Specific and non-specific effects of psychotherapeutic interventions for depression: Results from a meta-analysis of 84 studies

https://doi.org/10.1016/j.jpsychires.2016.12.015Get rights and content

Abstract

There is a long-standing and very active debate regarding which psychotherapeutic intervention should be used in depressive disorders. However, the effects of psychotherapies may result majorly from non-specific factors rather than from specific factors related to the type of psychotherapeutic intervention. We performed a systematic review and meta-analysis on aggregated data to understand how the effects of different psychotherapies are impacted by non-specific factors. We included randomized controlled trials that assessed the efficacy of psychotherapeutic interventions in the treatment of adult depressive disorders. The primary outcome was the change in depression score from baseline to the latest follow-up visit (i.e. response). A meta-regression was performed to predict response according to the type of intervention and non-specific factors (e.g. number of treatment sessions, length of follow-up, therapeutic allegiance of the investigator). The main analysis included 214 study arms from 84 trials. The effects of psychotherapies compared to the waiting list control condition failed to remain significant after adjusting for non-specific factors. Response increased with the number of treatment sessions (β = 0.03, 95% CI [0.01; 0.04]) and the length of follow-up (β = 0.01, 95% CI [0.00; 0.02]). Response also improved in case of presumed therapeutic allegiances among investigators (β = 0.29, 95% CI [0.07; 0.52]). Response to psychotherapies seems to be closely related to non-specific effects. The development of a well-designed trial that controls for non-specific factors might help disentangle the effects of psychotherapies.

Introduction

There is a long-standing and very active debate concerning which treatment should be used in major depressive disorder (MDD) (Gøtzsche, 2014, Nutt et al., 2014). This debate goes beyond a mere opposition between pharmacological and psychological interventions, and there are many controversies regarding the type of psychotherapy that would be the best for most patients (Cuijpers et al., 2008a, Ekers et al., 2008, Gloaguen et al., 1998, Leichsenring et al., 2004, Wampold et al., 2002). Despite considerable evidence in favor of an equivalency of psychotherapeutic interventions in depression, in the sense of the so-called “Dodo Bird Verdict”, this topic is still controversially discussed among clinicians and scientists (Budd and Hughes, 2009). All these controversies arise from divergent interpretations of the evidence derived from the numerous randomized controlled trials (RCTs) that are available (Cuijpers et al., 2008b), against a backdrop of ideological conflicts of interest. But one must bear in mind that while this debate is focused on hypothetical “specific effects” (the treatment/control difference) of these distinct interventions (for example arising from a theoretical framework underpinning a given psychotherapeutic intervention), it fails to consider the complexity of the phenomenon resulting in the benefit experienced by a given patient (i.e. the pre/post treatment difference, also called response). Some determinants of response, such as empathy shown by caregivers who are actively involved in an intervention in which they trust, are considered as experimental noise and as non-specific effects. These non-specific effects have been shown to be important determinants of antidepressant response (Naudet et al., 2011), and may be even more crucial in the response observed after a psychotherapeutic intervention (Ahn and Wampold, 2001, Chatoor and Krupnick, 2001, Parker et al., 2003). For example, it has been suggested that the improvement in cognitive psychotherapy may occur before the introduction of cognitive restructuring (the specific part of the therapy) (Ilardi and Craighead, 1994). One can hypothesize that this early effect may result from the behavioral techniques that are typically applied early in cognitive therapy (Jacobson et al., 1996), but it might also be hypothesized that the initiation of a rather “unspecific” empathic attitude by the clinician could trigger this improvement. An earlier meta-analysis (Barth et al., 2013) found no differences between psychotherapies, but did between different types of control condition (usual care and placebo groups evidenced better responses than waiting list groups). Although mostly based on indirect evidence, this result suggests that the differences that were evidenced resulted from non-specific factors rather than from specific factors related to the type of psychotherapeutic intervention. Since these non-specific effects are obviously not well balanced across control and experimental groups, an analysis of response determinants in psychotherapeutic interventions for MDD adjusting for these potential confounders is thus a crucial issue 1/to understand how psychotherapies work, and 2/to find ways of maximizing these factors in care that is delivered to patients. It is also a methodological challenge in the promotion of alternative ways of evaluating psychotherapies. We therefore performed a meta-analysis on aggregated data to understand how the effects of different psychotherapies and control conditions are impacted by non-specific factors in MDD psychotherapeutic research.

Section snippets

Methods

The protocol of the study was notified prior to implementation on an international register for systematic reviews (Registration-PROSPERO, 2014:CRD42014009522).

Study selection

A total of 198 studies were identified. After reviewing titles and abstracts, 82 articles were discarded because they did not meet the eligibility criteria. Of the remaining 116 studies, 84 were included in the meta-analysis (one article reported results from two distinct trials), corresponding to 217 study arms for which data were collected. Three arms were then excluded (two arms presented missing data on depression scores, and one arm corresponded to pooled data from two distinct

Summary of evidence

Without adjusting on non-specific factors (univariate analysis), we found a superiority of all psychotherapies and control conditions compared to the waiting list condition. These results are in accordance with those obtained in the Barth et al. meta-analysis (2013). The effects associated with psychotherapies and control conditions compared to the waiting list reference condition were considerably reduced after adjusting on non-specific factors (multivariate analysis), and all the confidence

Implications for research

It is clear that psychotherapeutic interventions are useful, but more research is needed to understand how they work. The development of a well-designed trial that controls for non-specific factors might help disentangle the effects of psychotherapies. In addition, we believe that it would be interesting to explore non-specific effects in future prospective studies, which could provide stronger evidence than meta-regressions and/or indirect comparisons.

Implications for practice

In psychotherapies, the most important

Contributors

Conceived and designed the experiments: NF, PC, GL.

Performed the experiments: PC, GL, NF.

Analyzed the data: PC, GL, NF.

Contributed reagents/materials/analysis tools: PC, GL, NF.

Wrote the paper: PC, GL, NF.

Revised the paper critically for important intellectual content: DD, FB, RJM.

Final approval of the version to be published: PC, GL, DD, RJM, FB, NF.

Role of the funding source

The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Conflicts of interest

There are no conflicts of interest regarding this paper. All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare that (1) No authors have support from any company for the submitted work; (2) P.C. was a trainee in Servier (pharmacokinetics department) for 6 months in 2013; G.L. is a trainee in cognitive behavioral therapy; DD has relationships (board membership or consultancy or

Acknowledgement

This work was supported by a local grant from Rennes CHU (D-PSY-D-META) (CORECT: COmité de la Recherche Clinique et Translationelle). We would like to thank Jean-Yves Lannou for his help on bibliographic research, Clémence Pontoizeau for her help in formatting manuscript, and Angela Swaine Verdier for revising the English.

References (31)

  • P. Cuijpers et al.

    Efficacy of cognitive–behavioural therapy and other psychological treatments for adult depression: meta-analytic study of publication bias

    Br. J. Psychiatry

    (2010)
  • P. Cuijpers et al.

    Psychotherapy for depression in adults: a meta-analysis of comparative outcome studies

    J. Consult. Clin. Psychol.

    (2008)
  • P. Cuijpers et al.

    Psychological treatment of depression: a meta-analytic database of randomized studies

    BMC Psychiatry

    (2008)
  • D. Ekers et al.

    A meta-analysis of randomized trials of behavioural treatment of depression

    Psychol. Med.

    (2008)
  • J.P.T. Higgins et al.

    The Cochrane Collaboration's tool for assessing risk of bias in randomised trials

    BMJ

    (2011)
  • Cited by (0)

    1

    These authors contributed equally to the work.

    View full text