Most studies have found clear gender differences in the prevalence of depressive disorders. Typically, studies report that women have a prevalence rate for depression up to twice that of men (Bebbington, 1996; Nolen-Hoeksema, 1987).
Most studies have found clear gender differences in the prevalence of depressive disorders. Typically, studies report that women have a prevalence rate for depression up to twice that of men (Bebbington, 1996; Nolen-Hoeksema, 1987). For example, Kessler et al. (1994) reported that women in the United States are about two-thirds more likely than men to be depressed, and a national psychiatric morbidity survey in Britain showed a similar greater risk of depression for women (Meltzer et al., 1995). Gender differences in depression appear to be at their greatest during reproductive years (Bebbington, 1996).
Recently, my colleagues and I set out to explore some of the possible explanations for gender differences in depression (Nazroo et al., 1997, 1998). A starting point to our research was Brown and Harris (1978), who showed that episodes of depression were almost always preceded by a major life event, a finding that has been confirmed in a number of other studies.
Our sample consisted of 97 couples recruited from an inner city area in Britain who had experienced a total of 115 shared crises, made up of one or more major life events. The members of each couple were interviewed separately by different interviewers. The interviews covered the nature of the crisis, responses to the crisis, role behaviors, responsibilities and role commitment, quality of marriage, childhood experiences, and psychiatric disorders. The latter were assessed using a shortened version of the Present State Examination, which covers a wide range of psychiatric symptoms and allowed the use of clinically validated diagnostic thresholds.
It has been suggested that apparent gender differences in the rate of depression are the result of one or more possible artifacts. In particular, it has been suggested that perceived differences in rate are the result of the use of assessments of depression that do not draw a distinction between clinical depression and subclinical symptoms. Typically, such approaches count the number of symptoms that each respondent reports and then average the number of symptoms across a population. This average score might be higher among women because they are more likely to report subclinical symptoms (Newmann, 1984, as cited in Nazroo et al., 1998).
We examined this theory by exploring whether any detected gender difference in depression was sensitive to a raising of the threshold for diagnosing depression. Contrary to the hypothesis, we found that, if anything, the size of the gender difference in depression increased, rather than decreased, as the threshold was increased (Nazroo et al., 1997, 1998). This suggests that the gender difference was neither a consequence of using too low a threshold nor of using an average score that emphasized women's hypothesized greater reporting of milder symptom states.
Another possibility is that gender differences in depression rates may be the result of men developing alternative disorders in response to stress, such as antisocial behavior and alcohol abuse. In particular, women may be more likely to have been socialized to express dysphoria in response to stress and men may be more likely to have been socialized to express anger or other forms of acting out. In support of this, studies have shown that expected gender differences in depressive disorders were balanced out by higher male rates of alcohol abuse and drug dependency (Kessler et al., 1994; Metzler et al., 1995).
Very few respondents, however, met the criteria for alcohol or drug abuse (two of the 97 men reached diagnostic criteria, and two women were just below this threshold) (Nazroo et al., 1998). Furthermore, it did not appear that the gender differences in depression were the result of men being more likely to externalize their anger. If anything, women reported both feeling and expressing more anger in response to the crises theyexperienced.
Our research suggested that these are not alternative gender-typed disorders developed in reaction to the same situation. Rather, such a pattern is likely to reflect responses to different life situations. Indeed, in the data presented by Meltzer et al. (1995), there is a strong suggestion that substance abuse is more frequent among young men and those who are single or separated, while gender differences in depression are at their greatest among those who are married (Bebbington, 1996).
It has been argued that gender differences in social roles and in the experiences, stresses and expectations that surround them are heavily involved in women's greater risk for depression. To explore this possibility, we classified the crises experienced by the couples in this study according to the role domain in which they occurred (Nazroo et al., 1997, 1998).
Overall, women had an 80% greater risk of an episode of depression following a crisis (Nazroo et al., 1998). This risk was more than five times greater for women following crises involving children, housing and reproduction. There was no gender difference in risk for crises involving finances, work and the marital relationship (Figure 1).
The possibility that this difference was a consequence of gender differences in roles was tested further by directly assessing role differences. Measures of role involvement, responsibility and commitment were used to identify differences in both role behavior and aspiration (Nazroo et al., 1998).
As expected, women were much more likely to be involved in child care, housework and financial management, while men were more likely to be responsible for financial provision. In contrast, men appeared to be just as likely to be committed to the idea of parenthood and having a good home environment as women were.
We then hypothesized that the impact of an event would be dependent on its salience to role identities (Nazroo et al., 1997, 1998). So, using the assessments of role involvement, responsibility and commitment, the relative salience of crises involving children, housing or reproduction (i.e., those crises that produced a gender difference in outcome) to the role identities of the couple was directly estimated. We found that women did not have a greater risk of depression onset following crises that did not have a greater role salience for them, but had a 10-fold and significantly greater risk following those that did have a greater role salience for them (Nazroo, 1998) (Figure 2).
Our study suggested that even though both women and men experienced similar levels of stress, gender differences were present. The greater effect of particular crises on women was a result of their greater salience to women's role identities. This suggests that the role strain effect is a consequence of differential sensitivity to events, as a result of role differences, rather than women experiencing more events.
Some researchers have argued that the greater sensitivity of women to particular types of events is a consequence of socialization (Kessler and McLeod, 1984). We suggested that such a structural interpretation of a socialization effect is inadequate because it leaves little room for individual difference (Nazroo et al., 1998). Although most couples interviewed followed a stereotyped gender-based division of roles, a significant minority did not follow this pattern. This implies a need to include an understanding of context and individual agency when considering role identity. Importantly, it was current role identity that predicted relative risk of depression, rather than the global greater sensitivity of women to particular types of events, which would be expected to result from a direct socialization effect.
It has been suggested that biological factors are responsible for the gender differences found in depression. First, there is the possibility that women may be at a greater risk because of the biological consequences of pregnancy andchildbirth. There has been a failure to specify what these biological consequences might be, however, and evidence to support biological explanations has been inconsistent.
In some contexts, having had children appears to be a risk factor for depression (Brown and Harris, 1978), while it appears to have no effect in others (Gaminde et al., 1993). It is possible that parity may operate through culturally influenced psychosocial pathways, rather than biological ones.
We found that parity among women was unrelated to risk (Nazroo et al., 1997, 1998). Thirty-eight percent of women with children had an episode of depression, compared to 33% of women without children.
While this suggests that biological changes related to parity are not important, the role-determined nature of the difference in risk between men and women (described earlier) suggests that the social changes that follow childbirth are of some importance. The implication is that any difference related to having had children is a consequence of increased exposure to life events rather than biological changes.
Another possibility is that sex hormones are responsible for gender differences in depression. Despite the fact that gender differences in the experience of depression appear to emerge around puberty (Bebbington, 1996), there is no clear evidence that hormonal changes associated with the menstrual cycle, resulting in perimenstrual mood changes, are involved (Harris et al., 1991; Popay et al., 1993).
We found that only certain crises were associated with a greater risk of depression for women but only when there were clear role differences between the men and women in associated domains (Nazroo et al., 1997, 1998). The fact that the gender difference turned out to have such a specific relationship with the nature of the stress suggests that such a general biological difference was not involved.
It thus seems that the gender difference in depression reported here was unlikely to be the result of biological differences related either to sex hormones or to parity.
In summary, there was no evidence to support the possibility that the higher risk was the consequence of a measurement artifact. The higher rate of depression among women was not a consequence of gender-typed responses to crises. The very specific context in which gender differences for risk of depression occurred suggested that they were not the consequence of a general difference in biological risk either.
Rather, the higher risk of depression among women was the consequence of gender differences in roles, with women only having a higher risk following crises involving children, housing and reproduction, rather than those involving finances, work and the marital relationship. Following such crises, there was only a gender difference in risk if there was a clear difference in the role salience of the crisis to the man and woman. Where the crisis had a similar role salience for both, the man had a similar risk of depression to the woman.
Bebbington P (1996), The origins of sex differences in depressive disorder: bridging the gap. International Review of Psychiatry 8(4):295-332.
Brown GW, Harris T (1978), Social Origins of Depression: A Study of Psychiatric Disorder in Women. London: Tavistock Publications.
Gaminde I, Uria M, Padro D et al. (1993), Depression in three populations in the Basque country-a comparison with Britain. Soc Psychiatry Psychiatr Epidemiol 28(5):243-251.
Harris T, Surtees P, Bancroft J (1991), Is sex necessarily a risk factor to depression? Br J Psychiatry 158:708-712.
Kessler RC, McGonagle KA, Zhao S et al. (1994), Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry 51(1):8-19.
Kessler RC, McLeod JD (1984), Sex differences in vulnerability to undesirable life events. American Sociological Review 49:620-631.
Meltzer H, Gill B, Petticrew M, Hinds K (1995), OCPS Surveys of Psychiatric Morbidity in Great Britain, Report 1: The prevalence of psychiatric morbidity among adults living in private households. London: Her Majesty's Stationery Office.
Nazroo JY, Edwards AC, Brown GW (1998) Gender differences in the prevalence of depression: artefact, alternative disorders, biology or roles? Sociology of Health & Illness 20(3):312-330.
Nazroo JY, Edwards AC, Brown GW (1997), Gender differences in the onset of depression following a shared life event: a study of couples. Psychol Med 27(1):9-19.
Nolen-Hoeksema S (1987), Sex differences in unipolar depression: evidence and theory. Psychol Bull 101(2):259-282.
Popay J, Bartley M, Owen C (1993), Gender inequalities in health: social position, affective disorders and minor physical morbidity. Soc Sci Med 36(1):21-32.