Seventy RCTs were included in the review; 48 assessed biologically based therapies including phytoestrogens, nine assessed mind-body therapies, one assessed a manipulative or body-based therapy, two assessed energy therapies and ten assessed whole medical systems.
Biologically based therapies (48 studies).
Phytoestrogens, including soy isoflavones extract (8 studies) and red clover (6 studies).
There were mixed results from trials examining biologically based therapies including phytoestrogens. One good-quality study in women with breast cancer (n=175) found no difference between treatment and placebo groups, but improvements in both groups. The largest reasonable quality trial of phytoestrogens in women without breast cancer (n=241) found no differences between groups given two doses of isoflavones or placebo on any outcome. Nine other trials found no differences between the groups, mixed results, or were considered to be of a poor quality.
Two fair-quality trials of soy isoflavone supplements in women with breast cancer (n=182 and n=177) showed evidence of improvement in either hot flash severity or frequency in the treatment groups. One fair-quality trial in women without breast cancer (n=75) showed evidence of improvement only in hot flash frequency. The other studies were of a poor quality and showed mixed results.
None of the good- or fair-quality trials of red clover reported any between-group differences in hot flashes.
Other phytoestrogen preparations.
Two studies assessed women with breast cancer (one fair-quality trial, n=12; one poor-quality trial, n=52) and neither found between-group differences. One poor-quality study (n=70) compared genistein with placebo or estrogen over one year and found a significant and persistent reduction in hot flash incidence in both treatment groups; the effect was significantly greater in the estrogen group than in the genistein group. Two poor-quality studies (n=30 and n=50) found no treatment effects with phytoestrogen creams.
Black cohosh (4 studies).
One large (n=304) fair-quality study showed improvements in menopausal symptoms in the treatment group. However, no differences were observed in a smaller (n=62) fair-quality trial. One poor-quality study (n=136) and one fair-quality study (n=85) of breast cancer survivors showed no improvement in hot flashes with treatment.
Dehydroepiandrosterone (2 studies).
One fair-quality study (n=60) found no differences between the treatment and placebo groups. One small poor-quality study (n=22) did not report between-group comparisons.
Other therapies (11 studies).
One fair-quality study (n=125) assessed vitamin E in women with breast cancer and found no between-group differences. One fair-quality study (n=80) assessed kava and found greater improvement in anxiety compared with placebo, while another fair-quality study n=64) assessed phospholipids liposome injections and found greater improvement in anxiety and menopausal symptoms compared with placebo. The remaining studies were considered to be of a poor quality and were not discussed in detail.
Mind-body and behavioural therapies (9 studies).
Two fair-quality (n=173 and n=81) and one poor-quality (n=30) trial examined exercise. Both of the fair-quality trials reported improvements (one in quality of life and one in menopausal symptoms) with treatment. Two poor-quality trials examined relaxation breathing, but neither reported between-group differences. One small poor-quality study (n=14) assessed progressive muscle relaxation and found an improved time for the onset of hot flushes for the intervention group (p<0.1). No between-group differences were found for audiotape relaxation versus usual care in one poor-quality trial (n=40). One fair-quality trial of stress management versus usual care (n=86) and one fair-quality trial of counselling support for women with breast cancer versus usual care (n=76) found no between-group differences in menopausal symptoms. However, women in the usual care group were more likely to experience aches and pains than those in the stress management group (p<0.1).
Manipulative or body-based therapies (1 study).
One small fair-quality study (n=30) evaluated low-force osteopathic manipulation of the pelvis, spine and cranium compared with sham treatment and reported improvements in hot flashes, night sweats, urinary frequency, depression and insomnia in the treatment group.
Energy therapies (2 studies).
One poor-quality study (n=80) found no difference between reflexology and routine foot massage on any outcome. Another poor-quality study (n=15) found a greater improvement in hot flashes in a placebo group than in a group treated with magnets at acupressure points (p=0.02).
Whole medical systems (10 studies).
Four studies assessed acupuncture, three comparing it with sham acupuncture and one with acupuncture intended for general well-being. No trial found any difference between these groups in hot flashes, although one poor-quality trial (n=30) reported improved mood in the treatment group. One of the fairer quality trials (n=45) also contained a group treated with conjugated oestrogen; this group showed significant improvement in self-reported symptoms compared with acupuncture and sham acupuncture (p<0.001).
Six studies assessed traditional Chinese medicinal herbs; three of these used combination therapies. Only one fair-quality trial (n=384), which used a standardised dose of ginseng compared with placebo, showed greater improvement in the treatment group: there were between-group differences in depression, well-being and health scores, but not hot flashes.