Your past medications may still be shaping your gut microbiome?
A 2025 study published in the American Society for Microbiology (ASM) has found that medications don't just affect the gut microbiome while you're taking them, some leave a signature that lingers for years after the last dose.
“The effects of antibiotics, psycholeptics, antidepressants, proton pump inhibitors, and beta-blockers are detectable several years after use.”
What did they find?
● Researchers used electronic health records from the Estonian microbiome cohort (n = 2,509), a population-based volunteer group, to track five years of prescription drug use alongside stool microbiome sampling.
● Of 186 drugs analysed, 167 (89.8%) were associated with the microbiome, and 78 of these (46.7%) showed long-term "carryover" effects, still detectable more than three years after last use.
● The effect appeared dose-dependent (additive): the more a drug was used in the past, the stronger its lasting imprint.
● Long-term effects weren't limited to antibiotics (macrolides, penicillins). They also showed up with benzodiazepines, beta-blockers, glucocorticoids, proton pump inhibitors (PPIs), biguanides (e.g. metformin) and antidepressants.
● A second time-point in 328 participants provided evidence consistent with a causal, rather than purely correlational, long-term effect for combination penicillins, macrolides, PPIs and SSRIs.
● Benzodiazepines stood out: their impact on microbiome composition, and how long it persisted, rivalled or exceeded some broad-spectrum antibiotic classes. Alprazolam (Xanax) showed a broader effect than diazepam (Valium), a pattern also reported elsewhere.
Why does it matter?
Most microbiome research has focused on drugs taken at the time of sampling. This study suggests ignoring drug history could be masking a real, lasting influence on microbiome composition, and may confound disease-microbiome associations when past medication use isn't accounted for. Because drug burden varies with age, population and clinical setting, unaccounted-for drug history may also help explain some of the inconsistencies seen between microbiome studies in different cohorts.
But worth noting the limitations: The study captured prescription drugs only (not over-the-counter use), assumed a filled prescription equalled use, relied on relative rather than absolute abundance data, and the cohort skewed female and health-conscious, as volunteer biobanks tend to.
Alvina, “We've come a long way, and there's still a long way to go. That's the nature of science: questions beget questions.”
“This is a fascinating area of research, but it’s worth flagging that most drug-microbiome-nutrient triads are backed largely by correlational human data, and in some cases the findings are genuinely contested.”
What Alvina finds most interesting is the overlap. Metformin and PPIs, for example, sit in both categories: they're linked to nutrient depletion and to microbiome disruption. That raises a real question. Is vitamin B12 deficiency on these drugs a direct absorption or pharmacokinetic effect? A downstream consequence of microbiome shifts affecting B12-producing (or competing) flora? Or some combination of both, and if so, in what proportion?