Article Text

Download PDFPDF

Cognitive behaviour therapy reduces long term risk of relapse in recurrent major depressive disorder
  1. Zindel V Segal, PhD, CPsych1,
  2. Lucio Bizzini, PhD2,
  3. Guido Bondolfi, MD, PD2
  1. 1Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Canada
  2. 2University Hospitals of Geneva, Department of Psychiatry, Geneva, Switzerland

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

OpenUrlCrossRefPubMedWeb of Science

Q Does adding cognitive behaviour therapy to pharmacotherapy reduce the long term risk of relapse of recurrent major depressive disorder?


Embedded ImageDesign:

Randomised controlled trial.

Embedded ImageAllocation:


Embedded ImageBlinding:

Single blinded (assessor blinded).

Embedded ImageFollow up period:

6 years.

Embedded ImageSetting:

University of Bologna, Italy; time frame not stated.

Embedded ImagePatients:

Forty five outpatients successfully treated with antidepressant drugs (tricyclics or SSRIs) for recurrent major depressive disorder. Excluded were: people with fewer than three prior episodes of depression; previous episode of depression over 2.5 years ago; history of substance abuse, personality disorder, or manic, hypomanic, or cyclothymic symptoms; or active medical comorbidity.

Embedded ImageIntervention:

Pharmacotherapy plus cognitive behaviour treatment (CBT); pharmacotherapy plus clinical management. CBT and clinical management consisted of 10 fortnightly 30 minute sessions. Both groups had antidepressant drugs reduced by 25 mg amitriptyline or equivalent fortnightly until drug free. Tapering off the drug dose was not possible in five participants (three in CBT group, two in clinical management group), who were excluded from the analysis.

Embedded ImageOutcomes:

Relapse (onset of a major depressive episode according to Research Diagnostic Criteria).

Embedded ImagePatient follow up:

88.9% at 6 years.


Cognitive behaviour treatment (CBT) significantly reduced risk of relapse compared with clinical management (absolute risk (AR) for at least one relapse: 8/20 (40%) with CBT v 18/20 (90%) with clinical management; p = 0.001). CBT also significantly delayed relapse compared with clinical management (mean time to relapse 235.0 weeks with CBT v 95.5 weeks with clinical management; p<0.001).


Adding CBT to pharmacotherapy can reduce the risk of relapse of recurrent major depressive disorder over a 6 year period.


The prognosis for major depressive disorder (MDD) suggests an illness course that averages four lifetime episodes of 20 weeks duration each, along with a 16% risk for relapse or recurrence associated with each successive episode.1,2 Routine antidepressant management of MDD targets symptom reduction within the acute episode as its primary goal, and recommends maintenance treatment for prophylaxis. The report by Fava et al suggests that non-pharmacological prevention in MDD is feasible, something that is of wide interest in light of the high rates of drug default, possibly related to a pharmacological tolerance or discontinuation by patients once they begin to feel better. What is innovative about this work is that a preventive intervention is sequenced with pharmacotherapy induced remission, thereby affording a greater number of people protection against relapse/recurrence at just that point when they might be tempted to stop treatment.

Forty five people with MDD were treated to remission with either a tricyclic antidepressant or a selective serotonin reuptake inhibitor over a period of 3–5 months. They were then tapered off their medication and received either CBT (with a lifestyle modification/wellbeing focus) or clinical management. The results of a 6 year follow up indicated a significant difference in the percentage of patients who relapsed (CBT = 8/20; clinical management  = 18/20) and in the number of multiple recurrences (CBT  = 12; clinical management  = 34). The authors ascribe the protective effects of CBT to the targeting of residual anxious or irritable symptoms, often the prodromal signs of a relapse. However, it would be important to distinguish which component of CBT (automatic thoughts/lifestyle modification/wellbeing focus) played the most active role in preventing relapses. While these data nicely complement those from Frank et al’s seminal study of psychotherapeutic prophylaxis following medication withdrawal,3 some caveats do bear mentioning. This paper is one of a series of publications from this group and as yet, there is no replication of this work at an independent site. At present the impetus for alternatives to maintenance pharmacotherapy is strongest on the patient side. Yet, it is unlikely that practice patterns would change in the absence of information on the effect sizes of CBT wellbeing based prevention compared with antidepressant continuation. At best, we have a compelling notion with some strong preliminary data. This additional information would make it easier to consider medication withdrawal and psychological prophylaxis as a defensible deviation from orthodoxy.



  • For correspondence: Dr Giovanni A Fava, Department of Psychology, University of Bologna, viale Berti Pichat 5, 40127 Bologna, Italy

  • Sources of funding: Mental Health Evaluation Project (Instituto Superiore di Sanità), Ministero dell’Università e della Ricerca Scientifica e Tecnologica.