Changes in gut bacteria linked to GLP-1RA and SGLT-2i treatments in type 2 diabetes: clinical and biological insights

In the gastrointestinal system GLP-1RAs delay stomach emptying, increase satiety signals, decrease intestinal motility, and regulate bile acid metabolism [1, 5]. It is possible that GLP-1RAs may directly affect the gut environment, as the presence of GLP-1 receptors on enteric neurons, immune cells, and enteroendocrine L-cells throughout the gut suggests [5, 20]. This local gut distribution of GLP-1 receptors supports a potential bidirectional interaction between GLP-1 signaling and the gut microbiota, which could potentially contribute to the therapeutic benefits observed.

Preclinical and clinical studies demonstrate that GLP-1RAs modulate the gut microbiome, restoring dysbiotic patterns and encouraging bacterial diversity. Notably, liraglutide treatment was associated with an increase in beneficial taxa such Lactobacillus, Bacteroides, and A. muciniphila, as well as a decrease in the Firmicutes to Bacteroidetes ratio, a crucial dysbiosis marker in metabolic diseases [6, 10]. These changes are associated with weight reduction and better glucose tolerance, which suggests that changes in the microbiota may both reflect and encourage metabolic improvement [6].

Furthermore, by creating niches for specific bacterial populations, GLP-1RAs may influence microbial richness through altered pH profiles, bile acid modulation, and altered gut motility [5, 11]. GLP-1Ras treatment was associated with increased numbers of two types of bacteria that make SCFA, Roseburia and F. prausnitzii, in both human and animal investigations. These bacteria are linked to better blood glucose regulation, a stronger intestinal barrier, and less inflammation [9, 12].

Changes in the microbiota brought on by GLP-1RAs are associated with changes in key microbial metabolite patterns that may impact host metabolism. Following GLP-1RA treatment, SCFAs, such as butyrate and propionate, are elevated [9, 10]. These SCFAs are known to improve insulin sensitivity, regulate appetite and hormones such as peptide YY (PYY), and reduce systemic inflammation [6, 12]. Reduced levels of gut-derived endotoxins, such as LPS, indicate that the intestinal barrier is intact and that pro-inflammatory chemicals are not translocating as much [5, 10, 11].

There is a bidirectional interaction between GLP-1/biliary acids and the gut microbiota [21, 22]. Bacterial-derived hyodeoxycholic acid (HDCA) promotes GLP-1 secretion by upregulating the expression of the bile acid receptor Takeda G protein-coupled receptor 5/G-protein-coupled bile acid receptor 1 and downregulating the expression of the bile acid receptor farnesoid X receptor in the gut [22]. In addition, liraglutide increase the serum levels of deoxycholic acid [19] and thus, may also have an impact on bile acid signaling. Hepatic lipid metabolism, glucose regulation, and gut hormone release are all affected by these relationships [3, 11]. Together, these results indicate that gut microbiota and its metabolites may actively contribute to the systemic effects of GLP-1RA therapy, rather than serving as passive responders.

Emerging clinical evidence also suggests that gut microbiota may play a mechanistic or modulatory role in the therapeutic efficacy of GLP-1 receptor agonists in T2DM, although findings remain heterogeneous across studies. In a pilot study involving 52 patients with T2DM treated with GLP-1RAs, 16S rRNA amplicon sequencing was used to characterize gut microbial composition and identify taxa associated with glycemic response. Responders and non-responders exhibited significantly different beta diversity profiles, with specific microbes such as Bacteroides dorei and Roseburia inulinivorans positively correlating with HbA1c reduction, while taxa including Prevotella copri and Mitsuokella spp. were negatively associated, suggesting that baseline gut microbiota may serve as a predictive biomarker for GLP-1 RA efficacy in clinical management of T2DM [23].

However, some studies have reported contrasting results. For example, in a 12-week randomized, double-blind, placebo-controlled trial (NCT01744236), 51 adults with T2DM, already receiving metformin and/or sulfonylureas, were assigned to liraglutide (1.8 mg), sitagliptin (100 mg), or placebo to assess the impact of GLP-1 receptor agonism and DPP-4 inhibition on gut microbiota composition. Despite significant improvements in glycemic control with both agents, particularly a 1.3% HbA1c reduction with liraglutide, and modest weight loss, 16S rRNA sequencing of fecal samples showed no significant changes in either alpha or beta diversity of the gut microbiota across groups. In addition, microbiota composition did not correlate with clinical outcomes, although shifts in fecal bile acid profiles (e.g. increased deoxycholic acid with liraglutide, and cholic/chenodeoxycholic/ursodeoxycholic acids with sitagliptin) were observed [24]. On the other hand, while no significant changes in microbial diversity or composition were observed after the first week of dulaglutide treatment, long-term (>48 weeks) administration led to a significant reduction in microbial abundance (as indicated by Chao1 and Ace indices) and a distinct shift in community structure (beta diversity), without changes in alpha diversity in a longitudinal clinical study of 41, newly diagnosed, treatment-naïve T2DM patients. Notably, key metabolic markers, including fasting glucose, HbA1c, BMI, and fasting C-peptide, were significantly correlated with the abundance of specific bacterial genera such as Bifidobacterium, Bacteroides, Ruminococcus, and Blautia [25]. Overall, these findings might suggest that the metabolic benefits of liraglutide, at least in the short term, may occur independently of detectable changes in gut microbial diversity or composition. Meanwhile, dulaglutide, for example, may exert its metabolic benefits in part through long-term stepwise modulation of gut microbial composition, highlighting a potential bidirectional relationship between GLP-1 receptor agonism and the intestinal microbiome in T2DM treatment.

In this context, another consideration is that while GLP-1RAs analogs influence the composition and diversity of the gut microbiota in both animal and human models, effects might potentially vary significantly by drug type, population characteristics, and treatment duration. A recent systematic review analyzed 38 preclinical and clinical studies to assess the impact of GLP-1RAs, including liraglutide, exenatide, dulaglutide, and semaglutide, on gut microbiota composition, diversity, and metabolic relevance, showed that while liraglutide consistently promoted the enrichment of metabolically favorable genera (e.g. Akkermansia, Bacteroides), dulaglutide was associated with increases in Ruminococcus and Akkermansia, while semaglutide produced mixed outcomes, enhancing A. muciniphila abundance but reducing microbial diversity, therefore suggesting that GLP-1RAs exert compound- and context-specific effects on the microbiome that may depend on host factors, treatment duration, and baseline microbial composition [26].

Microbiota effects may also contribute to the extra-glycemic benefits of cardiovascular and kidney protection, and anti-inflammatory activity. For instance, improved endothelial function and a slower rate of atherosclerosis progression are linked to increased abundance of A. muciniphila [10, 26]. SCFA enrichment is linked to fewer oxidative stress markers and more favorable glycemic variability [3]. Thus, shifting gut microbiota could be a sign of therapy responsiveness and a possible therapeutic enhancer.

Potential future applications in microbiome-personalized diabetes care may be supported by a few early studies that also indicate the baseline microbiota composition may predict the extent of the metabolic response to GLP-1RAs, suggesting that treatment of diabetes in a microbiome-specific manner should be explored [3, 12]. However, more extensive and continuous human studies are still required to validate these findings.

SGLT-2is, including dapagliflozin, empagliflozin, and canagliflozin, act primarily by inhibiting renal glucose reabsorption in the proximal tubules, thus promoting glycosuria. However, in addition to decreasing serum glucose, they have a wide range of metabolic effects. Notably, SGLT-2is induce natriuresis and osmotic diuresis, causing a contraction in plasma volume, reduction in blood pressure, and improvement in cardiac preload and afterload [6, 28]. They also decrease glomerular hyperfiltration, increase hemoglobin levels and promote the production of the kidney antiaging protein Klotho, which delays cardiovascular aging [29]. These effects are believed to contribute to the observed cardiovascular and renal protective outcomes in major clinical trials.

SGLT-2is promote a metabolic transition toward ketogenesis by enhancing hepatic lipid oxidation and increasing circulating ketone bodies, which serve as efficient energy substrates for the kidneys and heart [30]. These ketones, especially β-hydroxybutyrate, may enhance mitochondrial efficiency and possess anti-inflammatory characteristics. Furthermore, SGLT-2is promote uricosuria, which lowers the blood uric acid levels. This action may be significant because hyperuricemia is linked to endothelial dysfunction and cardiovascular disease [31]. Together, these systemic effects create a metabolic environment that may positively influence the gut microbial environment. Despite increased uricosuria, they decrease the risk of urolithiasis in part through increased urinary citrate [32].

Recent research indicates that SGLT-2is directly modifies the composition and function of the gut ecosystem. Preclinical and human research studies have documented an increased abundance of bacteria producing SCFA such as F. prausnitzii, Roseburia, and Bifidobacterium following SGLT-2i administration [1, 33]. These bacteria are linked to better gut-barrier function and anti-inflammatory effects. Also, SGLT-2is have been associated with a lower Firmicutes/Bacteroidetes ratio and a lower number of pathobionts such Enterobacteriaceae, which suggests that the body is moving toward a more eubiotic state [34].

Furthermore, SGLT-2i have been linked to elevated SCFA levels in the colon, specifically butyrate and acetate, which are known to affect insulin sensitivity, glucose homeostasis, and appetite control and protect from kidney and cardiovascular disease [14, 35]. Reductions in systemic LPS levels, which serve as a marker of endotoxemia and gut-barrier dysfunction, support these metabolic benefits [6]. SGLT-2is may affect microbiota through modifications in intestinal pH, bile acid composition, and local nutrition availability, all of which influence microbial habitats.

The gut microbiota may have a mechanistic impact in improving the renoprotective and cardioprotective benefits of SGLT-2is. Microbial profiles associated with improvement in SCFAs are connected to diminished systemic irritation, improved endothelial work, and decreased oxidative stress, all critical for vascular and renal wellbeing [2, 33]. In murine models, changes in microbiota initiated by SGLT-2is were connected to diminished renal fibrosis and decreased generation of pro-inflammatory cytokines. Research shows that modifications in microbiota can be related with protection from kidney injury and improved kidney function, as shown by lower albuminuria levels and stability of glomerular filtration rate [1].

Alongside improvements in fasting glucose and weight maintenance, dapagliflozin was associated with a progressively greater shift in gut microbiota composition over time, particularly influencing taxa such as Lactobacillus and Muribaculaceae in a diabetic kidney disease mice model [36]. A previous preclinical mice study investigated whether the therapeutic benefits of empagliflozin in T2DM-related diabetic nephropathy were mediated via modulation of the gut microbiota [37] Using a high-fat diet and streptozotocin (HFD/STZ)-induced mouse model of T2DM-associated diabetic nephropathy, male mice were randomized to receive empagliflozin or vehicle for four weeks. Empagliflozin promoted renal protection by significantly reducing blood glucose, urine albumin-to-creatinine ratio, and renal histopathologic injury while restoring gut microbiota diversity and epithelial barrier integrity. 16S rRNA sequencing revealed reductions in LPS-producing bacteria (e.g. Oscillibacter) and increases in SCFA-producing taxa (e.g. Bacteroides, Odoribacter), accompanied by elevated fecal SCFA levels and reduced fecal and serum LPS concentrations. Meanwhile, antibiotic-treated cohorts showed diminished renal and intestinal effects of empagliflozin, supporting a microbiota-dependent mechanism [37].

There is also some clinical evidence that empagliflozin's cardiometabolic benefits in early T2DM may be partly mediated through gut microbiota modulation and systemic metabolic remodeling. For example, empagliflozin treatment uniquely improved cardiovascular disease risk profiles, increased microbiota richness and SCFA-producing taxa, and shifted circulating metabolites toward a more favorable cardiometabolic profile compared to metformin in a randomized clinical trial of 76 treatment-naïve adults with T2DM and at least one cardiovascular disease risk factor [38]. Similarly, reduced abundance of Fusobacterium following dapagliflozin treatment was associated with improved glycemic, lipid, inflammatory, and endothelial parameters, as observed in an 8-week, single-arm clinical trial involving 12 patients with T2DM treated with metformin [39].

Current clinical evidence from human studies exploring the microbiota-related metabolic, clinical, or prognostic effects of SGLT-2is remains limited. In a double-blind, randomized trial of 44 metformin-treated patients with T2DM, neither dapagliflozin nor gliclazide significantly altered gut microbiome composition over 12 weeks, despite exhibiting divergent metabolic effects such as reduced weight and fasting insulin with dapagliflozin and increased insulin with gliclazide [40]. Both agents improved glycemic control comparably, but the absence of changes in microbial diversity or taxa suggests their metabolic actions are not mediated via alterations in the fecal microbiota.

In addition, SGLT-2i use in pre-dialysis CKD patients was associated with reduced circulating levels of protein-bound UTs of bacterial origin (indoxyl sulfate and p-cresyl sulfate), alongside distinct shifts in gut microbial composition, particularly enrichment of Bacteroides stercoris and Bacteroides coprocola, and changes in microbial functions linked to protein and carbohydrate metabolism in a matched case-control study of 60 CKD patients (SGLT-2i treatment, n = 30; untreated, n = 30) and 30 non-CKD controls. Although serum SCFA levels were lower in SGLT-2i treated patients, the study also identified gut taxa–metabolite correlations implicating a host–microbe–metabolite axis in CKD [41].

The glucose lowering response to SGLT-2is and GLP-1RAs may be anticipated by the nature of the gut microbiota, according to recent studies. A. muciniphila and F. prausnitzii have been associated to improved glycemic response in people taking GLP-1RAs [26, 43]. These associations support a novel hypothesis that proposes that gut microbial patterns may serve as predictive biomarkers of therapeutic efficacy. More broadly available metagenomic sequencing may allow the use of machine learning to stratify patients based on baseline gut microbiome composition and expected response to therapy. Microbiome-informed patient selection and customized treatment plans may result from this research.

In addition to being a biomarker, the gut microbiota might also become a modifiable component that can improve the therapeutic impacts of SGLT-2is and GLP-1RAs. Intervention that promote a healthy microbiota, such as a heathy diet rich in fiber, or probiotics, prebiotics, and even postbiotics may improve gut-barrier integrity, boost the synthesis of SCFAs, and potentially potentiate the benefit of GLP-1RAs [5, 6, 44]. FMT may also be used. It is already used to treat severe diarrhea caused by Clostridium difficile (https://www.nice.org.uk/guidance/mtg71/chapter/1-Recommendations↗). Recent trials have confirmed the feasibility of using pills (dubbed "poo pills" by the lay press) for FMT to modify the gut microbiota [45]. There is less information on SGLT-2is and the influence of the gut microbiota [33]. Combination techniques that use microbiome-directed therapeutics to supplement pharmaceutical treatment become possible as a result.

In more advanced renal disease, the potential changes in microbiota during the progression of kidney disease [16] may alter the effects of either drug or their combination. Alternatively, SGLT-2is and GLP-1RAs may correct these microbiota alterations and possibly induce improvements in metabolic profile seen in more advanced kidney disease, e.g. below an eGFR of 30 ml/min [16, 46]. These questions should certainly be addressed in future studies.

An emerging area of diabetes management is precision and personalized diabetes treatment based on gut microbiota analysis. Individualized therapy regimens based on microbiome profiles might be possible with a better understanding of host–microbe interactions. Depending on their microbial composition, some patients may benefit more from SGLT-2is, whereas others with low Akkermansia abundance may theoretically be candidates for GLP-1RA therapy [26, 43].

While microbiota-informed personalization is an appealing concept, translation into clinical practice remains preliminary. Current evidence primarily supports associations rather than causality, and predictive microbial signatures require validation in independent, large-scale cohorts before guiding therapy selection. Although developing frameworks propose integrating multi-omics data—including clinical indicators, and metabolomic and microbial profiles—to design personalized regimens, these approaches remain experimental. Robust validation in prospective clinical trials is essential before microbiota-informed therapy can be incorporated into standard care [5, 33].