Menopausal Status and Depression
Much of the research that gives rise to the perception that depression is an inevitable concomitant of the menopausal transition derives from clinic or patient samples of women who have sought treatment for menopause-related problems. These patient samples present a biased view of the natural menopause transition, as studies have shown that women who seek treatment tend to report more life stress and suffer more from clinical depression, anxiety, and psychological symptoms than nonpatient samples (Ballinger, 1985; Morse et al., 1994). To determine whether the menopausal transition is associated with depression in the general population, it is necessary to examine research from nonpatient samples.
Epidemiologic studies of menopausal status and depression do not provide consistent evidence of an association between the menopausal transition and depression among the general population of women (Avis, 2000; Nicol-Smith, 1996; Pearce, Hawton, & Blake, 1995). Using a variety of measures, most cross-sectional studies do not find an association between menopausal status and depression (e.g., Cawood & Bancroft, 1996; Dennerstein, Smith, & Morse, 1994; Dennerstein et al., 1993; Hällstrom & Samuelsson, 1985; Kaufert, Gilbert, & Tate, 1992; Kuh, Wadsworth, & Hardy, 1997; McKinlay, McKinlay, & Brambilla, 1987; Porter, Penney, Russell, Russell, & Templeton, 1996; Woods & Mitchell, 1997). On the other hand, several studies have found an association between menopausal status and depression or dysphoria mostly in the perimenopause (Avis, Kaufert, Lock, McKinlay, & Vass, 1993; Ballinger, 1975; Bromberger et al., 2001; Collins & Landgren, 1995; Hunter, Battersby, & Whitehead, 1986). Except for Avis et al. (1993) these studies used a symptom score of general negative mood or general psychiatric morbidity and did not specifically study depression.
Longitudinal studies find no evidence that onset of the transition leads to increased depression or that postmenopausal women are more depressed than pre-menopausal women (Avis, Brambilla, McKinlay, & Vass, 1994; Dennerstein, Lehart, Burger, & Dudley, 1999; Hunter, 1990; Kaufert et al., 1992; Kuh et al., 1997; Hällstrom & Samuelsson, 1985; Holte, 1992; Matthews et al., 1990; Woods & Mitchell, 1996). One longitudinal study that examined length of the perimenopause found that women who remained perimenopausal for at least 27 months showed a significantly higher rate of depression (Avis et al., 1994). Subsequent analyses revealed that this was largely explained by symptoms associated with menopause (i.e., hot flashes, night sweats, and menstrual problems).
Other Factors Related to Depression During the Perimenopause
Numerous studies have shown that psychosocial factors account for more of the variation in depressed mood among women at the time of menopause than does menopause itself. Studies have shown that stress and stressful life events (Ballinger, 1975; Dennerstein, Dudley, & Burger, 1997; Greene & Cooke, 1980; McKinlay et al., 1987), socioeconomic status (Dennerstein et al., 1997; Holte & Mikkelsen, 1991), negative attitudes toward menopause and aging (Dennerstein et al., 1993), and negative expectations of menopause (Holte & Mikkelsen, 1991) are related to more negative mood during the menopausal transition. Other studies have found that health is significantly related to depression (Dennerstein et al., 1994; Kaufert et al., 1992; McKinlay et al., 1987; Woods & Mitchell, 1996). Longitudinal data suggest that prior depression is the primary factor related to depression during the menopausal transition (Avis et al., 1994; Hunter, 1990; Kuh et al., 1997; Porter et al., 1996).
Studies have also found that women who report greater mood disturbances at the time of menopause also report a history of menstrual cycle or reproductive related problems (Bancroft & Backstrom, 1985; Pearlstein, Rosen, & Stone, 1997). These include previous or current premenstrual symptoms or complaints (Collins & Landgren, 1995; Dennerstein et al., 1993; Stewart & Boydell, 1993; Woods & Mitchell, 1996), dysmenorrhea (Collins & Landgren, 1995) and postpartum depression (Stewart & Boydell, 1993). It is not clear from these findings, however, whether this pattern of symptom reporting related to reproductive events is hormonally based or a result of greater symptom sensitivity or greater symptom reporting in general.
Research has also shown cultural differences in the association between menopausal status and depression with Western women having higher rates of depression in general and more likely to show an association with menopausal status than non-Western women (Avis et al., 1993; Boulet, Oddens, Lehert, Vemer, & Visser, 1994). Attempts to associate menopause and depression through cross cultural comparisons, however, are complicated by variations in the meaning of menopause as well as the presentation of psychological symptoms.
Endogenous Hormones and Depression
The majority of epidemiologic studies of menopausal status and depression use menstrual cycle changes to measure status. Menstrual cycle characteristics, however, are an indirect measure of hormonal change. The biochemical theories of menopause and depression would suggest that hormonal levels provide a more direct measure. Several studies have directly examined endogenous hormones (specifically estradiol) and depression and have not found a significant association between estradiol (E[sub2]) and depression (Ballinger, Browning, & Smith, 1987; Cawood & Bancroft, 1996; Chakravarti, Collins, Thom, & Studd, 1979; Dennerstein et al., 1999). These studies, however, had small samples and may not have had sufficient power to detect associations. Avis, Crawford, Stellato, and Longcope (2001) examined the relation between estradiol and menopausal symptoms and depression among 309 women during the menopausal transition and found an unadjusted association between E[sub2] and depression. However, when statistical models adjusted for hot flashes/night sweats and trouble sleeping, the association between E[sub2] and depression was no longer significant. The absence of support for an association between estradiol and depression may in part be a function of methodological limitations. Frequent repeated measurement of gonadal hormones are necessary to obtain reliable measures, as well as hormonal fluctuations, which may be more important than absolute levels.
Although not specifically studying depression, researchers have shown that estrogen can impact brain neurotransmitter activity (Halbreich & Kahn, 2001; Schmidt & Rubinow, 1991). A few small studies provide some support for the effects of estrogen on aspects of the serotonin system in menopausal women. Klaiber, Kobayashi, Broverman, and Hall (1971) report that postmenopausal women are more likely to have increased monoamine oxidase activity (which leads to a decrease in serotonin) in comparison with premenopausal women. Gonzales and Carrillo (1993) found that post-menopausal women have lower blood serotonin levels than premenopausal women and that blood serotonin levels increase following estrogen replacement therapy in post-menopausal women. Zarate, Fonseca, Ochoa, Basurto, and Hernandez (2002) showed a significant increase in serotonin and β-endorphin, and dopamine among symptomatic postmenopausal women who received low-dose conjugated equine estrogen. Decreases in gonadal hormones, however, are not related in a simple or direct way to psychological distress during the menopausal transition. Sex steroid levels per se have not distinguished symptomatic from asymptomatic women (Charney, 1996; Schmidt & Rubinow, 1991).
Exogenous Estrogen and Depression
Estrogen replacement therapy has long been touted as a mental tonic that improves well-being (Halbreich, 1997; Kopera & Utian, 1990; Utian, 1972) A number of observational studies have compared mood or depression scores of women taking estrogen therapy with those of nonusers. These observational studies show inconsistent results in which some studies report less depression or dysphoric mood among hormone users and others report greater depression among hormone users. These studies, however, are subject to numerous methodological problems such as selection bias (hormone users tend to be healthier to begin with), not considering reasons for initiating hormone use (women who are more depressed may be more likely to initiate use) and inability to examine change in depression in relation to starting hormone therapy. It is thus not possible to draw any conclusions on the role of estrogen from these studies. We therefore need to look at randomized clinical trials.
In randomized, placebo-controlled clinical trials, the most consistent findings with respect to the positive effects of estrogen on mood have been found in studies conducted among women undergoing a surgical menopause (when a surgical procedure stops menstruation, either through a bilateral oophorectomy or hysterectomy; Ditkoff, Crary, Cristo, & Lobo, 1991; Sherwin & Gelfand, 1989; Sherwin & Suranyi-Cadotte, 1990). However, given the complex hormonal changes associated with surgical menopause, the experience of surgery itself, and medical factors leading up to the surgery, one cannot generalize these results to the general population (Palinkas & Barrett-Connor, 1992). Further, women who have a surgical menopause often experience more distress than women who experience a natural menopause (Kaufert et al., 1992; Kuh et al., 1997; McKinlay et al., 1987).
Research conducted among non-hysterectomized women show less consistent results. Studies conducted among women reporting vasomotor or other symptoms are more likely to show a positive effect of estrogen therapy on psychological well-being (Hlatky, Boothroyd, Vittinghoff, Sharp, & Whooley, 2002; Limouzin-Lamouth, Marion, Joyce, & LeGal, 1994; Wiklund, Karlberg, & Mattsson, 1993) than studies conducted among asymptomatic women (Greendale et al., 1998; Strickler et al., 2000). These studies thus suggest that the benefits of estrogen therapy on reducing depressive symptoms may be due to the reduction of other symptoms such as vasomotor symptoms and sleep problems.
Research suggests that on a population level, the menopausal transition is not necessarily associated with increased depression. Clearly, many women transition through menopause without signs of depressive symptoms. However, some studies find that women who experience severe vasomotor symptoms may be at risk for depressive symptoms and that alleviation of these symptoms can improve depressive mood.
Women with other characteristics such as a history of PMS or previous affective disorders that are cyclic or associated with reproductive events may also be at risk for depressive symptoms or depression during the perimenopause. It is premature, however, to conclude that this pattern is related to a hormonal imbalance. Other pre-existing factors such as coping style or greater sensitivity to symptoms may be related to increased vulnerability. Unfortunately, these vulnerability factors have not been prospectively studied. Further, because most studies involve retrospective reporting of reproductive and psychiatric history, there is the inherent problem of selective recall among women experiencing problems at the time of the study.
What does this research suggest in terms of the previously described theories or hypotheses? The domino hypothesis that vasomotor symptoms can lead to depressed mood has considerable support. Although the biochemical hypothesis remains a leading hypothesis regarding mood disorders during the menopausal transition, research suggests that this hypothesis is very complex and involves more than ovarian estrogen decline. Not all women suffer from these biochemical changes and if estrogen decline was the critical factor, postmenopausal women would show higher rates of depression. Furthermore, the precise relation between estrogen and serotonergic function has yet to be fully delineated (Joffe & Cohen, 1998), and there is no clear understanding of the mechanisms whereby hormonal changes mediate change in mood (Brace & McCauley, 1997). As for the psychoanalytic view, research has not shown that onset of the menopausal transition leads to depression or that perimenopausal women suffer from a sense of loss. Further, there is no evidence of higher rates of depression among postmenopausal women. The social circumstances hypothesis has received considerable support in numerous studies. However it is important to point out that the social circumstances associated with depression do not differ from those related to depression generally. Thus, there does not seem to be a specific event concurrent with the menopausal transition that leads to depression. The social circumstances associated with depression during the menopausal transition are the same ones generally associated with depression.
The integrative model of depression presented by Nolen-Hoeksema (2001) provides a useful way to view depression during the menopausal transition and integrate previous theories. Nolen-Hoeksema proposes that stress, stress reactivity, and biological factors (e.g., hormones, HPA axis dysregulation, genetics) can all lead to depression. Stress can have a direct effect on depression or operate through biological factors or increasing stress reactivity. This model provides a useful way to view depression during the transition in that multiple factors can increase women's vulnerability to depression. Women may be depressed because of stress or other factors unrelated to menopause, because of heightened reactivity to stress, and/or because of biologic factors. These factors may also interact with each other to cause depression. Thus, depression can be viewed as multifaceted and not necessarily caused by only one factor. One addition to Nolen-Hoeksema's model should include cultural factors, which may influence how women respond to events and/or how depression is manifested.
- Avis, Nancy; Depression during the menopausal transition; Psychology of Women Quarterly; Jun 2003; Vol. 27; Issue 2.
Reflection Exercise #6
The preceding section contained information
about research on the association between menopause and depression. Write three case study examples
regarding how you might use the content of this section in your practice.
Online Continuing Education QUESTION 13
Numerous studies have shown that psychosocial factors account for more of the variation in depressed mood among women at the time of menopause than does menopause itself. What are these psychosocial factors?
Record the letter of the correct answer the