Why Do We Continue to Doubt Antidepressants' Efficacy?

Jeffrey A. Lieberman, MD


June 27, 2018

Only in psychiatry would the therapeutic benefits of one of the great pharmacologic breakthroughs in medical history be questioned over a half-century after its introduction to clinical practice. However, a recent meta-analysis in The Lancet[1] illustrates the great lengths that psychiatry must go to justify the utility of what should be a proven, evidence-based, standard-of-care treatment.

Psychiatry has the dubious distinction of being the only medical specialty with an anti-movement that is constantly challenging and undermining the field.

Pathways to a Breakthrough

When the Swiss psychiatrist Roland Kuhn was engaged by the CIBA-Geigy Pharmaceutical Company (now part of Novartis) in 1955 to study a new chemical entity (G22355), he administered it to a series of depressed patients in his clinic with dramatic results. However, the company did not believe him until Robert Boehringer, a member of the CIBA Board and principal of the eponymous company (Boehringer Ingelheim), asked him to treat his wife, who was suffering from depression, with the experimental compound. When she miraculously got better, Boehringer pressed CIBA to market the drug, and thus was born the first antidepressant, imipramine.[2]

Other tricyclic antidepressants (TCAs), which were named after their characteristic biochemical structure, followed with different dose potencies and side-effect profiles but comparable efficacy. The pharmacology of this new class of "wonder" drugs allowed researchers to formulate the biogenic amine hypotheses of depression.[3]

When Gershon and Shopsin reported their seminal findings implicating serotonin in the therapeutic mechanism of action of these medications, companies rushed to develop a selective serotonin uptake inhibitor (SSRI).[4] The first to be synthesized was zimelidine, through the work of Arvid Carlsson at Astra in Sweden (Carlsson won the Nobel Prize in 2000 along with Eric Kandel and Paul Greengard). However, its development was discontinued after cases of Guillain-Barré syndrome were reported. At about the same time, Ray Fuller, David Wong, and Brian Molloy at Lilly developed a compound eventually named fluoxetine, which, after clinical testing and FDA approval, was marketed as Prozac and popularized the treatment of depression.[5]

The ostensible advantage of fluoxetine was its therapeutic efficacy without the noxious side effects of TCAs. Numerous SSRIs followed (fluvoxamine, sertraline, paroxetine), as did pharmacologic variations, including selective norepinephrine uptake inhibitors (reboxetine, mianserin), serotonin-norepinephrine uptake inhibitors (SNRIs; venlafaxine, desvenlafaxine), triple reuptake inhibitors (duloxetine), and other various mechanisms (bupropion, nefazodone).

The medications that followed the TCAs were not so much innovations acting at novel biologic targets with superior therapeutic efficacy as they were refinements in pharmacology, mostly reflected in advances in safety and tolerability. This pattern of incremental improvements is not unusual in the trajectory of drug development; the discovery of a prototype is followed by refinement of its pharmacologic properties and improvement in its safety profile. We have seen this with antipsychotic and anxiolytic drugs as well as with the antidepressants.

Failures in Ethics and Trial Design

Despite their "me too" nature, this plethora of medications that came to market from 1990 to the present should have been hailed as a boon to clinicians and patients. However, several developments occurred that clouded their perceived value.

During this period of prolific development, pharmaceutical companies began aggressive marketing strategies in which they sought alternative indications for their drugs, engaged in promotional activities disguised as educational programs (eg, holding dinner lectures at fancy restaurants), and expanded their relationships with "key opinion leaders." All of this served to foster the potential for conflicts of interest and aroused the suspicion of the government, media, and the public.[6]

In addition, the emergence of clinical research organizations and expanded use of professional clinical trial sites and international sites for phase 3 and 4 studies compromised the integrity of studies, as reflected by rising placebo response rates, failed studies, lack of replication of results, and reduced effect sizes for active treatments.[7] At the same time, studies moved from predominantly inpatient to ambulatory settings, and the patients included in trials became more heterogeneous and shifted toward the milder end of the depression spectrum.[8]

Concurrently, reviews and commentaries by authors with ideologic or organizational biases weighed in to discredit the efficacy of antidepressant medications.[9,10] Psychiatrists, patients, and drug companies contributed to the controversy by the injudicious and broader use of antidepressants in patients, including in adolescents and children.[11,12]

The recent Lancet mega meta-analysis[1] of 552 studies (116,477 patients) was celebrated as demonstrating the efficacy of all classes of antidepressants compared with placebo, and superiority of the efficacy and tolerability of some drugs compared with others. While this ambitious review is to be commended, there is less news there than meets the eye. The overall effect size between drug and placebo is moderate at best, diluted by the aforementioned reasons. Moreover, the active drug comparisons were ad hoc, not systematic or sufficiently rigorous to draw firm conclusions.

Countering the Critics

In a recent Medscape commentary, Dr Nassir Ghaemi understandably questions these findings. However, citing an array of statistics, he comes to the wrong conclusions, a case of not seeing the forest for the trees.

His claims that effect sizes are so small as to be clinically insignificant, and that the older TCAs (particularly amitriptyline) are more effective than SSRIs and newer medications, ignore the broader context and evolution of psychotropic drug development. His former criticism is explained by the previously stated factors that have compromised the integrity of phase 3 and 4 studies in recent years that were previously stated, and his comment that TCAs are superior to newer medications has not been adequately tested.

The only real large-scale, rigorous comparative-effectiveness study conducted was the STAR-D trial,[13] which showed that most antidepressants work more or less comparably. STAR-D researchers advised that if the first drug fails, clinicians should try another and persevere with treatment, using drugs of different pharmacologic classes as monotherapy or in combination.

Dr Ghaemi concludes his commentary by saying, "Our profession seems devoted to believing that antidepressants work. They don't, at least not for [major depressive disorder]."

Why the Skepticism?

It must be unsettling for patients to see doctors argue amongst themselves about issues of such clinical and scientific importance. Physicians take an oath to "first do no harm" and always place the interests of the patient first. Scientists live by the credo of the scientific method, meaning, never take anything on faith; base conclusions on empirical evidence; require replication and verification. So why is the question of whether antidepressants work still controversial?

I believe that the enduring skepticism and distorted views about antidepressant drugs are due to the stigma of mental illness and prejudice toward the medical specialty responsible for its study and care. This historical stigma is perpetuated by lay and professional groups, who oppose the use and deny the efficacy of psychotropic medication for ideologic reasons or organizational biases. Psychiatry has the dubious distinction of being the only medical specialty with an anti-movement that is constantly challenging and undermining the field.[14]

The other source of opposition comes from anti-scientific or anti-medical groups who draw the battle lines along whether medical or psychotherapeutic treatment modalities can or should be used. The former challenge the notion of mental illness and the validity underpinning psychiatric nosology; the latter seek to deny or diminish the evidence that mental disorders have biological bases, and are effectively treated with somatic (medications, brain stimulation) forms of treatment, in favor of psychological explanations and psychotherapeutic approaches.

It is certainly appropriate—indeed, warranted—to require solid evidence for the efficacy and safety of medical treatments to justify their clinical use, but it is prejudicial and disingenuous to keep moving the threshold of proof higher and higher because of dogmatically held views. Doctors should not be fooled by the pharmacologic naysayers, and no patient with major depressive disorder should be denied the effective treatment that can be hugely beneficial for them.

As Mark Twain once said, "There are three kinds of lies: lies, damned lies, and statistics." Don't believe them.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.