Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Cognitive therapy for depression

Trial design (first 16 weeks [1])
Results (first 16 weeks)
Trial design (next year [2])
Results (continuation year)

Does talking make medical conditions better? Preconceptions do not prevent us thinking out criteria necessary to test whether talking works. For talking here, read cognitive therapy, and for a condition, depression.

Trials should include active and placebo comparators to demonstrate superiority of talking to pill taking. That ensures sensitivity. One might choose several outcomes, like some form of depression measure, a defined level of improvement, say, plus a higher level of improvement equivalent to being better (or no worse than if your football team lost on Saturday). A long trial would examine what happened when cognitive therapy stopped.

This is what we have in two reports of a complex randomised trial [1, 2]. The reports indicate that cognitive therapy may be better than we have thought.

Trial design (first 16 weeks [1])

The trial recruited 240 patients who were randomised into three groups, 120 treated with antidepressants, 60 with placebo, and 60 with cognitive therapy. Inclusion criteria were diagnosis of major depressive disorder using standard criteria, and age 18-70 years. All patients had scores of 20 or more on a modified 17-item Hamilton depression rating scale on each of two visits separated by at least a week.

Antidepressant therapy used paroxetine, or identical placebo. The dose was set initially at 10-20 mg a day, and raised in increments based on response and adverse events, to a maximum of 50 mg a day by the sixth week of treatment. Patients had weekly treatment sessions with a psychiatrist for the first four weeks, and every other week thereafter. The duration of treatment with placebo was limited to eight weeks because of ethical considerations, though the whole initial part of the study was conducted over 16 weeks.

While psychiatrists conducted antidepressant therapy, therapists, who were mostly post-doctoral psychologists, conducted cognitive therapy. Guidelines were used with the 50 minute sessions twice weekly for the first four weeks, once or twice weekly for the next eight weeks, and once weekly for the final four weeks.

Outcome was Hamilton depression rating score assessed by a blinded observer. The criterion for response was a score of 12 or less for most of the second eight-week period of treatment. Remission was defined as a score of 7 or less over the same period.

Results (first 16 weeks)

Patients in the trial were mostly (60%) women, with an average age of 40 years. Most (90%) had chronic or recurrent major depressive disorders, had depression for almost 20 years, and about 60% had received previous treatment with antidepressant drugs. The initial mean depression rating score was 23. Trial groups were similar.

Over the whole 16 weeks, 16% of patients treated with antidepressants and 15% of those treated with cognitive therapy withdrew from the trial. These were mainly due to adverse events with antidepressants, and dissatisfaction with treatment with cognitive therapy, and most occurred early. The mean paroxetine dose at the end of 16 weeks was 39 mg daily. The placebo pill group had similar withdrawal rates at eight weeks.

At the end of eight weeks response rates were 50% for antidepressants, 43% for cognitive therapy, and 25% for placebo. Both active treatments were significantly better than placebo, but not each other.

At the end of 16 weeks, 58% of patients in both the antidepressant and the cognitive therapy group had a response. Remission rates were 46% for antidepressant and 40% for cognitive therapy (Figure 1). These were not statistically different from one another.

Figure 1: Response and remission at 16 weeks

Trial design (next year [2])

By 16 weeks 104 patients had responded to treatment. After 16 weeks those responding on antidepressant treatment were re-randomised between continuing antidepressant therapy, or placebo, with the change made over a period during which dose was tapered off, and in a blinded fashion. These patients continued seeing their same psychiatrist, usually every two weeks for the first month and every month thereafter for a year. Patients on cognitive therapy had three booster sessions, at least one month apart.

Outcomes were Hamilton depression rating score assessed by a blinded observer weekly for two weeks, every other week for the next six weeks, and then monthly thereafter. Patients had relapsed if they had a score of 14 or greater on consecutive weeks.

Results (continuation year)

There were 34 patients on ongoing antidepressant therapy, 35 on placebo, and 35 who had had cognitive therapy. Absence of relapse was highest at 76% with cognitive therapy at one year, 47% with antidepressant, and 31% with placebo (Figure 2). Both active treatments were better than placebo. The proportion of patients with a sustained response, defined as completing and responding to acute treatment and staying free from relapse across the one year continuation phase, was better for cognitive therapy and antidepressants than for placebo (Figure 2). Rates were 37%, 27% and 16% respectively.

Figure 2: Relapse and sustained response rates over one year of continuation therapy


These are exciting results from a delightfully designed and conducted study of an explanatory and pragmatic nature. The numbers in the continuation phase were small, and that is one reason why one cannot be over-confident about the results, and why statistical testing and calculating NNTs is probably premature.

The results we have show that cognitive therapy appears to be at least as good in major depression as antidepressants, that the effects are long lasting, and are at least as good as ongoing use of antidepressants.

Some folk might argue about costs, but that is also premature. Much depends on how often people with major depressive disorder are expected to see their psychiatrist, because there were no more visits to the cognitive therapist than to psychiatrist, less over the whole of the study period, and CBT has less drug cost. It makes you think, this. Now we need to try and understand which patient with major depression would do better with which type of therapy.


  1. RJ DeRubis et al. Cognitive therapy vs medication in the treatment of moderate to severe depression. Archives of General Psychiatry 2005 62: 409-416.
  2. SD Hollon et al. Prevention of relapse following cognitive therapy vs medication in moderate to severe depression. Archives of General Psychiatry 2005 62: 417-422.

previous or next story