St. John's Wort for Depression: Which Extracts Help and Which Medicines Interact?

What is St. John's wort?

St. John's wort (Hypericum perforatum) is a yellow-flowered herb sold as tea, dried plant material and concentrated extracts.

People take St. John’s wort mainly for symptoms of depression. If you take medicine, the herb can interfere with contraceptive pills, transplant medicines or anticoagulants.

How did St. John's wort get its name, and how was it used?

Hold a leaf up to the light and its translucent glands look like tiny holes, hence the species name perforatum. In Europe its yellow flowers appear around June 24, the feast of St. John the Baptist. People hung sprigs over doorways, put them under pillows or wore them around the neck for protection against evil. Another tradition connected the red juice of the bruised plant with the saint's blood. Cambridge University Botanic Garden^[1] · Museum of the Order of St John · USDA plant guide^[2]

The Roman author Pliny described seeds of a plant called hypericon given in wine for sciatica and associated its seeds with recurring fevers. He distinguished several kinds. ^[3]

Early English records tell us how differently people used the plant. London's official pharmacopoeia of 1618 described an oil made from crushed plant tops with olive oil, wine and other ingredients. Nicholas Culpeper described applying an ointment to wounds, and in 1718 the apothecary John Quincy recorded the oil's use for bruises. Royal College of Physicians' recipe collection^[4] · Culpeper's Complete Herbal^[5]

A modern trial tested an ointment on cesarean incisions. The researchers reported better wound scores on day 10 and better scar scores on day 40 than with placebo ointment or no ointment. That finding concerns the preparation applied to a surgical incision. ^[6]

Why do people take St. John's wort?

Menopause, anxiety and seasonal depression

In one study, 80 postmenopausal women took tablets containing St. John's wort or placebo for two months. Seventy finished. Those taking the herb reported fewer and less intense hot flashes, and their depression ratings improved more. The published abstract gives the tablet amount in micrograms without enough information to match a product. In a separate trial of 47 women, hot flashes fell by an average of 2.3 per day with the extract and 1.0 with placebo after 12 weeks. The difference could be due to chance. Participants taking the extract reported better sleep and menopause-related quality of life. Together these studies leave real uncertainty about dependable relief from hot flashes or night sweats. Two-month trial^[13] · Twelve-week pilot^[10]

A larger menopause study found symptom relief with a combination of St. John's wort and black cohosh. It cannot tell us how much either ingredient contributed. ^[14]

In a 12-week trial of 40 adults with generalized social anxiety, those taking St. John's wort improved no more on the main anxiety scale than those taking placebo. In a seasonal depression study, all 20 participants received the herb and were assigned to bright or dim light. Both groups improved. Without a group receiving no herb, the trial couldn't separate its effect from other changes over the four weeks. Social anxiety trial^[11] · Seasonal depression study^[12]

Premenstrual symptoms and other uses

In a crossover trial, 36 women with mild premenstrual syndrome took a 900 mg daily extract and placebo for two cycles each. Their daily diaries showed greater relief of physical and behavioral symptoms on the herb. Mood symptoms and pain did not improve more than with placebo. A second trial, with 125 women completing two treatment cycles, found no clear advantage on its main measure of premenstrual anxiety symptoms. Crossover trial^[15] · Parallel trial^[16]

Placebo-controlled studies found no clear benefit for ADHD symptoms in children, obsessive-compulsive symptoms in adults, or quitting smoking. In a 12-week irritable bowel syndrome study, the placebo group actually reported better bowel symptom scores than the group taking the herb. ADHD trial^[17] · OCD trial^[18] · Smoking cessation trial^[19] · IBS trial^[20]

One six-week trial found a different result in people diagnosed with specific somatoform disorders, conditions involving persistent physical symptoms. In the analysis of 173 outpatients, about 45% responded to LI 160 compared with 21% taking placebo. The researchers excluded major depression and studied specific diagnoses. Someone with new or unexplained physical symptoms still needs an assessment of their cause. ^[21]

Common dosage and choosing an extract

There is no recommended daily intake for St. John's wort. 900 mg per day is one research dose. In a six-week trial of 375 adults with mild to moderate depression, researchers gave a specific extract called WS 5570 in three 300 mg doses each day. More participants improved with that preparation than with placebo. LI 160, another extract, was tested at 900 to 1,500 mg per day in a different group and fell short of its main outcome. WS 5570 trial^[22] · LI 160 trial^[9]

Look for the species, plant part, named extract, amount per full daily serving and any declared amounts of hypericin and hyperforin. These are different substances. A label standardized to 0.3% hypericin would supply 2.7 mg of hypericin in 900 mg of extract. That number tells you nothing about the amount of hyperforin. Some research extracts had several percent hyperforin. In one six-week trial, depression scores fell an average of 10.3 points with a 5% hyperforin extract, 8.5 points with a 0.5% extract and 7.9 points with placebo. The researchers found an overall trend and a clear difference between the 5% extract and placebo. This comparison leaves the size of a direct advantage over the lower-strength extract uncertain. ^[23]

Does low hyperforin change the decision?

A six-week trial in 324 adults with mild to moderate depression compared 250 mg of the extract Ze 117 twice daily with imipramine. Average depression scores improved similarly. Four of 157 participants assigned to the herb stopped because of adverse effects, compared with 26 of 167 assigned to imipramine. The comparison had no placebo group. ^[24]

A separate experiment asked whether a low-hyperforin Ze 117 product altered how volunteers processed test medicines. Each 500 mg tablet supplied 0.96 mg of hyperforin per day. Nineteen volunteers finished. Researchers measured how much of each medicine was present in the blood over time, before and after repeated doses of Ze 117. For several medicines, including midazolam and fexofenadine, the results stayed within the range the researchers had set for a comparable amount. For dextromethorphan, the amount rose by about 48%. These were healthy volunteers taking selected test drugs. The study didn't test protection against pregnancy, transplant rejection or blood clots. A study of 96 volunteers compared products more directly using the heart medicine digoxin. After 14 days, hyperforin-rich LI 160 lowered digoxin exposure by about 25%. Several other preparations, including a hyperforin-free extract, showed no clear interaction with digoxin under those conditions. Bring the actual label to your pharmacist even if it says “low hyperforin.” Digoxin product comparison^[25] Ze 117 interaction study^[26]

How should benefit be assessed?

Persistent low mood, loss of interest, or changes in sleep and daily functioning deserve an assessment. Tell the clinician about any past period of unusually high energy or very little need for sleep.

If you and your clinician choose treatment, agree on what improvement would look like in daily life and when to check it. Most of the evidence here concerns changes over several weeks. Ask about the next step if symptoms persist or worsen.

Drawbacks and side effects

A recent case report also describes a young adult who developed manic and psychotic symptoms, including very little sleep and delusions, while taking St. John's wort. Individual reports cannot tell us how often this happens.

Why the medicine and the timing matter

Hyperforin activates a receptor that signals liver cells to make more of a drug-processing enzyme, CYP3A4. Some medicines also depend on a transporter called P-glycoprotein, which can move them back into the intestine. When a preparation increases the activity of these systems, less medicine may reach or remain in the blood. Spacing doses by a few hours cannot reliably prevent a change in the amount of enzyme the body makes. The serotonin interaction works differently. Combining substances that increase serotonin activity can produce toxicity even when a medicine's level hasn't fallen. Human liver-cell experiment^[41] · Human interaction study^[30] · Medsafe guidance^[35]

The strongest reason to act depends on the medicine. The pill experiment found more bleeding but no evidence of ovulation. Separately, the UK regulator received 19 suspected pill or implant interactions, including 15 unplanned pregnancies. Those reports establish a reason for caution without telling an individual user the odds. In 12 healthy volunteers, St. John's wort reduced blood exposure to rivaroxaban and lowered its measured anticoagulant activity. Researchers measured the effect on a blood sample rather than counting strokes or clots. The cyclosporine reports, by contrast, documented a clinical event in actual transplant recipients. Contraceptive trial^[27] · UK safety reports^[28] · Rivaroxaban trial^[30] · Transplant reports^[31]

The size of the interaction can be substantial. In a volunteer study, indinavir exposure fell 57%. In five cancer patients, exposure to SN-38, the active breakdown product of irinotecan, fell 42%. These experiments raise concerns about losing treatment effect. Indinavir study^[32] · Irinotecan study^[33]

Stopping also changes the equation. After 12 healthy men had taken the herb for 14 days, researchers tracked how fast they cleared midazolam. The extra enzyme activity took roughly 46 hours to decline by half, and clearance approached its starting level after about a week. As induction fades, another medicine's level can rise, especially if its dose was increased while the herb was being taken. This small study cannot supply a universal washout or taper. Ask the prescriber how to monitor your medicine through the change.

What does the depression research show?

Across 35 studies enrolling 6,993 adults, the overall picture was most favorable for short-term treatment of mild to moderate major depression. In the ten placebo comparisons specific to that severity range, participants taking St. John's wort were about 45% more likely to meet the studies' definition of response (relative risk 1.45, 95% confidence interval 1.09 to 1.92). That is a relative difference. The number of additional people helped depends on how many improve with placebo. Results varied considerably by trial. Pooled results for remission, in which few symptoms remain, were too imprecise for a clear conclusion. Systematic review^[7] · RAND companion report from the same review^[42]

Participants on particular extracts had similar average response to those on antidepressants and reported fewer routine side effects. One six-week, manufacturer-supported trial in 251 adults with moderate to severe major depression asked whether WS 5570 performed at least about as well as paroxetine. The researchers specified in advance how much worse the herb could perform and still meet that standard. It met the standard, and average improvement favored the herb. Average Hamilton depression scores fell 14.4 points with the herb and 11.4 with paroxetine. Because the study had no placebo group, it cannot show how much of either group's improvement was due to treatment. Systematic review^[7] · WS 5570 and paroxetine trial^[8]

Other results temper that finding. In a US trial involving 340 adults with substantial depressive symptoms, neither LI 160 nor sertraline clearly beat placebo on the main outcomes. Another US placebo-controlled trial in 200 patients found no clear advantage for St. John's wort on its main depression score or response rate, although more patients taking the extract reached remission on a secondary measure. These mixed findings matter most when someone with severe symptoms is considering treating themselves or changing a prescription. Three-arm trial^[9] · Second US trial^[43]

Does improvement last?

One longer trial enrolled people with recurrent depression who had already improved after six weeks on WS 5570. During the next 26 weeks, 51 of 282 people who continued the extract relapsed (18%), compared with 37 of 144 who switched to placebo (26%). The review estimated a relative risk of 0.70, with a 95% confidence interval of 0.49 to 1.02, which leaves open the possibility of no difference. A separate analysis of how soon people relapsed favored the extract numerically but missed the trial’s stated statistical threshold. Those results concern people who already responded to this extract. ^{[44] [7]}

What do the laboratory findings explain?

In laboratory preparations of nerve endings, hyperforin reduced the uptake of serotonin, norepinephrine and dopamine. Hypericin did not show that effect in the same experiment. Researchers also observed receptor changes in rats. These findings suggest ways the plant could affect signaling. The laboratory findings cannot tell us which changes, if any, relieved symptoms in the people who took the extract. The clinical trials measured how people felt and functioned. Laboratory study^[45] · Clinical review^[7]

The bottom line

For mild to moderate depression, some named extracts helped more adults improve over several weeks than placebo did. The evidence is less dependable for severe depression and for hot flashes, social anxiety or seasonal symptoms. If you're considering St. John's wort and you're routinely taking medication that might conflict with it, it's advisable to discuss it with your doctor first. If you're already taking it with a transplant medicine, anticoagulant or hormonal contraceptive, tell the relevant care team promptly so they can plan a safe change and any monitoring you need.

Source notes

• NIH guidance on St. John's wort^[37]
• ^[7]
• Original placebo-controlled WS 5570 study^[22] · LI 160 study^[9]
• ^[26]
• Transplant case reports^[31] · Recovery after stopping^[46]

Help us improve this article

Spotted an error or found a source we should consider? Tell us which part of this article needs another look and include a link if you have one. Please leave out personal health details and lab results; we can’t answer individual medical questions.

Your name

Email address

What would you like to share?Select a topicCorrectionSource or research suggestionArticle requestAccessibility issueBusiness inquirySomething else

Your message

Continue to email →Email us directly