Benefits
Menopausal symptom relief
In one randomized, placebo-controlled trial in 90 early postmenopausal Chinese women aged 45 to 60, hot flash frequency fell 44.3% on 84 mg a day and 48.5% on 126 mg a day over 24 weeks, compared with 27.8% on placebo. The Kupperman menopause index fell 57.8% and 56.7% on those doses, compared with 34.6% on placebo. Both differences were statistically significant (p<0.01), but a large share of the improvement also occurred in the placebo group on its own.
Bone health: not measured here
The one trial cited on this page did not measure bone in any way, so there is no bone evidence here for this extract. Soy isoflavones are phytoestrogens and bone has been looked at in other research, but that work is not on this ingredient and is not cited on this page.
Honest evidence note
The same trial found no effect on blood lipids, and no change in estradiol, FSH or LH. That is reassuring for safety, but it also means the hormonal mechanism was not demonstrated. The higher 126 mg dose was no better than 84 mg, everything on this page rests on a single trial, and the women studied were Chinese, whose usual soy intake is typically much higher than a Western diet's.
Mechanism of action
Phytoestrogen activity
Soy isoflavones can bind estrogen receptors, which is the proposed explanation for the hot flash results. In the trial cited here, blood estradiol, FSH and LH did not change, so this mechanism is a proposal rather than something the study confirmed.
Equol conversion
Daidzein can be converted by gut bacteria into equol, a more active phytoestrogen. Not everyone's gut bacteria do this. The trial cited here does not report equol status, so its role in these results is unknown.
Clinical trials
Randomized, placebo-controlled trial in 90 early postmenopausal Chinese women aged 45 to 60, given soy germ isoflavones at 84 mg or 126 mg daily for 24 weeks. (Ye et al. 2012, Menopause)
90 early postmenopausal Chinese women aged 45 to 60. This is the only trial cited on this page.
Hot flash frequency fell 44.3% at 84 mg and 48.5% at 126 mg, compared with 27.8% on placebo. The Kupperman index fell 57.8% and 56.7%, compared with 34.6% on placebo. Both were significant against placebo (p<0.01). There was no effect on blood lipids, no change in estradiol, FSH or LH, and no extra benefit from the higher dose. Bone was not measured.