Six RCTs were included in the review. Two trials with 169 participants were placebo controlled, and four trials with 482 participants compared Hypericum with tricyclic antidepressants.
For Hypericum versus placebo:
The scores for the Hypericum groups showed improvement as did the scores for the placebo groups. The rates of change for antidepressant effects in each of the studies were more dramatic for Hypericum, with lower HAMD scores at study conclusion. Due to heterogeneity (chi-square 9.76, p < 0.001) the rates of change in two trials of Hypericum versus placebo could not be combined.
Response rates were homogeneous (chi-square 3.15, p = 0.076). The pooled response rate for Hypericum was significantly more effective than placebo with 73.2% responding to Hypericum and 37.9% responding to placebo (fixed RR = 1.48, 95% CI: 1.03, 1.92).
The drop-out rates were homogeneous (chi-square 1.07, p = 0.30). Hypericum showed a lower patient drop-out rate than placebo, with 12.2% dropping out from the Hypericum groups and 19.5% dropping out from the placebo groups (fixed RR = 0.53, 95% CI: 0.12, 0.93). Only one of the studies reported side effects and these were minimal.
For Hypericum versus standard antidepressants (TCAs):
The HAMD scores for the Hypericum groups showed improvement in each of the four studies as did the scores for the TCA groups. The rates of change for antidepressant effects were not homogeneous at 4 weeks, (chi-square 42.63, p < 0.001) nor at 6 weeks (chi-square 19.98, p < 0.001).
Response rates were homogeneous (chi-square 1.01, p = 0.60). The pooled response rate for Hypericum was significantly more effective than placebo with 64.0% responding to Hypericum and 66.4% responding to the TCAs (fixed RR = 1.11, 95% CI: 0.92, 1.29; random RR = 0.98, 95% CI: 0.67, 1.28).
The drop-out rates were homogeneous (chi-square 2.53, p = 0.47). Hypericum showed a lower patient drop-out rate than the TCAs, with 12.6% dropping out from the Hypericum groups and 16.2% dropping out from the TCA groups (fixed RR = 0.65, 95% CI: 0.36, 0.94). Reported reasons for drop-outs included intolerable side effects and insufficient efficacy, with either worsening of depression or no change in condition.
The side-effect rates were homogeneous (chi-square 1.72, p = 0.65). The TCAs were 1.7 times more likely to cause side-effects than Hypericum (fixed RR = 1.72, 95% CI: 1.30, 2.14). Nearly 47% of the TCA group reported side-effects, whereas 26.4% of the Hypericum group reported side-effects. Notable side-effects in both groups included dry mouth, fatigue, GI upset, dizziness and headaches.