Minocycline
An antibiotic that also calms the brain’s immune cells. Early trials suggested it eases negative symptoms; the two largest trials found no benefit.
At a glance
- Type
- Medicine: tetracycline antibiotic
- Used for
- Tested as an add-on to antipsychotics, mainly for negative symptoms
- Status
- Experimental, the two largest trials found no benefit
- Evidence
- EarlyHow evidence is rated
- Administration
- Capsules, added to an antipsychotic. In the trials, for 4 months to a year
- Common side effects
- Digestive complaints
- In Europe
- Tested in the UK, Israel and Romania, among other countries
Clinical summaryFor health professionals
- Proposed mechanism: inhibits activated microglia. In cells from patients, it reduced the uptake of synapses by microglia (Sellgren 2019).
- Chaudhry 2012 (Brazil, Pakistan): 144 patients within 5 years of onset, 12 months; negative symptoms improved by 9.2 points against 4.7 on placebo.
- BeneMin (UK, Deakin 2018): 207 patients within 5 years of onset, 200 to 300 mg a day for 12 months; no effect on negative symptoms, IL-6 or grey matter.
- Weiser 2019 (Israel, Romania): 200 patients, 200 mg a day for 16 weeks; no effect on PANSS or cognition.
- Meta-analysis of 10 trials, 895 patients: moderate benefit for negative symptoms (Shen 2026), but the effect is likely inflated by publication bias (Li 2026).
- People at risk of psychosis (NAYAB, Pakistan): 326 people, 6 months; no reduction in transition to psychosis (Qurashi 2024).
What it is
Minocycline is an antibiotic of the tetracycline family that passes from the blood into the brain. Besides killing bacteria, it calms microglia, the brain’s own immune cells.1 It was tested in schizophrenia, added to antipsychotics, mainly for negative symptoms such as low motivation and withdrawal, which antipsychotics barely improve.2
How it works
Microglia help remove unused synapses as the brain matures. In 2019, researchers grew microglia and nerve cells from people with schizophrenia and found that the microglia removed synapses more avidly than those from healthy people, partly in connection with the C4 gene. Minocycline reduced this removal in the laboratory. In health records, young adults treated with minocycline were slightly less often diagnosed with schizophrenia later than those given other antibiotics.3 How these findings relate to the illness is explained in Beyond dopamine (opens in a new tab).
How well it works
The first trials were encouraging. In Israel, 54 people in an early phase of schizophrenia took minocycline or a placebo for six months, and negative symptoms and planning skills improved more on minocycline.1 In Brazil and Pakistan, 144 people took it for a year: negative symptoms improved by 9.2 points, against 4.7 on placebo.4
The two largest trials did not confirm this. In the BeneMin trial, at 12 hospitals in the UK, 207 people whose illness had begun within the previous five years took minocycline or a placebo for a year. Negative symptoms did not differ, and the trial found no sign of an inflammatory process for minocycline to treat.2 In a trial of 200 people in Israel and Romania, symptoms and thinking skills did not differ after 16 weeks.5
Pooled across 10 trials with 895 people, minocycline improved negative symptoms moderately.6 But smaller trials reported larger effects, and another analysis found signs that trials with disappointing results had gone unpublished, which makes the benefit look larger than it is.7
In Pakistan, 326 young people at high risk of psychosis took minocycline, omega-3 fatty acids, both or neither for six months. Minocycline did not lower the number who developed psychosis.8
Side effects
Across trials, side effects were about as frequent as on placebo.6 In BeneMin, digestive complaints were reported 19 times on minocycline and 12 times on placebo, and admissions to hospital for worsening symptoms 15 and 10 times.2
Alternatives
Other medicines tested against inflammation in schizophrenia include celecoxib and aspirin (opens in a new tab), statins (opens in a new tab) and N-acetylcysteine (opens in a new tab). None is an established treatment.
What we don’t know yet
- Whether minocycline helps people with schizophrenia who have clear signs of inflammation.
- Whether it changes the loss of synapses in people, not only in the laboratory.
