Drugs vs. Microbiota: How Chronic Pharmacotherapy Affects Gut and Probiotic Bacteria
This narrative review describes directional gut microbiota changes induced by chronically used drugs (metformin, statins, PPIs, antihypertensives, levothyroxine, semaglutide) but does not establish definitive causal relationships or quantify effects with statistical precision sufficient to guide clinical decisions.
| Outcome | Grade | Direction | Effect | Studies |
|---|---|---|---|---|
| Gut microbiota composition with metformin | C | ▲ Favorable | Narrative: increase Akkermansia muciniphila, SCFA-producing bacteria; decrease Faecalibacterium, Ruminococcus. No pooled effect size or 95% CI reported. | — |
| Gastrointestinal adverse effects of metformin | C | ▼ Unfavorable | Incidence range 2-30% across primary studies; no pooled RR or 95% CI reported. | — |
| Gut microbiota composition with statins | C | ▲ Favorable | Narrative: increase Lactobacillus, Bifidobacterium; LDL reduction reported in primary studies. No pooled effect size or 95% CI. | — |
| C. difficile infection risk with PPIs | B | ▼ Unfavorable | Narrative reference to documented increased risk; no pooled OR or 95% CI calculated in this review (supported by external meta-analyses cited). | — |
| Gut microbiota composition with semaglutide | D | — Insufficient | Narrative: rodent and limited human data; no effect size or 95% CI reported. | — |
| Efficacy of adjunctive probiotics with pharmacotherapy | C | — Insufficient | Narrative: small isolated studies; no pooled effect size or 95% CI; strain-specific and drug-class-specific data insufficient. | — |
| Microbial diversity with PPIs | C | ▼ Unfavorable | Narrative: decrease in alpha-diversity; increase Streptococcaceae; no pooled effect size or 95% CI reported. | — |
Context
Chronic medications for metabolic diseases are taken by growing patient populations for decades. The gut microbiota modulates drug efficacy and toxicity, and these drugs in turn alter microbiota composition. Understanding these interactions is relevant to managing adverse effects and to the rationale for adjunctive probiotic use.
What the study showed
Metformin increases Akkermansia muciniphila and SCFA-producing bacteria while reducing Faecalibacterium and Ruminococcus; gastrointestinal side effects occur in 2–30% of users. Statins (atorvastatin, rosuvastatin) increase Lactobacillus and Bifidobacterium and reduce LDL, but primary studies have small samples and heterogeneous designs. PPIs increase Streptococcaceae and reduce microbial diversity, and are associated with higher risk of C. difficile infection. Semaglutide modifies microbial composition in rodents and humans, but controlled clinical data remain scarce. No outcome is presented with a consolidated 95% CI in this review.
How it was done
Narrative review with systematic database search in PubMed, Scopus, and Web of Science (October 2024–August 2025). Included experimental (in vitro, in vivo), clinical trials, and review articles on chronic drugs and microbiota. No meta-analysis was performed; no standardised data extraction or formal risk-of-bias assessment was conducted.
Effect magnitude
No consolidated effect size (RR, OR, SMD, MD with 95% CI) is calculated in this review; findings are qualitative and semi-quantitative descriptions from heterogeneous primary studies.
Limitations
Narrative review without a registered protocol (no PROSPERO or equivalent), with no formal risk-of-bias assessment (no tool such as AMSTAR-2, RoB 2, or ROBINS-I applied). Simultaneous inclusion of in vitro, animal, and clinical studies without certainty-of-evidence stratification. Heterogeneity in populations, doses, durations, and microbiota analysis methods precludes quantitative synthesis. Publication bias not assessed. Conclusions on probiotics are based on small, short-duration studies.
In clinical practice
Clinicians should note that chronically used drugs — particularly PPIs and antibiotics — alter the microbiota in clinically meaningful ways (C. difficile risk is documented). For metformin, pre-treatment gastrointestinal symptom screening and slow dose titration are strategies supported by moderate evidence. Adjunctive use of specific probiotics (e.g., Lactobacillus rhamnosus GG with PPIs, Bifidobacterium with metformin) is supported only by isolated grade C studies; no standardised indication exists.
What is still missing
Controlled RCTs are needed that assess concrete clinical outcomes (not merely microbiota composition) of adjunctive probiotic use for each drug class, with standardised metagenomic analysis methodology and a minimum follow-up of 12 months.