Thirteen RCTs (997 patients) were included. Eight of these were used in the meta-analyses (632 patients).
The validity scores ranged from two to five out of a possible five points.
Maximal workload (4 RCTs, 310 patients): hawthorn extract significantly increased maximal workload compared with placebo; the WMD was 7 watts (W) (95% CI: 3, 11, P<0.01). No significant heterogeneity was detected (P=0.5). The inclusion of data from the low-dose instead of the high-dose treatment arm made little difference to the results; the WMD (311 patients) was 6 W (95% CI: -1, 14). The WMD for studies using hawthorn plus other medications (212 patients) was 5 W (95% CI: 0.2, 10). The WMD for studies in which the use of other medications was unclear (98 patients) was 12 W (95% CI: 4, 21).
There were too few studies to reach any conclusions about the possibility of publication bias.
Hawthorn extract significantly reduced the pressure-heart rate product; the WMD (6 RCTs, 264 patients) was -20 (95% CI: -32, -8).
Hawthorn extract improved exercise tolerance, but the increase was not statistically significant; the WMD (98 patients) was 117 W minutes (95% CI: -1, 235).
Four RCTs showed that hawthorn extract improved symptoms (dyspnoea and fatigue). A meta-analysis of two RCTs using the Zerssen symptom score showed that hawthorn extract significantly improved the score; the WMD (169 patients) was 6 (95% CI: -9, -2).)
The most common adverse effect was dizziness or vertigo (reported in 8 patients). Five RCTs reported no adverse effects with hawthorn extract.