St John's wort as an antidepressant

BMJ 1996; 313 doi: http://dx.doi.org/10.1136/bmj.313.7052.241 (Published 03 August 1996) Cite this as: BMJ 1996;313:241
  1. Peter A G M De Smet,
  2. Willem A Nolen
  1. Clinical pharmacologist Royal Dutch Association for the Advancement of Pharmacy, 2514 JL The Hague, Netherlands
  2. Psychiatrist HC Rumke Group, Academic Hospital, Utrecht, Netherlands

    Longer term studies are needed before it can be recommended in major depression

    Among the many remarkable differences between German and British medicine is the extensive use of herbal medicines in Germany. An example is the use of Hypericum perforatum (St John's wort) for psychological complaints. In 1994 German physicians prescribed almost 66 million daily doses of preparations containing hypericum alone, worth a total of DM61m (£26m).1 In this issue Linde et al (p 253) conclude from a meta-analysis that hypericum extracts are more effective than placebo for treating mild to moderately severe depressive disorders.2 How should this evidence be weighed?

    Diagnostically, depressive disorders form a heterogeneous group which includes major depression as well as depressive states not satisfying the criteria for the full syndrome. These diagnostic types are defined not by severity but by clusters of symptoms and by the length of time they last. Since the efficacy of antidepressants in major depression is well documented, this syndrome is useful when considering treatment with antidepressants. Antidepressant drugs are particularly recommended for moderate and more severe forms of major depression.3 4

    One reason why German physicians may prescribe so many herbal medicines is because they are easily available. Marketing authorisation for hypericum extracts and other crude herbal medicines is obtained in Germany be reference to special phytotherapeutic monographs. When the monograph on St John's wort came out in 1984 it identified the constituent hypericin as an experimental monoamine oxidase inhibitor and permitted the medicinal use of the herb in average daily doses of 2-4 g of drug, or 0.2-1.0 mg total hypericin, for psychogenic disturbances, depressive states, anxiety, or nervous excitement.5 However, the clinical relevance of the monoamine oxidase inhibiting properties of hypericum extracts has not been confirmed,6 and hypericin is not their only bioactive constituent.7 Standardisation of hypericum extracts on hypericin content may therefore offer no guarantee of pharmacological equivalence.

    To be accepted as an antidepressant in the strict sense—that is, to be applicable in major depression—hypericum extracts should meet the same rigorous demands that have been laid down for synthetic antidepressants.8 The serious nature of major depression would exclude the special licensing route that has been proposed for herbs used in minor illnesses.9 Linde et al list several short term trials in which hypericum monopreparations were more effective than placebo in less severe forms of depressive disorders. The methodological quality of these studies is variable,10 and most lack diagnostic precision because they used no classification system for psychiatric disorders or only a global classification such as the international classification of diseases (ICD-9).7 Only one research group (Hansgen et al) evaluated a hypericum monopreparation in a homogeneous group of patients with a mild to moderate form of major depression according to accepted diagnostic criteria.11 This study was republished in 1996 with 36 additional patients.12 After four weeks of treatment, hypericum extract was more effective than placebo, with a relevant difference on the Hamilton depression scale.12 Positive results were also found in another recent trial, which compared a different brand of hypericum extract with placebo for six weeks in patients with a moderate depressive episode (according to ICD-10).13 Whether the observers in these multicentre trials received joint training to minimise variations in rating and whether diagnoses were reached by structured interview were not documented.

    Although promising, these studies are not sufficient to accept the use of hypericum extract in major depression when judged against the criteria of the European Union guidelines for regulatory evaluation of antidepressants.8 For example, specific studies in severely depressed patients are still missing, as are longer term studies to assess the risk of relapse and the possibility of late side effects. Of the four trials that have compared a hypericum monopreparation with a synthetic antidepressant, none lasted longer than six weeks. All tested the comparator drug in daily amounts below or at the lower end of the usual dose range, and none of the reports described a homogeneous patient group with major depression defined by the criteria of the Diagnostic and Statistical Manual of Mental Disorders.7 Like synthetic antidepressants, hypericum extract needs two to four weeks to develop its mood elevating effects.

    The herb may offer an advantage, however, in terms of relative safety and tolerability, which might improve patient compliance. So far, its extensive use in Germany has not resulted in published case reports about serious drug interactions or toxicity after overdose. In an open study of 3250 patients the most commonly noted side effects were gastrointestinal symptoms (0.6%), allergic reactions (0.5%), and fatigue (0.4%). Whether the allergic reactions could have been related to the photosensitising potential of hypericin was not investigated. The study lasted only four weeks, and laboratory values were not systematically monitored.14

    All in all, it is for good reasons that Linde et al ask for additional studies with better delineation of participants and for trials lasting longer than eight weeks. When future studies compare hypericum extracts with a synthetic antidepressant this comparator should be tested in therapeutic doses and the size of the trial should be large enough to detect a clinically meaningful difference.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.