Antibiotics you took years ago still show up in your gut

Gut bacteria differ enormously between people, and medications are one known reason why. Most microbiome research, though, only asks what someone is taking at the moment the stool sample is collected. That ignores how medicines are actually used: repeatedly, in courses, over years, then stopped. If a drug taken long ago still leaves a mark, then studies linking gut bacteria to disease could be picking up old prescriptions rather than the disease itself. This paper set out to test how far back that shadow reaches.

The team combined the observational Estonian microbiome cohort, using shotgun metagenomic sequencing of stool from 2,509 participants, with retrospective medication records drawn from electronic health records. That let them look not just at current use but at the full history of prescriptions for antibiotics and a wide range of human-targeted drugs. They then systematically tested which drug classes were associated with gut microbiome features and how long after use those associations persisted. A subgroup of 328 participants had a second microbiome measurement, allowing the researchers to look at changes around the point where a treatment was started or stopped.

Past medication use, not just current use, was associated with the gut microbiome. Effects were detectable several years after use for antibiotics, psycholeptics, antidepressants, proton pump inhibitors and beta-blockers. The authors describe the pattern as additive: the more a medication had been used, the stronger the apparent impact. Carryover and additive effects also showed up for non-antibiotic classes including benzodiazepine derivatives and glucocorticoids. In the 328 people with two samples, similar microbiome changes accompanied both starting and discontinuing treatment, which the authors say possibly indicates causal effects.

Earlier work by this group and others had already shown that current antibiotic and antidepressant use tracks with gut composition, and hinted that past use matters too. This is described as the first study to use detailed retrospective prescription data from health records to assess long-term medication effects systematically. It extends the picture from a snapshot to a history, and it complicates the interpretation of any study that links gut bacteria to a disease without accounting for years of drug exposure. What it cannot settle is how much of the difference is caused by the drugs themselves rather than by the conditions they were prescribed for.

If you have ever wondered why your gut profile looks unusual, this is a reason to think the answer may lie further back than your current diet or current prescriptions. It also gives you a reason to treat headlines linking gut bacteria to a specific disease with a little more caution, since medication history is a plausible hidden factor. The study does not show that any of these changes harmed health, and it says nothing about whether the changes can be reversed. It is a mapping exercise, not a guide to action.