New Data Show Lack of Efficacy for Antidepressants

Existing evidence for antidepressant effectiveness is minimal and at a high risk for bias


An article published this month in the journal BMC Psychiatry suggests that there is a lack of efficacy for SSRIs and that they significantly increase the risk of serious side effects.

SSRIs—Selective serotonin reuptake inhibitors—are the most commonly prescribed antidepressant medications, including such brand names as Prozac, Lexapro, Celexa, and Zoloft. A recent study found that 13% of the US population was being prescribed an antidepressant in 2012. According to a 2015 study, more than two-thirds (69%) of those prescribed an antidepressant do not actually meet the criteria for the diagnosis of major depressive disorder.

The current study involved the analysis of 131 randomized controlled trials—the gold standard for research study design—with a total of 27,422 participants. The authors were led by Janus Christian Jakobsen, MD, PhD, at the Centre for Clinical Intervention Research at Copenhagen University Hospital. The goal of the study was to determine the efficacy of SSRIs for depression, based on the largest statistical analysis possible. According to the authors:

“SSRIs might have statistically significant effects on depressive symptoms, but all trials were at high risk of bias and the clinical significance seems questionable. SSRIs significantly increase the risk of both serious and non-serious adverse events. The potential small beneficial effects seem to be outweighed by harmful effects.”

Photo Credit: Flickr

The authors make clear that when taken together, the studies do show a small benefit for antidepressant treatment. For trials that examined depression that was mild to moderate in severity, the benefit was just 1.29 points on the 53-point Hamilton Depression Rating Scale (HDRS). The difference for trials that studied severe depression was 2.69 points on the HDRS. Previous researchers suggested that a 3-point difference corresponds to “no clinical change”—that is, neither a doctor nor a patient would notice that change. Other researchers showed that at least a 7-point difference was necessary for “minimal improvement.” However, because of the large sample size, the 2.69-point increase was still considered “statistically significant” despite its lack of real-world significance.

These results add to a growing body of literature questioning the efficacy of antidepressant medications. One analysis in 2010 found that the benefit of antidepressant medications “may be minimal or nonexistent, on average, in patients with mild or moderate symptoms.” Although, the researchers suggested that for “very severe depression, the benefit of medications over placebo is substantial.”

However, researchers in 2008 found that although antidepressants appear more efficacious for severe depression, the difference is “relatively small even for severely depressed patients.” The researchers go on to state that this benefit is attributable to the placebo effect becoming less powerful for patients with severe depression, rather than an increase in the efficacy of the drug.

The dangers of antidepressant medications have also come into the spotlight. A 2016 article in Scientific American publicized the results of a Nordic Cochrane study which suggested that antidepressant treatment was causing increases in suicidality and aggression. The most common side effects are sexual dysfunctions (which can persist for years after stopping the medication), but a significant number of side effects have been observed, some of which are deadly. Withdrawal effects are also a problem—numerous physical and mental problems can develop upon discontinuing an antidepressant medication. Mad in America recently presented an overview of the large body of evidence on this topic.

Finally, there is the danger that antidepressant treatment potentially worsens outcomes. An NIMH study in 2006 found that 85% of untreated depressed patients recovered spontaneously within one year, while a 2010 article documented that out of 4,041 participants taking antidepressants, in an NIMH study, only 108 patients remitted and remained in the trial for the 1-year follow-up. A 1994 article showcased the evidence of “the likelihood that psychotropic drugs actually worsen, at least in some cases, the progression of the illness which they are supposed to treat.”

Dr. Jakobsen and his colleagues found other troubling concerns in their current analysis of the studies of SSRI efficacy. For instance, they found that every single one of the 131 studies they examined had a high risk of bias in study design and publication.

Studies consistently omitted critical data about outcomes and side effects. In all but 4 of the 131 cases, it was likely that patients and researchers were effectively unblinded—able to determine (because of side effects) whether the subject was taking a placebo or an actual SSRI. Placebo control groups rely on blinding to be effective since the placebo effect relies on the subject’s belief in the efficacy of the drug. If the blinding is ineffective, the drug will appear superior to the placebo group simply due to this design flaw.

Additionally, the researchers theorize that publication bias could also account for the statistically significant result. Publication bias refers to the tendency for only positive results to be reported in published articles. This occurs for several reasons: journals want to publish positive results, and so do study authors. The pharmaceutical companies funding industry research also rely on positive results for public perception of their drugs as well as FDA approval. In fact, when researchers in 2008 examined both published and unpublished studies, they found that 49% of the total studies had negative results. That is, almost exactly half of the studies that were conducted found no beneficial effect for antidepressants.

The rating scales used in the studies are themselves concerning. The HDRS, although commonly used, has been criticized by researchers for over-diagnosing depression. For instance, it is possible for an HDRS score to increase by 2 points when a participant does not think they have depression. You can also get 2 points for having difficulty falling back asleep after getting out of bed in the morning. The researchers also note that only 6 of the 131 studies utilized any quality-of-life measure, and they used different methods for measuring this, making the results essentially unanalyzable. The researchers note that quality of life is a highly relevant metric for the effectiveness of treatment.

Even in the best-case scenario, the evidence suggests that improvements in depression due to SSRI use are not detectable in the real world. Given the high risk of biased study design, publication bias, and concerns about the validity of the rating scales, the evidence suggests that the effects of SSRIs are even more limited. According to this growing body of research, antidepressant medications may be no better than sugar pills—and they have far more dangerous side effects.



Jakobsen, J. C., Katakam, K. K., Schou, A., Hellmuth S. G., Stallknecht, S. E., Leth-Møller, K…. Gluud, C. (2017). Selective serotonin reuptake inhibitors versus placebo in patients with major depressive disorder. A systematic review with meta-analysis and Trial Sequential Analysis. BMC Psychiatry, 17(58). doi: 10.1186/s12888-016-1173-2 (Abstract)

Previous articleTransformations to Liberation
Next articleSchool, Community and Online Resources
Peter Simons
Peter Simons was an academic researcher in psychology. Now, as a science writer, he tries to provide the layperson with a view into the sometimes inscrutable world of psychiatric research. As an editor for blogs and personal stories at Mad in America, he prizes the accounts of those with lived experience of the psychiatric system and shares alternatives to the biomedical model.


  1. How about when people go for treatment for alcohol dependency and get put on SSRIs, its usually the first thing they do after detox.

    Hi, we are the psychiatric industry, we know from statistics about 80% of you are going to drink heavily again so here is a prescription for pills that are dangerous to mix with lots of alcohol.

    Report comment

  2. This is what happens when you try to stretch the parameters for a particular class of drugs uses. Initially, SSRI’s were used for OCD symptoms, for which they worked. Then, someone got the bright idea to use them for depressions featuring compulsive and obsessive thinking, which is where we are now. Perhaps the notion for using them after leaving alcohol rehab is the same mentality in a different setting, one more aptly served by having alcoholics using niacin or inositol nicotinate (no flush niacin) for the same purposes.

    Report comment

  3. Peter, I’m glad you mentioned the proposed explanation for the supposed efficacy of the drugs in severe depression. That’s a tired refrain: “…but they do work in cases of very severe depression.” No, they don’t.

    Here’s the HAM-D. It captures a lot of anxiety-related issues, most of which should be expected to diminish as the trial progresses. Each item has its own scale, from 0 to 2, or from 0 to 4.

    1. Depressed mood

    2. Feelings of guilt

    3. Suicide

    4. Insomnia early [falling asleep]

    5. Insomnia middle [after falling asleep, wakefulness or restlessness during the normal sleeping hours]

    6. Insomnia late [early morning awakenings, and unable to go to sleep after getting out of bed.*]

    7. Work and activities [no difficulty, difficult or no longer working]

    8. Psychomotor retardation [moves slowly]

    9. Agitation [0=none, 1=fidgets, 2=plays wth hands, hair, 3=moving about, can’t sit still, 4: hand-wringing, nail-biting, hair pulling, biting of lips]

    10. Anxiety [psychological] [0-no apparent difficulty, 1=subjective tension and irritability, 2=worrying about minor matters, 3=apprehension apparent in face or speech, 4. Fears=expressed without being asked]

    11. Anxiety somatic [Anxiety with physiological involvement. (It says to avoid asking about symptoms that could be side effects of drugs, like dry mouth.)]

    Report comment

  4. Sigh. The only real long-term benefit I received from “antidepressants” is the obliteration of my sex drive. Thanks to the bigotry I have experienced from my “bipolar” label I’m still a virgin at 43. (My parents told me I would never marry after my diagnosis at age 20. I was too damaged for a decent man to want me.)

    If you can’t eat, it’s better not to feel hunger. Now if I could only be happy living in isolation so I wouldn’t have to deal with social rejection of any kind.

    Report comment

  5. I have been diagnosed with major depression, social anxiety, generalized anxiety, and OCD. When I read these articles rarely do they mention spectrum disorders, or any reference is vague. To say someone has major depression, suggests they only have one issue, and not several that fall on a scale, or a spectrum. So when I took Zoloft for the first time at age thirty-five it allowed me to enjoy reading, and I could read a book from start to finish for the first time. I also developed quit a sense of humour, as the depression symptoms lifted. The disadvantage was I started to put on weight.

    Report comment

  6. Is there mention in the actual medical literature yet of the massive misdiagnosing of the adverse effects of the antidepressants as “bipolar”?
    The “childhood bipolar epidemic” is an example of malpractice on a massive scale, malpractice even according to the DSM-IV-TR.

    And what is the benefit to humanity as a whole, of the medical community raping our entire country with higher medical costs, so they may have malpractice insurance, if the psychiatrists don’t utilize their malpractice insurance for what it was intended? Unless, the reason the doctors have malpractice insurance is only so doctors can get trust that they don’t actually deserve? Obviously, the psychiatrists made a Faustian deal with the legal community to not take such cases.

    Report comment

  7. Since all antidepressants are tested against a placebo group that is actually in total cold turkey withdrawal from their previous antidepressant, we have no idea of their efficacy. Even with this unethical design , the placebo group reaches the same level as the drug group just a few days later, even in heavily industry sponsored trials.

    Why risk all the side effects, including possibly a 60x increased risk of death by suicide, and getting hooked on a life shortening drug when you could have waited just a few days and got just as good result without drug (actually even when you are in cold turkey withdrawal)

    The reason the drugs seem to work a bit in very depressed patients is that these patients most probably had heavier doses and longer use than the others and therefore had heavier cold turkey reactions.

    Report comment

    • Good question Kjetil!

      I’m not doing the cold turkey thing, because after over 22 years on “antidepressants” I have become so physically dependent on these unnatural chemicals, if I did anything but a painfully slow taper it would send my system into shock. I could become violent and incoherent like a rabid animal. It might even kill me. This occasionally happens to people with a dependency like mine.

      Report comment

  8. Although I trained in psychiatry it was just at the beginning of the chlorpromazine induced drugs era and I discovered that the vast majority of the patients improved just by talking with them about their problems.
    Unfortunately though, psychiatry , despite a lack of scientific evidence, wanted to think of mental disorders as physical rather than mental and have created vast industries that have hurt countless numbers of people.
    There are many studies that now confirm this, but in our mechanized, dehumanized world this is ignored in favour of profits. We need to bring back human understanding.
    Robert Whitaker has shown us the history of psychiatry, so that we can learn from it.
    Eliot Valenstein has shown us the bad science that has promoted this disaster.

    Report comment