Recently, patients have been asking me about reports floating around the Internet that question the effectiveness of antidepressants. The source of most of these reports is a recently published study entitled “Antidepressant drug effects and depression severity: a patient-level meta-analysis.” The report concludes (emphasis mine):

“The magnitude of benefit of antidepressant medication compared with placebo increases with severity of depression symptoms and may be minimal or nonexistent, on average, in patients with mild or moderate symptoms. For patients with very severe depression, the benefit of medications over placebo is substantial.”

Most of my patients needed some clarification, but one even asked whether she should go off her antidepressant in view of the conclusions drawn in the study. The short answer is “No.” The study does not conclude that antidepressants are ineffective. In addition, this particular study has some severe limitations, which make its clinical correlation at least a bit questionable, including the following:

  • The authors of the article did not report data from their own clinical research. This article is based on calculations they performed on OPS’s (Other People’s Studies). They implemented several inclusion criteria and ended in selecting six studies in which patients took only one of two antidepressants: Paxil (an SSRI antidepressant) or Imipramine (an old Tricyclic antidepressant). None of the many other antidepressants were included.
  • The studies were done in a trial setting rather than in a clinical setting, thus eliminating other key ingredients of successful treatment, including the compassionate, empathetic relationship with a clinician who frequently uses supportive measures and/or psychotherapy in combination with medications. In other words, antidepressants may be more effective in combination with other non-medication interventions.
  • Selection bias: Study directors frequently use their discretion regarding who they included in the study. Unfortunately researchers, at times, try to include patients even if they do not meet the criteria fully, as indicated on page 52 of the article: “…when a minimum score at intake is required for study entry, study diagnosticians sometimes inadvertently inflate the scores of patients whose true score is just below the cutoff.”
  • Compared to clinical setting (outpatient clinic, private practice) where many criteria are used to evaluate improvement with treatment, the present study analysis relies exclusively on one questionnaire called Hamilton Depression Rating Scale. The assumption is that improvement in scores on the questionnaire truly correlates with clinical improvement. Many experts in the field doubt the accuracy of this assumption. Even the authors of the present study quote some articles that express this doubt specifically (references # 22 and 23).

It is therefore surprising to me that with so many limitations in the present study design, the authors decided to summarize their findings in such strong language, stating that in patients with mild or moderate depression the effect of antidepressants is “minimal or nonexistent.” When quoted in the media, such statements often lead people to quickly jump to premature conclusions that are not in their best interests.

I therefore advise caution in reading medical news and offer the following advice: Always consult your doctor before deciding on any treatment changes or discontinuing any medications.