The review determined the efficacy of monoamine oxidase inhibitors (MAOIs) for first-line treatment of atypical depression. The authors concluded that MAOIs may be more effective for atypical major depressive disorder than tricyclic antidepressants. While this cautious conclusion appears to reflect the data presented, small datasets, uncertain quality of the included studies and the lack of a reported review process limit confidence in interpreting these results.

To determine the efficacy of monoamine oxidase inhibitors (MAOIs) for first-line treatment of patients with atypical depression.

MEDLINE, EMBASE, PsycINFO, and PSYNDEX were searched from inception to 2004; the MeSH terms were reported. In addition, reference lists from identified articles were checked for relevant articles, and an expert in the field was contacted in order to locate any unpublished studies.

Randomised, controlled, double-blind clinical trials (RCTs) were eligible for inclusion in the review.

Studies comparing MAOIs with placebo or with selective serotonin re-uptake inhibitors (SSRIs) or tricyclic antidepressants (TCAs) in the acute phase treatment, which refers to a minimum treatment period of 6 weeks and a maximum treatment period of 12 weeks, were eligible for inclusion. The MAOIs included in the review were phenelzine and moclobemide.

Adults suffering from depression with atypical features, as defined by the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-IV or DSM-III). Atypical depression had to be the primary diagnosis.

The primary outcome in this review was efficacy of treatment, as measured by differences in response rates (in relation to the severity of atypical depression). Clinical trials without a standardised indication of responder rates were excluded. The included trials used a number of psychometric instruments: the Clinical Global Impression scale, the Hamilton Depression Rating Scale (HAMDS), the Schedule for Affective Disorders and Schizophrenia-Change, and the Hopkins Symptom Checklist.

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

Aspects of quality were assessed as part of the inclusion criteria: all included studies were required to be RCTs and to be double-blinded. The authors did not state how many reviewers performed the quality assessment.

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction.

Effect sizes were calculated for each study using the phi coefficient.

Individual and pooled effect sizes and their 95% confidence intervals (CIs) were calculated for the drug groups, weighted by sample size. Publication bias was examined using funnel plots.

Statistical heterogeneity was assessed using the chi-squared test.

Eight RCTs (n=670) were included in the review, resulting in 11 comparisons.

A significantly greater improvement in clinical symptoms from baseline was found with MAOIs (phenelzine) in comparison with placebo (effect size, d=0.45, 95% CI: 0.352, 0.605), based on 4 comparisons (n=250). Similarly, when compared with TCAs (imipramine), MAOIs (phenelzine) demonstrated a small but significantly greater improvement in clinical symptoms from baseline (d=0.27, 95% CI: 0.161, 0.421), based on 4 comparisons (n=236). However, no significant difference was shown when MAOIs (phenelzine and moclobemide) were compared with SSRIs (d=0.02, 95% CI: -0.104, 0.139), based on 3 comparisons (n=265). No statistical heterogeneity was found in these analyses. A funnel plot, looking at the dependency of the single effect sizes, did not reveal any pronounced publication bias.

MAOIs may be more effective for atypical major depressive disorder than tricyclic antidepressants. However, insufficient data were available for a comparison of MAOIs and SSRIs.

The research question was supported by clear inclusion criteria in terms of the population, study design and intervention. Several electronic databases were searched using broad MeSH terms; the authors did not state whether their search was restricted by language. Attempts were made to assess publication bias and locate unpublished studies. The procedures undertaken to select studies and extract the data were not reported and, as such, the likelihood of reviewer error or bias cannot be assessed.

The analyses undertaken were generally appropriate, although it might not have been appropriate to pool the data from MAOI and SSRI comparisons, owing to clinical heterogeneity. In addition, it was unclear which psychometric tests were included in the analyses. Other limitations, as acknowledged by the review authors, are the small datasets and that the HAMDS may not adequately assess the symptom profile found in atypical depression. While the authors' cautious conclusion regarding MAOIs and TCAs appears to reflect the data presented, small datasets, uncertain quality of the included studies and the lack of a reported review process limit confidence in interpreting these results.

Practice: The authors did not state any implications for practice.

Research: The authors stated that further studies with more recently developed antidepressants, including irreversible MAOIs, with an improved safety record are warranted.

This record is a structured abstract written by CRD reviewers. The original has met a set of quality criteria. Since September 1996 abstracts have been sent to authors for comment. Additional factual information is incorporated into the record. Noted as [A:....].