Do painkillers ‘stop Prozac working?’

“Aspirin and other popular painkillers could prevent Prozac from working properly,” the Daily Mail reported. This news story was based on research that was predominantly in mice, examining the effects of combining anti-inflammatory painkillers such as ibuprofen and aspirin with a class of antidepressants called selective serotonin reuptake inhibitors (SSRIs), to which Prozac belongs.

The study found that SSRIs work by increasing the levels of a depression “biomarker” called p11. The anti-inflammatory drugs prevent SSRIs from increasing the levels of this protein. They also blocked the mouse behavioural response to SSRIs, but had no effect on other types of antidepressant.

The researchers followed up their study by looking at medical records of people who had been in a clinical trial and who were receiving the SSRI, citalopram. They found that people who had taken anti-inflammatories were less likely to have improvements in their depression at 12 weeks. However, the human follow-up study can only show an association, and cannot tell us whether the anti-inflammatory drugs caused the SSRIs to be less effective.

This was well-conducted basic research but, at the moment, there is insufficient evidence of its application to humans. The observations from analysing human data suggest that further follow-up examining how anti-inflammatory painkillers alter the effectiveness of SSRI antidepressants is warranted.

Where did the story come from?

The study was carried out by researchers from Rockefeller University New York. Funding was provided by the Skirball Foundation and grants from the United States Army Medical Research Acquisition Activity (USAMRAA) and the National Institutes of Health, Mental Health and Aging. The study was published in the peer-reviewed medical journal Proceedings of the National Academy of Sciences.

The Daily Mail highlighted that this research was in mice with follow-up work using human medical records. There are a couple of points that may lead to misinterpretation. The newspaper said that “many complain that the ‘happy pills’ do nothing to lift their depression and now scientists have worked out why”. However, this study does not resolve the issue of why some people do not respond to antidepressants and there are likely to be many reasons underlying this. The report also said that antidepressants were successful in just 40% of people taking aspirin-like medications, while the research paper put this figure at 45%.

What kind of research was this?

This animal study used mice to investigate the effects of antidepressants when combined with anti-inflammatory medications (painkillers). The researchers followed up their findings by looking at data from a study that followed a cohort of people taking antidepressants to see whether those who took anti-inflammatory medications alongside their antidepressants had different outcomes to those who did not.

The researchers said that chemicals called cytokines, involved in the body’s immune response, may play a role in depression. This theory follows the observation that many patients undergoing certain cytokine treatments develop depressive symptoms, and some cytokines can regulate brain chemicals such as serotonin, which is linked to depression.

The researchers were interested in a protein called p11, which is a biochemical marker of depression. They said that mice that have been genetically modified so that they don’t produce this protein show some symptoms of depression. However, mice that have been genetically modified to produce more p11 show anti-depressant responses in mouse behavioural tests. They said that in tests in rodents, three types of antidepressant treatment (selective serotonin reuptake inhibitors [SSRIs], tricyclic antidepressants [TCAs] and electroconvulsive therapy [ECT]) have been shown to cause an increase in p11 levels in the rodent brain. The researchers wanted to see what effect anti-inflammatory drugs would have on the p11 protein.

What did the research involve?

The researchers treated the mice with the SSRI citalopram (brand name Cipramil) either by itself or alongside the anti-inflammatory drug ibuprofen. They then measured the levels of certain cytokines and p11 in the front area of the mouse brain.

They used genetically modified mice that did not produce the cytokine IFNγ or the cytokine TNFα to see if either of these were necessary for citalopram to increase p11. They also looked at the effect of these cytokines on the level of p11 in normal, non-genetically modified mice by injecting the mice with these cytokines.

The researchers then looked at the effect of three anti-inflammatory drugs (ibuprofen, naproxen or aspirin) and a different class of painkiller (paracetamol) on mouse behaviour after the mice had received one of a variety of antidepressants.

The mice received either:

  • an SSRI – either citalopram or fluoxetine (Prozac)
  • a tricyclic antidepressant (TCA) – either imipramine or desipramine
  • a monoamine oxidase inhibitor (MAOI) called tranylcypromine
  • an “atypical antidepressant” called bupropion

The researchers then analysed data from a trial of antidepressants in humans. The trial, called the “sequenced treatment alternatives to relieve depression (STAR*D)”, looked at data from 1,546 participants taking the antidepressant citalopram. The study also collected records of their symptoms at week 12 and whether they had taken an anti-inflammatory medication during the 12-week period.

What were the basic results?

In the mouse research, giving the mice either aspirin or ibuprofen with the antidepressants citalopram or Prozac blocked the increase of p11 normally seen with these antidepressants. However, aspirin or ibuprofen did not block the increase of p11 caused by a tricyclic antidepressant (desipramine).

The researchers found that two cytokines, IFNγ and TNFα, were regulated by both the antidepressant citalopram and by ibuprofen. They showed that in mice that did not produce either IFNγ or TNFα, citalopram no longer increased p11 levels. Injecting mice with either of these cytokines raised p11.

In the mouse behavioural tests, all types of antidepressant caused the mice to be less hesitant (an antidepressant response). However, giving ibuprofen alongside either of the SSRIs (citalopram or Prozac) reduced their antidepressant effects on the behavioural tests. Ibuprofen had less effect on the behavioural response to tricyclic antidepressants and it had no effect on the response to other types of antidepressants.

All three of the anti-inflammatory painkillers and paracetamol reduced the antidepressant effects of citalopram in mice.

In the human part of the study the researchers found that, of 1,546 participants, 810 were in remission from their depression by 12 weeks. Of these, 182 had taken an anti-inflammatory during the 12 weeks while the other 628 had not. There were 227 participants that were treatment resistant (not in remission) and had taken an anti-inflammatory at least once during the 12 weeks of treatment. The remaining 509 participants who were treatment resistant had not taken any anti-inflammatory drugs during this period.

This meant that of those participants who took an anti-inflammatory, 45% were in remission and 55% were treatment resistant at 12 weeks. Of those participants who did not take an anti-inflammatory 55% were in remission and 45% were treatment resistant. The difference between the remission rates was statistically significant (p=0.0002).

The researchers also looked at people taking painkillers other than anti-inflammatories (such as paracetamol). People taking other painkillers were also less likely to achieve remission (37% in remission) compared with those not taking painkillers (54% in remission, p=0.0002).

How did the researchers interpret the results?

The researchers said that their research shows that anti-inflammatory drugs inhibit the SSRI-antidepressant-induced increases in p11 and antidepressant-like behaviours in rodents.

They say this association is confirmed “in a dataset from a large-scale real-world human study (STAR*D) underscoring the clinical significance of these results”. They say that they are currently trying to understand the mechanisms by which anti-inflammatories and other painkillers have this effect on the SSRI class of antidepressants, but they suggest that doctors should “consider these findings when designing treatment strategies for their patients that include SSRIs”.


This predominantly animal research found that non-steroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen reduce the increase in p11 (a marker for depression) when mice are treated with SSRIs, which are a commonly used class of antidepressant. Furthermore, anti-inflammatories and other painkillers have been shown to reduce antidepressant behaviour in mice.

The researchers followed up this study by looking at patient data. However, although they found some associations between use of anti-inflammatories and other classes of painkiller and a reduced remission rate in patients taking a particular SSRI drug (citalopram), they point out that they cannot say whether the painkillers caused this effect.

They go on to say that to evaluate fully the effects of anti-inflammatories and other painkillers on SSRI antidepressant response in humans would require a prospective double-blind randomised clinical study. This would also allow standardised treatments to be assessed. This is important as the people in the clinical trial from which the data were analysed may have been on different antidepressant treatment regimens and different quantities of painkiller medication or other medications.

The researchers say that they are currently looking at the biological mechanisms underlying the observed effect.

This was well-conducted basic research, and the observations from analysing human data suggest that further follow-up is strongly warranted to investigate the effect of painkillers on the effectiveness of SSRI antidepressants.

NHS Attribution