Seventeen RCTs (n=1,503) were included: 5 studies (n=458) of red clover and 12 studies (n=1,045) of soy.
The average percentage reduction in the number of daily flushes from baseline in the placebo group was -29% (range: -1 to -59).
Isoflavone was associated with a statistically significant reduction in the number of daily flushes compared with control (effect size -0.28, 95% confidence interval, CI: -0.39, -0.18); significant heterogeneity was found (p<0.0001). The effect size remained significant when using a random-effects model (-0.49, 95% CI; -0.81, -0.17, p=0.001).
Soy isoflavone was associated with a statistically significant reduction in the number of daily flushes compared with control; the effect size was -0.54 (95% CI: -0.96, -0.13, p=0.004) when using a random-effects model. There was no significant difference between red clover isoflavone and control; the effect size (random-effects model) was -0.35 (95% CI: -0.86, 0.14, p=0.082). No significant heterogeneity was found for either of these subgroup analyses. Effect sizes for soy interventions appeared to range from favouring isoflavones (4 studies), to no statistically significant difference (7 studies), to one study favouring placebo.
Multiple regression analysis showed that a greater number of baseline flushes (p<0.0001) and increased isoflavone dose (p<0.0001) were associated with a greater response to treatment.
Twenty studies would be required to negate the statistically significant result.