This twelve-week, randomized, placebo-controlled, double-blind study compared the effects of a multi-strain probiotic versus a placebo in women with insulin resistance who were recommended by an internal medicine physician to take 1000 mg of metformin (500 mg twice daily, during or after meals). The study was registered on ClinicalTrials.gov (NCT06092060) on October 23, 2023. Blood samples were collected for biochemical analyses before and after the 12-week intervention, body composition assessments were performed, and participants completed questionnaires regarding gastrointestinal symptoms.

Participants were recruited through advertisements on the Polish Insulin Resistance Foundation–Healthy Diet and Healthy Life website. Eligible participants were included in this study if they met the following criteria: female, aged 25–45 years, BMI 25-39.9 kg/m², and insulin resistance (based on medical qualification: HOMA-IR and insulin curve were considered). All patients were required to be treated with metformin (1000 mg/daily). Patients were excluded if they had received metformin treatment before the study, had type 1 or type 2 diabetes, PCOS or had any acute or chronic clinically apparent inflammatory processes (connective tissue and joint diseases, respiratory tract inflammatory processes, urogenital inflammatory processes, inflammatory processes in the head and neck area), acute infection within the last month, cancer, alcohol abuse, drug addiction, immunodeficiency, chronic treatment with immunosuppressants, glucocorticoids, or use of antibiotics and non-steroidal anti-inflammatory drugs, steroids, antifungals within the last 30 days; daily consumption of synbiotics, prebiotics, and/or probiotics within the past 90 days; or travel to tropical countries in the last 4 weeks before the intervention. Other conditions that could pose any risk to the patient during the observation period were also exclusionary.

Consecutive potentially eligible patients attending the Department of Treatment of Obesity, Metabolic Disorders and Clinical Dietetics, Poznan University of Medical Sciences were recruited by a diabetologist for participation in this study.

All volunteers were introduced to the study’s purpose and protocols and provided written consent to participate. During the screening visit, the inclusion/exclusion criteria were confirmed based on interviews and medical records, and baseline demographic data were collected.

The study flow diagram is presented in Figure 1.

Before participation, all candidates underwent medical qualification. Patients from Poznan University of Medical Sciences, who met the inclusion criteria and were prescribed metformin by an internal medicine specialist, were invited to participate in the scientific project. A complete medical history and physical examination were conducted at the Department of Treatment of Obesity, Metabolic Disorders, and Clinical Dietetics of the Poznan University of Medical Sciences. Volunteers were informed about the purpose and methodology of the study and provided written informed consent.

Before starting the study, fasting venous blood samples were collected from all participants to measure glucose and insulin levels and calculate the HOMA-IR index (Matthews et al., 1985). Patients whose fasting glucose levels were between 3.0 and 25.0 mmol/L and fasting insulin levels between 3 and 55 mU/mL had their glycated hemoglobin (HbA1c) measured to exclude diabetes or prediabetes, based on the following criteria (Carson et al., 2010):

Thirty women diagnosed with insulin resistance, aged 25 to 45, were randomized into two groups using randomization software (www.randomizer.org), by the probiotic manufacturer in a 1:1 (probiotic: placebo) ratio. Allocation concealment was ensured with sequentially numbered drug containers of identical appearance. Blinding included both the authors of the study and the participants. The distribution of participants between the probiotic group (n = 16) and the placebo group (n = 14) in the final study group was impacted by elevated dropout rates and the presence of outliers, predominantly in the treatment group.

The probiotic group (Met+PROB) received two capsules twice daily (in the mornings and before bedtime) containing a lyophilized powder of the following probiotic strains: Bifidobacterium lactis W52, Levilactobacillus brevis W63, Lacticaseibacillus casei W56, Lactococcus lactis W19, Lactococcus lactis W58, Lactobacillus acidophilus W37, Bifidobacterium bifidum W23, Bifidobacterium lactis W51, and Ligilactobacillus salivarius W24 (Sanprobi Sp. z o.o. Sp. k., Szczecin, Poland) at a daily dose of 2 × 10⁹ colony-forming units (CFU), starting the day they began taking metformin (1000 mg/day) for 12 weeks. The probiotic preparation was microbiologically controlled to ensure consistency in strains and CFU counts throughout the study. Supplemented probiotics were considered safe and well-tolerated in the healthy population, and the tested product held the necessary safety certificates.

The placebo group (Met+PLAC) received two capsules twice daily containing only the probiotic carrier, corn starch and maltodextrins. All participants were instructed to take two capsules twice daily, once before breakfast and once before bedtime, starting on the same day they began metformin (1000 mg/day). Participants returned after four weeks to submit empty and partially used blister packs for compliance monitoring. They were also instructed not to change their usual physical activity or diet. Any side effects from metformin and/or the probiotic were to be recorded or reported by phone.

After initial qualification (based on inclusion and exclusion criteria), each participant underwent two sets of tests: at baseline (PRE) and after 12 weeks (POST).

Measurements were performed at the University of Physical Education in Poznan and Poznan University of Medical Sciences.

The sample size was calculated for HOMA-IR and was determined based on a similar study (15) using GPower version 3.1.9.4. Sample size analysis indicated that a minimum of 12 subjects per group was required to achieve at least 80% power to detect a statistically significant intervention effect at α = 0.05. To confirm the similarity between the Met+PROB and Met+PLAC groups, tests for comparisons of independent groups were performed before the intervention, using either the Student’s t-test or the Mann–Whitney U test, depending on the distribution of the variable. Anthropometric variables, dietary intake, and biochemical indicators were compared using two-way repeated measures ANOVA. For statistically significant results, Bonferroni post hoc tests were additionally conducted. To compare responses regarding gastrointestinal side effects, mean percentage values between groups were compared separately before and after the intervention. Quantitative variables were analyzed using the Student’s t-test or the Mann–Whitney U test (depending on variable distribution), and qualitative variables using Pearson’s Chi-square (χ²) test. The eta squared coefficient (η²) is presented as an indicator of effect size.

The primary outcome of this study was improvement in gastrointestinal symptoms assessed by the Rome IV questionnaire. After 12 weeks, participants receiving probiotics reported significantly lower frequencies of abnormal stool consistency and abnormal stool frequency during episodes of abdominal pain compared with the placebo group. These findings are clinically relevant, as gastrointestinal intolerance remains one of the most common reasons for metformin discontinuation and dose reduction, negatively affecting adherence and long-term metabolic outcomes (8).

Our results are consistent with recent clinical trials and meta-analyses indicating that probiotic supplementation can reduce metformin-associated gastrointestinal adverse events. A 2024 meta-analysis by Szymczak-Pajor et al. including randomized trials in patients with type 2 diabetes demonstrated significant reductions in diarrhea and abdominal discomfort with probiotic co-administration (14). Similarly, the ProGasMet trial reported improved gastrointestinal tolerability in metformin-intolerant patients receiving probiotic therapy (13). These data support the hypothesis that probiotics may enhance intestinal barrier function, normalize intestinal motility, and reduce luminal inflammation, thereby mitigating drug-related gastrointestinal disturbances.

The probiotic preparation used in the present study contained nine strains belonging to the genera Lactobacillus, Bifidobacterium, and Lactococcus. Several of these strains have previously been associated with modulation of intestinal permeability, short-chain fatty acid production, and regulation of bowel motility, which may explain the observed improvement in stool-related symptoms (16).

Strains of Lactobacillus acidophilus, Lactobacillus brevis, and Lactobacillus casei have been shown to support mucosal barrier integrity and reduce visceral hypersensitivity in functional bowel disorders, potentially contributing to reduced abdominal discomfort and improved stool consistency (17, 18). Bifidobacterium lactis and Bifidobacterium bifidum are frequently associated with enhanced production of acetate and lactate, which can stabilize intestinal transit and promote epithelial health (19).

Lactococcus lactis strains, while traditionally associated with dairy fermentation, have also demonstrated immunomodulatory properties that may attenuate low-grade intestinal inflammation (20). Importantly, the total probiotic dose used in this study (2 × 10⁹ CFU/day) was moderate compared with doses used in some metabolic studies. While this dose appears sufficient to influence gastrointestinal symptoms, it may be insufficient to induce measurable systemic metabolic effects, particularly in the context of concurrent metformin therapy, which itself strongly alters gut microbiota composition and host metabolism (9, 11, 12).

The results of studies investigating the effects of probiotic supplementation on glucose and lipid metabolism in patients with insulin resistance remain inconclusive. While some studies report beneficial effects, others—including the present study—have not demonstrated significant changes. For instance, a meta-analysis of 12 randomized controlled trials found that probiotic supplementation was associated with a significant reduction in glycated hemoglobin (HbA1c) levels and fasting insulin concentrations in patients with type 2 diabetes. However, the effects on fasting glucose levels, C-reactive protein (CRP), and lipid profiles were either non-significant or exhibited considerable heterogeneity (21). Despite theoretical synergy between probiotics and metformin on gut-mediated metabolic pathways, in our study no significant probiotic-related improvements in fasting glucose, insulin, HOMA-IR, RBP4, lipid profile, or body composition were observed. These findings align with several recent randomized trials reporting neutral metabolic effects of probiotic supplementation when combined with standard glucose-lowering therapy (21, 22).

Recent meta-analyses also indicate heterogeneity in metabolic responses to probiotics. A 2023 systematic review, which included 14 randomized controlled trials involving 1009 patients by Memon et al. found modest reductions in fasting glucose and HbA1c when probiotics were combined with metformin. Importantly, the frequency of gastrointestinal adverse effects was also found to be lower in the group receiving probiotics compared to the group receiving metformin alone (9). A 2025 network meta-analysis comparing different probiotic formulations in type 2 diabetes suggested that metabolic benefits are strain- and dose-dependent and may require higher CFU doses and longer treatment durations (12). Moreover, baseline metabolic severity appears to influence responsiveness, with larger effects observed in patients with established diabetes compared with those with early insulin resistance.

In the present study, participants had elevated HOMA-IR but normal glucose and HbA1c levels, which may have limited the potential for detecting metabolic improvement. Additionally, both groups demonstrated significant time effects for glucose, HOMA-IR, RBP4, and total cholesterol, reflecting the expected therapeutic response to metformin initiation. Any incremental probiotic effect may therefore have been masked by the dominant pharmacological action of metformin.

From a clinical perspective, improving tolerability of metformin may be as important as enhancing its metabolic efficacy, particularly during treatment initiation. Gastrointestinal side effects are a major contributor to poor adherence, and even mild symptoms may lead to dose reduction or discontinuation. The observed reduction in stool-related symptoms suggests that probiotic co-administration could represent a simple, low-risk strategy to improve patient comfort and persistence with therapy, especially in individuals initiating metformin for insulin resistance or early metabolic dysfunction.

Strengths of this study include the randomized, double-blind, placebo-controlled design, standardized assessment of gastrointestinal symptoms using Rome IV criteria, objective measurement of metabolic outcomes, and good adherence to both metformin and study supplementation.

Several limitations should be acknowledged. The sample size was small, limiting statistical power to detect small-to-moderate metabolic effects. Gastrointestinal outcomes were based on self-reported questionnaires, which may be subject to reporting bias. No gut microbiota, short-chain fatty acid, or bile acid analyses were performed, precluding mechanistic interpretation of probiotic effects. Additionally, metformin and probiotics were initiated simultaneously, making it difficult to distinguish whether probiotics prevented symptom development or merely attenuated ongoing intolerance. Finally, the study population was limited to women of reproductive age, which may restrict generalizability to men and older adults.

Future trials should include larger sample sizes, longer follow-up periods, and incorporation of microbiome and metabolite analyses to clarify mechanistic pathways. Stratification by baseline gastrointestinal sensitivity and insulin resistance severity may help identify subgroups most likely to benefit from probiotic co-administration. Comparative studies of different probiotic formulations and higher CFU doses may further optimize adjunctive strategies to improve metformin tolerability and metabolic outcomes.