Alternatives to Estrogen for Treatment of Hot Flashes: Title and subTitle BreakAre They Effective and Safe?

Hot flashes (or flushes) are the major symptom of menopause. Hot flashes are common, occurring in approximately 50% of women during the menopause transition,¹ and approximately 20% of women who are affected request treatment.² Women describe a hot flash as the sudden onset of a sensation of heat spreading over the face and chest and lasting a few minutes.² The frequency and intensity of symptoms varies widely among women, with some experiencing an occasional sensation of warmth while others experience frequent episodes of intense heat and drenching sweat that disrupt daily activities and sleep. Hot flashes are more prevalent in black and Latin American women than in white women, but are less common in Chinese and Japanese women.³ Cigarette smoking seems to increase the likelihood of experiencing hot flashes,¹ but most other potential risk factors such as obesity, exercise, and alcohol intake have not consistently been associated with hot flashes.⁴ Hot flashes resolve in most women after a few years, but in 10% to 15% of women, the hot flashes persist for decades or are even lifelong.⁵

In humans, body temperature is regulated by inducing vasodilation and sweating to release heat, and vasoconstriction and shivering to conserve heat. A hot flash is very similar to a heat dissipation response, as both result in vasodilation, sweating, and reduction in core body temperature. The core body temperature at which postmenopausal women with hot flashes vasodilate and sweat is lower than in women without hot flashes.⁶ Thus, normal fluctuations in core body temperature may trigger hot flashes.

The cause of altered thermoregulation in postmenopausal women with hot flashes is not known. One theory suggests that changes in estrogen levels associated with menopause alter central nervous system adrenergic neurotransmission and cause abnormal thermoregulation. This theory is supported by studies that show that systemic administration of yohimbe, an α-adrenergic antagonist that increases norepinephrine release, provokes hot flashes.⁷ In the systematic review of clinical trials of nonhormonal treatments for hot flashes by Nelson et al⁸ published in this issue of JAMA, treatment with clonidine, an α-adrenergic agonist that decreases central norepinephrine release, reduces the frequency of hot flashes.

Alternatively, there is some evidence that decreasing estrogen levels during the menopause transition result in changes in serotonergic neurotransmission that might cause hot flashes. Lower estrogen levels are associated with lower levels of serotonin in blood, resulting in increased sensitivity of hypothalamic serotonin receptors. Stimulation of these receptors can alter the thermoregulatory set point in animals. Mild stressors, such as heat or anxiety, cause a brief release of serotonin that may stimulate central receptors, lower the thermoregulatory set point, and cause hot flashes.⁹ The systematic review by Nelson et al⁸ also supports this hypothesis. It demonstrates that selective serotonin reuptake inhibitors (SSRIs), which increase central serotonin levels, are effective in the treatment of hot flashes.

Hormone therapy is the most effective treatment currently available for hot flashes. Overall, women randomized to hormone therapy have 60% to 85% average reductions in hot flash frequency.^{10 - 11} There is a dose-response effect, but low doses are often effective.¹² Given the marked efficacy of estrogen for treatment of hot flashes, are other therapies needed? Postmenopausal estrogen therapy reduces osteoporotic fractures and until recently was thought to reduce coronary events.¹³ Hormone therapy is associated with uterine bleeding, breast tenderness, and headache but these adverse effects were thought to be well worth the dual benefits of symptom relief and disease prevention. This reasoning has been largely responsible for a lack of research on the etiology of hot flashes and alternatives to estrogen for treatment.

Recently, large randomized trials have demonstrated that postmenopausal hormone therapy does not decrease the risk of coronary events.^{14 - 16} Among generally healthy women in the Women's Health Initiative trials, estrogen treatment increased risk for venous thromboembolic events and stroke, and progestin added to estrogen increased risk for venous thromboembolic events, stroke, coronary events, and breast cancer.^{14 - 15} The absolute increase in risk is small (<1 excess serious adverse outcome per 1000 50- to 60-year-old women who are treated with hormone therapy for a year).⁵ Nevertheless, many women would prefer safe alternative treatments for hot flashes, even if the therapy is not quite as effective as estrogen.

The review by Nelson et al⁸ is a technically rigorous systematic review of placebo-controlled randomized trials of nonhormonal therapies. All elements of the QUOROM statement recommendations for reporting systematic reviews have been addressed.¹⁷ The literature review was comprehensive and the inclusion and exclusion criteria were clearly defined. The authors excluded non-English publications, but this is unlikely to significantly impact the conclusions and there was no evidence of publication bias. However, the authors reported significant heterogeneity in the study results and the use of a random-effects model to combine study results may not adequately address the problem of heterogeneity.

Heterogeneity is particularly evident in the trials of antidepressants. One trial¹⁸ of venlafaxine reported a significant reduction in hot flashes and another trial¹⁹ reported a trend toward more hot flashes in the venlafaxine group. Despite this mixed evidence and the fact that venlafaxine is a selective norepinephrine reuptake inhibitor (SNRI), the authors combine these results with those of trials of the SSRIs paroxetine and fluoxetine. Stratified analyses suggest that the mixed results of these trials might be due to differences in the study populations. The antidepressants appeared effective primarily in trials that included women with a history of breast cancer using tamoxifen. Interestingly, the same difference was apparent in the trials of clonidine, which appeared effective only among women with a history of breast cancer and taking tamoxifen. The reason for this apparent difference is not clear, but it would be prudent to conduct additional trials in women without a history of breast cancer to document efficacy for these medications. Based on the systematic review,⁸ paroxetine is the only antidepressant shown to statistically significantly relieve hot flashes overall (a reduction of 1.7 hot flashes per day more than placebo) in women with a history of breast cancer and those without.

The review also included studies of soy and red clover extracts but excluded other dietary soy interventions, although the majority of the placebo-controlled soy extracts studies carefully documented and monitored total isoflavone content. Isoflavones are chemicals found in certain plants that are hypothesized to act like weak estrogens. Prior systematic reviews of soy²⁰ and the better quality randomized trials of dietary soy^{21 - 23} found little evidence of efficacy for treating hot flashes. Given the fact that the proposed mechanism of action of red clover extracts and soy extracts are similar (more similar than the SSRIs and SNRIs), the lack of efficacy for red clover extracts also decreases the likelihood that the effects of soy extracts are real. The preponderance of the evidence indicates little clinical benefit from isoflavone-based products, including soy extracts.

Additional nonhormonal therapies that have been studied in clinical trials were excluded in the review by Nelson et al.⁸ Black cohosh has been studied in many trials of varying quality and the results have been mixed. An ongoing National Institutes of Health–supported trial of black cohosh should provide better data. Small clinical trials of evening primrose oil,²⁴ dong quai,²⁵ ginseng,²⁶ and wild yam²⁷ do not support their use for relief of hot flashes.

The systematic review by Nelson et al⁸ suggests that paroxetine, gabapentin, and clonidine are modestly effective for relief of hot flashes. But are these drugs safe? Paroxetine can cause headache, nausea, insomnia, anxiety, and sexual adverse effects, and many women do not like the idea of taking an antidepressant. Clonidine appears to be about as effective as the antidepressants, but causes dry mouth and drowsiness. Despite the lack of head-to-head trials, gabapentin appears to be somewhat more efficacious than the antidepressants and clonidine but is associated with somnolence and dizziness and must be taken 3 times a day.

In addition to symptomatic adverse effects, do the nonhormonal treatments have long-term adverse effects? Treatment duration was only a few months in the trials included in the meta-analysis. Antidepressants, clonidine, and gabapentin have been marketed for a long time but large, long-term trials (similar to the Women's Health Initiative trials of hormone therapy) are lacking. Safety is a particular concern for isoflavone extracts because they contain estrogenic compounds and thus may be subject to some of the same long-term adverse effects as hormone therapy.²⁸

Women with hot flashes should understand that most symptoms resolve over several months to several years. Those women with mild symptoms may find adequate relief by wearing layered clothing and keeping the home and bedroom cool. For women with more bothersome symptoms, clinicians should understand the advantages and disadvantages of both hormone therapy and nonhormonal alternatives. Hormone therapy is more effective than nonhormonal alternatives but should probably be avoided by women at high risk for venous thromboembolic events, cardiovascular disease, and breast cancer.^{14 - 16} Nonhormonal alternatives are less effective than estrogen, generally have more symptomatic adverse effects, and long-term adverse effects are not as well documented. With all medicines or dietary supplements used for symptomatic treatment, the lowest effective dose should be used and stopped as soon as symptoms improve or resolve. A better understanding of the pathophysiology of hot flashes will likely be necessary for the development of nonhormonal therapies that equal or surpass the efficacy of hormones.