Intended for healthcare professionals


Cognitive behavioural therapy or antidepressants for acute depression?

BMJ 2015; 351 doi: (Published 08 December 2015) Cite this as: BMJ 2015;351:h6315
  1. Mark Sinyor, psychiatrist1, assistant professor2,
  2. Mark Fefergrad, psychiatrist1, assistant professor2,
  3. Ari Zaretsky, psychiatrist-in-chief1, associate professor2
  1. 1Department of Psychiatry, Sunnybrook Health Sciences Centre, Toronto, ON M4N 3M5, Canada
  2. 2Department of Psychiatry, University of Toronto, Canada
  1. Correspondence to: M Sinyor mark.sinyor{at}

Both options look equally effective, although evidence is limited

Depression is an important and underappreciated cause of global morbidity and mortality. It is the second leading cause of years lost to disability, accounting for approximately 10% of all such losses.1 2 More than half of the roughly 1 million people worldwide who die from suicide each year have depression.3 Although reported point prevalence rates for major depressive disorder in primary care are of the order of 10%,4 the literature is unclear about which evidence based treatments are most effective.

In a linked paper (doi:10.1136/bmj.h6019), Amick and colleagues report a systematic review and meta-analysis of studies comparing second generation antidepressants and cognitive behavioural therapy (CBT) for the initial treatment of major depressive disorder.5 They found no substantial differences in short term efficacy and tolerability between the treatments. This is generally consistent with findings from previous meta-analyses, although (uniquely) this one is confined to second generation antidepressants. The authors highlight the fact that, despite its methodological rigor, their review includes only 11 trials and a total of 1511 patients. Their evaluation of remission and response—both key outcomes—are based on data from just three and five trials respectively; the authors judge that existing evidence comparing these two treatments is of low to moderate strength.

Some people will question the absence of a placebo group in these trials. We do have convincing evidence, however, that antidepressants outperform a placebo for patients with moderate or severe depression, such as those included in this review.6 Placebo comparisons are methodologically challenging in trials of psychotherapy and may be ethically questionable in moderate to severe depression, where evidence for the superiority of active treatment is strong. Although the lack of placebo comparators is a limitation, the fact that outcomes after CBT were comparable to outcomes after second generation antidepressants could be interpreted as indirect, circumstantial evidence that CBT works better than a placebo.

More intriguing is the signal from some trials that CBT could be associated with a lower risk of relapse than antidepressants, once treatment stops. This finding, although statistically uncertain, aligns with the orientation of CBT towards training patients to be their own therapists. Given high rates of relapse in major depressive disorder,7 this is a meaningful finding that contributes to an established literature showing CBT to be an effective treatment for preventing recurrence of depression.8 9 Rates of relapse may be similar if antidepressants are continued, but some patients will prefer a finite course of psychotherapy to ongoing drug treatment.

Ultimately, the results of this analysis should be simultaneously reassuring and disappointing for patients and their doctors. Reassuring because if only one of these two treatments is available, affordable, or preferred, choosing that option is unlikely to affect a patient’s chances of improvement, remission, or tolerance of treatment. Disappointing because there is such a limited reliable evidence base to evaluate these basic treatments for a ubiquitous and often devastating condition. One solution is simply to do more trials. The relative paucity of good comparative evidence is in part because psychotherapy trials are more labour intensive for both patients and researchers. It is also because, in general, no one stands to profit substantially from expensive psychotherapy trials with positive outcomes, in contrast to the financial rewards to pharmaceutical companies that often follow positive trials of new drugs.

In their discussion, the authors mention an exciting recent study identifying a potential biomarker that could help to distinguish patients more likely to respond to CBT than to escitalopram.10 Although the study was small, short term, and at high risk of bias, we must hope that bigger and better studies of this and other biomarkers will lead eventually to better targeting of treatments.

In the meantime, policy makers must acknowledge the World Health Organization’s projection that major depressive disorder will be the leading cause of disease burden worldwide by 2030 by taking more meaningful steps towards primary prevention.11 These steps should include efforts to correct social antecedents of major depressive disorder such as poverty and lack of education, along with improved mental health curriculums in schools. Students could be taught basic CBT or other interventions such as mindfulness, with the aim of preventing symptoms rather than relying on treatment once symptoms start. Ultimately, such efforts could decrease the financial, personal, and interpersonal burden of this important illness while helping to free up resources for other mental health services and treatments.


Cite this as: BMJ 2015;351:h6315


  • Research, doi:10.1136/bmj.h6019
  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: MS has received grant support from the American Foundation for Suicide Prevention, the Physicians’ Services Incorporated Foundation, the Dr Brenda Smith Bipolar Fund, and the University of Toronto Department of Psychiatry Excellence Fund; MF receives residuals from two CBT books published by W W Norton and Company; AZ receives residuals from one CBT book published by W W Norton and Company.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription