Skip to main content

Table 3 Summary of findings and quality of evidence table

From: A systematic review of St. John’s wort for major depressive disorder

Outcome Study design: number of studies, number of participants Findings: direction and magnitude of effect GRADE
Comparison: SJW vs. placebo
 Depression, number of treatment responders 18 RCTs, N = 2922 RR 1.53 (1.19, 1.97), favors SJW Moderatea
 Depression scale score 16 RCTs, N = 2888 SMD 0.49 (0.23, 0.74), favors SJW Moderatea
 Depression remission 9 RCTs, N = 1419 RR 1.69 (0.63, 4.55), n.s. Lowa,b
 Depression relapse 1 RCT, N = 426 RR 0.70 (0.49, 1.02), n.s. Very lowa,b
 Quality of life—mental 2 RCTs, N = 358 SMD 0.48 (0.24, 0.73), favors SJW Lowb
 Quality of life—physical 2 RCTs, N = 358 SMD 0.28 (−1.03, 0.47), n.s. Very lowa,b
 Number of patients with adverse events 13 RCTs, N = 2600 OR 0.83 (0.62, 1.13), n.s. Moderateb
 Serious adverse events 6 RCTs, N = 1358 OR 0.26 (0.04, 1.23), n.s. Moderatea
 Gastrointestinal/metabolic/nutritional adverse events 16 RCTs, N = 2490 OR 1.06 (0.83, 1.41), n.s. Lowa,b
 Neurologic/nervous system adverse events 14 RCTs, N = 2404 OR 1.56 (1.08, 3.32) SJW more AE Lowa,b
 Skin/musculoskeletal adverse events 10 RCTs, N = 1978 OR 1.47 (0.98, 2.21), n.s. Very lowa,b
 Photosensitivity 4 RCTs, N = 1054 OR 1.10 (0.36, 3.56), n.s. Lowa,b
 Respiratory/infectious adverse events 7 RCTs, N = 1081 OR 1.48 (0.95, 2.33), n.s. Lowa,b
 Other organ system (eye, ear, liver, renal, reproductive) adverse events 5 RCTs, N = 1054 OR 1.87 (1.08, 3.32), SJW more AE Lowa,b
 Cardiovascular adverse events 4 RCTs, N = 759 OR 6.81 (0.92, 304.08), n.s. Very lowa,b,d
 Psychiatric adverse events 3 RCTs, N = 608 OR 1.61 (0.34, 10.21), n.s. Very lowa,b,d
 Sexual dysfunction adverse events 2 RCTs, N = 428 OR 1.92 (0.94, 4.00), n.s. Very lowa,b,d
Comparison: SJW vs. antidepressant
 Depression, number of treatment responders 17 RCTs, N = 2776 RR 1.01 (0.90, 1.14), n.s. Moderateb
 Depression scale score 14 RCTs, N = 2248 SMD 0.03 (−0.15, 0.21), n.s. Moderateb
 Depression remission 7 RCTs, N = 787 RR 1.17 (0.84, 1.62), n.s. Lowb
 Depression relapse 1 RCT, N = 241 RR 4.17 (0.47, 33.33), n.s. Very lowa,b
 Quality of life—mental 1 RCT, N = 216 SMD −0.11 (−0.15, 0.38), n.s. Very lowa,b
 Quality of life—physical 1 RCT, N = 153 SMD 0.35 (0.01, 0.70), favors SJW Very lowa,b
 Number of patients with adverse events 11 RCTs, N = 1946 OR 0.67 (0.56, 0.81), favors SJW Moderatea
 Serious adverse events 4 RCTs, N = 805 OR 0.62 (0.05, 5.46) n.s. Lowa,b
 Gastrointestinal/metabolic/nutritional adverse events 15 RCTs, N = 2491 OR 0.43 (0.34, 0.55) favors SJW Lowa,b
 Neurologic/nervous system adverse events 15 RCTs, N = 2492 OR 0.29 (0.24, 0.36) favors SJW Lowa,b
 Skin/musculoskeletal adverse events 10 RCTs, N = 1587 OR 1.18 (0.79, 1.78) n.s. Lowa,b
 Respiratory/infectious adverse events 2 RCTs, N = 352 OR 1.25 (0.70, 2.25) n.s. Very lowa,b
 Other organ system (eye, ear, liver, renal, reproductive) adverse events 4 RCTs, N = 761 OR 0.85 (0.52, 1.38), n.s. Lowa,b
 Cardiovascular adverse events 5 RCTs, N = 750 OR 0.55 (0.26, 1.16), n.s. Lowa,b
 Psychiatric adverse events 4 RCTs, N = 552 OR 0.41 (0.19, 0.87) favors SJW Very lowa,b
 Sexual dysfunction adverse events 2 RCTs, N = 301 OR 0.51 (0.30, 0.88) favors SJW Lowa,b
Effect of depression severity
 Depression, treatment responders 18 RCTs, N = 2922 Meta-regression did not suggest differences between patient subgroups
(p = 0.798)
Very lowa,c
 Depression scale score 16 RCTs, N = 2888 Meta-regression did not suggest differences between patient subgroups
(p = 0.365)
Very lowa,c
 Depression remission 9 RCTs, N = 1507 Meta-regression did not suggest differences between patient subgroups
(p = 0.159)
Very lowa,c
 Number of patients with adverse events 13 RCTs, N = 2600 Meta-regression did not suggest differences between patient subgroups
(p = 0.480)
Very lowa,c
  1. Quality of evidence was downgraded (by 1 or by 2, depending on the severity) for the following
  2. aInconsistency (heterogeneity, direction of effects; no replication)
  3. bStudy limitations (no good quality study; effect not present when excluding poor quality studies; studies not designed or not powered to assess outcome)
  4. cIndirectness (no head-to-head trials, effect based on indirect comparison)
  5. dImprecision (wide confidence interval)
  6. SJW St. John’s wort, n.s. not statistically significantly different, AE adverse event