No benefit for adding antidepressants to CBT in severe depression


In people hospitalized for severe depression, cognitive-behavioural therapy (CBT) was significantly effective, researchers write—and adding antidepressants (AD) didn’t improve outcomes. The researchers note that their finding contradicts the advice offered in clinical practice guidelines.

“The inpatient CBT was effective in depression. The effectiveness of CBT is not improved by the additional use of AD. The current prescribing practices of AD should be questioned,” they write.

The study was led by Reinhard Maß Kerstin Backhaus, Katharina Lohrer, Michael Szelies, and Bodo K. Unkelbach at the Center for Mental Health Marienheide, Germany. It was published in the journal Psychopharmacology.

Their participants came from the Aaron T. Beck inpatient unit at the Center for Mental Health Marienheide. The Beck unit is focused on providing CBT to all patients but also offers pharmacological intervention as needed, following a medical model of psychiatric treatment.

Several years ago, the Beck unit operated following the German guidelines, which recommend combining antidepressants with psychotherapy for severe, chronic, or recurrent depression. However, after a 2019 study revealed that the patients who received antidepressants did no better than those who did not, the unit’s policy changed—only those who specifically requested antidepressants were considered for those drugs. (All patients continue to receive CBT.)

Thus, the researchers were able to conduct this study by comparing outcomes from the previous years (Phase A, in which 60.3% of patients were prescribed antidepressants by the time they left the unit) with more recent outcomes (Phase B, in which only 27.9% of patients were prescribed antidepressants).

Unlike clinical trials, which typically have stringent criteria for participants—often excluding those with suicidal thoughts, those with severe or recurrent depression, and those with other mental health problems—the current study included a much more realistic group of patients. That’s because the participants comprised the actual patients treated in the Beck unit during this time. All the participants had depression severe enough to require hospitalization, with features such as a high risk of suicide, inability to work, and having had ineffective outpatient treatment already. Many patients had comorbid conditions, such as anxiety (about 20%), personality disorders (about 20%), OCD, eating disorders, psychosis, or PTSD (about 16% combined).

Thus, this study tells us what treatment works for real-life patients who are being treated in an inpatient hospital setting.

To measure depression severity, the researchers compared scores on the Beck Depression Inventory (BDI-II). The average score on the BDI was the same for both groups, as was the number of people who “responded” to treatment (their BDI score decreased by at least 50%) and the number of people who “remitted” (BDI score of 12 points or less). This was true both at the time the patients left the inpatient unit, as well as at the six-month follow-up.

In both groups—whether patients received an antidepressant plus CBT or CBT alone—a little more than 70% found the treatment to be effective (responded or remitted by the time they left treatment). After six months, this number dropped somewhat—closer to 50%—but remained similar in both groups.

And again, this is for patients hospitalized with the highest depression severity, suicidality, and other mental health problems at the same time.

The researchers sum up their results:

“At discharge (T2), there were 28.4% non-responders, 16.0% responders, and 55.6% remitted patients in phase A, and 28.3% non-responders, 14.8% responders, and 56.9% remitted patients in phase B (χ2 = 0.239, df = 2, p = 0.887). At follow-up (T3), there were 49.7% non-responders, 8.8% responders, and 41.5% remitted patients in phase A, and 48.8% non-responders, 12.3% responders, and 38.9% remitted patients in phase B (χ2 = 1.945, df = 2, p = 0.378).”

Four patients died by suicide during Phase A. All four were taking antidepressant drugs. No patients died by suicide during Phase B.

These outcomes demonstrate that whether patients received an antidepressant or not, they generally improved with CBT, and most of them maintained that improvement at six months. Some further analyses added preliminary evidence that the CBT-alone condition was slightly better than the combined condition: In the CBT-alone condition, treatment time was shorter, there were fewer dropouts, and the effect size for the decreased depression scores was higher.

“Some superiority of treatment is evident in phase B: (1) The treatments in phase B were 4½ days shorter; (2) in phase A, the proportion of dropout was greater than in phase B; (2) phase B shows somewhat higher effect sizes for the decreases in depression at T2 and T3,” the researchers explain.

The researchers write that their finding is consistent with other studies, including a large 2023 meta-analysis that found CBT alone was just as good as combined therapy.

One limitation of the study may involve the expectation effect: Patients who decide to go to the Aaron T. Beck inpatient unit know that they will receive psychotherapy and are likely to believe that it will help them. This may differ from patients in other units, who may not have that belief. It may also vary by country—it’s possible that patients in the US don’t have as high an opinion of CBT and thus wouldn’t get the boost from having positive expectations. It’s also possible that a hospital inexperienced in CBT might do a worse job at psychotherapy than the Beck unit, which specializes in that approach.

However, the researchers also mention that the slight benefit of antidepressants over placebo seen in clinical trials may be due to the methodological flaws of those studies. They cite the recent finding that only about 15% of patients may experience a unique benefit from antidepressant drugs—and note that even that could be due to the breaking of the blind in drug trials.

“In general, significant methodological problems have been highlighted in AD efficacy trials. A major problem is the breaking of the double-blind condition in RCTs which seriously questions the validity of these studies. Stone et al. pointed out that the 15% proportion of patients who seem to have a pharmacological benefit could also be explained by the effects of functional unblinding,” the researchers write.

CBT alone has been found to be just as effective as antidepressants in the short term and better than drugs in the long term. Moreover, psychotherapy avoids the harmful effects of antidepressant drugs, including sexual dysfunction for up to 88% of those taking them, weight gain and metabolic problems, emotional numbing, and more. Psychotherapy also doesn’t cause withdrawal, which is common after stopping antidepressants and can be long-lasting and severe.

Studies have found that antidepressant drugs may actually worsen outcomes in the long term, even after controlling for the baseline level of depression severity.

And one study found that those with more severe depression, those with comorbid anxiety, and those who were suicidal were least likely to benefit from antidepressant drugs.

The researchers in the current study write that clinical practice guidelines should be updated to reflect the lack of benefit for add-on antidepressant treatment. However, they note that this is unlikely because authors with financial ties to the pharmaceutical industry craft the guidelines.

“The treatment guidelines should therefore be updated accordingly. However, this could be hindered by the fact that many authors of the guideline have financial relationships with the pharmaceutical industry. This fact leads to a mixture of scientific and commercial perspectives and has been repeatedly subjected to severe criticism,” they write.




Maß, R,. Backhaus, K., Lohrer, K., Szelies, M., & Unkelbach, B. K. (2023). No benefit of antidepressants in inpatient treatment of depression. A longitudinal, quasi‑experimental field study. Psychopharmacology. (Link)

Editor’s Note: Part of MITUK’s core mission is to present a scientific critique of the existing paradigm of care. Each week we will be republishing Mad in America’s latest blog on the evidence supporting the need for radical change.

Previous articleMy resignation from Mental Health America
Next articleEveryone’s Afraid of an Angry Woman: Honoring Sinéad O’Connor
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.