Hypericum perforatum


St John’s wort (Hypericum perforatum) is a native flowering plant of Europe and Asia which produces attractive yellow flowers. According to Kiple and Ornelas (2000) its lemon-scented leaves have been used for thousands of years as human food and have also been used to make a form of tea. Extracts of the flowers and leaves of this plant are now widely taken in the belief that they are mood enhancing and have beneficial effects in the treatment of clinical depression. In Germany hypericum extracts are widely prescribed by physicians for the treatment of clinical depression and it is the best selling antidepressant there.

What is depression?

Clinical depression is a common, painful and disabling condition which is more severe than the normal downward fluctuations in mood that we all regularly experience. The American Psychiatric Association lists the following symptoms for depression:

• Depressed mood

• Loss of interest in and lack of pleasure derived from activities that the patient usually finds pleasurable

• Disturbed sleep patterns

• Abnormal activity patterns, either agitation or being uncharacteristically inactive

• Loss of drive and energy, loss of sex drive and reduced appetite

To make a formal diagnosis of clinical depression the first two of these symptoms must be present as well as most of the others. These symptoms should have been present for at least two weeks and should not be attributable to other disease, to drug use or be associated with bereavement. As many as one in five adults may be affected by depression during the course of their lives and rates are much higher in women than men.

Conventional drug treatment of depression

Medicinal drugs used to treat depression work by raising the amounts of serotonin (5HT or 5-hydroxytryptamine) and/or noradrenaline in synapses of the central nervous system. Monoamine oxidase (MAO) inhibitors such as iproniazid work by blocking the enzymes that are responsible for the breakdown of several nerve transmitters including noradrenaline and serotonin. These drugs have now been largely superseded by other drugs with less side-effects. People taking MAO inhibitors were required to avoid consuming foods, such as mature cheese and red wine, that contain a substance called tyramine which could precipitate large and dangerous rises in blood pressure in those taking MAO inhibitors. Tricyclic antidepressants, such as imipramine, inhibit the reuptake of both serotonin and noradrenaline into the presynaptic terminals which is the major route for curtailing the actions of these transmitters after release. Together with the monoamine oxidase inhibitors, the tricyclic antidepressants are termed first generation antidepressants. The selective serotonin reuptake inhibitors (SSRIs), such as fluoxetine and paroxetine, are examples of the so-called second generation antidepressants and, as their name suggests, they selectively inhibit the reuptake of serotonin into pre-synaptic terminals thus specifically increasing its actions. SSRIs are now by far the most widely used drugs in the treatment of depression; they have fewer side-effects than the first generation tricyclics. A readable account of antidepressants and their likely modes of action can be found in Parrott et al. (2004). The guidelines of the British Association for Psychotherapy for the diagnosis and treatment of depression can be found in Anderson et al. (2000).

Possible actions of Hypericum extracts

Hypericum extracts contain well over twenty bioactive substances, including the chemically complex polyphenols, hypericin and pseudohypericin, and the substance hyperforin which is described as a prenylated phloroglucinol (the chemical structures of hypericin and hyperforin may be found in Barnes 2002c). Early studies in the 1980s suggested that hypericin was an monoamine oxidase inhibitor and this led to widespread assumptions that hypericin was the active agent responsible for the antidepressant effects of Hypericum extracts, and that MAO inhibition was its mode of action. Later studies failed to confirm this effect of hypericin and have further suggested that hyperforin is probably the component responsible for most of the antidepressant effects of the extract. It is no longer believed that Hypericum extracts have significant monoamine oxidase inhibitory effects. This early assumption that hypericin was the key ingredient has resulted in the standardisation of commercial preparations of Hypericum according to their hypericin content, probably not to the active component(s). Typically St John’s wort tablets claim to supply between 300 and 1200 µg of hypericin per tablet. When a sample of tablets, purchased in London, was analysed they were found to contain between a third and two-thirds of the stated hypericin content even when the total hypericin and pseudohypericin content was used. This means that not only are Hypericum extracts standardised to the wrong component, but also that even the claimed content of this ingredient often does not accurately reflect the true content. Biochemical and pharmacological studies suggest that hyperforin acts as a non-selective reuptake inhibitor in the brain. It not only inhibits the reuptake of serotonin and noradrenaline but also that of other brain transmitters and does it in a different way to other antidepressant drug. Other reuptake inhibitors work by competing with the monoamine transmitter for the carrier molecules responsible for reuptake. Hyperforin, however, non-competitively inhibits the uptake of several monoamines by affecting the sodium transporting system and raising the intracellular concentration of sodium in the pre-synaptic terminals. The reuptake of monoamine transmitters is sodium dependent: it requires a low intracellular sodium concentration.

Testing the antidepressant effects of Hypericum extracts

There have been dozens of clinical trials of Hypericum extracts that have tested its effects against a placebo and/or against first generation antidepressant drugs such as imipramine. There have been few trials comparing its effectiveness with second generation SSRIs. When testing the effectiveness of antidepressant treatments and other treatments for psychiatric or psychological problems there are two major difficulties as listed below.

• There is almost always a large but variable placebo effect in such studies. At least a quarter of patients usually respond positively to the dummy treatment. This makes it more difficult to demonstrate a statistically significant effect of treatment and makes it particularly difficult to demonstrate a statistically significant difference between the two moderately effective treatments.

• There are no objective measures of treatment efficacy such as a change in a blood parameter. The severity of depression is measured using a numerical scale that is based upon patients’ responses to a number of questions about the severity of a list of symptoms (for example, the Hamilton depression scale). This is measured before and at various times after the start of treatment. Other measures that may be used include the physician’s and the patient’s global assessment of the change in severity of their condition using a sliding scale from, for example, ‘very much improved’ to ‘very much worse’. The subjective nature of the outcome measures makes it particularly important that the double-blinding of the different treatments is rigorous.

Does it work and is it safe?

Despite the dozens of clinical trials of St John’s wort conducted over the past two decades, including several large multi-centre studies, it is still not possible to make a definitive judgement on the usefulness and safety of Hypericum extracts for the treatment of mild to moderate depression. The confusion over the information available to members of the general public about St John’s wort is exemplified from the following headlines taken from the BBC website:

• 10/12/1999 Herb ‘helps ease depression’

• 1/3/2000 St John’s wort warning (relating to its possible interaction with prescription drugs)

• 31/8/2000 Herb ‘as effective as antidepressants

• 9/4/2002 Herb ineffective as antidepressant

• 11/2/2005 Herb ‘as good as depression drug’

There is fairly general agreement that it is not an appropriate treatment for severe depression, and in general that self-medication for severe depression is not appropriate because of the high suicide risk of sufferers. There also seems to be a consensus that the acute side-effects experienced by those taking Hypericum extracts are less than those experienced by patients taking the older tricyclic antidepressants such as imipramine. However, in March 2000, the UK Department of Health issued a warning about the possible dangers of combining Hypericum extracts with several prescription drugs. This was based upon evidence submitted to it from the independent Committee on the Safety of Medicines. Patients were advised to tell their doctor or pharmacist if they were taking St John’s wort and a prescription medicine. St John’s wort induces detoxification enzymes in the liver which can increase the rate at which a number of drugs are metabolised and thus render them less effective, and it may interact with other drugs in different ways (for example SSRI antidepressants). Hypericum extracts should not be used together with:

Anticoagulant drugs such as warfarin

• The heart drug digoxin

• Oral contraceptives

• Anti-rejection drugs such as cyclosporine

• Drugs used in the treatment of HIV infection

• Anti-convulsants used to treat epilepsy

• A number of drugs used to manage migraine

• Some anti-asthmatic drugs

• SSRI antidepressants.

There are isolated reports that Hypericum extracts may increase photosensitivity some preparations now carry a warning to avoid direct sunlight exposure when taking St John’s wort. This could be a particular problem for people suffering from seasonal affective disorder (SAD) who are also being treated with light therapy.

Before 2000, the bulk of the many published clinical trials supported the proposition that St John’s wort did have beneficial effects in treating mild to moderate depression. Many of these studies were conducted in Germany where St John’s wort was and is the biggest selling antidepressant ‘drug’. These studies generally concluded that St John’s wort was more effective than a placebo and of comparable efficacy to older antidepressant drugs such as imipramine, and had less side-effects and lower drop-out rates than with these tricyclic antidepressants. Several systematic reviews and meta-analyses of clinical trials support these general conclusions. Many reviewers and commentators criticised these early clinical trials for a variety of reasons such as those listed below.

• Many of these early trials were of short duration (often only about four weeks) and so the longer term effectiveness could not be determined.

• Some of these trials used less than optimal dosing of antidepressant drugs.

• Variability of dosing and lack of standardisation for hyperforin in the different trials.

• Inadequate matching of base-line severity and several studies used patients who did not meet current formal diagnostic criteria for clinical depression.

• Some studies did not use placebo controls despite the known very high level of placebo effect in such studies. For example Woelk et al. (2000) compared the effects of St John’s wort and imipramine and reported no discernible difference in their efficacy; this study caused great impact in the UK when it was published but did not include a placebo group.

• Some studies were said to have used inadequate outcome measures.

• Most comparative studies had compared St John’s wort with older first generation antidepressants rather than with modern SSRIs.

Barnes (2002) noted that a Cochrane review of 27 randomised controlled trials of St John’s wort extracts in patients with ‘neurotic depression’ and mild to moderate clinical depression lasting from four to twelve weeks had results consistent with the earlier summary of the consensus from these early trials. However, Barnes goes on to say that another meta-analysis conducted around the same time using stricter criteria for inclusion had included only six trials of patients who met formal diagnostic criteria for clinical depression. Four had compared St John’s wort with a placebo and two with a tricyclic antidepress-ant. Although the overall direction of the findings were similar in the two meta-analyses, the differences in the number of trials meeting the different inclusion criteria suggest that the ‘quality’ of many of these early trials was low.

In 1999, three major well-funded clinical trials of St John’s wort were started in the USA, one of them funded by the National Institutes of Health (NIH). The NIH-funded study was conducted by the same research group at Duke University Medical Center that published one of the positive meta-analyses of earlier St John’s wort trials but the conclusions were much less positive. The Hypericum Depression Trials Study Group (2002) had three groups each with over 100 patients; one was treated with hypericum, one given a placebo and the third group given sertraline an SSRI antidepressant. The first phase of the trial lasted for eight weeks but patients who responded to initial treatment were offered a further 18 weeks of treatment in order that longer term effects of treatment could be monitored. Neither the St John’s wort nor the SSRI antidepressant produced statistically significant differences to the placebo when outcome was assessed using the Hamilton Depression Scale. The SSRI did produce a bigger improvement in the Clinical Global Impression – Improvements scale than either St John’s wort or the placebo. The authors suggest that it is not uncommon in trials of antidepressant drugs for the active treatment not to produce a significant effect upon Hamilton score ratings, which rather undermines the criticism of some early trials of St John’s wort which have not used this outcome measure. To a non-psychiatrist, these data are hardly a ringing endorsement of either treatment but they do emphasise the importance of a placebo control.

Werneke et al. (2004) attempted to sum up the position on the efficacy of St John’s wort after the publication of the three recent large and largely negative trials. They reproduced a meta-analysis based upon literature searches conducted in June 2000 and then re-analysed this data adding to it the results of the three most recent studies. Addition of the more recent data substantially reduced the apparent effect of St John’s wort. They concluded that it may be less effective for treating depression than the earlier studies had suggested and that if future trials follow the trend set by the other more recent trials it may finally be shown to be ineffective. These negative conclusions are consistent with much of what has appeared in the British and American literature regarding St John’s wort since 2002. However a meta-analysis of 20 clinical trials of St John’s wort recently published in German still concludes that St John’s wort is more effective than placebos and of similar effectiveness to synthetic antidepressants with lower drop-out rates. They supported the positive view of the German health authority that St John’s wort is a first-line treatment for milder forms of depression. A recent trial conducted in Germany but published in English compared Hypericum extracts to the second generation antidepressant drug (SSRI) paroxetine in patients with moderate to severe depression. They concluded that Hypericum was at least as effective as paroxetine and that it was better tolerated.

St John’s wort is marketed in Britain and the USA as a dietary supplement but is a prescription drug in Germany; it has recently been banned for over the counter sale in Ireland and is only available there on prescription. The available data suggest that, even if it does have some benefit in treating depression, it should not be taken with other prescription medications and probably should not be used as a routine, long-term supplement for people not suffering from depression. It should be used as a medicine of herbal origin rather than a dietary supplement. It was included in this chapter largely because of the scale of its usage, with perhaps as many as two million Britons having tried it at some time.