|
St John's wort (e.g. JARSIN): an antidepressant effect only in German-speaking countries? A Cochrane review of the benefit of St John's wort (e.g. JARSIN) in depression first presented ten years ago was republished in October. Unlike the previous versions (cf. a-t 2005; 36: 48), the current version only analyses studies that include exclusively patients with "major depression" according to the US classification system DSM* IV (similar to depressive episode in ICD** 10). Most patients suffer from mild to moderate symptoms. The response rate with St John's wort is according to the meta-analysis overall significantly higher than with placebo (Rate Ratio [RR] 1.48; 95% confidence interval [CI] 1.23-1.77), and no significant differences are found from standard antidepressants. Despite the strict inclusion criteria, however, there are major differences (heterogeneity) between the results of the individual placebo-controlled studies. The effects do not even all point in the same direction. In two studies St John's wort comes out worse than placebo. The study of different St John's wort preparations does not explain the differences. Additional analyses intended to elucidate the causes of this marked heterogeneity highlight a peculiar constellation which requires explanation: only studies conducted in German-speaking countries including Austria and Switzerland show an effect (RR 1.78; 95% CI 1.42-2.25), but not those from other countries (RR 1.07; 95% CI 0.88-1.31). As in a previous version of the review, a weaker effect is found in larger studies than in smaller studies - evidence of possible bias due to the non-publication of negative studies or due to methodological deficiencies in the published small studies. In the authors' opinion it cannot be ruled out that some small studies from German-speaking countries are also misconducted and overestimate the effect (LINDE, K. et al.: St. John's wort for major depression. The Cochrane Database of Systematic Reviews 2008, Issue 4; date of last revision Oct. 2008). We feel that a meta-analytical review of the study results seems to be doubtful under these circumstances. In our opinion it cannot provide any evidence of benefit. -Ed.
|
