Editor—In contrast to Geddes and Cipriani's assertion of a lack of clinical uncertainty about the efficacy of antidepressants,1 a careful look at the evidence shows that this may not be true, even of the tricylic antidepressants. Although many trials show tricylic antidepressants to be superior to placebo, many find no difference, including some of the largest and best conducted older trials.2,3 In addition several factors are likely to have inflated the apparent superiority of antidepressants over placebo.4
These include unblinding because of side effects of active drugs compared with inert placebos; and inclusion of items on rating scales, such as sleep and agitation, which are likely to respond to non specific sedative effects of drugs. Since differences often amount to only a few points on rating scales, antidepressants may seem superior to placebo due to these effects alone. The remaining factors are selective reporting of positive outcomes, excluding early withdrawals from analysis, and publication bias.
The fact that many agents not classified as antidepressants—including benzodiazepines, barbiturates, buspirone—and various antipsychotics are superior to placebo or equivalent to antidepressants in some trials shows that “antidepressant effects” may indeed be attributable to non-specific pharmacological or psychological factors.4
The lack of substantial differences between antidepressants and placebo can-not be explained away by assuming that effects are diluted by inclusion of patients with mild conditions. Research on who responds best to antidepressants is inconsistent and inconclusive. The belief that patients with more severe depression respond better is based on limited data, and other evidence shows the reverse.5
Neither new nor older antidepressants have been shown to have specific “antidepressant” effects. Neither is consistently distinguishable from placebo, and the superiority sometimes observed may be attributable to non-specific effects or other methodological artefacts.
Competing interests: None declared.
References
- 1.Geddes JR, Cipriani A. Selective serotonin re-uptake inhibitors remain useful drugs which need careful monitoring. BMJ 2004;329: 809-10. (9 October.) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Medical Research Council. Clinical trial of the treatment of depressive illness. BMJ 1965;i: 881-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Raskin A, Schulterbrandt JG, Reatig N, Chase C, McKeon JJ. Differential response to chlorpromazine, imipramine and placebo. Arch Gen Psych 1970;23: 164-73. [DOI] [PubMed] [Google Scholar]
- 4.Moncrieff J. Are antidepressants over-rated? A review of methodological problems in antidepressant trials. J Nerv Ment Dis 2001;189: 288-95. [DOI] [PubMed] [Google Scholar]
- 5.Moncrieff J. A comparison of antidepressant trials using active and inert placebos. Int J Methods Psychiatr Res 2003;12: 117-27. [DOI] [PMC free article] [PubMed] [Google Scholar]