not (Dai & Chau, 2017). Low‐MW DFs include oligosaccharides (DP 3–9) and inulin (DP 4–80) (Mensink et al., 2015) regardless of chain length. 2023. High‐MW DFs consist primarily of nonstarch polysaccharides like cellulose (DP ∼10,000) (Tekin et al., 2014), hemicelluloses, and resistant starch. While all DFs resist digestion in the small intestine, their molecular size significantly impacts their fermentability in the colon—low‐MW fibers tend to be rapidly fermented in the proximal colon, while high‐MW fibers are fermented more slowly throughout the colon (Stribling & Ibrahim, 2023). This molecular size classification provides important insights into the physiological effects and functional properties of different DFs.

TABLE 2.

Classification of dietary fiber by molecular size (Stribling & Ibrahim, 2023).

Molecular size	Types of dietary fibers	Examples	Degree of polymerization	Reference
Low molecular weight	Oligosaccharides	Fructo‐oligosaccharides, oligofructose, galacto‐oligosaccharides, xylo‐oligosaccharides	3–9	Dai & Chau, 2017
	Inulin	Inulin	4–80	Mensink et al., 2015
High molecular weight	Nonstarch polysaccharides	Cellulose, hemicellulose, pectin, beta‐glucan	Varies, 10,000 for cellulose	Tekin et al., 2014
	Resistant Starch	Type I–V resistant starch	9600–15,900	Wang et al., 2023
	Other substances	Lignin	323–449	Ganewatta et al., 2019

Chemical groups in DFs also play a significant role in determining DFs’ physical and biological properties. For example, pectin and cellulose can both be methoxylated or acetylated, which are determined by the degree of methoxylation (DM) and degree of acetylation (DA) (Lin et al., 2011; Nasatto et al., 2015; Song et al., 2021; Wang et al., 2018; Xu et al., 2020). In general, pectin and cellulose with greater DM have higher viscosity in water (Wang et al., 2016), which could hinder enzymatic digestion and fermentation in the large intestine. The impact of DM on prebiotic potential has been systematically studied by Dongowski et al. (2002), who reported that with the increase in DM, the rates of fiber fermentation and SCFA production decreased in rats. In both in vitro and in vivo fermentations of pectin, low‐DM pectin was fermented faster than high‐DM pectin.

2.2. Polymorphism and crystallinity

A crystalline structure is formed as molecules or atoms are tightly packed in a particular and repetitive order, whereas amorphous structures refer to randomly arranged chemical units. Many biological materials, including DFs, consist of both crystalline and amorphous portions, which could be quantified using the crystallinity index (CI) 2019 (Liu et al., 2019). The CI of DF plays a vital role in determining its physicochemical properties but is often not reported, except for cellulose (Park et al., 2010). There are only a few types of DF compounds that contribute crystallinity, such as cellulose and some resistant starches (Liu et al., 2019; Ma & Boye2018). As a chain polymer composed of d‐glucopyranose units and linked by β‐1,4 glycosidic bonds, cellulose usually presents as a stiff rod‐like structure (Lin et al., 2011). These rods are interconnected by hydrogen bonds, and as the packing density of cellulose increases, crystalline regions are formed (Rowell et al., 2012). The importance of cellulose crystallinity to its physicochemical properties was reported 70 years ago, and a high crystallinity was associated with poor water adsorption capability (Ward, 1950). The crystallinity of cellulose varies between different sources, for example, most wood‐derived celluloses exhibit a high degree of crystallinity of about 65% (Rowell et al., 2012). Other than the crystalline regions, the remaining portion of cellulose with a lower packing density is referred to as amorphous cellulose. It has been also reported that crystallinity plays a major role in cellulose hydrolysis. One study correlated the initial enzymatic hydrolysis rate with the CI, and a near‐linear negative relationship was reported indicating the hydrolysis resistance of highly crystalline cellulose (Hall et al., 2010). Other studies also found that amorphous cellulose is hydrolyzed faster than partially crystalline cellulose (Fan et al., 1980; Hattori & Arai, 2016; Zhang & Lynd, 2004). Similar to cellulose, starch with high crystallinity is usually found in uncooked potato starch or corn starch, due to the presence of B‐ and C‐type polymorphs in amylose that inhibit enzymatic hydrolysis (Birt et al., 2013). Therefore, highly crystalline starch rich in amylose that is resistant to digestion is named type II resistant starch. However, digestibility can be improved by cooking via the disruption of crystalline regions during starch gelatinization.

2.3. Mesoscopic structure

Mesoscopic structure refers to the structural features between 10 and 1000 nm. For SDFs like pectin, the mesoscopic structure specifically depicts the compactness and configuration when it is dispersed in an aqueous phase (Zhao et al., 2020). In an aqueous phase with dynamic pH and ionic strength fluctuation, as well as the presence of other chemicals, extensive molecular interactions can occur between adjacent chains leading to either self‐association or heterotypic cooperation. Through such interactions, DFs could form a variety of complex structures, such as gels and biopolymer–biopolymer complexes (Alba et al., 2017). As those DFs are fermented in the human gut, the evolution of their mesoscopic structures could substantially impact the attachment of enzymes and microorganisms to the fiber surface, which further affects fiber degradation. Pectin is one of the examples whose mesoscopic structure alters depending on environmental conditions. The mesoscopic structure of pectin was studied using SAXS and fluorescence probing techniques (Alba et al., 2017; Zhao et al., 2020). Both methods indicated that the pectin at high acidic and conditions tend to be highly compacted due to the decreased intermolecular and intramolecular electrostatic repulsions. In this case, a smaller‐sized mesoscopic structure is formed with fewer hydrophobic regions. The compact mesoscopic structure at acidic conditions may hinder enzymatic degradation and microbial fermentation of pectin.

In this review, we categorize DF into four main mesoscopic structural types, each may present distinct characteristics and impacts on gut microbiota interactions (Table 3). Fibrillar structures, exemplified in cellulose (Aschner & Hestrin, 1946), feature long thread‐like formations that may provide surfaces for bacterial adhesion but could result in slower fermentation due to their compact nature. Network structures, found in pectin and β‐glucan gels, form interconnected polymer chains in a mesh‐like arrangement (Zsivanovits et al., 2004) that may trap water and nutrients, potentially slowing bacterial access while enabling sustained fermentation. Lamellar structures, present in some hemicelluloses within plant cell walls (Xiao et al., 2021), have a layered configuration that may require specific enzymes for degradation and typically result in slower fermentation rates. Finally, aggregated structures, such as those found in retrograded resistant starch (Kapelko‐Zeberska et al., 2024), consist of aggregated polymer chains that can provide varied fermentation rates depending on their degree of aggregation. These different mesoscopic structures could play crucial roles in how DFs interact with and are processed by gut microbiota.

TABLE 3.

Summary of fiber mesoscopic structures.

Mesoscopic structure	Characteristics	Examples	Potential impact on gut microbiota interactions	Reference
Fibrillar	Long, thread‐like structures	Cellulose	May provide surfaces for bacterial adhesion; slow fermentation due to compact structure	Aschner & Hestrin, 1946
Network	Interconnected polymer chains forming a mesh‐like structure	Pectin gels, β‐glucan gels	Can trap water and nutrients; may slow bacterial access but provide sustained fermentation	Tardy et al., 2021 YZsivanovits et al., 2004
Lamellar	Layered structures	Some hemicelluloses in plant cell walls	May require specific enzymes for degradation; slower fermentation rates	Xiao et al., 2021
Aggregated	Clusters of polymer chains	Retrograded resistant starch	Fermentation rates can be depending on the degree of aggregation	Kapelko‐Zeberska et al., 2024

2.4. Surface area and pore sizes

An understanding of the three‐dimensional microstructure of DF under a hydrated condition is vital to determine whether the chemical bond is accessible to enzymatic attack and microbial degradation (Zhang & Lynd, 2004). All DF particles have both external and internal surfaces, and the internal surface area is usually much larger than the external surface area. For example, cellulose possesses hundreds of times internal surface area larger than its external surface area, and further expand at full swelling (Chang et al., 1981). Jin et al. (2024) reported that the fermentability of corn fiber is highly correlated with the fiber's specific surface area. With the advance of microscopic techniques, the visualization of DF has been performed with a great resolution to characterize different structures. Ding and Michael (2006) visualized the cell wall microfibril arrangement using an atomic force microscopy, where the interspacial space could be identified among microfibril layers on the cell wall surface, ranging approximately from 5 to 200 nm. In addition to visualization, the nitrogen adsorption/desorption technique based on the Brunauer–Emmett–Teller (BET) theory has been used to reveal quantitative porosity information. Previous studies used this technique to measure the specific surface area of fiber (Nsor‐Atindana et al., 2020) as an indicator to confirm the size reduction treatment. In fact, the nitrogen adsorption/desorption technique can also precisely map the pore size distribution (Kang et al., 2018). Although it has not been extensively investigated, it is possible that enzymatic reactions may occur at different rates depending on pore size distribution. DF with high bacterial accessibility is usually desired because it usually indicates that the fiber is highly digestible. Guillon et al. (1998) suggested that pores with diameters >1 µm are accessible to bacteria, and pores with diameters >5 nm are accessible to enzymes. In their study, the fermentation rate as a function of porosity was studied with experiments, then mathematically modeled, and a strong positive correlation was found between the total pore volume accessible to enzymes and the fermentation rate. However, the correlation was less prominent between fermentability and micropores only accessible to enzymes. Their results suggest that fiber structure, particularly pore size, plays a significant role in bacteria–fiber interactions during degradation.

2.5. Assembly, architecture, and interaction of components in DFs

The natural architecture of DFs was recognized in the early 1970s, when Trowell first defined “DFs” as components of the plant cell wall that resist digestion by secretions of the human alimentary tract (Trowell, 1972). The distinct structure and complex packing of cell wall components (e.g., microfibrils) also greatly contribute to their recalcitrance to chemical and biological degradation (Raud et al., 2016). The assembly of DF architecture is achieved at several scales. Taking lignocellulose as an example, the first scale involves the intermingling of molecular chains packing in layers that are held together by weak van der Waals forces (Rowell et al., 2012). After packing into rigid cellulose fibrils, hemicelluloses are attached to cellulose intimately through hydrogen bonds. In the outermost layer, lignin is covalently linked to hemicelluloses, acting like a glue, to form a lignin–carbohydrate complex (Giummarella et al., 2019). The lignin–carbohydrate complexes can be formed by five types of connecting bonds: glycosides, benzyl ethers, γ‐esters esters, ferulate/coumarate esters, and hemiacetal/acetal bonds (Giummarella et al., 2019). All these interactions, including cellulose–hemicelluloses bonding and lignin–carbohydrate linkage, retard its enzymatic hydrolysis (and thus microbial degradation) by reducing the area of fiber accessible to enzymes and gut microbes (Du et al., 2014). Many naturally derived fibers are covered by a discrete and compact lignin packing, which is a ubiquitous structure for plants to protect themselves from pests and microorganisms, because lignin is considered the most recalcitrant compound for fiber degradation (Vanholme et al., 2010).

DFs can also interact with food components. For example, pullulan and pectin can significantly impact the structural properties and digestibility of fried starchy foods (Chen et al., 2019). When these fibers interact with starch during frying, they compete with starch granules for available water and form protective coatings around the granules, which helps preserve the starch's internal structure and reduces gelatinization during frying. As a result, the starch becomes more resistant to enzymatic digestion, leading to higher levels of slowly digestible starch and resistant starch. Pectin, being an anionic polysaccharide, was found to be more effective than pullulan at protecting the starch structure and reducing digestibility, likely due to its ability to create both electrostatic and steric barriers. This fiber–food interaction suggests potential applications in developing healthier fried foods with reduced digestibility. In many fruit‐ and cereal‐based products, DFs are often linked with phenolic compounds via covalent and noncovalent bonds (Rocchetti et al., 2022). As a result, the concept of “antioxidant dietary fiber” was introduced because of the potential antioxidant properties of phenolic compounds anchored on DFs.

2.6. Particle size

The particle size of DF plays a vital role in its hydration property, colonic function (e.g., transit time, fecal bulking), and microbial fermentability. A study found that both water‐holding and water‐binding capacities of pea hulls, which are primarily composed of cellulose and hemicellulose, increased after grinding, possibly due to the increased surface area and pore volume, and grinding led to a decrease in water‐binding capacity of wheat bran, citrus, and sugar‐beet DF probably by alteration and collapsing of the fiber matrix (Auffret et al., 1994). Another study was conducted to evaluate the effect of particle sizes of wheat bran, which is rich in arabinoxylan, cellulose, and residual starch, on the colonic function of young adult men (Heller et al., 1980). The results showed that wheat bran with a fine particle size (particle size ranges from 63 ‐ 500 µm) was less effective in holding water in feces and promoting transit of digesta through the gut compared to coarse‐ground (particle size ranges from 125 ‐ 2,000 µm) wheat bran. Thakkar et al. (2020) studied the impact of the particle size of maize bran, primarily consisting of arabinoxylan and cellulose, on SCFA production and microbiota community evolution. The maize bran samples with different particle size ranges (180–250, 250–300, 300–500, and 500–850 µm) were in vitro fermented. It was found that maize bran with size ranges of 250–300, 300–500, and 500–850 µm exhibited similar low SCFA production, while further reduction to 180–250 µm significantly improved the SCFA production.  More recently, Wang et al. (2024) studied the size effect of functionality and microbial fermentability of IDF derived from wheat bran, a processing byproduct rich in arabinoxylan and cellulose. They reported that the key functional properties of DFs, such as oil‐ and water‐holding capacity, were positively correlated with particle size. However, DFs with large particle sizes (above 420 µm, 200 mesh size) exhibited lower fermentability compared to those with medium (74–149 µm, 100‐200 mesh size) and small particle sizes (<74 µm, 200 mesh size). Moreover, the clinical study revealed that consumption of medium‐size DF increased the level of d‐alanyl‐d‐alanine and decreased the level of kynurenic acid, which showed a significantly positive correction with the abundance of Romboutsia and NK4A214_group, respectively. Particle size also determines the practical applications of DFs in food systems. Coarse fibers that exhibit high water‐holding capacity and alxative effect but slower hydration rates are suitable for applications where visible fiber content and texture attributes are desired in the final market product (McRorie & McKeown, 2017). Fine fibers with poort water holding capacity and mucosa irritating effect would not be desirable for irritable bowel syndrome patients 2017.

3. ADVANCED TECHNIQUES FOR CHARACTERIZING THE STRUCTURE OF DF

Table 4 summaries a list of advanced techniques for characterizing the structure of DF. Here, rather than listing all characterization techniques, we focus on the ones that may have a prominent potential to reveal new information regarding the structural properties of DF related to their microbial accessibility and fermentability.

TABLE 4.

A summary of advanced techniques for fiber characterization.

Fiber properties		Techniques	Advantages	Reference
Chemical composition	Molecular weight	Gel permeation chromatography	Precise	Yoo et al., 2017
	Degree of polymerization	Gel permeation chromatography	Precise	Yoo et al., 2017;
Structural properties	Crystallinity	X‐ray diffraction (XRD)	High sensitivity	Chen et al., 2018
	Particle size	Laser granulometry	Facile	Pang et al., 2019
	Surface area	Brunauer–Emmett–Teller (BET)	Sensitive	Liu et al., 2019
	Water holding capacity	Water retention value (WRV)	Indicating hydration and swelling properties	Weiss et al., 2018
	Porosity and pore size distribution	Ultra‐small‐angle neutron scattering (USANS/SANS)	Comprehensive structural information	Clarkson et al., 2013
	Compactness	Ultra‐small‐angle scattering (USAXS/SAXS)	Comprehensive structural information	Alba et al., 2018
Surface properties	Surface roughness	Atomic force microscopy (AFM)	High resolution	Isaac et al., 2018
	Surface hydrophobicity	Contact angle	Facile	Shi et al., 2020
Morphology	Imaging	Micro‐CT	Three‐dimensional structure	An et al., 2019
		SEM	High resolution	Liang et al., 2018

3.1. X‐ray diffraction

The degree of crystallinity in DFs can be quantified through a CI, which measures the proportion of ordered crystalline regions compared to disordered amorphous regions in the fiber structure (Thielemans et al., 2023; Ward, 1950). Scientists can precisely measure these differences using X‐ray diffraction (XRD) techniques, where the CI is determined by comparing the relative intensities of crystalline and amorphous diffraction patterns (Liu et al., 2019). XRD introduces constructive interference of monochromatic X‐rays scattered in a range of user‐defined angles and collects the interference patterns. The XRD pattern can then be analyzed to fingerprint the periodic structural arrangement in a sample. XRD has long been a ubiquitous approach to investigating the polymorphism of materials. Crystallinity is one of the important parameters that XRD can analyze. Given the XRD patterns of a cellulose sample, the degree of crystallinity can be calculated using the CI equation (1):

$$\mathrm{Crystallinity}\mathrm{index}=1-h_{\mathrm{am}}/h_{\mathrm{cr}},$$ (1)

where $h_{\mathrm{cr}}$ is the height of crystalline cellulose 002 peak at 2θ = 22.5°, and $h_{\mathrm{am}}$ is the peak height of amorphous cellulose. In addition to crystallinity, XRD also provides information to calculate crystal grain size. According to Gardner and Blackwell (1974), the cellulose from Valonia ventricosa has a crystal grain dimension of 16.34 × 15.72 Å. The repeating distance between two adjacent cellulose crystal domains was also calculated at 10.38 Å.

XRD has emerged as a valuable tool for characterizing the crystallinity of DFs from various food sources before and after processing. In a notable study, Tu et al. (2014) investigated how dynamic high‐pressure microfluidization and fermentation processes affect the crystalline structure of soybean residue fiber. Their XRD analysis revealed distinct structural modifications: fermented soybean fiber exhibited an amorphous structure with significantly reduced overall crystallinity, demonstrating that microbial fermentation effectively altered the crystalline regions. In contrast, microfluidization‐treated soybean fiber maintained its original crystalline structure, suggesting that while this process could disrupt amorphous regions, it left the crystalline domains intact. These findings have important implications for DF processing, as fermentation could serve as an effective pretreatment strategy to reduce fiber crystallinity, potentially enhancing its accessibility to gut microbiota and improving its functional properties.

3.2. Nitrogen adsorption/desorption technique

Porosity and accessible surface area are critical factors that influence the fermentability of DFs (Dhingra et al., 2012). Based on their accessibility to the external environment, pores can be categorized as open pores, closed pores, and interconnected pores (Chen et al., 2018; Zhang et al., 2017; Ziaiifar et al., 2010). To date, the impact of pore types on fiber functionality has not been systematically investigated. Technically, open pores maintain direct contact with the external surface (Thibault et al., 2024), allowing potential access to both enzymes and bacteria. These pores could thus serve as primary entry points for digestive enzymes and provide colonization sites for gut microbiota. In contrast, closed pores are isolated and inaccessible to the external environment (Inagaki, 1996), and therefore, they may not directly participate in enzymatic digestion or microbial fermentation. Interconnected pores form networks (Thibault et al., 2024; Brünler et al., 2016), potentially creating pathways that facilitate the movement of enzymes and bacterial metabolites. The Micro‐CT and mercury prosimeter technique can quantitatively characterize different pore types,   (Thibault et al., 2024), while the nitrogen adsorption/desorption technique based on BET theory is more widely used to study the specific surface area. This technique can also be used to assess the total surface area of fiber, where a positive correlation between the BET specific surface area and enzymatic accessibility was reported (Kang et al., 2018). Jin et al. (2024) used the nitrogen adsorption/desorption technique to quantify the specific surface area of corn fiber, which is rich in cellulose, hemicellulose, and lignin. In their study, corn fiber was treated with a combined dilute alkali and milling process to reduce particle size and partially remove hemicellulose to create a porous structure and enhance the surface area of corn fiber. The corn fibers subjected to different process conditions (1% NaOH or NaHCO₃, disk milling, or ball milling) were evaluated by in vitro fermentation using porcine fecal inoculum. They found that the BET specific surface area highly positively correlated with SCFA production in in vitro fermentation. Besides specific surface area, the nitrogen adsorption/desorption technique can also be used to quantify the pore size distribution of DFs. Researchers proposed that not all the pores are accessible to water during hydration, and the pore wettability could be associated with the local hydrophobicity, pore geometry, and pore size, because the wetting process involves a complicated converging–diverging capillary process (Rabbani et al., 2018). Therefore, water‐retention value (WRV), which is the amount of water retained in fiber after centrifugation, is regarded as an alternative parameter to indicate the likelihood of enzymatic accessing of fiber pores (Weiss et al., 2018). Previous studies found that fiber WRV is highly correlated with and can be an effective indicator of the enzymatic hydrolysis rate (Crowe et al., 2017; Noori & Karimi, 2016). In the future, it would be interesting to evaluate the WRV as an indicator to predict the fermentation potential of DF in the gut.

3.3. Small‐angle X‐ray scattering

SAXS is a powerful nondestructive technique for uncovering detailed structural features of DF on a broad size scale. The basic principle of SAXS is to analyze the scattering pattern of X‐ray when it interacts with the sample, and the pattern forms as a result of electron density inhomogeneity (Turovsky et al., 2015). When being used for characterizing DF, SAXS is able to interpret a variety of structural properties, including polymorphism, shape, dimension, compactness, and pore distributions (Cheng et al., 2015). Since polymorphism and pore distribution have already been widely studied with XRD and BET techniques, this section focuses on analyzing fiber mesoscopic structures using SAXS. Alba et al. (2018) and Khan et al. 2017 used SAXS to study the compactness of pectin and polysacchrides in aqueous environment. Figure 2a plotted the Porod SAXS pattern of pectin. In the low q‐region (q < 0.003 Å⁻¹), the structural information provided is associated with the length scale of the clusters where the pectin chains form random aggregates, and the size of the cluster could be estimated using the Guinier equation (R _cluster ∼100–200 nm). The shape of pectin could be evaluated using fractal dimensions, which refer to the slopes in the Porod plots. According to Figure 2a, the slopes from low‐ and high‐q regions are around −1.9 and −1.2, respectively. In theory, a fractal dimension of 2 describes a random walk, which indicates that pectin chains randomly fill the space (Alba et al., 2018). Herein, the fractal dimension in the low‐q region (∼1.9) indicates the random pectin cluster on a large scale. In the high‐q region, the fractal dimension reached ∼1.2, which suggests the scattering of stiff rods and corresponds to the radius of gyration (R _g) of the pectin chains. The fractal dimension at the low‐q region is also used to reflect the compactness (Alba et al., 2017). If the fractal dimension is close to 2 in the low‐q region, the pectin is regarded as less compact; if the fractal dimension is away from 2, the pectin cluster is compact.

FIGURE 2.

(a) Characteristic Porod plots of pectin samples. The first level of structure (R _cluster) is associated with the length scale of clusters, which are indicated with large purple circle in atomic force microscope (AFM) image. The shoulder in the scattering curve reflects charges on the polyelectrolyte backbone. The transition to the higher length scales (dashed black line) is associated with the radius of gyration (R _g) of individual pectin chains (small turquoise circle in AFM). Figure was adapted from Alba et al. (2018) with permissions. (b) Guinier plot, ln(qI(q)) vs. q ² for nata‐de‐coco (NdC), tunicate cellulose (TC), and bacterial cellulose (BC) showing two distinct regions having different slopes. BC, NdC (food‐grade bacterial cellulose), and TC have been investigated. Figure was adapted from Khandelwal and Windle (2014) with permissions.

SAXS has also been a robust method to extract detailed configuration information of cellulose microfibrils. Khandelwal and Windle used SAXS to characterize the dimensions of three types of cellulose (Khandelwal and Windle, 2014)). The SAXS Guinier plots of bacterial cellulose, nata‐de‐coco, and tunicate cellulose are shown in Figure 2b. As cellulose is a rod‐like structure, the Guinier approximation equation (2) could be used to calculate the dimension:

$$qI\left(q\right)=G\exp \left(\frac{-q^2{r}_{\mathrm{c}}^2}{2}\right),$$ (2)

where $G$ is the scaling constant, $r_{\mathrm{c}}$ is the radius of gyration of the cross section of the scattering element and $q$ is the scattering vector. By solving the equation, the dimension of cellulose can be calculated. For BC, the calculated cross section from slope 1 and slope 2 is 32 × 16 and 21 × 10 nm, respectively. The results are highly consistent with the width and height obtained from microscopic results (Khandelwal & Windle, 2014).

Recent studies have demonstrated SAXS as a particularly valuable tool for understanding how fiber structure influences microbial accessibility and fermentation patterns in the gut environment. The technique can track structural evolution during hydration and enzymatic modification, providing insights into how fibers respond in digestive conditions. For example, Larsson et al. (2022) used complementary SAXS and simulation analyses of cellulose‐rich materials to reveal how interstitial spacing and pore accessibility change when exposed by swelling—factors that directly impact microbial access to fiber surfaces. Recent work by Garina et al. (2024) demonstrated how SAXS and SANS can differentiate between protein and polysaccharide fiber alignment patterns during processing, showing that these biopolymers develop distinct structural anisotropies that could affect their digestibility. The technique is sensitive enough to detect transitions from densely packed states to more open structures with increased interfibrillar spaces that facilitate enzymatic and microbial penetration. When combined with fermentation studies, SAXS analysis can demonstrate correlations between structural parameters like crystalline lamellae, amorphous lamellae and fermentation consequences (Tu et al., 2021Garina's work used SAXS and SANS to reveal how cellulose form local alignment in transition zones, which likely can influence their fermentation patterns in the gut. These structure–function relationships provide valuable insights for designing DF ingredients with targeted fermentation profiles and optimized microbial accessibility.

3.4. Fluorescence probe and label

Fluorescence is an emerging technique to characterize plant‐based materials for visualization and quantification. Using fluorescence labeling, the binding between carbohydrates and enzymes becomes visible. Although characterization techniques, such as isothermal titration calorimetry and quartz crystal microbalance, provide insightful structural information to predict fiber–enzyme and fiber–bacteria interactions, the fluorescence probe allows direct observation and verification of fiber–enzyme and fiber–bacteria interactions. Donaldson & Vaidya (2017) used fluorescent protein‐tagged carbohydrate‐binding modules to label cellulose and lignin, along with a dye‐labeled enzyme to visualize the micro‐scale distribution of amenable and recalcitrant sites. In this regard, it would also be interesting to learn about fiber–bacteria interaction using fluorescence‐labeled bacteria to reveal more insights about bacteria selectivity and specific site binding. In addition to imaging, the fluorescence labeling technique can also be used to characterize the folding state of DF molecules. For example, the pyrene probe has been used to indicate the compactness of pectin, and Zhao et al. (2020) used pyrene to interact with the hydrophobic portion on pectin's surface. A greater fluorescence intensity of pyrene refers to greater exposure of the hydrophobic portion, which is an indication of high compactness since the polymer chain could not freely re‐fold to minimize its interfacial energy.

4. THE INTERPLAY BETWEEN FIBER AND MICROBIOME

4.1. DF degradation by the gut microbiome

Degradation of DF to SCFAs via microbial fermentation in the large intestine is a complex enzymatic and biological reaction and it involves the collaboration of multiple bacteria. A proposed flow chart of DF fermentation from DF to butyrate is shown in Figure 3 (adapted from Baxter et al., 2019). The degradation of resistant polysaccharides typically goes through two stages: the primary degradation and the secondary degradation. Primary degraders (e.g., Bifidobacterium spp. and Ruminococcus bromii) depolymerize resistant polysaccharides into mono‐, di‐, and oligo‐saccharides. Afterward, other bacteria carry the secondary degradation and further convert degraded polysaccharides into SCFAs including acetate, butyrate, propionate, and lactate. According to Hamaker and Tuncil (2014), fiber utilization requires carbohydrate‐active enzymes to cleave specific types of chemical bonds. For example, resistant starch primarily contains α‐1,4 and α‐1,6 glycosidic bonds, which are resistant to amylolytic enzymes. Therefore, only a limited number of gut bacteria capable of producing α‐amylase, glucoamylase, isoamylase, and pullulanase can degrade resistant starch into smaller segments. In the large intestine, the scenario is complex because multiple bacteria compete for the same DF. Ze et al. (2012) investigated the competition of four different amylolytic bacteria (R. bromii, Bifidobacterium adolescentis, Eubacterium rectale, and Bacteroides thetaiotaomicron) to degrade resistant starch and monitored the growth of each species. Their results revealed that the growth of these four amylolytic bacteria significantly varied with the type of starch and the amylose‐to‐amylopectin ratio.

FIGURE 3.

Proposed model of metabolites and microbes that catalyze the flow of carbon from resistant polysaccharides to butyrate. Modified from Baxter et al. (2019).

Depending on the complexity of DF molecules, the interaction between fiber and microbiome vary significantly. If the monomer composition of DF is identical, its degradation process may follow reproducible procedures, ultimately leading to a less diverse microbiome community in the gut (Cantu‐Jungles & Hamaker, 2020). However, the scenario changes when heterogeneous DF (e.g., plant pectic polysaccharide rhamnogalacturonan‐II containing 13 different sugars and 21 distinct glycosidic linkages arranged as a backbone) is orally administrated. A recent study reported the degradation pathway of rhamnogalacturonan‐II by B. thetaiotaomicron (Ndeh et al., 2017). In addition to various essential enzymes required to cleave 21 distinct glycosidic linkages, the successful removal of the borate steric barrier is also necessary to ensure complete degradation. Although B. thetaiotaomicron is capable of conducting every step to depolymerize rhamnogalacturonan‐II, such a comprehensive process more likely involves a collaboration of different species, resulting in a more diverse bacterial community. Xue et al. (2020) characterized three SDFs from Lentinula edodes byproducts with different MWs and evaluated the changes in the digestive system and the effects on large intestinal fermentation. Their results indicated that branched chain structure could increase the fermentation rate of the SDFs but not affect the concentration of total SCFAs produced.

For the degradation of cellulose, which is connected solely by β‐1,4 glycosidic bond, its degradation pathway still involves a collaboration of different microorganisms and enzymes. As shown in Figure 4, the primary degradation of cellulose to cellobiose (two glucose units) usually occurs outside of bacteria cells. Then, cellobioses are transported into bacteria cells, where the β‐1,4‐glucosidase enzyme breaks down the cellobioses into glucose molecules. Afterward, the glucose could be directly fermented. During the primary degradation, four enzymes involved are endo‐β‐1,4‐glucanase, exo‐β‐1,4‐d‐glucanase, cellobiase, and β‐glucosidase. In detail, endo‐β‐1,‐4‐glucanase specifically breaks the internal bonds to disrupt the cellulose structure and expose individual polysaccharide chains; exo‐β‐1,4‐d‐glucanase can access the chains from both reducing/nonreducing ends of the exposed chains and produce tetra‐saccharide or disaccharide such as cellobiose; finally, cellobiase or β‐glucosidase hydrolyzes cellobiose to release d‐glucose units. Throughout this process, all enzymes are essential in breaking insoluble cellulose into fermentable glucose.

FIGURE 4.

(a) The breakdown of cellulose by endocellulase, exocellulase, and cellobiase, reprinted from Hii (2012) with permissions (Hii et al., 2012) (b) Degradation pathway of cellulose. Reprinted from Datta (2024) with permissions (Datta, 2024)

It should be noted that none of the enzymes alone can break down the complex crystalline cellulose structure, but instead, a synergistic collaboration is required (Chandra et al., 2015). Even so, degradation of crystalline cellulose is still considered a highly rate‐limited process, during which the binding between enzymes and the surface of substrate particles is crucial (Zhang & Lynd, 2004). Many factors could impact the binding efficiency, such as micro–macro structures, polymorphism, and surface properties, which could play significant roles in determining the accessibility for enzymes to attack cellulose. In the meantime, nonspecific binding also occurs competitively to hinder the enzymatic reactions, for example, the presence of lignin on fiber surface could nonspecifically absorb enzymes and impede the degradation process (Vanholme et al., 2010). As DFs encompass a complicated packing architecture of different components, enzyme accessibility could be more difficult to predict.

Table 5 summarizes the fermentation of different fiber types and the distribution of their corresponding products. Across different fiber types, the fermentation product distribution varies significantly, with acetate typically being the dominant SCFA produced. Arabinoxylan from various sources produces mostly acetate (60%), followed by propionate (30%) and butyrate (10%) (Rumpagaporn et al., 2015), while resistant starch from potato yields a different ratio with higher propionate (45%) and lower butyrate (5%) (Warren et al., 2018). Citrus and sugar beet pectin fermentation results in a more balanced SCFA distribution with acetate (55%), propionate (25%), and butyrate (20%), suggesting that fiber source and structure influence fermentation outcomes (Larsen et al., 2019).

TABLE 5.

Fiber type and fermentation products.

Fiber type	Origin	Fermentation type	Approximate products distribution	Key findings	Reference
Arabinoxylan	Wheat bran, corn, rice bran	In vitro batch fermentation using human fecal inoculum	Acetate (60%), propionate (30%), butyrate (10%)	Terminal xylose residues attached to the xylan backbone is associated with slow fermentation	Rumpagaporn et al., 2015
Resistant starch	Potato	In vitro fermentation using human fecal microbiota	Butyrate (5%), acetate (50%), propionate (45%)	Native potato starches showed highest butyrate production	Warren et al., 2018
Pectin	Citrus and sugar beet	In vitro fermentation using human gut microbiota	Acetate (55%), propionate (25%), butyrate (20%)	Propionate was largest in fermentations of the high methoxyl pectins	Larsen et al., 2019

4.2. Impact of DF on the composition of gut microbiota

While there is a debate on the definition of a healthy gut microbiome, generally, Bifidobacterium spp., Lactobacillus, Lachnospiraceae, and Ruminococcaceae are considered beneficial to human health, and Escherichia coli, Clostridium perfringens, and Enterococcus spp. are potential pathogens in the gut (Cui et al., 2019). DF has a capability to promote beneficial bacteria but inhibit pathogens through several known mechanisms, including providing specific nutrients to beneficial bacteria, lowering pH through SCFA production to suppress pathogenic bacteria, and modulating the gut‐associated immune system to recognize and combat pathogens (Makki et al., 2018). Inulin and pectin are the two well‐studied DFs to modulate the composition of gut microbiota and their metabolites. For example, it was reported that inulin‐type fructans (ITFs) were able to modulate the composition and activity of the gut microbiota in obese women (Salazar et al., 2015). Specifically, the administration of 16 g/day of ITF for 3 months significantly increased Bifidobacterium spp. populations while reducing total SCFA in the gut. Xu et al. (2019) reported that pectin‐enriched diets increased the richness of microbiota, increased the relative abundances of Firmicutes and Proteobacteria, and decreased the abundance of Bacteroidetes in pig colons. In recent years, DFs from different plant sources have been evaluated for their capabilities to modulate the composition of gut microbiota. Liu et al. (2020) reported that DF isolated from sweet potato residues resulted in a significant increase in Bifidobacterium and Lactobacillus and a decrease in Enterobacillus, Clostridium perfringens, and Bacteroides. Yang et al. (2020) demonstrated that both IDF and SDF extracted from soy hulls were able to regulate the gut microbiota by increasing the abundance of Bifidobacteriales and Lactobacillales in in vitro fermentation. Praveen et al. (2019) systematically reviewed the use of seaweed DF and concluded that DF is a potential untapped bioresource to be explored as a prebiotic for immunomodulation on gut microbiota. Pérez‐Burillo et al. (2020) reported that the incorporation of DFs (citrus fiber, arabinogalactan) into salami formulation promoted an increase in the abundance of polysaccharide‐degrading genera, resulted in a reduction in Escherichia genera, and led to an increased amount of SCFA production in in vitro gut microbial fermentation.

Although it is generally agreed that DF has the ability to change the composition of gut microbiota, little is known about how this is achieved in a precisely predictable way. This is partially because of the complexity of both DF structure and gut microbiota, and because there are other factors determining microbial composition, such as the other food components (e.g., protein, fat) co‐taken with DFs, host immune system, host genetics, and gut environmental factors. To connect the DF structure, gut microbiota, and human health, Hamaker and Tuncil (2014) introduced the concept of “discrete fiber structure.” The concept emphasizes the importance of identifying the enzymes secreted by bacteria for fiber degradation. As illustrated in Figure 5 using pectin as an example, DF is regarded as a highly heterogeneous molecule. Its utilization by gut microbes depends on the presence of diverse monomers and functional groups, which require specific carbohydrate‐active enzymes encoded by microbial genes to cleave particular chemical bonds. Since gut microbes live in a highly competitive environment, the capability to degrade or partially degrade specific types of fiber for energy uptake is favored for survival. In this regard, the selective growth by introducing specific types of DF could lead to a shift in the microbiota population (Baxter et al., 2019), and a linkage can be built to connect fiber chemistry, microbe genomics, and the evolution of the gut microbiota as a consequence of different fiber intake. Accordingly, targeted shifting of gut microbiota could be achieved through the oral administration of selected DFs. This concept was later supported by Deehan et al. (2020), who found that crystalline and phosphate cross‐linked starch structures induced divergent and highly specific effects on gut microbiota composition in a human trial with three type‐IV resistant starches. However, inconsistent results have also been observed across individuals (Cantu‐Jungles & Hamaker, 2020), which could be due to several reasons. First, several different microbes can utilize the same DF, which can induce more competition stress; second, the complex structure of DF could result in inaccessible sites for enzymatic degradation; and third, individual microbiota response varies among the population.

FIGURE 5.

The polymeric chain structure and specific chemical bonds on pectin. Reprinted from Hamaker and Tuncil (2014) with permissions.

Table 6 provides an overview of various DFs, their origins, fermentability levels, and their impacts on gut microbiota. Highly fermentable fibers such as inulin, fructooligosaccharides (FOS, both branched and linear), and β‐glucans significantly promote beneficial gut bacteria like Bifidobacterium, Anaerobutyricum, and Bacteroides while reducing harmful ones like Bilophila and Proteobacteria. Pectin and resistant starch show varied impacts on microbial communities, with resistant starch particularly encouraging the growth of Bifidobacterium species and reducing potentially pathogenic bacteria like Alistipes putredinis. Moderately fermentable hemicellulose also supports beneficial microbes such as Bifidobacterium and Lactobacillus. Low‐fermentability fibers, including cellulose, lignin, and chitin, are less impactful but still contribute to increasing specific beneficial bacteria such as Enterorhabdus, Akkermansia, and Lachnospiraceae. This highlights the diverse effects of fiber types on gut microbial composition, underlining their role in gut health management.

TABLE 6.

Fiber type, origin, and their impacts on gut microbiota.

Fiber type	Origin	Fermentability level	Impact on gut microbiota	Reference
Inulin	Chicory	High	↑Bifidobacterium ↑Anaerostipes ↓Bilophila	Vandeputte et al., 2017
Fructooligosaccharides (FOS)—branched	Grass	High	↑Butyricicoccus ↑Erysipelotrichaceae ↑Phascolarctobacterium ↑ Sutterella	Popov et al., 2024
Fructooligosaccharides (FOS)—linear	Chicory	High	↑Anaerobutyricum ↑Lachnospiraceae ↑Faecalibacterium	Popov et al., 2024
β‐Glucans	Lentinus edodes	High	↑Bacteroides ↓Proteobacteria ↑Muribaculum ↓Helicobacter	Liu et al., 2023
Pectin	NA	High	↑Firmicutes ↑Proteobacteria ↓Bacteroidetes	Xu et al.2019,
Resistant Starch	Maize	Various fermentability	↑Bifidobacterium adolescentis ↑Bifidobacterium longum ↑Ruminococcus bromii ↓Alistipes putredinis ↓Bacteroides vulgatus ↓Odoribacter splanchnicus ↓Parabacteroides merdae	Li et al, 2024
Hemicellulose	Wood	Moderate	↑Bifidobacterium ↑Lactobacillus ↑Bacteroides	La Rosa et al., 2019
Cellulose	NA	Low	↑ Bifidobacterium	Nsor‐Atindana et al., 2020
Lignin	Birch	Low	↑Enterorhabdus ↑ Akkermansia	Lahtinen et al., 2023
Chitin	Crickets	Low	↑Lachnospiraceae	Refael et al, 2022

4.3. DFs, gut microbiota, and the gut–brain axis

DF plays a crucial role in modulating the gut microbiota and influencing brain function through the gut–brain axis (Berding et al., 2021; Ito et al., 2024; Sun et al., 2021). The fermentation of DFs by gut bacteria produces important metabolites, particularly SCFAs, which are key signaling molecules in microbiota–gut–brain communication (La Torre et al., 2021). The gut microbiota influences brain function and behavior through multiple pathways, including immune, endocrine, neural, and humoral routes (Berding et al., 2021). SCFAs can affect the brain by improving central, peripheral, and systemic immunity, enhancing intestinal barrier integrity, and modulating neurotransmitter systems (Berding et al., 2021). For example, SCFAs can stimulate the production of serotonin and other neurotransmitters, regulate the hypothalamic–pituitary–adrenal axis, and influence neuroinflammation (Ito et al.,2024). Recent evidence suggests that DF intake can positively impact cognitive function and mental health through these microbiota‐mediated mechanisms (Berding et al., 2021).

Different types of DFs, each with unique physicochemical properties including solubility, viscosity, and fermentability, can influence brain function through distinct pathways (La Torre et al., 2021). Soluble fibers, such as inulin and pectin, are readily fermented by gut bacteria, leading to the production of SCFA and supporting the growth of beneficial bacteria like Bifidobacterium and Lactobacillus (Ferreira‐Lazarte et al., 2018), which have been associated with improved cognitive performance (Ruiz‐Gonzalez et al., 2024). The impact of insoluble fiber on gut–brain axis was not extensively conducted; a recent work revealed a surprising negative impact of cellulose on the gut–brain axis, contrary to previous assumptions about DF benefits (Ito et al., 2024). The study found that cellulose‐rich diet decreases SCFAs and causes gut dysfunction, including increased intestinal permeability, decreased motility, and heightened intestinal hypersensitivity, through upregulation of TRPA1 receptors. These gut disturbances trigger vagus nerve signaling that activates the brain's opioid system and increases dopamine in the amygdala, ultimately leading to increased anxiety‐like behaviors in mice. While the evidence supporting the role of DF in brain health is growing, many questions remain unanswered. Future research needs to better understand how specific fibers affect brain function, determine optimal intake levels, and account for individual variations in response. This understanding could lead to more targeted dietary interventions for supporting cognitive health and emotional well‐being through fiber supplementation.

5. MODIFICATIONS OF DF STRUCTURES

5.1. Physical treatments

Grinding is a conventional approach to reduce fiber size. The size has been reported to greatly impact the fermentability of DFs and SCFA production in the gut (Nsor‐Atindana et al., 2020). Size reduction is also associated with increased surface area and pore volume, which further affect the hydration properties and enzymatic accessibility of DF. Raghavendra et al. (2006) grinded coconut residue to a particle size ranging from 550 to 1127 µm and found that the water‐holding capacity peaked when the particle size was close to 550 µm. In another study, Chitrakar et al. (2020) applied low‐temperature ball milling to reduce the particle size of asparagus leaf‐derived DF from 62.3 to 7.5 µm. The authors reported that ball‐milled products exhibited better hydration properties, especially water solubility, and a higher fraction of SDF compared with unmilled samples. In a similar study, Song and others applied enzyme‐assisted ball milling treatment to modify the structural properties of citrus fibers. They found that the treated citrus fibers had looser and more porous microstructure, which leads to a higher enzymatic and microbial accessibility (Song et al., 2021).

Extrusion is another popular method to modify DF's structure to enhance its functionality and fermentability. It is a complex process where the raw material is subjected to a high‐temperature and high‐pressure treatment for a period of time. During extrusion, a series of chemical reactions and thermochemical conversion can occur, including starch gelatinization and the Millard reaction (Feng & Lee, 2014). Recently, extrusion has been used to treat DFs to improve their functional and fermentation properties. Arribas et al. (2017) investigated the impact of extrusion on whole grains and found that the quantity of resistant starches was significantly increased due to the binding between lipids and amylose at a high temperature. Furthermore, there was also a significant increase in the fraction of SDF after extrusion. Andersson et al. (2017) found that the extrusion increased the extractability of DFs from wheat and rye and can potentially enhance the fiber's fermentability in the gut. Demuth et al. (2021) reported that extruded wheat bran arabinoxylan promoted butyrate production and growth of butyrate‐producing Faecalibacterium in the butyrogenic microbiota and inhibited the growth of Prevotella in the propiogenic microbiota. A recent study found that the particle size of okara fiber can be reduced from 283.2 µm to around 78.76 µm after six extrusion cycles at 350 rpm of screw speed with 1 L/h water addition (Kantrong et al., 2024). The particle size reduction was influenced more by the number of extrusion cycles than by screw speed or water content, with multiple passes through the extruder leading to progressively smaller particles, demonstrating the effectiveness of the shear forces generated during extrusion for particle size reduction.

Thermal treatment is simple but effective to modify fibers. The thermal modification of DF has been studied using a water bath treatment (98 ‐ 100°C) and wet and dry heat treatments (Bader Ul Ain et al., 2019). These treatments disrupted the original plant cell wall structure (Crowe et al., 2017; Jin et al., 2018) and altered the soluble‐to‐insoluble fiber ratio (Bader UI Ain et al., 2019). It is commonly accepted that high temperature treatments can break glycosidic bonds in polysaccharides, leading to the formation of small molecules (e.g., oligosaccharides) (Bader Ul Ain et al., 2019).

Ultrasonic treatment is an environmentally friendly physical technique to extract, degrade, and modify biopolymers at relatively low cost (Wang et al., 2021) and has been used for DF modifications (Martinez‐Solano et al., 2021). A recent study investigated the impact of high‐intensity ultrasound effects on both citrus and apple fibers and found that the soluble fiber content increased from 5.5% to 17% for citrus fiber and from 8% to 17.6% for apple fiber. Also, the water‐binding capability of citrus fiber increased from 18.2 to 41.8 g/g. Furthermore, the viscosity of the citrus and apple fiber increased from 1.4 and 1.34 to 84.4 and 31.7 Pa·s, respectively, at the shear rate of 100 s⁻¹ (Kalla‐Bertholdt et al., 2023). A similar trend was also reported in another study using chia fibers (Hassan et al., 2021). Ultrasound generates cavitation effects where the rapid formation and collapse of bubbles create intense localized pressure (Zhang et al., 2013), temperature, and shear forces that physically disrupt fiber structures through two main pathways, breaking down larger fiber aggregates into smaller particles and disrupting chemical bonds and crosslinks between polysaccharide molecules (Islam et al., 2014). As such, these mechanical and chemical modifications lead to increased surface area and exposure of functional groups (Chen et al., 2011; Hu et al., 2013), enhanced solubility through partial conversion of insoluble to soluble fiber, and improved water‐binding capacity through greater accessibility to hydrophilic groups.

High‐pressure processing is a nonthermal approach that has been used to pasteurize food products with a pressure range from 300 to 600 MPa (Arshadi et al., 2016). Compared to ultrasound treatment, the impact of high‐pressure processing on DF has not been extensively explored. A recent work has found that high‐pressure processing was an effective method for modifying DF, particularly in kelp powder (Zhao et al., 2024). High‐pressure processing significantly improved the fiber's functional properties, increasing a 1.31‐fold increase in water holding capacity at 600 MPa for 10 min, a 0.12‐fold increase in swelling capacity with a 10‐min treatment at 450 MPa, and 1.33‐fold increase in oil holding capacity at 600 Mpa for 10 min. The treatment also reduced particle size, created a more porous structure, and exposed more binding sites, which is correlated to enhanced water‐holding capacity and adsorption properties. High‐pressure processing also converted IDF to SDF and increase the SDF content by 63% at 600 Mpa for 10 min. Overall, the optimal processing conditions were found to be 450 MPa for 10 min for improving the functionality of kelp powder fiber. Future research is needed to explore more applications of high‐pressure processing‐treated DFs as a functional ingredient in various food products like baked goods and beverages, including its effects on product quality and consumer acceptance.

5.2. Chemical and thermochemical treatments

Chemical treatments using acids and alkalis have been used to modify the composition and structural properties of DF, mainly through removing impurities (e.g., starch, protein, and minerals), hemicellulose, and lignin. Qi et al. (2016) treated rice bran IDF with different concentrations of H₂SO₄, followed by KOH, to modify its chemical and structural properties. The authors found that the treatment reduced the amount of xylan, rhamnan, and fructan in the fiber mainly due to the removal of hemicellulose by strong H₂SO₄. These alterations in chemical composition led to changes in the microstructure of the DF, which consequently affected its functional properties, including oil‐binding capacity and water‐holding capability. Alkaline hydrogen peroxide (AHP) is another popular chemical treatment to modify DFs as it can solubilize hemicellulose and lignin (Cui et al., 2019). Zhang et al. (2020) reported that the AHP treatment significantly degraded hemicellulose, lignin, and pectin, as well as increased the total fiber content, damaged the crystal structure, and loosened the microstructure of citrus fiber. All these changes led to the improved physicochemical properties and fermentability of the fiber. The AHP treatment was also used to improve the physicochemical properties of DF from buckwheat straw (Meng et al., 2019). Although chemical treatments have the capability to modify the chemical and structural properties of DFs, the main drawback of these treatments is the heavy use of nonenvironmentally friendly chemicals, such as strong acids and alkaline. Recently, Jin et al. (2024) treated milled corn fiber with a dilute alkali solution. They reported that the alkali‐treated corn fiber had enhanced water retention, swelling, and oil‐holding capacities. Moreover, the alkali‐treated corn fiber showed significantly increased SCFA production in in vitro fermentation, which was probably due to its loose structure and decreased ester linkages after the treatment.

Subcritical water processing technology has recently emerged as an eco‐friendly, low‐cost, and effective method to modify DFs. At the subcritical condition, water is heated above its boiling point (100°C and 0.1 MPa) but below its critical point of 374°C at 22.1 MPa. In its case, several properties of water, such as viscosity, diffusivity, polarity, and density, are changed (Cheng et al., 2021). Such changes in thermochemical properties help subcritical water to perform several kinds of chemical transformations, including a dramatic increase in mass transfer and hydrolytic reactions. Besides the property changes, the water molecules at the subcritical state are dissociated to produce hydrogen ions (H⁺) and hydroxyl ions (OH⁻) with a dissociation constant (K _w) of 10⁻¹¹ mol/L, which is three orders of magnitude higher than the K _w of water at room temperature. The high levels of H⁺ and OH⁻ at subcritical conditions allow water to act both as a solvent and an acid/base catalyst to modify and break down the recalcitrant DFs. Recently, Su et al. (2024) applied subcritical water treatment to modify IDF from brewer's spent grain for enhancing its functionality and fermentability. Their results showed that the DFs treated with subcritical water had increased surface area and porosity, enhanced functional properties (e.g., water and oil holding capacities), and improved fermentability in in vitro fermentation. Moreover, by combining the subcritical treatment with a low concentration of weak acid (lactic acid), the fermentability of DF can be further enhanced. In another study, Rudjito et al. (2020) used a subcritical water technique to tune the molar mass and substitution pattern of complex SDF, glucuronoarabinoxylan (GAX). They reported that elevated temperatures in unbuffered conditions led to a significant pH decrease and autohydrolysis, resulting in reduced molar mass (approximately 10⁴ Da) and cleavage of arabinose substituents. By mitigating the pH reduction through mild buffered neutral and alkaline conditions, higher molar masses of the extracted GAX (approximately 10⁵ Da) were achieved. This approach also protected the labile arabinose substituents, leading to a greater prevalence of more complex glycan side chains.

5.3. Enzymatic treatment

Enzymatic treatments are ecofriendly approaches particularly suitable for the food industry (Spotti & Campanella, 2020). During enzymatic reactions, DFs are partially segmented into shorter polymer chains, exposing more surface functional groups and resulting in different physicochemical properties. Shen et al. (2020) used a combination of enzymes (α‐amylase, neutral protease, and glucosidase) to modify the DF extracted from black soybean hulls, and the modified fiber presented significantly higher cholesterol binding capacity. Recently, enzymatic treatment was applied to cocoa bean shell‐derived DF to improve its fermentability (Disca et al., 2024). The DF was treated with a cellulase mixture to induce a structural change, leading to a boosted production of SCFAs in colonic fermentation.

Enzymatic treatments are also used in combination with other treatments to enhance their capacity to modify DFs. For example, Lamothe et al. (2021) applied the combined enzymatic and microwave treatment to modify isolated pearl millet fiber for increasing its accessibility to gut bacteria. The results showed that the treatment increased the amount of insoluble fiber fermented in vitro from 36% to 59%. More promisingly, the treatment doubled butyrate production and almost tripled acetate production after 6 h of fermentation compared to the native millet fiber. In addition, the treated millet fiber had an increased Firmicutes/Bacteroidetes ratio with relative abundance increases in Blautia and Coprococcus genera and a decrease in Bacteroidetes.

All the aforementioned studies have demonstrated that enzymatic treatments can effectively modify DFs to improve their microbial accessibility and functionality. Compared with physical and chemical treatments, enzymatic treatment has clear advantages of low environmental pollution and energy consumption. Moreover, they are mild processes with lower byproduct generation compared with physical and chemical treatments. However, a major challenge is the high cost of enzymes, whose unit cost can be 10 times higher than chemicals (e.g., acids and bases). Therefore, enzyme dosage optimization and recycling are crucial for using enzymes to modify DFs.

6. CONCLUSION AND FUTURE PERSPECTIVES

This review discussed the structural properties of DF in a broad scope, from molecular composition and functional groups to polymorphism (angstrom to nanometer), mesoscopic structures (10–1000 nm), porosity (submicron), and particle size (microscale). The structural properties of DF greatly impact its physicochemical properties, enzymatic digestibility, and interactions with the gut microbiome based on the summarized literature. Moreover, advanced characterization techniques (e.g., XRD, nitrogen absorption/desorption, SAXS, fluorescence labeling) used to reveal the structural insights of fibers were elaborated upon with related examples. Finally, appropriate processing techniques to modify the structural and chemical properties of DF were discussed in detail.

• The properties of fibers are covered in a broader scope, from molecular composition and functional groups to polymorphism (angstrom to nanometer), mesoscopic structures (10–1000 nm), porosity (submicron), and particle size.

• Innovative characterization techniques to reveal the structural insights of fibers were elaborated, including XRD, BET, SAXS, as well as those examples.

• The impact of structural parameters on enzymatic and microbial activities was thoroughly discussed. For example, the CI of fiber has been found to remarkably impact its enzymatic reaction rates, while its fermentation fate and rate have not been systematically evaluated. In addition to crystallinity, DFs also differ from each other in structural compactness, porosity, degree of branching, pore wettability, and many other aspects. These structural differences may also significantly affect the fermentation performance and bacterial selectivity, due to different enzymatic accessibility and microbial affinity.

• The interplay between fiber and gut microbiome was covered. We use cellulose as an example to demonstrate the fiber degradation process in the gut, where a series of microorganisms and enzymes function collaboratively and competitively to interact with the fiber molecule.

• Processing strategies to modify fiber structures were discussed. We introduced and compared key processing technologies for modifying the chemical and structural properties of DFs, as well as their subsequent effect on the fermentability and capability to modulate the gut microbiome.

Although numerous studies have been conducted to understand the interaction between DF and gut microbiota, there are still many uncertainties due to the complex nature of DF and inconsistencies in the response of gut microbiota to DF. To date, most studies emphasize the composition of fiber, while the identity and importance of structural features in DFs were less addressed. Future work may focus on resolving the intriguing puzzle with a combination of advanced material characterization methods, unique fiber processing techniques, and advanced computational techniques. As mentioned, one major challenge in DF research is how complex and diverse the chemical and structural properties of DF are. Therefore, it is critically important to apply advanced material characterization methods to thoroughly understand the chemical and structural properties of DF at different scales before, during, and after microbial digestion. This information is vital for shedding light on the interaction between DF and gut microbiota, which in turn will lead to the development of fiber processing techniques that target specific chemical and structural properties.

Moreover, advanced computational techniques, such as machine learning, will play a foreseeable bigger role in future studies. Because the bacterial community within the gut is an extremely complex system that involves unpredictable dynamic competition and cross‐talk, digestion of DF by gut microbiota is often treated as a black box (Marchesi & Prosser, 2008). To date, sequencing 16S ribosomal RNA is the most widely applied and cost‐effective approach to reveal the gut microbiota community, and whole‐genome sequencing has been increasingly used to reveal the gut microbiota composition at a strain level. However, due to the vast data size of gene sequencing, the analysis of sequencing data requires exceptionally robust and efficient algorithms. Applying powerful machine learning tools based on big data will be an effective approach to gaining a deep understanding of the interplay between DF and the gut microbiome and to precisely predict the consequence of specific DF consumption.

AUTHOR CONTRIBUTIONS

Yiming Feng: Conceptualization; formal analysis; data curation; investigation; methodology; writing—original draft. Qing Jin: Investigation; methodology; writing—review and editing. Xuanbo Liu: Investigation; writing—review and editing. Tiantian Lin: Investigation; writing—review and editing. Andrea Johnson: Investigation; writing—review and editing. Haibo Huang: Conceptualization; supervision; funding acquisition; project administration; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Feng, Y. , Jin, Q. , Liu, X. , Lin, T. , Johnson, A. , & Huang, H. (2025). Advances in understanding dietary fiber: Classification, structural characterization, modification, and gut microbiome interactions. Comprehensive Reviews in Food Science and Food Safety, 24, e70092. 10.1111/1541-4337.70092