Fournier JC, DeRubeis RJ, Hollon SD, et al. Antidepressant drug effects and depression severity: a patient-level meta-analysis. JAMA. 2010;303:47-53.
A patient-level meta-analysis of randomized-placebo controlled trials (RCTs) of antidepressant medication (ADM) in the treatment of major or minor depressive disorder.
This study was designed significantly different than previous meta-analyses of ADM in depressive disorders.1,2 Previous meta-analyses looked at RCTs that only included patients with baseline depression scores reflecting a diagnosis of very serious depression (Hamilton Depression Rating Scale [HDRS] greater than or equal to 23). Unlike the current meta-analysis by Fournier et al, previous meta-analyses did not adequately explore the efficacy of ADM for individuals with mild-to-moderate depression.
The study included combined data from 6 large-scale placebo-controlled RCTs that were composed of patients with a broad range of baseline symptom severity. The individual HDRS scores for this analysis included mild-to-moderate depression and ranged from 10 to 39. The pooled sample in the analysis included 434 patients in the treatment group and 284 patients in the placebo group.
The authors evaluated the published literature from January 1980 through March 2009. The criteria for inclusion required studies to be RCTs of an FDA-approved ADM in the treatment of minor thorough major depressive disorder. Studies that examined only major depression or any other special sub-population were not included. The studies were restricted to adult outpatient treatment and treatment duration of greater than 6 weeks. All studies were required to record HDRS scores at intake and at the end of the treatment.
For this analysis the authors also excluded studies that included a placebo washout period. During a placebo washout period all patients are given a placebo in a single-blinded fashion. At the end of this period patients experiencing a placebo effect are eliminated from the study prior to randomization. Because early placebo responders are removed from the trial before they contribute data, the true rate of placebo response can be underestimated. A placebo washout is designed to enhance the ability to detect a difference in efficacy between medication and placebo. It also severely limits the ability to accurately assess the placebo response rate.
This study is considered a “mega-analysis” because the authors pooled the individual patient data as opposed to treatment groups, which is how traditional meta-analyses are performed. The mega-analysis is considered more appropriate and more powerful than a standard meta-analysis when original data is available. Therefore studies were excluded if the individual patient-level data was not available for fine-grained multivariate analysis.
Primary Outcome Measures
The primary statistical analysis investigated the relationship between baseline symptom severity (HDRS score) and treatment efficacy (measured as symptom change from intake to the end of treatment).
True drug effects (advantage of drug over placebo) were nonexistent to negligible for patients with baseline mild-to-moderate depression and similarly ineffective for lower-scoring severe depression. For baseline HDRS scores of less than 25, the magnitude of drug/placebo differences did not meet clinical significance.
For patients with the highest levels of baseline depression severity, medication was markedly superior to placebo.
The majority of clinicians are unaware that the established efficacy for ADM is based on studies that only include patients with more severe depression. Prior to the current mega-analysis by Fournier et al, there has been virtually no systematic evaluation of the true treatment effect of ADM in patients with less severe depression. This is in part because such data are limited in the FDA database and published literature. The inclusion criteria for the majority of studies impose a baseline HDRS cutoff score that reflects a diagnosis of severe depression to increase the sensitivity of ADM/placebo comparisons.
The current study is significant because it demonstrates there is little evidence to suggest ADM produce a specific pharmacological benefit for the majority of patients to whom they are prescribed. A recent survey of 503 depressed patients seeking treatment for depressions showed that the majority of patients (71%) had a HDRS score less than 22. 3 The current analysis showed that for HDRS scores of less than 25, ADM may have no-to-minimal treatment effect. One may extrapolate from this report that approximately 71% of the millions of patients that are prescribed ADM are not achieving a clinically significant benefit. The lack of evidence showing efficacy for ADM in less severe depression is not clearly reflected in the marketing messages to doctors and the public. In addition, the potential misguided prescribing of ADM has significant economic impact to consider.
Prescribers, policy makers, insurance companies, and consumers need to take stock in the many implications of these results. For integrative practitioners these results reinforce the importance of a multipronged integrative approach to mild-to-moderate depression. The results of this study also inform conventional evidence-based physicians to modify their prescribing practices and increase recommendations for other underutilized therapies such as counseling, dietary changes, exercise, and mind-body modalities. These noninvasive therapies have the opportunity to become the primary treatment choice for symptoms of mild-to-moderate depression. These modalities also have health-promoting side effects, which positively impact an individual’s overall health.
The results of the current study also open the door for botanicals and nutrients found within the integrative medicine tool kit. For milder forms of depression it may be appropriate to start with an adaptogenic botanical such as Rhodiola rosea. Emerging animal and human evidence suggest Rhodiola may be efficacious in stress, anxiety, and fatigue.4,5,6 There is also a strong body of evidence that consists of 29 studies from a variety of countries with 5,489 patients that shows St. John’s wort to be more effective than placebo and to have fewer side effects than standard ADM for treatment of depression.7 Recent studies have shown that vitamin D receptors exists in the brain, and evidence is rapidly emerging that correlates vitamin D deficiency and incidence of depression.8,9,10 Testing and normalizing vitamin D levels should be considered a first line of therapy for all practitioners treating depression. S-adenosyl methionine (SAMe) has a strong body of evidence that shows that it is comparable to or more effective than ADM, and with fewer side effects.11,12
Many practitioners consider depression a highly individualized and multifactorial condition. No 2 cases of depression are the same, and therefore a one-size-fits-all approach is likely not the most effective. The current study indirectly supports the use of non-pharmacological integrative management of mild-to-moderate depression. Evidence-based physicians need to return to treatments with an evidence-base supporting efficacy in milder forms of depression. For patients with more sever forms of depression (HDRS greater than 25) pharmaceutical management may be appropriate and the evidence-base supports this use. Until additional data disputes the current meta-analysis, practitioners accustomed to prescribing ADM for mild-to-moderate depression need to reevaluate their prescribing practices.
- Kirsch I, Deacon BJ, Huedo-Medina TB, et al. Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLoS Med. 2008;5:e45.
- Khan A, Leventhal RM, Khan SR, et al. Severity of depression and response to antidepressants and placebo: an analysis of the Food and Drug Administration database. J Clin Psychopharmacol. 2002;22:40-45.
- Zimmerman M, Posternak MA, Chelminski I. Symptom severity and exclusion from antidepressant efficacy trials. J Clin Psychopharmacol. 2002;22:610-614.
- Olsson EM, von Schéele B, Panossian AG. A randomised, double-blind, placebo-controlled, parallel-group study of the standardised extract shr-5 of the roots of Rhodiola rosea in the treatment of subjects with stress-related fatigue. Planta Med. 2009;75:105-112.
- Mattioli L, Funari C, Perfumi M. Effects of Rhodiola rosea L. extract on behavioural and physiological alterations induced by chronic mild stress in female rats. J Psychopharmacol. 2009;23:130-142.
- Chen QG, Zeng YS, Qu ZQ, et al. The effects of Rhodiola rosea extract on 5-HT level, cell proliferation and quantity of neurons at cerebral hippocampus of depressive rats. Phytomedicine. 2009;16:830-838.
- Linde K, Berner MM, Kriston L. St John's wort for major depression. Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD000448.
- Cherniack EP, Troen BR, Florez HJ, et al. Some new food for thought: the role of vitamin D in the mental health of older adults. Curr Psychiatry Rep. 2009;11:12-19. 9. Berk M, Sanders KM, Pasco JA, et al. Vitamin D deficiency may play a role in depression. Med Hypotheses. 2007;69:1316-1319.
- Bertone-Johnson ER. Vitamin D and the occurrence of depression: causal association or circumstantial evidence? Nutr Rev. 2009;67:481-492.
- Alpert JE, Papakostas G, Mischoulon D, eta l. S-adenosyl-L-methionine (SAMe) as an adjunct for resistant major depressive disorder: an open trial following partial or nonresponse to selective serotonin reuptake inhibitors or venlafaxine. J Clin Psychopharmacol. 2004;24:661-664.
- Papakostas GI, Shelton RC. Use of Atypical Antisychotics for Treatment Resistant Major Depressive Disorder. Curr Psychiatry Rep. 2003;5:460-466.