Study Question: Which nonhormonal therapies are effective in the management of menopausal hot flashes?

Setting: Various (meta-analysis)

Study Design: Meta-analysis (randomized controlled trials)

Synopsis: Recent concerns about the adverse effects of hormonal therapy have increased interest in alternative treatments of menopausal hot flashes. The investigators searched MEDLINE, the Cochrane Registry, other large databases, and reference lists of recent systematic reviews and relevant articles, and they consulted experts. To be included, studies had to be English-language, randomized, double-blind placebo-controlled trials evaluating the treatment of menopausal hot flashes with nonhormonal interventions. Two reviewers rated the quality of trials independently; disagreements were resolved by consensus with a third reviewer. Formal assessment found no evidence of publication bias. A total of 43 trials met inclusion criteria, including 10 trials of antidepressants, 10 of clonidine (Catapres), six of other prescribed medications, and 17 of isoflavone extracts. Among various antidepressants, high-quality studies supported only the effectiveness of paroxetine (Paxil). Of the 10 trials comparing clonidine with placebo, three met criteria for fair quality; four reported reduced frequency of hot flashes with clonidine compared with placebo; and six found no difference, with the overall meta-analysis reporting a small benefit. The effectiveness of clonidine is strongest in women taking tamoxifen (Nolvadex) for breast cancer. Two fair-quality trials of gabapentin (Neurontin) reported significantly reduced hot flash frequency compared with placebo. One of six trials of red clover isoflavones reported reduced hot flash frequency compared with placebo. Soy isoflavones were compared with placebo in 11 trials, most of which were of poor quality. Although the results were heterogeneous, overall results showed that soy isoflavone extract was significantly better than placebo.

Bottom Line: Evidence supports the nonhormonal treatment of menopausal hot flashes with paroxetine, clonidine, gabapentin, and soy isoflavone extract. The overall effect size of all nonhormonal treatments is less than that of estrogen. Treatment should be individualized according to symptom severity and risk profiles. (Level of Evidence: 1a–)