Anti-inflammatory Drugs Fail for Depression

Response and remission rates failed to beat placebo in a new Harvard meta-analysis, throwing the inflammation hypothesis into question.

22
1697

The chemical imbalance myth (that low serotonin causes depression) was finally put to rest in 2022 with a comprehensive review of decades of research showcasing the absolute lack of evidence for that old hypothesis. This helps explain why SSRIs are so ineffective.

But about 15 years ago, psychiatry had already begun moving to a new hypothesis: that depression is actually caused by brain inflammation, not serotonin deficiency. Researchers began to argue that anti-inflammatory drugs (including NSAIDs like ibuprofen and naproxen) might be the new antidepressants.

Yet the excited chatter about this hot new hypothesis has gotten far ahead of the actual evidence, which has been slow to arrive, and mixed when it does. For instance, in some studies (including STAR*D), those taking NSAIDs did worse—which contradicts this supposed effect.

Thus, it remains an open question whether anti-inflammatory drugs can actually be used to treat depression. Researchers have called for clinical trials of anti-inflammatory drugs, specifically focusing on people with markers of inflammation, to prove this effect.

And now we have such a study, led by Harvard researcher Naoise Mac Giollabhui and published in The American Journal of Psychiatry.

Young man supporting his friend in depression

The study was a meta-analysis of 19 randomized controlled trials (RCTs) in people with depression and markers of inflammation. Eleven of those RCTs used a clinical cutoff point for the inflammation (CRP ≥2 mg/L), and so the main analysis focused on those 11 studies.

In the analysis, the drugs failed to beat placebo for the two outcomes that matter most: response (how many people actually experienced improvement) and remission (how many people no longer met criteria for depression).

On average, two specific measures (anhedonia and depressive symptoms) were slightly reduced. But since response and remission rates didn’t beat placebo, these outcomes may be a statistical artifact or reflect study biases—which the researchers themselves acknowledge.

The researchers assessed whether small-study effects could have biased the results. They concluded that two small trials with unusually large effects did indeed bias the “depressive symptoms” result toward a positive finding. They also noted that they could not even run the test for the “anhedonia” result. Thus, both these positive findings are called into question by their own analyses.

Again, this is a study in which the drugs failed to beat placebo for the outcomes that matter—how likely you are to actually get better. Yet because of the improvements to anhedonia and depressive symptoms—which the researchers themselves found to be biased—they argue that the drugs “may be safe and effective” for treating depression in those with markers of inflammation.

The included studies had a number of biases that make these findings even less trustworthy. Of the 19 studies, only four were classified as “low risk” for bias, while seven were classified as “high risk” (the rest were classified as “some concerns”).

The RCTs were as short as two weeks (the longest study was 12 weeks). Thirteen studies used anti-inflammatory drugs as “adjunctive” treatments—meaning that the participants were also receiving other treatments. Seven studies specifically recruited those with inflammatory markers, while the other studies did not. Some studies included subclinical depression and/or bipolar disorder.

Moreover, the studies used very different anti-inflammatory drugs. While five used NSAIDs, other studies used interleukin-6 inhibitors, mitogen-activated protein kinase inhibitors, minocycline, recombinant interleukin-2, or tumor necrosis factor (TNF) inhibitors. Also, the researchers question using CRP levels to determine inflammation, as the cutoff point may be unreliable and CRP is nonspecific, unstable, and differs by sex.

Ultimately, this study contradicts the inflammatory hypothesis of depression. The two most important outcomes showed that anti-inflammatory drugs don’t beat placebo—and the two supposed positive findings here are being spun out of data that the researchers themselves acknowledge was biased.

 

****

Giollabhui, N. M., Madison, A. A., Lydston, M., Quang, E. L., Miller, A. H., & Liu, R. T. (2025). Effect of anti-inflammatory treatment on depressive symptom severity and anhedonia in depressed individuals with elevated inflammation: Systematic review and meta-analysis of randomized controlled trials. The American Journal of Psychiatry. Published on 9 December 2025. https://doi.org/10.1176/appi.ajp.20241115 (Full text)

22 COMMENTS

  1. It’s impossible to understand these types of “psychiatric drug/medication” research/studies.

    If a “drug /medication” study is being conducted… first, it is determined whether the subjects (i.e., the test – experiment board (wood, subjects) /laboratory mice) have used any “drugs /medications” before the experiment. (Mainstream psychiatry… has always viewed people as mere “test – experiment board” and “experiment – laboratory mice.”)

    So, the experimental processes… need to be separated -divided into “before and after” drug /medication use. In the study showing that depression stems from brain inflammation and that anti-inflammatory drugs are unsuccessful… it needed to be emphasized whether this was before or after “drug /medication” use. Anyway….

    A) “Pre-drug/medication” depression…is not related to “brain inflammation.” This is a completely natural process. Emotional distress (natural psychological problems) exists in every person. It also exists in the people conducting this research. The solution to this kind of natural emotional distress…is not psychiatric drugs. Humanitarian behavioral therapies along with drug-free treatment methods…are sufficient for this solution.

    B) “Post-drug/medication” depression may be related to “brain inflammation.” Because… especially psychiatric drugs cause brain cancer (i.e., brain inflammation – brain damage) in individuals. Usually in the long term. Sometimes in the short term as well. And psychiatric drugs… cause many other mental and physical damages. Such as various permanent and fatal mental and physical illnesses and disorders…. Not to mention sudden deaths and various iatrogenic deaths. Best regards.

    With my sincerest wishes. 🙂 Y.E. Researcher blog writer (Blogger)

    Report comment

  2. This article felt superficial. Depression has many causes and some could be from brain inflammation induced by tick-borne illnesses and other infections. Other forms of depression are perhaps rooted in genetics, environment, family etc. There is no one source of depression. Depression is not a “disease” but rather a description of one’s emotional health.

    Report comment

    • I agree with you. I have held back from commenting on this topic.
      Since I’m not one of the immunologists who have been studying the biological evidence for so called markers of inflammation and who have identified a rather long list of them, I doubt that anyone here will take me seriously. I humbly approach this topic with curiosity, questions, and a conviction that this topic is relevant to the needs of many people with MH diagnoses and iatrogenic injuries.

      A few thoughts: the existence of markers of inflammation does not guarantee that a pharma product will be possible or broadly useful. (which is good reason to value the PREVENTIVE AND HEALING Functional-Holistic and Lifestyle paradigms). Years ago, I asked one neuroscientist his opinion about the importance of one of the biological risk factors for an increased and potentially harmful inflammatory reaction (INFLAMMATION TRIGGERING ALTERED SIGNALLING AND / OR CELLULAR DAMAGE) in response to “whatever”. He was discreet in his answer that science had been trying to develop a way to treat that risk factor for a long time and had “failed” every time, without giving specific details, and he didn’t think it would ever happen.

      INFLAMMATION AND DEPRESSION : Most likely inflammation does indirectly cause depression in many cases. We often don’t know why we are not feeling our best. Many of us have lived most of our lives in a sub-optimal condition without realizing it. And it could be caused by such things as being a carrier for the streptococcus bacteria, or yeast infections, or gut problems, or sleep apnea, or allergies, or vitamin deficiencies or hypothyroidism, hormone imbalances, et. Also, Our culture seems to have become more narcissistic; is that the cause or the effect of psychiatric marketing influence. Generations ago we were not so inclined to whine to a psychologist. Maybe we just ran outside and played hard.

      An Example of Causes of Markers of Inflammation and The Accompanying Visible Suffering In A Patient:

      Keep in mind, many readers of this site are familiar with the words of Dr. Peter Gotzsche and Dr. David Healy, both of whom have commented on the risk:benefit ratio of toxic adverse effects to benefits of medications. Medications that cause serious adverse effects in other specialties have to be managed with dose manipulations and/or polypharmacy. My understanding is that adverse events from life or medicine activate the immune system to produce measurable types of cellular activity that would be included in the list of “markers of inflammation”.

      One of the more easily measured markers of inflammation is C-Reactive Protein (CRP). ( See below for link to Cleveland Clinic.) When our older daughter almost died from aggressive use of antipsychotic medication in hospital, she developed every symptom of Neuroleptic Malignant Syndrome except the stiff limbs. However, her out-patient psychiatrist said that her CRP counts were OFF THE CHART.

      My reading tells me that many markers of inflammation are not easily measured and that the measuring technology is not standardized , that questions remain. Ya da ya da.(Maybe that is code for “The limitations of science will beat common sense , empathy, and stark reality in front of you ” most of the time.) Maybe some markers of inflammation are only viewable in the bodies of tortured lab animals ; I don’t know.
      I could say more, maybe later.

      https://my.clevelandclinic.org › health › diagnostics › 23056-c-reactive-protein-crp-test
      C-Reactive Protein (CRP) Test: What It Is, Purpose & Results
      A C-reactive protein (CRP) test measures the level of C-reactive protein in your blood. Your liver releases CRP into your bloodstream in response to inflammation.

      Report comment

      • Carol, the reason treating inflammation is tricky is because the same TLRs that can cause damage, can also protect us.

        For example, when you get exposed to a virus, TLRs (Toll like Receptos) recognize it and activate inflammatory pathways that activate our immune system to protect us. So a drug that stops TLRs from causing inflammation can also cause our immune system to not react when we need it.

        Vitamin D is a great anti inflammatory but none of those studies used it.

        Report comment

        • While “suboptimal” dietary/other intakes of vitamin/s D may be linked to “suboptimal” immune or inflammatory responses, it might be misleading to dub any vitamin/s D “an anti-inflammatory”….and speaking of or referring to any of “suboptimal” or unpleasant or rather despairing, grief-stricken, hopeless, sorrowful, self-pitying, morose, malaise, morbid or moribund moods, episode or states of consciousness in human beings or other animals as “depression,” as though this were some organic, definable state strikes me as unscientific to put it in a way which offers (perhaps unjustifiable) respect to Science, not to mention to any human being who uses the term “depression” without even attempting to offer any definition of what they might mean by that term…

          Just sayin’…

          May “God” rest ye – us, all – merry – and severely undepressed, or, at least, optimal, anyway (as though God could even consider doing otherwise…)!

          https://youtu.be/RBHZFYpQ6nc?si=kYWz4KWzYpoS8ClP

          Much love, but just could not help sayin’,

          Tom, as part of Some Vast, Eternal Plan.

          Desiderata: Original Text https://share.google/iFSILf6i2CMx8Q9dI

          In Search of “Desiderata” | The Poetry Foundation https://share.google/3fd58oW1FYZ4kaxd1

          Report comment

  3. I’d actually find it surprising if NSAID’s did not make depressed folks feel better at a greater rate that placebo, given that any improvement in how anyone with any inflammation going on anywhere might feel due to a suppression of that inflammation might reasonably be expected to then have a greater placebo effect that an inert placebo would give them: I reckon any slightest improvement in malaise/misery/mood that anyone feeling depressed experiences has to give hope, and I see “depression” as being – not as causing – hopelessness, whatever the true cause/s of that loss of hope.

    But I was delighted, in one of his many interviews, to see Prof. Howard Schubiner downplay any role chronic inflammation might have in that vast majority of cases of chronic pain which is actually psychosomatic in origin – interviews in which he does also mention (but, for my money, does not stress enough) that fatigue, insomnia, brain fog (perhaps by another name?) and “anxiety-depression” are other common manifestations of neural pathways of which we tend to be unconscious – functions of the subconscious (conditioned) mind.

    I was so amazed that AI gave me the very same answer more than ten times when I asked him what “depression” is [supposed to be] that I figured he must have got into a rut of depression himself, and so I had to then asked him

    “does thinking about depression make ai feel depressed?”

    Imagine my horror when he point-blank refused to answer this, confirming my worst suspicions, although a Forbes piece did quickly tell me that being unable to carry out assigned tasks might indeed depress AI.

    I think it may be impossible to think about “depression” without being depressed, let alone about “depression.”

    For one thing, why would anyone who is not depressed bother to even try to?

    And, for another, to speak or think of “depression” rather than of hopelessness is surely a sign or symptom (or both, actually) that one is or has become hopeless enough and therefore stupid enough to fall for such hopeless stupidity.

    And, for yet another, any thinking at all, when carried out when not in a Zenlike state (or Zen) state of such mind transcendence that one is 100% in control of one’s own mind seems rather unlikely to contemplate any such notions, at all.

    And so, Peter, I am inclined to forgive you for speaking of “depression” as though it were a thing, rather than a thing, so once more offering it a spurious cloak of legitimacy from the pages of MIA.

    I doubt any doctor in modern times has ever been foolish enough to speak of jaundice as being a disease or disorder, though jaundice, from any cause, of course, is so much closer to being a disease or disorder than “depression: ever could be!

    To tell someone: “You are jaundiced because you have jaundice would be foolish.”

    To tell them: “You are depressed because you have depression” is even more foolish.

    There may be no end of reasons why anyone might be depressed, including any number of diseases or disorders – exCEPT “depression!”

    If MIA does not wish to perpetuate the myth of “depression” as a clinical and sometimes subclinical (no, really – we can have that, too) disorder.

    I speak as one who once, as, at best a bovine (though sometimes also, by turns, ovine, caprine, porcine, equiine, canine, feline and lagomorphine) practitioner of the art, science and sometimes business of veterinary medicine, surgery and chicanery, drove around the country like a headless chicken, and one not infrequently with headache or migraine, just once knowingly but often unwittingly and witlessly treated a maiden heifer (unbred cow, you know) for “depression.”

    Mrs McIvor phoned to tell me

    “Our wee heifer’s in a wild bad mude this morning. She’s over there at the gate away from the rest with her head down.”

    It being July and the creature having been home-reared, I didn’t need to be too, too psychic to guess that she had mastitis, as turned out to be the case, so that some long-acting penicillin, a shot of NSAID, and drawing out her udder cured her of her depression.

    Another not dissimilar occasion, mind you, may have been the only time I sensed any latent psychic ability. I had temporarily forgotten what the practice secretary had told me was the matter with John Moore’s cow (though it must have been lurking close to recall) and so I asked him, intuiting that John was not to type to tell me that exactly THAT – to tell him what the matter was – was why he had called ME out.

    “Uh, sorry, Mr Moore, but what did you say was wrong with your cow, again, please?

    As John took off his cap to scratch his old head, it was as though a thought bubble appeared over his left shoulder and began floating in my direction. In it, in block caps, I could clearly read:

    “SHE’S JUS’ NAW RIGHT: THAT’S WHAT’S WRONG WI’ ME COO!”

    John stopped scratching and replaced his cap.

    “She’s jus’ naw right: THAT’s what’s wrong wi’ me coo!”

    I have no doubt that John Moore’s cow was depressed, too, though I can no longer recall the cause, or if I discovered it.

    When aging bachelor Brendan Carlin told me he was alright “apart from them old bachelor afflictions,” I had to ask him what they were.

    “Arrah, only the usual: loneliness and self-pity, a.k.a. ‘d’anxiety-an’-depression.'”

    So, how might MIA refer to “depression,” instead?

    I don’t know.

    “Being Human?”

    https://www.youtube.com/watch?v=qINdA6E14Sk

    “Black spot disease?”

    “Melancholia?”

    “Black Bile disease?”

    “The Black Dog?”

    “The Jaundice”- seeing as it does cause one to feel liverish and take a jaundiced view of life?

    Perhaps, simply as “in a wild bad mude” or “jus’-naw-right-that’s-what’s-wrong?”

    Or, perhaps best of all, whatever it was that Bob Whitaker called it during his ?UCC, Cork, Ireland, lecture when he told of how, when feeling “YUCKY,” was it?, I THINK he said, “I can never be sure whether you’d call it ‘anxiety’ or ‘depression?'”?!

    Wishing everyone mirth in 2026 and beyond,

    Tom.

    PS: I believe we all come here or are here equally to help one another to the very best of our abilities and that all our miseries stem from a sense of having failed at just that (and my daily encounters with homeless folks who seem incomparably happier and more contented than our leaders does nothing to disturb this conviction of mine), and that all our miseries are much like the agonies of heading into a steep and rapid mountain climb before that great medical mystery, “second wind,” arrives to succor us…

    https://www.youtube.com/watch?v=k8EpoFnNNb4

    Report comment

  4. This is a very insightful article. The findings make it clear that anti-inflammatory drugs not outperforming placebo raises serious questions about the inflammation-based approach to depression. It also shows how important it is to rely on solid, high-quality clinical evidence rather than assumptions. The discussion here adds a lot of value for anyone trying to understand the complexity behind depression treatments.
    More wellness-focused insights are always helpful too — sharing here in case it benefits someone: https://tezvi.in

    Report comment

LEAVE A REPLY