Can non-pharmacological interventions prevent relapse in adults who have recovered from depression? A systematic review and meta-analysis of randomised controlled trials
Introduction
People who have recovered from an episode of depression (whether spontaneously or following the provision of treatment) are at an increased risk of becoming depressed again in the future. This risk can be as high as 60% for people who have experienced one episode, 70% in those who have had two episodes, and 90% in those who have had three previous episodes (American Psychiatric Association, 2000). These repeated episodes and the associated long-term suffering mean depression produces a large disease burden: the World Health Organisation estimates it to be the leading cause of disability worldwide (WHO, 2009) and is projected to rise to be the second biggest cause of burden of disease by the year 2020 (Murray & Lopez, 1997). Given the extent of the problem it is unsurprising that there are calls for the prevention and treatment of depression to be made a global public health priority (Whiteford et al., 2013).
Intervening after recovery from depression can prevent relapse, but has been the subject of comparatively little research. A meta-analysis of maintenance medication following recovery (Geddes et al., 2003) included only 31 trials, and there has not yet been, as far as we know, a comparable review in the non-pharmacological domain (one that considers randomised trials of all non-pharmacological options following recovery). A recent review looked at psychological therapies delivered to adults aged 18–54 (Biesheuvel-Leliefeld et al., 2015) and included 25 studies. Most reviews prior to this have been limited to one psychological intervention, for example mindfulness-based cognitive therapy (Piet & Hougaard, 2011), continuation cognitive behavioural therapy (, Clark, & Jarrett, 2009) or brief psychological interventions (Rodgers et al., 2012). In comparison, a series of recent reviews undertaken for a major national clinical guideline for depression (NICE, 2009), identified over 1500 trials of acute treatment using antidepressant medication and over 200 trials of acute psychological interventions. This serves to highlight the importance of further work in the area, particularly research that complements the existing evidence for acute treatments, and that extends our focus beyond ‘recovery’, towards prevention of relapse.
Clarity in the characterisation of relapse in depression is a difficult task given the nature of the disorder. There is (as yet) no biomarker to indicate presence of an episode, and symptoms tend to fluctuate over time rather than having dimensional shifts where there is a clear demarcation between illness and normal functioning (Clark and Beck, 1999, Hersen and Beidel, 2011, Ingram, 1998). For this reason, the criteria and level of symptoms needed to signify an episode of depression can vary. Diagnosis is important in indicating when a person may require treatment, but equally a person with symptoms that do not meet criteria for an episode may still have impaired functioning, and is at a greater risk of illness than someone with low or no symptoms (Cuijpers & Smit, 2004).
Taking into account this increased risk of further episodes of depression in those who have brief or incomplete recovery, Frank et al. (1991) identified four clinically important change points. They used the word ‘remission’ to indicate reduction of symptoms to below the diagnostic threshold for depression, and ‘recovery’ to indicate remission that lasted more than 6 months. They then differentiated between ‘relapse’, which was a diagnosable episode of depression occurring during remission, and ‘recurrence’ an episode that followed recovery. Although developed over twenty years ago with the aim of clarifying important points in the course of depression, these terms have not been universally adopted. This may be due to limited dissemination, but some have argued that the distinction between ‘relapse’ and a ‘recurrence’ is not clearly supported by evidence from intervention trials (Richards & Perri, 2010).
Inconsistencies in the terminology used to describe the course of depression present a challenge when attempting to combine data from multiple research studies. Indeed, a recent review of the lifetime approach to depressive disorder promotes the use of distinct terms for relapse and recurrence, yet these seem to be of limited utility, with the authors reverting to the broader ‘relapse/recurrence’ summary term to describe the majority of research (Bockting, Hollon, Jarrett, Kuyken, & Dobson, 2015). In this paper we use the term ‘recovery’ to denote any period where a person no longer meets diagnostic criteria for depression, the term ‘relapse’ to denote “any kind of significant deterioration in depression, which was preceded by clear improvement” (pp. 5 Richards & Perri, 2010). Without strong evidence of a dimensional shift between, say, a relapse at 5 months, and a recurrence at 7 months, it seems prudent to limit the number of categories we use. This also allows the inclusion of a broad range of studies, which may use varying terminology, in which we are able to consistently distinguish between the presence and absence of a depressive episode.
In contrast to medication, which seems to lose efficacy when discontinued (Anderson & Tomenson, 1995), acute phase psychological interventions may prevent relapse even when they are terminated at the point where a person has recovered (Gortner, Gollan, Dobson, & Jacobson, 1998). This is apparent with CBT (Hollon, Stewart, & Strunk, 2006), and there is a growing body of literature supporting the long-term impact of problem solving therapy given during the acute-phase (Nezu & Nezu, 2010). Perhaps these enduring effects are achieved through the amelioration of the causes of risk of relapse, or through the development of knowledge and skills that equip people to offset their risk (Hollon et al., 2006).
After the acute phase of treatment there has been some move to divide the post recovery period into ‘stabilisation’ or ‘continuation’ and ‘maintenance’ phases. In practice however, there has often been little to distinguish between these post-recovery phases, with patients identified at an elevated risk of relapse typically staying on medication for at least a two year period (Geddes et al., 2003, Thase, 2012). It is interventions delivered after the acute phase, when patients are no longer depressed, that are the focus of this paper.
Continued pharmacological treatment reduces relapse following successful pharmacological treatment during an acute episode (Geddes et al., 2003). This is the most common form of post-recovery intervention, but not all patients benefit, side-effects may increase over time, and people who are unwilling to stay on medication forever are likely to relapse when they discontinue it (Anderson & Tomenson, 1995), even if discontinuation is gradual (Viguera, Baldessarini, & Friedberg, 1998). Psychological approaches are often preferred by patients (Priest, Vize, Roberts, Roberts, & Tylee, 1996), and tend to be time-limited. Psychological interventions have been designed for, or in some cases been adapted for use after the acute phase.
One such intervention is mindfulness-based cognitive therapy (MBCT, Segal, Williams, & Teasdale, 2012), perhaps the most widely known psychological intervention developed specifically to prevent relapse. It was designed to be delivered after a person has recovered from depression, and can follow spontaneous recovery or acute treatment. It typically delivered in groups of 12 to 15 participants with two professionals in fortnightly sessions over three months. MBCT encourages people to process experience without judgement through mindfulness and meditation techniques, thereby helping to change their relationship with troubling thoughts and feelings (Segal et al., 2012). A meta-analysis of MBCT trials (Piet & Hougaard, 2011) suggests that it is effective (with a 34% reduction in risk of relapse compared to controls), particularly for patients who have had three of more previous episodes of depression (where the reduction was 43%).
Cognitive behavioural therapy (CBT) may also work through equipping people with specific skills to reduce their likelihood of relapse. As an acute treatment it aims to modify thoughts and behaviours in a way that alleviates key features of depression, such as identifying and challenging negative automatic thoughts and overcoming avoidance and reduced activity (Beck, 2011). There have been variations of CBT adapted to target relapse prevention, either as a continuation of prior acute treatment or as a standalone intervention. In contrast to MBCT where the delivery of the intervention follows a common format, the duration, mode of delivery (individual, group, self-help), and frequency of CBT has varied. Vittengl, Clark, Dunn, and Jarrett (2007) reviewed the effectiveness of CBT to prevent relapse, both from studies which terminate treatment at recovery, and studies which continued CBT after this point. They estimate that continuing with CBT reduces risk of relapse by 14% compared with discontinuing. Although this estimate of effect is lower than in the aforementioned MBCT meta-analysis, there are key differences in the acute treatment received and the length of included trials, as well as the treatment of missing data by reviewers.
Interpersonal psychotherapy (IPT) (Klerman & Weissman, 1994) can also be used after recovery with reduced frequency of sessions. The model of delivery and objectives of IPT remain essentially the same as in the acute phase, focussing on strategies to deal with interpersonal and social role problems which are held to be casual in the development and maintenance of depression (Klerman & Weissman, 1994). IPT ‘booster sessions’ may reduce relapse by providing continued support after the period where contact with a psychologist would typically end. As part of a wider review, Cuijpers et al. (2011) identified four studies which compared combination of maintenance IPT and pill placebo with pill placebo alone. Maintenance IPT was more protective against relapse than pill placebo.
Other non-pharmacological approaches have sought to prevent relapse by continuing to provide enhanced monitoring and care-planning regardless of whether patients continue to be symptomatic (Gilbody et al., 2006, Smit et al., 2007). Similarly there has been some research into the impact of continued monitoring of symptoms and coordination of care post-recovery (Unützer et al., 2002). Sequencing of interventions and support, as well as long-term management of care from different providers may have important effects (Wells et al., 2000) as well as interventions used in routine clinical practice such as exercise, symptom monitoring, and building social support. There are a broad range of intervention options for those who have recovered from depression, and a review of all randomised trials of these interventions may help to highlight what these are and whether they are effective.
This review will gather evidence about non-pharmacological interventions to prevent relapses in those who have recovered from depression. We firstly aim to describe the research that exists in this field, and highlight any potential gaps in evidence. We then aim to examine the efficacy of these interventions.
Section snippets
Method
We conducted a systematic review and meta-analysis following a pre-specified protocol (Clarke, Pilling, Kenny, & Mayo-Wilson, 2011).
Trial flow
We identified 20,531 potentially relevant citations and retrieved 389 papers for full-text review (Fig. 1). Of these, 29 trials were eligible for this review, and 22 trials were included in the meta-analysis. Seven of the trials either did not report or provide usable data on request, or did not have a control group that fitted into our definition.
Study characteristics
The 29 eligible trials were conducted in the USA (n = 12), the UK (n = 6), the Netherlands (n = 3), Italy (n = 2), Canada, Denmark, Germany, Sweden and
Findings
This review suggests that psychological interventions (specifically MBCT, CBT and continuation IPT) delivered when an adult has recovered from an episode of depression may reduce the risk of experiencing another episode within the following year by 22%. The baseline risk of relapse varied greatly between studies (for 12-month data, range = 0.2–0.79, IQR = 0.41–0.63), but as an indication of the clinical utility of these interventions, we estimate a number-needed-to-treat (NNT) of 8.33 for the study
Conclusion
Psychological therapies delivered following recovery from depression, including CBT, MBCT, and IPT, may reduce risk of relapse after one year by 22%. For CBT this effect is maintained after two years. This is similar to that previously reported for continuing medication, but without the need for continued treatment throughout the follow-up period. The difference in effectiveness between interventions and the mechanisms by which these interventions exert their effect are not well understood, nor
Role of funding sources
This study was supported by a grant from the British Psychological Society to SP (grant number 509517). The funding source had no involvement in the study design, collection, analysis, or interpretation of data, writing the manuscript, and the decision to submit the manuscript for publication.
Contributors
KC, SP and EMW designed the study and wrote the protocol. KC and JK conducted literature searches and extracted data. KC conducted the statistical analysis and all authors contributed to interpretation of results. KC and SP wrote the manuscript and all authors have revised and approved the final manuscript.
Conflict of interest statement
SP also receives funding from NICE for the production of clinical guidelines. KC, EM-W and JK have no financial or other conflicts of interest to declare.
Acknowledgements
The authors thank Katie Greenfield for assisting with sifting references, data extraction and management, Laura Gibbon for assessing risk of bias, and Rob Saunders and Ellie Bishop for commenting on revisions of the manuscript.
References (80)
- et al.
GRADE guidelines: 3. Rating the quality of evidence
Journal of Clinical Epidemiology
(2011) - et al.
Effectiveness of psychological interventions in preventing recurrence of depressive disorder: Meta-analysis and meta-regression
Journal of Affective Disorders
(2015) - et al.
Depression relapse prophylaxis with mindfulness-based cognitive therapy: Replication and extension in the Swiss health care system
Journal of Affective Disorders
(2010) - et al.
Relapse prevention with antidepressant drug treatment in depressive disorders: A systematic review
Lancet
(2003) - et al.
The effects of mindfulness-based cognitive therapy on recurrence of depressive episodes, mental health and quality of life: A randomized controlled study
Behaviour Research and Therapy
(2010) - et al.
Two-year outcome of internet-based relapse prevention for partially remitted depression
Behaviour Research and Therapy
(2013) - et al.
Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study
The Lancet
(1997) - et al.
The effect of mindfulness-based cognitive therapy for prevention of relapse in recurrent major depressive disorder: A systematic review and meta-analysis
Clinical Psychology Review
(2011) - et al.
Treatment of 70 +-year-olds with recurrent major depression: Excellent short-term but brittle long-term response
The American Journal of Geriatric Psychiatry
(1999) - et al.
Global burden of disease attributable to mental and substance use disorders: Findings from the Global Burden of Disease Study 2010
Lancet
(2013)