Treatment-resistant Depressed Patients Show a Good Response to Mindfulness-based Cognitive Therapy

The emotional, social and economic burden of depression for sufferers, their families and society is significant, with 12 month prevalence rates estimated at 2.9–12.6% and lifetime risk estimated at 17–19% (Kessler et al., 1994). The fact that depression is often a chronic relapsing condition, with relapse rates of 50–80% in those who have been depressed before (Judd, 1997) has contributed to the WHO prediction that, by 2020, depression will be the second biggest contributor to ill-health burden world-wide (Murray & Lopez, 1998). Individual CBT has been shown to be effective at treating acute depression and reducing relapse (DeRubeis et al., 2005; Hollon et al., 2005) but waiting lists for individual therapy are lengthy in most healthcare settings.

Even treatment by normally “effective” means leaves a substantial minority failing to meet criteria for remission. In a trial to examine relapse risk in the 12 months following antidepressant versus cognitive treatment, Hollon et al. (2005) found that, of the patients recruited for the first (acute treatment) phase, only 58% of the ADM group and the same proportion of the CT group met predefined criteria for recovery and were able to proceed to the relapse phase of the study. The fact that around 40% did not meet such criteria following an adequate “dose” of treatment show that there is still a great deal to do to help those patients who are often called “treatment resistant.”

Moreover, depression has another feature that is a considerable cause for concern: the development of a chronic clinical course, which also resists treatment. In these cases, patients report continuing symptoms of depression and accompanying distress about these symptoms. Some 15–39% of cases still meet criteria for Major Depressive Disorder (MDD) 1 year after symptom onset (Berti Ceroni, Neri, & Pezzoli, 1984; Van Valkenburg, Akiskal, Puzantian, & Rosenthal, 1984), and 22% of cases may continue to do so up to 2 years later (Keller, Lavori, Lewis, & Klerman, 1983). Of particular concern is the risk of suicidal behaviour in such patients. One in seven patients hospitalised for major depression die by suicide (Powell, Geddes, Deeks, Goldacre, & Hawton, 2000) and the Population Attributable Ratio (PAR) for depression in serious but non-fatal suicidal behaviour (that proportion of suicidal behaviour that would be removed if depression were taken out of the picture) is 80 per cent (Beautrais et al., 1996). If suicidal ideation occurs during one episode of depression, it tends to recur in later episodes (Williams, Crane, Barnhofer, van der Does & Segal, 2006), making the question of how best to treat recurrent and persistent depression particularly urgent for such at-risk patients.

What is it that keeps people depressed? Nolen-Hoeksema's Responses Styles Theory (1991) suggests that people who engage in “repetitive and passive thinking about one's symptoms of depression” tend to prolong the very symptoms they are trying to reduce. Ruminators often hold positive (but erroneous) beliefs that it will help, not realising that they are reducing their capacity to effectively problem-solve (Watkins & Moulds, 2005). Evidence suggests that the cognitive processes that increase vulnerability to future episodes are the same as those that maintain depression. These processes are depressive rumination (Jacobson, Martell & Dimidjian, 2001; Nolen-Hoeksema, 1991, 2000; Watkins & Teasdale, 2004) and high cognitive reactivity to mood shifts, where the experience of low mood more easily triggers negative thinking in previously depressed patients (Segal, Gemar, & Williams, 1999; Teasdale, 1999).

Mindfulness-based Cognitive Therapy (MBCT) arose out of an investigation into the cognitive processes that render depressed individuals vulnerable to repeated relapse and recurrence (Teasdale, Segal, & Williams, 1995) with a view to offering a programme that would target those cognitive vulnerabilities (Segal, Williams, & Teasdale, 2002). The intervention was designed as a class-based intervention to increase accessibility to effective relapse prevention. It incorporates, as a central component, mindfulness training as developed by Kabat-Zinn and his colleagues at the University of Massachusetts Medical Center (Kabat-Zinn, 1990), and adds components of cognitive behaviour therapy for depression (Beck, Rush, Shaw, & Emery, 1979). Two studies have demonstrated this group programme's efficacy in reducing relapse rates of depression at 12 month follow-up compared to treatment as usual (Ma & Teasdale, 2004; Teasdale et al., 2000).

MBCT teaches participants to observe their thoughts and feelings through the repeated practise of intentionally returning attention to a neutral object (e.g. the breath or body sensations) in the present moment. Participants are taught how to cultivate direct experiential awareness, together with an attitude of non-judgmental acceptance, towards whatever is present (including sad mood, which in previously depressed patients is likely to trigger patterns of global negative self-referent thinking). The cultivation of awareness during mindfulness practise enables patients to see more clearly when negative and ruminative responses are being triggered, and allows them to decentre from such patterns of thought, seeing them as mental events, rather than necessarily valid reflections of reality. Unlike standard CBT, where the focus is on changing the content of thoughts, MBCT's focus is on fostering meta-cognitive awareness and the modification of meta-cognitive processes that maintain unhelpful reactive or ruminative mind states. So although MBCT was designed for patients in remission, the above analysis suggests that it might be helpful for symptomatic patients who remain caught in such unhelpful thought processes.

The present article describes a clinical audit of a consecutive series of patients referred for MBCT, focussing on those who had continuing symptoms of depression despite treatment with antidepressant medication (ADM) or CBT or both. The two randomised controlled studies conducted on MBCT had shown that those with three or more episodes had a significant reduction in relapse rates but that those with two or less had a non-significant trend towards an increased risk of relapse in the year of follow-up (Ma & Teasdale, 2004; Teasdale et al., 2000). The inclusion criteria for the MBCT programme therefore retained the three-or-more-episodes feature for those who had been recurrently depressed, but was expanded to include patients whose mood disorder was running a chronic course, where it was hypothesised the depression was being maintained by a similar ruminative cognitive style.

The audit was conducted to examine the effects of this approach with this group, the acceptability of the class format, and the feasibility of including both recovered and symptomatic patients in the same classes. It was hypothesised that the class format, with a mix of recovered and symptomatic patients, would be acceptable as the meta-cognitive approach taken means there is no need to work with the unique content of each individual's automatic thoughts, whether currently negative in tone or not.

There are compelling reasons to wish to check out the possible relevance of MBCT for patients who remain in episode. The developers of MBCT have been very cautious about using MBCT for such patients. “It is important to note that MBCT was specifically developed for remitted patients and is unlikely to be effective in the treatment of acute depression, where factors such as difficulty in concentrating and the intensity of negative thinking may preclude acquisition of the attentional control skills central to the programme” (Teasdale et al., 2000, p. 622). However, since a considerable proportion of the patients in this study had already received CBT, we wished to examine patients' responses to an approach that might represent a back-up procedure if the currently available empirically supported treatments (ESTs) such as CBT and anti-depressant medication fail.

Finally, we were particularly interested in those patients who had suicidal ideation as part of their symptoms of depression. We wished to examine the hypothesis that such patients might respond significantly less well to MBCT. Such patients are known to be highly reactive to depressed mood and to react to small mood shifts with impaired problem solving (Williams, Barnhofer, Crane, & Beck, 2005), and we were concerned that this pattern might undermine the ability or motivation to practise mindfulness.

Of the 50 patients audited, 49 completed the programme (99%), and pre and post questionnaires were completed by 46 of these completers (94%).

The results of this preliminary audit suggest that MBCT is an acceptable treatment for patients who have only had a partial response to antidepressant medication and/or standard individual CBT. Further, for many patients it appears to be effective in significantly reducing levels of depression, even in those who start with a more severe pattern including suicidal depression. Although the numbers with bipolar depression are small (n=5), it appeared to have an effect in these patients also. Finally, MBCT proved to be an acceptable and useful treatment for one in-patient who was hospitalised due to a worsening of agitated depression just after the course began (BDI falling from 40 to 19).

The overall pre-post Effect Size was large, though smaller than that which would have been expected for pre-post CBT for naïve patients (those who have never had CBT before). In this case, we might expect a pre-post Effect Size of 2.2 (Ost, 2006). It is of note, however, that the proportion achieving remission was equivalent to a recent trial of 16 weeks of CBT (58%; DeRubeis et al., 2005), and that, for the more severe patients, the pre-post Effect Size (1.8) was of the same order of magnitude for individual CBT for naïve patients. Furthermore, in Baer's review of Mindfulness-based Interventions (Baer, 2003), the average Effect Size was 0.71 for pre-post change across a range of physical and psychiatric disorders, and 0.90 where depression was included as a dependant measure. The outcomes in the present study were comparable to these previous reports.

A number of methodological limitations need to be considered here. Firstly, the fact that there is no control group means that the benefits (and setbacks) seen in the participants cannot be conclusively assigned to the treatment. Secondly, these were patients whose treatment had not proved as effective as either they or their therapists had wished; it is therefore possible they were a highly motivated group. This means that MBCT may still not be effective as a front-line treatment for depression. Thirdly, there was no systematic data collected for these patients on other Axis I and Axis II disorders, their response to past treatments or their openness to non-traditional treatments that, in other settings, may be associated with effectiveness of pharmacological and psychological treatment.

Finally, four patients became worse in the BDI scores, and although they claimed that the programme was important to them, and that the rise in BDI scores was due to their normal mood fluctuations, further research needs to take account of the possibility of adverse effects. Overall though, the concern of Teasdale et al. (2000) that MBCT was unlikely to be effective in the treatment of acute depression was not borne out, even though 98% of patients identified problems with concentration and energy at the outset of the study. In addition, the argument that the sedentary focus of many of the practices (i.e. sitting meditation) may be de-activating and thus counter-productive for such patients again was not borne out. While, at times, some flexibility about the length of the meditation practices was required, participants were expected to and were mostly able to manage the 45 min as long as they learned to bring a kindly and non-judgemental attitude to this, which was repeatedly emphasised by the instructor. The yoga component of the course may also have acted to counteract the cycle of inactivity or agitation frequently encountered in depression.

If these results can be replicated, they may be seen to be consistent with the theoretical changes taking place within and outside cognitive theory and therapy that lay greater emphasis on the processes underlying cognition rather than its content (Hayes, Follette, & Linehan, 2004). The findings are similar to those in another study which examined the effects of mindfulness meditation on depression (Ramel, Goldin, Carmona, & McQuaid, 2004), and fit with the hypothesis that the processes (increased cognitive reactivity and ruminative responses), which play a role in causing relapse of depression, also play a role in the maintenance of depression once it has been established. Targeting these processes with an intervention like MBCT may then have an adjunctive role where standard treatments have failed to bring about a full therapeutic response.

In summary, we found that MBCT was associated with reductions in depression symptoms, with a large pre-post Effect Size (d=1.04). Even given the amount of time and commitment involved, the intervention had a high degree of acceptability and low dropout rate, despite the levels of severity as well as the disabling effects of numerous prior episodes of depression. Further research into the use of MBCT in the active phase of treatment of recurrent or chronic depression is indicated, given the promising results described, and the potential economic advantages of being able to offer this in a group format to a mix of both symptomatic and recovered patients.

The authors wish to thank the patients who participated in this research. Thanks also go to Queen Elizabeth Hospital staff, Dorothy Thompson for the data processing, and Jeff Martin for help with the graphics.

Mark Williams is supported by Wellcome Trust GR06779 (University of Oxford).