Elsevier

Journal of Affective Disorders

Volume 241, 1 December 2018, Pages 417-424
Journal of Affective Disorders

Research paper
Treating a broader range of depressed adolescents with combined therapy

https://doi.org/10.1016/j.jad.2018.08.027Get rights and content

Highlights

  • Response distributions of depressed adolescents were compared across treatments.

  • Cognitive-behavioral therapy was no more effective than placebo.

  • Fluoxetine was effective only in the middle range of response levels.

  • Combining cognitive-behavioral therapy and fluoxetine was broadly effective.

  • Combining the two treatments was also effective in low responders.

Abstract

Background

Traditional statistical analyses of clinical trials encompass the central tendency of outcomes and, hence, are restricted to a treatment's average effectiveness. Our aim was to get a more complete picture of the effectiveness of standard treatment options for adolescent depression, by analyzing treatment effects across low, middle, and high levels of response.

Methods

Secondary data analysis was performed of the Treatment for Adolescents with Depression Study (TADS, ClinicalTrials.gov, NCT00006286), a randomized controlled trial comparing fluoxetine (FLX), cognitive-behavioral therapy (CBT), and their combination (COMB) against placebo treating adolescents with major depression (n = 439). The proportional change from baseline to week 12 in the Children's Depression Rating Scale-Revised was used as an index of response. Response levels were analyzed via quantile regression models, thereby estimating treatment effects across the entire response level distribution, adjusted for baseline depression, study site, and patients’ treatment expectancies.

Results

Whereas CBT was no more effective than placebo across response levels, COMB was more effective than FLX in that its quantile treatment effects were both larger in magnitude and spread out across a broader range of response levels, including the low end of the response level distribution. Cohen's d of the difference was 1.39 (95% confidence interval 1.33–1.45).

Limitations

Ad-hoc analysis using data from a trial that was not originally designed to accommodate such analysis.

Conclusion

The combination of cognitive-behavioral therapy and fluoxetine was more effective than either treatment used alone, not just in average effectiveness, but in the breadth of patients in whom it was effective.

Introduction

Among young people, major depression accounts for a substantial portion of the burden of disease (Gore et al., 2011, Merikangas et al., 2010). Furthermore, depression has a variety of adverse effects, both immediately and in adulthood. In particular, depression during adolescence is associated with an increased risk for depression and other psychopathology in adulthood (Fergusson and Woodward, 2002, Jones, 2013, Rutter et al., 2006); an increased risk for self-harm and suicidal behaviors (Fergusson and Woodward, 2002, Gould et al., 1998, Hawton et al., 2012, Thapar et al., 2012); an increased risk for reduced social functioning and work difficulties (Costello and Maughan, 2015, Fergusson and Woodward, 2002, Thapar et al., 2012); and an increased risk of future adverse birth outcomes among women (Nkansah-Amankra and Tettey, 2015). It is clear, then, that depression in adolescents needs to be treated swiftly and effectively.

The ‘Treatment for Adolescents with Depression Study’ (TADS) was a corner-stone clinical trial that tested the effectiveness of standard treatment options for major depression in adolescents, including treatment with the selective serotonin reuptake inhibitor (SSRI) fluoxetine, treatment with cognitive-behavioral therapy, and these two treatments combined (March et al., 2004, Thapar et al., 2012). Statistical analysis of the acute phase indicated superiority of fluoxetine used alone, and of the combined treatment over a pill placebo, as well as of combined treatment over either fluoxetine or cognitive-behavioral therapy used alone; cognitive-behavioral therapy, however, failed to outperform the placebo (March et al., 2004).

Treatment effects were estimated using standard statistical methods, including mixed-effects and logistic regression (March et al., 2004). Whereas such analyses are indispensable to establishing a treatment's general level of effectiveness, they only truly examine outcomes for each treatment in terms of their central tendency, for example their mean value across all subjects within each treatment arm. However, looking at central tendency fails to reveal whether any detected treatment effect exists only on average or across the entire distribution of outcome values (Cleveland, 1994, Koenker and Hallock, 2001, Wilcox, 2012).

The quantile regression framework allows for a more complete look at the differences between the outcome distributions of two treatment conditions. With quantile regressions, the quantiles of an outcome's distribution are modeled as a function of the predictor variable(s), in contrast to the traditional regression framework, wherein only the mean of the outcome is modeled (Beyerlein, 2014, Briollais and Durrieu, 2014, Koenker and Hallock, 2001, Petscher and Logan, 2014). This allows for comparing all quantiles of the treatment condition's outcome distribution against the corresponding quantiles of the placebo condition's outcome distribution. Comparing quantiles is the most informative way of comparing two groups because it shows whether the groups differ in a simple or in a complicated way (Cleveland, 1994, Wilcox, 2012). In the simple case, one group has consistently higher values than the other. In a more complicated case, low, middle, and high outcome values differ by different amounts. It is possible that the groups differ in an even more complicated way, for example with one group outperforming the other at lower outcome values, but being outperformed at higher outcome values. Such complicated group differences cannot be uncovered by comparing the groups’ central tendencies.

Assessing a treatment's effect across the entire outcome range would be especially attractive if the outcome is an index of response (e.g., indexing the degree of change from baseline to the trial's end), since such an analysis would show whether a given treatment has a consistent versus differential effect on low, middle, and high response levels. Thus, rather than examining whether the average response level is higher among subjects receiving active treatment relative to controls, one could examine whether the treatment has some level of effectiveness across the entire range of response levels, including special subject subsets like low and high responders (Beyerlein, 2014, Briollais and Durrieu, 2014, Petscher and Logan, 2014). A treatment that is able to exert a consistent effect across response levels is more clinically valuable than a treatment with a more restricted breadth of effectiveness.

Our current aim was to gain a more complete picture of TADS treatment effects, by re-evaluating them across the entire response range. We did so by analyzing patients’ change in depression severity from baseline to the end of the TADS acute phase as an index of response, using quantile regression models (Beyerlein, 2014, Briollais and Durrieu, 2014, Koenker and Hallock, 2001, Petscher and Logan, 2014). This approach allowed us to analyze both central tendency and outcome extremes, thereby providing more insights into the breadth of the treatments’ effectiveness than traditional regression analyses. Since we are not aware of any comparable previous analyses, the study was exploratory. The research question was whether treatment effectiveness varied between low, middle, and high response levels.

Section snippets

Study design

The current study entailed secondary analysis of TADS trial data (ClinicalTrials.gov, NCT00006286). TADS is a randomized controlled trial that was designed to compare the effectiveness of cognitive-behavioral therapy CBT), a specific SSRI called fluoxetine, and the two treatments combined, all versus a pill placebo for the treatment of adolescents with a major depressive disorder. We analyzed data from the acute phase of the trial, which encapsulated the first 12 weeks of treatment. The data

Results

Table 1 provides the 10th to 90th response level quantiles of each treatment arm. Response levels were rather dispersed. In the placebo condition, the response level on the 10th quantile was 0.067 and as high as 0.61 on the 90th quantile. This corresponds to a reduction of the baseline depression score by 6.7% on the 10th quantile and by 60.1% on the 90th quantile. Similar ranges of response levels were found for CBT and FLX. In contrast, COMB had clearly higher response levels, achieving a

Discussion

In the TADS sample, whereas cognitive-behavioral therapy was found to be no more effective than placebo across response levels, our analyses revealed that combining cognitive-behavioral therapy and fluoxetine was not only effective, but had a more comprehensive effectiveness than fluoxetine used alone. This was indicated by higher AUCs for the combined-treatment effects, reflecting simultaneously that these effects were greater in magnitude and spread out across a broader range of response

Conclusion

We found that, in the TADS sample, the combination of cognitive-behavioral therapy and the SSRI fluoxetine was more effective than either treatment used alone, not just in average effectiveness, but in the breadth of patients in whom it was effective.

Declarations of interest

None.

Contributors

Conceptualization of study: SF and MMK. Data acquisition and preparation: SF. Statistical analysis: SF. Interpretation of data: SF and MMK. Drafting the manuscript: SF. Revising the manuscript for important intellectual content: MMK. All authors have approved the final manuscript.

Funding source

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Acknowledgements

Data used in the preparation of this manuscript were obtained and analysed from the controlled access datasets distributed from the National Institute of Mental Health (NIMH)-supported National Database for Clinical Trials (NDCT, https://data-archive.nimh.nih.gov/ndct/). NDCT is a collaborative informatics system created by the NIMH to provide a national resource to support and accelerate discovery related to clinical trial research in mental health. Dataset identifier: Clinical Trials #2145.

References (79)

  • H.J. Moller et al.

    Position statement of the European Psychiatric Association (EPA) on the value of antidepressants in the treatment of unipolar depression

    Eur. Psychiatry

    (2012)
  • S. Nkansah-Amankra et al.

    Association between depressive symptoms in adolescence and birth outcomes in early adulthood using a population-based sample

    Prev. Med. Rep.

    (2015)
  • M.A. Reinecke et al.

    Cognitive-behavioral therapy of depression and depressive symptoms during adolescence: a review and meta-analysis

    J. Am. Acad. Child Adolesc. Psychiatry

    (1998)
  • TaoR. et al.

    Early prediction of acute antidepressant treatment response and remission in pediatric major depressive disorder

    J. Am. Acad. Child Adolesc. Psychiatry

    (2009)
  • A. Thapar et al.

    Depression in adolescence

    Lancet (London, England)

    (2012)
  • T. Usala et al.

    Randomised controlled trials of selective serotonin reuptake inhibitors in treating depression in children and adolescents: a systematic review and meta-analysis

    Eur. Neuropsychopharmacol. J. Eur. College Neuropsychopharmacol.

    (2008)
  • K.D. Wagner et al.

    A double-blind, randomized, placebo-controlled trial of escitalopram in the treatment of pediatric depression

    J. Am. Acad. Child Adolesc. Psychiatry

    (2006)
  • V.R. Weersing et al.

    Cognitive behavioral therapy for depression in youth

    Child Adolesc. Psychiatr. Clin. N. Am.

    (2006)
  • C.J. Whittington et al.

    Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data

    Lancet (London, England)

    (2004)
  • M.K. Ali et al.

    Comparative efficacy of escitalopram in the treatment of major depressive disorder

    Neuropsychiatr. Dis. Treat.

    (2011)
  • Diagnostic and Statistical Manual of Mental Disorders

    (1994)
  • P. Bech

    Is the antidepressive effect of second-generation antidepressants a myth?

    Psychol. Med.

    (2010)
  • P. Bech

    The responsiveness of the different versions of the Hamilton Depression Scale

    World Psychiatry: Off. J. World Psychiatr. Assoc. (WPA)

    (2015)
  • A. Beyerlein

    Quantile regression-opportunities and challenges from a user's perspective

    Am. J. Epidemiol.

    (2014)
  • D. Brent et al.

    Switching to another SSRI or to venlafaxine with or without cognitive behavioral therapy for adolescents with SSRI-resistant depression: the TORDIA randomized controlled trial

    JAMA

    (2008)
  • J.A. Bridge et al.

    Clinical response and risk for reported suicidal ideation and suicide attempts in pediatric antidepressant treatment: a meta-analysis of randomized controlled trials

    JAMA

    (2007)
  • L. Briollais et al.

    Application of quantile regression to recent genetic and -omic studies

    Hum. Genet.

    (2014)
  • Canty, A., Ripley, B., 2016. boot: Bootstrap R (S-Plus) Functions, R package version 1.3-18....
  • CheungA. et al.

    Maintenance study for adolescent depression

    J. Child Adolesc. Psychopharmacol.

    (2008)
  • W.S. Cleveland

    The Elements of Graphing Data

    (1994)
  • E.J. Costello et al.

    Annual research review: optimal outcomes of child and adolescent mental illness

    J. Child Psychol. Psychiatry Allied Discip.

    (2015)
  • G.R. Cox et al.

    Psychological therapies versus antidepressant medication, alone and in combination for depression in children and adolescents

    Cochrane Database Syst. Rev.

    (2014)
  • G.R. Cox et al.

    Interventions for preventing relapse and recurrence of a depressive disorder in children and adolescents

    Cochrane Database Syst. Rev.

    (2012)
  • D. Curran-Everett et al.

    Explorations in statistics: the analysis of change

    Adv. Physiol. Educ.

    (2015)
  • B. Dubicka et al.

    Combined treatment with cognitive-behavioural therapy in adolescent depression: meta-analysis

    Br. J. Psychiatry: J. Ment. Sci.

    (2010)
  • B. Efron et al.

    An Introduction to the Bootstrap

    (1994)
  • G.J. Emslie et al.

    Fluoxetine versus placebo in preventing relapse of major depression in children and adolescents

    Am. J. Psychiatry

    (2008)
  • G.J. Emslie et al.

    Treatment of Resistant Depression in Adolescents (TORDIA): week 24 outcomes

    Am. J. Psychiatry

    (2010)
  • D.M. Fergusson et al.

    Mental health, educational, and social role outcomes of adolescents with depression

    Arch. Gen. Psychiatry

    (2002)
  • Cited by (0)

    View full text