Menopause and mood

Is depression linked with hormone changes?

Diana L. Dell, MD, FACOG; Donna E. Stewart, MD, FRCPC

VOL 108 / NO 3 / SEPTEMBER 1, 2000 / POSTGRADUATE MEDICINE

CME learning objectives

• To recognize that perimenopausal and menopausal women experience biologic changes that can alter brain, behavior, mood, and cognition
• To understand some common relationships between menopause and psychiatric disorders
• To learn which hormonal and nonhormonal approaches are used for women with menopausal symptoms

Dr Dell has received speakers' honoraria from Eli Lilly and Company, Pfizer, SmithKline Beecham Pharmaceuticals, and Tap Pharmaceuticals. She serves on the lecturers' bureau for the women's health division of Eli Lilly. She has current clinical trials sponsored by SmithKline Beecham and Lilly. Dr Stewart serves on the speakers' bureau and has received speakers' honoraria from Eli Lilly, Merck-Frosst, SmithKline Beecham, and Wyeth-Ayerst Pharmaceuticals. She has conducted clinical trials or served as a consultant for Eli Lilly, Merck-Frosst, SmithKline Beecham, and Wyeth-Ayerst.

This is the first of four articles on menopause

Preview: The perimenopausal period is a biologically vulnerable time, especially for women who have a history of depression associated with childbearing or the premenstruum. What help can you offer women who are struggling with negative feelings? When does hormone therapy make a difference, and when should other options be considered? Drs Dell and Stewart review the literature and explain various plans that might benefit women during this sometimes difficult period of their lives.
Dell DL, Stewart DE. Menopause and mood: is depression linked with hormone changes? Postgrad Med 2000:108(3):34-43

Menopause is a medical term indicating the absence of menstrual bleeding for 1 year. However, since neither the onset nor the departure of the final episode of menstrual bleeding is heralded by flashing neon lights, the term menopause is often used somewhat imprecisely. In most cases, we are actually talking about the "perimenopause," which more accurately describes the transitional period between full reproductive function and menopause. In literature from Europe and the United Kingdom, the term "climacteric" is used to describe the entire process, but that term has never been popular in the United States or Canada.

The average age for menopause in North American women is 51 years. Perimenopausal symptoms generally occur between the ages of 45 and 55 years and usually last 4 to 5 years. During this time, the hypothalamic-pituitary-gonadal axis begins to malfunction. The aging ovaries become progressively less sensitive to stimulation by follicle-stimulating hormone (FSH) and luteinizing hormone (LH). Ovulation occurs intermittently, if at all, and levels of gonadal hormones begin to change. The timing and character of menstrual bleeding also change, and fertility is markedly decreased.

As hormonal levels shift, 85% to 90% of women begin to experience symptoms of vasomotor instability described as "hot flashes" or "night sweats," which are attributable to the effect of decreasing estrogen levels on brain activity. Although the exact mechanism responsible for hot flashes remains unclear, numerous neuroregulators are involved: LH levels surge before the flash; hypothalamic norepinephrine activity increases; adrenocorticotrophic hormone and growth hormone levels rise after the flash; and beta endorphin levels decrease at the onset of a flash and rise significantly after it subsides (1). As ovarian function diminishes, hot flashes become more frequent, and some women experience progressively more severe disruption to nighttime sleep and daytime functioning.

It is during this time that susceptible women begin to have problems related to insomnia, irritability, and mood disturbances. Vasomotor symptoms tend to occur with the onset of rapid-eye-movement sleep, thus disrupting normal sleep architecture. The resulting daytime fatigue, poor concentration, and dysphoria may be difficult to differentiate from the somatic symptoms associated with depression.

Perimenopausal women may report difficulty falling asleep and maintaining sleep, as well as early morning awakening. Although multiple studies have shown insomnia to be estrogen dependent, the brain mechanism remains unclear, and circulating estrogen levels do not predict the intensity of subjective symptoms. Estrogen may have direct effects on sleep regulatory areas of the hypothalamus, preoptical area, and hippocampus.

Improved sleep is often the first change a woman experiences after starting estrogen therapy. At a neuroregulatory level in the brain, estrogen functions as both a serotonin agonist and a cholinergic agonist. It has mixed effects on norepinephrine and endorphins but decreases dopamine receptors and increases gamma-aminobutyric acid (GABA) activity. All of these suggest that estrogen might have a role in mediating mood in women (2).

Reports in the literature are inconsistent about whether there is an association between menopause and depressed mood. Studies from both gynecology and general practice have shown that while "clinic" populations endorse depression as part of the menopause, community-based populations do not. It appears that women are more tolerant of the physical changes associated with menopause but seek care for mood symptoms; mood issues are the most common concerns among women attending menopause clinics. Some gynecologic textbooks still list depressed mood, irritability, lack of confidence, and poor concentration as specific symptoms of the perimenopause.

The Massachusetts Women's Health Study (3) was a 5-year, community-based, longitudinal study of the association between menopause and depression. A cohort of 2,565 women, 45 to 55 years of age at baseline, were randomly selected from census data and surveyed by phone interviews every 9 months from 1982 to 1987. At the time of the last interview, 91.5% of the cohort was still available for follow-up. Like other community-based studies, this study did not show an increase in depression associated with menopause (3).

A history of depression was clearly the variable that was most predictive of subsequent depression (P<0.001). The onset of natural menopause was not associated with an increased risk for depression, except among women who experienced a long perimenopause (>27 months). In these women, transient depression was noted, and the onset of depressive symptoms was more strongly correlated with menstrual problems than with vasomotor symptoms. As reviewed over time, all the "action" was in the perimenopause; as women became postmenopausal their rates of depression declined, independent of any previous history of depression (3).

Only 5% of women in the Massachusetts Women's Health Study were using hormone medications. Neither the main effect of hormone therapy nor its interaction with previous depression was significant (3). Nonetheless, the interest in gonadal hormones as neuroregulators and the use of hormonal therapies for perimenopausal women with recurrent mood symptoms has been the subject of increasing interest over the past decade.

Demonstration of gender dimorphism in human brain development has helped emphasize the extent to which gonadal steroids affect the central nervous system. They profoundly influence brain development during prenatal life and brain function throughout life. As noted earlier, estrogen mediates the effect of multiple neurotransmitters in the human brain. It also has direct intracellular effects and may function as its own transcription factor at a molecular level (4).

As we learn more about the molecular biology of the brain, the role for gonadal steroid hormones in mediating mood and enhancing treatment strategies for mood disorders has become the object of intense scrutiny. We have documented, for example, that women with a history of premenstrual or postpartum mood disorders may be at higher risk for experiencing recurrent mood shifts when estrogen levels change premenopausally (5). Although data are inconsistent, there is evidence that estrogen improves mood in some postmenopausal women (6). In susceptible women, estrogen may decrease vulnerability to depression and may be an effective adjunctive therapy in treatment-resistant depression (7).

In a large meta-analysis, Zweifel and O'Brien (8) reviewed 38 studies of hormone replacement therapy (HRT) in menopausal women. The results suggest that HRT is effective in reducing menopausal depressed mood. On average, women using hormone replacement therapy had fewer symptoms of depression than 76% of control patients. Estrogen alone had a moderate to major effect on improving depressed mood. However, when progesterone was combined with estrogen, the positive effects of estrogen diminished. Androgen treatment, used alone or in combination with estrogen, appeared to be most effective for reducing symptoms of depression. The analysis was somewhat limited by variations in the methodologic approaches used in the various studies (8). (While the authors of the meta-analysis used the term "hormone replacement therapy," many current authors prefer to use the term "hormone therapy" or to refer to treatment with a specific hormone by name (9).)

Although investigations are ongoing, previous studies of progesterone and testosterone have been limited. In preclinical work, estrogen has been shown to induce changes in brain progesterone receptors; in humans, progesterone appears to act predominantly as an anxiolytic, with limited antidepressant activity. The central effect of testosterone, even in men, is probably secondary to estrogen, since it is dependent on the presence of aromatase, which converts the testosterone molecule into estrogen (10).

Mood and hormone therapy

In the United States, the Food and Drug Administration (FDA) has approved use of estrogen therapy for relief of vasomotor symptoms and prevention of osteoporosis, but not for use in psychiatric disorders. Nonetheless, estrogens, progesterones, androgens, and other compounds are commonly used "off-label" for perimenopausal mood disorders, to improve libido, and as adjunctive therapy in psychiatric illness, including depression, anxiety, and psychosis.

Estrogens
Estrogen therapy to improve mood is usually prescribed at the same doses used to control vasomotor symptoms and conserve bone density in postmenopausal women. Women who have had surgically-induced menopause may require higher estrogen doses in the initial phases of treatment. These women also appear to have greater improvement in mood with estrogen therapy (11).

Oral estrogens are used most often, since the oral route maximizes beneficial effects on cholesterol metabolism. Conjugated equine estrogen is the most commonly prescribed oral estrogen. Clinical experience suggests that women who develop depression, irritability, anxiety, or headaches while taking conjugated equine estrogen may not have these effects with other forms of estrogen therapy. Perimenopausal women who still need contraception and who tolerate combination oral contraceptive drugs are good candidates for using birth control pills that contain 20 micrograms of ethinyl estradiol, if there are no other contraindications.

Transdermal estrogen preparations are preferred for perimenopausal women who have migraine headaches or liver disease. Vaginal preparations are often recommended when a more intense local effect is desired; about half of the vaginal dose is reflected in subsequent systemic estrogen levels. We are not aware of any data that show a relationship between routes of dose delivery and mood symptoms.

Progestins
Since use of estrogen alone has been associated with unacceptably high rates of endometrial hyperplasia and endometrial cancer, estrogen plus progesterone is recommended for perimenopausal women who still have their uterus. Adding a progestin appears to depress mood in a dose-dependent manner (11). In spite of ongoing debate about whether micronized progesterone causes fewer mood symptoms than medroxyprogesterone acetate, both can cause similar symptoms in equivalent doses.

Estrogen and progesterone can be used continuously or in a cyclic fashion. Cyclic regimens cause scheduled bleeding and may precipitate symptoms associated with premenstrual syndrome. Continuous regimens can cause irregular spotting, which generally resolves within one year, but patient dissatisfaction with unscheduled bleeding may lead to discontinuation of therapy.

Combination therapies
The combination of oral conjugated estrogens plus medroxyprogesterone has been in common usage since 1994. Combination packaging allows women greater convenience with either continuous (Prempro) or cyclic (Premphase) estrogen-progesterone regimens. Some new oral hormone combinations use the synthetic estrogens and progestins developed for birth control (ethinyl estradiol plus norethindrone acetate [Activella, femhrt 1/5, Ortho-Prefest]). A transdermal patch containing these drugs (CombiPatch) is also available.

Androgens
Androgens clearly have mood-enhancing properties and suppress hot flashes in men who have been treated with gonadal suppression. Data for women are not as clear. Estrogen plus androgen therapy has been reported to increase libido, decrease breast tenderness, and decrease the frequency of migraine headaches. However, some women cannot tolerate androgen therapy and become irritable and depressed when they take it.

Adverse symptoms associated with androgen therapy are probably dose related, and the incidence of side effects is greatest at high doses. Side effects associated with androgen use in women include acne, hirsutism, voice changes, and weight gain. Androgen may also produce an undesirable change in lipid profiles, theoretically enhancing the risk for coronary artery disease (12).

Combination estrogen plus methyltestosterone is available in full-strength (Estratest) or half-strength (Estratest HS) doses. In clinical practice, the half-strength dose appears to produce fewer side effects that are likely to result in discontinuation of treatment.

Selective estrogen receptor modulators
The selective estrogen receptor modulators (SERMs) are a class of synthetic estrogen look-alikes that can act as agonists or antagonists in different tissues. The ideal SERM would protect against osteoporosis and cardiovascular disease and improve mood, memory, and cognitive ability without risk of endometrial or breast cancer. Unfortunately, drugs currently available in this class do not improve sleep or relieve hot flashes (and may make them worse). Currently available SERMs are probably mood neutral, and their effect on cognition has not been established.

Tamoxifen citrate (Nolvadex) is the grandmother of the SERM drugs. Initially used as adjuvant therapy for node-negative breast cancer and therapy for metastatic breast cancer, it is now FDA-approved for reduction in breast cancer incidence in high-risk women. Anecdotal reports of mood deterioration after starting tamoxifen have been reported, but randomized, controlled trials are lacking.

Raloxifene (Evista) was introduced in the United States and Canada for prevention of osteoporosis in 1998. It appears to be mood neutral, with the first published study noting "no significant differences between raloxifene groups and the placebo group at any assessment point" (13). Several other SERMs are currently under development or in clinical trials.

Complementary and alternative therapies
Phytoestrogens, which are weak estrogens of plant origin that are widely used by perimenopausal women, are presumed to competitively bind estrogen receptors, making them unavailable to more potent estrogens. Phytoestrogens have been isolated in more than 300 plants, including apples, carrots, coffee, potatoes, yams, soy products, bean sprouts, red clover sprouts, sunflower seeds, rye, wheat, sesame seeds, and linseed. They are also found in bourbon.

Epidemiologic data indicate that women in cultures with high dietary phytoestrogen consumption have fewer menopausal complaints. For example, anecdotal reports suggest that vasomotor symptoms are reduced in women using soy products, including dietary supplements. Beneficial effects of phytoestrogens on bone mass and cardiovascular disease have been demonstrated in small trials, but their value in endometrial cancer is unknown (14). We are unaware of good data indicating mood changes attributable to ingestion of phytoestrogens in perimenopausal women.

Panax ginseng (ginseng) is widely used by perimenopausal women for improving energy and relieving hot flashes. It contains an active ingredient that is chemically similar to estrogen. Side effects, including nervousness and insomnia, have been reported, along with hypertension, diarrhea, and vaginal bleeding (15).

Angelica sinensis root (dong quai) is a common Chinese herbal medicine used for "female complaints," but clinical testing has not shown any estrogenic changes in the endometrium or vagina. In double-blind trials, dong quai was no better than placebo for alleviating perimenopausal symptoms (14). It contains numerous coumadin derivatives and is contraindicated in women with increased menstrual bleeding and women using aspirin or anticoagulants (15). Its use has been discouraged in the United States and Canada.

Cimicifuga racemosa (black cohosh, squaw root) once was widely used by Native American women and was a primary constituent of Lydia Pinkham's Compound. Black cohosh binds estrogen receptors and suppresses LH but not FSH. Clinically, it appears to have some estrogen-like activity, and several clinical trials are in progress to determine its usefulness (14). We are unaware of any data indicating mood changes attributable to ingestion of black cohosh in perimenopausal women.

Vitex agnus-castus (chaste tree berry, monk's pepper), so named because it was once used by monasteries to maintain chastity among the monks, stimulates two of the dopamine receptors, has antiandrogenic effects, and inhibits prolactin secretion in vitro. One study has shown effectiveness for treatment of premenstrual symptoms, but we have seen no data about use in menopause. Side effects include itching, gastrointestinal symptoms, and headaches (15,16).

Valeriana officinalis (valerian root) does not have estrogenic properties but has been used for hundreds of years as a sedative and sleep-promoting aid. The active ingredient is probably a GABA-like compound that has limited toxicity and degrades rapidly (16). In Germany, valerian root is often used in combination with black cohosh by perimenopausal women.

Vitamin E has shown limited ability to reduce hot flashes, but its documented role as an antioxidant makes it a common choice for perimenopausal women. It is contraindicated in women using digitalis or anticoagulants.

Vitamin B₆ has been recommended to improve mood in women with premenstrual dysphoria, but data are mixed, with double-blind trials showing performance equal to placebo. A potential role for improvement of mood in perimenopausal women has been suggested but remains unverified.

Exercise improves both mood and bone mass in perimenopausal women. Relaxation therapy may promote a sense of well-being but has no consistently verifiable effects on hot flashes.

Perimenopausal treatment of mood symptoms

Use of drug therapy for perimenopausal depression is based on a number of clinical factors. For women with mild depression and prominent vasomotor symptoms at menopause, a trial of estrogen therapy is an appropriate first step. If alleviation of menopausal symptoms does not lead to a decrease in depression, psychotherapy or use of an antidepressant drug should be considered.

Antidepressant drug therapy should be prescribed for women with moderate or severe symptoms of depression. Concurrent therapy with estrogen may be indicated, but difficulties associated with starting multiple medications at one time can be an issue. One open-label trial indicated that low-dose venlafaxine (Effexor, Effexox XR) relieved hot flashes better than placebo in a few women with breast cancer (17). The mechanism of this effect has not been elucidated.

In women with treatment-resistant depression who are perimenopausal or postmenopausal, augmentation therapy with estrogen has been used with mixed results. The role of estrogen augmentation in psychotic disorders is also being investigated.

Summary

For most women, the perimenopause is not associated with depression. In those who do become depressed, mood symptoms tend to decrease at the end of menopause, regardless of previous history of depression. Many of the changes in brain function that affect mood are secondary to changing estrogen levels in the brain. Treatment with estrogen alone may improve mood in women with mild depressive symptoms, but those unresponsive to estrogen and women who have moderate to severe depression need antidepressant therapy.

References

1. Ginsburg ES. Hot flashes--physiology, hormonal therapy, and alternative therapies. Obstet Gynecol Clin North Am 1994;21(2):381-90
2. Halbreich U. Role of estrogen in postmenopausal depression. Neurology 1997;48(Suppl 7):S16-S20
3. Avis NE, Brambilla D, McKinlay SM, et al. A longitudinal analysis of the association between menopause and depression. Ann Epidemiol 1994;4(3):214-20
4. Stahl SM. Estrogen makes the brain a sex organ. J Clin Psychiatry 1997;58(10):421-2
5. Stewart DE, Boydell KM. Psychologic distress during menopause: associations across the reproductive life cycle. Int J Psychiatry Med 1993;23(2):157-62
6. Epperson CN, Wisner KL, Yamamoto B. Gonadal steroids in the treatment of mood disorders. Psychosom Med 1999;61(5):676-97
7. Rubinow DR, Schmidt PJ, Roca CA. Estrogen-serotonin interactions: implications for affective regulation. Biol Psychiatry 1998;44(9):839-50
8. Zweifel JE, O'Brien WH. A meta-analysis of the effect of hormone replacement therapy upon depressed mood. Psychoneuroendocrinol 1997;22(3):189-212
9. Speroff L. It's time to stop using the word 'replacement.' Maturitas 2000;34(1):1-3
10. Fink G, Sumner BEH, McQueen JK, et al. Sex steroid control of mood, mental state, and memory. Clin Exper Pharmacol Physiol 1998;25:764-75
11. Sherwin B. The impact of different doses of estrogen and progestin on mood and sexual behavior in postmenopausal women. J Clin Endocrinol Metab 1991;72(2):336-43
12. Castelo-Branco C, Vicente JJ, Figueras F, et al. Comparative effects of estrogens plus androgens and tibolone on bone, lipid pattern and sexuality in postmenopausal women. Maturitas 2000;34(2):161-8
13. Nickelsen T, Lufkin EG, Riggs BL, et al. Raloxifene hydrochloride, a selective estrogen receptor modulator: safety assessment of effects on cognitive function and mood in postmenopausal women. Psychoneuroendocrinol 1999;24(1):115-28
14. Lichtman R. Perimenopausal and postmenopausal hormone replacement therapy: part 2. Hormonal regimens and complementary and alternative therapies. J Nurse-Midwifery 1996;41(3):195-210
15. Seidl MM, Stewart DE. Alternative treatments for menopausal symptoms: systematic review of scientific and lay literature. Can Fam Physician 1998;44:1299-308
16. Blumenthal M, ed. The complete German commission E monographs: therapeutic guide to herbal medicines. Austin, TX: American Botanical Council, 1998
17. Loprinzi CL, Pisansky TM, Fonseca R, et al. Pilot evaluation of venlafaxine hydrochloride for the therapy of hot flashes in cancer survivors. J Clin Oncol 1998;16(7):2377-81

For a helpful guide to electronic and print resources on menopause for physicians and patients, see the Resource Guide in this issue.

Dr Dell is assistant professor, departments of psychiatry and obstetrics-gynecology, Duke University Medical Center, Durham, North Carolina. Dr Stewart is professor and chair, women's health, University Health Network Women's Health Program, Toronto, Ontario, Canada. Correspondence: Diana L. Dell, MD, FACOG, Assistant Professor, Departments of Psychiatry and Obstetrics-Gynecology, Duke University Medical Center, DUMC Box 3263, Durham, NC 27710. E-mail: dell0001@mc.duke.edu.

Symposium Index

• MENOPAUSE: Introduction to a four-article symposium by Michele G. Cyr, MD
• MENOPAUSE AND MOOD: Is depression linked with hormone changes? by Diana L. Dell, MD, FACOG, Donna E. Stewart, MD, FRCPC
• WEIGHT GAIN DURING MENOPAUSE: Is it inevitable or can it be prevented? by Laurey R. Simkin-Silverman, PhD, Rena R. Wing, PhD
• HRT TO PREVENT CARDIOVASCULAR DISEASE: What studies show, how to advise patients by Malissa J. Wood, MD, Jafna L. Cox, MD, FRCP
• POSTMENOPAUSAL OSTEOPOROSIS: Strategies for preventing bone loss, avoiding fracture by Kelly A. McGarry, MD, Douglas P. Kiel, MD, MPH

Please send technical questions related to the Web site to Ann Harste