Dietary fibers are complex carbohydrates that escape digestion in the upper gastrointestinal tract and reach the colon, where they serve as substrates for microbial fermentation. Fermentable fibers include a broad range of plant-derived polysaccharides, such as pectin, arabinoxylan, beta-glucans, fructo-oligosaccharides (FOS), galacto-oligosaccharides (GOS), inulin, xyloglucans, and resistant starches (20, 21). These fibers differ in solubility, structure, and fermentability, and have been studied for their influence on gut microbial composition, particularly on fiber-degrading taxa, and the concentration of microbial metabolites (22). While humans lack the enzymes required to break down plant cell wall components, the majority of gut microbes have the functional capacity to do so, and to use them as carbon and energy source (23, 24). The main fermentation end products from fiber are short-chain fatty acids (SCFAs), which have been extensively studied because of their role in metabolic health (25). SCFAs are organic acids containing two to six carbon atoms, with acetate (C2), propionate (C3), and butyrate (C4) being the most abundant, and produced in a ratio of approximately 3:1:1 (26). The key SCFAs, their microbial producers, and host-relevant functions are summarized in Table 1.

The ability to generate SCFAs is functionally redundant across taxonomically distinct bacteria (27): acetate is formed by the vast majority of gut bacteria while butyrate and propionate are produced by subsets of bacteria that form functionally distinct groups, and multiple products can be generated by the same species (28).

Cross-feeding also takes place; e.g., acetate can increase butyrate production, noted between acetate-producing Bifidobacterium and butyrate-producing Faecalibacterium (29). Examples of propionate-producing genera are Akkermansia, Bacteroides, Dialister, Phascolarctobacterium, and Phocaeicola via primarily the succinate pathway, and Anaerobutyricum (formerly Eubacterium halii), Blautia and Mediterraneibacter via the propanediol pathway (30, 31).

Butyrate is mostly formed by genera of the highly oxygen-sensitive anaerobic Clostridia families Lachnospiraceae and Ruminococcaceae, and examples include Agathobacter, Anaerobutyricum, Anaerostipes, Butyricicoccus, Coprococcus, Faecalibacterium, Gemminger, Lachnospira, Oscillibacter, Roseburia and Ruminococcus (31–33).

SCFAs serve as signaling molecules, energy supply, and regulators of metabolism (e.g., insulin sensitivity and fat storage), the immune system, and the gut barrier.

Butyrate provides 70% of the energy requirement of the colonic epithelium (34), facilitates tight junction assembly and promotes wound healing of the intestinal epithelium (35, 36). Butyrate (and acetate) can stimulate mucin 2 (Muc2) expression, mucus production and secretion (37).

SCFAs exert their functions by signaling through surface-expressed (free fatty acid) G-protein coupled receptors on epithelial cells, fat cells, and immune cells (38) or via histone deacetylase (HDAC) inhibition (39). Both signaling routes can regulate T-cell differentiation to induce IL-10-producing regulatory T-cells (40, 41). Butyrate is especially associated with intestinal and immuno-modulatory functions (19, 42, 43), due to its inhibitory effect on NF-κB (44) as well as IL-12 and IFN-γ (45), which play a role in chronic low-grade inflammation (46).

Similarly, propionate was shown to be inversely regulated by fasting and refeeding, and to reduce inflammatory CD4⁺ T cell responses by inhibiting NF-κB activity and histone deacetylase activity, leading to lower levels of IL-6, IFN-γ, and IL-17 (47).

While our focus is on fiber- and polyphenol-rich dietary interventions, it is important to understand the complete metabolic landscape of the colon, where the availability and type of substrates change along its length, influencing microbial metabolism and the resulting health effects.

In the first part of the large intestine (the proximal colon), microbes primarily ferment carbohydrates, the amount of which gradually decreases toward the descending colon (48), where microbes are specialized to harvest energy from residual peptides and proteins, yielding a more diverse array of metabolic products compared to the fermentation of dietary fibers (23).

Of these metabolites, branched-chain fatty acids (BCFAs) have gained attention because of their association with metabolic imbalances and poor colonic health (49–51). BCFA levels are influenced by diet, and reduced when changing from a Western to a Mediterranean diet, and negatively correlate to butyrate- and acetate generating bacteria, and microbial diversity (52).

Metabolite profiles of prediabetic individuals can feature increased levels of BCFAs (53). Prevotella copri and Bacteroides vulgatus are two examples of microbes that were found to drive the association between BCFA synthesis and insulin resistance (53). Besides metabolic imbalances, there is also a link between BCFAs and cancer development. A study found that BCFAs produced by Clostridium symbiosum led to increased cholesterol synthesis via mTORC1, in turn activating hedgehog signaling, resulting in increased colorectal cancer stemness and tumor growth in mice (54). Other (unfavorable) metabolites from amino acids include the uremic toxin p-cresol, produced from the fermentation of tyrosine, and associated with chronic kidney disease (55), generated by, e.g., Clostridium difficile (56).

Some amino acid metabolites are thought to benefit human health. The effects on colonic and systemic health have been noted for tryptophan metabolite indolepropionate and phenylalanine-metabolite phenylpropionate. Indolepropionate can be produced by members of the Clostridium genus or by a two-step transamination and reduction reaction by lactic acid bacteria and Bifidobacterium (57). Several Clostridium species were found to generate indolepropionate: C. sporogenes (58), three strains of C. cadaveris (59), the toxin-producing C. botulinum (60), as well as Peptostreptococcus anaerobius CC14N (59). An important link between the generation of beneficial tryptophan metabolites and dietary fiber intake was established by Qi et al., who found that the intake of fiber-rich foods was most strongly associated with indolepropionate concentrations (50). Interestingly, a study deciphered a mechanism by which specific tryptophan metabolites were generated, which was independent of abundances of tryptophan-metabolizing gut bacteria, but regulated in a substrate-specific manner of metabolic pathways that involved different species and cross-feeding mechanisms (57). In more detail, in an in vitro experiment, indole-producing E. coli and indolepropionate-producing C. sporogenes were shown to compete for tryptophan, and fiber-degrading Bacteroides thetaiotaomicron moved the scale in favor of C. sporogenes, by cross-feeding monosaccharides to E. coli, making more tryptophan available to C. sporogenes (57). The results suggest that fermentable fibers can regulate indole production as a beneficial tryptophan metabolite (57), which is in line with studies that found positive associations of indolepropionate with fiber intake (61) but also with polyphenols (62). Finally, it was shown that fiber-mediated regulation of indole generation was not limited to particular species or communities, but is more of a typical phenomenon taking place in the human gut microbiota (57). Indolepropionate concentrations were found to be inversely associated with T2DM risk (50), and lower concentrations were present in patients with heart failure with preserved ejection fraction (HFpEF) in two separate cohorts (63). Indolepropionate is an ligand of aryl hydrocarbon receptors (AhR), and engages with other receptors including pregnane X receptor (PXR) and Toll-like Receptor 4 (63, 64). In a mouse model of HFpEF it was shown that indolepropionate supplementation attenuated diastolic dysfunction, oxidative stress and inflammation by enhancing the nicotinamide adenine dinucleotide salvage pathway, suppression of NNMT (nicotinamide N-methyltransferase) expression, and restoration of nicotinamide, NAD+/NADH, and SIRT3 levels (63). Indolepropionate was also shown to upregulate occludin, which is a tight junction protein (63), thereby strengthening the gut barrier and decreasing intestinal permeability (59). A similar protective action was observed for Bacteroides fragilis-derived phenylpropionate in pigs, activating AhR signaling and maintaining the integrity of the intestinal epithelial barrier (65). These examples show how microbial amino acid fermentation can result in metabolites with either beneficial or unfavorable effects on host health, as summarized in Table 2.

Polyphenols are a diverse group of bioactive plant compounds long recognized for their antioxidative capacity and health benefits (66). Most polyphenols have low bioavailability and require microbial transformation for better bioaccessibility (67–69). There are different classes of polyphenols, such as flavonoids (including flavonols and flavanols), ellagitannins, isoflavones and lignans, differing in chemical structure, complexity and origin, that are converted to phenolic acids by the microbiota (70). Proanthocyanidins are oligomeric flavanols, which are themselves a subclass of flavonoids (71), and are abundant in foods like spices, cocoa, beans, legumes, nuts, berries, grains, tea and fruits (72). Flavonols, on the other hand, are broken down by microbiota-mediated ring-fission, resulting in low molecular weight phenolic catabolites including benzoic acid, phenylacetic acid and phenylpropionic acid, hippuric acid, and valerolactones (70, 72, 73).

Phenolic compounds can modulate gut microbial composition and barrier integrity. For example, mice supplemented with epigallocatechin gallate, a polyphenol found in tea, featured increased Faecalibacterium, Bifidobacterium and Akkermansia abundances and butyrate production and an anti-inflammatory effect as well as an enhanced gut barrier function in a model of inflammatory bowel disease (IBD) (74).

As described above, fiber played an important role in the generation of indolepropionate from tryptophan. The presence of fiber is often necessary for the beneficial action of polyphenols as well, as described below, but fiber is not a mandatory mediator in this case. For example, a fiber-free diet combined with polyphenols indicated that polyphenol-supplementation can suppress mucus barrier degradation (75). In this experiment, mice fed a high-fat/high-sucrose (HFHS) diet were supplemented with a cranberry-rich extract or water. Akkermansia abundances then reached 30% relative abundance upon the HFHS diet supplemented with the cranberry extract, associated with a number of protective effects, including reduced weight gain and visceral obesity and blunted oxidative stress and inflammation compared to the HFHS control group. A potential explanation for the protective effect of polyphenols was indicated by the significantly higher expression of mucus-encoding Muc2 and tumor suppressing transcription factor Krüppel-like factor 4 (Klf4) in the colon (76) compared to HFHS controls. Significantly higher Muc2 expression points toward the higher production of mucin, thereby maintaining the mucus lining and providing the fermentation substrate for Akkermansia (75).

Ellagitannins and their hydrolysis product ellagic acid, found, e.g., in walnuts, berries, pomegranates, pecans and almonds, are known to be metabolized by Gordonibacter and Ellagibacter, certain Lachnospiraceae members and Enterocloster species into urolithins (77, 78). Enterocloster species harbor a urolithin C dehydroxylase operon, recently identified as a key mechanism for converting urolithin C to urolithin A, potentially explaining inter-individual differences in urolithin A production (78). Urolithin A has attracted particular interest due to its broad health benefits. It has been shown to increase lifespan in model organisms such as C. elegans and mice by enhancing mitophagy, improving cellular function, and alleviating systemic inflammation (79, 80). In a mouse model of colitis, urolithin A (derived from caffeic and ellagic acid) was associated with reduced inflammation and improved gut barrier function (81). In humans, urolithin A acts as a direct AhR ligand, repressing the transcription of pro-inflammatory mediators, including IL6 and prostaglandin-endoperoxide synthase 2 (PTGS2) (82–84). Additional studies have reported its role in enhancing mitophagy, improving mitochondrial function and reducing inflammatory responses (85) as well as its association with increased loss of visceral fat (86).

Urolithin A supplementation (1,000 mg) in elderly individuals improved muscle endurance and reduced inflammatory biomarkers while being safe and well tolerated (87).

Polyphenol-rich foods, including strawberries, can enhance urolithin A production and increase abundances of SCFA-producing bacteria. For instance, daily consumption of 500 g strawberries for nine days significantly increased urinary urolithin A and plasma antioxidant capacity (88), and a 10-week dietary intervention based on strawberries increased diversity and Faecalibacterium and Prevotella abundances (16).

Hippurate is another microbially derived metabolite that has been positively associated with gut microbial diversity, polyphenol metabolism, and diets high in fruit and whole grains (89, 90). Elevated hippurate levels have been linked to reduced risk of metabolic syndrome and increased visceral fat loss (86, 89, 91). Additionally, hippurate correlates with higher abundance of Faecalibacterium prausnitzii and lower levels of Ruminococcus and Eubacterium, suggesting a specific role in metabolic health regulation (89). Elevated blood hippurate has also been associated with increased adipose tissue expression of neuroglobin, a neuroprotective oxygen-binding protein primarily expressed in brain neurons (89, 92).

These two microbially derived metabolites exemplify the diverse host benefits of polyphenol metabolism by the gut microbiota.

The microbial fermentation of plant fibers can release bioactive phenolic compounds from the plant matrix. Many polyphenols are associated with dietary fibers through hydrophobic interactions, hydrogen bonding via hydroxyl groups, or covalent ester bonds; they may also accumulate inside vacuoles or become integrated in the plant cell wall (93, 94).

Polyphenol-bound dietary fibers, such as those in whole-grain cereals, have been experimentally shown to have higher antioxidant capacity than polyphenols alone (95). In grains like wheat, corn, rice and rye, the primary structural polysaccharide is arabinoxylan, a pentose sugar polymer, composed of a xylose backbone with arabinose side chains (96, 97). The phenolic compound ferulic acid is ester-linked to these arabinose side chains where it contributes to the structural integrity of the grain cell wall by crosslinking adjacent arabinoxylan polymers (98). The bran, which includes the aleurone and outer grain layers, contains approximately 90% of whole grain phenolic acids, primarily ferulic acid. Most of it is concentrated in the aleurone, which also stores proteins, phytate, and inorganic nutrients (98, 99). Upon reaching the colon, the antioxidant fiber is fermented by gut microbes, for example Bacteroides species can cleave the ester bond linking ferulic acid to arabinoxylan, releasing ferulic acid for absorption (100).

Zheng et al., demonstrated that phenolic acid metabolites and SCFAs can act synergistically to reduce inflammation in Caco-2 cells, a commonly used model of the intestinal epithelial barrier (73). The study showed that the combination of butyrate with any of three phenolic metabolites (phenyl acetic acid, benzoic acid, and phenyl propionic acid) at physiologically relevant concentrations significantly reduced IL-8 production more effectively than the individual compounds alone. The anti-inflammatory synergy was mediated through the downregulation of IL-8, TNF-α, and VCAM-1, at both gene and protein expression levels (73). The findings suggest that SCFAs and polyphenol metabolites may work synergistically to restore gut homeostasis by reducing inflammation (73).

Studies have shown that impaired barrier function is associated with thinning of the mucus layer, increased intestinal permeability, and, in older individuals, a slower turnover of the intestinal epithelium. These changes are influenced by both dietary factors and the host microbiota (101, 102). The gut barrier plays a critical role in preventing the translocation of microbes, endotoxins and food antigens into systemic circulation, a function essential for avoiding inflammation and sepsis (103, 104).

Fecal microbiota transfer studies highlight the role of the microbiota in barrier function and host aging. When microbiota from aged mice are transferred to young mice, the young mice exhibit increased intestinal permeability along with inflammation in the central nervous system and retina, and altered cytokine signaling (103). Conversely, transferring microbiota from young to old mice can reverse these effects (103).

In humans, intestinal permeability is often assessed using plasma zonulin levels, a physiological modulator of tight junctions secreted by intestinal epithelial cells in response to dietary or microbial stimuli (105, 106). Zonulin lowers the expression of tight junctions and thereby increases intestinal permeability. Elevated zonulin levels have been linked to aging, frailty, chronic obstructive pulmonary disease (107, 108), arthritis (106) and cognitive impairment, particularly during the progression to Alzheimer’s disease (AD) (109). In AD, changes in gut barrier integrity have been suggested as a potential trigger before AD onset (109). Furthermore, microbial encroachment into the inner mucus layer is a feature observed in dysglycemia, independent of obesity (110). Zonulin levels also correlate with metabolic and microbial profiles. Higher zonulin is associated with increased waist circumference, fat mass, and systemic inflammation (111). In contrast, lower zonulin levels are linked to greater microbial diversity, particularly higher abundances of butyrate-producing Faecalibacterium prausnitzii, and reduced abundances of Bacteroidaceae, Veillonellaceae, Bacteroides, and Blautia (112). These microbial differences are associated with healthier dietary patterns, including greater intake of fiber, omega-3 polyunsaturated fatty acids, vitamins, and minerals (112).

Studies in both model organisms and humans have shown dietary fiber is essential for maintaining a homeostatic relationship between host and microbiota.

This becomes especially evident when examining the consequences of fiber deprivation on gut microbiome composition and barrier integrity. In mice, fiber-deprivation leads to significant compositional changes in both the small and large intestinal microbiota (113–115). In the small intestine, a decline in segmented filamentous bacteria has been observed, coinciding with impaired intestinal Th17 and intraepithelial T-cell development and enhanced susceptibility to infection. These alterations were also observed in the offspring, but could only be reversed through fecal microbiota transplantation, not with a high-fiber diet alone (115).

In the large intestine, low-fiber diets are associated with higher abundances of glycan-degrading microbes, particularly Akkermansia muciniphila, and intestinal barrier dysfunction (101, 113, 116).

Low-fiber intake has also been linked to increased abundances of IgE-coated commensals, reduced mucus thickness, exacerbated allergic responses and increased inflammation (116). In addition, antitumor immunity is adversely affected by low-fiber intake (117).

Desai et al. demonstrated that both chronic and intermittent fiber-deficiency shifted the microbiota from saccharolytic fermentation to use of host-derived mucus glycans as a fermentation substrate. This shift was characterized by a rapid expansion of glycan-degrading bacteria such as Akkermansia muciniphila and Bacteroides caccae, alongside a reduction in polysaccharide-fermenting species, without changing plasma levels of propionate or butyrate. The resulting thinning of the mucus layer favored pathogen expansion, increased inflammation and morbidity (113).

Holmberg et al. further showed that fecal microbiota transplantation from individuals on a low-fiber diet failed to induce adequate mucus production in mice. However, increasing fiber intake by 14 g significantly shifted the microbial composition and restored commensal-mucus interaction, mediated by Blautia through its production of acetate and propionate (101).

Although A. munciniphila is often considered a beneficial genus due to its propionate-generating ability, recent evidence points to a context-dependent role. Derosa et al. found that baseline A. munciniphila abundance was a negative predictor of 12-month survival in non-small cell lung cancer patients undergoing immune checkpoint inhibitor therapy (118). Patients with A. muciniphila abundances below 4.8% (considered normal) had the longest median survival (27 months), while those with levels above 4.8% had the shortest (8 months). Those patients with no detectable A. muciniphila had a median survival of 16 months after treatment (118). Those in the normal range also exhibited higher microbial diversity (Shannon Index) compared to patients with high or absent A. muciniphila levels (118). This represents a clear example of a U-shaped association of a microbial biomarker, in which extremes in one direction or the other are associated with the lowest benefits, or even with negative outcomes.

Before examining interventions beneficial for health and the microbiome, the following example aims to address the unintended effects on the host and microbiome of a high-protein, low-carbohydrate weight-loss diet. In a clinical trial, 80 overweight or obese postmenopausal women followed an 8-week very-low-calorie diet based on a meal replacement shake (121). The shake contained skimmed milk powder, milk protein, sodium caseinate, maltodextrin, canola and sunflower vegetable oils, artificial sweeteners, and was fortified with minerals, vitamins and inulin. The intervention led to weight loss and transient improvements in glucose regulation. Gut microbiota shifted significantly, with increased abundance of Akkermansia, and reduced abundances of plant-metabolizing genera Roseburia, Ruminococcus, Eubacterium, which are known to produce SCFAs (see section 2.1). Levels of SCFAs acetate, butyrate and valerate were significantly reduced (121). To explore the role of the microbiome, human fecal samples from trial participants were transferred into mice. This reproduced weight loss but revealed an unexpected enrichment of pathogenic C. difficile post-intervention, despite similar baseline abundances pre-intervention. This suggested a diminished capacity of the post-diet to restrict pathogen growth (121). These findings underscore how a complex change in diet composition can lead to multifaceted effects on host biomarkers, involving changes in the microbiome and its metabolic output.

In a recent study involving healthy individuals, the consumption of 23 g of walnuts twice daily increased alpha diversity, and the abundances of 13 genera (Table 3): Roseburia, Rothia, Parasutterella, Lachnospiraceae UCG-004, Butyricicoccus, Bilophila, Eubacterium eligens, Lachnospiraceae UCG-001, Gordonibacter, Paraprevotella, Lachnospira, Ruminococcus torques, and Sutterella (142). Notably, the intervention significantly increased urolithin A and iso-urolithin A, which were significantly associated with several taxa including Gordonibacter, Angelakisella, Defluviitaleaceae UCG-011, and Anaerovoracaceae Family XIII (142). These findings suggest that polyphenols in walnuts likely contributed to the observed rise in urolithins (see Section 2.5).

In another study with nuts, 57 g almonds consumed daily for 8 weeks increased alpha-diversity and reduced Bacteroides fragilis compared to an isocaloric cracker snack (143). A separate energy-restricted study found that a nut-enriched breakfast, compared to a nut-free breakfast, significantly reduced ghrelin levels, suggesting a role in appetite regulation among individuals at risk for cardiometabolic disease (144).

Two 8-week randomized controlled trials tested the effects of caloric restriction (CR) with or without the addition of nuts in women with high BMI and increased waist circumference and at least one other cardiometabolic risk factor.

In the first trial, participants consumed 30 g cashews and 15 g Brazil nuts daily. Both CR groups (with and without nuts) showed significant changes in beta-diversity, indicating a shift in community composition. However, only the CR-only group exhibited a reduction in alpha-diversity and a significant increase in the Firmicutes-to-Bacteroidetes ratio, changes that were not observed when nuts were added to the diet (145). A reduction in alpha-diversity points to a less stable microbial community.

Fecal SCFA analysis showed that CR and nuts significantly increased fecal propionic acid, and reduced acetate (145). Lactulose-secretion was used to measure intestinal permeability, which did not differ due to treatment, but lactulose secretion was significantly lower in the CR-nut-group compared to CR at the end of the intervention (145). Abundances that differed after the intervention versus baseline included increased Ruminococcus and UCG-002 in the CR-nut group and increased abundances of Blautia, and Ruminococcaceae genera UCG-005 and UCG-002, and a reduction in Ruminococcus in the CR group. Intestinal permeability correlated with fat-loss, IL-8 and Ruminococcus abundances (145). These findings suggest that the addition of nuts to caloric restriction may preserve community stability and support beneficial microbial taxa.

A second trial tested the effect of 8 g of Brazil nuts daily combined with CR to CR alone. The CR-nut group showed significantly decreased levels of CRP, TNF, IL-1β and IL-8 indicating an anti-inflammatory effect of the nut-enriched intervention (146).

The NU-AGE randomized controlled trial examined the effects of a one-year Mediterranean (Med) diet intervention in elderly subjects from five countries (Italy, United Kingdom, France, Netherlands, Poland) (147). Increased intake of fibers from vegetables, fruits and whole grains, plant proteins from legumes, polyunsaturated fatty acids from fish and vitamins and reduced fat, alcohol, salt and sugar intake consistently modulated microbiome composition across countries.

Greater adherence to the Med diet was associated with improved cognitive function, reduced frailty and lower inflammation. Key microbial responders included SCFA-producing taxa (see section 2.1) such as Faecalibacterium prausnitzii, Roseburia hominis, Eubacterium species (E. rectale, E. eligens, E. xylanophilum), as well as Bacteroides thetaiotaomicron, Prevotella copri and Anaerostipes hadrus. These taxa were negatively associated with inflammatory markers hsCRP and IL-17 and measures associated with increased frailty; and were further negatively associated with lower production of secondary bile acids and the uremic toxin p-cresol (147). Prior studies have similarly shown reductions in p-cresol following increased fiber intake (55).

The CORDIOPREV trial evaluated the long-term effects of a Med diet versus a low-fat diet in patients with established coronary heart disease over a seven-year follow-up (8, 148). The Med diet was superior in reducing the risk of major cardiovascular events (148). The analysis of fecal samples from a subgroup of 106 male participants revealed distinct baseline microbial compositions among individuals with and without metabolic syndrome. Those with metabolic syndrome had lower abundances of Bacteroides, Prevotella, Roseburia, Ruminococcus, and Faecalibacterium, and higher abundances of Streptococcus and Clostridia (8). Interestingly, the intervention only altered microbial composition in those with metabolic syndrome, and not in individuals without metabolic syndrome independent of obesity (8). Most of the compositional differences in participants with metabolic syndrome were reversed after follow-up 2 years later in addition to significantly decreased TG levels in both dietary intervention groups (8). Nutritional intake patterns shifted during the trial: the Med diet group increased their intake of olive oil, nuts and oily fish, and reduced consumption of carbohydrates and saturated fatty acids. The low-fat group increased carbohydrate consumption and reduced fat-intake; both groups reduced their intake of red and processed meats and carbonated drinks. Both groups increased dietary fiber consumption by 2.3 g (Med) and 3.2 g (low-fat) per 1,000 kcal (8, 148).

The DIRECT-PLUS randomized controlled trial integrated multiple anti-aging strategies and included 294 participants with abdominal obesity and dyslipidemia. Participants were randomized into three groups: healthy dietary guidelines, a Med diet, and a polyphenol-rich, low-red/processed meat Green-Med diet. Green-Med and Med diets were calorie-restricted and included daily consumption of 28 g walnuts. Green-Med participants were guided to consume 3–4 cups of green tea and a 500 mL shake composed of Mankai, an aquatic plant rich in protein and polyphenols, which increased the polyphenol content of the intervention by 800 mg gallic acid equivalents (17). Both interventions led to weight loss, improved cardiometabolic markers, body weight, blood pressure and fasting plasma leptin. Further, the interventions resulted in significant compositional changes: Green-Med increased abundances of Prevotella, Bacteroides, Lachnospira and genus DNF00809 from the Eggerthellaceae family and decreased abundances of Dorea, Collinsella and Bifidobacterium (Table 3). Green-Med intervention compositional changes were largely driven by non-core members (17). Interestingly, reduced Bifidobacterium abundances were significantly associated with greater weight loss. In the Med-diet group, Lachnospira, DNF00809 (Eggerthellaceae), Enterorhabdus, Intestimonas and genus UCG-003 from the Erysipelotrichaceae family significantly increased. Changes in metabolic pathways included a stepwise increase in BCAA (branched-chain amino acid) degrading pathways and a decrease in BCAA biosynthesis pathways, and an association analysis revealed that reduced cysteine biosynthesis was linked to, and mediated, weight loss and reduction in CHD risk (17). Positive effects were specifically attributed to the increased polyphenol content in the Green-med group, including the significantly higher reduction in intrahepatic fat compared to the Med diet (149) and higher circulating levels of hippuric acid, as well as Urolithin A (see sections 2.4 and 2.5) that related to greater visceral-adipose tissue reduction, mediated by increased consumption of walnuts and Mankai (86).

The Green-Med diet also impacted biological age, as assessed by DNA methylation age clocks. At the end of the intervention, a ~ 9-month favorable difference between observed and expected methylation age was observed (150). Greater age attenuation was linked to higher intake of Mankai and green tea, which correlated with elevated urinary metabolite levels of hydroxytyrosol, tyrosol, urolithin A, and urolithin C (150).

A personalized nutrition approach was tested in the post-prandial glucose targeting (PPT) trial in prediabetic individuals. The PPT diet was based on a machine learning algorithm that integrated multiple parameters (meal’s nutrients composition, blood data, anthropometrics, lifestyle and gut microbial features), to predict an individual’s postprandial glycemic response, and then provided recommendations for meals (151). The PPT diet was compared to a Mediterranean diet intervention (participants were encouraged to include whole-wheat bread and grains, legumes, fruits, vegetables, olive oil, fish, poultry and low-fat dairy products in their diet) (151). While both dietary interventions enriched Faecalibacterium species, the PPT diet had a more pronounced effect on the gut microbiome and metabolites than the control diet, including increased richness and diversity, and a significant decrease in human cell shedding. Further, the PPT diet increased Flavonifractor plautii, Roseburia hominis, Ruthenibacterium lactatiformans and three strains of Faecalibacterium prausnitzii (151) (Table 3).