Comparative efficacy of antidepressants on anxiety features in depression: A meta-analysis of double-blind studies of imipramine and moclobemide against placebo

Abstract

Listen

Translate article

INTRODUCTION: We carried out a meta-analysis of data for 950 depressed patients from all double-blind, placebo-controlled comparative studies of moclobemide against imipramine. The aim of the analysis was to determine: 1 if the two antidepressants reduce anxiety symptoms as effectively as symptoms of depression; 2. if there are differences in the anti-anxiety effects; and 3. if the severity of anxiety symptoms influences the outcome of treatment. METHOD: The analysis was done in sub-groups of patients classified according to different definitions of anxiety and to the degree (mild, moderate, severe) of psychic and somatic anxiety present at baseline. The definitions and the criteria for the severity of anxiety were derived from the HAMD-17 anxiety item scores (items 10 and 11) and HAMD-17 anxiety sub-scales (ANXP and SOMA). The application of various definitions of anxiety (single-item and composite sub-scale scores) served to check the consistency, reproducibility and generalizability of the results. The criteria of efficacy and of the outcome of treatment, which were set in advance, were either: 50% reduction of the HAMD anxiety sub-scale; or percentage of global improvement (CGEA, very much and much) at the end of treatment (4 weeks). RESULTS: The results showed that moclobemide and imipramine are equally effective in reducing psychic and somatic anxiety in depressed patients, independent of the severity of anxiety or how it was defined. The criterion of 50% decrease of anxiety scores (at week 4) was consistently reached by 60-70% of patients in all drug sub-groups. The time course of effects on anxiety symptoms was similar with both drugs. It ran parallel to the regression of depression, with significant improvement of anxiety symptoms by week 2 on all measures. The frequency of benzodiazepine (BDZ) co-administration was not systematically correlated with the severity of anxiety. Moreover, co-prescription of BDZ did not change the outcome of drug treatment, with respect either to anxiety symptoms or global therapeutic efficacy. In placebo groups, significant negative interactions were found between severity of anxiety and the efficacy of treatment, which decreased with increased anxiety. The incidence of adverse events with moclobemide was barely higher than in the placebo group. Insomnia (4.9%) was the only adverse event significantly higher with moclobemide than with other treatments. Drop-out rates in drug groups were similar (27%) and unrelated to the severity of anxiety. In the placebo group, drop-outs were significantly higher than in drug groups (39%), and were positively correlated with the severity of anxiety. CONCLUSION: The results of this study indicate that anxiety symptoms in depressed patients respond similarly to moclobemide or imipramine and that the severity of these symptoms (or the co-prescription of benzodiazepine) has no influence on either the magnitude of anti-anxiety effects or on overall outcome. (Int J Psych Clin Pract 2000; 4:111-117)