Research Article

Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration

  • Irving Kirsch mail,

    To whom correspondence should be addressed. E-mail:

    Affiliation: Department of Psychology, University of Hull, Hull, United Kingdom

  • Brett J Deacon,

    Affiliation: University of Wyoming, Laramie, Wyoming, United States of America

  • Tania B Huedo-Medina,

    Affiliation: Center for Health, Intervention, and Prevention, University of Connecticut, Storrs, Connecticut, United States of America

  • Alan Scoboria,

    Affiliation: Department of Psychology, University of Windsor, Windsor, Ontario, Canada

  • Thomas J Moore,

    Affiliation: Institute for Safe Medication Practices, Huntingdon Valley, Pennsylvania, United States of America

  • Blair T Johnson

    Affiliation: Center for Health, Intervention, and Prevention, University of Connecticut, Storrs, Connecticut, United States of America

  • Published: February 26, 2008
  • DOI: 10.1371/journal.pmed.0050045

Reader Comments (48)

Post a new comment on this article

Is the placebo effect reliable ?

Posted by plosmedicine on 31 Mar 2009 at 00:24 GMT

Author: Roland Dardennes
Position: Professor of Psychiatry
Institution: University Paris Descartes
Submitted Date: March 23, 2008
Published Date: March 25, 2008
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

Based on a selection of antidepressant randomized controlled trials (RCT), Kirsch and co-authors (1) concluded that “The relationship between initial severity and antidepressant efficacy is attributable to decreased responsiveness to placebo among very severely depressed patients, rather than to increased responsiveness to medication.” What is obvious from their meta-analysis is that responsiveness to placebo is not a stable phenomenon, opposed to the stable responsiveness to medication (see figure 3). My aim as a medical practitioner is to maximize the placebo responsiveness of any compound I expect to be efficacious, but is it feasible? Branded placebos give better results than unbranded placebos, red tablets are considered as stimulant and blue ones as sedative, and two pills of placebo are more efficacious than one (2). Frequent visits improve placebo response in 6-weeks antidepressant RCT (3). A further question is how much of the placebo response can I expect for my patients, given the results of placebo treatment in clinical trials ? Recent evidence showed that placebo response in Parkinson’s disease is related to the release of dopamine in both dorsal and ventral striatum (4) that may explain the role of expectation of clinical benefit in the placebo response through the activation of reward mechanisms. Sustained activation of ventral midbrain dopamine neurons is maximal when uncertainty is maximal (p=0.5), less pronounced at p=0.25 and 0.75, and absent at p=0.0 and 1.0 (5). De la Fuente-Fernandez et al. (6) suggested that clinical trials with a classical two-group parallel design maximize placebo response given that the a priori probability of clinical benefit is p = 0.5. Diminished dopaminergic transmission plays a role in the pathophysiology of depression; many dopamine agonists have demonstrated antidepressant activity (7). Kirsch et al.’s results may represent the maximal placebo response we can expect in treating patients with depression. In real-settings, this response may be lower than those obtained in RCT. The last point is to find alternative treatments if I do not expect efficacy from antidepressants and if I am not confident in maximizing placebo response. Alas, when placebo psychotherapies have a structure similar to the active method (e.g. same number of sessions), active psychotherapy has a very small superiority with an effect size of 0.149 (8). Mastering the placebo effect seems to be difficult. For the depressed patient who meets doctors and therapists with unknown expertise in placebo, antidepressants appear to give a constant answer.
1. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, et al. (2008) Initial severity and antidepressant benefits: A meta-analysis of data submitted to the Food and Drug Administration. PLoS Med 5(2):e45.
2. Moerman DE, Jonas WB (2002) Deconstructing the Placebo Effect and Finding the Meaning Response. Ann Intern Med 136:471-476.
3. Posternak MA, Zimmerman M (2007) Therapeutic effect of follow-up assessments on antidepressant and placebo response rates in antidepressant efficacy trials: meta-analysis. Br J Psychiatry 190:287-92.
4. de la Fuente-Fernández R, Ruth TJ, Sossi V, Schulzer M, Calne DB, et al. (2001) Expectation and Dopamine Release: Mechanism of the Placebo Effect in Parkinson's Disease. Science 293: 1164-1166.
5. Fiorillo CD, Tobler PN, Schultz W (2003) Discrete coding of reward probability and uncertainty by dopamine neurons. Science 299:1898 –1902.
6. de la Fuente-Fernández R, Schulzer M, Stoessl AJ (2004) Placebo mechanisms and reward circuitry: clues from Parkinson's disease. Biol Psychiatry 56:67-71.
7. Dunlop BW, Nemeroff CB (2007) The role of dopamine in the pathophysiology of depression. Arch Gen Psychiatry 64:327-37.
8. Baskin TW, Tierney SC, Minami T, Wampold BE (2003) Establishing specificity in psychotherapy: a meta-analysis of structural equivalence of placebo controls. J Consult Clin Psychol 71:973-9.

Competing interests declared: RD is professor of Psychiatry, prescribes antidepressants with positive expectations he shares with his patients and his students, and has received honoraria from Lilly, Lundbeck, Janssen-Cilag and Sanofi.