select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
A study published in the American Journal of Psychiatry looks at doxycycline use in adolescent psychiatric patients and schizophrenia risk.
Dr Katharina Schmack, Clinical Group Leader and Head of the Neural Circuits and Immunity in Psychosis Laboratory, The Francis Crick Institute, said:
“This is the first study to look at the link between doxycycline and schizophrenia risk in health records of adolescents seeking mental health care. This is an important population to study as these adolescents already face an elevated risk for schizophrenia.
“The results build on existing studies in the general population that link doxycycline treatment with schizophrenia risk, and a related antibiotic, minocycline, has previously been observed to improve symptoms in some individuals. However, while this current study shows an association, it does not demonstrate a causal relationship between doxycycline treatment and schizophrenia risk.
“The researchers have analysed existing health data rather than conducted a randomised controlled trial. Their ‘emulated trial’ approach is elegant but still limited because people prescribed doxycycline will differ from those who are not in many important ways, and those differences may also influence their risk for schizophrenia. And while the effects observed here are statistically significant, the absolute numbers are modest: at 15 years, doxycycline treatment at the highest dose was associated with a reduction of approximately 2.5% of absolute risk, meaning that instead of about 5 out of 100 people now roughly 2-3 out of 100 would develop schizophrenia.
“To further test whether doxycycline protects against schizophrenia, a randomized controlled trial would be needed, but this could be challenging. Schizophrenia risk unfolds over several years, so a clinical trial would need long-term follow-up in a large number of participants, adding to the complexity and cost of such a study. The optimal dose, timing and duration of the doxycycline treatment also remain uncertain.
“Given how challenging it is to design and conduct a randomised controlled trial, it may first be essential to better understand the underlying biological mechanisms. For example, some studies in animals show that the antibiotic minocycline has effects on synaptic pruning – the elimination of connections in the brain during development – which is thought to play a role in schizophrenia. In humans it is impossible to measure processes before schizophrenia develops or to manipulate them outside of a clinical trial, so experimental work in animal model systems is indispensable. Such studies could uncover how brain development, inflammation, and other biological processes shape schizophrenia risk.
“Uncovering clinical associations in studies like this is important because this can direct further biological investigations.”
Dr Dominic Oliver, Postdoctoral Researcher, Department of Psychiatry, University of Oxford, said:
“This is an exciting study by Lång and colleagues suggesting that the antibiotic doxycycline might help prevent the onset of schizophrenia in young people. The strongest evidence for new treatments comes from randomised controlled trials but these are very expensive and can take several years from start to finish. The authors instead used an ‘emulated target trial’ approach, which mimics a randomised trial using existing data from thousands of people prescribed doxycycline and compares them with similar individuals given other antibiotics.
“The results suggest that doxycycline might be protective against schizophrenia. The authors have used large, representative data covering everyone born in Finland between 1987 and 1997, which adds robustness to the findings. The idea is also biologically plausible: doxycycline can cross the blood–brain barrier and has anti-inflammatory effects, and increased inflammation is a feature of schizophrenia and of people at clinical high risk for psychosis.
“There are currently no approved preventive treatments for schizophrenia so this study is promising. However, many other treatments have shown early promise and have ultimately shown not to be effective in large-scale trials so we should interpret these findings with caution. The lack of a consistent dose-response relationship and the absence of similar effects in other forms of psychosis suggest the need for replication. This, in combination with doxycycline being prescribed for two weeks on average 10–15 years before psychosis onset, makes it important to consider alternative explanations for these findings, such as unmeasured differences in people prescribed doxycycline versus other antibiotics.
“Overall, this is an important study that opens the door to new avenues for preventive treatments for psychosis. Confirming these findings in other large datasets (and eventually in randomised controlled trials) is essential before doxycycline can be considered for preventive use in young people at risk of psychosis.”
Prof David Curtis, Honorary Professor, UCL Genetics Institute, University College London (UCL), said:
“The main problem with this study is that, by recruiting people at the time that they were prescribed doxycycline or a different antibiotic, the researchers ended up with two groups which clearly differ in a number of important ways. In particular, the patients prescribed doxycycline entered follow-up at a mean age of 18.8 rather than 15.4, were more likely to be female and had different psychiatric diagnoses as adolescents. Given the fact that there are marked differences between the two groups, even if we attempt to account for known confounders there’s really no way to tell that any difference in risk of developing schizophrenia is due to the doxycycline treatment or for some other reason. From everything we know, I find it quite hard to believe that prescription of doxycycline would reduce the risk of developing schizophrenia.”
‘Doxycycline Use in Adolescent Psychiatric Patients and Risk of Schizophrenia: An Emulated Target Trial’ by Ulla Lång et al. was published in American Journal of Psychiatry at 08:00 UK time on Wednesday 5 November.
DOI: Am J Psychiatry 2025; XX:1–9; doi: 10.1176/appi.ajp.20240958
Declared interests
Dr Dominic Oliver: No conflicts of interest to declare
Prof David Curtis: I have no conflict of interest