St. John's wort and depression
What 7 graded rows on st. john's wort agree on for depression, and where they differ. Counted side by side, never averaged.
Consensus finder Rows from the St. John's wort page, reviewed Oct 21, 2026 How it decides
What the rows agree on
Each block and row is one graded row from the St. John's wort page, with its PubMed record.
Your question
St. John's wort and depression
7 rows · 15,280 peopleMostly an improvement: 5 of 7 rows, and most of the people. 2 found no difference.
- BGrade B× 4
- CGrade C× 1
- ∅No effect× 2
- Rows
- 7
- Trials
- 74
- At least: a review without a trial count
- People
- 15,280
- Pooled reviews may count some twice
Where they differ
From the rowsThe best-graded row, grade B, is a systematic review of 5,489 people (2008) and found an improvement. 2 rows are graded No effect, our grade for adequately powered trials that found no effect.
The 5 rows that found an improvement were larger: a median of 2,922 people, against a median of 270 people for the 2 rows that found no difference.
Only the 5 rows that found an improvement include pooled reviews of several trials (4 of 5); the 2 rows that found no difference are single studies.
The 5 rows that found an improvement are newer (2005 to 2023) than the 2 rows that found no difference (2001 to 2002).
The rows
5 improved · 2 no difference- BGrade BMajor depression (response) · Improved
Systematic review, 29 studies · n 5,489 · 2008 · Adults with major depression in 29 trials
Varied across trials · Varied
Larger placebo-controlled trials: response rate ratio 1.28 (95% CI 1.10–1.49); versus SSRIs, response was similar (RR 1.00). Results from German-speaking countries were more favourable.
Row on the St. John's wort page · PubMed 18843608 (opens PubMed)
- BGrade BDepression response compared with SSRIs · Improved
Meta-analysis, 27 studies · n 3,808 · 2017 · Patients with depression in 27 trials comparing St. John's wort with SSRIs
Varied across trials · 4–12 weeks
Response (pooled RR 0.983) and remission (RR 1.013) were similar to SSRIs, and dropout was lower with St. John's wort (OR 0.587).
Row on the St. John's wort page · PubMed 28064110 (opens PubMed)
- BGrade BMild to moderate depression · Improved
Systematic review · n 2,922 · 2016 · Adults with mild to moderate major depressive disorder (placebo-controlled comparisons)
Varied across trials · Varied
More responders than placebo (RR 1.53, moderate-quality evidence, high heterogeneity); similar effectiveness to antidepressants with fewer adverse events.
Row on the St. John's wort page · PubMed 27589952 (opens PubMed)
- BGrade BModerate to severe major depression versus paroxetine · Improved
RCT · n 251 · 2005 · Adult outpatients with acute major depression (HAM-D ≥22) in 21 German practices
900 mg/day, raised to 1,800 mg/day in early non-responders (paroxetine 20–40 mg/day) · 6 weeks
HAM-D fell 14.4 points with WS 5570 and 11.4 with paroxetine, meeting non-inferiority and superiority margins; adverse events were less frequent (0.035 vs 0.060 events per day of exposure).
Row on the St. John's wort page · PubMed 15708844 (opens PubMed)
- CGrade CDepression compared with SSRIs and placebo · Improved
Meta-analysis, 14 studies · n 2,270 · 2023 · Patients with depression in 14 trials published 2000–2022
Varied across trials · Varied
The authors concluded that St. John's wort lowered HAM-D depression scores with fewer side effects than SSRIs; the headline pooled odds ratio was 2.44 (95% CI 1.33–4.45).
Row on the St. John's wort page · PubMed 36226689 (opens PubMed)
- ∅No effectModerately severe major depression versus placebo and sertraline · No difference
RCT · n 340 · 2002 · Adult outpatients with major depression and a HAM-D score of at least 20 (12 US clinics)
900–1,500 mg/day · 8 weeks
HAM-D fell by 8.68 points with St. John's wort and 9.20 with placebo (P = .59); full response 23.9% vs 31.9%. Sertraline also failed to beat placebo on the primary outcomes.
Row on the St. John's wort page · PubMed 11939866 (opens PubMed)
- ∅No effectMajor depression of at least moderate severity (HAM-D ≥20) · No difference
RCT · n 200 · 2001 · Adult outpatients with major depression and a baseline HAM-D score of at least 20 (11 US academic centres)
900 mg/day, raised to 1,200 mg/day after 4 weeks if response was inadequate · 8 weeks
No difference from placebo in the rate of change on the HAM-D or in response; remission was higher (14.3% vs 4.9%) but rare. Headache was more common with the herb (41% vs 25%).
Row on the St. John's wort page · PubMed 11308434 (opens PubMed)