Dieta, microbioma intestinale e diabete di tipo 1: dal rischio all'opportunità traslazionale

Introduction

The overall incidence of type 1 diabetes (T1D) has increased considerably in the second half of the 20th century, particularly in children under the age of five. , This rise in incidence, irrespective of genetic predisposition, suggests that early-life exposure to environmental factors, including diet, microbiome, viral infections, mode of birth, antibiotic exposure, and geographical location, may act as potential triggers for T1D onset. Emerging data implicate diet as a modifiable factor in both the risk and progression of T1D. While some factors may trigger or accelerate disease progression, others may confer protection against the development of T1D-related autoantibodies and disease progression. Several dietary compounds, including nicotinamide, zinc, and vitamins C, D, and E, have been investigated for their potential to delay or prevent T1D onset. Additionally, early-life exposure to foreign dietary antigens, such as gluten and milk proteins, has been studied for its impact on β-cell autoimmunity. Notably, dietary factors intricately modulate the composition of the gut microbiota, and their interactions with the host and their byproducts, including secondary bile acids and metabolites, can profoundly impact immune regulation. , Despite the potential impact of dietary factors, their underlying mechanisms in T1D onset and the extent to which these factors affect T1D remain unknown.

The review aims to summarize current evidence on the role of key dietary factors (e.g., breastfeeding, gluten, and cow's milk proteins) and microbiota-associated interventions (probiotics, prebiotics, and fecal material transfer) in shaping the gut microbiome, immune system, and modulating T1D risk and progression. In contrast to previous reviews that primarily catalog associations between diet and T1D, this article emphasizes mechanistic links through the microbiome–immune axis and critically evaluates the translational potential of microbiome-targeted interventions. By integrating evidence from animal models, observational human cohorts, and randomized controlled trials, we highlight both promising leads and key gaps that must be addressed before dietary or microbial interventions can be widely adopted in clinical practice.

Dietary influences on T1D

### The effect of breastfeeding on T1D

#### Impact of breastmilk on gut microbiota and immune function

Breast milk modulates the composition of the gut microbiota in infants, which further impacts immune responses and potentially T1D predisposition and pathogenesis. , Given the global variability in breastfeeding practices, it is important to evaluate the impact of breastmilk consumption on an infant's gut microbiome diversity and its potential association with T1D risk. With its combination of bacterial and bioactive components, human milk aids in developing the infant gut microbiome and the maturation of the immune system while competitively inhibiting the growth of pathogens. Breast milk harbors its unique microbiome comprising viable microorganisms from various genera, including Lactobacillus , Bifidobacteria , Staphylococcus, and Streptococcus in humans , ( ). Notably , breast milk has been shown to increase bacterial diversity within the infant gut microbiome. Microbial diversity is critical for immune system development, and reduced intestinal microbial diversity and abundance have been associated with various immune-mediated disorders, including T1D. Furthermore, SCFAs present in breast milk can stimulate receptors on regulatory T (Treg) cells and modulate bacterial gene expression, both of which play key roles in enhancing intestinal barrier integrity through tight junction expression and promoting anti-inflammatory responses. Multiple determinants, including the mother's ethnicity, weight, stage and course of lactation, and geographical location, are associated with the human milk composition and metabolome.

#### Mechanistic studies of key bioactive components in breast milk

Bioactive molecules, including leukocytes, triglycerides, and caseins, are important compounds in breast milk that strengthen the infant's immune system and may reduce T1D risk. Triglycerides are another important nutritional element within breast milk that combine with lingual and gastric lipases to become monoglycerides with a comprehensive defense against bacterial pathogens in the microbiome and build the innate immune system for breastfed infants. Moreover, readily digested caseins, which account for almost half of the protein content in human milk, particularly glycosylated κ-casein, show anti-microbial activity and have been shown to inhibit Helicobacter pylori infection. A number of studies using nonobese diabetic (NOD) mouse, a polygenic model for T1D, have found that dietary replacement of intact milk proteins with highly hydrolyzed casein (HC) can protect against T1D by reducing the production of reactive oxygen and nitrogen species (RONS), proinflammatory cytokines (IL17, interferon-gamma (IFNγ)), and intestinal T cell activation. , ,

Human milk oligosaccharides (HMOs) have been shown to stimulate the growth of certain Bifidobacterium species that often dominate infant gut microbiota composition ( ). HMOs have complex molecular structures composed of a lactose core elongated by type 1 and type 2 chains, frequently modified by fucose and sialic acid residues. The profiling of bacterial HMO consumption demonstrated the conservation of the type 1 chain in human milk and revealed the mechanisms by which Bifidobacterium secretes glycosidases to digest HMOs. Selective Bifidobacterium digestion of HMOs, despite their lack of nutritional value in human infant, suggests a coevolutionary mechanism at work. The impact of HMOs as a microbiome modulator was comprehensively assessed in NOD mice through a diet supplemented with HMOs. The introduction of HMOs resulted in significant changes in the fecal microbiota composition, including an increase in the relative abundance of Firmicutes and a decrease in Bacteroidetes . This shift led to a higher Firmicutes -to- Bacteroidetes ratio, which has been associated with lower glycemic levels and reduced insulitis, indicating a healthier microbiome. Furthermore, differences in the relative abundances of SCFA producers were observed in the HMO-supplemented group; increases in Akkermansia and Prevotella indicated improved gut integrity. Importantly, the introduction of HMOs also increased the relative abundance of segmented filamentous bacteria, which have been linked to diabetes protection. These findings highlight the potential of HMOs to modulate the microbiome and contribute to diabetes protection in NOD mice, where HMO supplementation led to reduced insulitis and improved gut integrity.

The most abundant immunoglobulin within human breast milk is secretory IgA (sIgA). sIgA shapes both the immune system and the microbiota composition of infants. , IgA is an essential immune component and has been shown to increase in the fecal microbiome of individuals with T1D. This increase in IgA-coated bacteria in individuals with T1D correlates with a decrease in SCFA-producing bacteria. The immunoactive components of milk provide beneficial effects beyond the lifespan of soluble factors in milk. A previous study supports “maternal educational immunity”, a mechanism through which immune cells transferred from an immunized mother to her nonimmunized foster pups instruct the development of the pups' T cell repertoire and lead to long-lasting, pup-derived CD8+ T cells specific to the mother's immunogens. This study demonstrates the crucial role of milk components in directing the T cell development of offspring. The individual components of breast milk indicate remarkable importance in forming the immune system and microbiome of breastfed infants, leading many to hypothesize the protective effects of breastfeeding against T1D incidence.

#### Intervention studies on the effects of breastfeeding and breastfeeding duration in T1D

Although breastfeeding has many benefits, exclusive breastfeeding may lack certain nutrients. For example, depending on the mother's diet, insufficient levels of vitamin D in breast milk during exclusive breastfeeding may lead to vitamin D deficiency in infants. Vitamin D plays an important immunomodulatory role by binding to its receptors on β-cells and immune cells. Vitamin D deficiencies have been linked to the exacerbation of immune diseases, including T1D. A small case-control study utilizing 60 patients with combined endocrine disorders (T1D, T2D, and autoimmune thyroiditis) showed that vitamin D3 supplementation led to improvements in carbohydrate, mineral, and lipid metabolism, which reduced pro-inflammatory Th1 cytokines elevated in T1D and T2D patients and increased anti-inflammatory Th2 cytokines. , Overall, evidence suggests that poor vitamin D levels may aggravate the onset of T1D. However, supplementation with vitamin D during pregnancy or infancy has been associated with a reduced incidence of T1D.

Similarly, antioxidants such as vitamin E are present in the early colostrum of breastfeeding mothers. During early- to mid-lactation, the content of vitamins A and E in breast milk declines, indicating worsening nutritional status in women already lacking essential nutrients. Vitamin E supplementation in both T1D and T2D patients has been shown to delay development and slow diabetic complications. Therefore, one could suggest that breastfeeding, along with the supplementation of essential nutrients, may be a key factor in T1D prevention. In addition, breastfeeding duration significantly affects the infant microbiome. A study from the New Hampshire Birth Cohort showed that breastfeeding for more than 6 mo mitigated the long-term impact of birth mode on the infant gut microbial composition. Cesarean-delivered (CD) infants who were breastfed for more than 6 months had a gut microbiota that was more similar to that of vaginally delivered infants than to that of CD infants who were breastfed for less than 6 months.

Several studies have attempted to demonstrate this relationship, with varying conclusions based on the length and exclusivity of breastfeeding. The first evidence of dietary manipulation affecting β-cell autoimmunity was observed in the BABYDIAB study, a long-term study in Germany that followed newborn children of parents with T1D. This study did not find any impact of breastfeeding duration on the risk of developing islet antibodies; however, the type of formula administered to infants was not accounted for as a contributing factor. This hypothesis was further explored in the Finnish Type I Diabetes Prediction and Prevention (DIPP) Study, which enrolled 2949 infants. This study identified that infants with exclusive breastfeeding for at least 4 months had a lower risk of seroconversion to positivity of insulinoma-associated protein 2 autoantibodies (IA-2A) in comparison to infants who breastfed for less than 2 months. This study also identified that early introduction to cow milk-based formula increased the risk of developing islet autoantibodies. Another pilot study later identified that individuals who received casein hydrolysate had significantly reduced hazard ratios for seroconversion to islet cell autoantibodies (ICA) positivity and the presence of at least one autoantibody. Moreover, a meta-analysis of 43 breastfeeding studies from 1996 to 2012 observed a reduction in diabetes risk if infants were exclusively breastfeeding for >2 weeks. This correlation is attenuated after three months of exclusive breastfeeding, and there is no clear association when breastfeeding is not exclusive.

Similarly, in a large Scandinavian cohort (Norwegian Mother and Child Cohort Study [MoBa] and the Danish National Birth Cohort [DNBC]), which followed infants for a mean of 10.2 y identified that infants who were not breastfed at all had a two-fold increased risk of T1D development compared to those who were breastfed for any duration during the first 12 months of life. Notably, this population-based study found no significant difference in risk by breastfeeding duration, even after analyses controlled for potential confounders, including family risk of T1D and ethnicity. A similar longitudinal study performed on a cohort of children in the United States, Diabetes Autoimmunity Study in the Young (DAISY), followed an early-life diet and observed the effect of the introduction of solid food and duration of breastfeeding on newborns with high genetic risk (human leukocyte antigen (HLA)-DR3/DR4-susceptible), and those with first-degree relatives. This study revealed that children who were still breastfed during the introduction of wheat/barley had a hazard ratio of 0.47 in T1D cohorts, indicating a potential protective effect against the development of T1D. Continuing breastfeeding while introducing wheat/barley reduced the chance of developing T1D by 34%, highlighting the importance of extended breastfeeding through solid food introduction after six months ( , represents all interventions study outcomes). The DIPP nutrition study, involving 5915 genetic high-risk newborns, categorized breastfeeding duration into quartiles and found a correlation between shorter breastfeeding duration and islet autoimmunity but not with overt T1D. This large cohort study also identified an association between early introduction (3–4 months of age) of root vegetables or solid foods and islet autoimmunity in children with high-risk HLA but not in children with moderate-risk HLA. Additionally, the study identified a weak link between INS and PTPN22 loci (non-HLA T1D risk genes) and the age of introduction of meat and fish, respectively, with islet autoimmunity. A large multinational longitudinal cohort study, the Environmental Determinants of Diabetes in the Young (TEDDY), found no significant association between longer breastfeeding duration and a reduced risk of islet autoimmunity or T1D in genetically susceptible high-risk children. , However, these associations differ by demographics, mode of delivery, length of gestation, parental age and education. , While large prospective longitudinal cohorts such as TEDDY have provided invaluable insights, important gaps remain. In particular, studies are needed that follow children from birth through the onset of T1D, including those born to parents both with and without T1D, and that incorporate randomized infant feeding interventions. Randomized infant feeding interventions could help mitigate challenges such as recall bias, which can otherwise lead to over- or underestimation of breastfeeding duration. For instance, a meta-analysis of 17 case-control studies on infant diet and T1D highlighted significant challenges, with inconclusive results often attributed to low participation rates and control selection bias. While concerns about the feasibility of such studies are valid, a previous feeding trial found that 69.9% of eligible parents consented to participate, indicating promising potential for similar research efforts. Nonetheless, large prospective studies such as TEDDY and DIPP have shown only modest or inconsistent associations between breastfeeding duration and T1D risk, suggesting that while breastfeeding confers many health benefits, its role in preventing islet autoimmunity or T1D progression remains uncertain.

### Gluten-containing dietary components

Gluten and gluten-based foods such as wheat, barley, and rye have been identified as potential diabetogenic factors capable of inducing T1D pathogenesis. , Constituents of gluten, particularly α-gliadin, were shown to increase gut permeability, induce the release of cytokines, and act as a general cytotoxin.

Findings from both animal and human studies exploring the correlation between early-life gluten exposure and T1D indicate an elevated risk of developing T1D with early exposure to gluten. In vivo studies have observed that both wheat-enriched and cereal-based diets increase the risk of developing islet autoimmunity in Biobreeding (BB) rats. , One study in the NOD mouse model has conversely shown that a gluten-enriched diet can also prevent T1D onset. NOD mice on a gluten-free diet (GFD) had a substantially lower incidence of diabetes (15%) compared to those on a standard chow diet (64%), and the onset of diabetes was delayed. Although no observable differences were found in the number of immune cells in the small intestine, such as CD3 + , TCRγδ + , IgA + , and IgM + cells, this indicates that the protective effect may not directly involve changes in intestinal immune cell populations. GFD also increases 10%–15% of Treg cells in Peyer patches and macrophages and CD103+ dendritic cells in the spleen of NOD mice. Similarly, another study found that mice fed a cereal-based diet (NTP-2000) exhibited increased mRNA expression of nitric oxide synthase and proinflammatory markers such as interferon γ (IFN-γ) and tumor necrosis factor α (TNF-α) compared to those on a diet based on soy protein isolate.

#### Mechanistic studies linking GFD and the microbiome

GFD decreased the number of aerobically and microaerophilically cultivated bacteria in their intestines and reduced the development of T1D when administered to 3-week-old NOD mice compared to standard-diet-fed mice. Furthermore, the GFD reduced the incidence of hyperglycemia. Most of these changes were linked to alterations in the gut microbiome, where Akkermansia species were more abundant in GFD-fed mice. In contrast, Bifidobacterium , Tannerella, and Barnesiella species increased in NOD mice fed with a gluten-containing diet. Notably, reintroducing gluten into the diet reversed these microbial changes and restored the higher incidence of diabetes. A similar effect was observed when a GFD was provided to pregnant and lactating NOD mice, which significantly reduced diabetes and insulitis in their offspring, even though the pups were weaned onto a standard chow diet. Again, this protection was associated with a marked difference in the gut microbiota, including increased Akkermansia, Proteobacteria , and TM7 species. Additionally, offspring from the GFD group showed higher levels of gut-primed α4β7 T cells in the pancreas and increased expression of anti-inflammatory markers and tight junction-related genes in the gut, suggesting that the diet modulates immune responses and reduces gut inflammation and enhances immune tolerance.

#### Intervention studies of GFD in T1D

Although the effect of the GFD in NOD mice is well documented and robust, findings from human studies have been more variable and inconsistent. Two studies focused on patients with both celiac disease and T1D reported a significant improvement in weight and serum ferritin in subjects on a GFD; however, no improvement in HbA1c was observed. , Nonetheless, other studies have not observed such effects. , Three studies conducted in high-risk individuals found that administration of a GFD did not markedly reduce autoantibody titers or diabetes incidence. However, one of these studies reported improved β-cell function following six months of GFD administration. The complete elimination of gluten from the diet has been considered as a potential intervention, since studies of GFD in older children diagnosed with T1D demonstrated improved diabetic metrics, including prolonged remission periods and reduced HbA1c levels. , GFD also improved glucose tolerance and insulin sensitivity in older children (median age of 16 y) after six months of administration. However, no improvements were observed in islet autoantibody titers. As all these studies carried out on a diet-based mechanism offer insight into the effect of GFD on T1D patients after disease development, mechanistic and large cohort longitudinal studies exploring the effect of GFD on inflammation during the autoantibody stage are still lacking and present a gap to explore. Together, these studies suggest that gluten restriction may modulate glycemic control and β-cell function in some individuals with established T1D, but the effects on islet autoimmunity and progression from preclinical stages remain modest and inconsistent across cohorts.

#### Timing of gluten introduction in T1D risk

Despite differences across models, one factor shown to ubiquitously impact disease development is the timing and mode of gluten introduction. Human studies have found that early exposure (<3 months of age) to gluten increases the risk of islet autoimmunity and that this risk is mitigated if gluten is introduced while the child continues to breastfeed. , , In agreement with this, BB rats had an increased risk of developing T1D if gluten was introduced postweaning compared to rats introduced to gluten during weaning (i.e., while still nursing).

Mothers' diet during pregnancy also influences the child's risk of developing T1D. Littermates of NOD dams fed a GFD show a reduced incidence of T1D. , One such study observed a 7-fold decrease in T1D incidence in NOD pups ( n = 20–22, 62.5%–8.3%), with reduced islet infiltration and lower Th17 expression in the intestine. Moreover, this antidiabetic effect of GFD has been identified across multiple generations. In a cohort of 6352 Danish women, higher maternal gluten intake during pregnancy was associated with an increased risk of T1D in their offspring. Women consuming more than 20 g/d of gluten during pregnancy had twice the risk of T1D onset in their children. Two large prospective cohort studies have also examined the relationship between maternal and early-life gluten intake and the risk of T1D in offspring. In the Finnish DIPP Study, which followed 4943 genetically susceptible children, maternal intake of gluten, cereals, or dietary fiber during pregnancy and lactation showed no association with islet autoimmunity or T1D risk, except for a possible higher risk linked to increased barley intake during lactation. Similarly, in the Norwegian Mother and Child Cohort Study (MoBa) involving 86,306 participants, maternal gluten intake during pregnancy was not associated with T1D in children. However, higher gluten consumption by the child at 18 months of age was associated with an increased risk of developing T1D. Together, these findings suggest that maternal gluten intake during pregnancy or lactation does not influence T1D risk, whereas higher gluten exposure in early childhood may be one of several context-dependent factors associated with increased risk in some cohorts but not others. ,

Cereal-based diets induce a proinflammatory response with increased levels of Th1 cells, IFN-γ, TNF-α, and nitric oxide (NO) synthase in NOD mice. , Similar effects of a cereal-based diet have already been discussed earlier. , Increased inflammation and an altered immune response are hallmarks of T1D pathogenesis and can be driven by changes in the gut microbiota. On the other hand, NOD mice fed a GFD had increased levels of anti-inflammatory forkhead box P3 (FOXP3+) regulatory T cells (Tregs) and M2 macrophages, along with decreased levels of proinflammatory cytokines. Mucosal integrin α4β7 + T cells were also increased in the pancreas. This integrin is a marker for T cells produced in the gut, suggesting that T cells present in the pancreas are likely primed by the gut microbiota. Additionally, in GFD-fed NOD mice, single-cell transcriptomic analysis of immune cells shows differentiation of Tregs, activated γδ T cells, and NKT cells along with IL-2-, IL-17-, and IL-15-induced genes, indicating that GFD affects broad immune cell populations.

Overall, gluten exposure has been shown to influence inflammation and T1D outcomes in animal models, though findings are inconsistent and sometimes contradictory. , , Mechanistic studies indicate that these effects are mediated through alterations in the gut microbiota composition and immune tolerance. In humans, epidemiological data demonstrate that elevated maternal gluten intake during pregnancy is associated with a 2-fold increased T1D risk in offspring. Notably, the child's own gluten intake at 18 months was associated with increased T1D risk in the Norwegian cohort. Early gluten introduction also increases islet autoimmunity risk; these effects are attenuated when gluten is introduced during continued breastfeeding. However, the lack of strong effect in human studies may be due to other cofounding factors, such as genetic heterogeneity, differences in baseline diet, geographic and environmental variation, or other modifying exposures that could mask a true but context-dependent effect of GFD.

### Cows' milk and cows' milk proteins

Many children mount an immune response against cows' milk early in life owing to the significant compositional differences between bovine and human milk proteins. The major allergens present in cow's milk include β-lactoglobulin, α-lactalbumin and caseins. While these are considered the most immunogenic compounds, other proteins, including bovine serum albumin (BSA) and lactoferrin, have also been identified as potential immunogenic candidates. Cow's milk products have also been have long been hypothesized as potential diabetogenic factors. Animal and human studies alike have found associations between T1D incidence and the consumption of cow's milk during both early weaning and later developmental stages.

A retrospective meta-analysis in children younger than 15 y across 40 countries demonstrated a positive correlation between a region's intake of animal-based energy sources (meat and dairy products) and T1D incidence; conversely, an inverse correlation was observed between plant-based sources and T1D incidence. , Likewise, while total protein intake was not correlated with T1D incidence, the specific consumption of cow's milk-derived β-casein A1 variant was correlated. This relationship was even more pronounced when the combination of β-casein A1 and B variants was considered.

Since the early introduction of cow's milk during weaning commonly occurs in conjunction with a decrease, or sometimes complete cessation of breastfeeding, it is difficult to determine whether the increased incidence of T1D in infants is due to the lack of breast milk or the early introduction of cow’s milk into the diet. Many human studies have shown the protective effects of breastfeeding, , yet very few studies have looked at the introduction of cow's milk and breastfeeding cessation as independent variables. Some studies have indicated that overall milk consumption is positively associated with T1D incidence rates. High milk consumption ( ≥ 3 glasses/d) in childhood was associated with more frequent T1D-related autoantibodies in the siblings of children diagnosed with T1D. Moreover, the administration of cow's milk-based infant formula before 3 months of age was strongly associated with an increased T1D risk compared to exclusive breastfeeding. In contrast, an intervention trial found no correlation between cow's milk-based formula and T1D incidence by the age of 8, as well as no observable correlation between the duration of breastfeeding and T1D incidence.

#### Mechanistic studies of key components of cow's milk linked to T1D development

#### β-Casein A1 variant and β-cell autoimmunity.

Human and cow's milk differ in several aspects; one key difference is protein content. Cow's milk contains a significantly higher protein concentration, with about 80% of its protein content consisting of caseins, compared to only about 30% in human milk. Among these caseins, β-casein exists in two primary variants, A1 and A2. While both A1- and A2-β-casein have been studied in animal models, only A2 has been consistently associated with protective effects against T1D development in animal models, , , but current human evidence is weak and inconclusive.

The A1 variant of β-casein has been strongly associated with T1D incidence. , Upon digestion, A1β-casein releases a bioactive peptide known as β-casomorphin-7 (BCM), which exhibits opioid-like activity and has been implicated in immunosuppressive and gut-modulating effects. , BCM-7 interacts with opioid receptors and has been shown to exhibit anti-proliferative effects, suppress lymphocyte proliferation, influence gut motility, and alter mucus secretion, as well as the expression of gastrin and somatostatin genes. However, some postulate that the observed reactivity to cow's milk proteins in T1D patients may not be causal, but rather reflect heightened gut permeability or a generalized immune response to the first dietary antigens encountered in early life. , Supporting the immune reactivity hypothesis, a German study found elevated A1-casein-specific antibodies in T1D patients and their siblings, compared to parents and healthy controls, suggesting a defect in immune tolerance to cow's milk proteins in T1D.

#### β-Lactoglobulin and BSA in β-cell autoimmunity.

Numerous case-control studies have been conducted to characterize the immunogenic properties of cows' milk and their association with β-cell autoimmunity. The DIPP study identified the overall consumption of cows' milk products as one of the few dietary factors associated with β-cell autoimmunity. Multiple studies have reported elevated levels of cows' milk-specific IgG and IgA antibodies in individuals with early-onset T1D. Bovine β-lactoglobulin, a major allergen and one of the most abundant proteins found in the soluble fraction of cow's milk, has been implicated in this immune response. Among studies reporting elevated cows' milk-specific antibodies in T1D patients, many also found elevated anti-β-lactoglobulin IgA and/or IgG antibodies. Interestingly, anti-β-lactoglobulin IgA antibody levels show an inverse correlation with breastfeeding duration, whereas anti-β-lactoglobulin IgG antibodies demonstrate a positive association with formula feeding. Although elevated antibody levels against multiple cow's milk proteins, including β-lactoglobulin, were initially observed, only β-lactoglobulin-specific IgA antibodies retained statistical significance after adjustment for formula feeding before 4 months of age ( ). T1D patients also have elevated IgA and IgG antibodies against BSA but not any other cows' milk protein, suggesting a specific immune response targeting bovine albumin. Moreover, cross-reactivity between anti-BSA antibodies and β-cell MHC class II surface proteins has been identified in BB rats, proposing a potential mechanism by which these antibodies may contribute to the development of anti-islet autoimmunity.

#### Cows' milk proteins and the microbiome

Few studies have explored the relationship between cows' milk consumption and gut microbiome alterations. However, a recent study in male Sprague-Dawley rats compared the impact of unfermented and fermented skim milk (yogurt) products from cows and sheep on the gut microbiota composition. They demonstrated that cows' milk consumption induced a distinct and highly variable taxonomic profile characterized by an increased abundance of Collinsella aerofaciens and enrichment of Firmicutes, Collinsella , and Lactobacillus taxa. An HC diet has also significantly impacted the gut microbiota composition. In HC-fed mice, the total bacterial load was markedly reduced, along with decreased production of butyric and lactic acid. Moreover, the Bacteroidetes:Firmicutes ratio increased, driven by an expansion of Clostridial clusters XIVa and the IV group, the latter of which includes beneficial probiotic species such as Faecalibacterium prausnitzii. , Overall, breastfeeding and the delayed introduction of gluten and cow's milk may reduce T1D risk, whereas high cow's milk intake during early childhood could be a potential risk factor.

#### Intervention studies on the effects of infant formula on the onset of T1D

Human studies attempting to associate cow's milk proteins with the gut microbiota composition and T1D were derived from two major longitudinal studies. These dietary studies include (i) the Trial to Reduce IDDM in the Genetically at Risk (TRIGR) and (ii) the Finnish Dietary Intervention Trial for the Prevention of T1D (FINDIA). While these studies do not specifically address the association between the gut microbiome and T1D incidence, a subsequent cross-sectional study derived from these studies was conducted to investigate this effect. The TRIGR study aimed to test whether supplementing breast milk with highly hydrolyzed milk formula could delay or prevent T1D onset. Subjects who received either hydrolyzed milk formula or conventional formula showed no significant difference in T1D incidence, indicating that, despite mechanistic plausibility, the hydrolyzed formula did not prevent T1D in practice.

In contrast, the FINDIA study suggested that bovine insulin-free formula modestly reduced β-cell autoantibody risk, though this effect was limited and has not been consistently replicated. The FINDIA study randomized newborns to receive either a bovine insulin-free cows' milk formula, a whey-based hydrolyzed formula, or a whey-based FINDIA formula from which bovine insulin was removed. Infants who received bovine insulin-free formula during the first six months had a reduced risk of developing β-cell autoantibodies by the age of three. Microbiome analysis shows an increased abundance of Bacteroides and a decreased abundance of Bifidobacterium in the seroconverted group (individuals who developed islet autoantibodies). Overall, weaning to an insulin-free formula reduced the cumulative incidence of autoantibodies in children at genetic risk of T1D. A cross-sectional analysis of TRIGR and FINDIA participants observed that autoantibody-positive (AAb+) children had a higher relative abundance of Bacteroidetes (genus Bacteroides ). Moreover, AAb+ subjects showed low abundances of Bifidobacterium adolescentis and Bifidobacterium pseudocatenulatum (<12% combined). A low abundance of Bifidobacteria and butyrate-producing species could compromise intestinal epithelial barrier function and induce inflammation, whereas the role of the Bacteroides genus in the development of T1D is insufficiently understood. Overall, these studies suggest that early-life exposure to insulin-free or highly hydrolyzed formula milk may influence the gut microbiota composition, potentially reducing β-cell autoimmunity risk in genetically predisposed children by promoting microbial communities that are associated with a lower T1D risk. However, it is important to note that the T1D risk remains suggestive but unproven.

### Red meat

Red meat is rich in heme iron, saturated fats, and advanced glycation end products (AGEs), all of which contribute to pro-inflammatory processes. Epidemiological studies indicate that animal-based foods are positively associated with T1D incidence, while plant-based foods show an inverse relation. , Red meat consumption has also been linked to autoimmune responses and the progression of various chronic diseases. In a global supply analysis of milk and meat from 1983–2000, increased meat supply correlated with rising T1D incidence. Similarly, a Sardinian case-control study in 298 children (ages 0–15), 145 children who developed T1D showed a dose-response relationship between meat consumption and T1D. Moreover, maternal consumption of both processed and red meat during the lactation period was associated with higher risks of both preclinical and clinical T1D of the offspring. , Excessive red meat intake has also been implicated in colorectal cancer risk, potentially due to heme, N-nitroso compounds, polycyclic aromatic hydrocarbons, and heterocyclic amines found during cooking. Red meat consumption also alters the gut microbiome composition. In rats, red meat consumption decreased the abundance of both Roseburia and Lactobacillus , two genera important for intestinal homeostasis and epithelial barrier integrity while increasing, Alloprevotella , a genus linked to proinflammatory responses ( ). Mechanistically, red meat has been linked to increased production of heme-derived reactive oxygen species, generation of proinflammatory N-nitroso compounds, and shifts in bile acid metabolism that may alter mucosal immunity. In addition, microbial metabolism of carnitine and choline in red meat can generate trimethylamine N-oxide (TMAO), a metabolite associated with systemic inflammation and insulin resistance. These pathways provide possible biological mechanisms by which red meat intake could influence T1D risk beyond epidemiological correlations. Although these studies reveal associations, further longitudinal and mechanistic research is needed to establish direct causal relationships between red meat-induced microbiome alterations and T1D development. At present, the current evidence is largely observational, and causality remains unproven.

### Fatty acids

Fatty acids are fundamental components of lipids and are essential substrates for energy production and cellular metabolism. Levels of adiponectin, an adipokine that plays a critical role in maintaining insulin sensitivity and exerting anti-inflammatory functions, are elevated in individuals with T1D. Although adiponectin is generally protective, these elevated levels in T1D are independently associated with all-cause and cardiovascular mortality, even in patients with normoalbuminuria, indicating a specific link to vascular injury. Supporting this finding, a previous study observed that dietary factors such as total lipids, saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs), and cholesterol intake were inversely correlated with adiponectin levels in T1D patients.

Several studies have examined the relationship between specific fatty acids and T1D risk factors.

The relationship between fatty acids and islet autoimmunity has also been extensively studied. A case-control Finnish Type 1 Diabetes Prediction and Prevention Study ( n = 7782) found that higher serum concentrations of pentadecanoic, palmitic, palmitoleic, and docosahexaenoic acids were associated with a decreased risk of islet autoimmunity, whereas higher ratios of arachidonic:docosahexaenoic acid and n-6:n-3 fatty acids were associated with increased T1D risk. The DIPP cohort of genetically susceptible children identified that higher intake of MUFAs, arachidonic acid, total n-3 fatty acids, and long-chain n-3 fatty acids was associated with reduced risk of developing islet autoimmunity. Moreover, higher total fat and SFAs intake correlated with a lower T1D risk. In contrast, the TRIGR cohort found that increased levels of pentadecanoic, heptadecanoic, stearic, and conjugated linoleic acids were linked to an increased islet autoimmunity risk after adjusting for age, HLA genotype, and maternal T1D.

Recent research has focused on the gut microbiota-fatty acid axis. In the streptozotocin (STZ)-induced T1D mouse model, the administration of n3 alpha-linolenic acid (ALA) and n6 linoleic acid (LA) improved blood glucose and lipid profiles, modulated the gut microbiota composition, and reduced inflammation. ALA administration increased the abundance of acetate- and butyrate-producing bacteria, such as Ruminococcaceae , whereas LA administration enriched taxa, including Alloprevotella, Prevotellaceae , and other Ruminococcaceae species. Both LA and ALA increased SCFA production, particularly acetate and butyrate. Notably, this study also identified reduced levels of lipoxin A4 (LXA4), a bioactive lipid derived from LA, in human T1D patients with diabetic retinopathy. In NOD mice, an anti-inflammatory diet enriched with soluble fiber (inulin) and omega-3 polyunsaturated fatty acids (PUFAs) increased colonic mucus layer thickness and upregulated the mRNA levels of mucin genes and the tight junction protein, claudin1. This diet also downregulated pro-inflammatory cytokine (IL-1β, IL-23, and IL-17) transcript expression while expanding regulatory T cells (FoxP3 + Treg and IL-10 + Tr1 cells) and promoting mucus-modulating bacteria such as Akkermansia muciniphila and Akkermansia glycaniphila.

A cohort of 2939 mother‒child pairs from a prospective T1D prediction and prevention study found that maternal use of vegetable oils was associated with an increased risk of preclinical T1D. Overall, these studies highlight the complex relationships between various fatty acids and T1D risk, progression, and complications, with particular emphasis on their effects on islet autoimmunity and the gut microbiota. Future research should focus on elucidating microbiome modulation and immune regulation following fatty acid supplementation in T1D patients to develop targeted therapeutic strategies.

### High glucose

High glucose levels have been linked with β-cell toxicity and stress, and numerous studies have investigated the relationship between sugar intake and T1D development. Studies in NOD mice demonstrated that elevated glucose levels induce β-cell endoplasmic reticulum (ER) stress, inflammatory cytokine release, and the upregulation of β-cell autoantigens, leading to an acceleration of the autoimmune response and increased T1D incidence. These findings align with human data, and research in a Swedish cohort confirmed that dietary disaccharides and sucrose are associated with T1D risk.

The timing of sugar exposure also matters. Analysis from the SEARCH for Diabetes in Youth study found that subjects with high-risk HLA genotypes were diagnosed approximately five months earlier if exposed to fruit juice during the first year of life. Additionally, weekly consumption of sugar-sweetened beverages (SSBs) and eggs was identified as a significant T1D risk factor.

T1D progression has also been linked to sugar intake patterns. Total sugar intake is associated with T1D progression in high-HLA-risk subjects and islet-autoantibody (IA)-positive subjects. Analysis of the DAISY cohort revealed a correlation where a higher glycemic index (GI) at IA development was associated with faster progression to clinical T1D, though causality remains unproven.

In a rat model of T1D, varying concentrations of glucose-sweetened beverages significantly altered the gut microbiome independent of body weight, adiposity, or caloric intake. In addition, sugar consumption reduced Prevotella and Lachnospiraceae incertae sedis , while increasing Bacteroides , Alistipes , Lactobacillus , Clostridium sensu stricto , Bifidobacteriaceae , and Parasutterella , which were elevated by effects consistent across different fructose-to-glucose ratios ( ). Overall, these studies demonstrate the potential detrimental effects of high sugar consumption on T1D development and progression. While animal studies show distinct microbiome shifts following sugar intake, future studies must determine whether these microbiome changes are causal in T1D pathogenesis or secondary, particularly in human cohorts. Both direct effects of glucose on β-cell function and immune modulation, as well as indirect effects via obesity, insulin resistance, and microbiome alterations, are supported by in vivo , in vitro , and human studies. However, their relative contributions to the initiation and progression of β-cell autoimmunity in humans remain to be clarified.

Microbiome intervention for T1D

### Probiotics

#### Mechanistic studies on probiotic modulation of the immune system in type 1 diabetes

A probiotic bacterium is defined as a nonpathogenic microorganism that is naturally present within the host's commensal microbiota and is believed to confer health benefits to the host. Probiotics, most commonly consisting of species from the genus Lactobacillus and Bifidobacterium , are found in yogurt and fermented food products and are frequently taken as oral supplements. Short-chain carbohydrates that are indigestible by humans serve as a primary energy source for the colonic microbiota, which ferment them to produce SCFAs. Prophylactic use of prebiotics and probiotics has long been thought to provide a range of health benefits by improving intestinal/digestive health, enhancing the immune response, reducing cholesterol levels, and potentially protecting the host against the development of autoimmune and allergy-related conditions when administered during early life. Probiotics have also been shown to be a viable adjunctive therapy in the treatment of infectious and chronic diseases such as infectious diarrheal diseases, inflammatory bowel disease, autoimmune diseases, including rheumatoid arthritis, multiple sclerosis, and T1D.

Perhaps the most widely studied probiotic is the commercialized formulation VSL#3, consisting of eight bacterial strains: three strains of Bifidobacterium, four strains of Lactobacillus, and Streptococcus salivarius subsp. thermophilus. To date, two studies have been conducted in NOD mice to determine the effects of VSL#3 on T1D onset and incidence. T1D onset was delayed, and the overall incidence of T1D was decreased in female NOD mice given VSL#3 at weaning (4 weeks of age) compared with controls. Severe insulitis and β-cell damage were also decreased, and this reduction was concurrently observed with an increase in the anti-inflammatory mediator, IL-10, in Peyer's patches and the spleen, along with increased IL-10 expression in the islet-associated mononuclear cells in the pancreas. However, this effect was not observed in female NOD mice given VSL#3 starting at 10 weeks of age, suggesting that the timing of probiotic administration is crucial in the prevention of T1D. VSL#3 has also shown promise as a preventative measure against various gastrointestinal diseases, but the evidence supporting its use as a treatment for T1D or its complications remains limited.

Similarly, another study examined the protective effects of VSL#3 against T1D onset, administered as either a stand-alone treatment or in combination with retinoic acid (RA), a metabolite of vitamin A that is important for immune tolerance and has been shown to prevent T1D in mice. In agreement with the previous study explained above, this study found that NOD mice treated with VSL#3, alone or in combination with RA, had a decreased rate of T1D and a reduced degree of insulitis compared with their placebo counterparts. Microbiome analysis revealed an increased relative abundance of Clostridia species known to enhance immune tolerance and reduce inflammation. The relative abundance of species from the Bacteroidaceae family varied by species/strain. Strain S24-7 was significantly decreased in VSL#3-treated mice, while other strains, including Rikenellaceae strains, were enriched. Moreover, VSL#3 was observed to induce a protolerogenic immune microenvironment by increasing CD103 + dendritic cells (DC), thereby decreasing the effector T cell (Teff)/Treg ratio, which results in reduced expression of proinflammatory IL-1β. In a double-blinded randomized trial in recent-onset T1D children (aged 2–12 y), the introduction of Visbiome probiotics (manufactured by Danisco-Dupont) improved fasting C-peptide levels, HbA1c, and the bolus insulin dose compared with those in the placebo group.

A more recent study observed similar improvements in NOD mice when treated with the Immune Regulation and Tolerance 5 (IRT5) probiotic, which is composed of Lactobacillus acidophilus, Bifidobacterium bifidum, Lactobacillus reuteri, Lactobacillus casei, and Streptococcus thermophilus . Compared with placebo-treated controls, NOD mice that were given IRT5 had a reduced incidence of diabetes, lower insulitis scores, increased β-cell mass, and reduced insulin autoantibody (IAA) levels. While the total abundance of Tregs remained unchanged, the proportion of Tregs that express the gut-homing receptor (CCR9 + ) increased in both the pancreatic lymph nodes and the lamina propria of the small intestine of IRT5-treated mice.

#### Intervention studies on the effects of probiotics in T1D

In the Environmental Determinants of Diabetes in the Young (TEDDY) study, 7473 genetically at-risk children were assessed to examine the association between early probiotic exposure and the development of islet autoimmunity. The study found that formula and/or dietary supplements containing probiotics administered between birth and 27 d of age were associated with a significantly lower risk of IA compared to children who either never received probiotics or received them after 27 d of life. Notably, among children carrying the high-risk HLA genotype DR3/4, early probiotic exposure was associated with a 60% reduction in IA risk relative to other genotypes. While the exact composition and abundance of bacterial species in probiotic supplements were not reported, most formulations were known to include Lactobacillus and Bifidobacterium species. Additionally, in a double-blind placebo-controlled clinical trial, the administration of a probiotic containing four strains Lactobacillus GG , Lactobacillus rhamnosus , Bifidobacterium breve Bb99 , and Propionibacterium freudenreichii ssp . shermanii, was not associated with the occurrence of T1D or β-cell autoimmunity up to age 5. Because large-scale prospective studies, including TEDDY, have reported only partial or subgroup-specific associations and randomized trials have not demonstrated strong protective effects, probiotics should currently be regarded as exploratory or adjunctive interventions in T1D, rather than established preventive or therapeutic strategies.

Studies have also explored the therapeutic potential of engineered probiotics in managing T1D in both mice and humans. Genetically modified Lactococcus lactis designed to secrete immunomodulatory molecules such as interleukin-10 (IL-10) and proinsulin has demonstrated efficacy in delaying diabetes onset in NOD mice by inducing antigen-specific tolerance and promoting Treg cells. Additionally, L. lactis engineered to express IL-4 and IL-10 was shown to reduce insulitis in NOD and STZ-induced diabetic mice and protect from T1D. In humans, combination therapy using anti-CD3 monoclonal antibody (teplizumab) and engineered L. lactis secreting proinsulin and IL-10 has been investigated for T1D treatment. Engineered L. lactis was sufficient to improve metabolic parameters, including C-peptide levels and reduced insulin requirements ( ). Both monotherapy and combination therapy reduced proinsulin-specific CD4⁺ and CD8⁺ T cells, suggesting the induction of immune tolerance.

### Prebiotics

#### Mechanistic and intervention studies on the effects of prebiotics in T1D

Prebiotics have also been investigated as a potential therapeutic approach for T1D. Combining prebiotics such as polydextrose with sitagliptin, a dipeptidyl peptidase-4 (DPP-4) inhibitor used as an antidiabetic drug to maintain GLP-1 levels, further improved insulin sensitivity and glycemic control in STZ-treated C57BL/6J mice. Similarly, the administration of inulin, another prebiotic, reduces diabetes incidence in the STZ model of T1D. Inulin-administered mice showed enhanced mucus production, SCFA, and the abundance of beneficial microbes, including Akkermansia muciniphila , Clostridium cluster IV , and Bifidobacterium . Moreover, inulin administration also increased CCR4-expressing Foxp3 + Tregs cells in pancreatic islets.

β-Glucans are a group of complex indigestible polysaccharides that have been implicated in the modulation of immune function and have therefore been explored for their prebiotic and immunomodulatory properties in the context of T1D. Purified forms of β-glucans, specifically yeast β-glucans (YBGs), have been previously shown to activate innate immune responses. Consequently, YBGs oral administration is also examined to temper autoimmunity in T1D. Indeed, low-dose oral administration of YBGs reduced T1D incidence and delayed the onset of hyperglycemia in prediabetic NOD mice, highlighting their potential as a prebiotic intervention in T1D. Finally, oral treatment with YBGs was found to result in an increased abundance of Bacteriodetes within the microbiome as well as increased levels of Foxp3 + Tregs, which supports its role in modulating immune homeostasis.

A single-arm pilot and feasibility trial investigated the effects of a 6-week administration of high-amylose maize–resistant starch modified with acetate and butyrate (HAMS-AB) prebiotics in individuals with long-standing T1D. This study identified that patients receiving HAMS-AB have better glycemic control and an increase in stool and plasma SCFA levels, including acetate, propionate, and butyrate. Moreover, patients who received HAMS-AB were also identified to have a shift in gut microbiome composition, such as Bifidobacterium longum , Bifidobacterium adolescentis , and vitamin B7, which was correlated with lower HbA1c, basal insulin requirements, and a more regulatory B and T cells phenotype. Furthermore, in another Australia-New Zealand Trial, HAMS-AB, utilized as SCFA-yielding biotherapy, delayed diabetes in NOD mice and remodeled the gut proteome and immune homeostasis in the gut of T1D patients. Moreover, fecal microbiota transfer from patients who responded to this therapy also improved diabetes incidence in the Germ-free NOD mice model.

In addition, Ismael H. et al. showed in a phase Ia clinical trial that daily consumption of acetylated and butylated high-amylose maize starch (HAMS-AB) was safe and well tolerated in youth with recently diagnosed T1D. HAMS-AB favorably altered the gut microbiome by increasing Bifidobacterium longum and Parabacteroides distasonis , elevated microbial metabolites such as hippurate and tryptophan, which are linked to improved glycemia and reduced inflammation, and downregulated the activation of mucosal-associated invariant T (MAIT) cells, suggesting an anti-inflammatory, immune-modulating effect.

A controlled trial involving children with T1D (aged 8–17 y) investigated the effects of oligofructose-enriched inulin, a prebiotic, administered over 12 weeks on intestinal dysbiosis and glycemic control. Children receiving the prebiotics experienced increased C-peptide levels and improved intestinal permeability relative to the placebo group. Additionally, the prebiotic-treated group had a higher abundance of Bifidobacterium , while the placebo-treated group had a significantly higher relative abundance of Roseburia inulinivorans, Streptococcus, Terrisporobacter , and Faecalitaleawhich. This finding indicates an inexpensive and potentially low-risk prebiotic treatment approach to improve glucose control in T1D patients. Research on T1D-related complications and the underlying mechanisms influenced by prebiotics remains limited and inconclusive, presenting a clear opportunity for further investigation. Although promising effects on immunomodulation and reducing inflammation, most prebiotic and probiotic studies in T1D remain small, short-term, or limited to preclinical models. While SCFA-mediated immunomodulation has been a leading focus, other mechanisms, such as direct bacterial–immune interactions and microbe-derived vesicles or secreted factors, remain insufficiently studied. Importantly, well-controlled clinical studies evaluating the effect of probiotics on T1D progression in early-stage T1D (in individuals with two or more AAb+) are still lacking.

### Dietary fibers

Dietary fiber is a critical nutritional component found in a wide range of commonly consumed foods and helps maintain gastrointestinal and metabolic health. A study involving 10 children with T1D identified an inverse correlation between mean maximum postprandial blood glucose and a fiber-supplemented diet. The coronary artery calcification in type 1 diabetes (CACTI) study, comparing 568 T1D patients and 1257 nondiabetic controls, revealed a significant inverse relationship between fiber intake and HbA1c levels at baseline in T1D patients, although the natural differences in HbA1c levels between these groups should be considered. Supporting these findings, additional research has demonstrated that high fiber intake not only improves HbA1c but also reduces hypoglycemic events in T1D patients. Notably, cardiovascular health benefits have also been documented.

Inflammatory markers and metabolic effects have been studied as well. Daily fiber intake exceeding 20 g/d resulted in an inverse correlation between high-sensitivity C-reactive protein (hs-CRP) levels (a key marker of cardiovascular risk and systemic inflammation) and fiber intake, even after adjusting for HbA1c and energy intake. Additional studies found that increased fiber intake reduces amino acid absorption and decreases protein degradation and synthesis while also lowering cholesterol levels, though it did not affect insulin requirements or insulin-mediated glucose disposal. , Notably, fiber intake during pregnancy has shown specific benefits. A study of pregnant women with T1D found that higher dietary fiber intake (20.5 g/d) was associated with 16 to 18% lower insulin requirements compared to women consuming 8 g/d fiber intake.

Recent research has highlighted the crucial relationship between dietary fiber and the gut microbiota composition in T1D. Studies have shown that individuals with multiple autoantibodies demonstrate distinct microbial patterns, particularly an increase in Bacteroides and a reduction in Prevotella and Firmicutes abundance. These patterns appear to correlate with dietary habits, where a “Western diet” (high in refined carbohydrates, low in plant fiber) promotes Bacteroides growth, while plant fiber-rich diets support Prevotella abundance. Notably, both Prevotella and Firmicutes possess enzymatic ability to digest plant polysaccharides and promote the production of SCFAs, which are known for their anti-inflammatory properties. Supporting these findings, intervention studies with fiber-rich foods like barley kernel-based bread have demonstrated that individuals with higher Prevotella/Bacteroides ratios show improved glucose metabolism. Moreover, when microbiota from these “responders” was transplanted into germ-free mice, the animals exhibited both enhanced glucose metabolism and increased liver glycogen content, suggesting a direct causal link between fiber-responsive gut bacteria and metabolic improvements.

### SCFAs and T1D

SCFAs are a class of fatty acids produced by the gut microbiota through the saccharolytic fermentation of indigestible dietary carbohydrates. They are primarily produced in the colon, where they are either used locally or transported across the epithelium into the blood circulation. T1D patients have lower levels of fecal SCFA, butyrate, and acetic acid compared to healthy controls. As signaling molecules, SCFAs engage in two main signaling pathways: histone deacetylases (HDAC) inhibition and G-protein-coupled receptors activation. There has been extensive research on SCFAs' roles as regulators of human metabolism and gut integrity, and perhaps most excitingly, the immune system. Because SCFAs are crucial in regulating inflammatory responses, decreases in SCFA-producing bacteria have been associated with proinflammatory and autoimmune disorders like T1D and inflammatory bowel disease. This dysbiosis is now recognized as a driver of disease progression and a target for SCFA-based therapies as a potential treatment strategy.

#### Mechanistic studies on the effects of SCFA in T1D

The mechanisms of individual SCFAs are also currently under investigation, as their properties must be known before attempting medical application. Significant progress has been made in testing individual SCFAs in NOD mice and rodents with virally-induced T1D. One study testing acetate and butyrate diet supplementation in NOD mice found a reduction of cytotoxic T cells (specifically, glucose-6-phosphatase catalytic subunit–related protein (IGRP+) T cells), delaying diabetes incidence from 10 weeks to about 25 weeks. This diet also alters B-cell differentiation and increases Treg cells, resulting from reduced gene expression of histone deacetylase (HDAC). Similar SCFA-producing prebiotics effects of T1D are also discussed in the earlier prebiotic section. Another experimental study on 10–12-week-old pregnant NOD females sought to find how butyrate supplementation during gestation and nursing would affect both mothers and their offspring. T1D susceptibility was further exacerbated in some mice with prolonged treatment of vancomycin, an antibiotic that reduced SCFA in mice and is associated with gut dysbiosis and disrupted immune homeostasis in humans. Maternal butyrate treatment significantly protected both mothers and their female offspring from insulitis and vancomycin-induced T1D by reducing the vancomycin-induced secretion of the proinflammatory cytokines INF-γ and IL-1β in the pancreas, thus effectively inhibiting pancreatic inflammation. Additionally, butyrate promoted Treg function while inhibiting vancomycin-induced recruitment of inflammatory pathogenic IFN-γ + T cells in the pancreas of female offspring. Butyrate administration also increases Treg cells in the colon, mesenteric lymph nodes, and Peyer's patches, and α4β7, CCR9, and GPR15-expressing Tregs in the pancreatic lymph nodes of NOD mice. A different study on LEW1.WR1 rats that were similarly treated with Kilham Rat Virus (KRV) to induce β-cell inflammation and T1D were then given SCFA therapy by integrating sodium butyrate, sodium formate, and sodium propionate into their diets. SCFA therapy in 6-week-old KRV-infected rats reduced T1D incidence from ~60% to ~20% and significantly decreased insulitis in pancreatic islets. When analyzing the fecal microbiota of the offspring of these rats, it was revealed that SCFA treatment in breeding mice caused significant changes in the composition and abundance of certain bacterial taxa in the gut microbiome of the offspring. In this way, SCFA therapy was able to override virally induced alterations of the gut microbiota.

Many experiments evaluating the effectiveness of SCFA therapy have also revealed a degree of vertical transmission of protection from mother to offspring. For example, the offspring of KRV-infected rat breeders exhibited significant alterations in gut microbiota composition, providing evidence that SCFAs may mitigate virus-induced islet autoimmunity in the next generation. Likewise, butyrate administration to pregnant NOD mice (at 10–12 weeks' gestation) conferred protection against both insulitis and vancomycin-induced T1D in mothers and their offspring. These findings underscore a critical perinatal window during which maternal gut microbiota alterations may influence the development of innate and adaptive immunity in offspring.

#### Intervention studies on the effects of SCFA in T1D

Emerging evidence consistently links T1D with reduced levels of SCFAs, particularly butyrate, and a corresponding loss of SCFA-producing bacteria. In children recently diagnosed with T1D, the fecal concentrations of total SCFAs, as well as specific components like acetate and butyrate, are significantly lower than those in healthy controls. Similar SCFA deficits are observed in adults with longstanding T1D. For instance, a 2021 study of Danish Finnish adults reported significantly reduced levels of fecal propionate and butyrate in individuals with T1D compared to non-diabetic controls, although the levels of other SCFAs such as acetate and valerate, remained comparable. Moreover, the overall ratio of total SCFAs to branched-chain SCFAs was diminished in the T1D group, reinforcing the association between T1D and altered gut metabolite profiles. These changes in SCFA levels are closely tied to shifts in the gut microbial landscape. Metagenomic analyses in pediatric T1D cohorts have revealed a marked decline in canonical butyrate-producing taxa such as Faecalibacterium prausnitzii , Eubacterium rectale , and Roseburia spp., alongside an increase in potentially pathogenic species like Escherichia coli and Bacteroides spp. Finally, a study from the Innovative Approaches to Understanding and Arresting Type 1 Diabetes (INNODIA) cohort found that the gut microbiome composition is linked to disease progression in newly diagnosed T1D patients. Over 2 y, 21 bacterial species increased in those with new-onset T1D, while greater Faecalibacterium prausnitzii abundance, a beneficial butyrate-producing commensal, correlated with lower HbA1c at diagnosis.

SCFAs are increasingly being studied in the context of T1D, a T cell-mediated disease, and have the potential to benefit greatly from SCFA administration. This research must begin with characterizing the changes in the gut microbiome and dysbiosis that occur throughout disease progression. The 2015 DIABIMMUNE cohort study included 33 genetically at-risk infants from Finland, Estonia, and Russia and utilized metagenomic sequencing to analyze gut microbiome changes across three stages of T1D development: nonconversion, seroconversion, and clinical onset. The microbiome of T1D subjects experienced decreased alpha diversity and a relative overabundance of certain taxonomic groups. In contrast, seroconverters showed intermediate abundances compared with patients at all stages of disease development, implying that there is a microbiome composition shift linked to the T1D diseased state. Another 2019 longitudinal study analyzing the gut microbiome of 47 children with islet autoimmunity or T1D showed data consistent with the aforementioned study: the children had clear microbiome dysbiosis, lacking in an abundance of bacteria such as Prevotella and Butyricimonas , which are known to have both anti-inflammatory and SCFA-producing properties.

### Fecal microbiota transplantation (FMT) and T1D

FMT is the transfer of processed donor stool to reshape a recipient's intestinal community that has moved from its ancient roots to a standardized therapy first proven effective against Clostridioides difficile infection. In T1D, a single‐patient case report described a 24-y-old woman with long‐standing T1D complicated by malnutrition and poor glycemic control who received one donor FMT via infusion, leading to gradual improvements in constipation, nutritional status, fasting glucose, and HbA1c over subsequent weeks. A randomized, double-blind pilot trial in 20 adults with established T1D and moderate-to-severe diabetic gastroenteropathy compared a single 25-capsule dose of encapsulated FMT versus placebo. The researchers found that, at four weeks, the FMT group experienced significantly greater reductions in gastrointestinal symptoms and quality-of-life scores without serious adverse events. Finally, in a study of 20 individuals aged 18–30 within six weeks of T1D diagnosis, participants were randomized to receive three FMTs (either autologous or healthy donors) over four months via nasojejunal infusion. At 12 months, those receiving autologous FMT showed significantly better preservation of stimulated C-peptide release, and specific gut bacteria and plasma metabolites correlated with β-cell function retention. These findings collectively indicate that microbiome modulation via FMT holds promise as a novel adjunctive approach to delay β-cell decline and manage complications in T1D. However, larger, controlled studies are needed to identify which patients are most likely to benefit and to elucidate the microbial and host mechanisms driving these therapeutic effects. In addition, regulatory, donor safety, and standardization challenges currently limit the broad application of FMT in T1D. While proof-of-concept studies suggest potential benefits in preserving β-cell function or improving metabolic outcomes, the evidence base remains preliminary, and FMT should be considered an experimental therapy pending further validation.

Conclusion and future directions

This review highlights the complex relationships among dietary factors, the gut microbiome, and T1D. Across numerous studies, specific dietary patterns have emerged as potential modulators of T1D risk through their influence on the gut microbial composition. Protective dietary practices, including extended breastfeeding, gluten-free diets, balanced fatty acid intake, and high fiber consumption, have been associated in some studies with beneficial microbiome signatures and reduced T1D risk, though findings remain inconsistent, and causality has not been established. Conversely, early exposure to red meat and cow's milk proteins, coupled with high glucose consumption, has been associated in some studies with increased T1D incidence. These dietary factors appear to exert their effects primarily by altering gut microbial communities, particularly by diminishing the abundance of SCFA-producing bacteria ( ).

The current evidence suggests that diet-driven microbiome alterations play a role in T1D pathogenesis. However, most findings are derived from observational studies, and the causal links between dietary factors, microbiome shifts, and T1D onset remain unconfirmed. Future research should focus on determining whether observed microbiome alterations are direct drivers of T1D progression or simply byproducts of dietary modifications. To address this causality question, we focus on the molecular mimicry mechanism and the gut microbiota. , Therefore, elucidating these mechanisms is critical for developing targeted dietary interventions and microbiome-based therapeutic strategies for T1D prevention and management. A critical gap in the current evidence base is the lack of dietary intervention or observational studies explicitly stratified by T1D disease stage. The recent staging framework of early-stage T1D, which defines stage 1 (autoantibodies alone), stage 2 (autoantibodies with dysglycemia), and stage 3 (overt clinical diabetes) will provide a valuable lens through which to examine disease etiology and progression. However, most dietary studies included in this review predate the formal adoption of this staging system or were not prospectively designed with stage-stratified analysis in mind. Future research should explicitly examine whether dietary factors differentially influence the transition between stages, progression rates within stages, or the likelihood of advancing to clinical diabetes. Such a stage-specific investigation would greatly enhance our understanding of the causal role of diet in T1D pathogenesis. Additionally, there is a notable absence of studies examining the relationship between specific dietary factors and T1D endotypes. T1D represents a heterogeneous disease, with emerging evidence suggesting distinct endotype clusters based on metabolic, immunological and genetic characteristics. , It is plausible that certain dietary exposures may preferentially affect particular endotypes or may modify disease progression differentially across endotype subgroups. For example, dietary antigens or bacterial metabolites may have divergent effects depending on underlying immune regulation profiles or genetic susceptibility variants. The current review identifies this as an important future research direction; studies integrating detailed immunophenotyping, metabolic profiling, or genetic stratification alongside comprehensive dietary assessment would strengthen our mechanistic understanding of the role of diet in T1D. Looking ahead, we propose several priorities for the field. First, dietary and microbiome interventions should be prospectively designed around T1D staging and endotypes rather than treating T1D as a single homogeneous entity. Second, mechanistic trials should move beyond broad “healthy diet” prescriptions to test specific microbial pathways, such as SCFA and secondary bile acid signaling, IL-2 and Treg modulation, and barrier integrity using defined fibers, metabolites, or microbial consortia. Third, future studies should integrate deep immune phenotyping, microbiome and metabolome profiling, and host genetics to disentangle causal pathways from correlates. Finally, regulatory frameworks will be needed to support the development of next-generation microbiome-based therapeutics, including engineered probiotics and metabolite analogs, as adjuncts to immunomodulatory therapies. Together, these approaches will be essential to translate diet-microbiome insights into precise, stage-specific strategies to prevent or delay T1D.