Medications taken years ago may still be shaping gut bacteria today, according to a large study — and the effect is not limited to antibiotics.
Researchers at the University of Tartu Institute of Genomics, led by Dr Oliver Aasmets, analysed stool samples and prescription records from more than 2,500 participants in the Estonian Biobank, publishing in mSystems.
What they found
Lasting changes in the gut microbiome were linked to several common drug classes: antibiotics (penicillins, macrolides), antidepressants (SSRIs), beta-blockers, proton pump inhibitors and benzodiazepines.
Some changes “could still be detected years after people had stopped taking the medication.”
Why the Estonian Biobank made this possible
The methodological advantage is worth noting, because it explains why this study could ask a question others could not.
Establishing that a drug taken years ago affects gut bacteria today requires knowing precisely what someone took and when — over years. Self-reported medication history is unreliable, particularly for drugs taken briefly in the past.
Estonia maintains linked national health records, so prescription histories can be matched to biological samples with a precision most cohorts cannot achieve. The finding depends on that infrastructure at least as much as on the microbiome analysis.
Why antibiotics were expected and the others were not
Antibiotics are designed to kill bacteria, so their effect on gut flora is unsurprising — and the persistence is the notable part, since the microbiome is often assumed to recover within weeks to months after a course.
Proton pump inhibitors have a fairly clear mechanism too. By suppressing stomach acid, they remove a barrier that normally kills swallowed organisms and shapes what reaches the intestine.
The others are harder. Beta-blockers act on cardiovascular receptors; SSRIs act on serotonin transport — though notably most of the body’s serotonin is in the gut, providing at least a plausible route.
The genuinely surprising result
Benzodiazepines — anxiety and sleep medications with no antimicrobial function — showed effects on gut bacteria comparable to those seen with broad-spectrum antibiotics.
That is a striking claim, and it has no obvious mechanism. Benzodiazepines act on receptors in the central nervous system, and there is no established reason they should reshape bacterial communities as profoundly as drugs designed to kill bacteria.
Possible explanations exist — effects on gut motility altering the environment, direct effects on bacteria that happen to carry related receptors, or confounding by the conditions for which the drugs are prescribed. None is established, and the finding is best treated as a signal requiring explanation rather than an established effect.
The confounding problem
This is where caution is required, and it applies across the whole study.
People take medications because they are ill, and illness itself affects the microbiome. Someone prescribed proton pump inhibitors has reflux; someone on SSRIs has depression, which is independently associated with microbiome differences; someone on beta-blockers has cardiovascular disease and typically several other conditions and drugs.
Disentangling drug effects from disease effects, and from the effects of everything else a person with that disease does differently, is extremely difficult in observational data.
The authors’ practical point
The most useful takeaway is methodological: microbiome-disease research should account for medication history, so drug effects are not mistaken for disease-related changes.
That is a pointed observation about a large and enthusiastic literature. Studies routinely report microbiome differences between people with a condition and healthy controls, and conclude the microbiome is involved in the disease. If the patient group is also taking medications that reshape gut bacteria for years, some of those differences may be pharmacological rather than pathological — and this study suggests the window over which that applies is much longer than researchers have assumed.
What this does not mean
This is an observational study showing associations, not proof that these drugs cause harm. A changed microbiome is not automatically a damaged one, and the study did not link the observed changes to any health outcome.
The longitudinal follow-up also involved a relatively small group, which is the part of the analysis supporting the persistence claim — the most novel and least well-powered element.
What a changed microbiome does and does not mean
Microbiome findings are routinely over-interpreted in both directions, and this study is a useful case for calibrating.
The gut contains an enormous and highly variable microbial community, and healthy people differ from one another far more than most studies’ effect sizes. There is no single healthy microbiome composition against which deviation can be measured, which makes “altered” a harder claim to interpret than it sounds.
What matters is function rather than composition. Different bacterial communities can perform the same metabolic work, so a shift in which species are present is not automatically a shift in what the microbiome does for its host.
This study measured composition and did not link changes to health outcomes. So the honest reading is that these medications leave a durable compositional signature — which is genuinely interesting for research methodology, and says nothing yet about whether patients are affected by it. Establishing that would require following people with and without those signatures and observing what happens to them.
It is emphatically not a reason to stop any prescribed medication. This summarises research and is not medical advice.