11 Depression in women

Depression in women is common and has attracted increasing interest. It affects women disproportionately and besides the emotional suffering, it has a major impact on their ability to function in their various roles.

An analysis of the effect of depression on disability adjusted life years (DALYs) by Murray and Lopez (1996) showed that depression was the second leading cause of disease burden after ischaemic heart disease. However, depression proved to be the leading cause of disability for women worldwide. The most likely reasons are the disproportionate prevalence of depression and the severity of its impact on a woman’s life course.

Possible reasons for the gender difference in depression prevalence could include a greater number of first-onset episodes, longer duration of depressive episodes, a greater recurrence of depression in women than in men, or a combination of these factors.

This question was addressed by three large epidemiological studies conducted in the United States, which showed that the greater number of first-onset depressive episodes in women and not gender differences in the duration or recurrence of depression is responsible for the gender difference (Eaton et al. 1997; Keller and Shapiro 1981; Kessler et al. 1993). It can therefore be concluded that women have greater rates of first-onset depression than men, but once they are depressed, both have episodes of similar duration and are equally likely to have recurrent depressive episodes.

The subsequent question arises why do women have greater rates of first-onset depression than men? This is most likely based on the fact that there are significant biological- and socialization-related differences between women and men. Women experience certain stressors more frequently than men and women may also react differently to these stressors, either due to different biological or socialization factors.

In the rest of this chapter I will describe epidemiological evidence about the difference in prevalence of depression, examine the phases in the life cycle of a woman, and consider social, psychological, and biological aetiological factors.

The lifetime prevalence of depression in women has been found to be consistently greater than in men in a number of countries and cultures, including the United States, Puerto Rico, New Zealand, France, Iceland, Taiwan, Korea, and Germany (Weissman et al. 1996).

The Office of Population Census and Surveys (OPCS) survey of 10 000 adults living in private households in the United Kingdom found that women were more likely than men to suffer from a neurotic health problem (Meltzer et al. 1995).

These findings were confirmed by Jenkins et al. (1997), who carried out a National Psychiatric Morbidity Survey of Great Britain; 10 108 of 13 000 selected adults were interviewed. The overall one-week prevalence of neurotic disorder was 12.3% in males and 19.5% in females. Unmarried and postmarital groups had a high rate of disorder, as did single parents and people living on their own. Unemployment was strongly associated with neurotic disorder. Individual neurotic disorders were all significantly commoner in women, with the exception of panic disorder.

The National Comorbidity Survey (NCS) in the United States (Kessler et al. 1993) showed that the lifetime prevalence of a major depressive disorder was 21.3% for women and 12.7% for men. These findings were confirmed in the NCS Replication (NCS-R) study, which showed similar findings (Kessler et al. 2003).

The very large, six-nation European study DEPRES (Depressive Research in European Society) (Angst et al. 2002) establishes the sex difference at the level of both major depressive disorder (female:male = 1.7) and of the various depressive symptoms. Very few studies have shown ratios close to unity, and they have usually been in restricted populations.

The largest study, which found minimal gender differences, was the North-Trondelag Health Study (HUNT), which was carried out in Norway on 92 100 individuals. They found minimal gender differences in dimensional depression scores and in prevalence rates of depression. Both these measures were found to increase continuously with age in both genders (Stordal et al. 2001).

A number of studies have shown that at certain periods during the life cycle of a woman the prevalence of depression is increased. These periods have been identified as the postpubertal, perimenstrual, perinatal, and perimenopausal phases.

Some studies of depression in children have shown that prepubertal boys and girls have similar levels of depressive disorder, while other studies have shown that prepubertal boys had greater rates of depression than prepubertal girls (Nolen-Hoeksema and Girgus 1994; Twenge and Nolen-Hoeksema 2002). Studies of postpubertal children showed that from age 12 onwards the rate of depression in girls increased substantially whereas the rates of depression in boys increased slightly or not at all (Angold et al. 1998; Twenge and Nolen-Hoeksema 2002). However, the rates of substance abuse and criminal behaviour increased in boys from the age of 12 onwards to a much greater degree than in girls (Angold et al. 1998).

Angold et al. (1999) investigated whether the morphological changes associated with puberty or the hormonal changes were more strongly associated with increased rates of depression in adolescent girls. The results showed that the increased rate of depression in adolescent girls was not related to the morphological changes which occurred during puberty and the authors therefore concluded that the causal factors of the increase in depression in females during the postpubertal phase are most likely associated with changes in androgen and oestrogen levels.

Henshaw (2007) defined perimenstrual syndrome as any constellation of psychological and physical symptoms that recur regularly in the luteal phase of the menstrual cycle, remit for at least one week in the follicular phase, and cause distress and functional impairment. The characteristics of functional impairment have been reported in various studies (Hylan et al. 1999; Pearlstein et al. 2000). However, the precise relationship between premenstrual syndrome/premenstrual dysphoric disorder and mood disorder has remained unclear (Henshaw 2007). Although premenstrual syndrome responds well to selective serotonin reuptake inhibitors (SSRIs), the onset of depression is more rapid than that of major depression (Kim et al. 2004).

The close temporal relationship between the symptoms of the perimenstrual syndrome and the menstrual cycle suggests a strong biological aetiological factor. The hypothalamic–pituitary–adrenal (HPA) axis, the gamma-aminobutyric acid (GABA) system, the serotonergic system, and the endogenous opioids have been implicated in the aetiology of the perimenstrual syndrome (Halbreich 2003). Craig et al. (2004) have established that oestrogens can affect the cholinergic, serotonergic, dopaminergic, and noradrenergic systems (Craig et al. 2004). I will discuss biological factors of depression in women in more detail later in this chapter.

About 50–80% of woman experience ‘postpartum blues’ within the first few days of giving birth. This condition is characterized by no more than two weeks of mild depressive symptoms of mood instability, tearful anxiety, and insomnia (Ahokas et al. 2001).

Perinatal depression encompasses depressive episodes, which occur either during pregnancy or during the first 12 months after delivery. A population study (Andersson et al. 2006) found that depression was more prevalent during mid-gestation than postpartum. In a population-based sample of 1555 women in Sweden, anxiety and depression were more prevalent during mid-trimester (∼30%) than postpartum (16%). A history of psychiatric disorder, being single, and obesity were risk factors for new-onset postpartum psychiatric disorder.

A systematic review published in 2004 (Bennett et al. 2004) showed the prevalence at first, second, and third trimesters to be 7.5%, 12.8%, and 12%, respectively. A more recent systematic review and meta-analysis of perinatal depression found the prevalence of depression in the first trimester at 11.0%, which dropped to 8.5% in the second and third trimester. After delivery, the prevalence of depression rose to its highest level in the third month at 12.9%. In the fourth through seventh months postpartum the prevalence of depression declined slightly, staying in the range of 9.9–10.6%, after which it declined to 6.5% (Gavin et al. 2005).

Apart from the effects on psychological health, antenatal depressive symptoms have been associated with increased risk of obstetric interventions such as epidural analgesia, operative deliveries, and admissions to neonatal units (Chung et al. 2001).

Postnatal depression has been associated with lower quality interaction between mothers and their infant (Stein et al. 1991), with subsequent greater child insecurity in attachment relationships (Marmorstein et al. 2004) and higher levels of psychiatric disturbances amongst their children (Murray and Stein 1989).

The perimenopause has been described as a ‘tumultuous and distressing endocrine event’ (Taylor 2002) with oscillations between episodes of hypo-oestrogenism and hyperoestrogenism (Shideler et al. 1989). Pior (1998) demonstrated that oestrogen variability is significantly higher in the perimenopause than in any other natural reproductive phase in women, including puberty.

Epidemiological studies found an increase in depressive symptoms in perimenopausal women compared with premenopausal women (Bromberger et al. 2001, 2003).

The results of these studies were confirmed in findings from the following prospective studies. Harlow et al. (2003) reported an association between menopausal transition and increased risk for clinical depression, particularly in women with a history of depression. Another longitudinal study found a significantly greater risk for episodes of clinical depression around menopause than when the women were premenopausal (Schmidt et al. 2004). Freeman et al. (2004) found that women in the menopausal transition were up to three times more likely to report depressive symptoms than premenopausal women.

In a longitudinal eight-year prospective cohort study of 231 women with no history of depression at the beginning of the study, Freeman et al. (2004) found the transition to menopause and its changing hormonal milieu were strongly associated with new onset of depressed mood.

Cohen et al. (2006) examined the association between the menopausal transition and onset of first lifetime episode of depression among women with no history of previous depressive episodes. In a longitudinal, prospective study of 460 subjects, after adjusting for age at study enrolment and history of negative life events, they found that premenopausal women with no lifetime history of major depression who entered the menopause were twice as likely to develop significant depressive symptoms as women who remained premenopausal. The increased risk for depression was somewhat greater in women with self-reported vasomotor symptoms.

Bebbington et al. (1998) carried out a prospective study on 9792 subjects. Social variables, which were likely to contribute to a postmenopausal decline in depressive disorders, were controlled in logistic regression analysis. The results showed that there was a clear reversal of the sex difference in prevalence of depression in those over the age of 55. This could not be explained for in terms of differential effects of marital status, child care, or employment status. This study adds to the view that sex difference in prevalence of depression is less apparent in later middle age. This could largely be due to biological causes or due to a combination of biological factors and social factors other than those controlled for in this study.

The interpretation of the evidence for specific biological or social vulnerabilities to depression in women can be ideologically driven, as the choice between social, psychological, and biological causes can be represented as the choice between seeing women either as socially disadvantaged, or as inherently vulnerable with all the associated implications of inferiority.

However, there is likely to be interplay between these two groups. Biological factors may influence psychological and social factors and vice versa. For instance, in the perinatal phase a woman is more likely to react with depressed mood to psychological and social stressors than at other times in her life. There is considerable research evidence of biological, psychological, and social aetiological factors, and the interaction of these factors.

The strong association between mood disturbance and the transitions in women’s lives, which is characterized by hormonal shifts, suggests a causal link. It is well established that oestradiol and progesterone modulate the neurotransmitter and neuroendocrine systems, including those involving monoamines.

A number of studies have examined the relationship between changing hormone levels in women and mood disorder. In this context oestrogen has been investigated most extensively. Oestrogen is a term used to include the three major hormones oestradiol, oestriol, and oestrone. Oestradiol is the most potent oestrogen and the predominant form circulating in the body from menarche to menopause (Shervin 1997). Premenopausally, 95% of oestradiol is produced by the ovaries and 5% from fat cells.

After menopause, oestrone is partly converted to oestradiol, but this amounts to much less oestradiol than was available premenopausally (Lobo 1997; Steiner et al. 2003). Oestriol is produced by the placenta. The most significant source of oestrogen after the menopause when ovarian function fails is the weaker oestrogen oestrone, which is produced in fat cells. The menopause has been described as ‘a period of low effective oestrogen’ (Douma et al. 2005).

Oestrogen affects more than 400 bodily functions. Among other brain functions oestrogen increases the rate of degradation of monoamine oxidase and intraneural serotonin transport, both of which serve to increase serotonin availability in the synapse, and as a result enhance mood (Carretti et al. 2005).

Some studies on the relationship of gonadal hormones and depression in women have shown that some women develop depression or experience an exacerbation of an existing depression during periods when hormone levels change substantially. These periods include the premenstrual phase of the menstrual cycle, during the postpartum period, and periods of unpredictable ovulation, which include puberty and perimenopause. (Lobo 1997; Shervin 1997; McEwen 1999; Carretti et al. 2005).

Kumar and Robson (1984) and Rubinow et al. (1998) have suggested that oestrogen may provide protection against depression. They reported an increased incidence of depressive symptoms during periods associated with low oestrogen concentrations, e.g. premenstrual and postpartum. However it has also been reported that depression may be equally high at times of high oestrogen concentration, such as the antenatal period (Evans et al. 2001). Furthermore, there is some evidence that the prevalence of depression decreases postmenopausally, at a time when oestrogen levels are significantly reduced (Kessler et al. 1993; Bebbington et al. 1998).

It has been proposed that the precipitous drop in oestrogen following delivery triggers depression (Deakin 1989; Gregoire et al. 1996; Ahokas et al. 2001). It has also been suggested that women with a history of depression are at increased risk for the mood destabilizing effects of gonadal steroids and therefore postnatal depression (Bloch et al. 2000).

Studies on women with polycystic ovary syndrome have been carried out in order to investigate the role of oestrogen in depression. Polycystic ovary syndrome is a common and pervasive endocrine disorder with varied presentation and usually reduced oestrogen production. The common clinical features include hirsuitism, infertility, menstrual disturbance, and obesity. Several studies have established that women with polycystic ovary syndrome are more likely to experience symptoms of depression than comparison healthy groups. (Elsenbruch et al. 2003; Weiner et al. 2004; Hollinrake et al. 2005; Lane 2006).

Studies, which tried to identify specific features of polycystic ovary syndrome, which might be the cause of depression, have found mixed results. Neither high androgen levels (Rasgon et al. 2003; Hollinrake et al. 2005), hirsuitism (Keegan et al. 2003), or infertility (McCook 2002; Himelein and Thatcher 2006) were found to significantly correlate with depression among women with polycystic ovary syndrome. In one study, higher free testosterone levels were associated with more positive mood states (Weiner et al. 2004). Higher body mass was associated with greater depression in one study (Rasgon et al. 2003), but unrelated to depression in another (McCook 2002).

It has also been suggested that oestrogen may have a role as an antidepressant. However, methodological difficulties have affected the few studies in this area. These include small number of subjects and the lack of control groups. In postpartum depression, oestrogen therapy may be useful both as prophylaxis in vulnerable individuals (Sichel et al. 1995) and as treatment (Klaiber et al. 1979; Gregoire et al. 1996). There is also some evidence that women taking oestrogen replacement therapy may respond better to fluoxetine (Schneider et al. 1997).

In the perimenopause, oestrogen replacement therapy has been found to be effective in reducing mild depressive symptoms (Epperson et al. 1999) and it has been reported to be an effective treatment for depression (Schmidt et al. 2000; Soares et al. 2001). However, it is difficult to determine from these studies whether the oestrogen is treating menopausal symptoms, such as sleep deprivation and anergia, or the depression per se.

According to Cutter et al. (2003) there is currently little evidence to suggest that oestrogen is a useful treatment for depression during the menopause or at any other time.

During the last 30 years there have been significant changes in family structures across Western Europe with divorce and remarriage rates increasing. Findings from the 1998 General Household Survey in the United Kingdom show that 38% of births were outside marriage compared with 26% 10 years earlier. Nevertheless, 61% of the 1998 births were registered jointly by parents living in the same address.

Between 1971 and 1990, the number of divorced and separated women with children rose from 290 000 to 650 000, and the number of lone mothers from 90 000 to 390 000. In the late 1980s, European figures showed that the United Kingdom had a lone parent rate of 17%, which was amongst the highest in Europe (Millar 1997).

Targosz et al. (2003) carried out a study on 5281 women and found that lone mothers had prevalence rates of depressive episode of 7%, about three times higher than any other group. The milder condition, mixed anxiety/depression, was also increased in frequency. These increased rates of depressive conditions were no longer apparent after controlling for measures of social disadvantage, stress, and isolation.

Lone mothers tend to suffer higher rates of material disadvantage than mothers who are in a relationship (Targosz et al. 2003). Poverty has been associated with a number of chronic uncontrollable negative life conditions, including inadequate housing, dangerous neighbourhoods, and financial uncertainties. The stresses of poverty can undermine parenting skills, increase relationship conflicts, reduce self-esteem, and undermine coping strategies. There is also an increased risk of a number of acute stressors, including exposure to crime and violence and physical and sexual assault (Belle and Doucet 2003). One study of low-income women in the United States showed that 83% of these women had a history of physical or sexual abuse (Bassuk et al. 1998). Heneghan et al. (1998) confirmed these findings in a study of inner-city mothers. They reported that poor financial status, poor health status, or activity limitation due to illness were associated with higher levels of depressive symptoms.

Belle and Doucet (2003) showed that poverty is the one stressor most consistently correlated with depression in women. Depressive symptoms were more common among people with low incomes, particularly mothers with young children.

Weich et al. (2001) investigated gender difference in rates of the most common mental disorders, anxiety and depression, on 9947 subjects in a seven-year, population-based cohort study. The odds ratio for the gender difference in the future prevalence of common mental disorders was 1.92 (95% confidence interval: 1.75–210). They also found that gender differences in common mental disorders were not explained by differences in the number or type of social roles occupied by men and women, or by reverse causality. They suggested that further studies should consider characteristics of social roles, such as demand, control, and reward.

Other studies have pointed to the importance of social factors, as the sex ratio for depression is not universally maintained across all sociodemographic categories. It is, for instance, much more marked in married than in never-married women (Bebbington et al. 1981; Weissman and Klerman 1977; Lindeman et al. 2000,) and young married women looking after small children appear to be particularly at risk, at least in some societies (Brown and Harris 1978, Ensel 1982).

However, marital status has different associations with affective disorders in different cultures. Married women are at low risk of depressive disorder in Mediterranean countries (Mavreas et al. 1986; Vazquez-Barquero et al. 1987), in rural New Zealand (Romans-Clarkson et al. 1988), and among British Orthodox Jews (Lowenthal et al. 1995). These societies all accord a high value to the home-making role of women. However, irrespective of cultural background, unsupported mothers appear to be at particular risk of depression (Targosz et al. 2003).

It is therefore important to take into account that not only are social variables important in influencing the sex ratio for depression, but the association with sociodemographic factors is itself affected by more subtle sociocultural influences. The pervasiveness of the sex ratio can be seen as a reflection of the all-pervading feature of social disadvantage experienced by women worldwide in different communities.

A number of studies have shown a clear relationship between the experience of stressful life events and onset of depression (Brown and Harris 1978; Mazure 1998). Several studies have tried to establish whether women experience more frequent or more severe life events or are more sensitive to the depressive effects of life events. Some studies found that women experience more negative life events than men (Brown and Birley 1968; Bebbington et al. 1991), others found no gender difference in life events occurrence (Uhlenhuth and Paykel 1973; Perris 1984), while some found that the risk for depression following life events is greater for women than for men (Uhlenhuth and Paykell 1973; Kessler and McLeod 1984; Nazroo et al. 1997; Maciejewski et al. 2001). Other studies found that women experience certain kind of negative life events more often than men (Wagner and Compas 1990; Weiss et al. 1999).

A study by Maciejewski et al. (2001), who analysed data from a nationwide community-based sample of 1024 men and 1800 women, found no gender differences in exposure to a number of stressful life events. However, women were approximately three times more likely than men to experience major depression in the wake of a stressful life event. It has been suggested that certain cognitive characteristics in women may contribute to this increased stress reactivity in women (Abramson et al. 2002).

The difference in reactivity to life events could be based on gender-related biological predisposing factors. However, there is also evidence that women are more likely than men to have had previous adverse experiences, which can act as vulnerability factors to depression.

A Finnish population study (Veijola et al. 1998) showed that a disturbed mother–child relationship and neurotic symptoms in childhood are stronger predisposing factors to depression in women than in men.

Women are at greater risk of suffering sexual and physical abuse, which includes child sexual abuse, rape, and male partner violence (Koss et al. 2003). Sexual and physical abuse tends to occur at various points across the lifespan. Prior victimization increases the risk of repeat victimization and adolescent victimization is the strongest risk factor of continued victimization (Koss et al. 2003).

Abuse is a potent risk factor for depression both immediately after the abuse and throughout the survivors’ lifetime (Weiss et al. 1999). The National Women’s Study (Saunders et al. 1999) found that women who had been the victim of completed rape in childhood had a lifetime prevalence of depression of 52%, compared to 27% in non-victimized women. A meta-analyses of 18 studies of depression and intimate violence found that mean prevalence rates of depression among abused women was 48% (Golding 1999). It has been concluded that childhood physical abuse is the strongest predictor of adult depression in all ethnic groups after controlling for background characteristics that are risk factors for both abuse and depression (Dube et al. 2001; Oddone-Paolucci et al. 2001).

Another vulnerability factor for women is the affiliative style of relationships, as women tend to have a stronger affiliative style in their social relationships than men. This style of social relationships includes close emotional communication and intimate personal relationships. This has been attributed to biological factors as well as gender role socialization (Feingold 1994). Although gender differences in affiliative style are apparent before puberty, they tend to intensify during pubertal transition (Maccoby 1990). Adolescent girls tend to spend more time talking than their male peers and with increasing age their conversations reflect an increasing interpersonal focus (Raffaelli and Duckett 1989).

It has been suggested that this process of gender role socialization may prepare adolescent girls for caregiving roles, but at the same time the pubertal intensification in affiliative orientation may sensitize postpubertal girls to the depressogenic effect of certain negative life events, and in particular events that represent conflicts, breaches, or losses within interpersonal relationships (Cyranowski et al. 2000).

The increased prevalence of depression amongst women compared to men and the heightened risk at various phases in the life cycle of women has been well documented. A number of studies have tried to establish the responsible aetiological factors for the increased rates of depression. No definite factor has been identified and it is likely that a combination of biological, social, and psychological factors interact to contribute to the increased rates of depression. Attempts to prevent and treat depressive episodes in women need to focus on a wide range of biological, social, and psychological risk factors.